UH Uralica 14 Helsingiensia14 Multilingual Finnic Multilingual This volume gathers together articles dealing with Finnic mi- nority languages and language contacts. The first part presents topics focusing on phonology, morphology, morphosyntax, Multilingual Finnic syntax as well as lexical relations. The second part of the book Language contact and change consists of non-peer-reviewed reports on archived and digital Finnic minority language materials as well as two field trip re- ports to Finnic-speaking or formerly Finnic-speaking areas. This collection is based on the work carried out during the project “Language change in multilingual Finnic”, funded by the Kone Foundation. Most articles are based on presentations given at a symposium organized at the XII International Congress for Finno-Ugric Studies in Oulu in August 2015. Language contact and change and contact Language

ISBN 978-952-7262-16-0 (print) EDITED BY ISBN 978-952-7262-17-7 (online) SOFIA BJÖRKLÖF & ISSN 1797-3945 SANTRA JANTUNEN UH Uralica 14 Helsingiensia14 Multilingual Finnic Multilingual This volume gathers together articles dealing with Finnic mi- nority languages and language contacts. The first part presents topics focusing on phonology, morphology, morphosyntax, Multilingual Finnic syntax as well as lexical relations. The second part of the book Language contact and change consists of non-peer-reviewed reports on archived and digital MULTI­LINGUAL FINNIC Finnic minority language materials as well as two field trip re- ports to Finnic-speaking or formerly Finnic-speaking areas. This collection is based on the work carried out during the project “Language change in multilingual Finnic”, funded by the Kone Foundation. Most articles are based on presentations given at a symposium organized at the XII International Congress for Finno-Ugric Studies in Oulu in August 2015. Language contact and change and contact Language

ISBN 978-952-7262-16-0 (print) EDITED BY ISBN 978-952-7262-17-7 (online) SOFIA BJÖRKLÖF & ISSN 1797-3945 SANTRA JANTUNEN Uralica Helsingiensia14

Multi­lingual Finnic

Language contact and change

EDITED BY SOFIA BJÖRKLÖF & SANTRA JANTUNEN

HELSINKI 2019 Sofia Björklöf & Santra Jantunen (eds): Multi­lingual Finnic. Language contact and change. Uralica Helsingiensia 14. Contents Most of the peer-reviewd articles in this publication are based on presentations given at the symposium “Change of in a multilinguistic environ- ment”, held at Congressus Duodecimus Inter­nationalis Fenno-Ugristarum at the University of Oulu in August, Sofia Björklöf & Santra Jantunen 2015. Orders • Tilaukset Introduction 7 Cover picture Sofia Björklöf The tadpoles on the cover were photographed in Tiedekirja Mäggär' village, Lake Jürgjär'v in 2014 when visiting Snellmaninkatu 13 I the Veps. They symbolize the development of the Finnic FI-00170 Peer-reviewed scientific articles languages – the subject matter of this article collection. ‹www.tiedekirja.fi› Language editor Uldis Balodis Santra Jantunen ISBN 978-952-7262-16-0 (print) Syntactic and aspectual functions of ISBN 978-952-7262-17-7 (online) Latvian verbal prefixes in Livonian 15 ISSN 1797-3945 Heini Karjalainen Printon AS Borrowing morphology: The influence of Russian 2019 on the Veps system of indefinite pronouns 55 Sofia Björklöf Uralica Helsingiensia Mutual contacts and lexical relations among the Finnic varieties of western and northeastern 89 The series Uralica Helsingiensia is founded in 2008. It features thematic volumes and mono- graphs on the history, typology and sociology of . Studies focusing on - Liina Lindström, Maarja-Liisa Pilvik, Mirjam Ruutma & Kristel Uiboaed tonian language and culture, and culture, as well as Saamic studies are equally welcome in the series. The series also seeks to support university studies on its focus On the use of and pluperfect in Estonian : areas. Uralica Helsingiensia is rated at level 1 by the Publication Forum of the Federation of Frequency and language contacts 155 Finnish Learned Societies. All submitted manuscripts are peer-reviewed. Torbjörn Söder Uralica Helsingiensia on perustettu 2008. Sarjassa julkaistaan temaattisia artikkelikokoelmia­ tai monografioita erityisesti uralilaisten kielten historian, typologian ja kielisosiologian The Finnish of Rautalampi and Värmland – A 195 alalta. Aihepiiriin kuuluvat myös Unkarin kielen ja kulttuurin, Viron kielen ja kulttuurin ja saamentutkimuksen yliopisto-opintoja tukevat julkaisut. Sarja kuuluu TSV: julkaisufoorumin Irina Novak luokituksessa kategoriaan 1, ja siinä noudatetaan vertaisarviointikäytäntöä. Venäjän kielen vaikutus tverinkarjalan murteiden Publisher • Julkaisija äännejärjestelmään 227 Finno-Ugrian Society • Suomalais-Ugrilainen Seura Vesa Koivisto Editors • Päätoimittajat Ulla-Maija Forsberg, Riho Grünthal Rajakarjalaismurteiden refleksiiviverbeistä 249 Editorial board • Toimitusneuvosto Márta Csepregi, Cornelius Hasselblatt, Magdolna Kovács, Johanna Laakso, Helle Mets- lang, Matti Miestamo, Irma Mullonen, Karl Paju­salu, Janne Saari­kivi, Anneli Sarhimaa, Elena Skribnik https://journal.fi/uralicahelsingiensia, https://www.sgr.fi/uh The publications are indexed in ARTO data base with the index Urbis. Julkaisut luetteloidaan ja indeksoidaan ARTO-tietokantaan tunnuksella Urbis. II SOFIA BJÖRKLÖF & SANTRA JANTUNEN Reports University of Helsinki

Fedor Rozhanskiy & Elena Markus A new resource for Finnic languages: The outcomes of the Ingrian documentation project 303 Introduction Liina Lindström, Pärtel Lippus & Tuuli Tuisk The online database of the University of Tartu Archives of Estonian Dialects and Kindred Languages and the Corpus of Estonian Dialects 327 Background Liis Ermus, Mari-Liis Kalvik & Tiina Laansalu The Archive of Estonian Dialects and Finno- at the Institute of the 351 This publication is an outcome of the project “Language change in multilingual Finnic”, or Itämerensuomalaisten kielten muutos moni- Anneli Sarhimaa kielisessä ympäristössä, funded by The Language Programme of the Finnic data sets in the ELDIAdata databank 367 Kone Foundation. Nina Zaiceva The aim of our project was to investigate the effects of multi­ heritage in 379 lingualism and language contact on language change. The focus Toni Suutari & Ulriikka Puura was on minority languages, such as Livonian and Veps, as well as Kotimaisten kielten keskuksen itämerensuomalaiset aineistot 401 on the historical language contact situation in northeastern Estonia and western Ingria. Additionally, the aim of the project was to pro­ Marjatta Palander, Helka Riionheimo, mote research data management and the availability of digitized ma­ Hannu Kemppanen & Jukka Mäkisalo Kielikorpuksia Suomen itärajalta 425 terials in en­dangered Finnic languages. The project was led by Riho Grün­thal, Professor of Finnic languages at the University of Helsinki. Uldis Balodis Santra Jantunen was involved throughout the project concentrating Expeditions among the Lutsi and on verbal derivation in Livonian. Ulriikka Puura contributed eight the design of Language Learning Materials 439 months by carrying out research on contemporary Veps communi­ Sofia Björklöf ties. After her departure, Sofia Björklöf joined the project focusing Vepsän mailla Herran vuonna 2014 479 on the lexical traces of mutual contacts between the Finnic languages in Ingria and northeastern Estonia. We all have been working on our doctoral theses during the course of this project. In addition, Santra List of contributors 494 Jantunen transcribed old unpublished recordings in Livonian and used them in her research. In 2014, a field trip was organized to Veps vil­ lages located in Vologda and Leningrad oblasts.

Multilingual­ Finnic. Language 6 contact and change. 7–12. Uralica Helsingiensia 14. Helsinki 2019. ‹https://doi.org/10.33341/uh.85048› SOFIA BJÖRKLÖF & SANTRA JANTUNEN INTRODUCTION

Along with investigating the impact of language contacts on lan­ Lindström et al., Söder, Koivisto) with some combining both types guage change, the project paid special attention to sociolinguistic fea­ of data (Karjalainen). In this volume, the strong Russian influence on tures as well as examined different manifestations of multilingualism. individual Finnic languages is covered in several articles (Karjalainen, One of the milestones of the project, the Twelfth International Koivisto, Novak). Parallel with it, Latvian has influenced Livonian Congress for Fenno-Ugric Studies, CIFU XII, was held in Oulu, Fin­ (Jantunen), whereas Swedish has influenced the Finnish in land, on August 17–21 in 2015. The project organized a symposium Värmland (Söder). Furthermore, the historical influence of German on dealing with Finnic minority languages. Most of the articles in this Estonian dialects illustrates a different type of language contact situ­ volume are based on the presentations held at the symposium. ation. The influence of Swedish, Latvian, Russian, and other Finnic languages can also be detected on a local or regional level. (Lind­ ström et al.) Given the diversity of language contact situations, mutual This publication contacts among various Finnic varieties spoken in the same area are discussed as well, namely the languages of Ingria and adjacent areas This volume includes articles on almost every Finnic minority lan­ including Vote, Ingrian, Estonian, and Ingrian Finnish (Björklöf), as guage: Livonian, already extinct as a first language in the traditional well as so-called Border Karelian and Savo dialects and Southeastern speech area; Vote or Votic, nearly extinct; Ingrian, soon facing the dialects of Finnish (Koivisto). same fate; Veps, with a couple thousand mainly elderly speakers; and Santra Jantunen (University of Helsinki) analyses the func­ Karelian, fragmented across Tver oblast and the Republic of Kare­ tions of Latvian-origin verbal prefixes in Livonian in her paper titled lia in . Dialectal varieties of Estonian and Finnish, otherwise Syntactic and aspectual functions of Latvian verbal prefixes in Livo- national languages, are included as well, because dialects and dialec­ nian. She concludes that in Livonian, verbal prefixes are used as both tal features are levelling out in the modern world. Värmland Finnish, lexical and grammatical elements, since, to some extent, some verbal for instance, is already extinct; also the dialect materials used as data prefixes express perfectivity and, thus, form a secondary strategy for are old. Historically, however, dialects have been a natural part of the distinguishing between aspectual properties. Moreover, the use of ver­ Finnic linguistic continuum. bal prefixes often corresponds to their Latvian parallels. The articles are organized according to the focused language Heini Karjalainen (University of Oulu) studies Russian influ­ area and speech community. Starting with Livonian we move to Veps, ence on Veps indefinite pronouns and their restructuring in her article Ingrian, and Vote, Estonian and Finnish dialects, and finally to Kare­ Borrowing morphology: The influence of Russian on the Veps system of lian dialects. All structural levels of language are represented in this indefinite pronouns. She demonstrates that several Veps indefiniteness volume including phonology (Torbjörn Söder, Irina Novak), mor­ markers were acquired as morpheme transfer (MAT) and morphologi­ phology (Santra Jantunen, Heini Karjalainen, Söder, Vesa Koivisto), cal pattern transfer (PAT) borrowings. Sociolinguistic factors, such as morphosyntax (Jantunen), syntax (Liina Lindström – Kristel Uibo­ the minority language status of Veps and extensive bilingualism, have aed – Maarja-Liisa Pilvik – Mirjam Ruutma), and lexicology (Sofia increased the motivation for borrowing. Björklöf, Söder, Koivisto). A common topic connecting all articles is Sofia Björklöf (University of Helsinki) describes the historical the effect of language contacts on Finnic minority languages. The data sociolinguistic situation and language contacts in western Ingria in her of individual articles originates from different sources. One article is article Mutual contacts and lexical relations among the Finnic varie- based on materials collected by the author herself (Novak), while an­ ties of western Ingria and northeastern Estonia. This article pays spe­ other part of the studies is based on earlier published materials as well cial attention to lexical borrowing and the identification of loanwords as unpublished materials preserved in archives (Björklöf, Jantunen, originating from closely related Finnic varieties. It includes extensive

8 9 SOFIA BJÖRKLÖF & SANTRA JANTUNEN INTRODUCTION lists of known borrowings from neighboring varieties and examines reflexive derivatives originally represent the primary type of reflex­ eight methodologically illustrative etymologies in detail; some of the ives in Border Karelian and they exhibit secondary Finnish influence etymologies are totally new. since some morphological derivative types are borrowed from Finn­ Liina Lindström, Kristel Uiboaed, Maarja-Liisa Pilvik, ish, while on the other hand, the reflexive conjugation resembles the and Mirjam Ruutma (University of Tartu) analyse the use and fre­ reflexive conjugation in Karelian dialects in Russia. However, due to quency of perfect and pluperfect, the two compound past tenses in Finnish infuence, it is less widely applied. Finnish influence can also Estonian dialects on the basis of language contacts in their article be seen in numerous lexical borrowings, which are used as reflexive On the use of perfect and pluperfect in Estonian dialects: Frequency verbs in Border Karelian. In comparison with this, Russian influence and language contacts. The authors claim that the frequency of per­ is considerably weaker than in other Karelian dialects. fect and pluperfect in Estonian dialects varies considerably between The second part of the book consists of non-peer-reviewed re­ dialect areas due to local language contacts and functional differ­ ports which, however, are invaluable for spreading information on ences between compound tenses. The two main regions where com­ available data and collections. Papers in this section compile informa­ pound tenses are used the most often are the Insular and Mulgi dialect tion on the archive materials on Finnic minority languages. We hope regions. that they will function as a resource for the future study of Finnic and Torbjörn Söder (Uppsala University) compares the Finnish that these reports will guide scholars to already existing materials. Fe- dialects of the old great-parish of Rautalampi in Savo, with dor Rozhanskiy (University of Tartu & Institute for Linguistic Stud­ those in Värmland in in his article The Finnish of Rautalampi ies RAS, St. Petersburg) and Elena Markus (University of Tartu & and Värmland – A comparison. Söder notes that these two varieties of Institute of Linguistics RAS, ) report on the outcomes of their the Savo dialect of Finnish share a common historical background, but Ingrian documentation project. In the general section of the congress different language contact situations have resulted in phonological, CIFU XII, Pärtel Lippus, Liina Lindström, and Tuuli Tuisk presented morphological, lexical, and syntactic differences. the materials archived at the University of Tartu. This volume includes Irina Novak (Karelian Research Centre of the Russian a report on the online database of the University of Tartu Archives of Academy of Sciences), a native speaker of Tver Karelian discusses Estonian Dialects and Kindred Languages and the Corpus of Estonian Russian influence in the phonology of Tver Karelian in her articleVe - Dialects by Liina Lindström, Pärtel Lippus, and Tuuli Tuisk näjän kielen vaikutus tverinkarjalan murteiden äännejärjestelmään (University of Tartu). (‘Russian influence on the phonetics of the Tver dialects of Karelian’). The rest of the reports have been written exclusively for this pub­ The influence on both and consonant systems is evident. She lication and, originally, were not presented at the symposium. Liis points out that the age of borrowed features varies because some of Ermus, Mari-Liis Kalvik, and Tiina Laansalu (Institute of them likely originate from Late Old Karelian, while others are more the Estonian Language) give an overview on the Archive of Estonian recent and have arisen in Tver dialects. The phonology of Dyorzha Dialects and Finno-Ugric Languages at the Institute of the Estonian was influenced by Russian more strongly in comparison with other Language in Tallinn. Nina Zaiceva (Karelian Research Centre of Tver dialects. the Russian Academy of Sciences) writes about the Veps materials at Vesa Koivisto (University of Eastern Finland) contributes a the Karelian Research Centre of the Russian Academy of Sciences. detailed study on reflexive verbs in Karelian dialects in the vicinity of Anneli Sarhimaa (University of Mainz) introduces the relatively the Finnish border – the so-called Border Karelian dialects – in his ar­ recent language sociological interview materials on minority Finnic ticle Rajakarjalaismurteiden refleksiiviverbeistä (‘On reflexive verbs languages collected during the ELDIA project, European Language in Border Karelian dialects’). He concludes that, on the one hand, Diversity for All (2010–2013). Toni Suutari and Ulriikka Puura

10 11 SOFIA BJÖRKLÖF & SANTRA JANTUNEN

(Institute for the ) present the Finnic materials archived at the Institute for the Languages of Finland in Helsinki. Marjatta Palander, Helka Riionheimo, Hannu Kemppa- nen, and Jukka Mäkisalo (University of Eastern Finland) present the corpora of Border Karelia, Ingrian Finnish, and Karelian Finnish newspapers at the University of Eastern Finland in Joensuu. The last two reports are accounts of field trips carried out recent­ ly in Finnic-speaking or formerly Finnic-speaking areas. Uldis Ba- lodis (University of Livonian Institute) writes about his Kone- funded field trips to the Lutsi Estonians in Latvia as well as the design of language learning materials for Lutsi descendants. Sofia Björklöf I (University of Helsinki) looks back on a field trip to the Veps villages Peer-reviewed scientific articles located in Vologda and Leningrad oblasts as a part of the project “Lan­ guage change in multilingual Finnic” in the summer of 2014.

Acknowledgements

We are very grateful to all of our contributors for their participation in creating this volume and also for their patience as well as to all the wonderful people who agreed to review individual articles for us. We thank the Kone Foundation for making it possible to organize the sym­ posium and, most of all, for funding this project and our research for four years. As editors we are grateful for the opportunity to edit this collection. We owe a heartfelt thanks to Professor Riho Grünthal, the leader of our project, for all his kind help.

12 SANTRA JANTUNEN University of Helsinki

Syntactic and aspectual functions of Latvian verbal prefixes in Livonian1

Abstract This paper presents an analysis of an assumed contact- induced change in the Livonian modes of expressing : the adoption of Latvian-origin verbal prefixes expressing perfective aspect. The main objective of this article is to determine whether long-standing contact between Livonian and Latvian has led to the introduction of verbal prefixes as both pure lexical ele­ ments and, in parallel, as markers of grammatical functions that distinguish Livonian from its closest cognate languages. The cur­ rent study is based on the data derived from unpublished record­ ings and published written material representing spoken Livonian, already extinct as a first language in the traditional speech area. There are a total of eleven Latvian-origin verbal prefixes in Livonian, a language which usually does not display this category. The prefixes are as follows: aiz-, ap-, at-, ie-, iz-, nuo-, pa-, pie-, pōr-, sa-, and uz-. In Latvian, most of these items can be used as bound verbal prefixes and also prepositions marking adverbial functions. In Livonian, these prefixes can be combined with both Livonian and Latvian verbs but, as a rule – except for pa- – they do not occur as prepositions. The frequency of their occurrence in the data varies considerably and, presumably, corresponds to the degree that a given prefix may derive perfective verbs. In fact, verbal prefixation can be considered, to some extent, a means for expressing perfective aspect in Livonian, thereby adding a second­ ary strategy to the inherent Finnic way of expressing aspectual op­ positions, namely the object case alternation and verbal particles. Keywords language contact, aspect, aspectual opposition, perfec­ tivity, verbal prefixes, Livonian, Latvian

1. This paper is written as a part of the research project “Language change in multilingual Finnic”, funded by the Kone Foundation.

Multilingual­ Finnic. Language contact and change. 15–53. Uralica Helsingiensia 14. Helsinki 2019. ‹https://doi.org/10.33341/uh.85032› SANTRA JANTUNEN SYNTACTIC AND ASPECTUAL FUNCTIONS OF LATVIAN VERBAL PREFIXES IN LIVONIAN 1. Introduction

In this article I will examine how intensive language contact between two genetically unrelated languages has influenced the grammatical system of one of these languages – Livonian – in its role as the re­ cipient language in this language contact situation. This article focus­ es on the expression of aspect, more precisely perfectivity, in spoken Livonian by means of Latvian verbal prefixes. Historically, the influence of language contact and the extensive Livonian-Latvian bilingualism characteristic of Courland Livonian speakers was very strong already during the time it was first documented. Therefore, the Latvian influence on Livonian has been considerable not only lexical­ ly but also grammatically. (See, .., Ernštreits & Kļava 2013, 2014; Grünthal 2003: 161–202, 2015; Rudzīte 1994; Suhonen 1973, 1974: 62–101; Wälchli 2000.) An example of this influence is the use of Latvian verbal prefixes in Livonian and the change they have induced in the Livonian aspect system. Courland Livonian was traditionally spoken in twelve villages located on the Livonian Coast along the northwestern coast of the Bal­ tic Sea and the Gulf of Rīga in Courland, Latvia. This area is part of the present-day municipalities of Ventspils and Dundaga. Map 1. Livonian villages on the Livonian Coast. Map from Kettunen (1938: III). 1.1. Aspect as a topic of research

In recent studies, aspect has been treated both at a more general level definition the “totality” view of perfectivity, that is, imperfective and and in language-specific studies. The large variety of terms and con­ perfective aspects are distinguished by the notion of totality. He devel­ cepts used to describe this phenomenon is striking. Aspect is described oped Comrie’ definition concentrating on defining perfective verbs as a temporal category that focuses on “the internal temporal constitu­ by adding that events denoted by the verbs have a result or end-state ency of a situation”, as stated in Comrie’s (1976: 3) classic definition, and they are located in the past. He also added that perfective catego­ while another temporal category, tense, locates a situation in time. ries have a strong tendency to appear only in past time reference. Klein Dahl (1985: 23), however, found Comrie’s definition based on seman­ (1994: 16) summarized aspect as the speaker’s viewpoint regarding tics problematic, because it makes it difficult to distinguish tense and the temporal course of an event or action. A relatively common view is aspect from each other. Therefore, temporal reference often limits, that on the lexical level, aspect is present in the Aktionsart of the verb for example, aspectual categories and this definition of aspect cannot lexemes, in other words, how the verb lexemes differ in their inherent be separated from time. Comrie (1976: 16) defined perfective aspect aspectual characteristics. Different means to express aspect include, as presenting a situation as a single whole, and not separating it into for example, verb morphology, adverbials, and specific particles. different phases that form that situation. Dahl (1985: 74) called this (Comrie 1976: 1–3, 5, 16; Dahl 1985: 24–27, 69–89; Klein 1994: 16.)

16 17 SANTRA JANTUNEN SYNTACTIC AND ASPECTUAL FUNCTIONS OF LATVIAN VERBAL PREFIXES IN LIVONIAN

In more recent studies, aspect has been treated from various 1.2. An overview of verbal prefixes in points of view. Bertinetto & Delfitto (2000) base their approach to Latvian and the Uralic languages aspect on Comrie’s (1976) view. They argue that the notions of as­ pect and Aktionsart (they use the term “actionality”) should be kept In standard Latvian, there are eleven verbal prefixes, aiz-, ap-, at-, separate, since these phenomena generally cover different domains, ie-, iz-, no-2, pa-, pār-, pie-, sa-, and uz-, which have been borrowed and because aspect proper is primarily an inflectional feature while into Livonian. In addition, two verbal prefixes appear in Latvian: a actionality is a lexical property. The opposing stance is that these are negative prefix ne-, and the prefix jā- that is used to form the debi­ both considered to belong to a single notion of aspect. Also, they em­ tive mood. The total number of prefixes in Latvian is greater than the phasize that aspectual properties are also present in tenses along with number borrowed into Livonian (see, e.g., LVG: 146–148, 212–215, temporal properties. 278–286, 486, 789). These prefixes are also used in the so-called Li­ Smith (1997) discusses aspect both from a general perspective vonian dialects of Latvian, spoken in Courland (Rudzīte 1980: 165). and in language-specific cases, including, for example, in English, These dialects as well as standard Latvian are possible donors of the Russian, and Navajo. Smith divides the aspectual meaning of a sen­ prefixes. The use of verbal prefixes is the greatest difference when tence into two parts: aspectual viewpoint (in her classification: perfec­ comparing the expression of perfectivity between Livonian and oth­ tive, imperfective, and neutral) and the situation type, or the classes of er Finnic languages. The inherent ways of expressing the opposition states or events. Thus, the viewpoint and the situation type result in the between imperfective and perfective actions primarily used in other aspectual meaning of a sentence, and form the two-component theory. Finnic languages are diverging case marking of direct objects and al­ In her treatment of individual languages, Smith analyzes the aspectual ternation between different types of adverbials. Diverging case mark­ systems of these languages concentrating on viewpoint. ing of direct objects, based on the genitive, nominative, accusative, In the treatment of aspect, several studies concentrate on individ­ or alternates between individual languages. In Finnish, some ual languages. For example, Borik (2006) seeks to define perfective personal pronouns such as minu- (I-acc), sinu-t (you.sg-acc), häne-t and in Russian, along with studying the interac­ ((s)he-acc) also have accusative forms, and accusative is the term tion between different types of aspect and developing a formal theory used for the non-partitive cases of object in Finnish (e.g., of (im)perfectivity. Borik distinguishes different levels of aspectual for Finnish, see VISK § 352, § 930–931, § 1498, § 1500; for Esto­ information in Russian: imperfective/perfective opposition, or view­ nian, see EKG II: § 487, § 506; Erelt & Metslang 2017; for Veps, see point aspect, appears on the level of outer aspect, while the dis­ Puura 2007, 2010). Estonian has also developed a secondary way of tinction exists on the level of inner aspect. Ziegeler (2006) emphasizes expressing perfectivity by using so-called bounders, such as the verbal in her versatile treatment of English aspect that verbal aspect is in particles ära ‘off, away’ and valmis ‘ready’ (Metslang 2001). Livo­ interaction with other grammatical categories and their development. nian employs this method too, and often uses the particle jarā ‘off, This viewpoint is present, for example, in her integration of studies of away’ (Tveite 2004). In addition to Estonian verbal particles, Wälchli aspect and modality. Ebert (2002) offers a description of the formerly (2001: 419) also considers Livonian and Latvian verbal particles as little studied aspectual system of Maltese. The aspectual system of well as Baltic prefixes as bounders, which have grammaticalized to Maltese is like that of the systems of other Arabic dialects. The two different degrees. Prefixes are the primary way to express perfectivity Maltese simple finite verb forms are combined tense-aspect forms. In in Latvian (see, e.g., Endzelin 1922; LVG: 533; Wälchli 2001: 433). addition, auxiliaries and preverbal particles have a crucial role in the Considering the intensive influence that Latvian has had on Livonian, Maltese tense-aspect system. 2. The Latvian letter stands for three phonemes: the /u͜ o/ and the /ɔ/, /ɔː/, which appear only in loanwords (Auziņa 2014‒2016; LVG: 23.).

18 19 SANTRA JANTUNEN SYNTACTIC AND ASPECTUAL FUNCTIONS OF LATVIAN VERBAL PREFIXES IN LIVONIAN prefixes presumably have a considerable role in marking perfectivity 2) With which verbs and verb types do these prefixes appear in in Livonian as a method parallel to the object case alternation. Livonian? Are the verbs that occur with verbal prefixes histori­ In Uralic languages, prefixation is not an inherent feature. Thus, cally Livonian words or are they borrowed from Latvian; in the in the Uralic framework, Livonian appears exceptional compared to latter case, perhaps as a combination of a prefix and verb. A fur­ most other Finnic languages as well as Uralic languages in general. ther question is, whether it is possible to separate some groups of In Finnish and Estonian, however, certain word formation strategies verbs on a semantic basis. and use of shared grammatical elements in head-initial marking can 3) How does expressing perfective aspect in Livonian with verbal be considered as the emergence of negative prefixes: in Finnish ei-, prefixes correspond to the same in Latvian. for example ei-toivottu ‘unwanted’ and epä-, for example epä|onni ‘misfortune’, in Estonian eba-, for example eba|õnn ‘id.’ respective­ A typical instance of Latvian-origin verbal prefixes in Livonian ly. Borrowed prefixes occur in similar instances, for example ir- in in this context is presented below in example (1). Two Latvian-origin Finnish irrationaalinen ‘irrational’, in Estonian irratsionaalne ‘id.’. verbal prefixes uz- ‘on; onto, to’, and iz- ‘out’ appear with the Li­ Nevertheless, these are mainly attached to nouns forming antonymic vonian verbs, kērat-õ (write-inf) ‘write’, and lugg-õ (read-inf) ‘read’. pairs and are more clearly used as a word formation strategy or even In comparison with Latvian, these are the same prefixes which are as fully lexicalized forms. Still, Livonian is not the only Finnic lan­ used with the semantically equivalent Latvian verbs rakst-īt (write-inf) guage, which has adopted verbal prefixes. Veps and Karelian have ‘write’ and las-īt (read-inf) ‘read’ expressing perfective aspect and borrowed some Russian prefixed verbs, which occur most typically completed function. The verb forms uz-rakstī-t (pvb-write-inf) ‘write’, in code-switching instances. However, as a rule, these prefixes are and iz-lasī-t (pvb-read-inf) ‘read’ are illustrative examples of the inter­ only occasionally used in conjunction with other verbs. Furthermore, twining of aspectual marking, eventually leading to lexicalized verb Leivu, an extinct dialect that was spoken in northeast­ derivations (examples (2) and (3)3). ern Latvia, also used borrowed prefixes from Latvian. Considering other Uralic languages, verbal prefixes are also found in the Ugric (1) Livonian (PK) languages Khanty and Mansi and especially in Hungarian as well as un pǟlõ um uz-kērat-õd in Selkup. (EKG I: § 254, § 399–400, § 448; Honti 1979: 12, 2007: and onto is on-write-ppp 49; Kiefer & Honti 2003: 137–138; Mägiste 1937: 3–4, 16, 20; VISK väggõ lǟlam siz vȯl̦ iz-lugg-õ very difficult then was out-read-inf § 172, § 1630–1631; Wälchli 2001: 418.) ‘and it was written on it, it was then very difficult to read it (make it out)’ 1.3. The objectives of the study (2) Latvian (Diena 17.3.2014) My main hypothesis is that verbal prefixation in Livonian is not mere­ Grāmat-u iz-lasīj-a Latvija-s vēstnieks ASV… ly a lexical feature but has a grammatical impact as well. book-acc pvb-read.pst-3 Latvia-gen ambassador USA This paper has a threefold objective. ‘The Latvian ambassador in the US read the book…’ 1) It seeks to determine whether Latvian-origin verbal prefixes actually introduce a new grammatical category expressing per­ fective aspect in Livonian, and if so, which prefixes are used in 3. Here and henceforth the Latvian examples are presented in the orthography of this function. Standard Latvian and are drawn from newspaper texts available on the internet.

20 21 SANTRA JANTUNEN SYNTACTIC AND ASPECTUAL FUNCTIONS OF LATVIAN VERBAL PREFIXES IN LIVONIAN

(3) Latvian (Latvijas Avīze 20.2.2013) More recent research also has focused on the Latvian verbal pre­ Vin̦ uz-rakstīj-a otr-u disertācij-u… fixes in Livonian. Wälchli (2001: 418) offers a very brief description of he pvb-write.pst-3 second-acc dissertation-acc the Latvian prefixes and their use in Livonian. He argues that the pre­ ‘He wrote a second dissertation…’ fixnuo - ‘of; from’ is used as a default prefix in Livonian. Ernštreits & Kļava (2013) describe the Latvian influence on and Ernštreits & Kļava (2014) discuss the mutual contacts between Lat­ 1.4. Previous research on Livonian verbal prefixation vian and Livonian and the grammatical changes these contacts have induced. Both papers offer a short description of the use of the prefixes Linguists observed verbal prefixation in Livonian already in the 19th in Livonian and emphasize that their use is a recent and not extensive century. For example, the phenomenon is mentioned in the first Livo­ development. Ernštreits & Kļava also note that Latvian has served as nian grammar by A. . Sjögren (1861: 43–45). Sjögren, however, does an example when Livonian adverbs have developed into prefixes, for not use the term prefix. The prefixes, which have historically evolved example ilz-nūz-õ (up-rise-inf) ‘get up’. from Latvian prepositions (see Table 1.), are labelled as prepositions, Verbal prefixation in Livonian is also a subject in a few typologi­ whereas prefixes having a different origin are labeled as particles. cal studies. Arkadiev (2014) and (2015) are areal-typological studies Sjögren’s main aim is to briefly describe this phenomenon in Livonian of prefixal perfectivization. Arkadiev (2014) mentions Livonian as with examples and translations into German. The grammar does not an example of a language without native prefixes that has borrowed discuss this topic more extensively and lacks a description of the func­ prefixes and their functions, including imperfective and perfective as­ tions of these prepositions and particles. It is noteworthy that Sjögren pectual opposition, as a result of intensive language contact. Arkadiev takes a critical attitude towards what he finds in Livonian by claiming (2015) aims to describe aspectual systems based on verbal prefixes that the use of these prepositions and particles is “against the spirit of by distinguishing typologically considerable features and to find areal the Finnic languages”. features and possible correlations in these features, for example in the The most extensive study on Latvian-origin verbal prefixes in degree of grammaticalization of aspectual categories. Livonian is dis­ Livonian thus far is de Sivers’ study (1971). This study is a general cussed in relation to the influence of language contacts in the field of overview of the verbal prefixes and their use in Livonian, analyzing, prefixal perfectivization. Arkadiev concludes that Slavic-like aspectu­ for example, the relationship between prefixes and adverbs that are al categories do not arise in languages using verbal prefixes as a result attested and more commonly used in Finnic languages. The assumed of contact with Slavic or , whether the prefixes are aspectual function of verbal prefixes is not discussed in more detail native or borrowed as in Livonian. The systems are either less gram­ as a subject of its own. According to de Sivers (1971: 76), perfectiv­ maticalized or grammaticalized but divergent from the Slavic system. ity is marked in Livonian only with the prefix nuo- ‘of; from’. Fur­ Arkadiev (2017) deals with contact-induced borrowing of verbal pre­ thermore, she states, that the perfective use originates mainly from fixes and their functions in a few languages, including Livonian. He the functional model of the adverbial synonym jara ‘off, away’ (also concludes that the new aspectual systems in the recipient languages jarā, jerā, järā). Regarding other prefixes, an exact definition of their are noticeably less grammaticalized or based on different principles aspectual functions is not clear. Nevertheless, as the more detailed than grammatical aspectual categories in the donor languages, despite analysis of individual cases shows below, the grammatical role of the intensive language contacts. investigated prefixes needs to be discussed from other perspectives as well.

22 23 SANTRA JANTUNEN SYNTACTIC AND ASPECTUAL FUNCTIONS OF LATVIAN VERBAL PREFIXES IN LIVONIAN

2. Verbal prefixes expressing and durative aspects are distinguished in Latvian. All above-mentioned perfective aspect in Latvian4 prefixes, except the negative prefix ne- and debitive marker jā-, mark perfective verbs. It must be noted, however, that perfective meaning In general, Latvian verbal prefixes are used to express various gram­ depends also on the lexical meaning of an unprefixed verb and the matical functions and employed in both inflection and derivation as a interaction of the lexical meaning and the prefix, as well as the seman­ means of word formation and indexing adverbial syntactic functions. tics of the context. Verb stems typically express activities and states, As stated previously, in Latvian, there are altogether thirteen verbal while, for example, the notions of achievement and accomplishment prefixes, aiz- ‘behind’, ap- ‘around, about’, at- ‘away; back’, ie- ‘in, are expressed by means of prefixes. Thus, in these compounds a pre­ into’, iz- ‘out’, no- ‘of; from’, pa- ‘along, on, in; by; under; slightly’, fix expresses completion of the process or the activity expressed by a pār- ‘over, across’, pie- ‘by, at’, sa- ‘together’, uz- ‘on; onto, to’, a verb stem, for example, darī-t (do-inf) ‘do’ – pa-darī-t (pvb-do-inf) ‘do, negative prefix, ne-, and jā- used to form the debitive mood. The lat­ make; accomplish’, redzē-t (see-inf) ‘see’ – sa-redzē-t (pvb-see-inf) ter two have no role in the expression of perfectivity. The prefixes ‘see (catch sight of)’, ēs-t (eat-inf) ‘eat’ – sa-ēs-t-ies (pvb-eat-inf-rfl) no- ‘of; from’ and sa- ‘together’ are particularly common in Latvian. ‘eat [a lot], eat to satiety’. (Endzelin 1922: 739; Holvoet 2001: 146; The verbal prefixes are highly grammaticalized in Latvian, for exam­ Kalnača 2014: 89, 91–93, 109, 112; LVG: 531, 534; MLLVG: 569– ple they cannot appear by themselves without the verb, and they are 570; Wälchli 2001: 414, 433.) only one-syllable-long units. This is in accordance with the general Second, two syntactic means expressing imperfective/perfective phonological typology of prefixes as they tend to be short, consist­ aspectual opposition are constructions of an unprefixed verb and an ing most often of a single syllable in various languages. Bisyllabic adverb (a phrasal verb) and, in certain contexts, biaspectual (aspect­ prefixes seem to be comparatively rare. The aspectual function of the less) verbs. The aspectual opposition is manifested by an imperfec­ prefixes in Latvian, however, is considered to have been grammatical­ tive construction formed by an unprefixed verb and a local adverb ized only to a lesser extent, especially compared to the Slavic languag­ and perfective prefixed verb, for example, ie-t iekšā (go-inf inside) es. This is particularly evident in many biaspectual verbs in Latvian. ‘go in’ – ie-ie-t (pvb-go-inf) ‘id’.5 The majority of these verb pairs are (Arkadiev 2015: 284; Hall 2008: 535–536, 538–539; Holvoet 2001: verbs of motion. Biaspectual verbs can be unprefixed verbs or pre­ 132, 147, 157; Kalnača 2014: 93; LVG: 217, 284–285, 486; VISK § fixed verbs, which express both imperfective and perfective actions, 172; Wälchli 2001: 414.) for example, do-t (give-inf) ‘give’, pār-dot (pvb-give-inf) ‘sell’. Third, Latvian employs a few means to distinguish imperfective and perfect tense forms are also used to express perfective aspect in Lat­ perfective aspect. First, verbal prefixes characteristically mark perfec­ vian, for example, rakst-u (write-1sg) ‘I write’ : esmu rakstīj-is (be.1sg tivity by adding a prefix to an imperfective unprefixed verb, for exam­ write.pst-pa.sg.) ‘I have written’. (Holvoet 2001: 146–147; Horiguči ple ie-t (go-inf) ‘go’ – ie-ie-t (pvb-go-inf) ‘go in’, and rakstī-t (write-inf) 2017: 70; Kalnača 2014: 92, 95–105, 111–112; Wälchli 2001: 414.) ‘write’ – uz-rakstī-t (pvb-write-inf) ‘write onto; finish writing; write Even though prefixed verbs usually express perfectivity, the com­ (all)’; see also example (3). Thus, prefixed verbs make a distinction bination of a prefix and a verb in Latvian does not necessarily refer to between two aspectual types: imperfective/perfective aspect and se­ a perfective action. A prefixed verb can also be imperfective in certain melfactive/iterative aspect. The latter in Latvian is expressed by suf­ cases. This applies, for example, to biaspectual (aspectless) verbs (see fixation along with several morphophonological features (see, e.g., Endzelin 1922: 738–739; Kalnača 2014: 102–105). Usually, when Kalnača 2014: 91, 105–109; LVG: 539–541). In addition, inchoative 5. A construction formed by an unprefixed verb and an adverb has also other interpretations besides that of imperfective, see, e.g., the discussion in Kalnača 2014: 4. I am grateful to Uldis Balodis for his valuable comments concerning Latvian. 96–97.

24 25 SANTRA JANTUNEN SYNTACTIC AND ASPECTUAL FUNCTIONS OF LATVIAN VERBAL PREFIXES IN LIVONIAN combined with a verb, a prefix simultaneously expresses perfectivity In the Latvian grammatical tradition, prefixal derivation has been and modifies the lexical meaning of the verb spatially, temporally, or ambiguously treated (see Kalnača 2014: 89–90). These items are char­ quantitatively. Spatial expressions are typical for the prefixes devel­ acteristically defined as prepositions, which are used as a first compo­ oped from the spatial prepositions, for example, at-nes-t (pvb-bring-inf) nent of a compound, or as a morpheme, a prefix, with which verbs, ‘bring’, ie-nes-t (pvb-carry-inf) ‘carry, bring in’, iz-nes-t (pvb-carry-inf) nouns, and adjectives are derived. The Latvian prefixes have evolved ‘carry, bring out’, uz-nes-t (pvb-carry-inf) ‘carry, bring up’. Tempo­ from identical prepositions and most of them still maintain this bi­ ral meaning can focus on the beginning of the action, for example, functional relation both structurally and semantically. However, the aiz-smēk̦ ē-t (pvb-smoke-inf) ‘start to smoke’, or point to a momen­ meanings and use of prefixes are not limited to their semantics and tary, sudden action, for example, ie-raudā-t-ies (pvb-cry-inf-rfl) ‘start functions as prepositions. The prefixes at- ‘away; back’, ie- ‘in, into’, to cry and quickly stop’, ie-smie-t-ies (pvb-laugh-inf-rfl) ‘burst out iz- ‘out’, and sa- ‘together’ are exceptions, as these are not used as laughing; laugh once and then stop’, sa-bī-t-ies (pvb-fear-inf-rfl) ‘get prepositions in an adnominal function in modern Latvian. The mean­ frightened, startle’. Quantitative meanings are, for example, pa-gulē-t ing of the prefixie - ‘in, into’ is close to the meaning of the adverb iekšā (pvb-sleep-inf) ‘sleep a little, take a nap’, pār-ēs-t-ies (pvb-eat-inf-rfl) ‘inside’, whereas the sociative meaning of the prefixsa - ‘together’ has ‘eat to excess’. It is not unusual for a prefix to be polysemic and ex­ a parallel in the adverb kopā ‘together’. (LVG: 216–217, 284–285.) press, in addition to perfectivity, additional lexical meanings. For ex­ The primary meanings of prepositions which most commonly ample, the verb pa-līs-t (pvb-crawl-inf) can have both a spatial perfec­ occur as verbal prefixes are presented in Table 1. tive meaning ‘crawl under sth’ and a quantitative perfective meaning preposition primary meanings ‘crawl a little’. It is also possible, that a prefix, functioning as apréver - aiz ‘behind’ be vide (the term used by Holvoet 2001), only derives a perfective ap ‘around, about’ verb without changing or modifying the lexical meaning of a verb. no ‘of; from’ Examples of this kind of verb pairs are, for example, cep-t (roast-inf) – pa ‘along, on, in; by; under; slightly’ iz-cep-t (pvb-roast-inf) ‘roast; fry; bake’, pirk-t (buy-inf) – no-pirk-t pār ‘over, across’ (pvb-buy-inf) ‘buy’, darī-t (do-inf) ‘do’ – pa-darī-t (pvb-do-inf) ‘do, pie ‘by, at’ make; accomplish’, maksā-t (pay-inf) – sa-maksā-t (pvb-pay-inf) ‘pay’. uz ‘on; onto, to’ This characteristic is especially typical for the prefixes iz- ‘out’, no- ‘of; from’, pa- ‘along, on, in; by; under; slightly’, and sa- ‘together’, Table 1. The primary meanings of certain Latvian prepositions. and less frequently also for aiz- ‘behind’, ap- ‘around, about’, pie- ‘by, at’, and uz- ‘on; onto, to’. The estimation on how widely verbal pre­ A characteristic feature of the prefixes is that the same item can fixes appear as préverbes vides varies between authors. According to have both an adnominal and an adverbial function. In their adnominal LVG (534), there are just a few of this type of verb pair, in which the function, they act as prepositions and require a particular case for their prefix adds only a perfective meaning; Kalnača (2014: 93) also limits headword. In their adverbial function, they generally modify the lexi­ this to a few verb pairs, while Holvoet (2001: 132, 147) considers this cal meaning of the verb and can also function as a perfectivity marker. phenomenon not to be rare at all. Holvoet (2001: 135–136, 147) also Their semantic variety is also broad. A characteristic feature for pre­ remarks that only verb pairs derived with préverbes vides are purely fixes when creating verb forms is that the same prefix can add differ­ aspectual, but often it is not clear, whether there is a purely aspectual ent meanings when combined with different stems and also different opposition or whether the prefix still has a lexical meaning. (Endzelin prefixes can add the same meaning. For example, the prefixat - ‘away; 1922: 739–740; Kalnača 2014: 93–94; LVG: 534; MLLVG: 567–569.) back’ can express both the meaning ‘direction here’, for example,

26 27 SANTRA JANTUNEN SYNTACTIC AND ASPECTUAL FUNCTIONS OF LATVIAN VERBAL PREFIXES IN LIVONIAN at-skrie-t mājās (pvb-run-inf home.loc) ‘run home (from somewhere prefix number of occurrences else)’ and the meaning ‘direction away, off’, for example, at-grūs-t no aiz- ‘behind’ 16 sien-as (pvb-push-inf of wall-gen) ‘push away from the wall’. (Kalnača ap- ‘around, about’ 26 2014: 92–93; LVG: 212–220, 267–269, 278–286, 534, 625–627.) at- ‘away; back’ 23 ie- ‘in, into’ 27 iz- ‘out’ 69 3. The data nuo- ‘of; from’ 133 pa- ‘along, on, in; by; under; slightly’ 6 This study is based on data derived from unpublished recordings and pie- ‘by, at’ 33 published written material of spoken Courland Livonian. The record­ pōr- ‘over, across’ 5 ings were made by Professor Seppo Suhonen in Tallinn, Estonia in sa- ‘together; multitude’ 67 the 1970s and early 1980s. In total, there are 51 hours of recordings, uz- ‘on; onto, to’ 56 of which only a small share has been published (see Suhonen 1975). Total 461 Of these, I have transcribed one hour from one informant to be used Table 2. The number of occurrences of each individual prefix in the data of this in this study (referred to later as PK). In addition, I have also used study. material from Mägiste’s (2006) collection of narratives. I have in­ cluded all prefix and verb combinations in my data. Altogether there As Table 2 shows, there are significant differences in the number of are 461 verbal prefix occurrences. The narrative stories of the applied occurrences of each individual prefix. One prefix,nuo - ‘of; from’, has data include recollections, for example, childhood memories, how to a particularly high frequency in the data (ca. 29% of all occurrences). celebrate different holidays, how managed to flee from the The frequent use of this prefix was also noted by de Sivers (1971: Livonian Coast during the Second World War, as well as depiction of 76–79). According to Wälchli (2001: 418, 428), the same prefixno - is customs and traditions (e.g., funerals, fishing, how to brew beer). also the most common one in Latvian. The prefixes iz- ‘out’, sa- ‘to­ In Livonian, the use of verbal prefixes varies depending on the gether; multitude’, and uz- ‘on; onto, to’ are quite frequent as well, speaker. Suhonen (1972: 218) mentions two observations regarding whereas the prefixes pa- ‘along, on, in; by; under; slightly’ and pōr- this. Younger speakers are more likely to use the prefixes and an indi­ ‘over, across’ are rare. In Mägiste’s data (1937: 13–14), these frequen­ vidual Livonian speaker can also deliberately avoid using them (see cies are slightly different. In his data, the three most frequent prefixes also Suhonen 1975: 3). Also, the phenomenon seems to be spread ir­ are pie- ‘by, at’, nuo- ‘of; from’, and sa- ‘together; multitude’. In my regularly along the Livonian Coast: on some occasions, in the Livoni­ data, the prefixes iz- ‘out’, and uz- ‘on; onto, to’ are highly frequent, an villages of Vaid and Sīkrõg, the prefixes are used less frequently or but in Mägiste’s data these prefixes, as well as ap- ‘around, about’, not at all (Ernštreits & Kļava 2014: 83). The language in both of these at- ‘away; back’, and ie- ‘in, into’, only have average frequency. The villages is defined as the Eastern dialect of the , but least frequent prefixes in Mägiste’s data are aiz- ‘behind’, pa- ‘along, the villages do not neighbor each other. on, in; by; under; slightly’ and pōr- ’over, across’, which are also the All eleven Latvian prefixes borrowed into Livonian occur in the least frequent in my data. data, some of them more often than others. The number of the occur­ rences of each individual prefix is shown in Table 2.

28 29 SANTRA JANTUNEN SYNTACTIC AND ASPECTUAL FUNCTIONS OF LATVIAN VERBAL PREFIXES IN LIVONIAN

4. Verbal prefixes in Livonian (pour-inf) ‘pour’. The verb kītõ is a Finnic verb, while the verb vallõ is used in the Finnic and Mordvin languages (SSA 1: 360; SSA 3: 397). In order to examine different combinations between verbs and verbal (6) Livonian (Mägiste 2006: 100) prefixes, I have grouped prefixed verbs into subgroups on an etymo­ nǟ- ku kovāl ta vȯl̦ , logical and semantic basis. The verbs are grouped on the basis of their see-2sg that wise (s)he be.pst.3sg origin, namely, whether the entire prefixed verb is a loan from Lat­ ku ta iz iz-kīt taisnip-t! vian or the verb stem is Livonian but has a Latvian prefix. (See also when (s)he neg.pst out-tell truth-prt Suhonen 1972: 218, 1985: 113–115.) ‘You see that (s)he was wise, when (s)he did not tell the truth!’ The borrowed Latvian prefixed verbs form a numerous group in Livonian. They have been borrowed as whole verb forms, for example, (7) Livonian (PK) at-burr-õ (back-cast_a_spell-inf) ‘cast a counterspell’ (see Latvian at- siz ie-valā- sie sūr-õz bal̦ l̦ õ sizzõl then in-pour-3sg it.nom-gen big-ill basin.ill into bur-t (pvb-cast_a_spell-inf) ‘id.’; example 4), iz-duod̦ -õ (out-give-inf) ‘Then pour it into the big basin.’ ‘succeed, manage’ (see Latvian iz-do-t-ies (pvb-give-inf-refl) ‘id.’), iz-tiekk-õ (out-arrive-inf) ‘subsist’ (see Latvian iz-tik-t (pvb-arrive-inf) ‘id.’), and nuo-plutsīn̦ -țõ (of-clean-inf) ‘clean’ (see Latvian no-plucinā-t The Livonian calques of Latvian-origin verbs are typically iden­ (pvb-clean-inf) ‘1. pick off, pluck off, plume; 2. scald’; example 5). tical with their cognates in the source language (Suhonen 1985: 113). Also, the semantic correspondence with parallel Latvian prefixed (4) Livonian (Mägiste 2006: 138) verbs is high for the resulting Livonian prefixed verbs (Rudzīte 1996: un siz um ikš tuoi burānikkā, and then be.3sg one other witch 4; de Sivers 1971; Suhonen 1972: 219). This applies also to the func­ se-m at-burr-õn, ku läpš tions of these prefixed verbs. According to Mägiste (1937: 22) these it-be.3sg back-spell-app that child verbs are half-translated, in which the first part, that is, the prefix, äb ūo räukk-õn jembet. is a lexical loan, whereas the second part is translated from Latvian. neg be.cng cry-app more Suhonen (1972: 218) calls them partial calques. These verbs could ‘And then another witch, (s)he has cast a counterspell, also be considered as loanblends, i.e., borrowings, which contain both (so) that the child has not cried anymore.’ borrowed and native elements (Haspelmath 2009: 39). The following (5) Livonian (PK) verbs illustrate this kind of word formation: uz-kērat-õ (on-write-inf) gadāg-vied-kõks no-plutsīn̦ -țõd ‘write’ (see Latvian uz-rakstī-t (pvb-write-inf) ‘id.’; example 1), iz- juniper-water-ins of-clean-ppp lugg-õ (out-read-inf) ‘read’ (see Latvian iz-lasī-t (pvb-read-inf) ‘id.’; ‘cleaned with juniper water’ example 1); unprefixed verbs being kērat-õ (write-inf) ‘write’ (see Latvian rakstī-t (write-inf) ‘id.’), lugg-õ (read-inf) ‘read’ (see Latvian In my data, 71 Livonian verb stems combine with the prefixes. lasī-t (read-inf) ‘id.’). Frequent Livonian stems in the data are, for example, kīt-õ (say-inf) Livonian prefixed verbs with Livonian verb stems also share a ‘say, tell’ (see example 6), lǟ-dõ (go-inf) ‘go’ (see examples 16, 18, 30, common feature when it comes to a link between certain verb stems 35, 37), pān-da (put-inf) ‘put, set’ (see examples 15, 18, 27, 30), tīe-dõ and prefixes. For a number of Latvian prefixed verbs, there is a formal (do-inf) ‘do, make’ (see examples 9, 20, 23, 25, 36), tūl-da (come-inf) connection between the prefix and verb, but not a semantic connec­ ‘come’ (see examples 11, 16, 36), and van̦ țl-õ (look-inf) ‘look, watch’. tion, i.e., the meaning of the prefixed verb is case-specific (Kalnača Examples (6) and (7) show the verb stems kītõ ‘say, tell’ and vall-õ 2014: 95). This is also seen in certain Livonian prefixed verbs, for

30 31 SANTRA JANTUNEN SYNTACTIC AND ASPECTUAL FUNCTIONS OF LATVIAN VERBAL PREFIXES IN LIVONIAN example, pie-and-õ (by-give-inf) ‘forgive’ (see Latvian pie-do-t When the basis for grouping Livonian prefixed verbs is seman­ (pvb-give-inf) ‘id.’) and pie-pall-õ (by-request-inf) ‘worship’ (see Lat­ tics, the verbs of motion constitute a distinctive group. Within this vian pie-lūg-t (pvb-request-inf) ‘id.’); the corresponding unprefixed group, there are both Livonian- and Latvian-origin verbs. The Livo­ verbs are and-õ (give-inf) ‘give’ (see Latvian do-t (give-inf) ‘id.’) and nian verb tūl-da (come-inf) ‘come’ : tunnõ-d (come.ppp-pl) ’have come’ pall-õ (request-inf) ‘request’ (see Latvian lūg-t (request-inf) ‘id.’). (See (example 11) is attested in all Finnic languages, in Ter Saami, and the also Wälchli 2001: 418.) Example (8) shows the Livonian verb pie- Mari languages as well as in some (SSA 3: 324). pallõ ‘worship’. The verb ratst-õ (ride-inf) ‘ride’ : ratst-õn (ride-app) ‘have ridden’ (ex­ ample 12), an early Germanic loanword, has cognates in all Finnic (8) Livonian (Mägiste 2006: 104) languages (SSA 3: 54). The verb broutš-õ (travel-inf) ‘travel’ (occur­ se vȯl̦ ikš pālandõks kūož, -app.pl it be.pst.3sg one worship place rences aiz-broutš-õnd (behind-travel ) ‘left’ and ie-broutš-iz-õm kus jumāl-d ui-ž pīe-pall-õ. (in-travel-pst-1pl) ‘drive in’ in examples 13 and 14) is borrowed from where God-prt can-pst.3sg by-request-inf the Latvian verb brauk-t (travel-inf) ‘go; travel; drive; ride’; inflected ‘It was a place for worshiping, where one can worship God.’ forms with the stem alternation are, for example, brauc-u (travel-1sg) ‘I travel’ : brauc (travel.3) ‘(s)he/they travel(s)’. The Latvian model has also influenced some Livonian prefixed (11) Livonian (Mägiste 2006: 58) verbs by encouraging the use of a prefixed verb instead of the original pierrõ passoul̦ suoddõ ne sǟl āt ie-tunnõ-d. unprefixed verb. A good example of this is the verb at-mǟdl-õ (back- after world war.prt they there be.3pl in-come.ppp-pl remember-inf) ‘remember’ (example 9); namely, the unprefixed verb ‘After the World War they came there.’ mǟdl-õ (remember-inf) also means ‘remember’ (example 10). Its Lat­ vian equivalent is at-cerē-t-ies (pvb-hope-inf-rfl) ‘id.’, which is formed (12) Livonian (Mägiste 2006: 18) with the prefix at- and the verb cerēt ‘hope’; the prefixed form has a ne vȯl̦ -tțõ ni tikkiž piški irbõ-st reflexive suffix. The reflexive ending is not unusual for Latvian verbs they be-pst.3pl now all small Īra-ela when a prefix is used, especially when expressing short-term activity, at-ratst-õn õbbiz-t-õks kuolkkõ. away-ride-app horse-pl-ins Kūolka.ill partially resultative activity, the beginning of an activity, an action ex­ ecuted in great amount, and a mostly finished activity (Kalnača 2014: ‘They all have now ridden with horses from Īra to Kūolka.’ 94; LVG: 278, 534–535). Thus, in Latvian the meaning ‘remember’ is (13) Livonian (Mägiste 2006: 64) derived with the help of a prefix. This has very likely also influenced ku mēg vȯl̦ mõ aiz-broutš-õnd jetspēn̦ … the Livonian verb to adopt a prefix. when we be.pst.1pl behind-travel-app.pl away ‘When we had traveled away…’ (9) Livonian (PK) at-mǟdlõ-b ku vel̦ te-i pastāl-tõ (14) Livonian (Mägiste 2006: 61) back-remember-1sg when still make-pst.3sg pastāl-prt Ȭdõ-n kīela seissõ- ie-broutš-iz-õm zvīedõr ‘I still remember when they made pastāl-shoes.’ evening-ess clock seven-ins in-travel-pst-1pl Swedish sadāmō-zõ, mis nuttā-b pa nīnõshammõ-ks. (10) Livonian (Mägiste 2006: 195) harbor-ill which call-3sg ins Nynäshamn-ins siedā ma mǟdl-õb, se vȯl̦ sūr salāndõm. ‘In the evening, at seven o’clock, we arrived at it.prt I remember-3sg it be.pst.3sg big theft a Swedish harbor, which is called Nynäshamn.’ ‘I remember that it was a great theft.’

32 33 SANTRA JANTUNEN SYNTACTIC AND ASPECTUAL FUNCTIONS OF LATVIAN VERBAL PREFIXES IN LIVONIAN

In my data, all the Latvian-origin verbs of motion are used in Livo­ Tveite 2004: 38). This is in accordance with the fact that pronouns nian in the same way as in Latvian. For instance, in (13) and (14), the are known to have a different syntactic behavior compared to other prefixed verbs aiz-broutš-õ (behind-travel-inf) ‘leave’ and ie-broutš-õ word classes (in Livonian, e.g., pronouns have more inflected forms (in-travel-inf) ‘arrive’ are used in connection with leaving from and also in than other nouns). Moreover, this naturally increases arriving at a harbor in the same way as they are used in Latvian: aiz- the frequency and relative proportion of partitive objects in Livonian brauk-t prom (pvb-travel-inf ptcl) ‘travel away’ and ie-brauk-t ostā transitive clauses. (pvb-travel-inf harbor.loc) ‘arrive at a harbor’. Certain verbal particles are used to express aspectual oppositions If one considers Haspelmath’s (2009: 35) claim that borrowing in Livonian. There are some examples of these verbal particles com­ verbs is more difficult than borrowing nouns, it is quite striking that bined with prefixed verbs in my data. The most commonly occurring Livonian has borrowed verbs to this extent. He also states that verbs verbal particle is jarā ‘off, away’ (also jara, jerā, järā). Also, ilzõ ‘up’, require more grammatical adaptation than nouns. In addition, accord­ jūrõ ‘to, close’, kubbõ ‘together’, and lebbõ ‘through’ appear in my ing to Haspelmath (2009: 37), loanwords are not usually analyzable in data. (See also de Sivers 1971: 56–60.) the recipient language in the sense that transparent compounds, for ex­ In the following I cover the expression of perfectivity in Livo­ ample, are understood as compounds in the donor language, but they nian by means of verbal prefixes. In example (15), the verb sīe-dõ are unanalyzable and monomorphemic in the recipient language. (eat-inf) ‘eat’ is inflected without a prefix as se-i-ti (eat-pst-3pl) ‘they ate’ and the action is perceived as imperfective, a continuous process. Example (16) shows the same verb sīedõ with a prefix: no-sei (of- 5. Aspectual functions of eat.pst.3sg) ‘(s)he ate’. verbal prefixes in Livonian (15) Livonian (Mägiste 2006: 116) …kuolm vežžõ tǟdud sa-pandõd In my data, prefixes can either mark pure perfectivity or their seman­ …three basket.prt full.pl together-put.ppp tic scope has been extended so that a given verb can be semantically lēba-d-õks un se vȯl̦ nust-õd modified by the prefix. In some cases both functions apply. In certain bread-pl-ins and it be.pst.3sg lift-ppp cases, the prefix is exclusively used to form a perfective verb giv­ min pǟ pǟl ilz un līnd-ud ing the entire sentence a perfective reading. This is particularly char­ I.gen head onto up and bird-pl acteristic of the prefix nuo- ‘of; from’. In general, Finnic languages sǟl se-i-ti, knäpp-īz-ti jūs. there eat-pst-3pl peck-pst-3pl at express aspectual oppositions by means of alternative case markings ‘…three baskets (were) filled with bread and it was lifted of direct objects and different types of adverbials. For direct objects, up above my head and birds ate there, pecked.’ Livonian employs nominative-genitive (in many stem types nomina­ tive and genitive are identical in singular, and for all nouns in plural), (16) Livonian (PK) nominative, genitive, and partitive (Tveite 2004 calls non-partitive ja pierrõ ku kālm-i-stõ tul̦ kuonnõ cases accusative following the Finnish grammatical tradition). Of and after when grave-pl-ela come.pst.3sg at_home­ these, generally, non-partitive cases are used with perfective mean­ siz set ikš kõrd no-sei then only one time of-eat.pst.3sg ings, and partitive with imperfective meanings. A considerable mat­ un lekš jõgā ikš en̦ š kuonnõ ter concerning partitive objects in Livonian is the favoring of parti­ and go.pst.3sg each one own at_home tive case for pronouns, at least personal and pronouns ‘and after (they) came home from the graves [graveyard], then (Kont 1963: 103–106; Larsson 1983: 112–113; Sjögren 1861: 241; (they) ate only once and each one left for their homes’

34 35 SANTRA JANTUNEN SYNTACTIC AND ASPECTUAL FUNCTIONS OF LATVIAN VERBAL PREFIXES IN LIVONIAN

In examples (17) and (18), the same prefixed verb, nūo-sīe-dõ (of- (19) Livonian (Mägiste 2006: 40) eat-inf) ‘eat’, appears three times. In example (17), a perfective inter­ pǟgin̦ perīmie-n vȯl̦ iz-brõuv-dõd vȯl̦ un pretation is supported by the verbal particle järā ’off, away’. In exam­ much householder-dat be.pst.3sg out-brew-ppp beer and ple (18), however, both verbs cannot be interpreted as perfective. The sīemizõ-ks kui-di jērn-idi sigā vȯzā pǟ-kõks nuo-kiett-õd. latter has a perfective meaning, implying the time when the process of eating-ins dry-prt.pl pea-prt.pl pig meat head-ins of-boil-ppp eating has ended or when the subjects have finished eating. The first ‘The householder had a lot of brewed beer and for food dry peas (and) pork boiled with the head.’ is more ambiguous and, in this example, is not clearly recognized as a perfective verb. (20) Livonian (PK) (17) Livonian (Mägiste 2006: 120) bēr-id̦ i tīe-b emmit pierrõ siedā funeral-prt.pl make-3sg more after it.prt no ne āt vȯnnõ-d sie ku um nuo-matt-õd ptcl they be.3pl be.ppp-pl it.nom-gen when be.3sg of-bury-ppp lēba järā nuo-sīe-nõd… ‘Funerals were rather held after the burial.’ bread.nom-gen ptcl of-eat-app.pl ‘They have finished that bread…’ Comparing example (20) with example (21) below, the unprefixed (18) Livonian (Mägiste 2006: 120) verb matt-õd bury-ppp ‘bury’ appears in the same form as …laz ānda-g amā-d-õn nūo-sīe-dõ the prefixed verbnuo-matt-õd (of-bury-ppp) ‘bury’ in the same context. imp give-imp.3sg all-pl-dat of-eat-inf Nevertheless, the situation depicted in the sentence is perfective with un ibbiz-t-õn ka laz ānda-g ainõ, and horse-pl-dat also imp give-imp.3sg hay.prt a nominative object mingi suodāmiez ‘a soldier’. Thus, it is possible pan-g jeddõ, un ku āt nūo-sīe-nõd, to conclude that the use of a prefixed verb is not obligatory when ex­ put-imp.3sg in_front and when be.3pl of-eat-app.pl pressing perfectivity because the perfective meaning can be present siz laz läk valā-m vil̦ l̦ õ kuoțțõ, also without a prefix. De Sivers (1971: 61) also mentions that a ver­ then imp go.imp.3sg pour-inf grain.prt sack.ill nacular, which has borrowed different kinds of Latvian elements, has amā-d-õn kuoțț-īd tǟduks. all-pl-dat sack-pl full occasionally rejected these prefixes. According to her data, there are ‘…let everyone eat and give even the horses hay, cases, in which an unprefixed verb is used where it would be possible put (it) in front of them, and when they have eaten, to have a prefixed verb instead and also vice versa. Furthermore, she go pour the grain into a sack, fill everyone’s sacks.’ adds that sometimes both variants are displayed in the same context or even in the same sentence. This shows that on occasion the prefix only In example (19), the prefix nuo- is attached to the verb kīet-õ emphasizes or confirms the meaning of the verb in question. Likewise, (boil-inf) ‘boil’, whereas in example (20) it combines with the verb Arkadiev (2015: 257) mentions Livonian as one of the languages in matt-õ (bury-inf) ‘bury’, forming the perfective verbs nuo-kīet-õ (of- which both unprefixed and prefixed verbs may appear in the same con­ boil-inf) : nuo-kīet-õd (of-boil-ppp) ‘boil’ and nuo-matt-õ (of-bury-inf) : text without a significant difference in their meaning. nuo-matt-õd (of-bury-ppp) ‘bury’. In example (19), the verb nuo-kīet-õd (21) Livonian (Mägiste 2006: 62) has two direct objects: kui-di jērni-di (dry-prt.pl pea-prt.pl) ‘dry peas’ in …kus sǟl mingi suodā-miez um matt-õd. the partitive, because the object is an uncountable plural, and sigā vȯzā where there some war-man be.3sg bury-ppp ‘pork’ in the nominative-genitive marking perfectivity. Example (20) ‘…where a soldier has been buried there.’ does not show a direct object despite the presence of a transitive verb.

36 37 SANTRA JANTUNEN SYNTACTIC AND ASPECTUAL FUNCTIONS OF LATVIAN VERBAL PREFIXES IN LIVONIAN

Thus, the prefix nuo- often derives perfective verbs, but this is not al­ The prefixesnuo - and iz- seem to truly have a perfective function, and ways the case. In my data, there are also other prefixes, which may be they often seem to indicate that the action or process expressed by the used to derive perfective verbs, at least in certain instances. In examples verb stem has been accomplished. (22), (23), and (24), the prefix iz-, the second most common prefix by In some cases, the prefixes both derive a perfective verb and frequency in my data, is used to form perfective verbs. In example (22), modify the lexical meaning of the verb, adding to it an additional, the prefixed verbiz-lugg-õ (out-read-inf) : iz-lugg-õnd (out-read-app.pl) adverbial meaning as in Latvian; also the adverbial meaning is often ‘read’ appears with a direct object bībõl ‘Bible’ in nominative-genitive identical with Latvian (see Kiefer & Honti 2003: 138, 147; Suhonen and a verbal particle lebbõ ‘through’, thus employing three means to 1972: 219). Still, in many cases in Latvian and Livonian, these lexical express perfectivity. In example (23), the prefixed verb is iz-tīe-dõ meanings of individual verbs are not identical. The prefix sa- often (out-make-inf) ‘do, make’: iz-te-i (out-make-pst.3sg) ‘(s)he did, made’, adds the Livonian verb adverbial meaning, such as ‘together; multi­ and, in example (24), the intransitive verb iz-dan̦ tš-õ (out-dance-inf) tude’. The verbal particle kubbõ ‘together’ often appears in connection ‘dance’ : iz-dan̦ tš-õnd (out-dance-app.pl) ‘(s)he danced’. Also, in ex­ with the prefix sa-. A large quantity is often emphasized by adding ample (19), the prefixed verb iz-brõuv-dõd (out-brew-ppp) ‘(s)he the adverb pǟgin̦ ‘a lot’ to the verb phrase, see example (27). In some brewed’ has a direct object vȯl̦ ’beer’ in nominative-genitive (see also cases, the prefix sa- ‘together; multitude’ also acts as a perfectivity examples 1 and 6). marker as in (25), (26), and (27). The direct objects in (25) and (26) are in nominative-genitive and nominative respectively, while in (27), (22) Livonian (Mägiste 2006: 42) the object is not displayed. (See also example (15), in which the pre­ vanā-miez kīt-, ku ne āt kuolm nēl̦ a old-man say-pst.3sg that they be.3pl three four fix sa- ‘together; multitude’ in the prefixed verb sa-pandõd (together- kõrd iz-lugg-õnd bībõl lebbõ un vel put.ppp) ‘put’ emphasizes the great amount of bread or the fullness of time.prt out-read-app.pl Bible ptcl and yet baskets in addition to perfectivity.) tō-b ikš-kõrd lugg-õ lebbõ. want-3sg one-time read-inf ptcl (25) Livonian (Mägiste 2006: 5) ‘The old man said that they have read the Bible through siz vȯl̦ sa-tie-dõd val̦ mõks three-four times and (they) want to read it once more.’ then be.pst.3sg together-make-ppp ready selliz-t āina-d: tīeme jou-t-kõks sog-dõd (23) Livonian (PK) such-pl medicine-pl yeast.gen flour-pl-ins mix-ppp un sie sǟr-stõ sīe-stõ tegīž kubbõ un siz pār päv-ḑi pid-tõd. together and then couple day-prt.pl keep-ppp and it.nom-gen leg-ela it-ela again iz-te-i pastāl̦ i ‘Then such medicines were made: yeast was mixed with out-make-pst.3sg pastāl.prt.pl flour and then it was left to stand for a couple of days.’ ‘A nd pastāl-shoes were made from the (boot) shaft.’ (26) Livonian (Mägiste 2006: 81) (24) Livonian (Mägiste 2006: 24) tämmõ-n vȯl̦ sa-kērat-õd āiga-rōntõz (s)he-dat be.pst.3sg together-write-ppp time-book ku vȯl̦ țțõ iz-dan̦ tš-õnd, siz laps-t kädst when be.pst.3pl out-dance-app.pl then child-pl post amā-d nim-īd-õks, touvvõ tǟd-õd-õks. all-pl name-pl-ins sky.gen star-pl-ins īrg-iz-t kizzõ pōțțeri sōrmō kīel-kõks. begin-pst-3pl ask.inf prayer.prt.pl language-ins ‘(S)he had a calendar written with all the names, the stars of the sky.’ ‘When (they) had danced, then (they) began to ask for prayers from the children in the language of Saaremaa [Estonian].’

38 39 SANTRA JANTUNEN SYNTACTIC AND ASPECTUAL FUNCTIONS OF LATVIAN VERBAL PREFIXES IN LIVONIAN

(27) Livonian (PK) For the prefixes ie- ‘in, into’ and ap- ‘around, about’ the main adver­ sūomõ-d jo vȯl̦ țõ väggõ pǟgin̦ sa-sōt-õnõd bial meanings are ‘in, into’, and ‘around’, respectively. In some cases, Finnish-pl already be.pst.3pl very much together-send-app.pl the prefixes are used to express perfectivity, as in examples (30), (31), ‘Finnish people had already sent quite a lot (of gifts).’ (32), and (33). Again, in examples (30) and (32), direct objects are in nominative-genitive and genitive, respectively (võrgõ-d (fishnet- The prefixuz - also has an additional lexical meaning besides express­ nom-gen.pl) ‘fishnets’, sel̦ l̦ i-z (such-gen) ‘such’), while examples (31) ing perfectivity. The secondary adverbial meaning is ‘on, onto’ as the and (33) lack objects. The perfective function appears infrequently for prefix often appears with the adverb or postposition pǟl, pǟlõ ‘on, these prefixes as well, as is also the case for the prefix uz- ‘on; onto, onto’ (see examples 1, 29, and 30). It seems, however, that uz- is used to’. to derive perfective verbs only sporadically. Nevertheless, there are (30) Livonian (Mägiste 2006: 21) some cases such as examples (1), (28), (29), and (30) that are excep­ ie-ētt-iz-õm võrgõ-d sūr-õ mierrõ, tions to the rule. In examples (28) and (30), the direct object is also in-throw-pst-1pl fishnet-nom-gen.pl big-ill sea.ill in the nominative-genitive (mǟdlõttõbkiuv ‘tombstone’, lōja ‘boat’), pūrit-iz-õm lōja-ks aigõ; ku sa-i-m and in example (29), in the nominative (tēd̦ ‘star; letter; symbol’), all sail-pst-1pl boat-ins shore.ill when get-pst-1pl of these refer to a perfective reading. Also, the verbal particle ilzõ aigõ, uz-vied-īz-õm lōja ilzõ ‘up’ appears in example (30). Hence, the prefixuz - primarily gives the shore.ill on-drag-pst-1pl boat.nom-gen up verb only the adverbial meaning ‘on, onto’ and does not have a clear rānda pǟl un lek-š-õm kuodāi. shore onto and go-pst-1pl to_home perfective function. ‘We threw fishnets into the Great Sea [Baltic Sea], sailed (28) Livonian (PK) the boat to the shore; when we got to the shore, we kālma-d-õn uz-pan̦ mǟdlõtõb-kiuv dragged the boat onto the shore and went home.’ grave-pl-dat on-put.pst.3sg memory-stone.nom-gen (31) Livonian (PK) un ka uz-pan̦ rānda-kīel-kõks and also on-put.pst.3sg shore-language-ins un sǟl um ie-pandõd pū sizāl and there be.3sg in-put.ppp tree into ‘(s)he placed a tombstone on the graves and also wrote in the coastal language [Livonian]’ ‘And there (it) is put into a tree.’

(29) Livonian (Mägiste 2006: 28) (32) Livonian (PK) jegā īdõ-n vȯl̦ uz-ummõl-tõd – siz tegiž jemā vȯl̦ ap-ēd̦ t-õn sel̦ l̦ i-z every one-dat be.pst.3sg on-sew-ppp then again mother be.pst.3sg around-dress-app such-gen mie-d-õn sǟlga pǟl un naiz-t-õn ‘Then again mother has put such (a skirt) on.’ man-pl-dat back.nom-gen onto and woman-pl-dat rīnda-d pǟl – tēd̦ , neikku (33) Livonian (PK) chest-nom-gen.pl onto symbol so_that siz vel razā-ks ap-ūd-iz või-b id-tuoiz-ta tund-õ. then still grease-ins around-fry-pst.3sg be.able-3sg one-other-prt know-inf ‘Then [the turbot (Scophthalmus maximus)] ‘A symbol had been sewn onto everyone – for men on the back was still fried with grease.’ and for women on the chest – so they can recognize each other.’

40 41 SANTRA JANTUNEN SYNTACTIC AND ASPECTUAL FUNCTIONS OF LATVIAN VERBAL PREFIXES IN LIVONIAN

Regarding the use of the less frequent prefixes in my data, their in the plural (Grünthal 2007: 408, 414; 2010: 101, 104–106; Kettunen use often corresponds to the use of their Latvian equivalents; for ex­ 1938: XXXVIII–XLI, LIV–LVII). Usually, it is only possible to spec­ ample, the prefix pie- ‘by, at’ in pie-pall-õ (by-request-inf) ‘worship’, ify the case of the object as non-partitive. In Finnic languages, the dis­ in example (8), and the highly infrequent prefixes pa- ‘along, on, in; tribution of non-partitive cases of direct objects varies depending on by; under; slightly’ and pōr- ‘over, across’ in pa-kīer-õ (along-turn-inf) the language in question, but in negative sentences the direct object is ‘turn slightly’: pa-kīer-õb (along-turn-3sg) ‘(s)he turns slightly’, in typically in the partitive. In Livonian, however, also non-partitive ob­ example (34), and pōr-lǟ-dõ (over-go-inf) ‘get across; cease, stop’ : jects in negative sentences are possible in certain conditions (see Kont pōr-lǟ-nd (over-go-ppp) ‘gotten across; ceased, stopped’, in example 1963; Larsson 1983; Sjögren 1861: 238–257; Tveite 2004). According (35). In example (34), the prefix pa- adds the verb the same meaning to Tveite (2004: 147, 150), these objects are exceptions, and my data, of ‘slightly, a little’, as in Latvian, and in example (35) the prefixpōr - though quite sparse, support his perception. In his comprehensive data implies the meaning ‘over’ as well as perfectivity (other possible lexi­ with nearly 7 000 sentences, there are only 635 negative sentences, of cal meanings for pār- in Latvian are e.g. ‘again’ and ‘wrong’ (LVG: which 63 have a non-partitive object. In other respects, Tveite (2004) 282–284). In Livonian, these prefixes rarely function as perfective only touches on negation in his study of the Livonian object. markers, although in Latvian they are used in this function. Affirmative predicates are also - over represented in my data, since there are only 24 examples with negative predicates in the total (34) Livonian (PK) of 461 examples of verbal prefixes. Prefixes occur in connection with siz lougõ lougõ ju pa-kīer-õb siedā then slowly slowly ptcl along-turn-3sg it.prt both intransitive and transitive negative verbs. In this study, negative pū-dõ mis sǟl sizāl um transitive verbs with a displayed direct object are of interest. They piece_of_wood-prt that there in be.3sg compose approximately half of the negative examples of my data. In ‘Then slowly, slowly one slightly turns that two examples, the object is a noun, in one example, it is a noun phrase piece of wood which is in there.’ with a pronoun and noun, and in the remaining examples, the object is a pronoun. No non-partitive objects in negative sentences appear (35) Livonian (Mägiste 2006: 133) in my data. For the sake of comparison, in example (36), the prefixed ku vȯl̦ pōr-lǟ-nd se laskimi, verb nuo-matt-õd (of-bury-ppp) ‘buried’ occurs in a negative sentence. when be.pst.3sg over-go-ppp it shooting siz mēg īe-krōm-iz-mõ vīl̦ a vaguonõz, In example (37), the prefixed verbnuo-tie-mõ (of-make-inf) ‘do, make’ then we in-crawl-pst-1pl property car occurs with the pronominal object siedā (it.prt) ‘it’; see also example amā vīl̦ a-ks un broutš-iz-mõ (6) with the partitive object taisnip-t (truth-prt) ‘truth’. all property-ins and travel-pst-1pl bōn̦ õ-ks sie tul mierrõ-n piddiz. (36) Livonian (PK) train-ins it.gen fire sea-dat along agā jedmõl kui iz ūo nuo-matt-õd ‘When the shooting had ended, then we crawled into the freight car, but before when neg.pst be.cng of-bury-ppp with all the wares, and traveled with the train along that sea of fire.’ siz nut-īz pa tōppin̦ õ-ks. then call-pst.3sg ins diluted_beer-ins ‘But before the burial, it was called diluted beer.’ As stated above, in Livonian, as in the Finnic languages generally, direct objects are marked with nominative, genitive, and partitive. Li­ vonian exhibits extensive case syncretism between the nominative and genitive in many inflectional types, in the singular and systematically

42 43 SANTRA JANTUNEN SYNTACTIC AND ASPECTUAL FUNCTIONS OF LATVIAN VERBAL PREFIXES IN LIVONIAN

(37) Livonian (Mägiste 2006: 135) Suhonen 1972: 220, 1985: 115–116; de Sivers 1971). In addition, the agā siedā iz sō nuo-tie-mõ, prefixes iz- ‘out’ and sa- ‘together; multitude’, and more or less also but it.prt neg.pst get.cng of-make-inf uz- ‘on; onto, to’, ie- ‘in, into’, and ap- ‘around, about’, can be used vȯl̦ lē-mõst näntõ-n uldzõ, to derive perfective verbs. Although, in addition to their perfective be.pst.3sg go-inf they-dat.pl out function, these prefixes also are used to express certain lexical mean­ saksā tul̦ sillõ. German come.pst.3sg in ings, except for the prefix iz- ‘out’. These lexical meanings are: sa- ‘But it cannot be done, they had to go [out], ‘together; multitude’, uz- ‘on; onto, to’, ie- ‘in, into’, and ap- ‘around, the Germans came in [to Latvia].’ about’. These prefixes seem to form an exception to Wälchli’s (2001: 418–419) observation that the use of verbal prefixes does not have a Thus, the verbal prefix does not seem to affect the choice of the case significant influence on the sentence on a semantic level. Although, of the direct object in negative sentences, but the choice is based more none of these prefixes derives a perfective verb in every case where likely on the negation itself and the specific conditions mentioned by it occurs with a verb. Also, in most cases, the use of verbal prefixes Tveite (2004). generally corresponds with the Latvian use of an equivalent prefix, and the Livonian prefixed verbs often have Latvian equivalents both in form and meaning (de Sivers 1971; Suhonen 1972: 219, 1985: 113). 6. Conclusions De Sivers (1971: 80) remarks, however, that Livonian prefixed verbs are gradually separating from the Latvian model because there are also Verbal prefixation in Livonian is a phenomenon that occurs lexically, some instances where they do not correspond formally or semanti­ with eleven Latvian-origin prefixes, and also grammatically, as at least cally to Latvian prefixed verbs. My data also point to that develop­ some of the borrowed prefixes have grammatical functions as well. ment, concerning especially the prefix nuo- ‘of; from’. Concerning Morphological elements, i.e., the verbal prefixes, can be borrowed as a the lexical meanings, the corresponding use is evident especially in lexical element, either by themselves or as a combination with a verb. the case of the rarely used prefixes. However, these prefixes typically Functional features, that is, aspectual functions and possible simulta­ do not have a perfective use in Livonian, though it might be possible neous adverbial meanings have been borrowed in addition to lexical to distinguish it in some situations, for example, with verbs of motion. elements, applying especially to the prefixes nuo- ‘of; from’ and iz- Usually, these prefixes only change or modify the lexical meaning ‘out’, and to some extent also to the prefix sa- ‘together; multitude’, of the verb. This applies especially to the prefixes aiz- ‘behind’, at- and to a lesser extent to the prefixes uz- ‘on; onto, to’, ie- ‘in, into’, ‘away; back’, pa- ‘along, on, in; by; under; slightly’, pie- ‘by, at’, and and ap- ‘around, about’. pōr- ‘over, across’. This also differentiates the use of these infrequent Differences in the frequency of occurrence of the different pre­ prefixes in Livonian from the Latvian use of the equivalent prefixes, fixes represent interesting findings. According to my data, it seems because in Latvian these prefixes are used as perfectivity markers. It that the most frequent prefixes are used as perfectivity markers in Li­ seems that the frequency of the prefix correlates with its aspectual, vonian. The most frequent prefix, nuo- ‘of; from’ especially can be perfective function in Livonian. In other words, the more frequent the considered not only a lexical element but also functioning as a per­ prefix, the more probable it is that it has an aspectual function. fectivity marker. This probably results from the high frequency of the In Livonian, prefixed verbs, and hence the sentences in which prefix. It might even be considered as a default perfectivizing prefix in they occur, are not necessarily perfective. Instead, the prefix often Livonian (see also Wälchli 2001: 418). Also, it seems to be developing gives the verb an adverbial, modifying, or emphasizing meaning. a use divergent from the Latvian use of the equivalent prefix no- (see Thus, one cannot say that verbal prefixation has fully developed to

44 45 SANTRA JANTUNEN SYNTACTIC AND ASPECTUAL FUNCTIONS OF LATVIAN VERBAL PREFIXES IN LIVONIAN express perfectivity in Livonian, or that is found References in Livonian (see also Arkadiev 2017). Also, the use of the verbal prefixes in Livonian does not appear systematically, since they are Arkadiev, Peter 2014: Towards an areal typology of prefixal perfectivization. not obligatory when expressing aspectual oppositions. Unprefixed – Scando-Slavica 60:2: 384–405. verbs may perfectly well occur in perfective contexts. My data sup­ — 2015: Arealʹnaja tipologija prefiksalʹnogo perfektiva (na materiale port Arkadiev’s (2015: 290) perception that a recipient language does jazykov Evropy i Kavkaza) [Areal typology of prefixal perfectiviza­ tion (Based on the material of languages of and the Caucasus)]. not adopt the system of a donor language completely, even when the Studia philologica. Rossijskaja Akademija Nauk. Institut slavjanove­ language contact in question has been long-standing and intensive. denija. Otdel tipologii i sravnitelʹnogo jazykoznanija. Centr balto-slav­ Still, as mentioned above, some of the Latvian-origin verbal prefixes janskix issledovanij. Moskva: Jazyki slavjanskoj kulʹtury. (nuo- ‘of; from’, iz- ‘out’, sa- ‘together; multitude’, uz- ‘on; onto, to’, — 2017: Borrowed preverbs and the limits of contact-induced change in ie- ‘in, into’, ap- ‘around, about’) are used to derive perfective verbs aspectual systems. – Rosanna Benacchio, Alessio Muro & Svetlana in Livonian in at least some instances. Therefore, a secondary strategy Slavkova (eds), The role of prefixes in the formation of aspectuality. has emerged to express perfectivity in Livonian, which often is used Issues of grammaticalization. Biblioteca di Studi Slavistici 39. Firenze: in conjunction with the inherent Finnic way of doing so, which uses Firenze University Press. 1–21. alternative case marking of direct objects and verbal particles. And, Auziņa, Ilze 2014‒2016: Fonētika. Skaņu un burtu atbilsme [Phonetics. naturally, the perfective use of these prefixes in Livonian does not The correspondence between the sounds and the letters]. – Ilga Jan­ in any way approach the extent of their use in Latvian where they sone & Anna Vulāne (eds), Latviešu valodas rokasgrāmata [Hand­ originate. book of the ]. Rīga: Latviešu valodas aģentūra. [Accessed 31 Jan 2019.] Available at: ‹http://valodasrokasgramata.lv/ materials/1/21D1A811-6B6E-5F10-894A-09ED1ABA1A87› Bertinetto, Pier Marco & Denis Delfitto 2000: Aspect vs. actionality: Why Abbreviations they should be kept apart. – Östen Dahl (ed.), Tense and aspect in the . Berlin – New York: Mouton de Gruyter. 1 first person ins instrumental 189–225. 2 second person neg negation Borik, Olga 2006: Aspect and reference time. Oxford Studies in Theoretical 3 third person nom nominative Linguistics. New York: Oxford University Press. acc accusative nom-gen nominative-genitive Comrie, Bernard 1976: Aspect. An introduction to the study of verbal aspect ade adessive pa active participle and related problems. Cambridge: Cambridge University Press. all allative PK Initials for a Livonian Dahl, Östen 1985: Tense and aspect systems. Oxford: Blackwell. app active past participle language consultant Diena = Kaspars Odiņš: Romāna Ļeņina harēms autors: Man svarīgi, lai šo cng connegative pl plural grāmatu pieņemtu latvieši un amerikāņi. – Diena 17.3.2014 [Latvian news­ cond conditional post postposition paper]. [Accessed 10 Sept 2015.] Available at: ‹https://www.diena.lv/raksts/ dat dative ppp passive past participle kd/intervijas/romana-_lenina-harems_-autors-man-svarigi-lai-so- ela elative prt partitive gramatu-pienemtu-latviesi-un-amerikani-14048424› gen genitive pst past Ebert, Karen . 2000: Aspect in Maltese. – Östen Dahl (ed.), Tense and ill illative ptcl particle aspect in the languages of Europe. Berlin – New York: Mouton de imp imperative pvb preverb Gruyter. 189–225. ine inessive rfl reflexive inf sg singular

46 47 SANTRA JANTUNEN SYNTACTIC AND ASPECTUAL FUNCTIONS OF LATVIAN VERBAL PREFIXES IN LIVONIAN

EKG I = Erelt, Mati, Reet Kasik, Helle Metslang, Henno Rajandi, Kristiina Ross, Holvoet, Axel 2001: Studies in the Latvian verb. Baltica Varsoviensia. Tom Henn Saari, Kaja Tael & Silvi Vare 1995: Eesti keele grammatika I. Morfo­ czwarty. Kraków: Wydawnictwo Uniwersytetu Jagiellońskiego. loogia. Sõnamoodustus [ I. Morphology. Word forma­ Honti, László 1979: Characteristic features of Ugric languages (observations tion]. Tallinn: Eesti Teaduste Akadeemia Keele ja Kirjanduse Instituut. on the question of Ugric unity). – Acta Linguistica Academiae Scien- EKG II = Erelt, Mati, Reet Kasik, Helle Metslang, Henno Rajandi, Kristiina tiarum Hungaricae 29: 1–26. Ross, Henn Saari, Kaja Tael & Silvi Vare 1993: Eesti keele gramma- — 2007: Syntaktische Konstruktionen fremden Ursprungs. Eine skep­ tika II. Süntaks [Estonian grammar II. Syntax]. Tallinn: Eesti Teaduste tische Stellungnahme. – Incontri Linguistici 30: 49–70. Akadeemia Keele ja Kirjanduse Instituut. Horiguči, Daiki 2017: Laika akuzatīva un lokatīva opozīcija un aspekts lat­ Endzelin, J. 1922: Lettische grammatik. Riga: Kommissionsverlag A. Gulbis. viešu valodā [Opposition of the temporal accusative and locative and Erelt, Mati & Helle Metslang (eds) 2017: Eesti keele süntaks [Syntax of the aspect in Latvian]. – Valoda: nozīme un forma 8. Valodas gramatiskās Estonian language]. Eesti keele varamu 3. Tartu: Tartu Ülikooli Kirjastus. un leksiskās sistēmas variatīvums. 70–78. Ernštreits, Valts & Gunta Kļava 2013: Dažas gramatiskas pārmaiņas lībiešu Kalnača, Andra 2014: A Typological perspective on Latvian grammar. War­ valodā latviešu valodas ietekmē [Some grammatical changes in Livo­ saw ‒ Berlin: De Gruyter Open. nian influenced by Latvian]. – Valoda: nozīme un forma. 3. Teorija un Kettunen, Lauri 1938: Livisches Wörterbuch mit grammatischer Einleitung. metodoloģija latviešu valodniecībā. 19–27. Lexica Societatis Fenno-Ugricae V. Helsinki: Suomalais-Ugrilainen Seura. — 2014: Grammatical changes caused by contact between Livonian and Kiefer, Ferenc & László Honti 2003: Verbal “prefixation” in the Uralic lan­ Latvian. – ESUKA – JEFUL 5: 77–90. Available at: ‹https://doi.org/ guages. – Acta Linguistica Hungarica 50 (1–2): 137–153. 10.12697/jeful.2014.5.1.05› Klein, Wolfgang 1994: Time in language. London: Routledge. Grünthal, Riho 2003: Finnic adpositions and cases in change. Mémoires de Kont, Karl 1963: Käändsõnaline objekt läänemeresoome keeltes [The nomi­ la Société Finno-Ougrienne 244. Helsinki: Finno-Ugrian Society. nal object in the Finnic languages]. Keele ja Kirjanduse Instituudi — 2007: Morphological change and the influence of language contacts uurimused IX. Tallinn. in Estonian. – Hans Fix (ed.), Beiträge zur morphologie: Germanisch, Larsson, Lars Gunnar 1983: Studien zum Partitivgebrauch in den ostseefin- Baltisch, Ostseefinnisch. North-Western European language evolution. nischen Sprachen. Studia Uralica et Altaica Upsaliensia 15. Uppsala. Supplement vol. 23. University Press of Southern Denmark. 403–432. Latvijas Avīze = Dace Kokareviča: Jāņa Endzelīna dzīve ir saistoša romāna — 2010: Sijasynkretismi morfologian koetinkivenä [Case syncretism as a vērta. – Latvijas Avīze 20.2.2013 [Latvian newspaper]. [Accessed touchstone of morphology]. – ESUKA – JEFUL 2: 91–113. Available at: 10.9.2015.] Available at: ‹http://m.la.lv/jana-endzelina-dzive-ir-saistosa- ‹https://doi.org/10.12697/jeful.2010.1.2.06› romana-verta-2/› — 2015: Livonian at the crossroads of language contacts. – Santeri Junt­ LVG = Auziņa, Ilze, Ieva Breņķe, Juris Grigorjevs, Inese Indričāne, Baiba tila (ed.), Contacts between the Baltic and Finnic languages. Uralica Ivulāne, Andra Kalnača, Linda Lauze, Ilze Lokmane, Dace Markus, Helsingiensia 7. Helsinki: University of Helsinki Finno-Ugric Lan­ Daina Nītiņa, Gunta Smiltniece, Baiba Valkovska & Anna Vulāne guage Section & Finno-Ugrian Society. 97–150. ‹https://journal.fi/ 2013: Latviešu valodas gramatika [Latvian grammar]. Rīga: LU uralicahelsingiensia/issue/view/uh7/uh7› Latviešu valodas institūts. Hall, Christopher J. 2008: Prefixation, suffixation and circumfixation. – Mägiste, Julius 1937: Laenatud indo-euroopa verbiprefikseist läänemere- Geert E. Booij, Christian Lehmann & Joachim Mugdan (eds), Morpho- soome keelis [On borrowed Indo-European verbal prefixes in the logie. Ein internationales Handbuch zur Flexion und Wortbildung. 1. Finnic languages]. Akadeemilise Emakeele Seltsi toimetised XXXII. Halbband. Handbücher zur Sprach- und Kommunikationswissenschaft Tartu: Akadeemilise Emakeele Seltsi kirjastus. 17.1. Berlin – New York: Mouton de Gruyter. 535–545. — 2006: Muistoja Liivinrannasta. Liivin kieltä Ruotsista [Recollections Haspelmath, Martin 2009: Lexical borrowing: Concepts and issues. – Mar­ about the Livonian Coast. Livonian from Sweden]. Translated and tin Haspelmath & Uri Tadmor (eds), Loanwords in the world’s lan- published by Anneli Honko. Mémoires de la Société Finno-Ougrienne guages. A comparative handbook. Berlin: Mouton de Gruyter. 35–54. 250. Helsinki: Finno-Ugrian Society.

48 49 SANTRA JANTUNEN SYNTACTIC AND ASPECTUAL FUNCTIONS OF LATVIAN VERBAL PREFIXES IN LIVONIAN

Metslang, Helle 2001: On the developments of the Estonian aspect. The ver­ Kotimaisten kielten tutkimuskeskuksen julkaisuja 62. Helsinki: Suoma­ bal particle ära. – Östen Dahl & Maria Koptjevskaja-Tamm (eds), The laisen Kirjallisuuden Seura & Kotimaisten kielten tutkimuskeskus. Circum-Baltic Languages. Typology and contact. Volume 2: Grammar Suhonen, Seppo 1972: Liivin prefiksiverbit [Livonian prefixed verbs]. – and Typology. Studies in Language Companion Series 55. Amsterdam Virittäjä 76: 217–20. Available at: ‹https://journal.fi/virittaja/article/ – Philadelphia: John Benjamins. 443–479. view/36120› MLLVG = Mūsdienu latviešu literārās valodas gramatika I. Fonētika un mor- — 1973: Die jungen lettischen Lehnwörter im Livischen. Mémoires de la foloģija [Grammar of Standard Latvian I. Phonetics and morphology]. Société Finno-Ougrienne 154. Helsinki: Finno-Ugrian Society. Atb. red. E. Sokols. 1959. Rīga: Latvijas PSR Zinātņu akadēmijas — 1974: Lähisukukielten muoto-oppia ja lainasuhteita [Morphology and izdevniecība. loan relationships of closely related languages]. Castrenianumin toi­ Puura, Ulriikka 2007: Vepsän kielen momentaaniset ja frekventatiiviset mitteita 9. Helsinki. verbijohdokset. Johdosten rakenteen, merkityksen ja käytön tarkaste- — 1975: Liivin kielen näytteitä [Samples of Livonian language]. Castreni­ lua [Momentary and verb derivatives in Veps. Study of anumin toimitteita 5. Helsinki: Finno-Ugrian Society & University of structure, meaning and use. Master’s thesis]. Helsinki: Helsingin yli­ Helsinki. opisto. Suomalais-ugrilainen laitos. Available at: ‹http://www.urn.fi/ — 1985: Ergänzungen zum Wortschatz im Livischen. – Studia Fennica URN:NBN:fi-fe200804181247› 28. Ostseefinnische Untersuchungen. 111–121. — 2010: Frequentative and momentative verbal derivation in Veps lan­ Tveite, Tor 2004: The case of the object in Livonian. A corpus based study. guage. – Linguistica Uralica 4 6:4: 2 61– 2 8 0. Av a i a b l e a t : ‹https://doi.org/ Castrenianumin toimitteita 62. Helsinki: The Finno-Ugrian Depart­ 10.3176/lu.2010.4.03› ment of Helsinki University & Finno-Ugrian Society. Rudzīte, Marta 1980: Bemerkungen über Präfixverben im Livischen. – Lin- VISK = Auli Hakulinen, Maria Vilkuna, Riitta Korhonen, Vesa Koivisto, gua Posnaniensis XXIII: 165–167. Tarja Riitta Heinonen & Irja Alho 2004: Iso suomen kielioppi [Com­ — 1994: Latviešu un lībiešu valodas savstarpējā ietekme [The mutual prehensive grammar of Finnish]. Helsinki: Suomalaisen Kirjallisuuden influence of Latvian and Livonian]. – . Boiko (ed.), Lībieši. Rīga: Seura. Verkkoversio. Available at: ‹http://scripta.kotus.fi/visk› Zinātne. 288–319. Wälchli, Bernard 2000: Livonian in a genetic, areal and typological per­ — 1996: Latviešu un lībiešu valodas kontaktu atspulgi [Reflections spective, or is Finnish better Finnic than Livonian? – Johanna Laakso of Latvian and Livonian language contacts]. – P. Vanags (ed.) Baltu (ed.), Facing Finnic. Some challenges to historical and contact linguis- filoloģija VI: Rakstu krājums. Rīga: Latvijas Universitāte. 3–7. tics. Castrenianumin toimitteita 59. Helsinki: Finno-Ugrian Society & Sivers, Fanny de 1971: Die lettischen Präfixe des livischen Verbs. Nancy: Department of Finno-Ugrian Studies of the University of Helsinki. Imprimerie Berger-Levrault. 210–226. Sjögren, J. A. 1861: Joh. Andreas Sjögren’s Livische Grammatik nebst — 2001: Lexical evidence for the parallel development of the Latvian and Sprachproben. Im Auftrage der Kaiserlichen Akademie der Wissen- Livonian verb particles. – Östen Dahl & Maria Koptevskaja-Tamm schaften bearbeitet und mit einer historisch-ethnographischen Einlei- (eds), The Circum-Baltic Languages. Typology and contact. Volume tung versehen von Ferdinand Joh. Wiedemann. Joh. Andreas Sjögren’s 2: Grammar and Typology. Studies in Language Companion Series 55. Gesammelte Schriften Band II. Theil I. St. Petersburg. Amsterdam – Philadelphia: John Benjamins. 413–441. Smith, Carlota S. 1997: The parameter of aspect. Second edition. Studies in Ziegeler, Debra 2006: Interfaces with English aspect: Diachronic and Linguistics and Philosophy, Volume 43. Dordrecht: Kluwer Academic empirical studies. Studies in Language Companion Series 82. Amster­ Publishers. dam – Philadelphia: John Benjamins. SSA = Itkonen, Erkki & Ulla-Maija Kulonen (eds) 1992–2000: Suomen sanojen alkuperä. Etymologinen sanakirja [Origin of Finnish words. Etymologi­ cal dictionary] 1–3. Suomalaisen Kirjallisuuden Seuran toimituksia 556,

50 51 SANTRA JANTUNEN SYNTACTIC AND ASPECTUAL FUNCTIONS OF LATVIAN VERBAL PREFIXES IN LIVONIAN

Latvialaisperäisten verbiprefiksien syntaktiset kehittymässä latvian vastineen no- käytöstä eroavaksi. Prefiksiverbi ja aspektuaaliset funktiot liivissä ei liivissä kuitenkaan aina ole perfektiivinen. Lisäksi prefiksittömät verbitkin voivat esiintyä perfektiivisissä yhteyksissä. Prefiksin käyttö Santra Jantunen ei siis ole pakollista perfektiivisyyttä ilmaistaessa. Voidaan siis todeta, ettei verbiprefiksaatio ole liivissä täysin ke­ Tutkimuksen kohteena olevaa Kuurinmaan liiviä luonnehtii pitkä­ hittynyt ilmaisemaan perfektiivisyyttä eikä prefiksien esiintyminen aikainen kontakti latvian kanssa sekä sen puhujien laaja kaksikieli­ ole systemaattista. Verbiprefiksit tarjoavat kuitenkin perfektiivisyy­ syys. Näiden seurauksena latvian vaikutus liiviin on ollut voimakasta den ilmaisemiseen uuden keinon alkuperäisten itämerensuomalaisten­ sekä sanaston että kieliopin osalta. Esimerkki tästä vaikutuksesta ovat ilmaisukeinojen,­ objektin sijanvaihtelun ja verbipartikkeleiden, yksitoista latviasta lainattua verbiprefiksiä:aiz -, ap-, at-, ie-, iz-, nuo-, rinnalle. pa-, pie-, pōr-, sa- ja uz-. Nämä prefiksit voivat esiintyä liivissä sekä omaperäisten että latviasta lainattujen verbien kanssa. Tutkimuksen aineisto on koottu aiemmin julkaisemattomista lii­ vin nauhoituksista sekä julkaistuista kielennäytteistä (Mägiste 2006). Eri prefiksien määrät vaihtelevat aineistossa (461 esimerkkilausetta)­ huomattavasti: yleisin prefiksi on nuo-, noin 29 % kaikista prefik­ sien esiintymistä, kun taas aineiston harvinaisimmista prefikseistä esiintymiä on vain muutama (pa- 6 kpl, pōr- 5 kpl). Aineiston pe­ rusteella prefiksin yleisyys näyttää olevan yhteydessä sen käyttöön perfektiivisyyden­ ilmaisemisessa: mitä yleisempi prefiksi on, sitä toden­näköisempää on, että sitä voidaan käyttää ilmaisemaan myös as­ pektuaalisia funktioita. Verbiprefiksit esiintyvät liivissä sekä leksikossa että osana morfo­logiaa. Funktionaalisia ominaisuuksia on lainattu leksikaalisen elementin, itse prefiksin, lisäksi. Aineiston perusteella voidaan sanoa, että eräitä latviasta lainattuja verbiprefiksejä voidaan käyttää muodos­ tamaan perfektiivisiä verbejä ja ilmaisemaan perfektiivisyyttä liivissä. Erityisesti prefiksiä nuo- käytetään tässä funktiossa. Myös prefiksejä iz-, sa-, uz-, ie- ja ap- (taajuusjärjestyksessä) voidaan käyttää perfek­ tiivisyyden ilmaisemisessa. Nämä prefiksit, lukuun ottamatta prefiksiä iz-, antavat lisäksi verbille myös adverbiaalisen lisämerkityksen: sa- ’yhdessä; paljon’, uz- ’päällä, päälle’, ie- ’sisällä, sisään’ ja ap- ’ym­ päri’. Muita prefiksejä (aiz-, at-, pa-, pie-, pōr-) ei tavallisesti käy­ tetä ilmaisemaan perfektiivisyyttä. Joissakin tapauksissa sekin tosin on mahdollista, esimerkiksi liikeverbien yhteydessä. Yleisesti verbi­ prefiksien käyttö liivissä näyttää vastaavan latvian vastaavien prefik­ sien käyttöä. Tästä on poikkeuksena prefiksin nuo- käyttö, joka on

52 53 HEINI KARJALAINEN University of Oulu

Borrowing morphology: The influence of Russian on the Veps system of indefinite pronouns

Abstract Veps is a Finnic minority language that has long been influenced by Russian, the prestige language in the speech area. The influence of Russian can be perceived in all subsystems of the Veps language, but hardly any research has been done on its impact on morphology. The current paper focuses on the influence of Russian on the Veps indefinite pronouns and their restructur­ ing. The contemporary Veps indefinite pronoun system is based on the use of different affixes and particles, i.e., indefiniteness markers, which are attached to interrogative stems. This article describes the various Veps indefiniteness markers, which have been acquired via morpheme transfer (MAT) and morphological pattern transfer (PAT) from Russian. The borrowing of indefinite­ ness markers is typical for languages under the very strong in­ fluence of another language. According to contemporary studies, the motivation for borrowing should primarily be attributed to so­ ciolinguistic factors and less to structural-typological similarities of the languages in question. In the Veps language community, such sociolinguistic factors are the minority status of the Veps language and the bilingualism of its speakers. Keywords Veps, Russian, morphology, indefiniteness markers, morpheme transfer, morphological pattern transfer

Multilingual­ Finnic. Language contact and change. 55 –87. Uralica Helsingiensia 14. Helsinki 2019. ‹https://doi.org/10.33341/uh.85033› HEINI KARJALAINEN BORROWING MORPHOLOGY: THE INFLUENCE OF RUSSIAN ON THE VEPS SYSTEM OF INDEFINITE PRONOUNS

1. Borrowing morphology as particles, which are attached to the interrogative stem and make up an a typological phenomenon indefinite pronoun, are called indefiniteness markers. It is commonly assumed that not all areas of morphology show This study explores the different ways Russian has influenced the Veps the same degree of propensity for borrowing. Derivation is perceived indefinite pronoun system from a typological viewpoint. The oldest as more borrowable than inflection and inherent inflection more bor­ of Finnic indefinite-interrogative pronouns was based on bare rowable than contextual inflection. (Gardani et al. 2015.) According to interrogatives, therefore ambiguity could not always be avoided. At contemporary typological studies and contrary to earlier claims, mor­ present, the Veps indefinite pronouns are usually based on interroga­ phological borrowing is surprisingly frequent in the languages of the tive pronouns with different indefiniteness markers, most of which are world. Overall, it is known to happen in languages under very strong borrowed from Russian. The Russian pronominal elements are attested influence from another language. (Haspelmath 2004 [1997]: 184–185.) in all the eastern Finnic languages as well as in some other Finno-Ug­ Traditionally, it has been argued that pronouns, pronoun para­ ric languages (van Alsenoy & van der Auwera 2015: 533; Alvre 1982: digms, and pronominal affixes are not susceptible to borrowing (Dixon 45, 2002: 161). The question of the degree of semantic and functional 1997: 22; Greenberg & Ruhlen 1992), but this view has lately been matching between the indefiniteness markers in the recipient language questioned (see Campbell 1997). Thomason & Everett (2001) have Veps and the source language Russian is discussed in more detail in found proof of numerous cases of borrowed pronouns in the languages Hienonen (2010). In the following, I argue that both morpheme trans­ of the world. As they state, pronoun borrowing is “nowhere near as fer and morphological pattern transfer influence the system. rare as one would suppose from reading the literature” (ibid. 301). Below, borrowing is used as “a cover-term for the adoption of Under certain social circumstances, pronouns or sets of pronouns are a structural feature into a language as a result of some level of bilin­ easily borrowed. Sometimes borrowing occurs when there is a per­ gualism in the history of the relevant speech community” following ceived gap in the pronominal paradigm (ibid. 304). Matras & Sakel (2007: 1). There is a main distinction between two As regards indefinite pronouns, there have not been any cases at­ types of borrowing which are labelled: morpheme transfer (MAT) and tested in the languages of the world in which a complete indefinite pro­ morphological pattern transfer (PAT) (for different definitions, see noun has been borrowed. Nevertheless, there is a considerable number Sakel 2007). MAT refers to the replication of linguistic matter consist­ of documented cases where languages, which have interrogative-based ing of actual phonological segments, whereas PAT refers to the replica­ indefinites, have adopted indefiniteness markers from another lan­ tion of functional or semantic patterns of the other language. In PAT, guage, via either MAT or PAT. This change takes place mainly under the form itself is not borrowed. (Matras & Sakel 2007; Sakel 2007.) heavy foreign interference. (Haspelmath 2004 [1997]: 184.) In the following I apply Haspelmath’s (2004 [1997]: 10–22) defi­ Some studies on language contact claim that structural and typo­ nitions of certain key concepts. Indefinite pronouns express indefi­ logical similarity between the source and the recipient language promote nite reference, the main functional characteristic of this subgroup of borrowability or is a precondition for it. Some more recent studies claim pronouns. Furthermore, there are four types of expressions, which in that the borrowability of structural categories is not dependent on the the Western grammatical tradition belong to the category of indefinite structural similarity of the languages but primarily on the sociolinguistic pronouns but which are excluded here. These are mid-scalar quanti­ factors, such as speakers’ deliberate choices, the intensity of the contact fiers, generic pronouns, universal quantifiers, and identity pronouns situation, and the level of bilingualism in the speech community (see, or determiners. The concept of a pronoun is understood in its broader e.g., Thomason & Everett 2001; Thomason 1999). sense, in which not only pro-nouns but also other pro-words such as Yaron Matras (2015) emphasises the usage-based model of bor­ pro-adjectives and pro-adverbs are covered by this term. Affixes and rowing and argues that the true locus of borrowing is the individual,

56 57 HEINI KARJALAINEN not the linguistic system. According to him, bilinguals do not organ­ ise their communication in the form of two “languages” or “linguistic systems”. Instead, they have an enriched and extended repertoire of linguistic structures among which they learn to select. (Ibid. 50.) Not every element exists in each language; therefore, some elements might be shared. In certain contexts, the “native” language of a speaker does not constitute a fully adequate means of communication. Therefore, bilingual speakers are motivated to draw on material from another lan­ guage. Using fusion or uniformity of form-function representations in both languages seems to be the most beneficial strategy in different interaction settings. As Matras points out, borrowing is not always de­ liberate or conscious, but it surely is purposeful and function-driven. (Ibid. 51−53.) Frank Seifart’s (2015; see also 2013) quantitative study ques­ tions the assumption that the borrowability of linguistic forms is con­ strained by structural properties. Seifart compares the extent of affix borrowing with structural similarity scores for the languages. The af­ fixes included both inflectional and derivational affixes. (Seifart 2015: 93−95.) The structural similarity scores were counted with the help of morphosyntactic features presented in the World Atlas of Lan­ guage Structures (Dryer & Haspelmath 2011). Seifart (ibid. 97−98) concludes that structural similarity between the donor and recipient language does not play a major role in determining the borrowability of affixes. Statistically, heavy affix borrowing is just as commonly attested between structurally dissimilar as well as structurally similar languages. Therefore, structural factors do not constrain the bilingual speakers’ creation of mixed varieties and borrowing. Typological distance between Veps and Russian is significant by several criteria. Since typological congruence does not play a major role in affix borrowing, a more plausible explanation for the motivation of borrowing is achieved by adopting a more user-oriented theory and by taking a glance at the sociolinguistic reality of the Veps speech commu­ nity. In the Veps speech community, bilingualism is unidirectional and Veps is used mainly in informal or private or domestic settings (Puura et al. 2013). According to Matras (2015: 52−53), this represents a typical situation in which the community is inclined to have a more lax attitude toward borrowing. Map 1 shows the Veps language area. Map 1. The Veps language area in the years 1900 and 2000 according to Grünthal (2011: 269). The language boundaries in the year 1900 based on the appendix of 58 Tunkelo (1946) with place names in Finnish. Map by Arttu Paarlahti. HEINI KARJALAINEN BORROWING MORPHOLOGY: THE INFLUENCE OF RUSSIAN ON THE VEPS SYSTEM OF INDEFINITE PRONOUNS

2. The data Päžarʹ and Pondal in the and in the village of Ladv in . The data of this article consist of 2 483 phrases with indefinite pro­ In addition to the described data, the study has taken advantage nouns. The older part of the data consist of published text samples of an elicitation test, which I carried out during fieldwork in 2014. (Kettunen 1920; Kettunen 1925; Kettunen & Siro 1935; Lönnrot 2002 For the test, I produced 55 sample phrases, each of which represents [1853]; Setälä & Kala 1951; Sovijärvi & Peltola 1982; Zaiceva & a different context and function where indefinite pronouns could be Mullonen 1969), which represent spoken Veps from approximately used. The grammar and the accuracy of the sample phrases were the late 19th century until the mid-20th century. These data encompass checked by Olga Žukova, the lector of Veps language at the Petro­ all main dialect areas of Veps. For the sake of clarity, the transcription zavodsk State University. I chose a total of seven informants for the of old dialect data has been simplified. test. They were from the villages of Päžarʹ and Pondal in the Vo­ Furthermore, the contemporary data represent the language situ­ logda oblast and from the village of Ladv in the Leningrad oblast. ation today and consist of both written and spoken language samples. I requested that they comment on the test phrases I read aloud for I have collected the contemporary written language samples from the them, as not all informants were literate in Veps. The informants’ Veps language corpus (http://vepsian.krc.karelia.ru/text/). The us­ main task was to fill in the empty gaps in the phrases with one inter­ ability of the corpus is unfortunately limited, since the lemmatisation rogative stem and with varying affixes and particles. In most cases, is occasionally imprecise. Certain original affixes may be missing, there was also a bare interrogative available. The informants were while in some instances the order of the morphemes has been rear­ allowed to choose multiple variants or forms and even to propose ranged, arguably for the sake of the new literary standards. Nonethe­ new ones, as in practice, the functions of pronoun series are often less, despite these incidental flaws, the corpus has been of great bene­ overlapping. The experiment proved to be fruitful but laborious, fit for this study. The corpus samples consist of texts mainly from the since for each informant, it took from one to two hours to go through New Testament (Uzʹ Zavet) and from the Veps newspaper Kodima. the sample phrases. As a result, the variation and multifunctionality It should be noted that the Uzʹ Zavet texts are translations and there­ of the pronoun series were clearly lower, as far as the informants fa­ fore differ from the rest of the language samples. As the number of miliar with Standard Veps were concerned. The speakers influenced contemporary written Veps language texts is not particularly vast, the by Standard Veps generally have a more normative attitude towards presence of the Uzʹ Zavet texts in this study can be justified. In this ar­ the language. On the other hand, speakers who do not use literary ticle, the translations of the Veps Uzʹ Zavet phrases are word-for-word Veps at all, might be unaccustomed to examining one’s language on translations and do not follow any official English Bible translations, an abstract level. because they do not include indefinite pronouns. The contemporary spoken language samples consist of record­ ings made by me during field trips to the Northern and Central Veps 3. MAT in the Veps indefinite pronoun system villages in 2009, 2011, and 2014. The first interviews were recorded in 2009 in the Central Veps dialect area in the region of Oyatʹ in Lenin­ The Veps indefinite pronoun system displays three clear cases of MAT grad Oblast. The recordings from 2011 were made in the city of Petro­ from Russian. There are three actual indefiniteness markers, which zavodsk and in the Northern Veps villages of Šoutarʹv and Kaleig as have been adapted from Russian: koje- ~ kojo-, ni-, and -ni ~ -nibudʹ. part of the ELDIA research project (https://www.eldia-project.org/). In Russian, these markers are not deeply embedded morphemes, thus The material from 2014 was recorded in the Central Veps villages of in this regard, borrowing such units is not difficult to carry out.

60 61 HEINI KARJALAINEN BORROWING MORPHOLOGY: THE INFLUENCE OF RUSSIAN ON THE VEPS SYSTEM OF INDEFINITE PRONOUNS

3.1. The marker koje- ~ kojo- As seen in the previous examples, the pronoun is usually accompanied by an enclitic particle -gi. Occasionally, the particle may be attached The Russian indefiniteness marker koe- is represented in Veps as to the prefix koje- instead of the pronominal stem (example 3). kojo- ~ koje- (see also Alvre 2002: 163; Blokland 2012: 4). In Russian, (3) VeC (Zaiceva & Mullonen 1969: 94) the koe-series is used when the referent is specific and the speaker Tulʹ kodi-he i ost-i koje-gi knows its identity (Haspelmath 2004 [1997]: 275). Furthermore, the come.impf.3sg home-ill and buy-impf.3sg indef-ptcl Russian koe-kto and koe-čto may also refer to a referent whose identity mi-dä vähä-iže-n. the speaker does not know (Bronnikov 2006: 5). Less commonly, the what-ptv little-dim-gen Russian pronoun koe-kak may be interpreted in a depreciative sense as ‘He/she came home and bought a little bit of something.’ ‘by whatever means, with great difficulty’ (Haspelmath 2004 [1997]: 190; Bronnikov 2006: 6). In a few cases, the pronoun occurs without the particle (example 4). In the Veps data, the marker koje- ~ kojo- is attested in both old and (4) VeC (Zaiceva & Mullonen 1969: 11) more recent data. The marker koje- ~ kojo- is attested in all main dialect jälgmäi koje-kut uspokoi-moi areas: in Northern, Central, and Southern Veps. The use in Southern afterwards indef-how calm.impf-1pl Veps is illustrated in example (1). However, koje- ~ kojo- is not attested ‘Afterwards we calmed down somehow.’ in Standard Veps, as no examples are found in the corpus data contain­ ing the New Testament (Uzʹ Zavet) and the newspaper Kodima. The meaning of the resulting indefinite pronoun is usually specific and (1) VeS (Kettunen 1920: 13) either known or unknown (as in example 5) to the speaker. Another nece-n tauve-n kaike-n sa-l-ii-n, sa-l-ii-n that-gen winter-gen all-gen get-freq-impf-1sg get-freq-impf-1sg typical meaning is a depreciative one, just as in Russian. kojo-mi-da-gi golʹu laps-ii-mu (5) Ve (Pronoun test 2014: Q4) indef-what-ptv-ptcl poor.pl.gen child-pl-prcom Minä kul-i-n koje-mi-dä, no e-n elʹgenda-nd, ‘For the whole winter I scraped, scraped up I hear-impf-1sg indef-what-ptv but neg-1sg understand-ptcp something with the poor children.’ kene-n änʹ nece ol-i. who-gen this be-impf.3sg According to the main data and excluding the pronoun test findings, ‘I heard something, but I didn’t understand, whose/what voice it was.’ the koje- ~ kojo-series is a marginal and not fully productive indefi­ nite pronoun series. In most cases, the prefixal markerkoje- ~ kojo- is However, on the basis of the pronoun test, there are stems other than combined with the stem kut ‘where’, but it may also combine with the kut ‘how’, mi ‘what’, mitte ‘which’, and miš ‘where’, which can be stems mitte ‘which’ (example 2), mi ‘what’, and miš ‘where’. combined with the prefixkoje- ~ kojo-. The informants approved of at­ taching koje- ~ kojo- to the stems konz ‘when’ (example 6), ken ‘who’ (2) VeS (Zaiceva & Mullonen 1969: 251) (example 7), and kus ‘where’ (example 8). In contradiction with the suks-i-l ajel-ii-ba poikpoliiž-i-d sa-m-ha, da da ski-pl-ade drive-impf-3pl willow-pl-ptv get-inf-ill and and main data, the koje- ~ kojo-series might not be as marginal and non- koje-mičč-i-d-gi. productive as first predicted (cf. Hienonen 2010). In this respect it indef-which-pl-ptv-ptcl resembles the Russian koe-series: in Russian the series has pronouns ‘They were skiing to get some willow representing all basic ontological categories except for amount, skolʹko (for a sleigh), and and something.’ ‘how much’ (see Haspelmath 2004 [1997]: 273). The difference is that

62 63 HEINI KARJALAINEN BORROWING MORPHOLOGY: THE INFLUENCE OF RUSSIAN ON THE VEPS SYSTEM OF INDEFINITE PRONOUNS the Veps koje- ~ kojo-series is used only in the spoken register, not in European languages, for example in Russian, Polish, Lithuanian, in the literary language. This must be due to the fact that in language Spanish, and non-standard English. (Ibid. 541.) planning the Russian marker has been dispreferred in favour of the The Veps indefiniteness markerni- is well-attested both in the old indefinite pronoun eraz, which is of Finnic origin (see also SSA s.v. (example 9) and the new data (example 10). eräs; Ojansuu 1922: 105). (9) Ve (Lönnrot 2002 [1853]: 24)1 (6) Ve (Pronoun test 2014: Q2) Ei ni-ken rodtʹe kirves käde-s. Koje-konz näg-i-n mei-den randa-l vedehiže-n, no neg.3sg indef-who born.cng axe hand-ine indef-when see-impf-1sg we-gen shore-ade water_sprite-gen but ‘No one is born with an axe in his hand.’ nece amu ol-i jo. that long_ago be-impf.3sg already (10) VeSt (Corpus: 489. Uzʹ Zavet 2006, Gospel of Luke 10: 42) ‘I once saw the Water sprite on our shore, Maria valič-i hüvä-n oza-n, si-dä häne-l but that was long ago already.’ Mary choose-impf.3sg good-gen part-gen, this-ptv she-ade ni-ken ei anasta. (7) Ve (Pronoun test 2014: Q6) indef-who neg.3sg steal.cng Koje-ken kolkota-b ikna-ha, no e-n näge, ‘Mary chose the good part, which no one will take away from her.’ indef-who knock-3sg window-ill but neg-1sg see.cng ken sigä om. who there be.3sg The marker ni- can be combined with any of the interrogative stems: ‘Somebody is knocking on the window, but I don’t see, who is there.’ ken ‘who’, mi ‘what’, mitte ‘which’, kus ‘where’, kuna ‘where’, miš ‘where’, konz ‘when’ (example 11), kut ‘how’, mit ‘how’, and kuverzʹ (8) Ve (Pronoun test 2014: Q7) ‘how much’. Minä jo kul-i-n nene sana-d koje-kus. I already hear-impf-1sg those word-pl indef-where (11) VeN (Sovijärvi & Peltola 1982: 10) ‘I already heard those words somewhere.’ Ei voi-nu män-da sluuž-ma-ha cerkva-ha viina-ta neg.3sg can-ptcp go-inf worship-inf-ill church-ill spirit-abe ni-konz. 3.2. The marker ni- indef-when ‘(He) could never go to the church to worship without spirits.’ The most firmly rooted indefiniteness marker in the Veps indefinite pronoun system is clearly the Russian prefixni- . The same marker ni- Much as in Russian, the Veps prefix ni- might precede not only a pro­ is attested in various other Uralic languages (see, e.g., van Alsenoy & noun but every negated coordinated element. In (12), the negative pre­ van der Auwera 2015; Alvre 2002: 163; Blokland 2012: 4). According fix ni- has been glossed as neg, except for the indefiniteness markers to van Alsenoy & van der Auwera (ibid.), the large number of bor­ before the pronoun. rowed negative markers in Uralic languages shows the heavy influ­ ence of Slavic. By using the negative indefiniteness marker together with the negative word, the Veps language, much as certain other Uralic languages, displays negative concord. Typological studies have shown negative concord to be an areal phenomenon found mostly 1. Lönnrot does not specify the dialects.

64 65 HEINI KARJALAINEN BORROWING MORPHOLOGY: THE INFLUENCE OF RUSSIAN ON THE VEPS SYSTEM OF INDEFINITE PRONOUNS

(12) VeC (Setälä & Kala 1951: 244) (14) Ve (Lönnrot 2002 [1853]: 7) Si-hesai mužik nece el-i, miše ni sö-da, ni Ei kudamo-s lidna-s ol-i carʹ nime-l Pirras. this-term man this live-impf.3sg that neg eat-inf neg indef which-ine city-ine be-impf.3sg tsar name-ade Pirras gʹo-da, ni pä-le, ni gʹaug-ha ni-kus ‘In one city there was a tsar called Pirras.’ drink-inf neg on-all neg foot-ill ind-where ni-mi-da ei-le. ind-what-ptv neg.3sg2-be.cng Examining the vast data, after excluding all modal existential and pos­ ‘Until then the man lived, so that (he had) nothing to eat, nothing to sessive constructions, there occurred very few negative phrases with­ drink, no clothes or shoes to wear, there isn’t anything anywhere.’ out negative concord (example 15). Such deviations are indeed very rare. As in Russian, Veps negated sentences always display negative con­ (15) VeN (Sovijärvi & Peltola 1982: 8) cord. In other words, the Veps indefinite pronouns are always marked Tö, sanu-b, e-t kuna kado-gii. with the indefiniteness marker ni- in negated sentences (cf. Savijärvi you.pl say-3sg neg-2pl where disappear-ptcl 1986: 58–59). The few exceptions to the rule are the modal existential ‘You, he says, won’t disappear anywhere.’ and possessive constructions, which do not contain a prefixni- in Rus­ sian, either (see Karjalainen 2016). In those constructions, the use of bare interrogatives is preferred, and ei- should be analysed instead as a These findings could also be supported by the pronoun test replies. As prefix and not as a negative verb, just like the prefix ne- in equivalent in the other data, the pronoun test informants were overwhelmingly constructions in Russian. This can be briefly illustrated by the modal of one mind: no one accepted any marker in negative sentences other existential construction (example 13). than ni- (example 16). (13) VeC (Setälä & Kala 1951: 278) (16) Ve (Pronoun test 2014: Q44) Uko-le teh-ta ei mi-da. Ei-le-nd leibä-d, ei-le-nd ni-mi-dä. man-all do-inf indef what-ptv neg-be-ptcp bread-ptv neg-be-ptcp indef-what-ptv ‘There is nothing the man can do.’ ‘There wasn’t bread, there wasn’t anything.’

The prefixal nature of ei- can also be observed in (14) (see also Again, only the modal existential and possessive constructions appear Karjalainen 2016). The resulting construction refers to a function as exceptions to the rule (example 17). Haspelmath (2004 [1997]: 4) calls specific known, rather than a nega­ (17) Ve (Pronoun test 2014: Q45) tive one as one would expect. Therefore, it is similar in its function to Perti-š om vilu, lämbita-da ei mi-l. one of the functions of the Russian prefixne- : v nekotorom gorode ‘in house-ine be.3sg cold heat_up-inf indef what-ade one city’. ‘It is cold in the house, there is nothing to heat up with.’

Therefore, as a conclusion, the use of negative concord in Veps ne­ 2. According to Kettunen (1943: 436), the negative verb and the verb olda ‘be’ are gated sentences should be considered just as compulsory as it is in merged in present and tense. Russian (cf. Savijärvi 1986: 58–59).

66 67 HEINI KARJALAINEN BORROWING MORPHOLOGY: THE INFLUENCE OF RUSSIAN ON THE VEPS SYSTEM OF INDEFINITE PRONOUNS

3.3. The marker -ni ~ -nibudʹ (19) VeC (Setälä & Kala 1951: 466) Sido min-dai ak kuna-nibutʹ lujo-mba, a As already discussed in Hienonen 2010, the Veps indefiniteness mark­ tie.imp.2sg I-ptv woman where-indef tight-adv.comp but er -ni ~ -nibudʹ can be related to the Russian suffix -nibudʹ (see also ei ka mina lenda-n. van Alsenoy & van der Auwera 2015). Alvre (2002), by contrast, pro­ neg.3sg and I fly-1sg pounds a view that suffixal-ni is the same affix as prefixalni- . Alvre’s ‘Tie me up, woman, tighter somewhere, if not, I will fly.’ argument is based on an idea that in strongly suffixing agglutinative languages a prefix might occasionally turn into a suffix. However, as Both -ni and -nibudʹ were attested in the pronoun test. The indefinite­ van Alsenoy & van der Auwera (ibid.) point out in their study on in­ ness marker -ni was approved by all seven informants, whereas only definite pronouns in Uralic languages, there are no documented cases five accepted the marker -nibudʹ (example 20). In Standard Veps, the of indefinites with prefixed ni-, which would have lost their negative normative form of this marker is -ni. However, it must be noted that meaning. With respect to the borrowed negative indefiniteness marker, both of the informants who did not approve any of the -nibudʹ-cases there is always an inherited negative concord pattern enclosed. Ac­ were frequent users of Standard Veps. Contrary to this, all five inform­ cordingly, in this study the affixesni- and -ni ~ -nibudʹ are considered ants accepting -nibudʹ lived in a village and clearly had less contact to be unique affixes. with the literary register. The Veps -ni ~ -nibudʹ pronoun series is used in several functions (20) Ve (Pronoun test 2014: Q17) defined by Haspelmath (2004 [1997]: 4): specific unknown, irrealis Konz-ni ~ konz-nibudʹ näge-moiš völ. non-specific, question and conditional protasis (see Hienonen 2010: when-indef see-refl.1pl again 286−287). The marker can be combined with any of the interrogative ‘We will see each other some time again.’ stems that are: ken ‘who’, mi ‘what’, mitte ‘which’, kus ‘where’, kuna ‘where’, miš ‘where’, konz ‘when’, kut ‘how’, mit ‘how’, and kuverzʹ ‘how much’ (example 18). As the example (21) from the Central Veps dialect reveals, -ni and -nibudʹ might occur in the same sentence. The informant uses the (18) VeS (Kettunen 1920: 114) marker -nibudʹ in the first clause, where the pronoun is used indepen­ Ka ii-k putuu-ž mei-le sigaa jauho-d well neg.3sg-q fall-cond we-all there flour-ptv dently. In the second clause, the pronoun is a determiner of a noun, kuverda-d-ni. and it seems that in such cases the informant finds the shorter marker how_much-ptv-indef -ni more convenient. ‘Well, wouldn’t some flour fall there for us.’ (21) VeC (Zaiceva & Mullonen 1969: 19) Dengo-i-d e-le ost-te-s, män-da kuna-nibudʹ In my data, the form -ni is, generally speaking, far more common money-pl-ptv neg-be.cng buy-inf-ine go-inf where-indef than -nibudʹ, but both forms do exist. The marker -nibudʹ occurs in the da mičče-he-ni bohata-ha mužika-ha -ks. Northern and Central Veps data, whereas in Southern Veps data there and which-ill-indef rich-ill man-ill night-tra were no matches. The use in Central Veps is illustrated in (19). ‘There is no money for buying [anything], one should go somewhere and to some rich man’s place for the night.’

68 69 HEINI KARJALAINEN BORROWING MORPHOLOGY: THE INFLUENCE OF RUSSIAN ON THE VEPS SYSTEM OF INDEFINITE PRONOUNS

As a rule, the marker -nibudʹ co-occurs in phrases with disjunctive 4. The rearrangement of morphemes conjunctions such as libo ‘or’ (example 22) and ili ‘or’ (example 23). In addition to -nibudʹ, bare interrogatives and indefinite pronouns with I proceed by analysing the rearrangement of morphemes encoding in­ other markers (-ni, -se, naku) are also permitted in such phrases (Kar­ definiteness in the Veps indefinite pronoun system. Usually, all the jalainen 2016). Veps affixal indefiniteness markers are extrafixes, either prefixes or suffixes. In light of the findings of recent typological studies, extrafix­ (22) VeN (Sovijärvi & Peltola 1982: 144) al indefiniteness markers, however, just like all other extrafixes, have mina si-li žäri-n mi-da-ni liha-d libo I you-all fry-1sg what-ptv-indef meat-ptv or a tendency to be rearranged (Haspelmath 2004 [1997]: 23). Indefi­ mi-da-nibudʹ siga. niteness markers tend towards having the narrowest possible syntactic what-ptv-indef there scope and therefore end up as close to the pronominal stem as pos­ ‘I will fry you some meat or something there.’ sible. That is to say, when an indefiniteness marker becomes an affix, it may give rise to a process in which suffixal indefiniteness markers (23) VeC (Setälä & Kala 1951: 341) switch places with suffixal case markers. Nece-n sunduga-n avai-tas aka-d lähiže-d, sizare-d The rearrangement of the morphemes applies to the Veps suf­ this-gen trunk-gen open-3pl woman-pl close-pl sister-pl fixal indefiniteness marker -ni, which in certain cases tends to move ili ken-nibudʹ nevesta-n pole-späi ak. or who-indef bride-gen side-ela woman closer to its pronominal stem. In the data, this tendency occurs only ‘This trunk is opened by close women, by the sisters, when the pronominal stems mitte ‘which’ and kuverzʹ ‘how much’ are or by some woman from the bride’s family.’ concerned. Since the singular has no ending in Veps, it is natural that in the nominative case no rearrangement of the mor­ As in any other context, only -ni is used in the older Southern Veps phemes can be observed. Therefore, only cases other than nominative data (example 24). are provided next. In the instances of rearranged morphemes, the affix -ni may ei­ (24) VeS (Kettunen 1920: 79) ther have a form -ni- or -nija-. In (25), mitte has the stem mičče-. The A laps-kuluuže-d mi-š-ni ikna-n aa But child-poor-pl what-ine-indef window-gen under.ade affix is represented as-nija- and followed by a ending. maga-tas, libo ke-he-ni män-hed raffaz-he (25) VeC (Zaiceva & Mullonen 1969: 127) sleep-pass or who-ill-indef go-ptcp people-ill magada-m-ha mičče-nija-n grivennika-n päiva-ks tači-b which-indef-gen dime-gen day-tra toss-3sg sleep-inf-ill ‘But poor children are sleeping somewhere outside, ‘Tosses a few dimes for the day.’ or have gone to some people’s (house) to sleep.’ In (26), the stem kuverzʹ is in the nominative. The affix is represented as -nija- and followed by a genitive case ending. (26) VeC (Zaiceva & Mullonen 1969: 127) kuverzʹ-nija-n surustoita-b da užina-u möst how_much-indef-gen feed-3sg and supper-ade again gö söta-b already feed-3sg ‘Feeds a little bit, and during supper feeds again.’

70 71 HEINI KARJALAINEN BORROWING MORPHOLOGY: THE INFLUENCE OF RUSSIAN ON THE VEPS SYSTEM OF INDEFINITE PRONOUNS

In (27), both the stem and the suffix have the ending. The survey of the data revealed some areal tendencies in the phenom­ enon of the rearranging of -ni- ~ -nija- and other morphemes. All re­ (27) VeN (Sovijärvi & Peltola 1982: 152) ported cases were attested in the Central and Northern Veps dialects. E-d-ik näh-nu (mi) mittuš-t-nija-d aka-d neg-2sg-q see-ptcp (what) which-ptv-indef-ptv woman-ptv In the old Southern Veps data, no such cases were found. proit-te-s pass-inf-ine ‘Didn’t you see (that) any woman passing?’ 5. PAT in the Veps indefinite pronoun system

Before the ending, both forms -ni- and -nija- occurred in As grammatical markers, indefiniteness markers are often not strongly the data. After -nija- the illative case ending -ha is used (example 28), grammaticalised and they can therefore be related to other elements whereas after -ni- the case ending is -he (example 29). This results in the grammar which enables calquing or, in other words, PAT. Thus from the fact that in the instances of rearranged morphemes, -nija- and far occurrences of PAT in indefiniteness markers have not yet been -ni- are understood as belonging to the stem. According to the phono­ sufficiently studied. There may be many occurrences of PAT in the logical rules of Veps, the illative case ending depends on the last vowel indefinite pronoun systems of the world’s languages, but usually it of the stem (Zaiceva 1995: 43). When the last vowel is -a-, the illative is difficult to show whether the resulting marker has its origin in lan­ case ending is -ha, and when the vowel is -i-, the ending is -he (ibid.). guage contact or is an independently developed formation. Therefore, the few attested cases are thought to represent only the tip of the ice­ (28) VeC (Zaiceva & Mullonen 1969: 183) berg. (See Haspelmath 2004 [1997]: 185). nu i nece-n sageda-n maido-iže-n amunda-n taga-ze well and this-gen thick-gen milk-dim-gen ladle-1sg back-ill The most clear case of PAT from Russian to the Veps indefinite mičče-nija-ha astka-iže-he i sö-m. pronoun system is the use of the indefiniteness marker -se. Further­ which-indef-ill pot-dim-ill and eat-1pl more, the use of naku in the Pondal dialect is presented here as an ‘Well, and I ladle this thick milk (= farm cheese) example of a locally invented indefiniteness marker. back into some pot and we eat (it).’ 5.1. The marker -se (29) VeC (Zaiceva & Mullonen 1969: 56) pida-b mičče-ni-he luht-ha ligota-da The pattern for adopting the Veps demonstrative pronoun se ‘this’ must-3sg which-indef-ill puddle-ill soak-inf as an indefiniteness marker comes from Russian (see also Hienonen ‘One must soak (it) in some puddle.’ 2010: 287). There seem to be striking similarities in the use of the In the following pronoun test sentence (30), completed by one of the Veps -se and the Russian particle -to. The particle -to may be added to informants, mitte again has the stem mičče-. It is followed by an infix nouns, pronouns, adjectives, pronominal adjectives, adverbs, quanti­ -ni-, the plural marker -i-, and the partitive case ending -d. fiers, , and finite verb forms (Leinonen 1998: 74–77). The prototypical meaning of the postpositive -to is demonstrative-empha­ (30) Ve (Pronoun test 2014: Q8) sising. It has been argued in the literature that it might also be used Kacu-hta, tat to-i lidna-späi as a definite article, but as Leinonen (ibid. 75) points out, this use is look-mom.imp.2sg dad bring-impf.3sg town-ela mičče-ni-i-d tavaro-i-d. not regular enough, and therefore the evidence does not validate this which-indef-pl-ptv ware-pl-ptv view. In Russian, -to serves also as an indefiniteness marker. Accord­ ‘Look, dad brought some wares from the town.’ ing to Haspelmath (2004 [1997]: 273), the origin of this indefiniteness

72 73 HEINI KARJALAINEN BORROWING MORPHOLOGY: THE INFLUENCE OF RUSSIAN ON THE VEPS SYSTEM OF INDEFINITE PRONOUNS marker is unclear, but he suggests a link between -to ‘now’ and the (32) VeS (Kettunen 1925: 48) Old Russian tŭ, ≠ to ‘that’. Van Alsenoy & van der Auwera (2015: 45) da sidbukark i basi-b: ”ken-se jo ku consider the link between and indefinites even more and dung_beetle and talk-3sg who-this already when straightforward. Evidence of demonstratives used as indefinites in tʹeni-n basi-b stukič?” specific contexts can be found, for instance, in tonight-essin talk-3sg push.impf.3sg (von Heusinger 2011). The grammaticalisation of deictic words into ‘And the dung beetle says: Who was the one, he says, that pushed tonight like that?’ indefiniteness markers can be seen both in Russian and Veps (see also naku in section 5.2.). According to the data, the indefinite use of pronouns with the In the oldest data, only very few pronouns with the suffix -se may marker -se is a very recent grammaticalisation in Veps. In the older be interpreted indefinitely. In the oldest Southern Veps samples (Ket­ samples, se ‘this’ may be interpreted mainly as a demonstrative or tunen 1920; Kettunen 1925) there were only three sentences where indicating emphasis. In attributive use, se usually follows the head of -se-pronouns had been used with an indefinite meaning. This is illus­ the phrase, though there are also a few cases where it may precede it trated by one of these sentences in (33). (Kettunen 1943: 397). When se follows the head, it is unstressed and, (33) VeS (Kettunen 1925: 63) by implication, it has no clear demonstrative function (ibid.). The ex­ priha-kuluune joksʹ, joksʹ kuna-se meca-le ample in (31) is fairly prototypical. boy-poor run.impf.3sg run.impf.3sg where-indef forest3-ade (31) VeS (Kettunen 1920: 88) ‘The poor boy ran, ran to some bloody place.’ A prihad se ii-le, ni-ke-da ii-le. but boy-ptv this neg.3sg-be.cng indef-who-ptv neg.3sg-be.cng ‘But the boy isn’t (there), there is nobody.’ Similarly to the Southern Veps data, there are very few cases of the indefinite use of -se-pronouns in the other old dialect samples. The few instances of such use were found in Kettunen & Siro (1935; 2 oc­ Lauri Kettunen has proposed a view that in suffixal cases, -se may currences) and Setälä & Kala (1951; 12 occurrences), and only in the even act as a definite article. However, similarly to Russian, this ar­ dialect area of Central Veps. As can be seen from these occurrences, ticle use is not regular enough in Veps to support this proposition. the use of the suffix -se was not yet productive: it was used only with Veps -se should instead be seen as a pronoun, which expresses the the pronominal stem kuverzʹ ‘how much’. This is presented in (34) and prominence of the referent in the discourse. This kind of use is similar (35). Again, as seen before, the stem kuverzʹ may be in the genitive to the Finnish demonstrative se in the 20th century (see, e.g., Leinonen (example 34) as well as in the nominative case (example 35). 1998: 75). It has been suggested that the use of the suffixal pronoun (34) VeC (Kettunen & Siro 1935: 46) se in old literary Finnish would be at least partly based on the model Kuverda-n-se si-da aiga-d kodi-š ol-i of Swedish (see Kiuru 1990: 289). However, as the same suffixal pro­ how_much-gen-indef this-ptv time-ptv home-ine be-impf.3sg noun can be found in the eastern Finnic languages, additional studies nece uk. are needed in order to develop a full picture of the suffixal-se . this man In the older Veps dialect samples, when -se follows a pronoun, ‘The man was at home for some time.’ the meaning of the sentences is usually either interrogative (exam­ ple 32) or relative (example 33) (see also Kettunen 1943: 402). 3. In Veps, mända mecale ‘go to a forest’ is an idiom meaning approximately ‘go to some bloody place’ (see, e.g., Zaiceva & Mullonen 1972 s.v. mec).

74 75 HEINI KARJALAINEN BORROWING MORPHOLOGY: THE INFLUENCE OF RUSSIAN ON THE VEPS SYSTEM OF INDEFINITE PRONOUNS

(35) VeC (Setälä & Kala 1951: 222) In the present-day spoken language, the use of -se-pronouns is well-at­ Kuverz-se aiga-d proidu-i tested in all dialects including the northern ones. The example phrases how_much-indef time-ptv pass-impf.3sg (39) and (40) are from two speakers of Northern Veps. ‘Some time passed.’ (39) VeN (Interview 2011, female) vot kus-se anketa-s ol-i mise, nu However, in the data by Zaiceva & Mullonen (1969) collected just a Well where-indef questionnaire-ine be-impf.3sg that well few decades later, the indefinite use of the -se-pronouns is far more ken-se ristit sanu-i numerous (32 occurrences). As in the historically older data, it only who-indef person say-impf.3sg occurs in Central and Southern Veps. The use of the marker -se has be­ ‘Well, it was somewhere in the questionnaire, come more productive, as this data show that it can be combined with well, that some person said.’ any of the interrogative stems: ken ‘who’, mi ‘what’, mitte ‘which’ (40) VeN (Interview 2011, male) (example 36), kus ‘where’, kuna ‘where’, miš ‘where’, konz ‘when’, kut ‘how’, and kuverzʹ ‘how much’. siiga-d-se ei-lä nügü-, kuna-se läht-nu whitefish-ptv-this neg.3sg-be.cng nowadays where-indef go-ptcp (36) VeS (Zaiceva & Mullonen 1969: 226) ‘There aren’t any whitefish4 nowadays, (they have) gone somewhere.’ mič-ii-l-se jüv-i-lʹ da hʹeenh-ii-lʹ mii-dʹ dʹo what-pl-ade-indef corn-pl-ade and feather-pl-ade we-ptv already hloputa-ba, e-n teda mi-lʹ. 5.2. The marker naku scatter-3pl neg-1sg know.cng what-ade ‘They are scattering us with some corn and The second case of a PAT borrowing from Russian to Veps is present­ feathers, I don’t know with which.’ ed here as an example of local borrowings and fast linguistic changes, which are especially typical for small segregated language communi­ When it comes to modern literary Veps, the use of the marker -se is ties. I discovered this areal peculiarity in Pondal, a village which is very common, and the present-day -se-series has a wide distribution of located in the eastern part of the Central Veps dialect area, and which I functions (see Hienonen 2010). visited during my fieldwork in 2014. In Pondal, a demonstrative pro­ noun naku ‘here’ is used together with different interrogative stems. (37) VeSt (Corpus: 455. Uzʹ Zavet 2006, Gospel of Luke 6: 22) In certain cases, it seems, that naku may be on a pathway of gram­ huigenzoit-tas tei-den nimi-ki kuti mitte-se paha. reject-pass you.pl-gen name-ptcl like what-indef evil maticalisation from an emphatic particle to an indefiniteness marker. ‘And your name will be rejected like something evil.’ The naku + pronoun construction closely resembles the collo­ quial Russian construction vot ‘here’. In some Veps dialects other than (38) VeSt (Corpus: 587. Uzʹ Zavet 2006, Gospel of John 7: 50) Pondal, this construction is represented as vot + pronoun. Another Siloi Nikodim, kudamb konz-se ö-l ol-i equivalent in colloquial Russian, the construction vona ‘look’ + pro­ then Nicodemus who when-indef night-ade be-impf.3sg noun, has been presented by Matti Larjavaara (1986: 180). kävu-nu Iisusa-nno da iče-ki ol-i farisei, The Russian particle vot may normally function as an ad­ go-ptcp Jesus-aprx and self-ptcl be-impf.3sg Pharisee verb ‘here, there’, as a demonstrative, or an emphasising particle sanu-i he-i-le say-impf.3sg they-pl-all (Kuosmanen 1999: 105; Kuosmanen & Multisilta 1999: 52–53;

‘Then Nicodemus, who had gone to Jesus at night 4. The European whitefish Coregonus( lavaretus) is a species of freshwater white­ and who himself was a Pharisee, said to them.’ fish, which is widespread from central and northwest Europe to Siberia. 76 77 HEINI KARJALAINEN BORROWING MORPHOLOGY: THE INFLUENCE OF RUSSIAN ON THE VEPS SYSTEM OF INDEFINITE PRONOUNS

Padučeva 1996: 161). In contemporary spoken Russian, vot mainly 6. Conclusions can be connected with the same functions as the Northern Russian -to (Leinonen 1998: 83). The construction vot + pronoun is usually used This study describes the various Veps indefiniteness markers acquired referentially, cataphorically (see, e.g., Hauenschild 1982: 174–177). from Russian via MAT (morpheme transfer) and PAT (morphological Unfortunately, due to space and time limits, we cannot discuss the pattern transfer). The underlying assumption of this study is adopted potential indefinite function of the colloquial Russian vot + pronoun from Frank Seifart (2015), who believes that the motivation for bor­ construction in this study. This might be an important issue for future rowing should primarily be traced to sociolinguistic factors and not as research. Moreover, a further study with more focus on the functions much to structural-typological similarities of the languages in ques­ and affinity betweenvot and -to is needed. tion. The oldest stratum of Finnic indefinite-interrogative pronouns The example phrases presented below are produced by two of my was based on bare interrogatives, so it is obvious that ambiguity could seven informants who attended my pronoun test in 2014. Both of these not always be avoided. Therefore, Veps speakers were motivated to informants originated from Pondal village. In Pondal, the construc­ transfer morphemes and morphological patterns into their pronoun tion naku + pronoun is used in various different contexts. The gener­ system and seek uniformity between the systems. However, the Veps ated meaning of the construction seemed to be sometimes demonstra­ system is not a full copy of the Russian system. Even though the sys­ tive, sometimes indefinite. The sentences (41) and (42) are examples tem might seem fusional, the distribution of the functions varies (see of contexts where an indefinite reading might be more reasonable than also Hienonen 2010). a demonstrative one. The evidence in this study is based on data containing 2 483 ex­ ample phrases, collected from various sources: old and new samples (41) Ve (Pronoun test 2014: Q6) representing both written and spoken registers. Furthermore, some Naku ken kolkota-b ikna-ha, no e-n näge, indef who knock-3sg window-ill but neg-1sg see.cng evidence has been drawn from an elicitation test. The test showed that ken sigä om. the variation and multifunctionality of the pronoun series is clearly who there is.3sg lower, as far as informants familiar with Standard Veps are concerned. ‘Someone is knocking on the window, The speakers influenced by Standard Veps tend to have a more norma­ but I do not see, who is there.’ tive attitude towards the Veps language. Both MAT and PAT from Russian have influenced the Veps sys­ (42) Ve (Pronoun test 2014: Q14) tem of indefinite pronouns. The system displays three clear cases of Ak kul-išt-i, jälg-i-l naku mi kävele-b. MAT from Russian, the indefiniteness markers koje- ~ kojo-, ni-, and woman hear-caus-impf.3sg track-pl-ade indef what walk-3sg -ni ~ -nibudʹ. The marker koje- ~ kojo- occurs in all main dialect areas ‘The woman heard: something is walking on the tracks.’ of Veps and in both older and more recent data. It is missing only in newer literary Veps. However, koje- ~ kojo- is not a fully productive Unfortunately, this topic is too broad in scope to be discussed as part indefinite pronoun series, as it is mainly combined with only a few of this study. There are still many unanswered questions regarding the stems. The most firmly rooted indefiniteness marker in the Veps sys­ functions of the construction naku ~ vot + pronoun. Further research tem is the Russian prefix ni-. The marker ni- occurs in both old and should be undertaken to cast light on this issue. new data and is fully productive. This study shows that, as in Russian, in negated sentences the Veps indefinite pronouns are always marked with the marker ni-. The only exceptions to the rule are the modal existential and possessive constructions, in which ei- should instead

78 79 HEINI KARJALAINEN BORROWING MORPHOLOGY: THE INFLUENCE OF RUSSIAN ON THE VEPS SYSTEM OF INDEFINITE PRONOUNS be interpreted as a prefix, not a negative verb. Therefore, the use of Abbreviations negative concord in Veps negated sentences should be considered just as compulsory as it is in Russian. 1 first person ine inessive The marker -ni ~ -nibudʹ is very common and fully produc­ 2 second person mom momentative tive. The form -ni is more common than -nibudʹ, which only occurs 3 third person neg negation in Northern and Central Veps. Furthermore, this article discusses the abe abessive pass passive ade adessive pl plural rearranging of morphemes as applied to the Veps suffixal indefinite­ adv adverb poss possessive ness marker -ni. Typologically extrafixal indefiniteness markers tend all allative prcom prolative-comitative to have the narrowest possible syntactic scope and thus typically oc­ aprx approximative ptcl particle cur as close to the pronominal stem as possible (see Haspelmath 2004 caus causative ptcp participle [1997]: 23). In Veps, the suffixal-ni may in certain cases switch places cng connegative ptv partitive with suffixal case markers. comp comparative q question The indefiniteness markers, like any other grammatical markers, cond conditional sg singular are usually not strongly grammaticalised, which enables morphologi­ dim diminutive term terminative cal pattern transfer. The most clear case of PAT from Russian to the ela elative tra translative Veps indefinite pronoun system is the use of the marker-se . The gram­ essin essive-instructive Ve Veps maticalisation of the Veps -se-pronoun is a fairly recent innovation, gen genitive VeC Central Veps since in the older data the indefinite use is almost non-existent and the ill illative VeN Northern Veps imp imperative VeS Southern Veps marker is not yet fully productive. This article also discusses an ex­ impf imperfect tense VeSt Standard Veps ample of a locally invented indefiniteness marker naku in the Central indef indefiniteness marker Veps dialect spoken in Pondal village. Veps is a minority language, which has been influenced by Rus­ sian for centuries. The strong impact of Russian can be seen in Veps phonology, morphology, syntax, and lexicon, but the influence on References morphology has not been sufficiently studied thus far. Further, this study has advanced the documentation and digitisation of the endan­ Alsenoy, Lauren van & Johan van der Auwera 2015: Indefinite pronouns gered Veps language. The description of basic Veps grammar found in Uralic languages. – Matti Miestamo, Anne Tamm & Beáta Wag­ in these materials is useful for both language developers of Veps and ner-Nagy (eds), Negation in Uralic languages. Typological Studies on linguists doing typological or comparative research. Language 108. Amsterdam: John Benjamins Publishing Company. 519−546. Alvre, Paul 1982: Läänemeresoome indefiniitpronoomeneist [On indefinite pronouns of Finnic languages]. – Antti Iivonen, Seppo Suhonen & Pertti Virtaranta (eds), Voces amicorum Sovijärvi: in honorem Antti Sovijärvi. SUST 181. Helsinki: Suomalais-Ugrilainen Seura. 45−55. — 2002: Russische Lehnelemente in Indefinitpronomen und Adverbien der Ostseefinnischen Sprachen [Russian loan elements in indefinite pronouns and adverbs in Finnic languages]. – Linguistica Uralica 48: 161–164.

80 81 HEINI KARJALAINEN BORROWING MORPHOLOGY: THE INFLUENCE OF RUSSIAN ON THE VEPS SYSTEM OF INDEFINITE PRONOUNS

Blokland, Rogier 2012: Borrowability of pronouns. Evidence from Uralic. – Hienonen (Karjalainen), Heini 2010: The Implicational Semantic Map for Cornelius Hasselblatt & Beata Wagner-Nagy (eds), Finnisch-Ugrische Veps Indefinite Pronouns. – Linguistica Uralica 46: 281–292. Mitteilungen 35. Hamburg: Helmut Buske Verlag. 1−34. Karjalainen, Heini 2016: Vepsän kielen paljaiden interrogatiivien indefiniit- Bronnikov, George 2006: Meaning postulates and logical form. The case tinen käyttö [Indefinite use of Veps bare interrogatives]. – Virittäjä 120: of Russian indefinite pronouns. Unpublished manuscript. Available at: 360–397. Available at: ‹https://journal.fi/virittaja/article/view/52695› ‹http://newstar.rinet.ru/~goga/work/pronouns-and-meaning-postulates.pdf› Kettunen, Lauri 1920: Näytteitä etelävepsästä 1 [Samples of Southern Campbell, Lyle 1997: Amerindian Personal Pronouns. A Second Opinion. – Veps 1]. Helsinki: Suomalaisen Kirjallisuuden Seura. Language 73: 339–351. — 1925: Näytteitä etelävepsästä 2 [Samples of Southern Veps 2]. Hel­ Corpus = Korpus vepsskogo jazyka [Veps language corpus]. Available at: sinki: Suomalaisen Kirjallisuuden Seura. ‹http://vepsian.krc.karelia.ru/text/› — 1943: Vepsän murteiden lauseopillinen tutkimus [Syntactic study of Dixon, M. . 1997: The Rise and Fall of Languages. Cambridge: Veps dialects]. SUST 86. Helsinki: Suomalais-Ugrilainen Seura. Cambridge University Press. Kettunen, Lauri & Paavo Siro 1935: Näytteitä vepsän murteista [Samples of Dryer, Matthew S. & Martin Haspelmath (eds) 2013: The World Atlas of Veps dialects]. SUST 70. Helsinki: Suomalais-Ugrilainen Seura. Language Structures Online. Leipzig: Max Planck Institute for Evolu­ Kiuru, Silja 1990: Mistä vanhaan kirjasuomeen tulivat jälkipronominit se ja tionary Anthropology. Available at: ‹http://wals.info› hän? [Where did suffixal pronouns se and hän in old literary Finnish ELDIA – European Language Diversity for All. Available at: ‹https://www. come from?]. – Virittäjä 94: 278–305. Available at: ‹https://journal.fi/ eldia-project.org/› virittaja/article/view/38375› Gardani, Francesco, Peter Arkadiev & Nino Amiridze 2015: Borrowed mor­ Kuosmanen, Anne 1999: On the relationship between the melodical struc­ phology. An overview. – Francesco Gardani, Peter Arkadiev & Nino ture and discourse functions of the particles NU and VOT in sponta­ Amiridze (eds), Borrowed morphology. Language Contact and Bilin­ neous Russian. In DIAPRO-1999. 105–110. Available at: ‹http://www. gualism 8. Berlin: Mouton de Gruyter. 1–23. isca-speech.org/archive_open/archive_papers/dia_pros/diap_105.pdf› Greenberg, Joseph H. & Merritt Ruhlen 1992: Linguistic Origins of Native Kuosmanen, Anne & Teija Multisilta 1999: Nu and vot in Spoken Rus­ Americans. – Scientific American 267: 94–99. sian. On discourse functions and prosodic features. – Scando-Slavica Grünthal, Riho 2011: Population decline and the erosion of the Veps lan­ 45: 1. Copenhagen: Munksgaard. 49–64. Available at: ‹http://dx.doi. guage community. – Riho Grünthal & Magdolna Kovács (eds), Eth- org/10.1080/00806769908601135› nic and linguistic context of identity. Finno-Ugric minorities. Uralica Larjavaara, Matti 1986: Itämerensuomen demonstratiivit I. Karjala, aunus, Helsingiensia 5. Helsinki: University of Helsinki, Department of Finn­ lyydi ja vepsä [Demonstratives of Finnic languages I. Karelian, Olo­ ish, Finno-Ugrian and Scandinavian studies & Finno-Ugrian Society. nets Karelian, Ludic, and Veps]. Suomalaisen Kirjallisuuden Seuran 267–293. Toimituksia 433. Helsinki: Suomalaisen Kirjallisuuden Seura. Haspelmath, Martin 2004 [1997]: Indefinite Pronouns. Oxford Studies in Leinonen, Marja 1998: The postpositive particle -to of Northern Russian Typology and Linguistic Theory. Oxford: Oxford University Press. dialects, compared with (Komi Zyryan). – Jyrki Hauenschild, Christa 1982: Demonstrative pronouns in Russian and Czech. Papinniemi, Jouko Lindstedt & Pekka Pesonen (eds), Studia slavica Deixis and anaphora. – Jürgen Weissenborn & Wolfgang Klein (eds), finlandensia.Studia slavica finlandensia in Congressu XII slavistarum Here and There. Cross-linguistic Studies on Deixis and Demonstra- internationali Cracoviae anno MCMXCVIII oblata. Helsinki: Venäjän tion. Pragmatics & Beyond, III: 2–3. Amsterdam: John Benjamins ja Itä-Euroopan instituutti. 74–90. Publishing Company. 167–186. Lönnrot, Elias 2002 [1853]: Om det nord-tschudiska språket. Elias Lönnro- Heusinger, Klaus von 2011: Specificity, referentiality and discourse prom­ tin väitöskirja vepsän kielestä vuodelta 1853 [On Veps language. Doc­ inence. German indefinite demonstratives. – Ingo Reich et al. (eds), toral dissertation by Elias Lönnrot on Veps language from the year Proceedings of Sinn & Bedeutung 15. Saarbrücken: Saarland Univer­ 1853]. Juminkeon julkaisuja 24. Kuhmo: Juminkeko. sity Press. 9–30.

82 83 HEINI KARJALAINEN BORROWING MORPHOLOGY: THE INFLUENCE OF RUSSIAN ON THE VEPS SYSTEM OF INDEFINITE PRONOUNS

Matras, Yaron 2015: Why is the borrowing of inflectional morphology dis­ SSA = Itkonen, Erkki & Ulla-Maija Kulonen (eds) 1992−2000: Suomen sa- preferred? – Francesco Gardani, Peter Arkadiev & Nino Amiridze nojen alkuperä. Etymologinen sanakirja [Origin of Finnish words. (eds), Borrowed morphology. Berlin: Mouton de Gruyter. 47–80. Etymological dictionary]. Suomalaisen Kirjallisuuden Seuran Toimi­ Matras, Yaron & Jeanette Sakel 2007: Introduction. – Yaron Matras & tuksia 556, Kotimaisten kielten tutkimuskeskuksen julkaisuja 62. Hel­ Jeanette Sakel (eds), Grammatical borrowing in cross-linguistic per- sinki: Suomalaisen Kirjallisuuden Seura & Kotimaisten kielten tutki­ spective. Berlin: Mouton de Gruyter. 1–13. muskeskus. Ojansuu, Heikki 1922: Itämerensuomalaisten kielten pronominioppia [Study Thomason, Sarah 1999: Speakers’ Choices in Language Change. – Studies of Finnic pronouns]. Turun suomalaisen yliopiston julkaisuja B I:3. in the Linguistic Sciences 29: 19–43. Helsinki: Suomalaisen Kirjallisuuden Seuran Kirjapainon Osake­yhtiö. Thomason, Sarah G. & Daniel L. Everett 2001: Pronoun Borrowing. – Pro- Padučeva, E. V. 1996: Neopredelennostʹ kak semantičeskaja dominanta ceedings of the Annual Meeting of the Berkeley Linguistics Society russkoj jazykovoj kartiny mira [Indefiniteness as semantic dominant 27: 1. Berkeley Linguistics Society. 301–315. Available at: ‹http:// of Russian linguistic picture of world]. – . Benacchio, . Fici & L. journals.linguisticsociety.org/proceedings/index.php/BLS/article/ Gebert (eds), Determinatezza e indeterminatezza nelle lingue slave. view/1107› Florence (): Unipress. 161–186. Tunkelo, E. A. 1946: Vepsän kielen äännehistoria [Phonological history of Puura, Ulriikka, Heini Karjalainen, Nina Zajceva & Riho Grünthal 2013: Veps]. Helsinki: Suomalaisen Kirjallisuuden Seura. The Veps language in Russia. ELDIA Case-Specific Report. – Studies Zaiceva, M[aria] I. & M[aria] I. Mullonen 1969: Obrazcy vepsskoj reči [Sam­ in European Language Diversity 25. Johannes Gutenberg-Universität ples of Veps dialects]. Leningrad: Nauka. Mainz, Research consortium ELDIA. Available at: ‹https://phaidra. — 1972: Slovarʹ vepsskogo jazyka [Dictionary of the Veps language]. univie.ac.at/detail_object/o:315545› Leningrad: Nauka. Sakel, Jeanette 2007: Types of loan. Matter and pattern. – Yaron Matras & Zaiceva, Nina 1995: Vepsän kelen grammatik. I Nimiden kändluz [Gram­ Jeanette Sakel (eds), Grammatical borrowing in cross-linguistic per- mar of the Veps language. I Inflection of nouns]. Petroskoi: Petroskoin spective. Berlin: Mouton de Gruyter. 15–29. valtion­yliopiston kustantamo. Savijärvi, Ilkka 1986: Venäjän partikkeli ni karjalassa ja vepsässä [The — 2000: Vepsän kelen Grammatik. II Verboiden kändluz. Verboiden Russian particle ni in Karelian and Veps]. – Ritva-Leena Heikkinen, nimiformad. Kändamatomad sanad [Grammar of the Veps language. Muusa Ojanen, Ilkka Savijärvi & Lea Siilin (eds), Rusistica 2. Tut- II Inflection of verbs. Verb categories. Invariable parts of speech]. kielmia venäjän kielen ja itämerensuomalaisten kielten kontakteista. Openduzabukirj. Petroskoi: Petroskoin valtionyliopiston kustantamo. Joensuu: Joensuun yliopisto. 52–80. Seifart, Frank 2013: AfBo: A world-wide survey of affix borrowing. Leip­ zig: Max Planck Institute for Evolutionary Anthropology. Available at: ‹http://af bo.info› — 2015: Does structural-typological similarity affect borrowability? A quantitative study on affix borrowing. – Language Dynamics and Change 5: 92–113. Setälä, E. N. & J. H. Kala 1951: Näytteitä äänis- ja keskivepsän murteista [Samples of Northern and Central Veps]. Published by E. A. Tunkelo & R. Peltola. SUST 100. Helsinki: Suomalais-Ugrilainen Seura. Sovijärvi, Antti & Reino Peltola 1982: Äänisvepsän näytteitä [Samples of Northern Veps]. SUST 171. Helsinki: Suomalais-Ugrilainen Seura.

84 85 HEINI KARJALAINEN BORROWING MORPHOLOGY: THE INFLUENCE OF RUSSIAN ON THE VEPS SYSTEM OF INDEFINITE PRONOUNS

Morfologiaa lainaamassa: venäjän kielen tunnus naku. Vepsän kielessä pronominin -se kieliopillistuminen in­ vaikutus vepsän indefiniittipronominien definiittisyyden tunnukseksi on tapahtunut melko hiljattain, sillä van­ järjestelmään hemmissa lähteissä käyttöä ei juurikaan esiinny eikä tunnus ole vielä produktiivinen. Tunnus naku on hyvin paikallisesti käytetty indefiniit­ Heini Karjalainen tisyyden tunnus. Indefiniittisyyden tunnusten lainaaminen on tyypillisintä kielil­ Vepsä on itämerensuomalainen vähemmistökieli, johon sen prestiisi­ le, jotka ovat joutuneet toisen kielen hyvin voimakkaan vaikutuksen kieli venäjä on vaikuttanut jo satojen vuosien ajan. Venäjän vaikutusta kohteiksi. Uusimpien tutkimusten mukaan lainaamisen motivaationa voidaan havaita kaikilla vepsän kielen tasoilla, mutta vaikutusta mor­ toimivat ennemmin erilaiset sosiolingvistiset tekijät kuin lähde- ja fologiaan ei ole tutkimuksessa tähän mennessä juurikaan käsitelty. Ar­ kohdekielen strukturaalis-typologinen samankaltaisuus. Vepsäläisessä tikkelin aiheena on venäjän kielen vaikutus vepsän indefiniittiprono­ kieliyhteisössä tällaisia tekijöitä ovat erityisesti kielen vähemmistö­ minien järjestelmään. Vepsän kielen tunnuksiset indefiniittipronominit asema ja sen puhujien kaksikielisyys. muodostuvat tyypillisesti interrogatiivivartalosta ja indefiniittisyyden tunnuksesta. Vartaloon liittyvä tunnus voi olla joko affiksi tai partikke­ li. Artikkelissa kuvataan erilaisia vepsän indefiniittisyyden tunnuksia, jotka on saatu kieleen niin sanottuina MAT- (morpheme transfer) ja PAT-lainoina (morphological pattern transfer). MAT tarkoittaa kielel­ lisen aineksen toisintamista toisessa kielessä, kun on kyse todellisista fonologisista segmenteistä, PAT toisen kielen funktionaalisten tai se­ manttisten mallien toisintamista, jolloin muoto itsessään ei lainaudu. Vepsän indefiniittipronominijärjestelmässä MAT-lainoja edusta­ vat koje- ~ kojo-, ni- sekä -ni ~ -nibudʹ. Osin epäproduktiivinen koje- ~ kojo- esiintyy kaikilla murrealueilla sekä vanhemmassa että uudem­ massa aineistossa standardoitua vepsää lukuun ottamatta. Produktii­ vinen ni- on vankimmin vepsän indefiniittipronominien järjestelmään juurtunut venäläislaina. Se esiintyy vanhoissa ja uusissa näytteissä. Tutkimuksessa osoitetaan, että ni- esiintyy vepsän kieltolauseissa ve­ näjän tapaan aina fakultatiivisena kieltoaineksena (vrt. Savijärvi 1986: 58–59). Ainoan poikkeuksen muodostavat modaaliset eksistentiaa­ li- ja possessiivirakenteet, joissa ei-aines on tulkittava pikemminkin prefiksiksi kuin kieltoverbiksi. Tunnus -ni ~ -nibudʹ on hyvin yleinen ja produktiivinen. Tunnuksen -ni yhteydessä esiintyy morfeemien uu­ delleenjärjestäytymistä, jossa sijapäätteet vaihtavat paikkaa suffiksaa­ lisen indefiniittisyyden tunnuksen kanssa. Indefiniittisyyden tunnukset, kuten muutkaan kieliopilliset tun­ nukset, eivät ole yleensä vahvasti kieliopillistuneita, jolloin PAT- lainaaminen on mahdollista. PAT-lainoina esiintyvät tunnus -se sekä

86 87 SOFIA BJÖRKLÖF University of Helsinki

Mutual contacts and lexical relations among the Finnic varieties of western Ingria and northeastern Estonia1

Abstract The aim of this article is 1) to describe the historical language contact situation between the genetically closely related Finnic varieties of western Ingria, 2) to give examples of the nu­ merous loanwords originating from mutual contacts among local Finnic varieties as well as areal diffusion, and 3) to discuss the method of investigating contacts and borrowing among closely related varieties. The data are taken from old dialectal materials published in vocabularies and dictionaries as well as preserved in archives. The words that are analysed and discussed etymologi­ cally in more detail are drawn from Vote, Ingrian, and Estonian. Although it is often difficult to confirm the direction of borrowing among closely related varieties, I seek to determine the direction of diffusion in the varieties whose development cannot be de­ scribed merely in terms of a traditional binary family tree model. Examples of mutual borrowing between Vote, Ingrian, Estonian, and Finnish are presented. Estonian loanwords in Vote and Ingrian can usually be recognised by their distribution. Most vocabulary originating as loans (in Vote, Ingrian, and Estonian) has been bor­ rowed from Finnish. Loans in both Vote and Estonian often have a distribution not only in Ingrian but also in Finnish. Because of the phonetic similarity of these varieties, the donor variety usu­ ally cannot be defined. Vote loanwords occur only sporadically in Ingrian and Estonian: they may also form a substratum.

1. This study is written as a part of the research project “Language change in multi­lingual Finnic”, funded by the Kone Foundation.

Multilingual­ Finnic. Language contact and change. 89–153. Uralica Helsingiensia 14. Helsinki 2019. ‹https://doi.org/10.33341/uh.85034› SOFIA BJÖRKLÖF MUTUAL CONTACTS AND LEXICAL RELATIONS AMONG THE FINNIC VARIETIES OF WESTERN INGRIA AND NORTHEASTERN ESTONIA

The speakers of Finnic varieties in western Ingria used to 1. Introduction live in old rural communities with long-term multilingualism, vil­ lages with a mixed population, and vague language boundaries. The vocabulary of the Finnic varieties in western Ingria and north­ The arrival of new inhabitants from the countries, which ruled eastern Estonia reflects the historical language contact situation of the this area and the foundation of St. Petersburg in 1703 changed the ethnographic balance between different peoples in Ingria. area in several ways. The closely related Finnic varieties spoken in the This increased linguistic diversity and altered the hierarchy of the area are Vote, Ingrian, Estonian, and Ingrian Finnish. When speaking languages leading gradually to accelerating language and identity about these, I use the word “variety” instead of “language” in order shift of the local peoples of Ingria. to the fact that their situation in their traditional speech areas Keywords language contact, loanwords, etymology, dialectology, resembles more a dialect continuum, especially when it comes to Vote family-internal borrowing, receptive multilingualism, Vote, Ingrian, and Ingrian. Estonian, Ingrian Finnish The traditional settlements of speakers of Vote and Ingrian were once located in western Ingria, the southeastern coastal area of the between the Narva River and the present-day metropolis of St. Petersburg. At the beginning of the 18th century, St. Petersburg was established at a location that had been an ancient trading centre of Finnic peoples. Over the centuries, both Russia and Sweden have gov­ erned the Ingrian area, whereas the local people have never formed a politically independent state there. In the 17th century, Finnish speak­ ers migrated to Ingria, and following the foundation of St. Petersburg in 1703 and the Treaty of Nystad in 1721, Russian migration increased considerably. The most recent Finnic-speaking newcomers, Estonian migrants, started to settle in the same area in the second half of the 19th century. Historically, German was spoken in Ingria to some extent and presumably, though more temporarily, Swedish was as well. The arrival of new inhabitants from the countries, which have historically dominated this region politically, changed the ethno­ graphic balance between different peoples in Ingria. This increased linguistic diversity and altered the language hierarchy leading gradu­ ally to accelerating language and identity shift of the local peoples of Ingria. Vote and Ingrian speakers were plurilingual and some of their villages in the Vaipooli area (located in the Lower Luga area), namely Jõgõperä, Liivtšülä, Luuditsa, and Rajo, had a mixed population. In the easternmost Estonian parish of Vaivara there were also contacts between Estonians and speakers of Vote, Ingrian, and Finn­ ish. The fishermen in the Vaipooli area met other fishermen speaking other Finnic varieties. (Ariste 1968: 14; 1981: 52–59, 79.) Presum­ ably, Vote was once spoken across a wider area than in the 19th and

90 91 SOFIA BJÖRKLÖF MUTUAL CONTACTS AND LEXICAL RELATIONS AMONG THE FINNIC VARIETIES OF WESTERN INGRIA AND NORTHEASTERN ESTONIA

20th centuries and speakers of Vote lived alongside Estonian speakers in the present-day Vaivara region and also somewhat more to the south (Ariste 1965a: 110; 1965b: 92; Ernits 2005: 82, 83). Ingrian, on the other hand, is supposed to have been spoken as far as the present-day location of St. Petersburg and even to the Sestra River on the south­ ern Karelian Isthmus as well as to the Nazija River in the east (Ernits 2007: 13).

2. Hypotheses and aims

My main hypothesis is that there are numerous loanwords originating from mutual contacts among local Finnic varieties; however, among closely related varieties it is often difficult to define the direction of borrowing. Nevertheless, the main aim is to find the origin of shared words, which in some of these Finnic varieties result from areal diffu­ Map 1. The location of Ingria. Map from the back cover of Teinonen & Virtanen sion rather than descending directly from the common protolanguage. (eds, 1999). Furthermore, I seek to determine the direction of the diffusion and to reconstruct the language contact situation in Ingria in the 19th and 20th centuries. This article has a twohold objective. In section 6, I describe the historical sociolinguistic situation in western Ingria, as this has not been done to this extent before. The data are collected from various contemporary writings. Section 7 deals with vocabulary and is divided into two parts. I have gathered all the words with at least two Finnic donor varieties found in previous studies and present them in the first part of the section. The latter part of the section is an attempt to de­ scribe the method of investigating mutual borrowing between closely related varieties by analysing eight selected words. My aim is to make Map 2. Parishes in northeastern Estonia and western Ingria, and the Gulf of Fin- the method more transparent and, if possible, to develop it somewhat. land islands. Map base from ALFE, markings by the author. Estonian parishes: Jõelähtme,­ Kuusalu, Kadrina, Haljala, Viru-Nigula, Lüganuse, Jõhvi, Vaivara. Finn- ish parishes in Ingria: Kallivere, [Kosemkina]2, [Narvusi], [Kattila], [Soikkola], [Kaprio], [Hevaa], Novasolkka, Moloskovitsa. Finnish islands: Suursaari, Tytärsaari, Lavansaari, Seiskari. Vaipooli is located at the head of the bay in the parish of Kattila.

2. Finnish was not the main Finnic language in parishes written in [square brackets].

92 93 SOFIA BJÖRKLÖF MUTUAL CONTACTS AND LEXICAL RELATIONS AMONG THE FINNIC VARIETIES OF WESTERN INGRIA AND NORTHEASTERN ESTONIA 3. Data and methods As for the old sociolinguistic language contact situation, the ex­ isting principal resources are the writings of scholars who carried out The empirical part of the current article is based on old dialectal word their fieldwork in the 20th century. These include fieldwork journals materials on all Finnic varieties published in vocabularies and dictionar­ written on the trips as well as articles based on the impressions while ies. In the case of Estonian and Finnish, the materials are still partly un­ out in the field but written and published later. My main sources are published and are stored in archives. The materials were mostly collected the posthumously published texts of Paul Ariste, who conducted his between the beginning of the 20th century and the 1970s and they repre­ fieldwork in Ingria in 1942–1980. His diaries did not appear in print sent the situation of these varieties at a point when the development of until 2005. Although Ariste conducted his work mostly during the So­ multilingualism as well as language and identity shift had been ongoing­ viet years, he is known to have written in a way which really did not for a long time. The analysis and detailed etymological discussion in this take Soviet censorship into consideration. Ariste’s notes can therefore section focuses on words in Vote, Ingrian, and Estonian. Some words be considered relatively reliable and the fact that his journals were are discussed in more detail in order to make the method more transpar­ written while out in the field and immediately after the situations he ent, because often older works in etymology and especially etymological describes makes them exceptionally important. Another source are dictionaries only give the conclusion, not the articulation leading to it. the memoirs written by Lauri Kettunen about his fieldwork. Kettunen The methods used are contact linguistics, etymology, lexicology as conducted fieldwork with Vote speakers in 1911–1915. His book, well as comparative dialectology. In the case of closely related varieties however, is written decades afterwards and it was most likely also in contact, the classical criteria of etymology – phonology, semantics, edited prior to its publication in 1945. Therefore, it cannot be consid­ and areal distribution – unfortunately are not sufficient alone. Recognis­ ered as authentic or reliable as Ariste’s notes. My third source is the ing borrowing between closely related varieties is complicated, because field journal of ethnologist Ilmar Talve from a trip in 1942, published there are only a limited amount of phonological features which can re­ in 1990. Talve translated the text from his native Estonian into Finnish veal loanwords. These are mainly found in Vote, for instance sound before publishing it. He does not mention editing the text and it has characteristics suggesting a loan origin (see, Suhonen 1986; Lauerma maintained its travel journal style. 1993: 160). Furthermore, defining the direction of diffusion or the pre­ The historical mutual contacts of genetically closely related cise loan source is often difficult and left unanswered in previous stud­ Finnic varieties have received little attention in modern Finno-Ugrian ies and vocabularies (Lauerma 1993; VKJo; VKKu; IMS). research. These three doctoral dissertations on morphology and pho­ When studying closely related varieties in contact, areal distri­ nology are the only larger existing works addressing this topic: Helka bution becomes a more significant factor. Additionally, knowledge of Riionheimo (2007) has studied the effects of contacts with Estonian in the (cultural) history of the area in question and any sociolinguistic the past tense formation of Ingrian Finnish speakers living in Estonia, information is crucial, because the context of the contact has to be Ossi Kokko (2007) investigated the use of some cases in the speech of taken into consideration. As for the relationship between the Estonian Ingrian Finnish speakers living mostly in Estonia, and Petri Lauerma’s Northeastern Coastal dialects and Finnish, Suhonen (1979: 360–364; (1993) dissertation focuses on Vote , but also exam­ 2000: 373) has pointed out that the most important criteria to be taken ines loanwords with both Finnic and Russian origin (especially pp. into consideration are the following: distributions of words, distribu­ 165–192). Viitso (1993: 526) has considered this study to be of great tions of their meanings, contacts with neighbouring dialects, the origin importance for the study of Vote etymology. I am currently conducting of the possible loan original (especially in the case of a young loan­ lexical research in the field for my doctoral dissertation (a monograph word), inflectional types and categories, and the historical likelihood in Finnish) on the language contact situation in western Ingria and of it being a loan. All these factors suggest what is possible and plau­ northeastern Estonia from the viewpoint of internal borrowing within sible when investigating the direction of linguistic diffusion. a . 94 95 SOFIA BJÖRKLÖF MUTUAL CONTACTS AND LEXICAL RELATIONS AMONG THE FINNIC VARIETIES OF WESTERN INGRIA AND NORTHEASTERN ESTONIA 4. The Finnic varieties analysed in this study and their speech areas in western Ingria

In a broader context, the Finnic languages are closely related languag­ es forming a subbranch of the Uralic language family. This article fo­ cuses on Vote, Ingrian, and Estonian with a special emphasis on the Estonian Northeastern Coastal and Eastern dialects.

4.1. Vote

Vote was spoken almost exclusively in western Ingria. Vote is tradi­ tionally considered to have four dialects: Western, Eastern, Kukkuzi, and Krevin Vote (see Map 3; classical division of the dialects Kettunen 1915; Heinsoo 1998: 19–22; VKS). Western Vote, as the largest sub­ dialect, was spoken over the most extensive area. Modern research divides Western Vote into two groups: Central Vote, which was spoken in the Kattila region, and Lower Luga Vote, which is still spoken by a couple of elderly people in the villages with a mixed population, such as Jõgõperä, Liivtšülä, Luuditsa, and Rajo (Ernits 2005: 77; Muslimov 2005; Markus & Rozhanskiy 2011a; Kuznetsova et al. 2015: 130). In the Finnish tradition, these groups have been referred to as the Kattila and Vaipooli (or Vainpuoli in Finnish) dialects. Eastern Vote was spo­ ken in the villages of Itšäpäivä, Mahu, Iivanaisi, Kliimettina, Koslova, and Kaprio. When a particular word occurs in the eastern area then it can be considered old and therefore especially noteworthy. The East­ ern dialect died out in the 1970s. Kukkuzi Vote spoken in Kukkuzi vil­ lage is, in fact, a mixed language: it has an Ingrian phonetic and lexical superstrate with a Vote grammatical base (Suhonen 1985; Muslimov 2005; Markus & Rozhanskiy 2011b; 2012). Krevin was the dialect of the Vote speakers who were relocated to Latvia in the 15th century and Map 3. Dialect areas of Vote and locations of villages. Extracted from the map by which became extinct during the 19th century. There is only a small Tiit-Rein Viitso in Heinsoo (1998: 28). Idavadja = Eastern Vote. The Kattila area is amount of material on this variety, but it is important for investigating divided into the Orko ‘valley’ and Mäči (= Mätši) ‘hill’ areas. Rajo is not marked on the map but it is located west of Jõgõperä on the same side of the Luga River. the history and earlier stages of development of Vote. Liivčülä = Liivtšülä, Luucca = Luuditsa, Ičäpäivä = Itšäpäivä. Laugaz = Luga River; Jaama = Kingissepp.

96 97 SOFIA BJÖRKLÖF MUTUAL CONTACTS AND LEXICAL RELATIONS AMONG THE FINNIC VARIETIES OF WESTERN INGRIA AND NORTHEASTERN ESTONIA

course of the Luga River. The latter dialect is only fragmentarily docu­ mented, but it may be useful to research, as it could have preserved old vocabulary, which in other dialects of Ingrian may have been replaced with words from Ingrian Finnish dialects (Laanest 1970: 109). It has been assumed that Oredež Ingrian was formed when speakers of Ingrian fled the Swedish regime to the distant Oredež area in the 17th century.

4.3. Estonian

Estonian was spoken throughout all of Ingria at the beginning of the 20th century. The areas from which the Estonian-speaking immigrants moved to Ingria remain unclear, but it is likely that they came from all over the Estonian-speaking region. Estonian was spoken in west­ ern Ingria in the Kattila region and surrounding areas by over 2 000 people at the beginning of the 20th century, at least in the Vote villages of Itšäpäivä (in the Eastern Vote area), Jarvikoištšülä, Kõrvõttula, Mati, Pummala, and in the mixed population Vote villages of the Lower Luga region (Jõgõperä, Liivtšülä, Luuditsa, Rajo, and Kukkuzi) (Ariste 1987: 21–30; 1998: 15–16). At the end of the 19th century, in some parishes in western Ingria even a third of the inhabitants were speakers of Estonian Map 4. Ingrian dialect areas and villages according to IMS (without the Oredež area). Extracted from the map in IMS (VII). Dialect areas: IV Lower Luga Ingrian, (Hakamies 1991: 201; repeated by Leskinen 1995: 172). There were I Soikkola Ingrian, II Hevaa Ingrian. The city of Narva should be marked on the already 64 1164 Estonians in Ingria in 1897 according to the popula­ western side of the Narva River. Jaama = Kingissepp; Hevaanjoki = Hevaa River. tion census conducted in Russia that year. According to Ariste (1998: 15), the contacts between speakers of Estonian and Vote did not be­ come intensive until the second half of the 20th century. According to 4.2. Ingrian Muslimov, the largest number of Estonians lived in Novasolkka5 parish (Mehmet Muslimov, p.. 16 & 17 October 2015; see Map 6). Speakers Ingrian has four main dialects: Lower Luga, Soikkola, Hevaa, and Oredež Ingrian (see Map 4; classical division of the dialects Porkka 4. Of which, 12 238 in the city of St. Petersburg. 3 5. Muslimov has had Estonian language consultants in at minimum the following 1885 ; Laanest 1961; Nirvi 1961; 1971). Lower Luga Ingrian is still villages in Ingria: Sakkola, Novesi, Zapalje, Ivanovskoje, Tikanpesä, Kattila, Kupanit­ spoken by some elderly people along the lower course of the Luga River sa, Viron-Priiskova (Rus. Krasnaja-Priiskova, not marked on the Finnish road map of (and a mixed Finnish/Ingrian variety partly on the Kurkola peninsula Ingria from 1992 (compiled by Roland Randefelt) but located between Vennäin-Priis­ (Kuznetsova et al. 2015: 131)). Soikkola Ingrian is spoken by some el­ kova and Hakuli), Lopitsa, Moloskovitsa, Spankkova, Serepetta, Muho­vitsa, Edasi (an Estonian kolkhoz in Kikkeri), Uusi-Hinkkala (near Spankkova), Markkusi (Rus. Mar­ derly people on the Soikkola peninsula, Hevaa Ingrian was spoken in guzi), Simetsa (a village founded by Estonians), Arokylä (Rus. Ara-), Reskutsa (Rus. the Hevaa River area, and Oredež Ingrian was spoken along the upper Treskovitsa), Marvitsa, Suur-Rutja or Pien-Rutja (in either of the villages), Raakovitsa, Saappola, Prömpeli, Mustapää, Keskikylä (Rus. Srednje, not marked on the map but 3. Although, according to Porkka (1885: 17–18), the Lower Luga dialect is not a located northeast from Klenna roughly by the number 12 on the Finnish road map dialect of Ingrian. For criticism, see Laanest (1961: 200–202). of Ingria), Ivanskoi, and possibly Asikka (Mehmet Muslimov, p.c. 17 October 2015).

98 99 SOFIA BJÖRKLÖF MUTUAL CONTACTS AND LEXICAL RELATIONS AMONG THE FINNIC VARIETIES OF WESTERN INGRIA AND NORTHEASTERN ESTONIA

dialects of Estonian and as a substrate in place names (Ariste 1965a: 109–110; 1965b: 92; Pall 1969: 6–7, 261–263, 303–304; 1970: 12–17; 1977: 16, 228–230). However, this assumption has been made mostly on the basis of onomastics and archaelogy and has been criticised by Grünthal (1997: 113–149, especially 119–121). The main Estonian dialects historically in contact with Vote and Ingrian are the Northeastern Coastal and Eastern dialects of Estonian (see Map 5). The Estonian Northeastern Coastal dialects were spo­ ken in the parishes of Jõelähtme, Kuusalu, Haljala, Viru-Nigula, Lü­ ganuse, Jõhvi, Iisaku, and Vaivara; the Estonian Eastern dialect was spoken in the parishes Iisaku, Torma, Kodavere, Maarja-Magdaleena, Palamuse, and Laiuse. The Northeastern Coastal dialects are divided in the Coast dialect (Jõe, Kuu, Hlj, Vai) and the Alutaguse dialect or Northeast dialect (VNg, Lüg, Jõh, IisR) (Pajusalu et al. 2012: 246).

5. Languages in contact with the Finnic varieties examined in this study

Map 5. Estonian dialect areas and parishes in the northeastern corner of Estonia ac- Ingrian Finnish, that is, the Finnish dialects spoken in Ingrian terri­ cording to Pajusalu (1999). Map extracted from the map in Pajusalu et al. (2009: 56). tory, is considered a variety with close contacts with the other Finnic The Estonian Northeastern Coastal dialects were spoken in the parishes from Jõe­ lähtme to Iisaku and Vaivara (Jõe, Kuu, Hlj, VNg, Lüg, Jõh, Iis (the abbreviation for the varieties of this area. Russian, although not examined in this article, dialect is IisR), Vai); the Estonian Eastern dialect in the parishes from Iisaku to Maarja- also had a substantial effect on the Finnic varieties, while German and Magdaleena (Iis, Trm, Kod, MMg, Pal, Lai). Kadrina and Rakvere (Kad, Rak) belong to Swedish did not. the Central dialect, Tartu-Maarja (TMr) and a part of Kodavere and Maarja-Magdalee- na (the abbreviations for the dialects are KodT, MMgT) belong to the Tartu dialect. 5.1. Ingrian Finnish and other dialects of Finnish of Vote, Ingrian, and Finnish came in contact with speakers of Estonian Ingrian Finnish was the largest Finnic variety in Ingria in terms of mostly in or nearby the city of Narva in eastern Estonia, at the Narva- number of speakers and was spoken across this entire area. Only in Jõesuu harbour, and in the city of Jaama6 in western Ingria, where the western Ingria was Finnish not the main Finnic variety. (Saloheimo Estonian congregation had 4 500 members. Inhabitants of Ingria also 1991: 81.) Ingrian Finnish dialects are classified as Southeastern dia­ worked in these places and went there to trade goods. (Ariste 1981: lects of Finnish. The speakers of Ingrian Finnish had originally moved 52; 1987: 30; 1998: 17; Talve 1990: 64.) Conceivably, the dialect of to Ingria in the 17th century after the Treaty of Stolbovo in 1617, when Vaivara parish is of special importance (Viitso 1993: 524). It has been Sweden began to colonise Ingria. The newcomers from Äyräpää par­ conjectured that an old Vote-speaking population in eastern Estonia has ish on the Karelian Isthmus were called äyrämöiset, while the others, left traces in the vocabulary of the Eastern and Northeastern Coastal whose precise origin is harder to define and is not discussed here, were 6. Jamburg in German, nowadays Kingissepp. called savakot and the Narvusi Lutherans. The main Ingrian Finnish

100 101 SOFIA BJÖRKLÖF MUTUAL CONTACTS AND LEXICAL RELATIONS AMONG THE FINNIC VARIETIES OF WESTERN INGRIA AND NORTHEASTERN ESTONIA

dialects were the äyrämöis-dialects, spoken mainly along the coast of Ingria, and savakko-dialects, spoken mainly inland. The smallest dia­ lect, the Narvusi or Kosemkina dialect, also classified as the Finnish dialects of Kurkola and Rosona (in Leppik 1966; 1972; 1973; 1975 and Mägiste 1925), is strongly influenced by the other Finnic varieties of the area. The background of the Finnish dialects of Ingria was not uniform and Ingrian Finnish in the 20th century was already a fusion of different dialects. (Leskinen 1991: 230; 1995: 170–171; Savijärvi 1996a: 8–10; Riionheimo 2007: footnote 8, 18). Speakers of other Southeastern dialects of Finnish were encountered while fishing on Seiskari, Lavansaari, and Tytärsaari – islands in the Gulf of Finland. In St. Petersburg, there likely were speakers of even more Finnish va­ rieties, for example the Southern Karelian dialects of Finnish. For the Map 6. Ingrian Finnish dialects (northern parishes only partially shown) accord- Ingrian Finnish dialects in western Ingria, see Map 6. ing to SMS. Map base from ALFE, markings by the author. The dialect spoken on Suursaari (Suu) belonged to the Kymi dialects; the dialects spoken on Tytärsaari, Lavansaari, and Seiskari (Tyt, Lava, Seis) belonged to the Southeastern dialects 5.2. Other languages of Ingria (South Karelian dialects) of Finnish proper. Dialects from Lüganuse to Vaivara are dialects of Estonian. The Ingrian Finnish dialects spoken in Kallivere7, Kosemkina, The most widely spoken language in Ingrian territory was, of course, 8 Narvusi , Kattila, Soikkola, Kaprio, Hevaa, Novasolkka, and Moloskovitsa (Kall, Kose, Russian. Other Indo-European languages also have been spoken Nar, Kat, Soik, Kap, Hev, Nov, Mol) are considered the western group of Ingrian Finnish by Muslimov (2009), followed by Kuznetsova et al. (2015: 132–133). Kosem- in this area: Hakamies (1991: 200) conjectures that as 6 600 of the kina, Narvusi, Kattila, Soikkola, Kaprio, Hevaa are parishes where Finnish was not 11 490 Germans in Ingria in 1848 lived in Kronstadt and Narva (data the main variety. Vote speakers lived in the parishes of Narvusi, Kattila, Hevaa; In- by von Köppen 1849; 1867), there were relatively few Germans in grian speakers lived in the parishes of Narvusi, Kattila (in the Vaipooli area), Soik- the countryside and that they did not have contacts with or impact on kola, and Hevaa. The circled area is the main area – aside from the cities – where speakers of the Finnic varieties spoken there. Presumably, at one point Vote and Ingrian speakers would encounter each other most: on the islands this was in connection with fishing, on the mainland it was due to the location of the there were also Swedish speakers in Ingria, at least Krjukov (1993 areas in which they lived. [1987]: 24) mentions that the and Germans who moved to Ingria in the 17th century, assimilated into other Lutherans (i.e., Finn­ ish- and Estonian-speaking people) in the 18th and 19th centuries. For example, the Swedish-speaking chronicler Thomas Hiärne (1638– 1678) of Livonia was born in Skuoritsa, central Ingria. In comparison to the number of speakers of German, there were 7. Interestingly, in Kuznetsova et al. (2015: 133), Kallivere is not mentioned in the discussion of Ingrian Finnish dialects of western Ingria. Tyrö is regarded as a west­ 5 148 , 17 800 Ingrians, 76 069 , and 3 522 Estonians living ern Ingrian Finnish dialect. in Ingria in 1848 when the peoples of Ingria were documented for the 8. Kuznetsova et al. (2015: 133) call the Ingrian Finnish dialect spoken in Narvusi first time by ethnologist Peter von Köppen (von Köppen 1867: 20, 41, parish the Lower Luga dialect. In fact, Kettunen (1930: 191, 193, 194, 195) also 92, 105, 114). writes in Finnish about the [Finnish] dialect of the Lower Lugans, the dialect of (the Lutherans of) Lower Luga, the [dialect] group of Lower Luga, the Lower Lugans, and the Lower Luga dialect.

102 103 SOFIA BJÖRKLÖF MUTUAL CONTACTS AND LEXICAL RELATIONS AMONG THE FINNIC VARIETIES OF WESTERN INGRIA AND NORTHEASTERN ESTONIA 6. The historical sociolinguistic situation in western Ingria

Sociolinguistically, in these Finnic-speaking rural communities, the language boundaries were extremely vague. In fact, both Vote and Ingrian speakers have repeatedly expressed the opinion that their dia­ lects are dialects of the same language (Ariste 2005: 47).

6.1. Plurilingualism and receptive multilingualism

In the Vaipooli area in western Ingria, the plurilingualism of the last speak­ ers of Vote and Ingrian can be described as receptive multilingualism: they used their own language when communicating with each other (Ket­ tunen 1945: 237; Ariste 1957: 122; 1958: 148; 1981: 58). The traditional use of the different Saami languages was also like this (Pasanen 2015), and using one’s own language for contacts between speakers of the Saami languages and speakers of Finnish was a traditional model in the Saami area (M. Aikio 1988: 73–74, 77). These languages are, in principle, ge­ netically related but not mutually intelligible, and receptive multilingual­ ism is based on speakers living in the same area for a long time. Due to a shared basic vocabulary and grammar the closely related Finnic varieties have to some extent. However, receptive multilin­ gualism in western Ingria was based on and very much supported by the fact that speakers of the different Finnic varieties of this region had lived Map 7. Finnic varieties spoken in the villages of western Ingria (extracted from the map in Laanest 1964: 4). The map shows the mixed Vote- and Ingrian-speaking for an extended period of time alongside each other in an area consisting villages in the Lower Luga area. Otherwise, the map is prepared from the point of of villages with mixed population and long-term multilingualism. Ariste view of the Ingrian variety, and therefore some of the Vote-speaking villages and (2005: 69) also notes that the Ingrians, who had had more contact with ei­ many of the Finnish-speaking villages are left unmarked, for example, Lempola, ther Finnish-speaking or Estonian-speaking people, better understood the Pummala, and Mahu, where both Vote and Finnish were spoken. Luutsa = Luu- language in question. Talve (1990: 64) notes that in 1942, almost all the ditsa. Luuga j. = Luga River; Kingissepp = Jaama. Votes in Vaipooli understood Estonian, because there had been Estonian Explanations: Ingrian | Vote | Finnish | Vote & Ingrian | Finnish & Ingrian inhabitants in Vaipooli. At the end of the 19th century, when even a third of the inhabitants were Estonian speakers in some parishes in western In­ In fact, many speakers of Finnic varieties – especially the Votes and gria, Finnish speakers in western Ingria would even read Estonian books Ingrians – in western Ingria also spoke each other’s varieties. Accord­ when they did not have Finnish books (Hakamies 1991: 201).9 ing to Kettunen (1930 [1915]: 7), this was the case for Vote and Finnish speakers in Lempola, Pummala, and Mahu. In Luuditsa (Luucca, Luutsa 9. On receptive multilingualism in contemporary Finnish–Estonian interaction see Härmävaara (2013; 2014; 2017; forthcoming). However, in spite of the studied their own countries and then trying to interact with one another can not be compared languages being closely related, the situation of Finns and Estonians brought up in with the situation of western Ingria directly.

104 105 SOFIA BJÖRKLÖF MUTUAL CONTACTS AND LEXICAL RELATIONS AMONG THE FINNIC VARIETIES OF WESTERN INGRIA AND NORTHEASTERN ESTONIA on maps 3 & 7), there lived “truly bilingual Votes”, who were equally low PRESTIGE high proficient in Ingrian and Vote (Ariste 1981: 62). In Soikkola, Finnish Vote Ingrian Estonian Finnish Russian speakers could understand Vote (M. Paulaharju 2010: 142 [original smallest rural rural newest rural largest rural urban source S. Paulaharju 1915]). However, the data available are somewhat language language language language language ambiguous. For example, it has been said that in the past, speakers of no standard no standard standard standard standard these four different Finnic languages usually did not learn each other’s language language language language language varieties because they got along by only speaking their own (Ariste 1957: no literary no literary (literary use education, education, 122). In the 20th century, Ingrian became the more common language of use use* in Estonia)** media media communication between Vote and Ingrian speakers, because there were no adminis­ no adminis­ no adminis­ (administr. adminis­ more speakers of the latter variety. Ariste states that the last speakers of tration tration tration in Finland) tration Vote in the Vaipooli area knew Ingrian but not vice versa. (Ariste 1981: no church no church no church Lutheran Orthodox 58.) It has also been said that services in Lutheran churches also had to church church be held in Estonian, because Estonian speakers did not understand Finn­ Table 1. Prestige, status, and use of the languages of western Ingria at the begin- ish (Kettunen 1957: 126–127); however, this is likely due to the fact that ning of the 20th century. speakers of Estonian were the most recent immigrants to the area. * With the exception of an attempt in the 1930s. ** In schools in so-called Estonian Ingria in the 1920s and 1930s. 6.2. Language prestige and asymmetry of contacts even an attempt to create a standard language and to use it as a language The prestige of Russian, the urban and a fully developed languge used of instruction in schools in the 1930s. In his often quoted article, Tsvet­ in all domains, was high, whereas the rural languages Vote and In­ kov (1925: 43) notes that when an Ingrian-speaking wife was brought grian had neither official status nor literary use and were used only to a Vote-speaking household, the whole family shifted their language in informal domains. Russian, on the other hand, was the language to Ingrian. Though, most likely they shifted to a convergent variety, of administration, education, and the Orthodox church10. Finnish was as suggested by Fedor Rozhanskiy (p.c. 20 August 2015; more about also a rural language in Ingria, but as the major Finnic variety it had this variety in Rozhanskiy & Markus 2013: 230; 2014). According to higher prestige than the smaller Finnic varieties also because of its use Rozhanskiy, the women were mostly from Soikkola, and, of course, the in the Lutheran church and later in the Finnish schools of Ingria. Its Ingrian spoken in the Lower Luga area is not the Soikkola variety. It use among speakers of Vote and Ingrian – considered “Finns” by the was also typical for speakers of Vote to speak Ingrian with speakers of Lutheran priests at one time – had even been supported by any means Ingrian despite the fact that Ingrian-speakers knew Vote as well (Tsvet­ possible in the 17th century (Savijärvi 1998: 274). kov 1925: 43), but this also may have been a mixed variety. Vote was Generally speaking, the language contacts and use of local varie­ very rarely used with children. At the last stage, it has been reported ties has been asymmetric, as also Markus & Rozhanskiy (2010; 2013) that there were children in only two families speaking Vote with their have pointed out in their studies of Vote and Ingrian. These two Finnic Vote-speaking father and Ingrian with their Ingrian-speaking mother at varieties had the lowest prestige in Ingria; however, Ingrian was still the beginning of the 20th century (Ariste 1968: 15; 2005: 98, 111, 112). in a more powerful position. Ingrian had more speakers and there was Usually, Russian was preferred with children because it was believed that they would absolutely need Russian in the future. 10. The language used in masses held at the Orthodox church was, of course, Church Slavic, which was not intelligible (even) for speakers of Russian. However, Estonian, like Finnish, had a standard language and literary use. the church was considered the Russian church without a doubt. These languages had a status in their homelands which added to their

106 107 SOFIA BJÖRKLÖF MUTUAL CONTACTS AND LEXICAL RELATIONS AMONG THE FINNIC VARIETIES OF WESTERN INGRIA AND NORTHEASTERN ESTONIA status in Ingria, but the autochthonous languages of Ingria – Vote and One of the few means used for describing belonging to a particu­ Ingrian – were spoken only in Ingria. In so-called Estonian Ingria, a lar Finnic group was a distinction Vote and Ingrian speakers would ten kilometre-wide strip on the eastern side of the Narva River, which make. Namely, the verb used for ‘speaking’ is different: in Vote pa- was ceded to Estonia in 1920, the language of education was Estonian jattaa, in Ingrian läkkää, and this difference was used to define the in 1922–1926 (in Kallivere, Vanaküla, Kullaküla) and in 1933–1940 language of the other speaker. The terms pajattaja and läkkääjä, both (in Kallivere, Vanaküla) (Alenius 2015). However, the Finnic varie­ meaning ‘speaker’, were used without the potential name of the lan­ ties in this article are examined as they were spoken during a period guage spoken. However, it seems that the speakers understood this mostly during the middle of the 19th century and this political develop­ difference to be dialectal. (Ariste 1981: 62; 2005: 66, 83, 92, 97, 124, ment is more recent. In the case of Russian, the position of the domi­ 125, 126, 142.) Another means for drawing this distinction was to use nant language should be discussed in more detail. Ultimately, political terms derived from local place names: names of villages, parts of vil­ power guaranteed a more prestigious status for Russian. lages, or village groups (Ariste 2005: 28, 30). On the other hand, some speakers have referred to songs in Finnic languages as Finnish songs 6.3. Ethnic and linguistic identities in order to draw a distinction with Russian culture (Ariste 2005: 55). The adjectives Vo. soomalain, Ing. suumelain ‘Finn’ have been used Accordingly, the ethnic and linguistic identities of the speakers of Vote for speakers of Finnish in Ingria, and people from Finland have been and Ingrian were vague: the main factor in defining identity was reli­ called Vo. soomõõmaakko (approximately ‘a person from Finland, gion, which meant that local people could define themselves as Rus­ Finlander’) or even Vo. finn(i) (a Russian loanword). All Lutherans sian because of their “Russian” religion. The Greek Orthodox religion may also have been called Finns or “of Finnish belief” – also Esto­ played a significant role and it in fact was the main connective factor nians and Germans – because the Lutheran church was the Finnish between the speakers of Vote and Ingrian and the . In spite of church in Ingria. (Ariste 2005: 31–32; IMS s.v.; VK s.v.) One repre­ their language, every person in the Orthodox faith could be considered sentative example is when on a field trip in 1942, Ariste asked in Vote a Russian. For example, mixed marriages were allowed only with oth­ if an old Russian man speaks Vote. The man answered in Russian: er Orthodox people, not with the Finnish-speaking Lutheran people. “Немного по лютерански говорю”, ‘I speak a little bit of Lutheran’ Being Russian became more prestigious as time went on and having a (my translation and italicisation). (Ariste 2005: 37.) The concepts of Finnic ethnic identity might even have been dangerous after the Rus­ language, nation, and religion were completely intermingled in Ingria. sian Revolution – at least if it was written in one’s passport. Usually speakers of Vote marked their ethnicity as Ingrian (ижорец, ижорка 6.4. Language attitudes in Russian) and nationality as Russian when answering cencuses be­ cause they simply did not have a Vote identity. Working in the cities, As for language attitudes, there has been very little, if any, awareness especially in St. Petersburg, contributed to the language shift to Rus­ or interest among the speakers of the smallest Finnic varieties of Ingria sian. These facts led to a gradual assimilation where Ingrian speakers – Vote and Ingrian – in their own local varieties or in raising their pro­ assimilated into the Russians as Vote speakers assimilated into both file. In 1942, Ariste noted that Vote speakers were totally uninterested Ingrian speakers and the Russians. (Tsvetkov 1925: 41–42, 44; Ariste in and passive about the possibility of the extinction of their language. 2005: 31, 40, 67, 74.) In 1967, Ariste’s (2005: 104) language con­ Ariste encountered only two language consultants who felt otherwise: sultants were of the opinion that Vote and Ingrian were disappearing a Vote-speaking consultant who despised her mother tongue and an­ because of the deportations of their speakers to Finland and Russia other consultant whom he classified as “a true Vote patriot”. One would during the war, which is also certainly true. think that most likely not until the repressions would Vote and Ingrian

108 109 SOFIA BJÖRKLÖF MUTUAL CONTACTS AND LEXICAL RELATIONS AMONG THE FINNIC VARIETIES OF WESTERN INGRIA AND NORTHEASTERN ESTONIA speakers have underrated their mother tongue and called it “non-cul­ 7. Vocabulary tural” or “the language of calves” or “sheep” – and in cases where this occurred, this probably would have been due to mistreatment. How­ This section will discuss earlier studies dealing with mutual borrow­ ever, it is interesting that in 1913 in Mahu, where Eastern Vote was ing among Finnic varieties in Ingria. In the portion focusing on em­ spoken, Kettunen was told that fifty years earlier Vote had been spoken pirical research, word examples found in Vote, Ingrian, and also Es­ all over the village but then the speakers had suddenly begun to scorn tonian are discussed from an etymological perspective. Despite the their language. One reason for this might have been that unlike Finn­ presence of individual words in these varieties, these are examples of ish and Estonian, Vote had no standard language, as one speaker of areal diffusion, not genetic inheritance. The words analysed here illus­ Vote interviewed by Kettunen stated: “kammuga tšeeli, bõõ tširjaa bõõ trate the etymologically most unambiguous cases, and are given here pappia” ‘such a language, no book, no priest’ (my translation). This in order to describe the method of etymologising borrowing among may sound slightly negative but can, nevertheless, be interpreted as a closely related varieties. Each is categorised according to its probable rather neutral statement concerning the situation of this variety. In any donor variety. In earlier research, some have noted the senselessness case, this development is likely to be somewhat earlier compared with of attempting to etymologise the direction of borrowing (Laakso in developments in the western speech areas of Vote, which Ariste visited VKJo 1995: IV–V). This is likely true in many cases; however, with in 1942, because Eastern Vote died out already in the 1970s. (Kettunen detailed and diligent work some proposals concerning the direction of 1945: 231; Ariste 2005: 43, 44, 65, 67, 74, 77, 87, 133.) borrowing can be posited using etymology.

6.5. Mixed varieties 7.1. Earlier research

Ariste often describes the language of his consultants as a mixed lan­ In the research literature, some Finnic loanwords with a distribution guage: Vote with Ingrian influence or even vice versa (Ariste 2005: across the three aforementioned varieties can be found. I have col­ 62, 65, 67, 73, 74–75, 95, 97, 101, 130, 138, 151, 157). This reflects lected these from the literature and will present them here with ­ the nature of plurilingualism and its effects on idiolects. However, the lish translations of their meanings (the translations from Finnish and question is, whether these varieties have always been mixed to some Estonian into English are my own except for the translations from extent. Like all of the dialectologists conducting fieldwork at that time, Söderman 1996). Ariste also looked for speakers of the so-called “purest” language or dialect possible. This idealised concept does not really reflect reality, as 7.1.1. Research on Vote no “pure” Vote-speaking or “pure” Ingrian-speaking communities have ever existed. A consultant might even say that someone speaks Vote, In his doctoral dissertation, Petri Lauerma (1993: 165–192) lists a but during his conversations with the person in question Ariste might total of 33 Ingrian, Finnish, or Estonian loans in Vote, which, accord­ instead say that their idiolect is Ingrian. The Kukkuzi dialect of Vote is ing to him, have a distribution in both Ingrian and Estonian as well. a topic of its own: it is a variety with substantial Ingrian influence on a However, I have already found wider Finnic distributions for some of Vote substrate, and it can be classified as a mixed language (Suhonen the words he mentions. These will be dealt with later in the ongoing 1985; Muslimov 2005; Markus & Rozhanskiy 2011b; 2012). study. Words given by Lauerma are the following: For more on the current situation of Ingrian Finnish, Ingrian, and Vote, see Rozhanskiy & Markus (2019), in this volume, and Kuznet­ sova et al. (2015). 110 111 SOFIA BJÖRKLÖF MUTUAL CONTACTS AND LEXICAL RELATIONS AMONG THE FINNIC VARIETIES OF WESTERN INGRIA AND NORTHEASTERN ESTONIA

Vo. ahteri ‘stern (of a ship)’, ennustā ‘to predict’, hame ‘skirt’, hert- to scratch’, kriŋgeli ‘pretzel’, kȫkki ‘kitchen’, лusti ‘grand, beauti­ tain ‘warm-hearted, kind, sweet, good’, heruttā ‘to get a cow to in­ ful, slender; joy, pleasure’, лustissi ‘beautifully’, *лustiuz ‘beauty’, crease its milk production when milking’, kaneppi ‘hemp’ (< Est. *лustivoitta ‘to make beautiful’, meisteri ‘master’, mekkiä ‘to Ariste 1941), kapteni ‘captain’, kena ‘nice, beautiful, slender, vain’, taste’, *millīn ‘what kind of’, narria ‘to tease, to jibe, to mock, to koer ~ koera ‘dog’, koppeli ~ koppeni ‘pen (for animals)’, kort- slander’, nikõrta ‘to do joinery (woodwork)’, passia ‘to fit, to be teli ‘apartment’, kovera ‘concave’, kukerpalli ‘somersault’, lajenki sufficient’, pekki ‘lard’, pikisell ‘to squash, to squeeze, to press’, ‘load’, lēkuttā ‘to swing’, meistari ‘master’, murhe ‘concern, worry, pikissä ‘to squash, to squeeze, to close’, pikisüz ‘squash, squeeze’, grief’, omena ‘potato’, panekki ‘a holiday during fast’, paperi polle ‘apron’, priski ‘sturdy, healthy, good’, pǟ-sūkkuri ‘loaf ‘paper, document, note’, pēntara ‘bank, edge’, puteli ‘bottle’, reh- sugar’, rīvõtoi ‘shameless, mean, obscene, disobedient, stubborn’, vata ‘to reef’, seilata ‘to sail’, seka ‘mix’, sekain ‘messy’, sekamī runni ‘heap (of pieces of ice blocks, stones)’, rūn ‘gelding’, rūnõt ‘messy’, tehota ‘to have an effect’, trehvata ‘to meet by accident, to ‘to geld’, rūto ‘link (in a chain), a piece of glass, windowpane; hit sth’, treijata ‘to roll, to turn, to twist’, venua ‘to stretch’, venuttā diamonds (a suit in playing cards)’, selettemin ‘explaining, expla­ ‘to stretch sth’, vettus ‘to get wet’ (Lauerma 1993: 165–192). nation’, sūkkuri ‘sugar’, sūkkurīn ‘sugary’, sūkkur-līv(õ) ‘granu­ lated sugar’, sūkkur-pǟ ‘sugar loaf’, taksi ‘charge’, tontti ‘evil Dmitri Tsvetkov’s Vocabulary of the Joenperä (Jõgõperä) dia­ spirit, devil, brownie, hobgoblin’, tōppi ‘tankard (also as an unit lect of Vote (Vatjan kielen Joenperän murteen sanasto, VKJo 1995, of measure)’, tōppīn ‘a tankardful (as a measurement)’, tramppia edited by Johanna Laakso) presents 87 loans in Jõgõperä Vote from ‘to stamp one’s foot; to limp’, tulli ‘rowlock’, turu ‘market place’, either Ingrian or Estonian. I have collected these from the VKJo re­ tükkü ‘piece’, tükküin ‘in pieces’, tükküitt` ä ‘by pieces’, tükük- verse vocabulary compiled by Johanna Laakso. An asterisk in front kõin ‘a small piece’, vāgõn ‘wooden plate, dish’, vokki ‘spinning wheel’, värski ‘fresh, raw’ (all < Ing./Est.). of a word indicates that the word form is an assumed nominative or infinitive construed by the Finnish editor. The words are: Vo. (J) āstaik ‘year’, āstaikõin ‘yearling’, bulli ‘bull’, bässi ‘ram’ 7.1.2 Research on Ingrian (+Fi.), eлusõлл ‘to live’, eлussa ‘to live’, halli ‘grey’, hallissua ‘to grey’, hapo ‘sour’, hapo-pīḿ ‘sour milk, ?curd’, hapo-rahk In his Dictionary of (Inkeroismurteiden sanakirja, IMS ‘?curd’, haria ‘to raise’, huлkkua ‘to wander around’, huлkkujõ 1971), R. E. Nirvi presents words borrowed into Ingrian from two ‘wanderer’, huлkkumin ‘wandering around’, huntti ‘wolf; seldom other Finnic languages: 40 loans from Estonian or Vote and 34 from used: sluggard, idler, loiterer’, hunttsu ‘sluggard, idler, loiterer’, Estonian or Finnish. However, many of these words share the same hursti ‘shroud, bed linen, sackcloth, tablecloth’, kaiлussa ‘to hug’, stem. I have collected these words from the Reverse Vocabulary of kaллõt ‘to pour (liquid)’, kaлõ ‘harsh, severe, rough (voice)’, kap- Ingrian dialects, Inkeroismurteiden käänteissanasto edited by Raimo teni ‘captain’, kena ‘beautiful, pretty, handsome, dandy’, kenap Jussila (1986). He uses the term donor language (Fi. lainanantaja­ kieli­ ) ‘more slender’, kenassua ‘to get beautiful, to get pretty’, ketti and indicates that the information concerning the donor languages is ‘chain’, ketti-koir ‘dog in chains’, kihittä ‘to chase away’, kihutta taken as such from Nirvi’s dictionary (Jussila 1986: IV). Nirvi, though, ‘to move, to lift, to throw (away), to speed’, kihutuz ‘speeding, throwing (away), removing’, kikertäss ‘to giggle, to trill, to war­ usually only compares words in different varieties with the Ingrian ble’, kīli ‘wedge’, kippua ‘to tend to, to hurry, to try impatiently’, words. An asterisk marks the words that are not entry words in IMS kittsi ‘small goat’, kleitti ‘dress’, kliŋkissä ‘to bolt’, kliŋkittä ‘to according to Jussila, an asterisk in parentheses marks the words that are lock, ?to bolt’, kliŋkittäss ‘to get closed’, kliŋkki ‘bolt’, kлoppia not entry words but are not marked as such by Jussila, and an exclama­ ‘to pile, to stack, to heap up’, kontturi ‘office’, krāmi ‘thing(s), tion mark shows the words which are not at all compared with other belongings’, krāppi ‘card, scraper’, krāppia ‘to scrape (leather), Finnic words by Nirvi himself. The words are as follows:

112 113 SOFIA BJÖRKLÖF MUTUAL CONTACTS AND LEXICAL RELATIONS AMONG THE FINNIC VARIETIES OF WESTERN INGRIA AND NORTHEASTERN ESTONIA

Ing. järkü ‘bench, stool’, !kalan|marja ‘caviar’, karjusti ‘herds­ 7.1.3 Research on Estonian man’, •kuhja ‘haystack; heap’, *kuhje ‘id.’, kǟvü ‘spool’, köhä ‘cough’, !köhähellä v., !köhähtäissä ‘to cough once’, !köhähtǟ ‘to In her doctoral dissertation, Tiina Söderman (1996) studied 698 cough once’, lauda ‘board, plan k; table’, !lauta|hurstˊi ‘table­cloth’, words found in at least three of the parishes of the Estonian north­ luikkoi ‘swan (Cygnus)’, •lännikkö ‘wooden tub’, !mäŋgü ‘display eastern coastal area not attested anywhere else in the Estonian lan­ (of birds)’, mäŋgüdellä ‘to learn to play sth’, (*)mäŋgüttǟ ‘id.’, guage area. The data source is VMS. The exact subdialects in the study •piho ‘hallow of the hand, palm; a handful of linen’, !(*)ristetti are the Northeastern Coastal dialects of Jõelähtme, Kuusalu, Haljala, ‘godfather’, !*rist|ätti ‘id.’, ruōja ‘dirt, rubbish’, !(*)ruōjahussa ‘to get dirty’, !ruōjakas ‘dirty’, (*)ruōjakkaine ‘id.’, !ruōjata ‘to make Viru-Nigula, Lüganuse, Jõhvi, Iisaku, and Vaivara, complemented dirty’, !(*)ruōjattua ‘to get dirty’, ruōma ‘trace (in harness)’, with the Central dialect of Jõelähtme, Kuusalu, Haljala, Iisaku, Kad­ rüppü ‘lap, hem’, sōja ‘warm’, !suōja|paita ‘sweater, cardigan’, rina, and Rakvere (see Map 5). Söderman (1996: 39, 157) gives one !(*)suōja|sāppād ‘shoes made of felt’, !(*)suōja|sǟ ‘warm weather Estonian (NE) dialect word, ilama, with the donor language being ei­ in the wintertime’, !suōjennella ‘to warm yourself up’, !suōjentā ther Ingrian or Vote. According to her (ibid.), in the eastern part of the ‘to warm up’, !sǟri|marja ‘calf’, •vāttia ‘to look’, väljǟz ‘outside’, area and especially in Vaivara, it is not always clear if the Estonian *väljǟ ‘away’, !ädidöin ‘father­less’, ätti ‘father’ (all < Est./Vo.; dialect words in question are borrowed from Finnish or Ingrian/Vote. words with a bullet in front of them are primarily borrowed from She gives eleven examples of these words: Vote). Est. (NE) akkun ‘window’, kaarn(a) ‘pine bark; float of a net; Ing. •ahteri ‘stern (of a ship)’, •*ahter|speili ‘id.’, •*ahteri|seili wart’, karsin ‘pen for pigs and calves’, kiista ‘race’, kimbur ‘bit­ ‘spanker, driver (the last sail of a ship)’, •*ahteri|tammes ‘stern ter, sour (e.g., milk)’, kokk(a) ‘hook, fork’, mutkas ‘playful, trick’, log of a ship’, •*ahteri|toppi ‘top stern sail’, •aŋkkuri ‘anchor’, parv ‘choir loft’, peugal ‘thumb’, pulika(s) ‘wooden plug, little •hālada ‘to hoist sails, to draw on land (a ship, for example); to stick’, raadu ~ raado ‘miserable; thin, bad creature, loathsome’ gather sth’, •jahti ‘ship or boat with one mast’, kurssi ‘direction; (all < Fi./Ing./Vo.). course’ (Nirvi: < Fi., comp. Rus.), !ledidä v. ‘of plaiting warp’, !ledittǟ ‘to plait’, letti ‘plait made when warping fabric; plait’, In fact, in the word articles, Söderman (op. cit., s.v.) has considered all !letti|mado ‘“plait of worms”, a lot of worms together’, !lettiä ‘to eleven of these words to be likely Finnish loanwords. All of the exam­ plait’, līkki ‘reinforcement of the outer edge of a sail’, mamma ples are not actually even attested in Vote and Söderman also does not ‘mother’, nōtti ‘note’, parkki ‘willow bark’, !parkkia ‘to bark hide take the extent of the distribution in Ingrian and Vote into considera­ into leather’, •pirkkeli ‘a piece of wood with which the sail of tion. However, in my opinion, some of the distributions are found over a ship is tautened’, posti ‘post office; postman’, puttēli ‘bottle’, such a limited range that there is no question about the donor variety. !putˊeli|kanta s., rākki ‘carcase, scrap, wreck’, !rākkia ‘to spoil Kaja Toikka (2003) studied the vocabulary of the eastern part by hitting; to maltreat’, rāmi ‘frame’, •rīkki ‘state’, •riŋki ‘circle’, of the northeastern coastal area of Estonia in her master’s thesis. Her •vint(t)eli ‘gimlet, auger’, •*vintelin ‘id.’, !•(*)vokka|seili ‘fore­ data consist of 432 dialect words from VMS all attested in two of sail’, ülöz ‘up’, *ülesse ‘id.’, *ülez ‘id.’ (all < Fi./Est.; words with a bullet in front of them are primarily borrowed from Estonian. the following parishes, Viru-Nigula, Lüganuse, Jõhvi, Iisaku, and Vai­ The words rāmi and vokkaseili may also have been borrowed vara (see Map 5), but nowhere else in the Estonian language area. from Russian.) Toikka (2003: 21–113) gives approximately 40 Estonian Northeastern Coastal dialect words in two parishes with a distribution in Vote and Ingrian as well. However, for example, for rengi ~ trengi ~ renki she gives no attestation in Ingrian, although this word is widely known in

114 115 SOFIA BJÖRKLÖF MUTUAL CONTACTS AND LEXICAL RELATIONS AMONG THE FINNIC VARIETIES OF WESTERN INGRIA AND NORTHEASTERN ESTONIA

Ingrian (IMS 473, 597). For most words, she assumes they are shared Ariste notes that there are several loans from Finnish, Ingrian, and among the different Finnic languages and descend from the common Vote in the Estonian dialect of Vaivara. However, in his three differ­ protolanguage. Additionally, there are Russian loans found in all of ent works from three different decades, the only examples he pro­ these varieties; however, these are not discussed in this article. The vides of the numerous loanwords he had mentioned are these seven following words were analysed by Toikka: words.11 This subject matter will be examined more carefully in the Est. (NE) ahas ‘narrow, tight’, ?alenema ‘to go down’, aruma ‘to future. Nevertheless, having examined a larger lexical corpus, it can rake’, arvutus ‘riddle’, eru ‘apart’, iha ‘sleeve’ (< Vo.), ?ilgu ‘totally’, be reported that there certainly is a large amount of words borrowed issunda(i)ne ‘young plant’, jurmakas ‘unable to comprehend’, from Finnish as well as from either Finnish or Ingrian in the Estonian ?kangane ‘of linen’, kardin ‘pen for animals’, karisema ‘to drop, dialect in Vaivara. to fall’, kehnu ‘shabby, poor’, kuja ‘street’, kukk ‘flower’, kutkuma? ‘tingle’, lamuma ‘to lie (on sth)’, lantu ‘rutabaga, swede (Brassica 7.2. Analysed examples of mutual borrowing napus)’, ?lebemed ‘cardfuls (of wool)’, lekku ‘open fire’, levitus? ‘layer of straw’, liemina ‘lambwool’, liugupäev ‘Shrove Tuesday’, In this article, five of the words mentioned above are discussed: treh- ?liugutama ‘to slide down a hill’, ?moningu(i)ne­ ‘some kind of, to vata from Lauerma (1993), hulkkua and polle from VKJo, ilama some (extent)’, ?muhverdama ‘to devour’, m u m m o ­k a ( i ) n e ‘bug’, from Söderman (1996), and roju (s.v. raju in this article) from Toikka mögetama ‘to bleat’, panetamma ‘to slander’, peli ‘game’, ?pigar (2003). The other words discussed in the following sections, heikko, ‘snaps’, ?pumbukene ‘water lily’, raagus ‘raw, rare’, ?raaked ‘rem­ kiuru, raju, and höyry, are words I have otherwise come across when nant’, roju ‘storm’, siint ‘from here’, takjaine­ ~ takkiaine ‘burdock’, tohestama ‘to open (an abscess)’, tolked ‘tuft, something hanging’, conducting my research. These words are chosen because they are vüht(i) ‘skein’, ümmargaine ~ ümmärgäine ‘round’. methodologically illustrative examples of mutual borrowing. Some of the etymologies are totally new. The rest of the words given above will In comparison, the master’s thesis of the present author (Björklöf be analysed in my forthcoming doctoral dissertation. 2012) deals with the vocabulary in the western subgroup of the Es­ In the next sections, the words are presented according to the fol­ tonian Northeastern Coastal dialects. The data consist of words at­ lowing principles. The variant of the words chosen for the numbered tested in at least two of the following parishes: Jõelähtme, Kuusalu, headings corresponds to their appearance in the donor variety; only and Haljala (see Map 5) but nowhere else in the Estonian language one variant is given. For example, the first heading is “EstN, EstS area. The data source is VMS. These three studies of Estonian dialect trehvama ‘to meet accidentally; to happen’ > Vo., Ing.”, because this vocabulary show clearly that influence of Vote and Ingrian is appar­ Estonian word has been borrowed into Vote and Ingrian. The words ent in the eastern part of the northeastern coastal area, but at the same are discussed in an order of increasing complexity. For example, the time, no influence of Vote or Ingrian exists in the western area. (See first Estonian loanword I discuss is trehvama, because its etymology also Pajusalu et al. 2009: 154.) is the most straightforward, I continue with põll, and conclude the In his articles, Paul Ariste (1962: 16–17; 1977: 9; 1981: 57) section with hulkuma, because its relationship with other Finnic varie­ points out seven loans in the Estonian Vaivara dialect (see Map 5) ties is more complicated. In order to make the text easier to follow, spoken on Estonia’s eastern border. These words may have been bor­ variants drawn from different Finnic varieties as well as comments rowed from any of the other Finnic languages of Ingria: on their areal distributions are given at the beginning of each word Est. (Vai) iha ‘sleeve’, kukkane ‘flower’, nauhha ‘ribbon’, paita section. The order of presenting the words is Vote, Ingrian, Estonian, ‘shirt’, papu ‘bean’, puikko ‘pin’, süksü ‘autumn’ (all < Fi./Ing./Vo). 11. Constructing hypotheses on language contacts in this way but not providing substantial evidence for them is, in fact, common to Ariste (see Junttila 2015: 26).

116 117 SOFIA BJÖRKLÖF MUTUAL CONTACTS AND LEXICAL RELATIONS AMONG THE FINNIC VARIETIES OF WESTERN INGRIA AND NORTHEASTERN ESTONIA and Finnish, followed by possible variants in other Finnic languages indicator the word is an Estonian loan in both Vote and Ingrian. In (Karelian, Lude, Veps, and Livonian). When quoting dictionaries, the Vote, the word is attested in the Vaipooli area and Kukkuzi, which are explanations of meanings given in Finnish and Estonian have been the typical areas most likely to have taken loanwords. Unfortunate­ translated into English; however, some German explanations have ly, only a few words are this clear and straightforward. Ariste (1987: been left untranslated in the text. Some Jõgõperä and Kukkuzi variants 23, 29) has indicated that his Vote-speaking language consultants use of Vote are given with different orthographies due to different prac­ words of Estonian origin when talking about their connections with tices in dictionaries. The dictionary sources are marked when giving Estonians, the word in question being one of the examples; Lauerma the words for the first time (see the list of sources of data at the end of (1993: 178) has also compared these three words. The second Finnish the article; all of them have been searched for these words). etymological dictionary, Suomen sanojen alkuperä (SSA), shares my view on the Estonian origin of this word in Vote and Ingrian (however, 7.2.1. Estonian loanwords in Vote and Ingrian SSA presents some peculiar forms of the words in question that are not found in other dictionaries: Vo. trehve̮ ta, Ing. trehvata12; the Vote Estonian loans in both Vote and Ingrian can be relatively easy to iden­ form is also given in SKES). In the first Finnish etymological diction­ tify because they usually are words not attested in the northern Finnic ary, Suomen kielen etymologinen sanakirja (SKES), it was suggested varieties, in (Ingrian) Finnish, for example, and in Ingrian, of course. that the word in Vote and the word trehvata used in so-called Estonian The areal distributions in the investigated varieties give us clear evi­ Ingria (“Viron Ink.”) – the area ceded to Estonia in 1920 – have been dence for the direction of diffusion – especially when it comes to Vote, borrowed from Estonian, German being the donor language of the Es­ another southern Finnic variety. tonian word. In SSA, the term “Viron Ink.” used in SKES has been interpreted to stand for a dialect area of Finnish – which would mean 1) EstN, EstS trehvama ‘to meet accidentally; to happen’ approximately the dialects of Kallivere, Kosemkina, and Narvusi – > Vo., Ing. although the term is not explained in SKES (SSA uses “InkVi” for Vo. trehvata Li Ra (Lu) ~ trehvõt J-Tsv, pr. trehvaan Li Ra J; trehvaa Estonian Ingria). In any case, nowadays this word does not seem to (P), pr. trehvaab Ku ‘to happen’, trehvama, trehve̮ t J-Tsv ‘id.’, be known in the Finnish dialect of Ingria; at least none of Muslimov’s trehvāb Ku ‘id.’ (VK; VKJo; VKKu) language consultants know this word (Mehmet Muslimov, p.c. 16 & Ing. trehvada Me ‘to happen; to meet (sth)’ (IMS) 17 October 2015). In (mainland) Finnish, the same word has been bor­ EstN, EstS trehvama ~ (t)rähvama ~ rehvama widespread ‘to meet rowed separately from Swedish: Fi. treffata ~ trehvata ~ rehvata ‘to someone (by accident); to happen; to meet’ (VMS) (< Germ. treffen) meet’ < Swe. träffa ‘to meet; to happen’ (SSA). Also according to the Fi. treffata LounSm EPohjanm, areally Häme Kpohjanm PSm ~ Estonian etymological dictionary EES, Vo. trehvata ~ trehvaa ‘to hap­ trehvata areally Häme dialects ~ rehvata areally Häme and Savo pen’ and Ing. trehvada ‘to happen’ are borrowed from Estonian. dialects; Peräp ‘to meet’ < Swe. träffa ‘to meet; to happen’ (SSA)

The verb ‘to meet accidentally; to happen’ is attested in three dialects of Vote and one dialect of Ingrian, in fact only in a single Ingrian par­ ish: Vo. trehvata ~ trehvõt ~ trehvaa, pr. trehvaab ‘to happen’, Ing. trehvada ‘to happen; to meet (sth)’. In Estonian, the German loan­ verb (treffen >) Est. trehvama ~ (t)rähvama ~ rehvama ‘to meet some­ 12. The source for the Ingrian word is Sovijärvi (1944: 39), who actually gives the one (by accident); to happen; to meet’ is widespread, which is a clear form trehvada.

118 119 SOFIA BJÖRKLÖF MUTUAL CONTACTS AND LEXICAL RELATIONS AMONG THE FINNIC VARIETIES OF WESTERN INGRIA AND NORTHEASTERN ESTONIA

2) EstN, EstS põll ‘apron’ even though Äimä (1901: 33) and Ojansuu (1916: 173) had already > Vo., Ing., Fi. written that Fi. polle in Tytärsaari is an Estonian loan. In VKJo there Vo. polle Lu Li Ra J-Tsv ~ pollõ Lu J (R-Eur), g. pollee Lu Ra J ~ pollõõ are doubts about the origin of the word and Vo. polle in Jõgõperä has R Lu ‘apron’ (VK; VKJo) been considered an uncertain Ingrian or Estonian loanword, which is Ing. polle Re Sa Vi (Paulaharju) Hev Kur Uus Vää Ro ~ polli Hev indicated with a question mark. In IMS, Ing. polle is compared only 15 ‘apron’ (IMS; IKHe) with the Estonian word põll. SSA and EES consider the direction 16 EstN, EstS põll ~ poll ~ põlles (gen. põlle ~ polle) widespread ‘apron’ of borrowing to be Fi., Ing., Vo. < Est., quite like I do, while Posti (VMS); in the Northeastern Coastal dialects poll Jõe Kuu13, polle (1970) considered it to be Ing., Vo. < Est. SKES states that the Esto­ VNg Vai, põll Lüg Jõh (EMSkk) nian dialect word in Vaivara would be polle and in Haljala poll, which Fi. polle Kanteletar, Gulf of Finland islands, Ink ‘apron’ (SSA) (at least is likely because the dialects in question do not have the midvowel /õ/. Tyt Lava Tyrö Mol Nov Kap Kat Kose14) According to the Word register of the Dictionary of Estonian dialects (EMSkk), the variant poll is attested in Jõelähtme and Kuusalu, polle The same word for ‘apron’ is attested in four Finnic varieties. In in Viru-Nigula and Vaivara (see also footnote 13). The background of Vote, polle ~ pollõ ‘apron’ is attested only in the Vaipooli area and in the Estonian word has been analysed in more detail in articles by Posti Ingrian, polle ~ polli ‘id.’ in three dialect areas out of four. In Estonian, (1970) and Oja (2005). the word põll ~ poll ~ põlles (gen. põlle ~ polle) ‘id.’ is widespread and it has several phonological variants. In Finnish, polle ‘id.’ is at­ 3) EstN, EstS hulkuma ‘to wander around’ tested only in Ingrian Finnish dialects in parishes where contacts with > Vo., Ing., Fi. speakers of Estonian and Ingrian have emerged. Consequently, it is Vo. ulkkua Kett. M J (K-Ahl. R-Lön. P) ~ hulkkua Lu Li J (Ku) ~ an Estonian loanword in Finnish, in Vote, and most likely in Ingrian xulkkua J ~ hulkkuag I ‘to wander around; to walk’, huлkkuma, as well, because the southern Finnic variety Estonian and the north­ huлkkua J-Tsv ‘id.’, huлkkua ~ xuлkkua Ku ‘id.’, huлkkuag Ma ‘id.’ ern Finnic variety Ingrian do not share words exclusive to them. The (VK; VKJo; VKKu; VKMa) phonetic systems of Ingrian and Finnish lack the midvowel /õ/ which Ing. hulkkua Me Sa L-Suu Ro ‘to wander around’ (IMS) has been substituted in these Estonian loans with the nearest vowel /o/. EstN hulkuma ~ (h)ulkma ~ ulkama widespread ‘to wander around’ The absence of initial syllable /õ/ in the variants polle ~ pollõ in Vote (VMS) refers to a loan origin: in this case the word might have been borrowed EstS hulkuma ~ hulkma ~ ulkima ‘to wander around’ (VMS) via speakers of Ingrian and Finnish if not from the Estonian variants Fi. hulkkua Lönnrot 187417; Ink (only Nar18 Kall), hulkutella Eastern poll ~ polle. The non-initial syllable /õ/ in the variant pollõ in Vote, dialects, Kall in Ink, hulkkia Uusikirkko Vpl., hulkehtia Ilomantsi on the other hand, is typical for Vote: it is an accommodation to the ‘to wander around / umherschlendern’ (SSA; SMS) variety’s own structure, also known as etymological nativisation (on Li. *ul̄k̀ pr. ul̄k̆ k ə̑ b SjW ’herumziehen, herumstreichen / to wander this phenomenon, see A. Aikio 2007). Lauerma (1993: 186) compares around’ (LW) the word in Vote only with Ingrian and with the Finnish of Kosemkina 15. According to EES, even the Estonian Swedish word päll ‘apron’ is a loanword 13. According to Posti (1970: 470), the variant poll would be known in Haljala, but in from Estonian. the Word register of the Dictionary of Estonian dialects (EMSkk) there is no such slip. 16. SSA records the word in Vote only from Jõgõperä (Tsvetkov). 14. According to Muslimov (2015), Fi. polle is known in the Finnish dialect of Mo­ 17. A name (and a year) given in SSA stands for the first occurrence of the word loskovitsa and also in the Finnish dialect of Tyrö, Novasolkka, Kaprio, and Kattila in written Finnish. (Mehmet Muslimov, p.c. 16 October 2015). I have found its use noted in different 18. According to Muslimov, Fi. hulkkua is known in the Finnish dialect of Narvusi sources from Tytärsaari, Lavansaari, and Kosemkina. and Kosemkina (Mehmet Muslimov, p.c. 16 October 2015).

120 121 SOFIA BJÖRKLÖF MUTUAL CONTACTS AND LEXICAL RELATIONS AMONG THE FINNIC VARIETIES OF WESTERN INGRIA AND NORTHEASTERN ESTONIA

The verb ‘to wander around’ is attested in all southern Finnic varieties, Kettunen) and that its “original” meaning in Finnish would be “höls­ namely Vote, North Estonian, South Estonian, and Livonian. It is found kyä”, ‘to swash, slosh, slop about’. However, it is more likely that the in three dialect areas in Vote: ulkkua ~ hulkkua ~ xulkkua ~ huлkkuag word is a homonym. EES19 follows SSA and considers the stem to be a ‘to wander around; to walk’, in Estonian hulkuma ~ (h)ulkma ~ ulki- Finnic stem and does not indicate any mutual borrowing. ma ~ ulkama is widespread (see above for the division of the North and South Estonian variants), and in Livonian *ul̄k̀ pr. ul̄k̆ k ə̑ b ‘herumziehen, 7.2.2. Finnish loanwords in Vote, Ingrian, and Estonian herumstreichen / to wander around’ was originally given in the Livisch- deutsches und deutsch-livisches Wörterbuch (‘Livonian-German and Etymologising loanwords in Vote, Ingrian, and Estonian with Finnish German-Livonian dictionary’) by Sjögren and Wiedemann from 1861. as the possible donor variety can be difficult due to lack of materials. The southern distribution, as there is a cognate also in Livonian, sug­ Even though the Finnish dialect collection in the Archive of the Dic­ gests this might be an inherited word also in Vote, as does its presence in tionary of Finnish dialects (SMSA) is considerable, the materials on two eastern Vote villages, Itšäpäivä and Mahu. However, taking a closer Ingrian Finnish dialects are unfortunately inadequate20 (see Elomaa look at the Livonian word with only one attestation raises the question 1989: footnote 2, 37–39; Lauerma 1993: 161). However, when a word of a more probable Estonian loan origin and Kettunen in LW in fact has is widespread in mainland Finnish, there is little reason to assume it compared the word with Estonian hulkuda. In addition, the word-initial would not be known in the Ingrian Finnish dialects. In such cases it /h/ has disappeared in the development of Vote and therefore is a sign of is reasonably safe to suppose the word in question was part of the a loan origin. The variants lacking /h/ appear only in older data, which Ingrian Finnish vocabulary as well. As for words with a limited distri­ are not the most accurate in their marking of the linguistic material and bution in Finnish, the lack of comparative data makes it impossible to therefore might not be the most reliable. In these cases, etymological elaborate on their etymology. nativisation is also possible (see A. Aikio 2007). In Ingrian and Finnish, the word is a loan from Estonian: Ing. hulkkua ‘to wander around’ in 4) Fi. heikko ‘weak’ two dialect areas and Fi. hulkkua ‘to wander around / umherschlendern’ > Vo., Ing., EstN (a totally new etymology) in two parishes, although the Ingrian word might also have been medi­ Vo. heikko K-Al. M-Set. J-Tsv. Ku ~ eikko K-Ahl., g. heikoo J ‘weak; ated via Vote and the Finnish word via Ingrian or Vote. In Finnish, the thin; poor’ (VK; VKJo; VKKu) word is attested only in Ingrian Finnish dialects in western Ingria, the Ing. heikko Me Re Sa Sä ‘weak’ (IMS) contact area with other Finnic varieties. Laanest (1964: 31) and also EstN (h)eik Kuu ‘weak’ ~ eiko Vai ‘fussy, mercurial’ (EMS) Krohn (1901: 238) have considered the Ingrian word (Krohn ibid.: in Fi. heikko Agricola; widespread ~ heiko Satak Häme EPohjanm Peräp western Ingria) to be a probable Estonian loanword and Savijärvi (1998: Länsip ~ heikeä LönnrotLis 1886; dial. ‘weak’, heiketä, heikentyä 276; repeated in Savijärvi 2001: 163) has proposed that the word in the ‘to weaken’ (SSA), widespread (SMS) Finnish dialect in western Ingria is “an item from Ingrian, Votian, or Ka. heikko rare ‘weak, poor’ (KKS; SSA) Estonian”. In VKJo Vo. huлkkua ‘to wander around’ is considered either Ve. heikembāńe southern dialects ‘weaker’ (VVS; SVJa; SSA) as an Ingrian or Estonian loan, and in IMS Ing. hulkkua is compared with Est. hulkuda. Ahlqvist (1856: 157) considers the words in Vote and 19. According to EES, even the Estonian Swedish word holka ‘to wander around’ Estonian to be equivalents and strangely enough, SSA considers all the is a loanword from Estonian. words to be equivalents. Ojansuu (1916: 138–139) considered this word 20. According to Jussila (ed., 1991), in SMSA, Archive of the Dictionary of Finnish dialects, the number of word slips from western Ingria are as follows: Moloskovitsa 469, in Finnish – with the meaning ‘to wander around; Fi. kuljeksia’ – to be Novasolkka 0, Hevaa 0, Kaprio 0 (however, there is one example from Kaprio in SMS, a literary loan from Estonian (borrowed in the 20th century by Lauri s.v. järven|emä), Soikkola 0, Kattila 38, Narvusi 64, Kosemkina 0, Kallivere 35 573.

122 123 SOFIA BJÖRKLÖF MUTUAL CONTACTS AND LEXICAL RELATIONS AMONG THE FINNIC VARIETIES OF WESTERN INGRIA AND NORTHEASTERN ESTONIA

The word for ‘weak’ is attested in three dialect areas in Vote and in one 5) Fi. kiuru ‘Eurasian skylark, Alauda arvensis’ dialect area in Ingrian and Estonian: Vo. heikko ~ eikko ‘weak; thin; > Vo., Ing. (a new etymology), EstN poor’ in Kattila (the materials are old), the Vaipooli area, and Kukkuzi; Vo. kiuru Kett. K P M Kõ S Lu Ra J I ~ kˊiuru J-Tsv. ~ tšiuru K-Ahl. Ing. heikko ‘weak’ in Soikkola, Est. (h)eik ~ eiko ‘weak; fussy, mercu­ Li (M) ~ Кïуро Tum. ‘Eurasian skylark’, kiuru ‘id.’ J-Tsv Ku, rial’ in the Northeastern Coastal dialects. The limited distributions re­ kiurukkõin JTsv. J dem. ‘id.’ (kiurukke̮ in J-Tsv), kiurulintu M ‘id’; fer to a loan origin from Finnish, where the word heikko is widespread kiuruttaa JTsv. J ‘to trill, to warble’, kiurutõlla JTsv. J freq. ← and has phonological variants as well as verbal derivations (heiko, kiuruttaa; kiurutte̮ ma, kiurutta, kiuruttē̮ ma, kiurute̮ л̄ J-Tsv ‘to trill, heikeä ‘weak’; heiketä, heikentyä ‘to weaken’). The word-initial /h/ to warble’ (VK; VKJo; VKKu) in Vote is an indication of a loan origin because historically it has Ing. kiuru Re Sa Ro ‘Eurasian skylark’ (IMS) disappeared in Vote and, additionally, at least eleven other words with EstN kiur Hlj VNg Lüg Jõh IisR Vai Iis Kod ~ kiuro Vai ‘Eurasian the meaning ‘weak’ have been attested in Vote according to VK. The skylark’; kiurama Hlj, tsiuruma IisR ‘to trill, to warble’ (VMS; Estonian dialect words have in fact been borrowed from two differ­ EMS) ent directions: Kuusalu, a parish located east of Tallinn, most likely Fi. kiuru Juslenius 1745; mostly E dial.21 ‘Eurasian skylark’, k i u r ( u ) i ­n e n , received the word across the Gulf of Finland from either the coastal kiuri SE dial. ‘Eurasian skylark’ (SSA); probably partially literary; area of Kymenlaakso or the Karelian Isthmus (about the borrowing vernacular especially in Kar Ink (SMS) areas, see Björklöf 2012; 2017; 2018), while Vaivara on Estonia’s Ka. kiuru ~ kˊiuru widespread, kiurinki Paatene ‘Eurasian skylark’; eastern border borrowed it from Ingrian Finnish. The Karelian rare kiuruta Paatene ‘to trill, to warble’ (KKS; SSA) heikko ‘weak, poor’ in two dialect areas out of three is also a Finn­ Lu. kiur Sn ~ kiuroi B ~ kˊiuru Lm ~ kiuruińe B ‘Eurasian skylark’; ish loan according to SSA. In the southern dialects of Veps only the kiuruta Sn ~ kiuruttada B ‘to sing like an E. skylark’ (; SSA) comparative heikembāńe ‘weaker’ is attested, which seems peculiar, but there actually are strange derivations in Veps without an attested The name of the bird ‘Eurasian skylark’ in Vote, Ingrian, and Esto­ stem, -Āne in fact being one of the derivational types. Lauerma (1993: nian is most likely a Finnish loan: Vo. kiuru ~ kˊiuru ~ tšiuru ~ Кïуро 165), oddly enough, compares the word in Vote only with Ingrian but ‘Eurasian skylark’ (attested also in Eastern Vote), Ing. kiuru ‘id.’ (in not with Estonian or even with Finnish. Ahlqvist (1856: 123), one of two dialect areas, Soikkola and Rosona22), EstN kiur ~ kiuro (NE, E the first linguists studying Vote, considered the words in Finnish and dial.) ‘id.’ < Fi. kiuru (Jusl 1745; mostly E dial.) ‘id.’, kiur(u)inen, kiuri Vote equivalents, which is not the case, because in Vote, there is no (SE dial.) ‘id.’. In Karelian, kiuru ~ kˊiuru is widespread and in Lude, word-initial /h/ and words with /h/ in this position are always borrow­ kiur ~ kiuroi ~ kˊiuru ~ kiuruińe are also attested. In Finnish, this word ings or at least are the result of later influence from other languages. has been regarded either as onomatopoeic (SSA) or a potential Bal­ Surprisingly, SSA also does not take borrowing into consideration and tic loanword (Kalima 1936; SKES; SSA). In studies on Finnish bird instead treats the words as cognates. names, Marttila (2010: 196) follows SSA and classifies the word as on­ omatopoeic or a loan, but Salminen (2010: 352) categorises the Finnish word as an onomatopoeic word. The most recent etymological study (Junttila 2015: 102, 108) regards the possible Baltic etymology given

21. Also in the Finnish dialect of central and northern Ingria (Rääp Sku Kup) (Meh­met Muslimov, p.c. 16 October 2015). 22. Rosonanjoki (Rosona River) dialect area according to J. Mägiste (1925): Ingrian dialects spoken in so-called Estonian Ingria.

124 125 SOFIA BJÖRKLÖF MUTUAL CONTACTS AND LEXICAL RELATIONS AMONG THE FINNIC VARIETIES OF WESTERN INGRIA AND NORTHEASTERN ESTONIA in SKES as controversial due to the way in which the etymology has The phonetic form of the latter Estonian variant in Iisaku, located west been received by the research community. The derivation of the word of the easternmost Estonian parish Vaivara, might have some influence is defined to be irregular and the word is categorised as part of a group from Vote. Ariste (1956: 15), on the other hand, has treated the Estonian of words listed as descriptive, contaminations, or onomatopoeic (ibid.). Northeastern Coastal kiur and the Finnish kiuru as cognates and Mäger The words could be cognates in all of the eastern Finnic varieties (1967: 128) has considered the Estonian word an onomatopoeic word. (Eastern Finnish dialects, Ingrian, Karelian, Lude). However, the lim­ ited distribution in Ingrian suggests a Finnish loan origin. The sound 6) Fi. raju ‘wild; severe, hard’; in compounds meaning ‘storm’ change *k > tš in front of front vowels in itself is the most characteristic > Vo., Ing., EstN (a totally new etymology) feature in Vote, but in the variant tšiuru in Vote it probably is only a Vo. raju P ‘storm’, comp. raju-ilma: Раю-Илма Pal. ‘storm’, raju|ragõ Votism in a loanword, as Lauerma (1993: 167) has also suggested; in M ‘hailing, hailstorm’, raju|vihma L ‘rainstorm’. Also roju Lu Li Ra modern terms it has to be classified as a result of etymological nativisa­ 1. Lu Li ‘storm’ 2. ‘stormy’; rojukaz Lu Li ‘stormy’. (VK) tion (see A. Aikio 2007). Suhonen (1986: 152) gives the words in Vote Ing. raju adj. ‘fierce’ Me usually raju ilma ‘bad weather’ Sa. Also as an example of loans with either Ingrian or Ingrian Finnish origin. roju n.: roju Ro ‘storm’; roju|ilma L-suu ‘rainy, windy, and stormy Kettunen (1930: 17 [1915: 16]) considered the path of borrowing to be weather’; rojugas adj. L-Suu ‘r-, w-y, and s-y (about weather)’; Sa Vo. kiuru < Fi.23 and SSA similarly lists it as Vo. tšiuru < Fi. However, ‘quarrelling’; rojustaa v. Sa ‘on a bad, r-y and s-y autumn; to quarrel’; the latter is the only borrowing of this word between the Finnic vari­ rojuz n. Sa ‘r-y, w-y, and wet autumn; bad-mannered’. (IMS) eties given by SSA. The word in Vote was also mentioned by Setälä EstN raju ~ rajo Lüg Jõh IisR Vai Khk VJg Trm Kod Äks Lai ‘storm’. (1964) as a cognate of the Finnish word. Mägiste (1927) considered the Also roju ‘storm’ VNg Lüg. (VMS) word to be a loan from Vote into Estonian, and the variant siuru (< Vo. Fi. raju Hemminki 1605; widespread ‘wild; severe, hard; (also areally tšiuru) in old folk songs in fact might be one, as Saareste (1952: 62) dial.) large, enormous’, comp. raju|ilma Finno 1583, raju|myrsky (SSA) also writes. Laakso regards Vo. kiuru (in Ku Ma, Setälä’s materials, Ka. raju adj. ‘on wind, rain: hard, heavy; on animals: fierce’ (KKS; SSA) and Kettunen 1915) as a loan from either Finnish, Ingrian, or Estonian in the Reverse Vocabulary of Vote (Laakso ed. 1989). In the reverse The word for ‘storm’ is attested in all three analysed varieties: EstN vocabulary of the Vocabulary of the Joenperä (Jõgõperä) dialect of raju ~ rajo ‘storm’ in northeastern and eastern parishes; Vo. raju ‘id.’ Vote (VKJo, Laakso ed. 1995), the nouns Vo. kiuru, kiurukke̮ in in the in one village, compounds raju|ragõ ‘hailing, hailstorm’, raju|vihma Jõgõperä dialect are considered loans from Ingrian and the verbs Vo. ‘rainstorm’ both in one village; Ing. raju (adjective according to IMS) kiurutta, kiurute̮ лл in Jõgõperä are regarded as having been borrowed ‘fierce’, usually raju ilma ‘bad weather’ in two villages. In Finnish, from either Ingrian, Finnish, or Estonian, although there are no such the word is widespread: Fi. raju ‘wild; severe, hard; (also areally Ingrian or Finnish verbs in the dictionaries. In Vote, Estonian, and also dial.) large, enormous’, and it is attested also in compounds raju|ilma, Lude, verbs used to describe the singing of this bird have even been raju|myrsky. In Vote, Ingrian, and Estonian, raju has to be considered attested in a couple of areas in each variety: Lu. kiuruta, kiuruttada ‘to a Finnish loan because of its wide distribution in Finnish and limited sing like an E. skylark’; Vo. kiuruttaa; kiurutõlla (freq.) (JTsv. J) ‘to distribution in the recipient varieties – and also due to its semantics: in trill, to warble’; Est. kiurama (Hlj), tsiuruma (IisR) ‘to trill, to warble’. Finnish, this word has more meanings, in the recipient varieties it only 23. In fact, Kettunen (1930: 17 [1915: 16]) writes: “Words like these [with k before a refers to a storm (in Ingrian, it is used only as an adjective). The Esto­ ] are borrowings from other Finnish dialects of Ingria - -” (my translation nian attestation in the westernmost parish of Kihelkonna on the island and italicisation): in the context of the time of writing in question (and considering the of Saaremaa is most likely also a Finnish loanword, because this is an­ pronoun other), this might be interpreted to mean other Finnic languages or varieties of the area. In SSA it simply has been interpreted to mean dialects of Finnish. other contact area with Finnish. In SKES, the Finnish, Karelian (in two

126 127 SOFIA BJÖRKLÖF MUTUAL CONTACTS AND LEXICAL RELATIONS AMONG THE FINNIC VARIETIES OF WESTERN INGRIA AND NORTHEASTERN ESTONIA dialect areas out of three), and Estonian words are considered to be The word for ‘steam’ is attested in all eastern Finnic varieties. Fi. equivalents, which I doubt, because an old word in North Estonian also höyry is widespread, except for in the Southwestern dialects and would be found in South Estonian. SSA does not give the variants in Länsi-Uusimaa, Ingrian höürü ‘fog; steam’ is known in two villages, Vote and does not take any borrowing into account, either, but consid­ while other attested variants include Karelian höyry ‘steam, vapour, ers the Finnish, Ingrian, Karelian, and Estonian variants cognates. EES fog’ (in all three dialect groups; ‘dispute’ in one group), Lude höür(ü) simply considers the stem to be a Finnic stem, while the distributions ‘water vapour, steam’, and Veps höür ‘steam’. This word has been are not taken into account at all. A word with /o/ in the initial syllable borrowed into Vote as höürü ‘steam’ in two villages in the Vaipooli and a couple of derivatives of it are known as well, attested only in area and in Kukkuzi and into Estonian as öür ~ öürü ~ öur ‘steam’ in Vote in the Vaipooli area, in Lower Luga and Soikkola Ingrian, and in a couple of northeastern parishes. I agree with SSA, which consid­ Estonian in two northeastern parishes: Vo. roju ‘storm; stormy’, roju- ers these both to be Finnish borrowings. However, SSA gives EstNE kaz ‘stormy’; Ing. rojuilma ‘rainy, windy, and stormy weather; storm’, forms öür (g. öüru), öüri, öürü, of which öüri is not known in VKS; rojugas adj. ‘rainy, windy, and stormy (about weather)’, rojustaa v. none of these forms are known in Wiedemann (1893 [1869]), which ‘on a bad, rainy and stormy autumn’, rojuz n. ‘rainy, windy, and wet usually is the source of Estonian dialect words for SSA. In VKJo, the autumn’; EstN roju ‘storm’. These could be considered phonological noun Vo. höürü in the Jõgõperä dialect is considered a loan from either variants. Toikka (2003: 83) simply writes that the word roju is a shared Ingrian or Finnish. As the form generally used in Estonian is aur, the word in Vote, Ingrian, and the Estonian Northeastern Coastal dialects. Estonian variant õur in Lüganuse parish might be a contaminated form in a dialect area (Alutaguse) where the midvowel /õ/ is especially pro­ 7) Fi. höyry ‘steam’ ductive and usually replaces the vowel /o/. In Finnish, the verb höyrytä > Vo., Ing. (a new etymology), EstN ‘to steam’ is widespread, höyryttää ‘to steam sth’ is sporadic. In Vote, Vo. höürü Lu J-Must. J-Tsv, g. höürüü J ‘steam’, höürü Ku ‘steam; air’; a couple of verbal derivations are known in the same villages as the höürüüssä: höürüssä Lu höürüss J-Tsv ‘to steam’, höürǖssÄ Ku ‘to noun: Vo. höürüssä, höürüss, höürǖssÄ, höürüä ‘to steam’, höürütell steam’, höürüä J-Tsv ‘to steam; to stagger, to totter (with fatigue)’, ‘to steam sth’. In Estonian, adjective öürüne ‘steamy’ and verbal deri­ höürütellä: höürütell J-Tsv ‘to steam sth’ (VK; VKJo; VKKu) vations öürämä ‘to steam’, öüritamma ~ öüritämmä ‘to steam sth’ are Ing. höürü ‘fog’ Me ‘steam’ Sa; höürüssä Me L-suu ‘to steam’, known in a couple of the same northeastern parishes where the noun höürüttää Me ‘to steam’, höürüdellä freq. Me ‘to steam sth’ (IMS) is known. In Ingrian, the distribution of a couple of verbal deriva­ EstN öür ~ öürü ~ öur VNg Lüg Jõh Vai ‘steam’, õur Lüg, aur tions is even more limited, one village in Lower Luga and one village widespread; öürüne ‘steamy’ VNg Vai; öürämä ‘to steam’ Lüg Vai, in Soikkola: Ing. höürüdellä (causative, frequentative) ‘to steam sth’, öüritamma ~ öüritämmä ‘to steam sth’ VNg Lüg (VMS) höürüssä, höürüttää ‘to steam’. The Ingrian word can be considered a Fi. höyry Juslenius 1745; widespread, except for SW dial. and LUus ‘Dampf, loan because of its limited distribution. Dunst’ (SSA); höyry widespread24; höyrytä ~ höyrätä no data from Ink ‘to steam’, höyryttää ~ höyryyttää sporadic ‘to steam sth’ (SMS) 7.2.3. Are there Ingrian loanwords in both Vote and Estonian? Ka. höyry ~ höyŕy ‘steam, vapour, fog; dispute’ (KKS; SSA) Lu. höür (B) Td ~ höürü B N Ph Lm ‘water vapour, steam (e.g., coming When it comes to possible Ingrian loans in Vote and Estonian, the from a machine, breath in cold weather)’ (LS; SSA) donor variety is difficult to define as these words are often found in Ve. höür ‘steam’ (SSA); not in VVS or SVJa Finnish as well and also due to the phonological similarities between 24. Also in the Finnish dialect of central and northern Ingria (Mehmet Muslimov, Ingrian and Finnish. Words borrowed into both Vote and Estonian are p.c. 16 October 2015). usually known in both Ingrian and Finnish, and therefore it is typically

128 129 SOFIA BJÖRKLÖF MUTUAL CONTACTS AND LEXICAL RELATIONS AMONG THE FINNIC VARIETIES OF WESTERN INGRIA AND NORTHEASTERN ESTONIA impossible to tell which variety was the source of the loan. In fact, 8) Vo. ilata ‘to clean up’ both varieties are in a position to have mediated loans in such cases. In > Ing. (a new etymology), EstN, Fi. (a new etymology) principle, almost every Finnish loan in Vote (which is also found in In­ Vo. ilata Al Kett R U L P M Kõ Po Lu Li J (K-Ahl. Ja-Len. V Ra Ku) ~ grian) may have been mediated via Ingrian. The Ingrian influence on ilat J-Tsv ~ ilatag I ‘to clean up, to put away, to cleanse, to arrange; Vote has been strongest in the mixed population villages in Vaipooli to take care of (animals)’, iлama, iлat J-Tsv ‘to clean up’, iлatA Ku (e.g., Lauerma 1993: 50), where Ingrian can be considered the primary ‘to clean up, to cleanse, to clear away’ (VK; VKJo; VKKu) donor variety. Vote has borrowed plenty of words from Ingrian, but as Ing. ilada Sa L-Suu Ro ‘to clean up (slightly), to repair, to cleanse’ of now, I have not found loans in both Vote and Estonian which I could (IMS) say with certainty were borrowed from Ingrian. EstN ilama Jõh Vai ~ ilamaie Lüg ‘to clean up; to take away’ (VMS; EMS) 7.2.4. Are there Vote loanwords in both Ingrian and Estonian? Fi. ilata ‘to clean up, to cleanse’ Kall (SMS)

Although there are some loans from Vote known in Ingrian and pos­ So far, I have managed to find only one probable borrowing from Vote sibly in Estonian as well, loans from Vote found in both Ingrian and into both Ingrian and Estonian: Ing. ilada ‘to clean up (slightly), to Estonian are exceedingly scarce. For comparison, Arvo Laanest repair, to cleanse’ in two villages and Rosona, EstN ilama ~ ilamaie (1988) has studied 490 Vote words from the materials found in the ‘to clean up; to take away’ in a couple of northeastern parishes. In Finnic atlas ALFE, which was being compiled at the time. The main Vote, the same verb is widespread and attested also in Eastern Vote: group consists of Vote-Ingrian-Finnish words. The remainder is com­ Vo. ilata ~ ilatag ‘to clean up, to put away, to cleanse, to arrange; to posed of two equally-sized groups. Laanest divides the words of these take care of (animals)’ and VK also gives many examples of its use. groups into those found in 1) Vote and Finnish (and usually Ingrian Söderman25 (1996: 39, 157) treats the Estonian word as a loan from as well) but not in Estonian and into those found in 2) Vote and Esto­ either Ingrian or Vote (or as derived from the Estonian word ilu ‘beau­ nian but not in Finnish. For example, in the Lower Luga area, Ingrian ty’), but an Ingrian origin is not very plausible when comparing the has a Vote substrate and the local Finnish dialect has both a Vote and distributions. A particular origin for a loanword is more likely when Ingrian substrate (Savijärvi 1998: 274; 2001: 165; Muslimov 2005; its distribution is like that of other borrowings. Thus, we see there are Rozhanskiy & Markus 2014). According to Kuznetsova et al. (2015: many borrowings in the same direction with a similar distribution in 133), all the Finnish varieties of western Ingria, except the dialects in Estonian. In Finnish, the verb ilata ‘to clean up, to cleanse’ is attested Kaprio and Tyrö, have substrate vocabulary from Vote. only in Kallivere, where it is an apparent loanword. The word could Although there is borrowing between Ingrian and Vote, the lack also be a substrate word in at least Ingrian and Estonian. of loanwords from Ingrian and Vote into each other and into Estonian might refer to the lower prestige of these two varieties compared to Estonian. In fact, words with an attestation only in Vote and Estonian might be worth examining more closely, because of the old assump­ tion that these two varieties have had even closer contacts in the past.

25. Söderman also gives a distribution in VNg, but the word attested there is sim­ ply a homonym with the meaning ‘Est. ilastama; to make slimy’.

130 131 SOFIA BJÖRKLÖF MUTUAL CONTACTS AND LEXICAL RELATIONS AMONG THE FINNIC VARIETIES OF WESTERN INGRIA AND NORTHEASTERN ESTONIA

8. Discussion example, with Ing. ennustā ‘to predict’ which is attested in Ingrian only in the Rosona area according to IMS). One might, of course, as­ Areal distribution is a crucial factor for the etymological research of sume that the word has not been registered anywhere else although it borrowings between closely related varieties (see also Björklöf 2018). “in fact” has been known more widely, but such an approach is non- When there are few sound criteria, the importance of distribution is scientific and should be avoided. emphasised. There can be other problems as well: There is a some­ Examining mutual borrowing in western Ingria and northeastern what limited amount of lexical material on the small Finnic variety Estonia shows that the hallmark of Estonian loanwords in Vote and In­ Ingrian (and Vote), although when compared with available Livonian grian is usually their relatively limited distribution. As for the Estonian materials, for example, the data from Vote are substantial (on Ingrian, distributions in the present study, it seems that a limited distribution see, e.g., Grünthal forthcoming: p. 23 in the manuscript). Despite the in the Northeastern Coastal and Eastern dialects usually indicates that size of the collection on Finnish dialects in the Archive of the Diction­ the word is borrowed from the other Finnic varieties spoken in west­ ary of Finnish dialects (SMSA) being impressive even on a worldwide ern Ingria. Whereas, Estonian loanwords in the other Finnic varieties level, the materials on Ingrian Finnish dialects are inadequate, as men­ in western Ingria usually have a more widespread distribution in Esto­ tioned earlier. In western Ingria, lexical material from Ingrian Finnish nian and may also be attested in South Estonian. Lembit Vaba (2014) dialects has been collected almost exclusively from Kallivere. This discusses distribution as an etymological criterion using the example material is worth closer scrutiny, because it contains many loanwords of Estonian, the donor languages being genetically distant languages. from other Finnic varieties of that area. Collecting newer text materi­ As his earlier research (Vaba 1977; 1997; 2015) and the work of Mari als has also revealed interesting facts: according to Savijärvi (1996b: Must (2000) have shown, loanwords borrowed from the same donor 187; 1998: 276), the Finnish dialects of the Narvusi region in west­ language usually have a specific, limited distribution in Estonian dia­ ern Ingria seem to have a considerable amount of lexical items from lects. There are also loanwords with small, punctual distributions that Ingrian, Vote, and Estonian, and words with these forms and mean­ appear almost scattered around. The studies of Söderman (1996) and ings are apparently not found in standard Finnish or the other Finnish Björklöf (2012; 2018) show the same distributional correlation in the dialects (of Ingria). It must be noted that already Merle Leppik – in Finnish loans in Estonian. her thorough survey on the development of the phonological system Finnish, the Finnic variety with the highest prestige in Ingria, has of the Ingrian Finnish Kurkola dialect – wrote that Vote and Ingrian been the donor variety for a large number of borrowings: most of the played a considerable role in the development of the Finnish dialects loanwords in Vote, Ingrian, and Estonian are borrowed from Finnish. in Kurkola and the Lower Luga area (Leppik 1975: 197). A portion of Loans found in both Vote and Estonian often occur not only in Ingrian the lexical items from Ingrian and Vote in the western Ingrian Finnish but also in Finnish. Due to the phonetic similarity of these varieties, dialects can probably be considered a substratum, because a portion it is usually impossible to determine the donor variety. Vote loans in of the Ingrian and Vote speakers there were converted to Lutheran­ Ingrian and Estonian only occur sporadically and these may also result ism and were therefore Fennicised in the 17th century (Krjukov 1993 in part from a substratum. [1987]: 22, 23; Savijärvi 1998: 274; 2001: 165). However, at this point there exists practically no literature on the Overall, when investigating Finnic varieties in this particular methodology for investigating mutual contacts among closely related area, careful attention must be paid to the extent and location of the varieties. Riionheimo (2013) discusses the use of comparison between distribution. We cannot take a distribution in only one or two parishes languages (Fi. kieltenvälinen vertailu) for studying languages in con­ and assume the word in question derives from the protolanguage (sur­ tact, but she is comparing languages, which are used in profoundly prisingly, this has sometimes been done when preparing the SSA, for different ways in education, legislation, and in relation to national

132 133 SOFIA BJÖRKLÖF MUTUAL CONTACTS AND LEXICAL RELATIONS AMONG THE FINNIC VARIETIES OF WESTERN INGRIA AND NORTHEASTERN ESTONIA identity. On the contrary, areal varieties are more typical in the tradi­ Abbreviations tional Finnic linguistic area. When it comes to the Finnic varieties in the Lower Luga area, we can even speak of mixed varieties. General: adj. = adjective, comp. = compound, dem. = diminutive, dial. = Jouko Lindstedt (2000: 242) has proposed that linguistic areas, dialectal/dialects, freq. = frequentative, g. = genitive, n. = noun, pr. = for example the Balkans but also many others, have shown that con­ present tense, v. = verb. vergence may be more than borrowing, because donor and recipient Other languages: Ger. = German, Rus. = Russian, Swe. = Swedish. languages cannot always be clearly distinguished and there is mutual Finnic varieties: Est. = Estonian, EstN = North Estonian, EstS = South reinforcement of change. This might be the case with language ar­ Estonian, Fi. = Finnish, Ing. = Ingrian, Ka. = Karelian, Li. = Livonian, eas formed by genetically related and typologically close varieties as Lu. = Lude, Ve. = Veps, Vo. = Vote. well. In the Balkans there is morphological convergence but hardly Dialects of Vote: Western Vote: 1) Lower Luga (Vaipooli) Vote: J = Jõgõperä, Li = Liivtšülä, Lu = Luuditsa, Ra = Rajo; 2) Central (Kattila) any lexical borrowing because the languages are used as an important Vote: Ja = Jarvikoištšülä, K = Kattila, Ke = Kerstova, Ki = Kikeritsa, means of (self-)identification and they are symbols of group identity Kõ = Kõrvõttula, L = Lempola, M = Mati, Mu = Muukkova, P = Pum­ (Lindstedt 2000: 239). Therefore, this sociolinguistic situation is quite mala, Pi = Pihlaala, Po = Pontizõõ, R = Rudja, S = Savvokkala, U = the opposite of that found in western Ingria. At the same time, this Undova, V = Velikkä. Eastern Vote: I = Itšäpäivä, Ii = Iivanaisi, Kl = counterexample sheds light on the impact of identity issues on lan­ Kliimettina, Ko = Kozlova, Ma = Mahu. Ku = Kukkuzi (dialect area); guage change: the identity of the speech community affects the bor­ Kr = Krevin (dialect area). rowing processes enormously – either on a conscious or a subcon­ Dialects of Ingrian: Ala-L = Lower Luga dialect area: Soi = Soikkola scious level. parish / villages: Har = Harkkola, Kol = Kolkanpää, Lo = Loka, Me = The Finnic varieties are a continuum spoken across a relatively Metsäkylä, Ou = Oussimäki, Re = Repola, Ru = Ruutsia, Sa = Saarove, broad area, in which Ingria forms the core area and the meeting point Sä = Säätinä, Tam = Tammikontu, Tar = Tarinaisi, Uus = Uusikylä, Vi of closely related varieties. The internal integration of the Finnic lan­ = Viistinä, Vää = Väärinoja (a Lutheran village); Kosemkina (= Nar = guage area has been difficult to describe, and in their traditional speak­ Narvusi) parish / villages: Kur = Kurkola, L-suu = Laukaansuu, Vibjä ing areas, the varieties of the present-day national languages Finnish (Vipiä). Hev = Hevaa dialect area: Kap = Kaprio parish / villages: Mur = Murtove, Ve = Vepsä. Ro = Rosonanjoki (Rosona River) dialect area and Estonian have considerably levelled off. In my opinion, the most according to J. Mägiste; Ylä-L = Oredež dialect area. fruitful way of investigating the varieties of western Ingria is to treat Dialects of North Estonian appearing in this study: Northeastern them simultaneously and with a comparative approach. This is be­ Coastal dialects: Jõe = Jõelähtme, Kuu = Kuusalu, Hlj = Haljala, VNg cause the language varieties of this region have never existed in a = Viru-Nigula, Lüg = Lüganuse, Jõh = Jõhvi, lisR = Iisaku (North­ monolingual environment. eastern Coastal), Vai = Vaivara. Eastern dialect: Iis = Iisaku (East­ ern), Trm = Torma, Kod = Kodavere, MMg = Maarja-Magdaleena, Pal = Palamuse, Lai = Laiuse. Central dialect: Kad = Kadrina, Rak = Rakvere, VJg = Viru-Jaagupi, Äks = Äksi. Island dialect: Khk = Kihelkonna. E = Eastern dialect, NE = Northeastern Coastal dialects. Dialects of South Estonian appearing in this study: Tartu dialect: TMr = Tartu-Maarja, KodT = Kodavere (Tartu). Dialects of Finnish appearing in this study: Kymi dialects: Suu = Suur­ saari. Southeastern dialects: South Karelian dialects of Finnish proper (Western): Tyt = Tytärsaari, Lava = Lavansaari, Seis = Seiskari.

134 135 SOFIA BJÖRKLÖF MUTUAL CONTACTS AND LEXICAL RELATIONS AMONG THE FINNIC VARIETIES OF WESTERN INGRIA AND NORTHEASTERN ESTONIA

Ingrian dialects of Finnish proper (Northern)26: HaaK = Haapakan­ Sources of data gas, Rääp = Rääpyvä, Kel = Keltto; (Eastern:) Mark = Markkova, Jär = Järvisaari; (Central:) Lii = Liissilä, Inkere = Inkere, Ven = Venjoki, EMS = Eesti murrete sõnaraamat [Dictionary of Estonian dialects] I–VI (a– Tuut = Tuuteri, HieM = Hietamäki, Tyrö = Tyrö (w), Rop = Ropsu osatama). 1994–2019. Tallinn: EKI. Available at: ‹http://www.eki.ee/ (c/w), Sku = Skuoritsa, Kolp = Kolppana, Spa = Spankkova, Hats = dict/ems/›. Installments 1–30 in PDF format available at: ‹http://www. Hatsina, Kop = Koprina, Kup = Kupanitsa, Ser = Serepetta (c/w); eki.ee/dict/ems/pdf.html› (Western:) Mol = Moloskovitsa, Nov = Novasolkka, Hev = [Hevaa]27, EMSkk = Eesti murdesõnavara koondkartoteek [Word register of the Dic­ Kap = [Kaprio], Soik = [Soikkola], Kat = [Kattila], Nar = [Narvusi], tionary of Estonian dialects]. Archive of Estonian Dialects and Finno- Kose = [Kosemkina], Kall = Kallivere. Dialectal areas appearing in Ugric Languages (AEDFUL) at the Institute of the Estonian Language. this study: Kar = Karjala ‘Karelian dialects of Finnish proper’, Ink = Tallinn: Eesti Keele Instituut. (See (in Estonian and English): ‹http:// Inkeri ‘Ingrian dialects of Finnish proper’. Abbreviations used in SSA emsuka.eki.ee/›) and appearing in this study: EPohjanm = Etelä-Pohjanmaa ‘southern IKHe = Laanest, Arvo 1997: Isuri keele Hevaha murde sõnastik [Vocabu­ Ostrobothnia’, Häme = Häme, KPohjanm = Keski-Pohjanmaa ‘cen­ lary of the Ingrian Hevaa dialect]. Tallinn: EKI. tral Ostrobothnia’, LounSm = Lounais-Suomi ‘southwestern Finland’, IMS = Nirvi, Ruben E[rik] 1971: Inkeroismurteiden sanakirja [Dictionary of LUus = Länsi-Uusimaa ‘western Uusimaa’, Länsip = Länsipohja, Ingrian dialects]. LSFU XVIII. Helsinki: SUS. Peräp = Peräpohjola, PSm = Pohjois-Suomi ‘northern Finland’, Satak = Jussila, Raimo (ed.) 1986: Inkeroismurteiden käänteissanasto [Reverse Satakunta, Savo = Savo. E = Eastern dialects, SE = Southeastern dia­ vocabulary of Ingrian dialects]. R. E. Nirvin Inkeroismurteiden sana­ lects, SW = Southwestern dialects. kirjan aineistosta toim. R. Jussila. LSFU XVIII:2, KKTKJ 41. Hel­ Dialects of Lude appearing in this study: B = Bošinkylä, Lm = Lidˊžmi, sinki: SUS & KKTK. N = Nuomoil, Ph = Pyhäjärvi, Sn = Sununsuu, Td = Tiudia. KKS = Karjalan kielen sanakirja [Dictionary of Karelian] 1–6. 1968–2005. Collectors of the Vote materials appearing in this study: -Ahl. = Ed. Pertti Virtaranta (1–3), Raija Koponen (4–6). LSFU XVI, 3–6 also A. Ahlqvist, -Al. = V. Alava, -Eur. = D. E. D. Europaeus, Kett. = L. KKTKJ 25. Helsinki: SUS & KKTK. Kettunen, -Len = Ja. Ja. Lensu, -Lön. = E. Lönnrot, -Must. = O. A. F. KKSV = Karjalan kielen verkkosanakirja [Online dictionary of Karelian]. Ed. Mustonen, Pal. = P. S. Pallas, -Se. = E. N. Setälä, -Tsv. = D. Tsvetkov, Marja Torikka, editor in chief from 8 September 2010 Leena Joki. Internet Tum. = F. Tumanski (abbreviations are usually given with an abbrevia­ application Jari Vihtari. Kotimaisten kielten tutkimuskeskuksen verkko­ tion of a village, e.g., “K-Ahl” ‘Ahqvist from Kukkuzi village’). julkaisuja 18. Helsinki: KKTK. Published 15 October 2009, updated 22 In sources of data and references: EKI = Eesti Keele Instituut February 2019 [Accessed February 2019]. ‹http://kaino.kotus.fi/kks› [Institute of the Estonian Language]; KKTK = Kotimaisten kielten Laakso, Johanna (ed.) 1989: Vatjan käänteissanasto [Reverse vocabulary of tutkimuskeskus [Institute for the Languages of Finland]; KKTKJ = Vote]. LSFU XXII, KKTKJ 49. Helsinki: SUS & KKTK. Kotimaisten kielten tutkimuskeskuksen julkaisuja [Publications of LS = Lyydiläismurteiden sanakirja [Dictionary of Ludic dialects]. 1944. Ed. the Institute for the Languages of Finland]; LSFU = Lexica Societatis Juho Kujola. LSFU IX. Helsinki: SUS. Fenno-Ugricae; MSFOu = Memoires de la Societe Finno-Ougrienne; LW = Kettunen, Lauri 1938: Livisches Wörterbuch. LSFU V. Helsinki: SUS. p.c. = personal communication; SKS = Suomalaisen Kirjallisuuden SMS = Suomen murteiden sanakirja [Dictionary of Finnish dialects] 1–8 Seura [Finnish Literature Society]; SKST = Suomalaisen Kirjallisuu­ (a–kurvottaa). 1985–2008. KKTKJ 36. Helsinki: Kotimaisten kielten den Seuran toimituksia [Publications of the Finnish Literature Soci­ tutkimuskeskus. ety]; SUS = Suomalais-Ugrilainen Seura [Finno-Ugrian Society]. SMS = Suomen murteiden sanakirja [Dictionary of Finnish dialects] (a– lysmä). 2012–. Kotimaisten kielten keskuksen verkkojulkaisuja 30. 26. The Ingrian Finnish dialects are divided according to Muslimov (2009), fol­ Helsinki: Kotimaisten kielten keskus. Updated 22 May 2019 [Accessed lowed by Kuznetsova et al. (2015: 132–133). February 2015]. ‹http://kaino.kotus.fi/sms› 27. The parishes in [square brackets] are parishes where Finnish was not the main Finnic language.

136 137 SOFIA BJÖRKLÖF MUTUAL CONTACTS AND LEXICAL RELATIONS AMONG THE FINNIC VARIETIES OF WESTERN INGRIA AND NORTHEASTERN ESTONIA

SMSA = Suomen murteiden sana-arkisto [Archive of the Dictionary of References Finnish dialects]. Helsinki: Kotimaisten kielten keskus. (See (in Finn­ ish): ‹http://www.kotus.fi/aineistot/sana-aineistot/suomen_murteiden_ Aikio, Ante 2007: Etymological nativization of loanwords. A case study sana-arkisto›) of Saami and Finnish. – Ida Toivonen & Diane Nelson (eds), Saami SSA = Itkonen, Erkki & Ulla-Maija Kulonen (eds) 1992–2000: Suomen Linguistics. Amsterdam studies in the theory and history of linguistic sanojen alkuperä. Etymologinen sanakirja [Origin of Finnish words. science. Series IV. Current issues in linguistic theory. Volume 288. Etymological dictionary] 1–3. SKST 556, KKTKJ 62. Helsinki: SKS Amsterdam. 17–52. & KKTK. Aikio, Marjut 1988: Saamelaiset kielenvaihdon kierteessä. Kielisosio­ SVJa = Zajceva, M[aria] I. & M[aria] I. Mullonen 1972: Slovarʹ vepsskogo loginen tutkimus viiden saamelaiskylän kielenvaihdosta 1910–1980. jazyka [Dictionary of the Veps language]. Akademija Nauk SSSR. [Saamis in the vortex of language shift. Sociolinguistic study of the Karelʹskij filial. Institut jazyka, literatury i istorii. Leningrad: Nauka. language shift in five Saami villages in 1910–1980. PhD thesis.] SKST VK = Vadja keele sõnaraamat. Vadˊdˊaa tšeelee sõna-tširja. Slovarʹ vodskogo 479. Helsinki: SKS. jazyka [Dictionary of Vote]. 2013. 2., täiendatud ja parandatud trükk. Äimä, Frans 1901: Äänneopillinen tutkimus Tytärsaaren murteesta [A pho­ Ed. Silja Grünberg. Tallinn: EKI. User interface 2019 available at: nological study on the dialect of Tytärsaari]. Vähäisiä kirjelmiä XXXI. ‹http://www.eki.ee/dict/vadja/›, PDF’s of the dictionary available at the Eripainos aikakauskirjasta Suomi (IV:1). Helsinki: SKS. same address. Ahlqvist, Aug[ust] 1856: Wotisk grammatik jemte språkprof och ordförteck- VKJo = Tsvetkov, Dmitri 1995: Vatjan kielen Joenperän murteen sanasto ning [Grammar of Vote with language samples and vocabulary]. Acta [Vocabulary of the Joenperä (Jõgõperä) dialect of Vote (with reverse Societas Scientiarum Fennicae. Tomi qvinti. Fasciculus I. Helsingfor­ vocabulary)]. Toimittanut, käänteissanaston ja hakemiston laatinut siae: Finska Vetenskaps-Societeten. Johanna Laakso. LSFU XXV, KKTKJ 79. Helsinki: SUS & KKTK. Alenius, Kari 2015: Language in education in Estonian Ingria between the VKKu = Posti, Lauri 1980: Vatjan kielen Kukkosin murteen sanakirja [Dic­ World Wars. Presentation given at the Congressus Duodecimus Inter- tionary of the Kukkosi (Kukkuzi) dialect of Vote]. Ainekset kerännyt nationalis Fenno-Ugristarum, Oulu 2015. Abstract in the Book of Lauri Posti. Painokuntoon toimittanut Seppo Suhonen Lauri Postin abstracts. 430. Abstract available at: ‹https://www.oulu.fi/sites/default/ avustamana. LSFU XIX, KKTKJ 8. Helsinki: SUS & KKTK. files/content/CIFU12-BookOfAbstracts_4.pdf› VKMa = Kettunen, Lauri 1986: Vatjan kielen Mahun murteen sanasto ALFE = Atlas Linguarum Fennicarum 1–3. Itämerensuomalainen kielikartasto. [Vocabulary of the Mahu dialect of Vote]. Eds Jarmo Elomaa, Eino Läänemeresoome keeleatlas. Ostseefinnischer Sprachatlas. Lingvističe­ Koponen & Leena Silfverberg. Castrenianumin toimitteita 27. Hel­ skij atlas pribaltijsko-finskix jazykov. 2004–2010. Helsinki: SKS & sinki: Castrenianumin laitokset & SUS. KKTK. Available at: ‹https://avaa.tdata.fi/web/avaa/-/kotus-kielikartastot› VKS = Vadja keele sõnaraamat [Dictionary of Vote] 1–7. 1990–2011. Eds Ariste, Paul 1941: Vadja keelenäiteid [Vote language samples]. Acta et Elna Adler & Merle Leppik (1–4), Silja Grünberg (5–7). Tallinn: EKI. Commentationes Universitatis Tartuensis B 49-6. Tartu: Tartu Riiklik VMS = Väike murdesõnastik [Shorter dialect dictionary (of Estonian)] I– Ülikool. II. 1982, 1989. Ed. Valdek Pall. Tallinn: Valgus. Väike murdesõnastik — 1956: Läänemere keelte kujunemine ja vanem arenemisjärk [The [User interface of the Shorter dialect dictionary of Estonian]. Tallinn: development and the earlier stage of the Finnic languages]. – H[arri] EKI. ‹http://www.eki.ee/dict/vms/› Moora (ed.), Eesti rahva etnilisest ajaloost. Tallinn: Eesti NSV Tea­ VVS = Vepsän verkkosanasto [Online vocabulary of Veps]. 2007. Helsinki: duste Akadeemia Ajaloo Instituut. 5–23. Kotimaisten kielten tutkimuskeskus. [Accessed June 2015.] ‹http:// — 1957: Vatjalaisten nykyisyydestä [About the current situation of Votes]. kaino.kotus.fi/sanat/vepsa/› – Virittäjä 61: 119–124. — 1958: Vadja rahvajutte Mati külast [Vote folk tales from Mati village]. – Keele ja Kirjanduse Instituudi uurimused II. Keelelisi küsimusi. Tallinn: Eesti NSV Teaduste Akadeemia & Eesti Riiklik Kirjastus. 148–166.

138 139 SOFIA BJÖRKLÖF MUTUAL CONTACTS AND LEXICAL RELATIONS AMONG THE FINNIC VARIETIES OF WESTERN INGRIA AND NORTHEASTERN ESTONIA

— 1962: Teiste Läänemere keelte elemente Vaivara murrakuis [Elements Södergård (eds), Ideologi, identitet, intervention. Nordisk dialektologi of other Finnic languages in Vaivara dialects]. – Emakeele Seltsi aasta­ 10. Nordica Helsingiensia 48. Dialektforskning 5. Helsingfors: Helsing­ raamat VIII: 11–18. Tallinn. fors universitet; Nordica vid Finska, finskugriska och nordiska institu­ — 1965a: Vadja keele ja eesti murrete suhetest [On the relations of the tionen. 369–394. Available at: ‹http://hdl.handle.net/10138/224299› Vote language and the Estonian dialects]. – Sõna sõna kõrvale. Paul — 2018: Uusia lainaetymologioita viron koillisrannikon rantamurteen Ariste teaduslikust tegevusest. Emakeele Seltsi toimetised 7. Tallinn. länsiryhmässä [New loan etymologies in the western subgroup of the 106–110. [Published for the first time in German in 1963: Die Bezie­ Estonian Coast dialect of the northeastern coast]. – Sampsa Holopainen hungen des Wotischen zu den estnischen Mundarten. – Congressus & Janne Saarikivi (eds), Περὶ ὀρθότητος ἐτύμων. Uusiutuva uralilainen Internationalis Fenno-Ugristarum I, Budapestini habitus 20.–24. etymologia. Uralica Helsingiensia 11. Helsinki: SUS. 357–487. Avail­ IX.1960. Budapest. 212–215.] able at: ‹https://journal.fi/uralicahelsingiensia/issue/view/uh11/489› — 1965b: Vadja kohanimedes [Vadja in place names]. – H[arri] Moora EES = Metsmägi, Iris, Meeli Sedrik & Sven-Erik Soosaar 2012: Eesti etü- & L[embit] Jaanits (eds), Slaavi-läänemeresoome suhete ajaloost. Tal­ moloogiasõnaraamat [Estonian etymological dictionary]. Tallinn: linn: Eesti NSV Teaduste Akadeemia Ajaloo Instituut. 91–106. EKI & Eesti Keele Sihtasutus. User interface available at: ‹http://www. — 1968: Welche Sprache spricht die ostseefinnische Bevölkerung in eki.ee/dict/ety/› Vaipooli? – Fenno-ugrica. Juhlakirja Lauri Postin kuusikymmen- Elomaa, Jarmo 1989: Vatjan astevaihtelu ja vatjan ja inkeroisen kontaktit vuotispäiväksi 17.3.1968. MSFOu 145. Helsinki: SUS. 12–19. – erityisesti erään inkeroislähtöisen vatjalaisidiolektin näkökulmasta — 1977: Estonian and Karelian (Izhorian) language contacts. – Sovjet- tarkasteltuna. [Vowel harmony in Vote and the contacts of Vote and skoje Finno-Ugrovedenije XIII: 7–11. Ingrian – especially from the point of view of one idiolect of Vote with — 1981: Keelekontaktid. Eesti keele kontakte teiste keeltega [Language an Ingrian origin. Licenciate thesis.] Helsingin yliopiston itämeren­ contacts. The contacts of Estonian with other languages]. Eesti NSV suomalaisten kielten laitos. Teaduste Akadeemia Emakeele Seltsi toimetised 14. Tallinn: Valgus. Ernits, Enn 2005: Vadja keele varasemast murdeliigendusest ja hilisemast — 1987: Vadjalaste ja eestlaste kokkupuuteid [Connections between hääbumisest [On the earlier dialect division and the later extinction of Votes and Estonians]. – Fenno-Ugristica 14. Uurali keelte sõnavara ja Vote]. – Karl Pajusalu & Jan Rahman (eds), Piirikultuuriq ja -keeleq. grammatiline ehitus. Tartu Riikliku Ülikooli toimetised 776. 21–31. Piirikultuurid ja -keeled. Võro Instituudi toimõndusõq 17. Võro: — 1998: Vadjalaste ja eestlaste suhteid 19. sajandil ja 20. sajandi alguses Võro Instituutˊ. 76–90, 183–184. Available at: ‹https://wi.ee/et/voru- [Relations between Votes and Estonians in the 19th century and the instituudi-toimetised-17/› beginning of the 20th century]. – Emakeele Seltsi aastaraamat 35–42 — 2007: Isurite asualadest ja algkodust [About the residential districts 1989–1996: 14–21. Tartu. and the cradle of the Ingrians]. – Helen Koks & Jan Rahman (eds), — 2005: Vadja päevikud 1942–1980 [Vote diaries 1942–1980]. Ed. Ergo- Õdagumeresoomõ kodo. Läänemeresoome kodu. Võro Instituudi Hart Västrik. Litteraria 22. Eesti kultuuriloo allikmaterjale. Tartu: toimõndusõq 20. Võro: Võro Instituutˊ. 13–32, 175–176. Available at: Eesti Kirjandusmuuseum. ‹https://wi.ee/et/voru-instituudi-toimetised-20/› Björklöf, Sofia 2012: Viron rantamurteen länsiryhmän sanaston alkuperä Grünthal, Riho 1997: Livvistä liiviin. Itämerensuomalaiset etnonyymit suomalaislainojen valossa. [The origin of the vocabulary of the western [From Olonets Karelian to Livonian. Finnic ethnonymes]. CT 51. Hel­ subgroup of the Estonian Coast dialect in light of Finnish loanwords. sinki: Suomalais-ugrilainen laitos & SUS. Master’s thesis.] Pro gradu -tutkielma. Helsingin yliopiston suomen — (forthcoming): Inkeroisen kieli []. – Lassi Saressalo, kielen, suomalais-ugrilaisten ja pohjoismaisten kielten ja kirjallisuuk­ Kati Kallio & Riho Grünthal (eds), Inkerikot, setut ja vatjalaiset. Hel­ sien laitos. Available at: ‹http://urn.fi/URN:NBN:fi-fe2012121410304› sinki: SKS. — 2017: Svenska lånord i den västra gruppen av estniska kustdialekter Hakamies, Pekka 1991: Inkeri monietnisenä alueena [Ingria as a multieth­ [Swedish loanwords in the western subgroup of the Estonian Coast dia­ nic area]. – Pekka Nevalainen & Hannes Sihvo (eds), Inkeri. Historia, lect]. – Jan-Ola Östman, Caroline Sandström, Pamela Gustavsson & Lisa kansa, kulttuuri. SKST 547. Helsinki: SKS. 197–204.

140 141 SOFIA BJÖRKLÖF MUTUAL CONTACTS AND LEXICAL RELATIONS AMONG THE FINNIC VARIETIES OF WESTERN INGRIA AND NORTHEASTERN ESTONIA

Härmävaara, Hanna-Ilona 2013: Kielten samankaltaisuus monikielisen The development of selected cases as reflectors of the individualisation suomalais-virolaisen vuorovaikutuksen resurssina [Cross-linguistic of a dialect. PhD thesis.] Joensuun yliopiston humanisisia julkaisuja similarities as a resource of multilingual interaction between Finns 48. Joensuu: Joensuun yliopisto, humanistinen tiedekunta. Available and Estonians]. – Annekatrin Kaivapalu (ed.), Lähivõrdlusi. Lähiver- at: ‹http://urn.fi/URN:ISBN:978-952-219-036-9› tailuja 23. Tallinn: Eesti Rakenduslingvistika Ühing. 60–88. Available Köppen, Peter v[on] 1867: Erklärender Text zu der ethnographischen Karte at: ‹http://dx.doi.org/10.5128/LV23.03› des St. Petersburger Gouvernements. St. Petersburg: Kaiserliche Aka­ — 2014: Facilitating mutual understanding in everyday interaction demie der Wissenschaften. between Finns and Estonians. – Applied Linguistics Review 5/1: 211– Köppen, Peter von (ed.) 1849 = Koeppen, P. v. (ed.) 1849: Ethnographische 245. [Special issue on Receptive multilingualism.] Available at (subject Karte des Peterburgischen Gouvernements. St. Petersburg: Kaiserli­ to a charge): ‹https://doi.org/10.1515/applirev-2014-0010› che Akademie der Wissenschaften. — 2017: Official language policy as a factor in using receptive multi­ Krjukov, Aleksej 1993 [1987]: Ingermanland och ingermanländarna [Ingria lingualism among members of an Estonian and a Finnish student and Ingrians. Translated from Finnish into Swedish by Sulo Huovinen]. organization. – Maarja Siiner, Kadri Koreinik & Kara D. Brown – Sulo Huovinen (ed.), Ingermanland – om land och folk. : (eds), Language policy beyond the state. Language Policy 14. Cham: Kulturfonden för Sverige och Finland. 19–32. [Published for the Springer. 201–221. Available at (subject to a charge): ‹http://dx.doi. first time in Finnish in 1987: Inkerinmaa ja inkeriläiset. – Punalippu org/10.1007/978-3-319-52993-6_11› 8/1987: 123–130.] — (forthcoming): Building the bridge in interaction – receptive multilin- Krohn, Julius 1901: Kantelettaren tutkimuksia [Studies on Kanteletar] 2. gualism among Finns and Estonians. [PhD thesis.] University of Hel­ Toimittanut ja täydentänyt Kaarle Krohn. SKST 95. Helsinki: SKS. sinki. Kuznetsova, Natalia, Elena Markus & Mehmet Muslimov 2015: Finnic minor­ Heinsoo, Heinike 1998: Vadjalased ja vadja keele kujunemine [Votes and ities of Ingria. The current sociolinguistic situation and its background. the development of the Vote language]. – Jaan Õispuu & Marje Joalaid – Heiko Marten, Michael Rießler, Janne Saarikivi & Reetta Toivanen (eds), Kaheksa keelt, kaheksa rahvast. Tallinn: Tallinna Pedagoogika­ (eds), Cultural and linguistic minorities in the Russian Federation and ülikool. 14–29. the European Union. Multilingual Education 13: Comparative studies Junttila, Santeri 2015: Tiedon kumuloituminen ja trendit lainasanatutkimuk- on equality and diversity. Berlin: Springer. 127–167. Available at (subject sessa. Kantasuomen balttilaislainojen tutkimushistoria. [Accumula­ to a charge): ‹https://www.springer.com/gp/book/9783319104546› tion of information and tendencies in the study of loanwords. Research Laanest, Arvo 1961: Isuri murrete rühmitamisest [On dividing the Ingrian history of Baltic loans in Proto-Finnic. PhD thesis.] Helsinki: Helsingin dialects]. – Keele ja Kirjanduse Instituudi uurimused VI. Keele, kirjan- yliopisto. Available at: ‹http://urn.fi/URN:ISBN:978-951-51-1842-4› duse ja rahvaluule küsimusi. Tallinn: Eesti NSV Teaduste Akadeemia Jussila, Raimo (ed.) 1991: Suomen murteiden aineslähteitä [Material sources & Eesti Riiklik Kirjastus. 200–216. on Finnish dialects]. KKTKJ 57. Helsinki: KKTK & VAPK-kustannus. — 1964: Isurid ja isuri keel – meie lähemaid sugulaskeeli [Ingrians and Kettunen, Lauri 1915: Vatjan kielen äännehistoria [Sound history of Vote]. the Ingrian language – one of our most closely related languages]. Tal­ Suomi IV:15. Helsinki: SKS. linn: Eesti NSV Teaduste Akadeemia Keele ja Kirjanduse Instituut. — 1930 [1915]: Vatjan kielen äännehistoria [Sound history of Vote]. — 1970: Kes on isurid? – Aino Valmet, Paula Palmeos & Juhan Peegel Toinen, uusittu painos. SKST 185. Helsinki: SKS. (eds), Saaremaast Sajaanideni ja kaugemalegi. Tallinn: Kirjastus Val­ — 1945: Tieteen matkamiehenä. Kaksitoista ensimmäistä retkeä 1907– gus. 100–110. 1918 [A scientific explorer. My first twelve trips 1907–1918]. Helsinki: — 1988: Vatjan sanaston suhteista lähisukukieliin [On lexical relations Werner Söderström. of Vote with closely related languages]. – Symposium 1988. Itämeren- — 1957: ”Isuri keel” [“Izhorian language”]. – Virittäjä 61: 124–133. suomalaisen filologian symposiumi, Turku 30.8.–2.9.1988. Referaatit. Kokko, Ossi 2007: Inkerinsuomen pirstaleisuus. Eräiden sijojen kehitys Turku: Turun yliopiston suomalaisen ja yleisen kielitieteen laitos, Fin­ murteen yksilöllistymisen kuvastajana. [Scattered Ingrian Finnish. ska institutionen vid Åbo Akademi, Suomen ja Neuvostoliiton välinen tieteellis-teknillinen yhteistoimintakomitea. 48. 142 143 SOFIA BJÖRKLÖF MUTUAL CONTACTS AND LEXICAL RELATIONS AMONG THE FINNIC VARIETIES OF WESTERN INGRIA AND NORTHEASTERN ESTONIA

Lauerma, Petri 1993: Vatjan vokaalisointu. [Vowel harmony of Vote. PhD — 2012: Votic or Ingrian: new evidence on the Kukkuzi variety. – Finn- thesis.] MSFOu 214. Helsinki: SUS. isch-Ugrische Mitteilungen 35: 77–94. Available at (subject to a charge): Leppik, Merle 1975: Ingerisoome Kurgola murde fonoloogilise süsteemi ‹https://buske.de/finnisch-ugrische-mitteilungen-band-35-9607.html› kujunemine [Development of phonological system in the Ingrian Finn­ — 2013: Correlation between social and linguistic parameters in model­ ish Kurkola dialect]. Tallinn: Eesti NSV Teaduste Akadeemia Keele ja ing language contact: evidence from endangered Finnic varieties. – Kirjanduse Instituut. International Journal of the Sociology of Language 221: 53–76. Avail­ Leskinen, Heikki 1991: Inkerin kielimuodot [The language varieties of able at (subject to a charge): ‹https://doi.org/10.1515/ijsl-2013-0023› Ingria]. – Pekka Nevalainen & Hannes Sihvo (eds), Inkeri. Historia, Marttila, Annu 2010: A cross-linguistic study of lexical iconicity and its kansa, kulttuuri. SKST 547. Helsinki: SKS. 222–233. manifestation in bird names. [PhD thesis.] Helsinki: University of Hel­ — 1995: Inkerin asutus ja väestöryhmät [The inhabitants and ethnic sinki. groups of Ingria]. – Mauno Jokipii (ed.), Itämerensuomalaiset. Heimo- Muslimov, Mehmed Zakirovitš 2005: Jazykovye kontakty v Zapadnoj Inger- kansojen historiaa ja kohtaloita. Jyväskylä: Atena Kustannus. 163– manlandii (nižnee tečenie reki Lugi). [Language contacts in Ingria (the 173. lower course of the Luga River). PhD thesis.] Sankt-Peterburg: Insti­ Lindstedt, Jouko 2000: Linguistic balkanization: contact-induced change by tute for Linguistic Studies of the Russian Academy of Sciences. mutual reinforcement. – Dicky Gilbers, John Nerbonne & Jos Schae­ Muslimov, M[ehmed] Z. 2009: K klassifikacii finskix dialektov Ingerman­ ken (eds), Languages in contact. Studies in Slavic and General Lin­ landii [On classification of the Finnish dialects in Ingria]. – Sergej guistics, vol. 28. Amsterdam: Rodopi. 231–246. Available at (subject Myznikov & Igorʹ Brodskij (eds), Voprosy uralistiki 2009. Naučnyj to a charge): ‹https://brill.com/view/title/30810› alʹmanax. Sankt-Peterburg: Nauka. 179–204. Mäger, Mart 1967: Eesti linnunimetused [Bird names in Estonian]. Ed. Val­ Muslimov, Mehmet 2015: Moloskovitsa dialect in Western Ingria. Presen­ dek Pall. Tallinn: Eesti NSV Teaduste Akadeemia Keele ja Kirjanduse tation given at the Congressus Duodecimus Internationalis Fenno- Instituut. Ugristarum, Oulu 2015. Abstract in the Book of abstracts. 204–205. Mägiste, Julius 1925: Rosona (Eesti Ingeri) murde pääjooned [Main charac­ Abstract available at: ‹https://www.oulu.fi/sites/default/files/content/ teristics of the Rosona (Estonian Ingria) dialect]. Tartu. CIFU12-BookOfAbstracts_4.pdf› — 1927: Paarist vadja laenust eesti keeles [On a couple of Vote loans in Must, Mari 2000: Vene laensõnad eesti murretes [Russian loanwords in Estonian]. – Eesti Kirjandus 21: 161–162. Eesti Kirjanduse Selts. Estonian dialects]. Tallinn: Eesti Keele Sihtasutus. Markus, Elena & Fedor Rozhanskiy 2010: Asymmetry in Votic-Ingrian Nirvi, Ruben E[rik] 1961: Inkeroismurteiden asema [The position of Ingrian language contacts. – SLE 2010: 43rd Annual meeting of the Societas dialects]. – Kalevalaseuran vuosikirja 41: 99–132. Helsinki: WSOY. Linguistica Europaea, Vilnius University, Lithuania, 2–5 September Oja, Vilja 2005: Sõna põll taustast [On the background of the word põll]. – 2010. Book of abstracts. 141. Keel ja Kirjandus XLVIII: 475–480. Markus, E[lena] B. & F[edor] I. Rozhanskiy 2011a: Sovremennyj vodskij Ojansuu, Heikki 1916: Suomen kielen tutkimuksen työmaalta. Sarja esitelmiä jazyk. Teksty i grammatičeskij očerk. V 2-h tomah. [Contemporary I [From the field of study of the . A series of presenta­ Vote. Texts and grammar. In 2 volumes.] Sankt-Peterburg: Nestor- tions I]. Suomalais-virolaiset kielelliset kosketukset [Finnish-Estonian Istorija. language contacts]. 98–202. Jyväskylä: Gummerus. Markus, Elena & Fedor Rozhanskiy 2011b: The development of a mixed lan­ Pajusalu, Karl 1999: Eesti murded ja murderühmad [Estonian dialects and guage in the multilingual environment (evidence from the Kukkuzi dialect groups]. – Emakeele Seltsi aastaraamat­ 43: 64–98. Tartu. dialect). – Uralic languages and multilingualism: contexts and mani- Pajusalu, Karl, Tiit Hennoste, Peeter Päll & Jüri Viikberg 2009: Eesti festations in a language family 2–3 June, 2011. Abstract available at: murded ja kohanimed [Estonian dialects and place names]. 2., täien­ ‹https://www.slm.uni-hamburg.de/ifuu/download/konferenzen/markus- datud trükk, toim. Tiit Hennoste. Tallinn: Tartu Ülikooli eesti ja üld­ rozhanskiy.pdf› keeleteaduse instituut & EKI.

144 145 SOFIA BJÖRKLÖF MUTUAL CONTACTS AND LEXICAL RELATIONS AMONG THE FINNIC VARIETIES OF WESTERN INGRIA AND NORTHEASTERN ESTONIA

Pajusalu, Karl, Péter Pomozi, Dezső Juhász & Tiit-Rein Viitso 2012: Socio­ Rozhanskiy, Fedor & Elena Markus 2014: Lower Luga Ingrian as a convergent linguistic comparison of the development of Estonian and Hungarian language. Presentation given at the FINKA Symposium. On the border of dialect areas. – Linguistica Uralica XLVIII: 241–264. Available at: language and dialect. University of Eastern Finland, Joensuu, 4–6 June, ‹http://dx.doi.org/10.3176/lu.2012.4.01› 2014. Handout of the presentation; abstract in Abstracts. 36–37. Pall, Valdek 1969: Põhja-Tartumaa kohanimed [Place names in Northern — 2019: A new resource for Finnic languages: The outcomes of the Tartumaa] 1. Ed. M[adis] Norvik. Tallinn: Valgus. Ingrian documentation project. – Sofia Björklöf & Santra Jantunen — 1970: Maahanmuuton jälkiä Pohjois-Tartumaan paikannimissä [Traces (eds), Multi­lingual Finnic. Language contact and change. Uralica Hel­ of immigration in the place names of Northern Tartumaa]. – Virittäjä singiensia 14. Helsinki: SUS. 303–326. Available at: ‹https://doi. 74: 12–22. org/10.33341/uh.85039› — 1977: Põhja-Tartumaa kohanimed [Place names in Northern Tartu­ Saareste, Andrus 1952: Kaunis emakeel. Vesteid eesti keele elust-olust maa] 2. Ed. M[adis] Norvik. Tallinn: Valgus. [Beautiful mother tongue. Discussions about Estonian]. Lund: Eesti Pasanen, Annika 2015: Kielipesiä, kielimestareita ja kieli-ideologioita. Kirjanike Kooperatiiv. Inarin­saamen kielen revitalisaatio [Language nests, language masters, Salminen, Tapani 2010: Huomioita lintujen nimityksistä [Observations on and language ideologies. The revitalisation of Inari Saami]. Presenta­ bird names]. – Sirkka Saarinen, Kirsti Siitonen & Tanja Vaittinen tion in the Finno-Ugrian Society, 18 September 2015. (eds), Sanoista kirjakieliin. Juhlakirja Kaisa Häkkiselle 17. marras- Paulaharju, Marjut (ed.) 2010: Samuli Paulaharjun Inkeri [Samuli Paula­ kuuta 2010. MSFOu 259. Helsinki: SUS. 343–358. harju’s Ingria. Edited articles about the journey to Ingria in the sum­ Saloheimo, Veijo 1991: Inkerinmaan asutus ja väestö 1618–1700 [The set­ mer of 1911 written by S. Paulaharju]. SKST 1292. Helsinki: SKS. tlement and inhabitants of Ingria in 1618–1700]. – Pekka Nevalainen Porkka, Volmari 1885: Ueber den Ingrischen Dialekt mit Berücksichtigung & Hannes Sihvo (eds), Inkeri. Historia, kansa, kulttuuri. SKST 547. der übrigen finnisch-ingermanländischen Dialekte. Helsingfors: J. C. Helsinki: SKS. 67–82. Frenckell & Sohn. Savijärvi, Ilkka 1996a: Länsi-Inkerin heimot, kielet ja murteet [The tribes, Posti, Lauri 1970: Eesti põlle algupärast [On the origin of the Estonian põll]. languages, and dialects of western Ingria]. – Ilkka Savijärvi, Muusa – Keel ja Kirjandus XIII: 470–472. Savijärvi & Janne Heikkinen (eds), Vot, ihminen tahtoo kotimaalle. Randefelt, Roland 1992: Inkeri / Ингерманландия / Ingermanland / Ingria / Länsi-Inkerin kieltä ja kohtaloita. Studia Carelica Humanistica 8. Ingeri: tiekartta / карта дорог / vägkarta / Straßenkarte / road map / Joensuu: Joensuu University Press. 7–17. teede kaart. 1 : 200 000. — 1996b: Western Ingria – where languages and dialects meet. – Ilkka Riionheimo, Helka 2007: Muutoksen monet juuret. Oman ja vieraan ris- Savijärvi, Muusa Savijärvi & Janne Heikkinen (eds), Vot, ihminen teytyminen Viron inkerinsuomalaisten imperfektinmuodostuksessa. tahtoo kotimaalle. Länsi-Inkerin kieltä ja kohtaloita. Studia Carelica [Multiple roots of change. Mixing native and borrowed influence in Humanistica 8. Joensuu: Joensuu University Press. 183–188. the past tense formation by . PhD thesis.] SKST 1107. — 1998: Cognate languages in contact. Ingria – where four Baltic-Finnic Helsinki: SKS. languages meet. – Jussi Niemi, Terence Odlin & Janne Heikkinen — 2013: Kieltenvälinen vertailu kielikontaktitutkimuksessa [Comparing (eds), Language contact, variation, and change. Studies in Languages between languages in the study of language contact]. – Leena Koleh­ 32. Joensuu: Joensuun yliopisto. 269–286. mainen, Matti Miestamo & Taru Nordlund (eds), Kielten vertailun — 2001: Lähisukukielet kontaktissa [Closely related languages in contact]. metodiikka. SKST 1387. Helsinki: SKS. 219–250. – Congressus Nonus Internationalis Fenno-Ugristarum 7.–13.8.2000 Rozhanskiy, F[edor] I. & E[lena] B. Markus 2013: O statuse nižnelužskogo Tartu. Pars VI: Dissertationes sectionum Linguistica III. Tartu. 161–166. dialekta ižorskogo jazyka sredi rodstvennyx idiomov [On the status of Setälä, E[mil] N[estor] 1964: E. N. Setälän vatjalaismuistiinpanot. Les notes Lower Luga Ingrian among related varieties]. – Lingvističeskij bespre- d’E. N. Setälä sur la langue vote. Rédigées et publicées par les soins de del – 2. Sbornik naučnyx trudov k jubileju A. I. Kuznetsovoj. Moskva: Lauri Posti et de Seppo Suhonen. – Memoria saecularis E. N. Setälä Moscow University Press. 219–232. 27.II.1964. MSFOu 135. Helsinki: SUS.

146 147 SOFIA BJÖRKLÖF MUTUAL CONTACTS AND LEXICAL RELATIONS AMONG THE FINNIC VARIETIES OF WESTERN INGRIA AND NORTHEASTERN ESTONIA

SKES = Toivonen, Y. H., Erkki Itkonen, Aulis J. Joki & Reino Peltola 1955– Toikka, Kaja 2003: Kirderannikumurde idaosa murde- ja keelekontaktid. 1981: Suomen kielen etymologinen sanakirja [Etymological dictionary [Dialect and language contacts in the eastern part of the (Estonian) of Finnish] I–VII. LSFU XII. Sanahakemiston (VII) koostaneet Satu Northeastern Coastal dialects. Printed master’s thesis.] Tallinna Peda­ Tanner ja Marita Cronstedt. Helsinki: SUS. googikaülikooli läänemeresoome keelte õppetool. Söderman, Tiina 1996: Lexical characteristics of the Estonian North East- Tsvetkov, Dimitri 1925: Vadjalased [The Votes]. – Eesti Keel IV: 39–44. ern Coastal dialect. [PhD thesis.] Studia Uralica Upsaliensia 24. Acta Vaba, Lembit 1977: Läti laensõnad eesti keeles [Latvian loanwords in Esto­ Universitatis Upsaliensis. Uppsala. nian]. Tallinn. Sovijärvi, Antti 1944: Foneettis-äännehistoriallinen tutkimus Soikkolan — 1997: Uurimusi läti–eesti keelesuhetest. [Studies on Latvian–Estonian inkeroismurteesta [Phonetic-soundhistorical study on the Ingrian dia­ language relations. PhD thesis.] Tallinn – Tampere: EKI & Tampereen lect of Soikkola]. Suomi-sarjan eripainoksia 5. Uusi sarja. Helsinki: yliopiston suomen kielen ja yleisen kielitieteen laitos. SKS. — 2014: Levik kui etümoloogiline kriteerium (eeskätt eesti sõnavara SSA = Itkonen, Erkki & Ulla-Maija Kulonen (eds) 1992–2000: Suomen näitel) [Areal distribution as an etymological criterion (mainly on the sanojen alkuperä. Etymologinen sanakirja [Origin of Finnish words. example of Estonian lexis)]. – Keel ja Kirjandus LVII: 700–707. Avail­ Etymological dictionary] 1–3. SKST 556, KKTKJ 62. Helsinki: SKS able at: ‹http://kjk.eki.ee/ee/issues/2014/8-9/541› & KKTK. — 2015: Sõna sisse minek [“Going in to the word”. Collection of articles Suhonen, Seppo 1979: Über die Beziehungen zwischen den Finnischen u. d. 1989–2015]. Emakeele Seltsi toimetised 73. Tallinn: Eesti Teaduste Estnischen Küstdialekten. – Erhard F. Schiefer (ed.), Explanationes Akadeemia Emakeele Selts. et tractationes Fenno-Ugricae in honorem Hans Fromm. Münchener Viitso, Tiit-Rein 1993: Tutkimus vatjan vokaalisoinnusta [A study on the Universitäts-Schriften. München. 357–366. vowel harmony of Vote. Review: Petri Lauerma, Vatjan vokaalisointu]. — 1985: Wotisch oder Ingrisch? – Dialectologia Uralica: Materialen des – Virittäjä 97: 522–526. ersten Internationalen Symposions zur Dialektologie der uralischen Wiedemann, Ferdinand Johann 1893 [1869]: Ehstnisch-deutsches Wörter- Sprachen 4.–7. September 1984 in Hamburg. Veröffentlichungen der buch. Zweite vermehrte Auflage. Redigirt von Jacob Hurt. St. Peters­ Societas Uralo-Altaica. Band 20. Herausgegeben von Wolfgang Veen­ burg: der Kaiserlichen Akademie von Wissenschaften. ker. Wiesbaden: Otto Harrassowitz. 139–148. — 1986: Zur Identifizierung der jungen Lehnwörter im Wotischen. – Lyökämme käsi kätehen. Beiträge zur Sprachkontaktforschung im Bereich des Finnougrischen und des Germanischen A. D. Kylstra zum 65. Geburtstag. Amsterdam: Rodopi. 151–161. — 2000: Kirderannikumurde suhted naaberkeeltega [Relations of the (Estonian) Northeastern Coastal dialects with neighbouring lan­ guages]. – Jüri Viikberg (ed.), Inter dialectos nominaque. Pühendus- teos Mari Mustale 11. novembril 2000. Eesti Keele Instituudi toimeti­ sed 7. Tallinn: EKI. 362–374. Talve, Ilmar 1990: Matka vatjalaisiin 1942 [Trip to the Votes in 1942]. – Pekka Laaksonen & Sirkka-Liisa Mettomäki (eds), Inkerin teillä. Kalevalaseuran vuosikirja 69–70. Helsinki: SKS. 46–66. Teinonen, Markku & Timo J. Virtanen (eds) 1999: Ingrians and neighbours. Focus on the eastern Baltic Sea region. Studia Fennica Ethnologia 5. Helsinki: SKS.

148 149 SOFIA BJÖRKLÖF MUTUAL CONTACTS AND LEXICAL RELATIONS AMONG THE FINNIC VARIETIES OF WESTERN INGRIA AND NORTHEASTERN ESTONIA

Itämerensuomen varieteettien keskinäiset virossa. Päähypoteesini on, että tutkimallani alueella on paljon kes­ kontaktit ja sanastolliset suhteet kinäisten kontaktien ja alueellisen diffuusion tuloksena saatuja laina­ Länsi-Inkerissä ja Koillis-Virossa sanoja, vaikka läheisten sukukielten tapauksessa lainaamisen suun­ taa on usein vaikea todistaa. Lainasanat tarkentavat kuvaa sellaisten Sofia Björklöf läheisten sukukielten keskinäisistä suhteista, joiden kehitystä ei voi kuvata yksin perinteisen binaarisen sukupuumallin keinoin. Esitän Länsi-Inkerin ja Koillis-Viron historiallinen kielikontaktitilanne hei­ esimerkkejä vatjan, inkeroisen, viron ja suomen murteiden välises­ jastuu alueen kielten sanastossa eri tavoin. Lähemmin tarkasteltavat tä lainautumisesta (virosta on lainattu sanat trehvama ’tavata; osua; kielimuodot ovat viron koillisrannikkomurteet28, viron itämurteet, sattua’, põll ’esiliina’, hulkuma ’kuljeskella’, suomesta heikko, kiu- vatjan murteet, inkeroismurteet sekä Inkerin suomalaismurteet. Kaik­ ru, raju, höyry ja vatjasta ilata ’siivota’). Näiden esimerkkitapausten ki nämä kielimuodot ovat geneettisesti läheisiä itämerensuomalaisia avulla kuvaan läheisten sukukielten välisen lainautumisen tutkimisen varieteetteja. metodia. Osa julkaistavista etymologioista on uusia. Virolaiset lainat Vatjan ja inkeroisen perinteiset puhuma-alueet sijaitsivat Länsi-­ vatjassa ja inkeroisessa ovat tavallisesti tunnistettavissa pienen levik­ Inkerissä, Suomenlahden kaakkoisrannikolla Narvajoen ja nykyi­ kinsä perusteella. Laina-alkuperää olevasta vatjan, inkeroisen ja vi­ sen metropolin, Pietarin, välisellä alueella. Pietari rakennettiin ron yhteisestä sanastosta suurin osa taas on peräisin suomesta. Sekä itämerensuomalaisten ikivanhalle kauppapaikalle 1700-luvun alussa. vatjassa että virossa esiintyvillä lainoilla on usein inkeroisen lisäksi Paikallinen väestö ei koskaan ole muodostanut omaa valtiota, vaan levikki suomessa, joten kahden viimemainitun äänteellisen saman­ aluetta ovat hallinneet Venäjä ja Ruotsi. 1600-luvulla Inkeriin muutti kaltaisuuden takia varmaa lainanantajakieltä on tavallisesti mahdoton suomenpuhujia ja venäläisten migraatio kasvoi huomattavasti Pietarin määritellä. Vatjasta inkeroiseen ja viroon on saatu vain yksittäisiä lai­ vuonna 1703 tapahtuneen perustamisen jälkeen. Lopulta virolaiset tu­ noja, jotka saattavat olla myös substraattia. lokkaat asettuivat samalle alueelle 1800-luvulla. Länsi-Inkerissä oli lisäksi saksan- ja todennäköisesti myös ruotsinpuhujia. Uusien asukkaiden saapuminen poliittisesti hallitsevilta alueilta muutti Inkerin väestörakennetta, lisäsi kielellistä diversiteettiä, muutti kielten välistä hierarkiaa ja johti asteittain Inkerin kanta-asukkaiden kielen ja identiteetin vaihtoon. Sekä vatjan että inkeroisen puhujat olivat monikielisiä ja joissain kylissä oli seka-asutusta. Yleisesti ot­ taen varieteettien väliset rajat vanhoissa maalaisyhteisöissä ovat olleet hyvin häilyviä: vatjan ja inkeroisen puhujat ovat toistuvasti ilmaisseet mielipiteenään, että heidän murteensa ovat saman kielen murteita. Tässä artikkelissa kuvaan yhtäältä Inkerinmaan historiallista sosio­lingvististä tilannetta, toisaalta Inkerin historiallista kielikontak­ titilannetta. Artikkelin sanasto-osuus pohjautuu vanhoihin murremate­ riaaleihin, jotka on julkaistu sanakirjoissa, sekä viron osalta edelleen osin julkaisemattomaan, arkistossa säilytettävään materiaaliin. Sanat, joita analysoin etymologisesti, esiintyvät vatjassa, inkeroisessa ja 28. Termin valinnasta ks. Björklöf (2018: 363).

150 151 SOFIA BJÖRKLÖF MUTUAL CONTACTS AND LEXICAL RELATIONS AMONG THE FINNIC VARIETIES OF WESTERN INGRIA AND NORTHEASTERN ESTONIA

Läänemeresoome keelte omavahelised kontaktid Peamiseks hüpoteesiks on, et on võimalik leida omavaheliste kontak­ ja sõnavara seosed Lääne-Ingeris ja Kirde-Eestis tide tulemusena levinud laensõnu, kuigi lähedaste sugulaskeelte puhul laenamise suunda on tihtipeale raske tõestada. Laensõnad täpsustavad Sofia Björklöf pilti lähedaste sugulaskeelte vahelistest kontaktidest, mille arengut ei saa üksnes traditsionaalse binaarse sugupuuga kirjeldada. Näiteid Lääne-Ingerimaa ja Kirde-Eesti ajaloolised keelekontaktid peegeldu­ tuuakse vadja, isuri, eesti ja soome murrete omavahelise sõnade laena­ vad selle geograafilise ala keelte sõnavaras erinevatel viisidel. Käes­ mise kohta (eesti keelest on laenatud sõnad trehvama, põll, hulkuma, olevas artiklis vaadeldakse lähemalt eesti kirderannikumurdeid, eesti soome keelest heikko ’nõrk’, kiuru ’lõoke’, raju ’metsik; äge’, mis esi­ idamurdeid, vadja keele murdeid, isuri keele murdeid ning Ingerimaa neb ka liitsõnades rajuilma, rajumyrsky ’torm’, höyry ’aur’ ning vadja soome keele murdeid. Mainitud keelekujud on ajalooliselt lähedased keelest on laenatud ilata ’koristama’). Nende näidete abil kaalutakse läänemeresoome sugulaskeeled. ka metoodiliselt lähedaste sugulaskeelte vahelist laenamisprotsessi. Vadja ja isuri keele kõnelejate traditsioonilised alad asusid Lää­ Artiklis avaldatud etümoloogiatest osad on uued. Eesti laenud vadja ne-Ingeris, Soome lahe kagukaldal Narva jõe ja praeguse Peterburi va­ ja isuri keeles on tavaliselt tuvastatavad oma kitsama levila alusel. helisel alal. Peterburi rajati läänemeresoomlaste igivanale kauplemis­ Seevastu suurem osa laenatud sõnavarast, mis on ühine vadja, isuri kohale 18. sajandi alguses. Kohalik rahvas pole kunagi moodustanud ja eesti keelele, on pärit soome keelest. Nii vadja kui ka eesti keeles enda oma riiki, vaid allus ala hõivanud Venemaale ja Rootsile. 17. esinevaid laene tuntakse tavaliselt peale isuri ka soome keeles. Kuna sajandil asus Ingerisse elama sisserändajatena põhjapoolt soome keele kahe viimase häälikuline ülesehitus on sarnane, on kindlat laenuand­ kõnelejaid. Ka venelaste migratsioon kasvas märgatavalt peale Peter­ jakeelt tihtipeale võimatu määratleda. Vadja keelest isuri ja eesti keel­ buri asutamist 1703. aastal. Viimastena asusid 19. sajandil Ingeri­ de on saadud ainult üksikuid laene, mille puhul võib olla tegemist ka maale uustulnukatena ka eestlastest migrandid. Peale eelmiste on substraadiga. Lääne-Ingeris elanud ka saksa keele ning tõenäoliselt ka rootsi keele kõnelejaid. Uute elanike saabumine poliitiliselt juhtivatest riikidest muutis Ingerimaa rahvastikku, kaasates keelelist mitmekesisust. Tekkinud olukord muutis keelte vahelist hierarhiat ja viis järk-järgult Ingeri põlis­elanike keele ja identiteedi assimileerumiseni. Nii vadja kui ka isuri keele kõnelejad olid kaua aega mitmekeelsed. Osades külades oli segaasustus. Üldiselt on keelte vahelised piirid vanades põllu­ majandus­ühiskondades olnud väga hägusad. Näiteks on vadja ja isuri keele kõne­lejad korduvalt väljendanud arvamust, et nende murded on ühe ja sama keele murded. Käesolevas artiklis kirjeldatakse Ingerimaa ajaloolist sotsio­ lingvistilist olukorda ja Lääne-Ingeri ajaloolisi keelekontakte. Ar­ tiklis vaadeldavad sõnad põhinevad vanadel murdematerjalidel, mis on avaldatud sõnaraamatutes. Peale selle analüüsitakse ka arhiivides leiduvaid avaldamata eestikeelseid sõnu. Etümoloogiliselt analüseeri­ takse lähemalt sõnu, mis esinevad nii vadja, isuri kui ka eesti keeles.

152 153 LIINA LINDSTRÖM, MAARJA-LIISA PILVIK, MIRJAM RUUTMA & KRISTEL UIBOAED University of Tartu

On the use of perfect and pluperfect in Estonian dialects: Frequency and language contacts1

Abstract The present paper examines the use of two compound tenses – perfect and pluperfect – in Estonian. Perfect and pluper­ fect have emerged due to the influence of the Baltic and Germanic languages and are used frequently in Estonian. However,­ while looking at the usage frequency derived from the Corpus of Esto­ nian Dialects, dialect areas display remarkable differences, which can be explained either by local language contacts with Swedish, Russian, Latvian, and Finnic languages (Votic, Ingrian, and Finn­ ish) or by functional differences in the use of compound tenses. It appears that there are two main regions where the compound tens­ es are used more often compared to other areas: the Insular dialect and Mulgi dialect regions. The increase of compound tenses in the Insular dialect could be a result of contacts with Swedish. How­ ever, the Insular dialect also exhibits a high number of negated utterances using the perfect reflecting changes in the formation of negation more generally in this area. The Mulgi dialect shows a high number of pluperfect forms that can be related to the abun­ dance of reported narratives in the data, but also as an increase of using pluperfect as an evidential strategy, which is probably a result of contacts with Latvian. Keywords compound tenses, perfect, pluperfect, Estonian dialects, frequency, corpus-based dialectology, language contacts

1. The previous version of this study has been published in Estonian (Lindström, Pilvik, Ruutma & Uiboaed 2015). This study was supported by the European Union through the European Regional Development Fund (The Centre of Excellence in Estonian Studies).

Multilingual­ Finnic. Language contact and change. 155–193. Uralica Helsingiensia 14. Helsinki 2019. ‹https://doi.org/10.33341/uh.85035› LIINA LINDSTRÖM, MAARJA-LIISA PILVIK, ON THE USE OF PERFECT AND PLUPERFECT IN ESTONIAN DIALECTS: MIRJAM RUUTMA & KRISTEL UIBOAED FREQUENCY AND LANGUAGE CONTACTS

1. Introduction During the development of the construction, the olema-verb had auxiliarized and the past event (i.e., the expression of tense) had been In the present article, we examine the usage frequency of Estonian foregrounded (Metslang 1993). A typical perfect concludes the past compound past tenses – perfect and pluperfect – in Estonian dialects, event from the present point of view; “the perfect indicates the con­ using the data from the Corpus of Estonian Dialects2 (CED). We take tinuing present relevance of a past situation” (Comrie 1976: 52). The a look at the links between the frequency of use and local language semantic shift from resultative to perfect thus includes the generaliza­ contacts to examine whether long-term language contacts explain dif­ tion from “current result” to “current relevance” (Lindstedt 2000). ferences between dialects with regard to their (dis)preference for the Estonian compound tenses still can carry both temporal and as­ compound tense forms. pectual meanings simultaneously with one of them becoming more Perfect and pluperfect are past tense forms that have emerged in salient in a sentence. The Estonian perfect has all the main functions Estonian and other Finnic languages most likely through the influence that are characteristic of typical perfects (cf. Comrie 1976; Lindstedt of surrounding Indo-European languages. Ariste has claimed that the 2000): perfect of result, experiential perfect, perfect of a persistent development of compound tense usage in Estonian mainly has fol­ situation, perfect of recent past, quotative (evidential) perfect (see lowed the example of the Baltic languages (Ariste 1956), later Sere­ Metslang 1997 for details). Example (1) represents the experiential brennikov (1959) has stressed that in addition to the Baltic languages, perfect; ex. (2) represents the pluperfect of the result. the Germanic languages have also played an important role. The same (1) Estonian assumption is shared more or less by several scholars: the role of the Ta on käi-nud mitme-s ülikooli-s. Baltic languages has been important in the initial emergence of com­ s/he be.3sg go-app many-ine university-ine pound past tenses in Estonian and other Finnic languages, but the later ‘He has gone to many universities.’ use and function of the perfect and pluperfect are attributed to Ger­ manic influence (Ikola 1960; Laanest 1975: 152). Laakso (2001: 191) (2) Estonian links the spread of the compound past tenses more generally with the Lapse-d ol-i-d toa juba ära korista-nud. emergence of copula constructions, which, in turn, bear the imprint of child-pl be-pst-3pl room.gen already pfv clean-app other (especially Germanic) languages. ‘The children had already cleaned the room’ The emergence of compound tenses can also be explained as a natural development of a language – grammaticalization. Likewise, Although the emergence of perfect and pluperfect has also been ex­ Estonian compound past tenses can be seen as a universal grammati­ plained as a language-internal phenomenon, language contacts have calization path from the aspectually used verb olema ‘to be’ and the clearly reinforced the influence on compound tense usage. For in­ predicative active past participle (-nud ) to the tense category where stance, literary Estonian has been under strong Russian and German the olema-verb and a past participle form the core of the clause. (Mets­ influence, especially during 1890–1990 (Metslang 1997). lang 1993.) Aspectual use is still present today and is sometimes re­ In Estonian dialects, the use of the past compound tenses has not ferred to as the retrospective, expressing the present state as a result of gained much attention. Lonn & Niit (2002) note that in the Insular a previous event (Erelt 2013: 86), e.g., Poiss on üllatunud ‘The boy is dialect, compound tenses are extensively used. As this grammatical surprised’, Koer on surnud ‘The dog is dead’. phenomenon exists in all dialects, the topic is not discussed in other papers dealing with dialects. Differences among dialects, however, have been attested for verbal constructions in general. Uiboaed et al. 2. ‹http://www.murre.ut.ee/murdekorpus/› (2013) have studied different constructions consisting of a finite and

156 157 LIINA LINDSTRÖM, MAARJA-LIISA PILVIK, ON THE USE OF PERFECT AND PLUPERFECT IN ESTONIAN DIALECTS: MIRJAM RUUTMA & KRISTEL UIBOAED FREQUENCY AND LANGUAGE CONTACTS a non-finite verb in the Corpus of Estonian Dialects (on which the 2. Estonian dialects and language contacts present article also is based) and found that the western dialects tend to use such constructions more extensively than eastern dialects. The Estonian dialects are traditionally divided into northern and southern western, higher frequency group was formed by the Mid, Western, groups, which have also been thought of as having developed from Insular, and sometimes also Coastal (depending on the method) dia­ separate languages. These two groups differ on phonological, mor­ lects. The lower frequency areas in the eastern part were formed by phological as well as on lexical levels. As the Northeastern and Coast­ the southern dialects (Võru, Seto, Tartu, Mulgi) as well as the Eastern al dialects have systematic differences from other northern dialects, and Northeastern dialects. (Uiboaed 2013: 89.) Uiboaed’s study, how­ they are often classified as a separate group. (Kask 1984.) ever, did not include olema + -nud compounds and the present article The present study follows the dialect classification by Pajusalu extends the knowledge of this phenomenon. et al. (2009) that has also been the basis for the Corpus of Estonian In this article, we discuss the use of perfect and pluperfect forms Dialects (Map 1). According to this classification, there are three main in Estonian dialects based on the Corpus of Estonian Dialects. We dialect groups: the North Estonian dialect group consists of the East­ concentrate on regional differences and explore whether there are ern, Insular, Central, and Western dialects, the Northeastern Coastal significant (dis)similarities between geographical areas, which could dialect group consists of the Northeastern and the Coastal dialects, and be seen in previous studies (Uiboaed et al. 2013; Uiboaed 2013). We the South Estonian group consists of the Mulgi, Tartu, Seto, and Võru therefore observe the compounds formed by the olema-verb and -nud dialects. A minor difference concerns Võru and Seto which are con­ participle in Estonian dialects, focusing primarily on usage frequency. sidered to be one dialect by Pajusalu et al., whereas the CED separates Quantitative findings are interpreted in the light of language contacts. these as two different geographical variants. This distinction is clearly Our main hypothesis is based on Uiboaed’s study about other ver­ justified on a syntactic level, as Seto has proved to be considerably bal constructions, so we expect higher usage frequency of compound different from Võru in various syntactic respects (see Uiboaed 2010; past tenses in the western dialects compared to eastern areas. Differ­ Lindström, Uiboaed & Vihman 2014; Lindström forthcoming). ences might emerge from the fact that in western dialects compound The influence of language contacts on Estonian morphosyntax tenses have similar functions as the simple past tense (imperfect) and has been examined mostly in connection with two linguistic areas the latter is preferred in eastern dialects. Discrepancies in usage fre­ (Sprachbund): the Circum-Baltic (see Dahl & Koptjevskaja-Tamm quency might be induced by language contacts: long-term interaction 2001; Wälchli 2011) and the European linguistic area, whose charac­ with Germanic languages (especially with Swedish) in western Esto­ teristics are usually subsumed under the term Standard Average Eu- nia and stronger contacts with Latvian and Russian in eastern Estonia. ropean (SAE3) (Haspelmath 2001). The main mediator of the SAE We now turn to the overview of the relevant contact languages features into Estonian has probably been German (both Low and High which most likely have had an impact on usage frequency of the per­ German), which was the language of the upper class in the region for fect and pluperfect in dialects. Additionally, we briefly present corre­ a long period of time. German has influenced Estonian on all linguistic sponding tense forms in these languages. levels, including the functions of the perfect and pluperfect (Metslang 1997), and this impact was constant from the 13th century up to the beginning of the 20th century. However, based on previous studies, it is hard to estimate the regularity and intensity of German influence on different Estonian dialects; there is no evidence that German influence

3. SAE features in Estonian have been discussed by Metslang (2009).

158 159 LIINA LINDSTRÖM, MAARJA-LIISA PILVIK, ON THE USE OF PERFECT AND PLUPERFECT IN ESTONIAN DIALECTS: MIRJAM RUUTMA & KRISTEL UIBOAED FREQUENCY AND LANGUAGE CONTACTS

language area has been described as a complex linguistic community where mutual influences of individual languages are ongoing and there is no single dominating language; therefore, it is also referred to as a contact superposition zone rather than a Sprachbund (Koptjevskaja- Tamm & Wälchli 2001: 626; Wälchli 2011: 325). There are no iso­ glosses covering all the Circum-Baltic languages (Koptjevskaja- Tamm & Wälchli 2001: 728–732). Estonian belongs to the eastern part of the Circum-Baltic lan­ guages together with the Baltic, Finnic, and (es­ pecially Russian). Although these languages belong to different language families, many morphosyntactic similarities have been attested among them (see, e.g., Dahl & Koptjevskaja-Tamm 2001; Klaas-Lang & Norvik 2014; Serzant 2012, 2015; Vaba 2011). All of the above-mentioned studies are characterized by the fact that at least when addressing the morphosyntax, Estonian is treated as a whole and the intralanguage variation is not accounted for. How­ ever, there are a multitude of phenomena in Estonian displaying significant regional differences that can be explained by contacts that have taken place locally, that is, they have affected only certain areas. Thus, for example, due to the Swedish influence in Swedish contact areas the use of the nominative subject instead of the parti­ tive has increased (Juhkam 1998, 2012), but the same has also been observed in the bilingual Estonian-Russian region in Iisaku4 (Must Map 1. The dialect classification in the Corpus of Estonian Dialects. Dialect names 1965). Swedish influence has increased the use of the nominative are highlighted in larger font. The map is the modified version of the parish map object (Juhkam 1983), and again the same tendency has been noted (EKI 2014). Place names marked on the map are referred to later in the text. in Iisaku (Must 1965), where the Swedish influence could not be the main factor addressing this change. Thus, similar phenomena may occur in a number of regions, but the languages driving these devel­ has been stronger on some areas than on others. Still, we can infer opments may vary. from geographical and historical facts that German influence has most The primary contact languages for dialects are Swedish, Latvian, likely been weaker in the southeastern part of Estonia, e.g., the Seto Russian, and Votic, to a lesser extent also Finnish and Ingrian, in the dialect area, due to its intense connections to Russian-speaking . earlier period also Livonian. These languages have a clear regional The Circum-Baltic language area includes the Finnic languages, spread, long-term contacts, and bilingual speakers in corresponding Russian, the Baltic languages (Latvian and Lithuanian), the Germanic areas. We now turn to the individual contact languages and briefly languages (particularly Swedish, German, Low German), and several describe the history of these contacts. other languages spoken in the region, see the review by Koptjevska­ 4. Place names mentioned here and in the rest of the article are marked on Map 1 ja-Tamm & Wälchli (2001) and Wälchli (2011). The Circum-Baltic in smaller font.

160 161 LIINA LINDSTRÖM, MAARJA-LIISA PILVIK, ON THE USE OF PERFECT AND PLUPERFECT IN ESTONIAN DIALECTS: MIRJAM RUUTMA & KRISTEL UIBOAED FREQUENCY AND LANGUAGE CONTACTS

Swedish. Estonian has had long-standing contacts with Swedish Latvian. Estonian-Latvian language contacts have been long and the Scandinavian languages, but more intensive and direct contacts and many corresponding structures can be found in both languages. were established from 12th century, when the Swedes moved to Es­ Latvian influence is the strongest in the Estonian enclaves in Latvia tonia’s western and northern coasts. Since then, the contacts between (Leivu, Lutsi) and in the dialect areas near Estonia’s southern state Swedish and Estonian have been strong and persistent, indicated by border (Vaba 2011). The number of Latvian loanwords is the highest both the number of loanwords as well as the mutual interactions in the in the Võru and Mulgi dialects as well as in Saarde and Häädemeeste pronunciation and morphosyntax (Juhkam 1998). Expanding during (the southern part of the Western dialect) (Vaba 1997). the Medieval period and reaching its peak in the 15th century, Swedish Latvian contacts are also noticeable on the islands; this has been settlement was continuous between Ruhnu and Viimsi (Juhkam 1992). described as the Sprachbund surrounding Irbe Strait – more specifi­ Until World War II, the Swedish-speaking population of western cally common traits have been attested in western Saaremaa, western Estonia (Vormsi, Noarootsi, Riguldi, Osmussaar), northern Estonia Hiiumaa, the Latvian Dundaga dialect, the Ruhnu , (Vihterpalu, Kurkse, the Pakri Islands, Tallinn), and on Ruhnu Island and Livonian. It is likely that Livonian has been an intermediary lan­ was rather stable. Estonian and Swedish dialect contacts, therefore, par­ guage at least on the lexical level. (Vaba 1997: 475–479.) ticularly affected the islands and the Western dialect area (mainly the Votic, Ingrian, and Finnish. Votic influences are apparent western Estonian coastal parishes) as well as the northwestern part of in the Eastern (Pall 1982; Univere 1988) and Northeastern dialects Estonia (Risti, Harju-Madise, Keila Parishes) (Juhkam 1998: 26–29). (Pall 1982; Alvre 2000). Pall (1982) explains Votic traits in the Eastern During World War II, most of the Estonian Swedes moved to Sweden. dialect with direct contacts that have arisen as a result of the Votes’ Russian. The most long-standing Russian influence has been immigration. Alvre (2000) has listed common features in the North­ present in southeastern Estonia, which has long been in the sphere of eastern dialect and Votic and highlights various phonetic and gram­ influence of Pskov. The main Russian influence has been attested in matical similarities. Ariste has foregrounded Vaivara Parish as a com­ the Seto area, which is visible in the material culture as well as in lexi­ plex contact area where the eastern part has been exposed to a number cal loans (Must 2000), pronunciation (Pajusalu 1999), and grammar of other Finnic languages – Votic, Ingrian, and Ingrian Finnish. The (Kask 1984: 27). Votic-­Ingrian-Finnish mixed villages have been situated on the other Close and long-term contacts have also been present in the North­ side of the Narva River. (Ariste 1962.) Contacts between the North­ eastern dialect, in the area of Iisaku, where during the 17th–18th cen­ eastern dialect of Estonian, Votic, Ingrian, and Ingrian Finnish have tury there arose a Russian-Estonian bilingual population (Must 1965, implied many loanwords from one Finnic language to another in the 2000). There has been long-term Russian settlement on the west coast area (see Björklöf 2019, in this volume). of Lake Peipus, which dates back to the end of the 16th century, when Based on the distribution of Finnish loanwords, it has been ar­ a large number of Russians migrated to Estonia. At the end of the 17th gued that the Finnish influence is evident also in some parishes of the century, most of the current villages already existed. In the beginning Coastal dialect, mainly Kuusalu, Haljala, Jõelähtme (e.g., Söderman of the 18th century, Old Believers, who had left the Russian Ortho­ 1996; Norvik 2000; Björklöf 2012, 2018). This is a result of inten­ dox Church, moved to Estonia and the former Estonian villages on sive oversea trading and fishing contacts (Must 1987: 13–15; Björklöf the coast of Peipus became Russian villages. (Must 2000: 8.) Russian 2012, 2018). contacts on the west coast of Lake Peipus, therefore, are not as long- standing as in area; however, they have been intensive and influenced especially the Eastern dialect of Estonian and the southern part of the Northeastern dialect.

162 163 LIINA LINDSTRÖM, MAARJA-LIISA PILVIK, ON THE USE OF PERFECT AND PLUPERFECT IN ESTONIAN DIALECTS: MIRJAM RUUTMA & KRISTEL UIBOAED FREQUENCY AND LANGUAGE CONTACTS

3. Perfect (and pluperfect) Thus, Swedish has developed the so-called HAVE-perfect or the pos­ in the contact languages sessive perfect, which is characteristic of the European linguistic area (Haspelmath 2001; Heine & Kuteva 2006), but the former BE-perfect Next we will examine the use of compound tenses in the contact lan­ is still preserved to some extent. Also, in Estonian Swedish dialects guages in order to identify the nature of their potential effects on Es­ the ha(va)-perfect is more frequent than the vara-perfect (Henrik tonian dialects. Rosenkvist, p.c.). As perfect and pluperfect are extensively used in Swedish. In modern Swedish, the perfect is mainly formed with Swedish and they have a wide functional range, one could assume that an auxiliary verb ha(va) ‘to have’ and the main verb in its supine form, the contact with Swedish has increased the use of compound tenses which historically is the neuter form of the perfective participle (Dahl also in the contact area. 1995: 61), examples (3)–(4). Additionally, a more constrained perfect- Russian. Russian has no special form for either the perfect or like construction is comprised of the auxiliary vara ‘to be’ and the pluperfect, but only one finite past tense, which historically is a form perfective participle (example 5). In modern Swedish, the use of the of the participle (Tommola 2000: 442). General past has functions that construction with vara is restricted and it expresses mainly resultativ­ are filled by the perfect or pluperfect in other languages. Perfect has, ity. In older Swedish, vara as an auxiliary was more widespread com­ however, existed in Old Church Slavonic (in example 6) where it was pared to contemporary Swedish, and its use was less limited. (Larsson formed with the copula ‘be’ and the participle. 2009: 156.) The gradual disappearance of the perfect formed with the (6) Old Church Slavonic (Tommola 2000: 459) vara-verb is related to the development of the perfect construction th čast’ post-a preminu-l-a est’ with the ha(va)-verb: vara dominated until the 17 century. A remark­ part fast-gen pass:pfv-pp-f be.3sg able decline in its frequency of use occurred sometime around the turn ‘A part of the fast-time has passed.’ of the 17th–18th century, when the ha(va)-perfect started to become significantly more frequent. (Larsson 2009.) Additionally, Russian has developed perfect-like constructions: a re­ (3) Swedish (Larsson 2009: 144) sultative passive construction (in example 7) that, when occurring Han har kommit hit. he have:prs come:sup here with the past copula, is ambiguous between a perfect (dynamic) and ‘He has come here.’ a present resultative (stative) reading (Tommola 2000: 463), and the so-called possessive perfect (in example 8), in which the actor is ex­ (4) Swedish (Larsson 2009: 144) pressed with the construction u + GEN (u syna ‘son’). This construc­ Artikeln har blivit färdig. tion is particularly common in the northern and northwestern Russian article:def have:prs become:sup ready:sg dialects, and it has spread also to the Finnic languages and Latvian ‘The article has become ready.’ (Seržant 2012). In Estonian, the corresponding construction is formed with an adessive agent, the verb olema ‘to be’, and a passive past par­ (5) Swedish (Larsson 2009: 272) ticiple (Lindström & Tragel 2010; Lindström 2015), more frequently De är redan anlända. they be:prs already arrive:ptcp.pl used in southern and eastern Estonia (Uiboaed 2013: 176, ex. 9), i.e., ‘They have already arrived.’ in the contact area of Russian.

164 165 LIINA LINDSTRÖM, MAARJA-LIISA PILVIK, ON THE USE OF PERFECT AND PLUPERFECT IN ESTONIAN DIALECTS: MIRJAM RUUTMA & KRISTEL UIBOAED FREQUENCY AND LANGUAGE CONTACTS

(7) Russian (Tommola 2000: 463) As the Latvian perfect and pluperfect are formally and functionally Magazin by-l otkry-t (v vosem’). quite similar to those in Estonian, it can be assumed that the use of the shop be-pst.m open:pfv-ppp (prep eight) perfect and pluperfect in the contact area is more or less the same as ‘The shop was open/opened (at eight o’clock).’ the Estonian average or has increased the use of compound tenses in the Latvian contact area of Estonian. (8) Russian (Tommola 2000: 464) Votic, Ingrian, and Finnish. In all Finnic languages, includ­ U syn-a institut za-konč-en/-o. ing Votic, Ingrian, and Finnish, the perfect and pluperfect are both prep son-gen institute pfv-finish-ppp:m/-nt formed with the verb ‘be’ and the past participle. ‘My son has graduated from the institute.’ In Votic, it has been noted that compound tenses are rare and in­ (9) Estonian (Seto dialect) stead the simple past tense is used. (Ikola 1960; Markus & Rozhanskiy Salmõ-l oll’ ka ostõ-t purgitäis me-tt 2011: 233). The tendency seems to apply when observing Votic texts Salme-ade be.pst.3sg also buy-ppp jarful honey-prt that are incorporated into the Corpus of Estonian dialects. ‘Salme had also bought a jarful of honey.’ Ikola (1960) has argued that the development of the Finnish tense system (including the perfect and pluperfect) in the old literary The focus of the current article lies on olema + nud-participle com­ language has been influenced by Swedish. However, a comparative pounds and therefore the passive constructions (including the posses­ study on the use of the perfect, based on translations of the Bible, has sive perfect) are not discussed further. revealed that in the first Finnish translations of New Testament texts Russian lacks perfect and pluperfect forms and uses a general (1776), the perfect occurs less frequently than in the Swedish version, past instead, which leads us to assume that similar tendencies appear and its functions also differ significantly from the Swedish perfect also in the Russian contact areas: simple past is used instead of perfect (Swedish representing a typologically “usual” use of perfect) (Dahl and pluperfect and the compound past tenses have a lower usage fre­ 2014). Following that, Finnish dialects that have been in contact with quency in comparison to other areas. Estonian dialects may have used compound tenses even less than what Latvian. The Latvian perfect is formed by much the same is attested in the Swedish-influenced translations of the Bible. means as the Estonian perfect, i.e., using the auxiliary būt ‘to be’ and As the Votic, Ingrian, and Finnish forms of the perfect and plu­ the active past participle, except that the participle agrees with the perfect are similar to Estonian, it can be expected that contacts with subject in both number and gender (example 10). The auxiliary verb those languages have not significantly affected the use of the perfect can occur either in present, past, or future form. In the present per­ and pluperfect in the contact area. fect (the auxiliary is in the present tense) when the subject is in third person singular, the auxiliary verb can be omitted from the sentence. (Kalnača 2015: 82.) (10) Latvian (Kalnača 2015: 82) Šodien Bille ar vecomāti ir gājušas ilgi. today Bille with grandmother:ins.f be.prs.3 walk:ptcp.pst.f long ‘Today Bille and her grandmother have been walking for a long time’

166 167 LIINA LINDSTRÖM, MAARJA-LIISA PILVIK, ON THE USE OF PERFECT AND PLUPERFECT IN ESTONIAN DIALECTS: MIRJAM RUUTMA & KRISTEL UIBOAED FREQUENCY AND LANGUAGE CONTACTS 4. The usage frequency of compound past tenses in Estonian dialects

4.1. Data and method

The current study presents frequency data from the Corpus of Esto­ nian Dialects (CED). The role and impact of frequency on the devel­ opment and change of linguistic phenomena has been studied quite thoroughly in recent decades (see, e.g., Arppe et al. 2010; Divjak & Gries 2012; Gries & Divjak 2012). The importance of frequency has been in central focus especially in usage-based approaches to lan­ guage (Barlow & Kemmer 2000; Bybee 2010). In studies concern­ ing dialect syntax, it has been found that syntactic variation usually cannot be described via categorical distinctions (the phenomenon X occurs in some area and Y does not), but rather in terms of frequen­ tial typicalities and atypicalities (Kortmann 2010). Moreover, text fre­ quencies have been claimed to reflect the perception of the language system more adequately than for instance the atlas data used widely in traditional dialectology (Szmrecsanyi 2013: 4). Frequency effects Map 2. Word frequencies per parish in dialect data. White indicates the absence of can hardly be ignored when we talk about language contacts: regional data from that parish. idiosyncrasies can be exposed by frequency. The usage pattern al­ ready existing in a language (or dialect) can acquire a higher usage retrieval (01.10.2014) the corpus contained 834 311 morphologically frequency due to a similar pattern in a contact (or model) language annotated words. The overall number of speakers in the data was 389 with which the speakers are constantly confronted (Heine & Kuteva and the number of interviews (= files) of various length was 323. 2005: 47). Even if a particular phenomenon exists in several areas, its The data were automatically retrieved with R (R Development usage frequency can be substantially lower outside the core of the ob­ Core Team 2013) scripts, searching for the lemma of the verb olema served contact region (see Koptjevskaja-Tamm & Wälchli 2001: 627). with a 7-word context from both left and right of the verb. Only obser­ Change in the usage frequency of some (previously existing) pattern vations with the -nud participle within the given context were selected has been considered the typical contact-induced grammatical transfer for further inspection. These criteria were met by 13 385 context units, (Heine & Kuteva 2005: 48). which in turn were inspected manually for true compound past tenses. We now turn to the usage frequency of perfect and pluperfect The final dataset for further analysis contained 6 242 utterances with tense forms in the CED, which consists mainly of spoken dialect in­ either perfect or pluperfect forms. terviews recorded in 1960–1970s. Speakers of the CED have been The CED does not cover all Estonian parishes and therefore the mostly elderly people considered to be good representatives of the research data are not fully representative of the whole country. Map 2 old local dialects. (For an overview of the CED, see Lindström et al. presents the parishes which are covered with data for the present anal­ 2019, in this volume.) The research material has been collected from ysis. The darker the region, the more material has been included from the morphologically annotated part of the corpus; at the time of data that area. White areas indicate parishes where no data was available.

168 169 LIINA LINDSTRÖM, MAARJA-LIISA PILVIK, ON THE USE OF PERFECT AND PLUPERFECT IN ESTONIAN DIALECTS: MIRJAM RUUTMA & KRISTEL UIBOAED FREQUENCY AND LANGUAGE CONTACTS

In the present study we examine the usage frequency of the com­ Absolute Normalized pound tense forms consisting of the verb olema ‘to be’ and the -nud Dialect Words frequency frequency participle. Different moods have not been distinguished except in Sec­ Coastal 51 667 316 510 tion 5.1. where we only discuss examples of indicative mood in order Eastern 45 280 339 625 to compare the proportions of perfect and pluperfect; however, moods Insular 166 898 1 723 861 other than indicative are used infrequently and thus do not consider­ Mid 130 086 860 552 ably distort the overall picture. At this stage, the functions of the com­ pound past tenses have not been tracked separately. We have, however,­ Mulgi 63 516 617 810 differentiated affirmative and negative sentences as well as perfect and Northeastern 47 660 193 338 pluperfect in order to explore the patterns which potentially affect the Seto 39 175 123 262 overall usage frequency of the phenomenon in the data. Tartu 65 591 428 544 We carried out analyses on three levels: dialects, parishes, and Võru 70 038 486 579 villages. Since the number of words from different dialects is repre­ sented unevenly in the corpus we utilized frequency normalizations Western 154 400 1 157 625 in order to level out frequency fluctuations and achieve comparability Table 1. The absolute and normalized frequencies of the compound tense con- across areas. In the first two analyses (dialects and parishes), we use structions across dialects. the mean size of the corpus as a normalization base5.

4.2. Results 800 700 600 500 4.2.1. Dialects 400 300 First, we examine the usage frequencies of the compound past tenses across dialects. Table 1 presents the absolute and normalized frequen­ cies of compound tense forms. Here, the frequencies are summarized within a single dialect, i.e., the total number of occurrences across all the parishes belonging to the respective dialect is presented. The normalized frequencies, therefore, reflect the usage frequencies of the compound past tenses as if the amount of annotated words from all the dialects were equal. Map 3 gives the same data spatially. Darker areas indicate higher frequency and lighter areas indicate lower frequency.

5. The formula of normalization: (form frequency/words in dialect or parish)*corpus Map 3. Normalized frequencies of compound tense constructions in the dialect mean. data.

170 171 LIINA LINDSTRÖM, MAARJA-LIISA PILVIK, MIRJAM RUUTMA & KRISTEL UIBOAED

Both the table and the map reveal that the usage frequency of the com­ pound tenses is particularly high in the Insular and Mulgi dialects, whereas in the Seto dialect it is the lowest. The frequency is low also in the Northeastern dialect. Thus, the differences between the dialects are remarkable and at least partially correspond to the language con­ tact situation: high frequency in the contact area of Swedish (Insular dialect), low frequency in the contact area of Russian and the Finnic languages (Seto, Northeastern dialects). However, the Eastern dialect also has had long-standing contacts with Russian as well as with Votic, which, however, are not reflected in decreased use of perfect and plu­ perfect. Likewise, the high frequency of compound tenses in the Mul­ gi area is hard to explain with Latvian contacts as we lack evidence concerning the higher prevalence of compound tense usage in Latvian compared to Estonian. However, based on current data and taking into account that there exist corresponding compound tenses in Latvian, we may conclude that the influence of Latvian may be responsible for increased use of compound tenses in Mulgi. Map 4. Normalized frequencies of compound tense constructions across parishes.

4.2.2. Parishes

We now turn towards a lower level of generalization, and have a look at the compound past tense frequencies by parish. The sparseness of the data in the CED prevents us from getting a more coherent geo­ graphical picture: the recordings do not cover all parishes and the map, therefore, is fragmented by white areas, indicating the absence of data from those parishes (see also Map 2 above). Next, we take a look at the normalized frequencies, this time the normalization base is the mean number of words from parishes. Map 4 does not show clearly distinguishable larger coherent areas, but simi­ larly to dialects, the construction frequency is higher on the islands, particularly in Hiiumaa and in the southwestern part of Saaremaa. On the mainland, the usage frequency of compound past tenses is higher in Karksi Parish, which belongs to the Mulgi dialect, and in Urvaste Parish in the Võru dialect region.

172 Map 5. Distribution of data points across villages (Google (2016) as a base map). LIINA LINDSTRÖM, MAARJA-LIISA PILVIK, ON THE USE OF PERFECT AND PLUPERFECT IN ESTONIAN DIALECTS: MIRJAM RUUTMA & KRISTEL UIBOAED FREQUENCY AND LANGUAGE CONTACTS

4.2.3. Villages frequency is rather high also in Halliste, Rõngu, and Urvaste, which could be indicating that we are not dealing with the idiosyncratic be­ Next, we will narrow the geographical unit and take a look at individual havior of only one parish (or one village/speaker). One could assume villages. This time we use absolute frequencies per village and visualize that the usage frequency of the compound tense forms has been influ­ the density of data points on Map 5 (see the previous page). The inter­ enced by contacts with Latvian where similar constructions exist. pretation of the map is similar to one known from regular geographical The Seto and the Northeastern dialects are low frequency areas. maps visualizing landscape surfaces. The number of lines on the map Low frequency in Seto can be explained by the long-term contacts indicates the proportion of constructions in the region. The higher the with Russian, which lacks the corresponding tense constructions. His­ concentration of lines, the more data points (i.e., occurrences of com­ torically, Seto has had less contacts with the German-speaking upper pound tenses) we have in our dataset for that particular area (these areas class, which definitely has reduced the overall German influence in are “higher”). The data reveal that the frequency of the compound tens­ that area and, which could serve as an additional factor leading to es is very high in the southwestern part of Saaremaa, northern Saare­ less frequency in compound tense usage. In contrast, the second Rus­ maa, and western Hiiumaa (Reigi Parish), i.e., in the Insular dialect. sian contact area – the Eastern dialect – exhibits rather the opposite The second distinctive core area is formed by the villages in the Karksi tendencies (Map 3). The newer and more restrained nature of Rus­ Parish, in the Mulgi dialect area, where the frequency of compound sian contacts in the Eastern dialect could have prevented the extensive tenses is very high, and in the Tartu dialect (Nõo and Rõngu Parishes). structural impact on the dialect. When interpreting­ the map, however, one has to keep in mind that the The Northeastern dialect is characterized as a contact area with construction frequencies depicted on the village map are not normalized Russian (especially Iisaku Parish, see Must 1965), on the one hand, and and the density of data points is at least partially also affected by the with other Finnic languages (Votic, Ingrian, Finnish), on the other hand. amount of raw data available from different regions (see Map 2). The usage frequency of the compound tenses is relatively low in the Northeastern dialect. Due to the complex past historical and geographi­ 4.2.4. Language contacts and the usage of cal situation in this dialect area it is difficult to assess whether the rare compound past tenses: summary use of compound tenses is a result of language contacts or some other factors, since compound tenses do exist in neighboring Finnic languag­ When the information from the three maps is converged, differences es. The contacts with Russian have taken place in a relatively restricted in usage frequencies of the compound tenses imply the influence of in­ area and thus have probably not influenced the whole region; however, dividual contact languages. Frequency of the compound past tenses is the Russian influence may have come via neighboring Finnic languages higher in areas where the Swedish language contacts have been stronger since there has been a strong influence of Russian especially on Votic and and longer; particularly in the Insular dialect (in the southwestern part Ingrian. Thus, the low frequency of compound tenses in the Northeastern of Saaremaa and the western part of Hiiumaa). The parish map (Map 4) dialect can be a result of contacts with other Finnic languages, indirect shows that the frequency of compound tenses is fairly high on the entire Russian influence, or it may be an indication of weaker influence from island of Hiiumaa, especially in Reigi Parish, a major Swedish populated Germanic languages, i.e., this area is more conservative in this respect. region in the past. Relatively high construction frequency in the Western We can thus conclude that the language contacts have influenced dialect also illustrates the influence of Swedish, although the smaller areas the usage frequency of the compound tenses in Estonian dialects. with long-term Swedish settlement were not included in the analysis. Intense contacts with Swedish have been constant on the northwest Surprisingly, the Mulgi dialect (especially Karksi) stands out coast and on the islands since the 12th century (or earlier), active over­ with a very high frequency of compound tenses (Maps 4 and 5). The seas communication is known also from earlier times. The increase in

174 175 LIINA LINDSTRÖM, MAARJA-LIISA PILVIK, ON THE USE OF PERFECT AND PLUPERFECT IN ESTONIAN DIALECTS: MIRJAM RUUTMA & KRISTEL UIBOAED FREQUENCY AND LANGUAGE CONTACTS usage frequency in the islands and in western Estonia can therefore be across dialects. The proportion of the pluperfect is highest in the Mulgi attributed at least partly to long-term contacts. The Seto area has long dialect (53% of all compound tense constructions). Other South Esto­ contacts with Russian, whereas similar contacts in the Eastern dialect nian dialects (Võru and Tartu) have a higher proportion of pluperfects, as region on the coast of Lake Peipus have been significantly shorter. well, implying that there might be a wider functional range for pluperfect This could possibly explain why the usage frequencies of the com­ in this region. Seto, however, differs from other southern dialects with a pound tenses differ significantly in the two regions with Russian con­ lower proportion of pluperfect tense usage. As Table 1 and Map 3 already tacts. Another possible explanation that applies more to the Seto and illustrated, Seto makes less use of the compound tenses in general. the Northeastern dialects is that both regions could possibly have been Dialect Perfect Pluperfect Pluperf (%) much less affected by German than other areas in Estonia. Mulgi 263 301 53 Võru 226 219 49 Tartu 206 143 41 5. Other factors influencing Eastern 190 93 33 construction frequency Western 697 325 32 Mid 556 220 28 In addition contacts, the overall increase in frequency Coastal 198 76 28 is associated with other factors, as well. In our data, two areas were Northeastern 124 42 25 Seto especially highlighted with high frequency: the Mulgi (especially 91 31 25 Insular 1 327 298 18 Karksi Parish) and Insular dialects. In both areas, the perfect and plu­ perfect have a variety of different functions, which explains the higher Table 2. Proportions of indicative perfect and pluperfect forms in the dialect data6. usage frequencies. In the next sections, we inspect differences between the use of 1) perfect and pluperfect, 2) negative and affirmative claus­ es to explore the potential influence of these aspects on frequency.

5.1. Perfect vs. pluperfect

While a typical perfect expresses a past event from the present point of view, the reference point of the pluperfect is in the past, i.e., it describes a past event from the point of view of an observation time which could be also in the past. One of the most salient functions of the pluperfect is its quotative (evidential) use. In contemporary (spoken) Estonian, the pluperfect is frequently used in reported narratives whereas the perfect is rather rare in this function. (Lindström & Toomet 2000.) To examine the proportions of the two compound tenses we only Figure 1. Proportions of indicative perfect and pluperfect forms in the dialect data. included the instances which were morphologically marked for indica­ tive mood, since only indicative makes a distinction between past tense forms (imperfect, perfect, and pluperfect) (Erelt et al. 1995: 74–79). 6. Note that numbers in this table differ compared to other tables, as only indica­ Table 2 and Figure 1 present the distribution of the perfect and pluperfect tive mood is included to calculate the proportions of perfect type.

176 177 LIINA LINDSTRÖM, MAARJA-LIISA PILVIK, ON THE USE OF PERFECT AND PLUPERFECT IN ESTONIAN DIALECTS: MIRJAM RUUTMA & KRISTEL UIBOAED FREQUENCY AND LANGUAGE CONTACTS

Compared to other dialects, Mulgi exhibits a disproportionately high (12) Võru, Põlva Parish use of pluperfects, which express (the reported nature of ja tul-nuq sõss hainaruga (.) läve-st sisse ja the information) and are particularly characteristic of reported narra­ and come-app then haystack threshold-ela in and tive texts (example 11). ast-no sinnäq sängü ette ja ütel-nü et (...) step-app there bed.gen ahead and say-app that (11) Mulgi, Karksi Parish kule Jaan’ et (.) miu vanaemä olli elä-nu (.) vanast Püügle listen.imp.2sg Jaan that I.gen grandmother be.pst.3sg live-app formerly Püügle ‘And (it has been told that) the haystack came in across the Savimäe-l ja (.) ja olli asute-n sõnna threshold and stepped by the bed and said that “listen, Jaan” /…/’ Savimäe-ade and and be.pst.3sg establish-app there ende-l (.) maja tetä ja (...) Much as in other areas, in Mulgi the bare past participle has a high self-ade/all house.gen do.inf and usage frequency; however, it often alternates with the pluperfect and ‘My grandmother lived in Püügle Savimäe and perfect tenses. In example (13), the narrative starts with the pluperfect had established herself a house there’ (first line), followed by a bare past participle (line 2,tahten ‘want-app’), which is replaced with a perfect in its negative form, and later the The abundance of the quotative uses of pluperfect in the Mulgi area speaker continues again with a bare past participle. Similar patterns is at least partially related to the fact that the Mulgi sample contains can be found in other dialects, as well, but for some reason the Mulgi more narrative texts (such as legends and reported narratives) than dialect clearly prefers the pluperfect for expressing evidentiality. general descriptions of everyday work and customs, which are char­ (13) Mulgi, Karksi Parish acteristic of most CED texts. Among the texts from Karksi, there are nuu vanast olli (.) talu esä ollu ja kangest two recordings collected by Paul Ariste in 1938, which mainly contain ptcl earlier be.pst.3sg farm.gen father be.app and very relatively short reported narratives (stories). The expansion of usage rikass (...) siss tahte-n (.) et eij ol tahte-n frequency may therefore be related to the higher-than-average amount rich then want-app that not be.cng want-app of narrative texts, which motivate the frequent use of the pluperfect for las-te-l se-dä raha (.) anda (.) ja mõtel-nu evidential purposes. child-pl-ade/all that-prt money.prt give.inf and think-app However, the large number of reported narratives in the data does t (.) ma mata ta maha või kohekil äräde that I bury.1sg it.gen ground.ill or somewhere pfv not alone explain the general frequent use of pluperfect in Mulgi or in seant ihne ollu kangest (...) other South Estonian dialects (Tartu and Võru) since narrative texts are such stingy be.app very not exceptional in other areas included in the sample. Moreover, the ‘Earlier, there was a farmer and (he was) very rich. Then (he) wanted pluperfect is not the only means for marking reported evidentiality and (repairs:) didn’t want to give the money to (his) children, and thought therefore it is not clear why the Mulgi dialect prefers the pluperfect to that “I will bury it in the ground or somewhere”, (he) was so stingy’ such an extent for expressing this meaning. Estonian has a variety of evidential constructions conveying evidential meaning, e.g., quotative A previous study on synthetic and analytic quotative constructions in mood, bare past , infinitives, evidentialpidama ‘must’ + in­ the Corpus of Estonian Dialects has also noted that quotative construc­ finitive constructions, etc. (Kask 1984; Kehayov 2008). The useof tions in general are quite frequent in Mulgi data and analytic quotative bare past participles (example 12), in particular, is common across all constructions (mainly indicative perfect and pluperfect) in particular of Estonia (Pilvik & Uiboaed 2014). are used there more frequently than in other dialects (Pilvik & Uibo­ aed 2014).

178 179 LIINA LINDSTRÖM, MAARJA-LIISA PILVIK, ON THE USE OF PERFECT AND PLUPERFECT IN ESTONIAN DIALECTS: MIRJAM RUUTMA & KRISTEL UIBOAED FREQUENCY AND LANGUAGE CONTACTS

The main contact language of the Mulgi dialect region, Latvian, has a special mood for marking reported evidentiality (oblique mood) as well as bare past participles that are used for the same purpose (Kalnača 2015; Muižniece, Metslang & Pajusalu 1999). The contacts with Latvian may have reinforced the need for expressing reported­ ness more generally in South Estonian, especially in the Mulgi area. The alternation between the compound tenses and bare participle is also common in Latvian (see Siegl & Kehayov 2006 for discussion on the relationship between the bare participle and compound tenses).

5.2. Affirmative vs. negative constructions

The data analysis revealed that there are remarkable differences in proportions of negative compound tenses between dialects. Figure 2. Proportions of affirmative and negative compound tense forms in the The inspection of the ratio between affirmative and negative utter­ dialect data. ances (Table 3 and Figure 2) reveals that the Insular dialect exhibits con­ siderably more negative sentences among the olema and -nud compounds Closer inspection of negated perfect cases in the Insular dialect shows (47%) than other dialects. The number of negated sentences is slightly that the use of this construction has been functionally extended and higher also in the Western dialect (27%). Higher usage frequency of the does not express only the typical meaning for perfects. In example compound tense forms in the Insular and Western dialects is, therefore, at (14), the first two affirmative clauses are in the simple past, while for least partly related to their increased use in negative sentences. the following negated clause the perfect is used instead of the simple past (as it is in Standard Estonian). In this context, however, the ne­ Dialect Affirmative Negative Negation (%) gated clause may be interpreted also as an experiential perfect, but Insular 912 811 47 such an interpretation is not evident in all cases. In example (15), the Western 838 319 28 first affirmative clause (kis juuwa täis mees oli ‘who was a drunk­ en man’) employs the simple past, while in the following negative Tartu 334 94 22 clause, the perfect is used (see pole sääld tulngid ‘that didn’t leave Eastern 266 73 22 from there’), followed by the affirmative clause again in the simple Mid 676 184 21 past (see=li teise umiguni veel ja ‘that was until next morning’). There Coastal 257 59 19 is no evident functional reason (other than negation) to use perfect in Võru 397 89 18 the second clause; in standard Estonian, in this context the simple past Mulgi 516 101 16 would be used. Thus it seems that for negation, often the perfect is preferred in contexts where (at least in standard Estonian) the simple Northeastern 166 27 14 past occurs. Seto 109 14 11 Table 3. The proportion of negative sentences among the uses of compound past tense forms.

180 181 LIINA LINDSTRÖM, MAARJA-LIISA PILVIK, ON THE USE OF PERFECT AND PLUPERFECT IN ESTONIAN DIALECTS: MIRJAM RUUTMA & KRISTEL UIBOAED FREQUENCY AND LANGUAGE CONTACTS

(14) Insular, Jämaja Parish (703 cases out of the 811 negated sentences, i.e., 87%); in comparison, moo isa surr-i nuoorelt ää (.) naad ela-sid neli the corresponding proportions in the Western dialect were 123 out of I.gen father die-pst.3sg young pfv (.) they live-pst.3pl four 319 (39%). Koit (1963) has referred to the negation word pole (his­ oasta-d kuoos (...) rohkem naad põle ela-n torically: ep ‘not’ ole be.cng) as a highly productive means of nega­ year-prt together more they neg.be.cng live-app tion in the Insular dialect. Also, Map 6, combining Andrus Saareste’s ‘My father died young. They lived four years (1955) data on the alternation of ei ole and pole in Estonian dialects together, they didn’t live any more’ (displayed utilizing symbols) and dialect corpus frequency data on the (15) Insular, Emmaste Parish occurrences of pole from a study on analytic and synthetic parallel kiss juuwa täis mees ol-i (.) see pole sääld constructions (Uiboaed & Pilvik 2014), show the clear preference for who drink.inf full man be-pst.3sg this neg.be.cng there_from pole in the Insular dialect. tul-n-gid (.) see= li teise umigu-ni veel ja come-app-cli this be-pst.3sg another morning-ter still and ‘Who was a drunken man, that didn’t leave from there, that was until next morning’

The negative perfect also alternates with a bare participle, which is used as an evidential strategy in Estonian. In example (16), in the first clause only a bare participle appears, but the second clause (syntacti­ cally coordinated with the first clause) uses the perfect tense. The bare participle does not have a negative counterpart of its own; in the nega­ tive clause, however, the simple past could be expected instead of the negative perfect in standard Estonian. (16) Insular, Kihnu Parish aga rahvas ol-n ikka jõrmu täüs ja but people be-app ptcl fear.prt full and põlõ toeht-n mitte uata-ma tulla neg.be.cng may-app neg look-sup come.inf ‘But people were very afraid and were not allowed to come and look’

It is not clear why the compound tenses are used more often in negated sentences in the Insular and Western dialects, or whether these tenden­ cies are driven by contact-induced extension of the compound tense functions or by a language-internal development. Information about clear parallels in the main contact language – Swedish – is not avail­ able to the authors. However, it is worth noting that the vast majority of the negative compound tense forms in the Insular dialect are formed with the synthetic negation word pole ‘is not’ (as in examples 14–15) Map 6. The spread and frequency of pole and ei ole in Estonian dialects. 

182 183 LIINA LINDSTRÖM, MAARJA-LIISA PILVIK, ON THE USE OF PERFECT AND PLUPERFECT IN ESTONIAN DIALECTS: MIRJAM RUUTMA & KRISTEL UIBOAED FREQUENCY AND LANGUAGE CONTACTS

This leads us to assume that in the Insular dialect, the compound 6. Summary pole + -nud has generalized into the regular negative construction re­ ferring to the general past also in contexts where otherwise the simple The perfect and pluperfect as contact-induced categories exist in all past would be used. The former past tense negative construction in Estonian dialects, but differences emerge in their usage frequency. By the Insular dialect is formed with the negation word es and the con­ examining the data in the Corpus of Estonian Dialects on three levels negative verb form (e.g., es ole ‘was not’), but its use has somewhat (by dialects, parishes, and villages), we found that high compound diminished (especially in Hiiumaa, Koit 1963), and given way to the tense frequency is particularly characteristic of the Insular and Mulgi general North Estonian ei ‘no(t)’ + -nud participle compound, e.g., ei dialects (especially in Karksi Parish), low frequency, in turn, is com­ ol-nud ‘was not’ (Lonn & Niit 2002) and pole + -nud participle (pole mon in the Seto and Northeastern dialects. These areas have also had ol-nud ‘has not been; was not’). One can argue that the changes of intensive local language contacts: the Insular dialect with Swedish, negation in the tense system have lead the pole + -nud compound to the Mulgi dialect with Latvian, the Seto dialect with Russian, and the express general past tense negation. This has potentially given rise to Northeastern dialect area with other Finnic languages and, to some the increased frequency of compound tense usage in the Insular dia­ extent, also with Russian. Thus, the contact language influences are lect. One might also be tempted to connect higher usage frequency of at least partially reflected in the frequency: contacts with languages, the negative compound tenses to the lower proportion of pluperfects in which make extensive use of similar categories (Swedish, Latvian) the data, as seems to be the case at least for the Insular dialect where have reinforced the use of compound past tenses in contact regions; the proportion of pluperfect forms is relatively low compared to other languages, which either only have simple past forms (Russian) or in dialects (as indicated in Table 2 and Figure 1). However, negative plu­ which the compound tense forms exist, but are less used (Votic, to perfect forms (e.g., polnud / ei olnud ol-nud ‘had not been’) can be some extent also Finnish and Ingrian), have had an opposite effect considered a rather rare and marginal phenomenon in both formal and on the usage frequency of the perfect and pluperfect in Estonian dia­ informal uses of Estonian. Due to the rarity of these forms in actual lects. Although German has also heavily influenced all levels of the language use, it is impossible to argue that the amount of pluperfect Estonian language for a long period of time, it is difficult to assess the forms in the data is also responsible for the proportions of affirma­ impact of German on a more local level; it can only be stated that due tive and negative sentences. In our data, altogether there were only to historical reasons, the Seto dialect is less influenced by German. 21 negative indicative pluperfect forms, with only 4 of them from the When comparing the frequencies of compound tenses in differ­ Insular dialect. Even when we take into account all possible moods (in ent dialects to those of other constructions consisting of a finite and a addition to the indicative), the corresponding figures rise to a mere 85 non-finite verb form (Uiboaed 2013; Uiboaed et al. 2013), the former and 14; i.e., negative pluperfect forms are still infrequent in the data. do not show the distinctions, which are otherwise strong and clear-cut The majority of negative compound tense forms, therefore, consists of between the western and eastern (including southern) dialects. indicative perfect forms with dialectal differences still holding. The high frequency regions, the Insular and Mulgi dialects in particular, are characterized also by additional functions of compound tense constructions. In the Insular dialect, higher frequency is related to a large proportion of negated sentences. The majority of these sen­ tences are negated using a synthetic form pole (instead of the analytic ei ole ‘is not’), which has become entrenched as a general negation word; instead of the simple past, the perfect is used in negative con­ texts, often receiving a general past tense reading. Texts from Karksi

184 185 LIINA LINDSTRÖM, MAARJA-LIISA PILVIK, ON THE USE OF PERFECT AND PLUPERFECT IN ESTONIAN DIALECTS: MIRJAM RUUTMA & KRISTEL UIBOAED FREQUENCY AND LANGUAGE CONTACTS

Parish in the Mulgi region are characterized by an abundance of quo­ References tative uses of the compound past tenses which is mainly reflected in the frequency of the pluperfect. The same tendency is also present Alvre, Paul 2000: Kirderannikumurde ja vadja keele ühisjooni [Common though to a lesser extent in other South Estonian dialects (Tartu and features of the Estonian Northeastern Coastal dialect and the Votic lan­ Võru). This may be related to the large amount of narrative texts in the guage]. – Jüri Viikberg (ed.), Inter dialectos nominaque. Pühendusteos research material as well as the overall tendency towards expressing Mari Mustale 11. novembril 2000. Eesti Keele Instituudi Toimetised 7. Tallinn: Eesti Keele Sihtasutus. 1–13. reported evidentiality more often; the latter may also be induced by Ariste, Paul 1956: Läänemere keelte kujunemine ja vanem arenemisjärk [The contacts with Latvian. development and the earlier stage of the Finnic languages]. – H. Moora (ed.), Eesti rahva etnilisest ajaloost. Tallinn: Eesti Riiklik Kirjastus. 5–23. — 1962: Teiste läänemere keelte elemente Vaivara murrakuis [Elements Acknowledgements of other Finnic languages in the Vaivara dialects]. – Emakeele Seltsi aastaraamat VIII: 11–18. Tallinn. This study was supported by the Estonian Research Council (project Arppe, Antti, Gaëtanelle Gilquin, Dylan Glynn, Martin Hilpert & Arne PUT90 “Estonian Dialect syntax”) and by the (European Union) Eu­ Zechsel 2010: Cognitive Corpus Linguistics: five points of debate on ropean Regional Development Fund (Centre of Excellence in Esto­ current theory and methodology. – Corpora 5(1): 1–27. nian Studies). Barlow, Michael & Suzanne Kemmer 2000: Usage-based models of lan- guage. CSLI publications Stanford. Björklöf, Sofia 2012: Viron rantamurteen länsiryhmän sanaston alkuperä Abbreviations suomalaislainojen valossa [The origin of the vocabulary of the west­ ern subgroup of the Estonian Coast dialect in light of Finnish loan­ words. MA thesis]. Pro gradu -tutkielma. Helsingin yliopiston suomen 1, 2, 3 person m masculine kielen, suomalais-ugrilaisten ja pohjoismaisten kielten ja kirjallisuuks­ ade adessive neg negation ien laitos. Available at: ‹http://urn.fi/URN:NBN:fi-fe2012121410304› all allative nt neuter — 2018: Uusia lainaetymologioita viron koillisrannikon rantamurteen app active past participle pfv perfective länsiryhmässä [New loan etymologies in the western subgroup of the CED Corpus of Estonian Dialects pl plural Estonian Coast dialect of the northeastern coast]. – Sampsa Holopainen cli clitic pp past participle & Janne Saarikivi (eds), Περı` o̓ ρθότητος ἐτύμων. Uusiutuva uralilainen cng connegative verb form ppp passive past participle etymologia. Uralica Helsingiensia 11. Helsinki: Suomalais-Ugrilainen def definite prep preposition Seura. 357–487. Available at: ‹https://journal.fi/uralicahelsingiensia/ ela elative prs present tense issue/view/uh11/489› f feminine prt partitive — 2019: Mutual contacts and lexical relations among the Finnic varieties gen genitive pst past of western Ingria and northeastern Estonia. – Sofia Björklöf & Santra ill illative ptcl particle Jantunen (eds), Multi­lingual Finnic. Language contact and change. imp imperative ptcp participle Uralica Helsingiensia 14. Helsinki: Finno-Ugrian Society. 89–153. ine inessive sg singular Available at: ‹https://doi.org/10.33341/uh.85034› inf infinitive sup supine Bybee, Joan 2010: Language, usage and cognition. Cambridge University ins instrumental ter terminative Press. Comrie, Bernard 1976: Aspect. An introduction to the study of verbal aspect and related problems. Cambridge: Cambridge University Press.

186 187 LIINA LINDSTRÖM, MAARJA-LIISA PILVIK, ON THE USE OF PERFECT AND PLUPERFECT IN ESTONIAN DIALECTS: MIRJAM RUUTMA & KRISTEL UIBOAED FREQUENCY AND LANGUAGE CONTACTS

Dahl, Östen 1995: The Tense System of Swedish. – Rolf Thieroff (ed.), Tense — 1992: Eestirootslastest [On the Estonian Swedes]. – Keel ja Kirjandus Systems in European Languages II. De Gruyter. 59–68. XXXV (7): 394–402. — 2014: The perfect map: Investigating the cross-linguistic distribution — 1998: Eesti-rootsi murdekontaktid [Estonian–Swedish dialect con­ of TAME categories in a parallel corpus. – Benedikt Szmrecsanyi tacts. Unpublished MA thesis]. Tallinn: Tallinn University, Faculty of & Bernhard Wälchli (eds), Aggregating Dialectology, Typology, and Philology. Register Analysis. Berlin: De Gruyter. 268–289. — 2012: Harju-Madise murrak [The Harju-Madise dialect]. Mari-Liis Dahl, Östen & Maria Koptjevskaja-Tamm (eds) 2001: Circum-Baltic Lan- Kalvik & Helmi Neetar (eds). Tallinn: Eesti Keele Sihtasutus. guages. Typology and Contact. Volume 1: Past and Present, Volume 2: Kalnača, Andra 2015: A Typological Perspective on Latvian Grammar. De Grammar and Typology. Studies in Language Companion Series Gruyter Mouton. 54–55. Amsterdam: John Benjamins Publishing Company. Kask, Arnold 1984: Eesti murded ja kirjakeel [Estonian dialects and the Divjak, Dagmar & Stefan Th. Gries 2012: Frequency effects in language standard language]. Tallinn: Valgus. representation. Vol. 2. Walter de Gruyter. Kehayov, Petar 2008: An areal-typological perspective to evidentiality: the EKI 2014 = Eesti Keele Instituudi kohanimeandmebaasi kihelkonnapiiride cases of the Balkan and Baltic linguistic areas. Dissertationes Lin­ andmestik [Map of the Place Name Database created by the Institute guisticae Universitatis Tartuensis 10. Tartu: Tartu Ülikooli Kirjastus. of Estonian Language]. Available at: ‹http://hdl.handle.net/10062/6524› Erelt, Mati 2013: Eesti keele lauseõpetus. Sissejuhatus. Öeldis [Estonian Klaas-Lang, Birute & Miina Norvik 2014: Balti areaali tüpoloogilisi sarna­ syntax. Introduction. Predicate]. Tartu Ülikooli eesti keele osakonna susi morfosüntaksi valdkonnas [Typological similarities of morpho­ preprindid 4. Tartu. syntax in the Baltic linguistic area]. – Keel ja Kirjandus LVII (08–09): Erelt, Mati, Reet Kasik, Helle Metslang, Henno Rajandi, Kristiina Ross, Henn 590–608. Available at: ‹http://kjk.eki.ee/ee/issues/2014/8-9/535› Saari, Kaja Tael & Silvi Vare 1995: Eesti keele grammatika I. Morfoloo- Koit, Enn 1963: Eitus saarte murdes [Negation in the Insular dialect]. – gia. Sõnamoodustus [Estonian grammar I. Morphology. Word forma­ Nonaginta. Johannes Voldemar Veski 90. sünnipäevaks 27. juunil tion]. Tallinn: Eesti Teaduste Akadeemia Eesti Keele Instituut. 1963. Emakeele Seltsi Toimetised 6. Tallinn. 136–147. Gries, Stefan Th. & Dagmar Divjak 2012: Frequency effects in language Koptjevskaja-Tamm, Maria & Bernhard Wälchli 2001: The Circum-Baltic learning and processing. Vol. 1. Walter de Gruyter. Languages: An Areal-Typological Approach. – Östen Dahl & Maria Haspelmath, Martin 2001: The European Linguistic Area: Standard Average Koptjevskaja-Tamm (eds), Circum-Baltic Languages. Typology and Con- European. – Martin Haspelmath, Ekkehard König, Wulf Oesterreicher tact. Volume 2: Grammar and Typology. Studies in Language Companion & Wolfgang Raible (eds), Language Typology and Language Univer- Series 55. Amsterdam: John Benjamins Publishing Company. 615–750. sals / Sprachtypologie und sprachliche Universalien / La Typologie Kortmann, Bernd 2010: Areal variation in syntax. – Peter Auer & Jürgen des langues et les universaux linguistiques: An International Hand- E. Schmidt (eds), Language and Space: An International Handbook book / Ein internationales Handbuch / Manuel international, 1–2. Ber­ of Linguistic Variation. Theories and Methods. Berlin: De Gruyter. lin: de Gruyter. 1492–1510. 837–864. Heine, Bernd & Tanja Kuteva 2005: Language Contact and Grammatical Laakso, Johanna 2001: The Finnic languages. – Östen Dahl & Maria Change. Cambridge: Cambridge University Press. Koptjevskaja­ -Tamm (eds), Circum-Baltic Languages. Typology and — 2006: The Changing Languages of Europe. Oxford: Oxford University Contact. Volume 1: Past and Present. Amsterdam: John Benjamins Press. Publishing Company. 179–212. Ikola, Osmo 1960: Perfektin ja pluskvamperfektin synnystä [On the origin Laanest, Arvo 1975: Sissejuhatus läänemeresoome keeltesse [Introduction of the perfect and pluperfect]. – Virittäjä 64: 364–368. Available at: to the Finnic languages]. Tallinn: Eesti NSV Teaduste Akadeemia ‹https://journal.fi/virittaja/article/view/34251› Keele ja Kirjanduse Instituut. Juhkam, Evi 1983: Rootsipärane objekt eesti murretes [The Swedish-origin Larsson, Ida 2009: Participles in Time. The Development of the Perfect Tense in object in Estonian dialects]. – Keel ja Kirjandus XXVI (3): 122–125. Swedish. Nordistica Gothoburgensia 29. Göteborg: Göteborgs Universitet.

188 189 LIINA LINDSTRÖM, MAARJA-LIISA PILVIK, ON THE USE OF PERFECT AND PLUPERFECT IN ESTONIAN DIALECTS: MIRJAM RUUTMA & KRISTEL UIBOAED FREQUENCY AND LANGUAGE CONTACTS

Lindstedt, Jouko 2000: The perfect-aspectual, temporal and evidential. – Metslang, Helle 1993: Verbitarind ajatähendust väljendamas [Verbal con­ Empirical Approaches to Language Typology (6): 365–384. struction in expressing temporality]. – Virittäjä 97: 203–221. Available Lindström, Liina 2015: Subjecthood of the agent argument in Estonian pas­ at: ‹https://journal.fi/virittaja/article/view/38574› sive constructions. – Marja-Liisa Helasvuo & Tuomas Huumo (eds), — 1997: On the use of the Estonian past tense forms through the last one hun­ Subjects in Constructions – Canonical and Non-Canonical. Construc­ dred years. – Mati Erelt (ed.), Estonian: typological studies II. Publications tional Approaches to Language 16. Amsterdam: John Benjamins Pub­ of the Department of Estonian of the University of Tartu 8. Tartu. 98–145. lishing Company. 141–173. — 2009: Estonian grammar between Finnic and SAE: some comparisons. — forthcoming: Seto lause põhijooned [Basic features of the Seto clause]. – Sprachtypologie und Universalienforschung 62, 1/2: 49–71. – Karl Pajusalu, Andreas Kalkun & Ergo-Hart Västrik (eds), Muižniece, Liena, Helle Metslang & Karl Pajusalu 1999: Past participle III: Keel, folkloor ja tänapäeva kultuur. Seto Instituut. finitization in Estonian and Latvian. – Mati Erelt (ed.),Estonian. Typo- Lindström, Liina, Pärtel Lippus & Tuuli Tuisk 2019: The online database of logical studies III. Publications of the Department of Estonian of the the University of Tartu Archives of Estonian Dialects and Kindred Lan­ University of Tartu 11. Tartu. 128–157. guages and the Corpus of Estonian Dialects. – Sofia Björklöf & Santra Must, Mari 1965: Vene-eesti kakskeelsus Kirde-Eestis [Russian–Estonian Jantunen (eds), Multi­lingual Finnic. Language contact and change. bilingualism in northeastern Estonia]. – H. Moora & L. Jaanits (eds), Uralica Helsingiensia 14. Helsinki: Finno-Ugrian Society. 327–350. Slaavi-läänemeresoome suhete ajaloost. Tallinn: Eesti Raamat. 107–131. Available at: ‹https://doi.org/10.33341/uh.85040› — 2000: Vene laensõnad eesti murretes [Russian loanwords in Estonian Lindström, Liina, Maarja-Liisa Pilvik, Mirjam Ruutma & Kristel Uiboaed dialects]. Tallinn: Eesti Keele Sihtasutus. 2015: Mineviku liitaegade kasutusest eesti murretes keelekontaktide Norvik, Piret 2000: Kirderannikumurde sõnavaralisi suhteid [Lexical rela­ valguses [Abstract: The use of the compound past tenses in Estonian tions in the Estonian Northeastern Coastal dialect]. – Jüri Viikberg dialects in the light of language contacts]. – Jüvä Sullõv (ed.), Aig õda- (ed.), Inter dialectos nominaque. Pühendusteos Mari Mustale 11. gumeresoomõ keelin. Aeg läänemeresoome keeltes. Time and tense in novembril 2000. Eesti Keele Instituudi Toimetised 7. Tallinn: Eesti Finnic languages. Publications of Võro Institute 29. Võro: Võro Instituutˊ. Keele Sihtasutus. 184–201. 40–71. Available at: ‹https://wi.ee/et/voru-instituudi-toimetised-28/› Pajusalu, Karl, Tiit Hennoste, Ellen Niit, Peeter Päll & Jüri Viikberg 2009: Lindström, Liina & Piret Toomet 2000: Eesti suuliste narratiivide keelelisi Eesti murded ja kohanimed [Estonian dialects and place names]. 2. ed. erijooni [Linguistic features of oral narratives in Estonian]. – Tiit Hen­ Tallinn: Eesti Keele Sihtasutus. noste (ed.), Eesti keele allkeeled. Tartu Ülikooli Eesti Keele Õppetooli Pall, Valdek 1982: Idamurde vahekorrast naabermurretega ja vadja keelega Toimetised 16. Tartu. 174–203. [The Relation of the Estonian Eastern dialect with neighboring dialects Lindström, Liina & Ilona Tragel 2010: The possessive perfect construction and the ]. – Keel ja Kirjandus XXV (5): 246–251. in Estonian. – Folia Linguistica: Acta Societatis Linguisticae Euro- Pilvik, Maarja-Liisa & Kristel Uiboaed 2014: Grammatical evidentiality in paeae 44 (2): 371–400. Estonian: areal evidence of variation. Paper presented at the Syntax of Lindström, Liina, Kristel Uiboaed & Virve-Anneli Vihman 2014: Varieeru­ the World’s Languages VI, Pavia. mine tarvis-/vaja-konstruktsioonides keelekontaktide valguses [Varia­ R Development Core Team 2013: R: A Language and Environment for Sta- tion of the tarvis/vaja-constructions in light of language contacts]. – tistical Computing. R Foundation for Statistical Computing. Available Keel ja Kirjandus LVII (08–09): 609–630. Available at: ‹http://kjk.eki. at: ‹http://www.R-project.org/› ee/ee/issues/2014/8-9/536› Saareste, Andrus 1955: Väike eesti murdeatlas. Petit atlas des parlers Lonn, Varje & Ellen Niit 2002: Saarte murde tekstid [Insular dialect texts]. estoniens [Small dialectal atlas of Estonian]. Uppsala: Kungliga Gus­ Eesti murded VII. Tallinn: Eesti Keele Instituut. tav Adolfs Akademien. Markus, Elena B. & Fedor I. Rozhanskiy 2011: Sovremennyj vodskij jazyk. Serebrennikov, B. A. 1959: Pluskvamperfekti ja perfekti päritolu problee­ Teksty i grammatičeskij očerk. V 2-h tomah [Contemporary Votic lan­ mist läänemeresoome keeltes [On the problem of the origin of the plu­ guage. Texts and grammar. In 2 volumes]. Sankt-Peterburg: Nestor- perfect and perfect in the Finnic languages]. – Emakeele Seltsi aasta- Istorija. raamat IV (1958): 249–255.

190 191 LIINA LINDSTRÖM, MAARJA-LIISA PILVIK, ON THE USE OF PERFECT AND PLUPERFECT IN ESTONIAN DIALECTS: MIRJAM RUUTMA & KRISTEL UIBOAED FREQUENCY AND LANGUAGE CONTACTS

Seržant, Ilja A. 2012: The so-called possessive perfect in North Russian and Perfekti ja pluskvamperfekti kasutusest the Circum-Baltic area. A diachronic and areal account. – Lingua 122: eesti murretes: sagedus ja keelekontaktid 356–385. — 2015: The Independent Partitive as an Eastern Circum-Baltic isogloss. Liina Lindström, Maarja-Liisa Pilvik, Mirjam Ruutma & Kristel Uiboaed – Journal of Language Contact (8): 341–418. Siegl, Florian & Petar Kehayov 2006: The evidential past participle in Esto­ Artiklis vaadeldakse perfekti ja pluskvamperfekti esinemissagedust nian reconsidered. – Études finno-ougriennes 38: 75–117. eesti murretes korpusandmete põhjal. Murretevahelistele sageduserine­ Söderman, Tiina 1996: Lexical characteristics of the Estonian North East- ern Coastal dialect. [PhD thesis.] Studia Uralica Upsaliensia 24. Acta vustele otsitakse seletusi eelkõige lokaalsetest keelekontaktidest. Liit­ Universitatis Upsaliensis. Uppsala. ajad arvatakse olevat eesti keelde tekkinud balti keelte mõjul, nende Szmrecsanyi, Benedikt 2013: Grammatical variation in British English dia- edasiste funktsioonide arengut on mõjutanud peamiselt germaani kee­ lects: a study in corpus-based dialectometry. Cambridge: Cambridge led (Serebrennikov 1959, Ikola 1960). Eesti murrete puhul on aga liit­ University Press. aegade kasutusele mõju avaldanud lisaks ka lokaalsed kontaktid rootsi, Tommola, Hannu 2000: On the perfect in North Slavic. – Östen Dahl (ed.), vene, läti ja läänemeresoome keeltega (peamiselt vadja, soome ja isuri Tense and Aspect in the Languages of Europe. Empirical Approaches keelega). Eesti murrete korpuse andmete põhjal kasutatakse perfekti to Language Typology 20. Berlin: De Gruyter. 441–478. ja pluskvamperfekti teiste murretega võrreldes eriti sagedasti saarte ja Uiboaed, Kristel 2010: Ühendverbid eesti murretes [Compound verbs in Mulgi murdes, kasutussagedus on aga oluliselt väiksem Seto ja kirde­ Estonian dialects]. – Keel ja Kirjandus LIII (1): 17–36. murdes. Sagedused peegeldavad vähemalt osaliselt keelekontaktide — 2013: Verbiühendid eesti murretes [Verb compounds in Estonian dia­ mõju: kontaktid keeltega, kus samad kategooriad on olemas ning laial­ lects]. Dissertationes Philologiae Estonicae Universitatis Tartuensis daselt kasutuses (rootsi ja läti keel), suurendavad liitaegade kasutussa­ 34. Tartu: Tartu Ülikooli Kirjastus. Available at: ‹http://hdl.handle. net/10062/34499› gedust ka kontaktalal (resp. saarte ja Mulgi murdes); tihedad kontaktid Uiboaed, Kristel, Cornelius Hasselblatt, Liina Lindström, Kadri Muischnek keeltega, milles vastav kategooria puudub (vene) või on küll olemas, & John Nerbonne 2013: Variation of verbal constructions in Estonian ent harvemini kasutuses (vadja ja isuri keel), aga pigem vähendavad liit­ dialects. – Literary & Linguistic Computing 28 (1): 42–62. aegade kasutussagedust ning kontaktalal võivad liitaegade asemel olla Uiboaed, Kristel & Maarja-Liisa Pilvik 2014: Exploring the borders of ana­ enam levinud teised väljendusvahendid (nt lihtminevik). lyticity and syntheticity. Paper presented at the conference Methods in Lisaks keelekontaktide mõjule tulevad sagedusi uurides esile ka piir­ Dialectology XV, Groningen, August 11–15 2014. kondlikud funktsionaalsed erinevused liitaegade kasutuses. Saarte murde Univere, Aili 1988: Idamurre [The Estonian Eastern dialect]. – Emakeele andmestikus oli eitavate lausete proportsioon liitaegade kasutusjuhtudest Seltsi aastaraamat 32 (1986): 59–93. märkimisväärselt kõrgem kui teistes murretes. Valdav enamik eitavatest Vaba, Lembit 1997: Uurimusi läti-eesti keelesuhetest [Studies on Latvian– lausetest olid omakorda pole-eitussõnaga moodustatud perfekti­vormid, Estonian language relations]. Tallinn – Tampere: Eesti Keele Instituut mis annab alust oletada, et pole + nud-partitsiibi ühend on saarte murdes & Tampereen yliopiston suomen kielen ja yleisen kielitieteen laitos. kujunenud üldminevikku tähistavaks eitusvormiks, kuna seda kasutatak­ — 2011: Kuidas läti-eesti keelekontakt on mõjutanud eesti murdekeele grammatikat ja sõnamoodustust [How Latvian–Estonian language se ka kontekstides, milles vastav jaatav vorm oleks lihtminevikus. Mulgi contact has influenced grammar and word formation in Estonian dia­ murde suur liitaegade kasutussagedus­ on seotud kvotatiivsete pluskvam­ lects]. – Emakeele Seltsi aastaraamat 56 (2010): 204–246. perfekti vormide suure osakaaluga andmestikus. See võib tuleneda sel­ Wälchli, Bernhard 2011: The Circum-Baltic languages. – Bernd Kortmann lest, et Mulgi murde materjalis esines keskmisest enam narratiivseid & Johann van der Auwera (eds), The Languages and Linguistics of tekste (muistendeid, pärimusi), ent ka kontaktimõjulisest­ evidentsiaalsete Europe. A Comprehensive Guide. De Gruyter Mouton. 325–340. strateegiate laialdasest kasutamisest kontaktalal läti keelega.

192 193 TORBJÖRN SÖDER Uppsala University

The Finnish of Rautalampi and Värmland – A comparison

Abstract In this paper some of the features of the Finnish spo­ ken in the former great-parish1 of Rautalampi and Värmland are discussed. These two Finnish varieties are, owing to their com­ mon historical background, closely affiliated, both being defined as Eastern Finnish Savo dialects. These varieties were isolated from each other for hundreds of years and the circumstances of the two speech communities are in many respects different. The language contact situation in Värmland involves Swedish varie­ ties and Värmland Finnish but limited contact with other Finnish varieties. The contact situation in what once was the great-parish of Rautalampi involves mainly Finnish varieties, both dialects and Standard Finnish. Social and economic circumstances have also had an effect on the contact situations, especially since the 19th century when Finland was separated from Sweden and in­ dustrialization made its definite entrance into both areas. The dif­ fering language contact situations have in some respects yielded different results. This paper provides examples of the differences, which have occurred as an effect of the circumstances in Rau­ talampi and in Värmland. I provide examples from the phonologi­ cal, morphological, lexical, and syntactic levels and discuss them in light of the language contact situation in these areas. Keywords Värmland Finnish, Rautalampi Finnish, Finnish dialects, language contact, language change

1. In this paper, the term “great-parish” is used as an equivalent to Fi. suurpitäjä and Swe. storsocken, which both literally mean ‘large parish’. It refers to a geo­ graphically large parish which once was established as the result of new settlements in a previously sparsely inhabited area. Over time and up until the present day, such parishes have in general been divided into smaller administrative units.

Multilingual­ Finnic. Language contact and change. 195–225. Uralica Helsingiensia 14. Helsinki 2019. ‹https://doi.org/10.33341/uh.85036› TORBJÖRN SÖDER THE FINNISH OF RAUTALAMPI AND VÄRMLAND – A COMPARISON

1. Introduction century to the end of the 17th century when the migration began to diminish. It is quite clear from the study that Savo is the most fre­ In the late medieval period, people in the historical province of Savo, quent area of origin for the , especially the Värmland in eastern Finland, began to move from the central areas of the prov­ Finns. The areas around Pieksämäki and Mikkeli in western Savo ince and settle in more peripheral areas. Eventually, this movement stand out together with the great-parish of Rautalampi as the most expanded to neighboring provinces and in the late 16th century the ex­ important point of origin (Saloheimo 2009: 26). The areas around pansion reached the central parts of . This expansion has Juva, Ranta­salmi, and Sääminki farther east also appear as important played an important role in the demographic history of both Finland points of origin. What is noteworthy in Saloheimo’s study is that the and Sweden. great-parish of Rautalampi appears to be a less important point of The northern part of the contemporary province of Central Fin­ origin than what it is often considered to be. According to Saloheimo land and the northwestern part of the province of Värmland in west­ (ibid. 12) only 15 percent of the persons in the study had their origin ern Sweden are two areas which were much affected by the migra­ in the great-parish of Rautalampi. The origin of the residents of the tion from eastern Finland. The settlements in central Finland led up great-parish seems to be quite similar to that of the Värmland Finns to the establishment of the great-parish of Rautalampi in the 1560s. (Saloheimo 1986: 11). The northwestern part of Värmland and the adjacent Norwegian areas received their first eastern Finnish settlers in the middle of the 17th 1.2. The Finnish of Rautalampi and Värmland century. The settlers mastered different techniques which made them suc­ The Finnish varieties of Rautalampi and Värmland are generally con­ cessful in their new environment. The slash-and-burn technique is sidered as closely affiliated (Kettunen 1930; Tuomi 1989). The lack often mentioned in connection with these settlers and it has become of the suffix-p(i) as a marker for third person singular, which appears one of their emblematic features. However, in order to be successful farther east, is one of several shared features (Hormia 1970: 58). The farmers in the forest areas of Finland and Sweden a wide range of Finnish spoken in the area which originally formed the great-parish of techniques were used. Rautalampi is not considered to form one single dialect, but as belong­ ing to several eastern dialects. The central and by far the largest part 1.1. The settlers’ origin of the former great-parish belongs to the North Central Finland dialect area, the very western part to the Middle Bothnian dialect area, the In this comparison between the Finnish in Rautalampi2 and in Värm­ very eastern part to the North Savo dialect area, and the southern part land it is assumed that the two varieties are closely affiliated because to the South Central Finland dialect area (see Map 1). of the speakers’ common ancestry. The great-parish of Rautalampi Värmland Finnish, which is regarded as a separate eastern dia­ has usually been thought of as being the most important area of origin lect, is usually not divided into subdialects, but smaller differences by far of the Värmland Finns (cf. Bladh 2001: 24; Kettunen 1930: within this area are sometimes mentioned (Kettunen 1910: 3; Mägiste 149). Veijo Saloheimo (2009) has tried to identify the point of origin 1948: 5). of 1 430 Forest Finns. The period of investigation stretches from the first Forest Finnish migration from Finland to Sweden in the late 16th

2. In this article I use Rautalampi with reference to the area originally covered by the former great-parish of Rautalampi.

196 197 TORBJÖRN SÖDER THE FINNISH OF RAUTALAMPI AND VÄRMLAND – A COMPARISON

1.3. The language contact situations

Considering the common cultural and linguistic background of the settlers in Rautalampi and Värmland, and the similar geographic and climatic conditions in the great-parish of Rautalampi and northwest­ ern Värmland, one can assume that the settlers’ lives initially were quite similar in the two areas. The settlers established themselves as self-sufficient forest farmers and continued an existence based on the livelihood of earlier generations. The two areas were, however, geo­ graphically isolated from each other and were affected by phenomena and changes that occurred locally. The language contact situations in Rautalampi and Värmland were different. In Rautalampi, the situation involved Rautalampi Finnish and other Finnish varieties, whereas the situation in Värmland mainly involved Värmland Finnish, Swedish, and Norwegian. The impact of these differences became increasingly important during the 19th century, when industrialization and other dramatic socio-econom­ ic changes reached the forest regions of Finland and Sweden.

1.3.1 The situation in Rautalampi

The separation of Finland from Sweden in 1809 meant a definite po­ litical divide between the two areas, which ultimately would have an effect on the local situation, especially in Finland. The Grand Duchy of Finland ultimately gained an increasing amount of influence over its domestic affairs, which, in conjunction with strong nationalistic striving, led to the establishment of Finnish social institutions. The Finnish language had an important role within this nationalistic striv­ ing, both as a symbol and as a means. During the 19th century Finn­ ish developed into a language which was used in all domestic institu­ Map 1. The great-parish of Rautalampi was established in the 1560s. Since then the tions and domains. These changes certainly had an impact, not only area has been divided into smaller administrative units on several occasions and on the situation in the political centers of Helsinki and Turku, but also currently 22 municipalities (Fi. kunta) constitute the former great-parish. on the situation elsewhere in Finland. The endeavor to make Finnish a language for all societal institutions in Finland created a need for words and expressions for a number of conceptual systems, and, fur­ thermore, a written standard language to bridge differences between different Finnish varieties (Laine 2007: 78).

198 199 TORBJÖRN SÖDER THE FINNISH OF RAUTALAMPI AND VÄRMLAND – A COMPARISON

The most important steps towards the status of a fully-fledged na­ tional language were taken in the latter part of the 19th century. At the beginning of the century, severe Russian censorship limited the sub­ jects of published texts. Despite this, newspapers played an important role early on as language developers, by publishing texts on subjects which had not been written about in Finnish before. Finnish literature also enriched the vocabulary. The significant increase in the publica­ tion of Finnish literature, predominantly of translations, created a base for the vocabulary of many scientific disciplines (Laine 2007: 81). The Finnish vernaculars and the old literary language lacked terms for many concepts which were required in different sections of society. The concepts lacking a Finnish term were usually known in other lan­ guages and the challenge was to find a suitable Finnish equivalent. At the beginning of the 19th century, the attitude towards foreign words was rather positive and borrowing foreign words was considered one of many ways to bring new words into the language. However, as the idea of a Finnish nation grew stronger during the century, purism became a more distinguished feature as well, and efforts were made to restrict the number of loanwords by replacing them with words of Finnish origin. In the search for a Finnish identity, attention was paid to parts of the old great-parish of Rautalampi. In 1837, the city of Jyväskylä was founded in the southern part of the former great-parish. Jyväsky­ lä became increasingly important for trade and manufacturing in the area (Jokipii 1988: 13). The city also earned a reputation as a center Map 2. In the middle of the 20th century, Värmland Finnish was still spoken in the for Finnish-medium education, when several educational institutions parishes of Södra Finnskoga, Nyskoga, and Östmark in the northwestern part of Värmland. The Värmland Finnish material presented in this article has been re- were founded around the middle of the century. As the idea of a Finn­ corded in these parishes. ish identity grew stronger among students and intellectuals, attention was paid to the Finnish spoken in Jyväskylä and its surroundings. Ac­ cording to the ideas of that time, the language in the area was consid­ 1.3.2. The situation in Värmland ered pure and not ruined by any corrupting Swedish influence. Schol­ ars praised the Finnish spoken in the southern part of the great-parish In Värmland, after the 17th century settlement phase, there probably and adjacent areas (see Topelius 1908: 135). There were differences was a situation where individuals among the first settlers’ descendants between different areas in Rautalampi, between rural and urban areas, spoke only Finnish, were bilingual in Finnish and Swedish3, or only but altogether the development of education, media, and the economy spoke Swedish (Bladh 1995: 222). The contacts between the Värm­ increased exposure to Standard Finnish. land Finns and the Swedish population required Swedish-speaking 3. On the Norwegian side, bilingualism included Norwegian.

200 201 TORBJÖRN SÖDER THE FINNISH OF RAUTALAMPI AND VÄRMLAND – A COMPARISON persons in the group, but it was not necessary for all Finns to speak for Höbråten and Baskaboro for Lortbäcken. When grandmother Swedish. Bilingual individuals could function as intermediaries be­ was being prepared for her confirmation, she had to interpret tween the two groups by communicating directly with Swedes or by the teaching. The priest did not speak Finnish and some of the functioning as interpreters between Värmland Finns and Swedes. As children did not speak Swedish. My grandmother was bilingual. contacts between Swedish-speaking society and the Finns became When I grew up at the turn of the century there were still many more intense, among other reasons due the establishment of churches Finnish-speaking people in “Söfinnskogen” [The Southern Finn­ ish Forest]: Röjden, Bjurberget, and Mackatjärn. (Finnson 1976: and extended transportation, the proportion of bilingual individuals 8, my translation.) increased. However, as long as there were domains where Finnish was in use, the increased number of bilingual individuals was not a threat The dramatic socioeconomic changes of the 19th century changed the to Värmland Finnish. Locations in nature, farms, and other places had outlook for the preservation and development of Värmland Finnish. Finnish names (Mägiste 1955), of which some are still in use today. By the end of the century, many Finns had been forced to leave the In addition, traditional livelihood with its many occupations was also traditional Värmland Finnish domains to find their income elsewhere. a Finnish-language domain. Finnish had a strong position on Finnish This development was to some extent, and temporarily, counter­ farms and in the activities that were connected with them. Children balanced. Marriage statistics show that the Finns to a great extent mar­ were naturally raised in a Finnish-speaking environment with some ried within the group (Bladh 1995: 221). Finns who married Swedes eventually learning Swedish through contacts with the Swedish- were predominantly crofters. Yeomen married more often within the speaking population. group, which in Swedish gave rise to the notion Finnadel ‘Finnish The church was an institution where the Värmland Finns would nobility’. It seems that the Värmland Finnish culture and language come in contact with the Swedish-speaking community, but the clos­ were best preserved on larger farms, which functioned as hubs in a est church was a considerable distance away for some Finns (Broberg Värmland Finnish network. 1952: 126; Bredin 2009: 19). A visit to the church was commonly At the beginning of the 20th century, the use of Värmland Finnish paid only when one had other errands to run in the church village. reached a stage where the language was no longer transferred to the Therefore, events such as the baptism of a child could be delayed for younger generation. From a language endangerment perspective such years. The difficulty for people in the most remote villages to attend a situation implies that the future and survival of a language is severely church had implications for confirmation. Sometimes an adult, func­ threatened. In the assessment of language endangerment, scales refer­ tioning as a protector and an interpreter, accompanied the children to ring to language use are commonly used to determine to what extent a the church village, where the children stayed for a period of education language is threatened (cf. Fishman 1991). Ethnologue (Ethnologue) in the spring. The priest John Finnson, who was born in Bograngen in uses a scale with 13 levels, extending from the most vigorous level the southern part of the Värmland Finnish area, recounts how relations International (level 0) down to Extinct (level 10). According to Eth­ were between the church and the Finns. This quote is also an account nologue’s language endangerment scale Värmland Finnish took a step of the asymmetric bilingualism (Thomason 2001: 4, 9) in Värmland: down from level 6b4, Threatened, to level 75, Shifting, at the beginning th When Swedish was introduced in the 1870s to the southern parts of the 20 century and steps further down the scale later in the century. of the parish in connection with the establishment of the public school, it happened with a lot of crying and agony. The whole vil­ lage of Skallbäcken moved to North America, as the people in the 4. “The language is used for face-to-face communication within all generations, village did not want to learn Swedish. In my childhood, people but it is losing users.” (Ethnologue.) 5. “The child-bearing generation can use the language among themselves, but it is still more generally used the Finnish placenames, like Heinaho not being transmitted to children.” (Ethnologue.)

202 203 TORBJÖRN SÖDER THE FINNISH OF RAUTALAMPI AND VÄRMLAND – A COMPARISON

2. Perspectives 2.1. Anaptyxis

The Finnish spoken in Rautalampi and in Värmland gives us the op­ Anaptyxis or svarabhakti (Fi. svaavokaali) is a feature common in the portunity to study how the language of two historically closely af­ Savo dialects in Finland. The anaptyctic vowel appears after an unstressed filiated Finnish speech communities is affected by different language syllable between two successive consonants, of which the first isl and the contact situations involving opposed circumstances. The Finnish in second k, p, m, or v. An anaptyctic vowel may also appear between n and Rautalampi is part of the Finnish dialect continuum, which makes it h (ISK § 33). In example 1, the anaptyctic vowel separates l and k. susceptible to influences from surrounding dialects. The people in (1) FiR (Lehtinen 1982: 67) Rautalampi are also members of Finnish society, which exposes them ̆ ja küllä ne ol valàkos-i-(a ja to Standard Finnish. The Värmland Finns, on the other hand, lived and indeed they be.sg3.pret white-pl-part and isolated from other Finnish-speaking societies and had a more limited komme-i-ta ne alùsvoattē-t contact with Standard Finnish. Instead they were exposed to different beautiful-pl-part they underwear[nom]-pl Swedish varieties, both standard and non-standard. ‘and indeed they were white and beautiful, The exposure to Standard Finnish through contacts with Finnish those [pairs of] underwear’ scholars, however, must not be neglected. Carl Axel Gottlund made research trips to Värmland in the years around 1820 and he also dis­ The anaptyctic vowel as it appears in the Savo dialects is not a gen­ tributed Bibles among the Värmland Finns (Söder 2012: 226). Torsten eral feature in Värmland Finnish (Kettunen 1910: 90; 1930: 149). The Aminoff went on a research trip in 1871 (Aminoff 1876) and in first consonant clusters lk (illustrated in example 2), lp, lm, lv, and nh are three decades of the 20th century several scholars collected linguistic not separated by an anaptyctic vowel, but there are, however, other and folkloric material among the Värmland Finns with perhaps Lauri instances of this phenomenon. Mägiste (1948: 6) mentions that the Kettunen having been the most frequent visitor (Kettunen 1960). successive consonants hn, hm, hl, and hj quite often are separated by By comparing and putting into perspective linguistic material an anaptyctic vowel (cf. Kettunen 1910: 90). In example 3, the past from both areas, I illustrate how the different contact situations have participle nähäny ‘seen’ (cf. FiSt nähnyt) appears with an anaptyctic affected the Finnish in these areas. The material that appears in this vowel between h and n. The phrase is an extract from Mörtberg’s re­ article derives from consultants born in Rautalampi and Värmland cords which tell about Ukko, a mythical figure controlling thunder and (Mörtberg 2011; Mägiste 1948) between 1854 and 1900 and for Värm­ rain to whom one must pay respect when trying to bring rain. land’s part they represent the last generation of Värmland Finns who (2) FiV (Mägiste 1948: 11) spoke Värmland Finnish. The records from Rautalampi represent the sitten ne brūga-s valgos-i-a voatte-i-da parishes of Saarijärvi (Yli-Luukko 1997), Pihtipudas (Lehtinen 1982), then they use-sg3.pret white-pl-part garment-pl-part and present-day Rautalampi (Yli-Luukko 1983). The Värmland mate­ vede-ä bello-n ü:lttö rial comes from the parishes of Nyskoga, Södra Finnskoga, and Öst­ pull-inf field-gen over mark (Mörtberg 2011; Mägiste 1948) in northwestern Värmland. ‘then they used to pull white clothes across the field’ Stress and linking symbols have been omitted in the records of Lehtinen, Mägiste, and Yli-Luukko. In order to make the morphologi­ (3) FiV (Mörtberg 2011: 129) cal analysis clearer, long vowels are sometimes rendered with two jotta ukko ois nähä-ny so_that Ukko[nom] be.sg3.cond see-pptcp vowels instead of the original placed above a single vowel. ‘so that Ukko would have seen’

204 205 TORBJÖRN SÖDER THE FINNISH OF RAUTALAMPI AND VÄRMLAND – A COMPARISON

2.2. Primary be systematic and there are substantial differences between individuals (Söder 2011: 113–116; Mörtberg 2011: 19). Example 6 is from a consult­ A feature common in the Savo dialects in Finland is primary gemination ant who almost exclusively uses voiced stops. The words dulˊ, guiva, and (Fi. yleisgeminaatio), which implies the lengthening of a short consonant in gesä in the example correspond to Standard Finnish tuli, kuiva, and kesä. positions between a short stressed vowel and a long unstressed vowel, e.g., (6) FiV (Mägiste 1948: 36) kalaa fish-part.sg ‘fish’ >kallaa (Hormia 1970: 31). This feature is present ja se dulˊ nīn guiva gesä in the Rautalampi material (see example 4) but absent in the material from and it come.sg3.pret so dry[nom.sg] summer[nom.sg] Värmland (Kettunen 1930: 149), which is illustrated in example 5. ‘and a very dry summer arrived’ (4) FiR (Yli-Luukko 1983: 38) The use of the grammatical subject se, in example 6, is an apparent kun̬ sihen nüt ruot̀ē-t tullō when thereto now rib[nom]-pl come.sg3.pres instance of syntactic influence from Swedish, cf. Swe. och det kom ‘when the ribs come there’ en så torr sommar ‘and a very dry summer arrived’, where det is the grammatical subject. (5) FiV (Mörtberg 2011: 66) Kala-hao-t, kala-risu-t, pitt 2.4. Palatalized consonants fish-twig[nom]-pl fish-brushwood[nom]-pl have_to.sg3.pret pan-na pitkä-nä perjantai-na jott pare-mp A phonological feature common in many Savo varieties and also in put-inf long-ess.sg Friday-ess.sg so_that good-comp.nom.sg both Rautalampi and Värmland is the palatalization of consonants. In lykky tuloo Värmland, the palatalized sounds [lj], [nj], and [sj] are common in posi­ luck[nom.sg] come.sg3.pres tions which correspond to the Standard Finnish sequence consonant≈+ i ‘You had to put fish twigs and fish brushwood [into (see examples 7 and 8). There is a variation between palatalized and the water] on Good Friday for better luck at fishing’ non-palatalized instances, but palatalized sounds appear as a rule. Primary gemination is considered to have, at least partly, its origin in (7) FiV (Mägiste 1948: 15) Savo (Hormia 1970: 31) and since it is absent in Värmland it must be garhu dulˊ sīnä ja mäń a relatively recent innovation. bear[nom.sg] come.sg3.pret there and go.sg3.pret ülös hoŋka-an up pine_tree-ill.sg 2.3. Stops ‘the bear came there and went up to the pine tree’

A phonological feature that separates Värmland Finnish from Rautalampi (8) FiV (Mägiste 1948: 11) Finnish is the abundant occurrence of voiced stops. Since Mörtberg has nīn se olˊ üks toinen gub adapted the orthography of his materials to Standard , so it be.sg3.pret one other man[nom.sg] all stops, with a few exceptions, are written , and consequently the joga ne galla-ś valdia 6 7 variation between voiced and unvoiced stops is not visible. In phonetical­ who they call-sg3 .pret Valdia ly more accurate records, this feature is evident. Julius Mägiste discusses ‘so there was another man whom they called Valdia’ the feature in his work Metsäsuomalaismurteiden nykyisestä vaiheesta 6. Similarly to modern colloquial Finnish, the demonstratives se and ne function (1948: 6–7: cf. Kettunen 1930: 151) and notes a general variation between as personal pronouns (cf. Karlsson 2009: §97:4). 7. As in colloquial Finnish, verb forms in the third person singular are commonly voiced, semi-voiced, and voiceless stops. This variation does not seem to used even though the subject is in the third person plural (cf. Karlsson 2009: §97:6).

206 207 TORBJÖRN SÖDER THE FINNISH OF RAUTALAMPI AND VÄRMLAND – A COMPARISON

Palatalized consonants are phonetically rendered quite imprecisely in (11) FiV (Mörtberg 2011: 330) Mörtberg’s records. Forms corresponding to Mägiste’s olˊ are some­ Se oil kupp, joka voitele-iti times spelled oll and sometimes oil (cf. example 11) It be.sg3.pret man[nom.sg] who grease-sg3.refl.pret In the Rautalampi Finnish material, the situation varies more. In tuva-, jotta se lento-on pääs-is most parts of the area, regressive palatalization appears as in Värm­ hut-ine.sg so.that he flight-ill.sg get-sg3.cond j j j hän=nin, vaan se aina puto-is alas land: pan , ol , halus pro Standard Finnish pani, oli, halusi (9). In the he=too8 but he always fall-sg3.pret down records from Saarijärvi (Yli-Luukko 1997), farther southwest in the ‘There was a man, who greased himself in the hut, area corresponding sequences appear as a rule without palatalization so that he too could fly, but he always fell down’ but instead with apocope (10). (12) FiV (Mörtberg 2011: 237) (9) FiR (Lehtinen 1982: 28) Min=en tahto-nna tä-tä se hirvi ni se olˊ män-nä I=neg.1sg want-conneg.pret this-part.sg that elk[nom.sg] so it be.sg3.pret go-pptcp isännä-ks, sillä mina tällä-isi-n, ‘the elk, it had gone’ master-transl.sg therefore I arrange-pret-sg1 jotta se tappa-itiin (10) FiR (Yli-Luukko 1997: 53) so_that he kill-sg3.refl.pret kyllä se ol kummallis-ta ‘I didn’t want him as master, therefore indeed it be.sg3.pret strange-part.sg I arranged it so that he killed himself’ ‘indeed it was strange’ (13) FiV (Mägiste 1948: 14) 2.5. Reflexive conjugation Ne sano-i jotta ne kulkoo jälleen they say-sg3.pret that they go.sg3.pres back tuloo kotiin ja näüttele-itöö A morphological feature attributed to Savo and eastern Finnic in gen­ come.sg3.pres home and show-sg3.refl.pres eral is the occurrence of the so-called reflexive conjugation, which ‘They said that they come back, come home, and show themselves’ is a morphologically many-sided structure used most commonly to give a reflexive meaning to the verb stem (for details, see J. Koivisto (14) FiV (Mägiste 1948: 19) 1990). In the description of reflexives in Finnish dialects, Värmland Kun olˊ jogii jog olˊ hakka-ina-an When be.sg3.pret someone who be.sg3.pret chop-refl.pptcp-px.sg3 and Rautalampi Finnish, the latter as part of the Central Finland Dia­ leikka-ina-an niin buara se sa-i lect, are described as having similar reflexive conjugation systems cut-refl.pptcp-px.sg3 then only he can-sg3.pret (V. Koivisto 1995: 71). In Värmland Finnish, the reflexive conjuga­ diede-ä niin veŕ staana-ś tion is a regular way of expressing reflexivity (see examples 11–14). know-inf so blood[nom.sg] stop-sg3.pret In the Rautalampi material, on the other hand, I have found only a ‘When there was someone who had chopped himself, cut few exceptional instances of reflexive conjugation, e.g., elä näöttäete himself then only he could know how to make the blood stop’ ‘don’t show yourself’ (Lehtinen 1982: 15).

8. In Värmland Finnish, the particle -kin ‘too’ assimilates regurlarly with the word to which it is attached, cf. FiSt hänkin.

208 209 TORBJÖRN SÖDER THE FINNISH OF RAUTALAMPI AND VÄRMLAND – A COMPARISON

2.6. Phrasal verbs (17) FiV (Mägiste 1948: 12) Se emä jog on gasvatta-na A feature of Värmland Finnish is the appearance of constructions for that mother[nom.sg] who be.sg3.pres raise-ppcpt which a Swedish phrasal verb, cf. Swe. partikelförbindelse (SAG III: ülös minu-d §4), is seemingly the model. In these constructions it is often the ad­ up I-acc.sg verb ylös ‘up’, which forms a lexical unit together with the verb. In ‘The mother who has raised me’ example 15, the unit söi ylös ‘ate up’ shows clear correspondence with the Swedish phrasal verb äta upp ‘eat up’. In these phrasal verb-like expressions, the adverb ylös, as in examples 15–17, appears as an equivalent to Swedish upp ‘up’. Another exam­ (15) FiV (Mörtberg 2011: 296) ple of an adverb which appears in these kinds of constructions is the Se-n pitt ol-la aita it-gen.sg have_to.sg3.pret be-inf fence[nom.sg] adverb teliin (see examples 18 and 19). ympäriis aama-n, elles ne (18) FiV (Mägiste 1948: 24) around stack-gen.sg otherwise those lehmä-t ja muu-t eläime-t Sitten se doimitt-i midengä se gäv deliin Then he tell-sg3.pret how it go.sg3.pret to cow[nom]-pl and other-pl animal[nom]-pl sö-i ylös ‘Then he told how it happened’ eat-sg3.pret up (19) FiV (Mägiste 1948: 38) ‘There should be a fence around the [hay]stack, otherwise the cows and other animals would eat [it] up’ Se güsü si-ldä: ”mideŋgä seiso-o he ask.sg3.pret she-abl.sg how stand-sg3.pres Another example with ylös is shown in example 16, where the adverb su-n deliin nüd?” you-gen.sg to now is combined with the verb kirjoittaa ‘write’ forming a unit equivalent to the Swedish phrasal verbs skriva upp ‘note’. This construction also ‘He asked her: ”How are you now?”’ appears in modern Standard Finnish (cf. ISK § 455). In example 17, the expression kasvattaa ylös does not correspond to a phrasal verb, The adverb appears in constructions which have Swedish phrasal but to the compound uppfostra ‘raise’, where upp is the prefix to the verbs including till ‘to’ as the model. In example 16, käydä teliin cor­ verb fostra ‘raise, foster’. responds to Swe. gå till ‘happen’ and in example 17, seisoa teliin cor­ responds to Swe. stå till, as in the expression Hur står det till? ‘How (16) FiV (Mörtberg 2011: 94) are you?’. The adverb has apparently occurred under the influence of Sen poika ei huoli-nna oppi-a, it-gen.sg boy[nom.sg] neg.sg3 want-conneg.pret learn-inf Swedish till, cf. SweV te ‘to’ (Warmland 1997 s.v. te; Söder 2011: vaikka isä kirjoitt-i ylös 127), and has been adapted to Värmland Finnish morphology. Other even_though father[nom.sg] write-sg3.pret up examples of Värmland Finnish adverbs of Swedish origin are ramil- ja meina-is opetta-a len (< Swe. fram ‘there, through, forward’, cf. SKES s.v. prami) and and intend-sg3.pret teach-inf ahtiin (< Swe. akt ‘attention’; cf. Mägiste 1960: 166) appearing in ‘His son did not want to learn [the father’s knowledge], even examples 20 and 21. though the father wrote [it] down and intended to teach [him]’

210 211 TORBJÖRN SÖDER THE FINNISH OF RAUTALAMPI AND VÄRMLAND – A COMPARISON

(20) FiV (Mörtberg 2011: 121) In the Värmland Finnish material, I have not found examples where Niin kevää-llä oll-tiin sitten flyötingi-ssä, the passive is used as the form for first person plural. Instead forms so spring-ade.sg be-pass.pret then timber_floating-ine.sg with the suffix -mmo(n)/-mmö(n) are used (see example 24). niin jotta ne sa-i flyötingi-n ramillen so that they get-sg3.pret timber_floating-acc.sg through (24) FiV (Mägiste 1948: 43) ‘So in spring you would be engaged in the timber müö ommon müö-nä kaikki_tǖni floating, in order to get the timber floating through’ we be.pl1.pres sell-pptcp everything_completely ‘we have sold everything’ (21) FiV (Mörtberg 2011: 218) Mikkeli-stä joulu-un ne saa tiete-ä, Michaelmas-ela.sg Christmas-ill.sg they get.sg3.pres know-inf The Värmland Finnish suffix -mmo(n)/-mmö(n) has its origin in the se-n pitä-ä pan-na ahtiin Savo dialects and it corresponds to Standard Finnish -mme. The la­ (s)he-gen.sg have_to-sg3.pres put-inf attention bial vowel of the suffix has developed due to influence of the labial si-llä väli-llä that-ade.sg interval-ade.sg nasal consonants (Nirvi 1947: 38). The widespread use of the suffix ‘From Michaelmas to Christmas they can tell9 from it [the Milky -mmo(n)/-mmö(n) in Värmland Finnish indicates that such forms were th Way], you have to pay attention to it during that period’ in use in the 16 century in Savo, but as illustrated in examples 22 and 23, the use of the suffix has decreased in Rautalampi Finnish to the advantage of passive forms. Nirvi explains this change as having 2.7. Passive forms and first person plural a sociolinguistic reason, meaning that because of its labial vowel and perhaps also because of the length created by -n, the suffix would have A common feature in colloquial Finnish with regard to the verbal sys­ conflicted with that of the neighbor dialects, which ultimately led to its tem is the use of passive forms instead of active verb forms including loss (Nirvi 1947: 39). The passive forms were already in use as the first the suffix -mme in the first person plural, e.g., me sanotaan ‘we say’ person plural imperative and could therefore easily be transferred to pro me sanomme and me mentiin ‘we went’ pro me menimme (Karls­ the use in the active voice (Nirvi 1947: 39–40). In Värmland, the pas­ son 2009: 302; ISK § 1326). In Rautalampi Finnish, as is shown in sive was also used as the first person plural imperative, e.g., otetaan examples 22 and 23, this feature is common, too. ‘let us take’ (Standard Finnish ottakaamme) (Aminoff 1876: 218), but the use of passive as the first person plural active voice was never (22) FiR (Yli-Luukko 1983: 25) introduced. In Finland, the passive as the form of first person plural ’er rei( ä-stä müö lähe-ttī other hole-ela.sg we leave-pass.pret has spread, according to Mielikäinen (1991: 50), to a much wider area ‘we went through another hole’ than what appears to be the case according to Kettunen’s survey from 1940 (map 164). (23) FiR (Lehtinen 1982: 17) müö män-tīn tuone’ 2.8. Lexicon we go-pass.pret there ‘we went over there’ The lexicon of the two varieties has been formed by their differing contexts. Swedish loanwords in Värmland Finnish bear witness to the contact with Swedish. In the Värmland Finnish examples, several 9. You can tell whether there will be a lot of snow or not. Swedish loanwords appear: buara (example 14) < Swe. bara ‘only’

212 213 TORBJÖRN SÖDER THE FINNISH OF RAUTALAMPI AND VÄRMLAND – A COMPARISON

(Fi. vain); fintti(example 31) < Swe. fin‘nice’ (Fi. hieno); fiuli(exam ­ Other instances of päivä meaning ‘sun’ are viskais hopeata vasten ple 32) < Swe. fiol ‘violin’ (Fi. viulu); flyötinki (example 20) < SweV päiveä ‘they threw silver towards the sun’ and päiveä ihailemaan ‘in flöting ‘timber floating’ (Fi. uitto); kallata (example 8) < Swe. kalla order to admire the sun’ (Mörtberg 2011: 130, 318). ‘call’ (Fi. kutsua, nimittää); kup (examples 8, 11 and 32) < Swe. gubbe Some of the lexical differences between the two varieties have ‘man’ (Fi. mies, ukko); meinata (example 16) < Swe. mena ‘mean’ occurred as the result of deviant lexicalization of new concepts which (Fi. aikoa); staanata (example 14) < Swe. stanna ‘stop’ (Fi. pysähtyä). were introduced to the speech communities after the Forest Finns’ set­ Värmland Finnish, on the other hand, has preserved Eastern tlement of Scandinavia. Finnish and non-standard lexical features such as hukka ‘wolf’ pro The potato became a very popular crop in the beginning of the susi and päivä ‘sun’ pro aurinko (see examples 25 and 26), whereas 19th century. In Värmland Finnish, putis(i) is used to denote the crop Rautalampi Finnish shows many lexical features that instead are West­ and in Rautalampi peruna, which etymologically correspond to Stand­ ern and Standard Finnish, e.g., vuosi ‘year’ (cf. FiV ajastaika), tuli ard Swedish potatis and Standard Finnish peruna, respectively. In ‘fire’ (cf. FiV valkea). both Finnish and Swedish dialects there exist words for ‘potato’ with the opposite etymology (cf. SKES s.v. potaatti; SAOB P3105). (25) FiV (Mörtberg 2011: 46) As rail transport was introduced in the 19th century a word mean­ Ku “hukka” sano-i, se sö-i kaikki when wolf[nom.sg] say-sg3.pret it eat-sg3.pret all ing ‘train’ became necessary. In Värmland Finnish, the word paana is eläime-t, mutt en muista found and in Rautalampi juna (see examples 27 and 28). The Värm­ animal-acc.pl but neg.sg1 remember.conneg.pres land Finnish word is a loan from Swedish (bana ‘path, line, track’), mi-ksi si-tä sano-ttiin=kaan, vaan which has been lexicalized as the means of transportation. The Finn­ what-transl.sg it-part.sg say-pass.pret=at_all but ish juna originates from a word meaning ‘queue, row’ and its modern hukka ei saa-nu sano-a meaning ‘train’ has been in use since the middle of the 19th century wolf[nom.sg] neg.sg3 get-conneg.pret say-inf (SSA s.v. juna; Hakulinen 1979: 428). ‘When they said “hukka”, it ate all animals, but I don’t remember what they called it, but you could not say hukka’ (27) FiV (Mörtberg 2011: 232) (26) FiV (Mörtberg 2011: 222) Ne piili-llä kävv, ja se makso-i they car-ade.sg go.sg3.pret and it cost-sg3.pret Ja ne ott-i kolt naula-a ja yltä 200 kruunu-a se ois and they take-sg3.pret three nail-part.sg and more_than 200 crown-part.sg it be.sg3.cond kaivo-vat hampaa-sta ver-ta, ja naula-si-vat tul-lu paljo halve-mpi paana-lla dig-pl3.pret tooth-ela.sg blood-part.sg and nail-pret-pl3 become-pptcp much cheap-comp.nom.sg train-ade.sg seittyse-en kuuse-en jok oll ikään ‘They went by car, and it cost more than 200 crowns, such-ill.sg spruce-ill.sg which be.sg3.pret like it would have been much cheaper to go by train’ pohoise-ssa, jo-hon ei paljo p ä i v ä north-ine.sg which-ill.sg neg.3sg much sun[nom.sg] (28) FiR (Yli-Luukko 1983: 46) paista-nu shine-conneg.pret tulˊ juna’-onnettom̀ ūs Pieksäm̀ ä̀ e-llä become.sg3.pret train_accident[nom.sg] Pieksämäki-ade.sg ‘And they took three [wooden] plugs and dug blood from the tooth, and they plugged it into such a spruce, ‘a train accident happened in Pieksämäki’ which faced north, which didn’t get much sun’

214 215 TORBJÖRN SÖDER THE FINNISH OF RAUTALAMPI AND VÄRMLAND – A COMPARISON

In the 20th century, the automobile became an increasingly common (29) FiR (Yli-Luukko 1983: 18) means of transportation and in both Rautalampi and Värmland words ne mua-t olˊ si-llä lālla originating from the German word Automobil, which first occurred in those land[nom]-pl be.sg3.pret that-ade.sg manner print in 1886 (Häkkinen 1987 s.v. auto), are found. In Värmland, the järjeste-ttü enne, oja-t si-llä word for ‘automobile’ is piil(i) (see example 27), an adaptation of the arrange-pass.pptcp before ditch[nom]-pl that-ade.sg th lāla että Swedish bil ‘automobile’ attested in Swedish in the early 20 century manner that (SAOB B2511). In Rautalampi, auto is used, but in other Finnish varie­ ‘the lands were arranged in that manner before, ties in Finland, there is also evidence of piili being used (SSA s.v. piili). the ditches, in that manner that’ Some of the lexical differences between the two varieties are not only due to new concepts being lexicalized in different ways, but have (30) FiR (Yli-Luukko 1983: 22) occurred as a result of neologisms coined in 19th century Finland, when se=kkī pit ol-lan nīj a new written standard language was created to meet the needs of the that=too have_to.sg3.pret be-inf such järestükse-ssǟ developing Finnish society. The increasing use of Finnish in contexts order-ine.sg where Swedish earlier had been used and changes in society created a ‘that too had to be in such order’ need for a renovation of the Finnish vocabulary. There existed unwill­ ingness among scholars to use Swedish loans to fill the gaps. Instead el­ ements from the Finnish language were used in the renewal. More than In the Värmland Finnish material, I have not found the verb or the noun, 150 of the 1 000 most common Finnish words are, judging from written but the examples below show ways of expressing similar meanings sources, neologisms of which the majority originates from the 19th cen­ in Värmland Finnish. In example 31, the verb tällätä (< Swe. ställa tury (Häkkinen 1985: 155). These neologisms have also found their way ‘put, set’; SKES s.v. tällätä) is used. In example 32, oornink(i) (< Swe. into the language of speakers, whose language appears in the records. ordning ‘order’) is used together with soana (cf. FiSt perfect partici­ The verb järjestää ‘organize, arrange’ (see example 29) and ple saanut < inf. saada ‘get’) and peällä ‘on’ (cf. FiSt päällä ‘on’) to the corresponding noun järjestys ‘order, system’ (see example 30) form an expression like Swedish få ordning på något ‘put something in both originate from järki ‘sense, reason, row’ (SSA s.v. järki). The order’. older concrete meaning of the word is still preserved in the adverb (31) FiV (Mörtberg 2011: 78) järjestään ‘in a row’ and an earlier meaning of the verb is ‘put in a ”Ei=pä se niin fintti oo, vaan row’ (Hakulinen 1979: 404). In the 1850s, the verb appears in print neg.sg3=quite that so nice[nom.sg] conneg.pres but with the modern meaning ‘organize’ (Europæus 1853 s.v. organisera). ko minä se-n tällöö-n”, mehtä-herra10 The noun järjestys was initially more widely spread than the verb in when I it-acc.sg arrange.pres-sg1 forest_lord[nom.sg] 19th century dictionaries (Hakulinen 1979: 460 footnote). sano-i say-3sg.pret ‘“Not quite so fine, but when I arrange it”, the lord of the forest said’

10. A mythical character appearing in Värmland Finnish folklore.

216 217 TORBJÖRN SÖDER THE FINNISH OF RAUTALAMPI AND VÄRMLAND – A COMPARISON

(32) FiV (Mägiste 1960: 56) of kone and kapine in Mörtberg’s records is a result of contacts with Se kuṕ olˊ osta-na uuve-n fiuli-n Finnish linguists who frequently paid visits to the Värmland Finns in that man[nom.sg] be.sg3.pret buy-pptcp new-acc.sg violin-acc.sg order to make records. siinä, van sei soa-na yhtää The occurrence of commercial companies in the increasingly in­ there but he.neg.3sg get-pptcp any dustrialized 19th century required a suitable vocabulary. In Finland, oorninki-a se-n peällä order-part.sg it-gen.sg on the old noun yhtiö was introduced in the 1840s. Initially, it was used ‘This man had bought a violin there, but he could not get it in order’ in several meanings, but towards the end of the century, the mean­ ing ‘(commercial) company’ (36) was established (Hakulinen 1979: The word kone is an old Finnish word and the meaning ‘tool’ (SSA s.v. 455, 464). In Värmland, the word puulaak (Mägiste 1960: 174), from kone) has been attested in the dialects. The modern meaning ‘machine’ Swedish bolag ‘(commercial) company’, is used in the same mean­ is mentioned in print in 1838 (Helenius s.v. maschin), but the word ing (see example 37). Mörtberg’s spelling bolaag is not phonetic. To was probably in use in the modern meaning already in the 18th century indicate Swedish loanwords Mörtberg sometimes uses a spelling in­ (Hakulinen 1979: 442 footnote). In example 33, there is talk about fluenced by Swedish orthography. Another example from Mörtberg’s haymaking in the times before mowers were introduced. records is djurit ‘animals’ (Mörtberg 2011: 46) from Swedish djur ‘animal’, cf. jyyr ‘animal’ (Mägiste 1960: 169). (33) FiR (Yli-Luukko 1983: 12) ku e ol-t konne-i-ta sielä (36) FiR (Lehtinen 1982: 40) when neg.sg3 be-conneg.pret machine-pl-part there ühti jö-m pomò, joka olˊ ‘when there were no machines there’ company-gen.sg boss[nom.sg] who be.sg3.pret ‘the company’s boss, who was’ In Mörtberg’s records the word kone appears once, in a comment about Mörtberg’s recording equipment (see example 34). Another word for (37) FiV (Mörtberg 2011: 63) ‘machine’ in Värmland Finnish is masina, which appears in example 35. Ne men sitten Härjedaali-in se they go.sg3.pret then Härjedalen-ill.sg that (34) FiV (Mörtberg 2011: 101) bolaag ott-i nii-ltä torpa-t Niin mu-n niin pitä-ä sano-a jonniin company[nom.sg] take-sg3.pret they-abl cottage-acc.pl so I-gen.sg so have_to-sg3.pres say-inf some.acc.sg ‘Then they went to Härjedalen, the company took their cottages’ sana-n tuo-hon konee-seen word-acc.sg that-ill.sg machine-ill.sg ‘So I have to say some words into that machine’ 3. Conclusions (35) FiV (Mörtberg 2011: 277) vaan nyt nii-llä on hakku-masina-t This short comparison between the Finnish of Rautalampi and the but now they-ade be.sg3.pres chopping-machine[nom]-pl Finnish of Värmland shows that the differences in their respective ‘but nowadays they have chopping machines’ contact situations have had an impact. The lexical level shows salient differences. Värmland Finnish has retained, on the one hand, East­ The word kapine appears in Mörtberg’s records as the word for ‘tool’ ern Finnish lexical features (e.g., päivä ‘sun’, hukka ‘wolf’) and, on (ibid. 134). According to SSA (s.v. kapine) the word does not exist in the other, incorporated words from Swedish (e.g., piil(i) ‘automobile’, Värmland Finnish. It seems quite plausible to think that the appearance oornink(i) ‘order’). Some lexical features of clear Swedish origin are

218 219 TORBJÖRN SÖDER THE FINNISH OF RAUTALAMPI AND VÄRMLAND – A COMPARISON more difficult to define as loans into Värmland Finnish. The Värmland References Finnish adverb ramillen ‘there, through, forward’, for instance, has etymological counterparts in Finnish in Finland, cf. tulla (p)ramille Aminoff, Torsten 1876: Tietoja Wermlannin Suomalaisista [Information on Finns in Värmland]. – Suomi 11:2: 161–244. ‘come out’ (SKES s.v. prami ) and it cannot be ruled out that the word Bladh, Gabriel 1995: Finnskogens landskap och människor under fyra already existed in the language at the time of the settlement of Scan­ sekler. En studie av natur och samhälle i förändring [Landscape and dinavia; although the contact with Swedish probably has strengthened people in Finnskogen during four centuries. Study of nature and soci­ its position. In Rautalampi Finnish, the impact from Standard Finnish ety in the process of change]. Göteborg: Kulturgeografiska institu­ is evident in the lexicon. Neologisms coined in the 19th century often tionen & Göteborgs universitet. contrast with Swedish loans in Värmland Finnish. — 2001: Bakgrund och historia [Background and history]. – Maud Wedin The influence from Swedish is evident in Värmland Finnish con­ (ed.), Det skogsfinska kulturarvet. Falun: Finnsam. 11–26. structions similar to Swedish phrasal verbs, e.g., syö ylös ‘eats up’, Bredin, Inge 2009: Skogsfinnarna och kyrkan [Forest Finns and the church]. even though similar constructions are not unattested in the Finnish of – Värmländsk kultur 3–4: 18–20. Finland (cf. ISK § 455). Broberg, Richard 1952: Kyrkan och värmlandsfinnarna [The church and the The isolation of Värmland Finnish from other varieties of Finnish Forest Finns]. – Erik Palin (ed.), Karlstads stift i ord och bild. Stock­ has in some respects preserved archaic features, such as the use of verb holm: Idun. 113–130. Ethnologue = Lewis, M. Paul, Gary F. Simons & Charles D. Fennig (eds), forms with -mmo(n)/-mmö(n) in Värmland instead of passive forms for Ethnologue. Languages of the World. Eighteenth edition. Dallas: SIL first person plural and the conservation of the reflexive conjugation. International. [Accessed 15 May 2015.] Available at: ‹https://www. ethnologue.com/about/language-status› Abbrevations Europæus, Daniel 1853: Svensk-finskt handlexikon. Ruotsalais-suomalainen sanakirja [Swedish-Finnish lexicon]. Suomalaisen Kirjallisuuden Seu­ 1 first person ill illative ran Toimituksia 16. Helsinki: Suomalaisen Kirjallisuuden Seura. 3 third person ine inessive Finnson, John 1976: F. Prästen John Finnson berättar om gångna tider i abl ablative inf infinitive Södra Finnskoga [Former pastor John Finnson talks about the old days acc accusative nom nominative in Södra Finnskoga]. – Finnskogen 2: 8. ade adessive part partitive Fishman, Joshua A. 1991: Reversing language shift. Clevedon: Multilingual comp comparative pass passive Matters. cond conditional pl plural Häkkinen, Kaisa (ed.) 1987: Nykysuomen sanakirja 6. Etymologinen sana- conneg connegative pptcp past participle kirja [Dictionary of Modern Finnish 6. Etymological dictionary]. Hel­ ela elative pres present tense sin ki: WSOY. ess essive pret Hakulinen, Lauri 1979: Suomen kielen rakenne ja kehitys [Structure and Fi. Finnish refl reflexive development of Finnish]. Helsinki: Otava. FiR Rautalampi Finnish sg singular Hormia, Osmo 1970: Finska dialekter. En översikt [Finnish dialects. An FiSt Standard Finnish Swe. Swedish overview]. Lund: LiberLäromedel. FiV Värmland Finnish SweV Swedish dialect of Värmland ISK = Hakulinen, Auli, Maria Vilkuna, Riitta Korhonen, Vesa Koivisto, gen genitive transl translative Tarja Riitta Heinonen & Irja Alho 2004: Iso suomen kielioppi [Com­ prehensive grammar of Finnish]. Suomalaisen Kirjallisuuden Seuran Toimituksia 950. Helsinki: Suomalaisen Kirjallisuuden Seura. Jokipii, Mauno 1988: Keski-Suomen historia II [History of Central Finland II]. Jyväskylä: Keski-Suomen maakuntaliitto. 220 221 TORBJÖRN SÖDER THE FINNISH OF RAUTALAMPI AND VÄRMLAND – A COMPARISON

Karlsson, Fred 2009: Suomen peruskielioppi [Basic grammar of Finnish]. Mielikäinen, Aila 1991: Murteiden murros. Levikkikarttoja nykypuhekielen Suomalaisen Kirjallisuuden Seuran Toimituksia 378. Helsinki: Suo­ piirteistä [Change of dialects. Distribution maps of features of col­ malaisen Kirjallisuuden Seura. loquial language]. Suomen kielen laitoksen julkaisuja 36. Jyväskylä: Kettunen, Lauri 1910: Descendenttis-äännehistoriallinen katsaus Keski- Jyväskylän yliopisto. Skandinavian metsäsuomalaisten kieleen [Sound historical overview Mörtberg, Matti 2011: Matti Mörtbergs värmlandsfinska uppteckningar. of the language of the Forest Finns of Central Scandinavia]. – Suomi Sammanställda och kommenterade av Torbjörn Söder [Värmland 8:4. 1–149. Finnish notes by Matti Mörtberg. Compiled and commented on by Tor­ — 1930: Suomen murteet II. Murrealueet [Finnish dialects II. Dialect björn Söder]. Acta Academiae Regiae Gustavi Adolphi 118. Uppsala: areas]. Suomalaisen Kirjallisuuden Seuran Toimituksia 188. Helsinki: Kungl. Gustav Adolfs akademien för svensk folkkultur. Suomalaisen Kirjallisuuden Seura. Nirvi, Ruben Erik 1947: Passiivimuotojen aktiivistumisesta [On the activi­ — 1940: Suomen murteet III A [Finnish dialects III A]. Suomalaisen Kir­ zation of passive forms]. – Suomi 104:4. 1–47. jallisuuden Seuran Toimituksia 188. Helsinki: Suomalaisen Kirjalli­ SAG III = Teleman, Ulf, Lisa Christensen, Staffan Hellberg & Erik Anders­ suuden Seura. son 1999: Svenska akademiens grammatik. 3, Fraser [Grammar of — 1960: Kahdeksan matkaa Vermlannin metsäsuomalaisiin [Eight trips Swedish Academy. 3, Phrases]. Stockholm: Svenska akademien. to the Forest Finns in Värmland]. Helsinki: Kauppakirjapaino. Saloheimo, Veijo 1986: Savolaxarnas rörlighet och expansion på 1500- och Koivisto, Jouko 1990: Suomen murteiden refleksiivitaivutus [The reflexive 1600-talen [Mobility and expansion of Savo people in the 16th and 17th conjugation in Finnish dialects]. Suomalaisen Kirjallisuuden Seuran centuries]. – Sulo Huovinen (ed.), Värmlandsfinnar. Om finnskogens Toimituksia 532. Helsinki: Suomalaisen Kirjallisuuden Seura. historia och kultur. Stockholm: Nordstedts Tryckeri. 7–22. Koivisto, Vesa 1995: Itämerensuomen refleksiivit [Reflexives in Finnic]. — 2009: “Det ser så ut”: försök att identifiera svedjefinnarnas ursprung Suomalaisen Kirjallisuuden Seuran Toimituksia 622. Helsinki: Suo­ [“It seems so”: an attempt to identify the origin of Finns in Sweden]. malaisen Kirjallisuuden Seura. [Accessed 17 Nov 2015.] Available at: ‹http://www.migrationinstitute. Laine, Päivi 2007: Suomi tiellä sivistyskieleksi. Suomenkielisen maantieteen fi/files/pdf/oppaat/Svedjefinnarnas_ursprung.pdf› sanaston kehittyminen ja kehittäminen 1800-luvulla [Finnish on its way SAOB = Ordbok över svenska språket [Dictionary of the Swedish language] to becoming a language of education. The development of and developing (1898‒). Lund: Svenska akademien. Available at: ‹https://www.saob.se/› Finnish geographical vocabulary in the 19th century]. Turun yliopiston SKES = Toivonen, Y[rjö] H[enrik], Erkki Itkonen, Aulis J. Joki & Reino Pel­ suomalaisen ja yleisen kielitieteen laitoksen julkaisuja 77. Turku: Turun tola 1955–1981: Suomen kielen etymologinen sanakirja [Etymological yliopisto. Available at: ‹https://www.utupub.fi/handle/10024/29030› dictionary of Finnish] I–VII. LSFU XII. Sanahakemiston (VII) koosta­ Lehtinen, Raija 1982: Pihtiputaan murretta [The dialect of Pihtipudas]. neet Satu Tanner ja Marita Cronstedt. Helsinki: Suomalais-Ugrilainen Suomen kielen näytteitä 13. Helsinki: Kotimaisten kielten tutkimus­ Seura. keskus. Söder, Torbjörn 2011: Henrik Hanssonin vermlanninsuomalainen käännös Mägiste, Julius 1948: Metsäsuomalaismurteiden nykyisestä vaiheesta [On maaherra Henrik Adolf Widmarkin viisivuotiskertomuksesta vuosilta the present state of Forest Finnish dialects]. Suomi-kirjan eripainoksia. 1881–1885 [Report by Governor Henrik Adolf Widmark from 1881– Uusi sarja 21. Helsinki: Suomalaisen Kirjallisuuden Seura. 1885 translated into Värmland Finnish by Henrik Hansson]. – Sanan- — 1955: De värmlandsfinska ortnamnens vittnesbörd [A record of place jalka 53: 109–137. Available at: ‹https://doi.org/10.30673/sja.86707› names in Värmland Finnish]. Småskrifter utgivna av Värmlands — 2012: “Skriv som du talar”: Carl Axel Gottlund och den finska museum 1. Karlstad: Värmlands museum. ortografin [“Write as you speak”: Carl Axel Gottlund and the Finnish — 1960: Vermlannin sammuvaa savoa. Kielennäytteitä vuosilta 1947‒51 ortography]. – Eberhard Winkler (ed.), Lapponicae investigationes et [The dying Savo dialect of Värmland: language samples from 1947– Uralicae: Festschrift zum 65. Geburtstag von Lars-Gunnar Larsson. 51]. Suomalaisen Kirjallisuuden Seuran toimituksia 265. Helsinki: 223–231. Suomalaisen Kirjallisuuden Seura.

222 223 TORBJÖRN SÖDER THE FINNISH OF RAUTALAMPI AND VÄRMLAND – A COMPARISON

SSA = Itkonen, Erkki & Ulla-Maija Kulonen (eds) 1992–2000: Suomen Vertailussa Rautalammin ja Vermlannin suomi sanojen alkuperä. Etymologinen sanakirja [Origin of Finnish words. Etymological dictionary]. Suomalaisen Kirjallisuuden Seuran Toi­ Torbjörn Söder mituksia 556, Kotimaisten kielten tutkimuskeskuksen julkaisuja 62. Helsinki: Kotimaisten kielten tutkimuskeskus & Suomalaisen Kirjal­ Keskiajan loppupuolella väestöryhmiä alkoi siirtyä Savon historial­ lisuuden Seura. lisen maakunnan keskiosasta maakunnan reuna-alueille. Muuttoliike Thomason, Sarah Grey 2001: Language contact. An Introduction. Washing­ ulottui maakunnan rajojen ulkopuolelle ja lopulta Skandinavian keski­ ton, D.C.: Georgetown University Press. osaan. Savolaisen asutuksen takia Rautalammin suurpitäjä perustet­ Topelius, Zacharias 1908: Maamme kirja. Lukukirja Suomen alimmille oppilaitoksille [Book of our land. Reading book for elementary schools tiin 1560-luvulla Hämeen historiallisen maakunnan pohjoisosaan ja in Finland]. Helsinki: Edlund. 1600-luvun puolivälissä samaa alkuperää olevaa väestöä asettui Ruot­ Tuomi, Tuomo 1989: Suomen murteiden sanakirja. Johdanto [Dictionary of sin Vermlantiin. Finnish dialects. Introduction]. Kotimaisten kielten tutkimuskeskuk­ Vermlanninsuomi ja rautalamminsuomi ovat läheisiä varie­ sen julkaisuja 36. Helsinki: Kotimaisten kielten tutkimuskeskus. teetteja, jotka olivat erillään toisistaan yli 300 vuotta. Vermlannin­ Warmland, Knut 1997: Värmländsk ordbok. De värmska dialekternas ord- suomi eristyi, toisin kuin rautalamminsuomi, suomen kielen murre­ skatt [Dictionary of Värmland Swedish. Vocabulary of the Värmland jatkumosta ja standardisuomesta, jonka merkitys puolestaan kasvoi dialect]. Stockholm: Wahlström & Widstrand. Suomessa 1800-luvun loppupuolella. 1900-luvun toisella puoliskolla Yli-Luukko, Eeva 1983: Rautalammin murretta [The dialect of Rautalampi]. verm­lanninsuomi­ sammui. Suomen kielen näytteitä 18. Helsinki: Kotimaisten kielten tutkimus­ Erilaiset olosuhteet Rautalammilla ja Vermlannissa kehittivät keskus. varieteetteja eri suuntaan. Molemmat kielimuodot luetaan suomen — 1997: Saarijärven murretta [The dialect of Saarijärvi]. Suomen kielen itämurteisiin, mutta vermlanninsuomessa näkyy myös vaikutusta näytteitä 47. Helsinki: Kotimaisten kielten tutkimuskeskus. ruotsista ja samalla arkaaisia piirteitä kun taas rautalamminsuomi on saanut vaikutteita suomen muista varieteeteista. Sanastossa esiintyy eri kielikontaktitilanteesta johtuvia eroja. Leksikaalisesti vermlanninsuomi eroaa joissakin suhteissa rautalam­ minsuomesta, koska vermlanninsuomi on säilyttänyt vanhoja murre­ sanoja (esim. ajastaika pro vuosi, valkea pro tuli) ja toisaalta saanut lainoja ruotsista (esim. oornink(i) pro järjestys, piil(i) pro auto). Sanastollisten erojen lisäksi vermlanninsuomen eristynyt asema on johtanut siihen, että eräät morfosyntaktiset piirteet, jotka leimaa­ vat suurta osaa suomen kielen puhuma-alueesta, puuttuvat tältä varie­ teetilta. Esimerkkejä verbintaivutuksesta ovat monikon ensimmäisen persoonan suffiksin-mmo(n)/-mmö(n) käytön ja refleksiivitaivutuksen säilyminen. Rautalamminsuomessa käytetään monikon ensimmäises­ sä persoonassa finiittimuotojen sijaan passiivimuotoja ja refleksiivi­ taivutus esiintyy varieteetissa katoavana piirteenä.

224 225 IRINA NOVAK Venäjän tiedeakatemian Karjalan tiedekeskuksen Kielen, kirjallisuuden ja historian instituutti

Venäjän kielen vaikutus tverinkarjalan murteiden äännejärjestelmään1

Abstrakti Tverinkarjala kuuluu varsinaiskarjalan eteläisiin murteisiin ja jakautuu kolmeen alamurteeseen: Tolmatšun, Ves­ jegonskin ja Djoržan jo sammuneeseen alamurteeseen. Tverin­ karjalaiset ovat asuneet jo neljä vuosisataa venäläisten keskuu­ dessa ja puhuneet venäjää pitkään. Venäjän kieli on vaikuttanut huomattavasti tverinkarjalan äännejärjestelmään sen kehityk­ sen eri vaiheissa. Tästä johtuvat muutamat alamurteita toisis­ taan erottavat piirteet. Venäjän vaikutusta tverinkarjalan äänne­ järjestelmään on sekä vokaali- että konsonanttijärjestelmässä. 1) vokaalijärjestelmässä: pitkien uu-, yy-vokaalien ja u-, y-lop­ puisten diftongien jälkikomponenttien redusoituminen; sekun­ daaristen pitkien vokaalien esiintyminen; väljenevien diftongien fonemaattiset erot; Djoržan alamurteessa loppu- ja sisäheitto; 2) konsonanttijärjestelmässä: f- ja c-konsonanttifoneemien esiin­ tyminen; sibilanttien distribuutio; geminaattojen kk, tt, pp, čč reduktio l-, r-, n-konsonanttien jäljessä; runsas liudentuminen ja kvalitatiivisen astevaihtelun epäproduktiivisuus. Avainsanat karjalan kieli, tverinkarjala, venäjän vaikutus, äänne­ op­pi, murteet

1. Artikkeli on valmistunut Venäjän tiedeakatemian Karjalan tiedekeskuksen budjettirahoituksella.

Multilingual­ Finnic. Language contact and change. 227–248. Uralica Helsingiensia 14. Helsinki 2019. ‹https://doi.org/10.33341/uh.85037› IRINA NOVAK VENÄJÄN KIELEN VAIKUTUS TVERINKARJALAN MURTEIDEN ÄÄNNEJÄRJESTELMÄÄN

1. Johdanto 200 000

180 000 Kielikontaktit vaikuttavat kielen muuttumiseen huomattavasti: ne vaikuttavat erityisesti kielen sanastoon ja äännerakenteeseen, morfo­ 160 000 logiaan taas vähemmässä määrin. Rakenteidenkin siirtymistä kielestä toiseen kuitenkin tapahtuu. Tässä artikkelissa annetaan yleiskatsaus 140 000 venäjän vaikutuksesta tverinkarjalan alamurteiden äännejärjestelmään. 120 000 Artikkeli tarkastelee eri-ikäistä venäjän vaikutusta tverinkarjalassa­ ja erityisesti tverinkarjalan alamurteille ominaisia piirteitä. 100 000 Karjalaiset ovat asuneet Tverin alueella noin 400 vuotta. Ensim­ 80 000 mäinen karjalaisten muuttoaalto Novgorodin Karjalasta Sisä-Venäjälle (entisiin Novgorodin ja Tverin lääneihin) ajoittui vuosiin 1581–1595, 60 000 jolloin Ruotsi valtasi Käkisalmen. Toinen muuttoaalto oli Stolbovan rauhan jälkeen v. 1617–1661 (ks. Virtaranta 1961: 32–36). 1900-lu­ 40 000 vun alussa tverinkarjalaisia oli Tverin alueella noin 141 000 henkeä 20 000 (Golovkin 2001: 19–42). Nykyään Tverin Karjalassa asuu noin 7 000 karjalaista ja vain kolmannes heistä osaa karjalaa (Vserossijskaja pe­ 0 repis naselenija 2010). 1834 1859 1873 1895 1897 1926 1933 1937 1939 1959 1970 1979 1989 2002 2010 Ensimmäinen Tverin Karjalassa käynyt tutkija oli suomalainen Kuvio 1. Tverinkarjalaisten määrä Tverin alueella (Golovkin 2001: 225–226; Vse­ David Emanuel Daniel Europaeus vuonna 1848. Kesällä 1882 alueella rossijskaja perepis naselenija 2010). matkasi Theodor Schvindt. Vuonna 1895 Tverissä kävi kaksi nuorta tutkijaa: Vihtori Alava ja Kusti Karjalainen. Vuosina 1912–1913 mer­ kittävän työn teki Juho Kujola. Tutkijoiden keräämä aineisto sisälsi sanastoa, runoja ja kielennäytteitä (Virtaranta 1961: 7–8; KKS 1968: XV–XXIV). Neuvostoaikana tverinkarjalan äänne- ja muoto-oppia Tverinkarjala jakautuu kolmeen alamurteeseen. Pohjoisin näistä tutkivat Aleksandr Beljakov, Tiit Kukk, Grigori Makarov, Konstantin sijaitsee Vesjegonskin piirissä ja sen naapuripiireissä (Krasnyi Hol­ Manžin, Paula Palmeos, Aleksandra Punžina ja Jaan Õispuu. Vuonna min ja Sandovon piirit). Vesjegonskin alamurretta puhuu noin viisi 1957 ja useana kesänä sen jälkeen Tverin Karjalassa retkeili suoma­ prosenttia tverinkarjalaisista. Eteläisin ja pienin tverinkarjalan puhu­ lainen tutkija Pertti Virtaranta, joka keräsi materiaalia aihepiirien mu­ ma-alue sijaitsi alueen lounaisosassa (Zubtsovin piiri), Djorža-joen kaan, yhteensä 6 600 sanalippua ja 300 tuntia haastatteluja (Joki 2011: varrella. Tämä seutu kärsi pahoin toisessa maailmansodassa, sillä 70–73). rintama kulki sen läpi. Nykyisin Djoržan alamurre on jo hävinnyt. Tverinkarjala kuuluu varsinaiskarjalan etelämurteisiin samoin Suurin osa karjalaa puhuvista tverinkarjalaisista, noin 95 prosenttia, kuin Tihvinän ja Valdain saarekemurteet, Raja-Karjalan murteet (Ilo­ puhuu Tolmatšun alamurretta. Tverin alueen Lihoslavlin, Spirovon, mantsi, Korpiselkä, Suojärvi, Suistamo, Impilahti) ja Karjalan tasa­ Rameškin ja Maksatihan piireissä (alueen keskiosassa) kieli on säily­ vallassa Pohjois-Aunuksessa puhuttavat murteet (Tunkua, Suikujärvi, nyt ja karjalaa kuulee puhuttavan vielä nykyäänkin. Juuri Tolmatšun Repola, Rukajärvi, Paatene, Mäntyselkä, Porajärvi, Voijärvi) (Torikka alamurteen pohjalta kehitettiin 1900-luvun alkupuolella tverinkarjalan 2004). kirjakieltä (Golovkina toim. 2001: 11–13).

228 229 IRINA NOVAK VENÄJÄN KIELEN VAIKUTUS TVERINKARJALAN MURTEIDEN ÄÄNNEJÄRJESTELMÄÄN

Tverinkarjalan kolmen alamurteen äänne- ja muoto-oppi eroa­ vat toisistaan. Djorža sijaitsee kaukana Tolmatšusta ja Vesjegonskista, ja sen alamurre eroaa huomattavasti näistä kahdesta muusta alamur­ teesta (SKG: 5). Tolmatšun ja Vesjegonskin alamurteiden foneettiset isoglossit ja näiden alamurteiden välinen raja eivät sulaudu yhteen. Vesjegonskin alamurre jaetaan vielä itä- ja länsiryhmään, joista jäl­ kimmäinen on foneettisesti samanlainen kuin Tolmatšun alamurre. Tolmatšun ja Vesjegonskin alamurteiden muoto-opillisen järjestelmän vertaileva tutkimus ei kuitenkaan paljasta merkittäviä eroja (Novak 2014: 48–49). Tverinkarjalaismurteiden foneettiset ja muoto-opilliset ominais­ piirteet ovat peräisin kahdesta lähteestä. Osa ominaispiirteistä on pe­ rua muinaiskarjalan murteista, sillä siirtolaiset Novgorodin Karjalasta toivat uudelle asutuspaikalle oman paikallismurteensa. Osa piirteistä on venäjän kielen vaikutusta (Novak 2016: 28–89). 1900-luvun alkupuolella Tverin Karjalassa kaksikielisyys oli vakiintunut, ja vuosina 1939–19891 assimilaatio venäläisiin voimis­ tui huomattavasti (Gromova 1997: 13). Näistä syistä venäjän kieli on vaikuttanut merkittävästi Sisä-Venäjän karjalan murteisiin. Kieleen on tullut hyvin huomattava määrä venäläisiä lainasanoja, ja muutamilla aloilla, esimerkiksi kulttuurin, politiikan, tekniikan ja talouden alal­ la, lainasanakerrostuma on vallitseva. Lainasanojen kautta venäjä on vaikuttanut murteiden äänne- ja lauseoppiin. Morfologia ja taivutus­ järjestelmä eivät tavallisesti ole yhtä herkkiä muiden kielten vaikutuk­ sille, mutta tverinkarjalan alamurteissa myös ne ovat silti joutuneet venäjän kielen vaikutuksen alaisiksi. Tverinkarjalan muoto-opillisen järjestelmän tärkeimpiä venäjän vaikutuksesta aiheutuneita muutoksia ovat marginaalisten sijamuotojen väheneminen ja tämän seuraukse­ na toisien sijojen merkitysten laajeneminen, omistusliitteiden kato, kieliopillisen superlatiivin kato ja menneen ajan tempusjärjestelmän yksinkertaistuminen (Zlobina 1967: 9–10; 1971: 31–37). Seuraavassa käsittelen sitä, miten venäjä on vaikuttanut tverinkarjalan alamurtei­ den vokaali- ja konsonanttijärjestelmään.

1. Ajanjakso 1939–1989 oli tverinkarjalaisten täydellisen unohtamisen aikaa. Karjalankieliset kirjat oli poltettu ja karjalan kielen opetus kouluissa oli lopetet­ Kartta 1. Karjalan kielisaarekkeet Sisä-Venäjällä (Kukk 1989: 81 OKR:n ja Virtarannan tu (Golovkin 2006: 18). Tverinkarjalaisia ei mainittu virallisesti missään vuosien (1973: 44) mukaan). 1939–1989 aikana (Golovkina toim. 2001: 17).

230 231 IRINA NOVAK VENÄJÄN KIELEN VAIKUTUS TVERINKARJALAN MURTEIDEN ÄÄNNEJÄRJESTELMÄÄN

2. Venäjän vaikutus tverinkarjalan ʼköyhäʼ, löwdˊiä ~ Drž. löwdˊi ʼlöytääʼ, löwvän ʼlöydänʼ, powda ~ Drž. alamurteiden vokaalijärjestelmään powd ʼpoutaʼ, riwgu ~ Drž. riwg ʼriukuʼ, šiwla ~ Vsg. śiwla ~ Drž. šiwl ʼsinullaʼ, kuwži ~ Drž. kuwž ʼkuusiʼ, kuwven ʼkuudenʼ, luw ʼluuʼ, Tverinkarjalan vokaalijärjestelmä on esitetty taulukossa 1. pyw ʼpyyʼ, tuwlˊi ~ Drž. tuwlˊ ʼtuuliʼ, tywńi ~ tywń ʼtyyniʼ. Karjalan kielen muissakin murteissa klusiilien heikon asteen edellä (v:n edellä) etinen keskinen takainen uu:n, yy:n ja u-, y-loppuisen diftongin jälkikomponentti on muuttunut lavea pyöreä lavea lavea pyöreä puolivokaaliksi (Kujola 1910: 19; E. Leskinen 1933: 62, 69), mutta suppeat i y [i̮ ] u vahvassa asteessa nämä ovat säilyneet muuttumattomina. Vesjegons­ puolisuppeat e ö [e̮ ] o kin itäryhmässä ja paikoin Tolmatšussa alkuperäiset diftongit ja pitkät väljät ä a vokaalit ovat kuitenkin säilyneet, mikä vuorostaan viittaa siihen, että Taulukko 1. Tverinkarjalan vokaalijärjestelmä. ilmiö ei ole vanha, sillä se esiintyy säännönmukaisesti vain Valdain murteessa (Palmeos 1962: 8–11). Alkuperäisten pitkien uu-, yy-vokaa­ Vokaalit e ja i on esitetty kahdella allofonilla, mutta tällä ominai­ lien ja u-, y-loppuisten diftongien redusoituminen on ilmeisesti venä­ suudella ei ole merkitystä erottavaa tehtävää. Etisen tai keskisen jän vaikutusta, koska kummatkaan eivät ole venäjän kielen vokaali­ allo­fonin esiintyminen riippuu useimmiten siitä, onko sana taka- tai järjestelmälle tyypillisiä. Sitä paitsi keskivenäjän murteessa esiintyy etuvokaalinen. w-puolivokaali samassa asemassa, nimenomaan vokaalin jäljessä (vrt. Karjalan kielessä on monoftongien lisäksi pitkiä vokaaleja sekä lu­ ven. murt. pra[w]dа ʼtosiʼ, ven. murt. la[w]kа ʼpenkkiʼ) (Orlova toim. kuisia väljeneviä ja suppenevia diftongeja ja triftongeja. Tverinkarjala­ 1970: 34). ei ole tässä suhteessa poikkeus, vaikka venäjä onkin vaihtelevassa Tverinkarjalan alamurteissa, samoin kuin monissa muissa karja­ määrin vaikuttanut tverinkarjalan alamurteiden vokaalijärjestelmään. lan kielen murteissa, puolivokaalia tavataan myös preesensin yksikön Tässä artikkelissa annettujen esimerkkisanojen lähteet ovat OKR kolmannen persoonan päätteessä, esim. ottaw ʼhän ottaaʼ, kuččuw ʼhän 1963, DKT 1990, NKK 1994, SKJ 1994, DKV 1995, SKG 2001, kutsuuʼ, keźŕiäw ~ Drž. keźŕiw ʼhän kehrääʼ, šuaw ʼhän saaʼ, šyöw ~ KPKS 2014, kirjoittajan litteroimat kielennäytteet vuosilta 1950–1970 Vsg. syöw ʼhän syöʼ, rubiew ʼhän rupeaaʼ. Puolivokaali tässä ase­ ja kirjoittajan vuosina 2009–2016 keräämät kenttäaineistot. Murre on massa palautuu kantasuomen kolmannen persoonan *pi-tunnukseen määritelty tarkemmin vain silloin kun muoto esiintyy ainoastaan jos­ (*pi > vi > v > w > u) (Zaikov 2000: 50–51). On mahdollista, että täl­ sakin tietyssä murteessa. laiset puolivokaalit ovat laajentaneet w:n esiintymisalaa alkuperäisten u-, y-loppuisten diftongien ja pitkien vokaalien jälkikomponenttina. 2.1. w-puolivokaalin esiintyminen pääpainollisten alkuperäisten pitkien uu-, yy-vokaalien ja u-, y-­ 2.2. Väljenevien diftongien fonemaattisia loppuisten diftongien jälkikomponenttina eroavaisuuksia tverinkarjalan sisällä

Kaikille tverinkarjalan alamurteille on luonteenomaista w-puoli­ Karjalan kielen luonteenomaisten piirteiden joukkoon kuuluu alku­ vokaalin esiintyminen. w-puolivokaali on läheinen u-, y-vokaaleille, peräisten pitkien vokaalien aa, ää, oo, öö, ee diftongiutuminen (*aa > ja se on kehittynyt juuri näistä labiaalivokaaleista niiden redusoitu­ oa > ua, *ää > eä (> iä), *oo > uo, *öö > yö, *ee > ie), mikä saattaa hy­ misen seurauksena. w-puolivokaali esiintyy suppenevien diftongien vinkin olla muinaiskarjalaista alkuperää (H. Leskinen 1998: 370–371, ja alkuperäisten pitkien vokaalien jälkikomponenttina. Esimerkkejä 379). Tämä muutos koskee myös pitkiä supistumavokaaleja. tästä ovat kawha ~ Drž. kawh ʼkauhaʼ, kewhä ~ köwhä ~ Drž. kewh

232 233 IRINA NOVAK VENÄJÄN KIELEN VAIKUTUS TVERINKARJALAN MURTEIDEN ÄÄNNEJÄRJESTELMÄÄN

Kaikissa tverinkarjalan alamurteissa tavataan diftongeja iä, ie, 2.3. Sekundaariset pitkät vokaalit r:n edellä uo, yö. Vesjegonskin itäryhmässä (Potšep, Djudikovo, Mosejevskoje, Petrjajka, Benjakovo, Trufanovo, Golovkovo, Pokryškino, Ostolo­ Pitkät vokaalit eivät ole tyypillisiä venäjän vokaalijärjestelmälle, mikä povo, Možaika sekä naapurikylät) ja Tolmatšun lounaisryhmässä on todennäköisesti vaikuttanut pitkien uu:n, yy:n, ii:n katoamiseen (Ploskoje, Timoškino, Dalʹnitši sekä naapurikylät) esiintyy väljene­ tverinkarjalan­ alamurteista (Beljakov 1949: 71–72; Palmeos 1966: vä i̮ a-diftongi, mutta Vesjegonskin länsiryhmässä sekä Tolmatšun 1–2). Pitkät vokaalit uu, yy ja ii ovat useimmiten diftongiutuneet (*uu > ja Djoržan alamurteessa esiintyy sen sijaan diftongi ua (Bubrih ym. uw, *yy > yw (ks. 2.1.), *ii > ij (paino on alkukomponentilla, jälkikom­ 1997: kartat 4, 6, 15, 17, 19; OKR: 6). Esimerkkejä diftongeista ovat ponentti on painoton samoin kuin i-loppuisella laskevalla diftongilla), Vsg. hi̮ aba ʼhaapaʼ, mi̮ a ʼmaaʼ, mi̮ amo ʼäitiʼ, ni̮ agla ʼneulaʼ, ri̮ adaja eikä tverinkarjalassa enää eroteta vokaalien pituusastetta. Kuitenkin ʼtyöntekijäʼ, ti̮ aš ʼtaasʼ, ti̮ atto ʼisäʼ (vrt. Tlm. Drž. huaba ~ huab, mua, tverinkarjalan alamurteissa vartaloissa, joissa on kvalitatiivisen aste­ muamo ~ muam, nuagla ~ nuagl, ruadaja ~ ruadai, tuaš, tuatto ~ tuat). vaihtelun alaisten konsonanttiyhtymien rg : rr > r, rd : rr > r heikko Diftongin i̮ a ensimmäinen komponentti (keskivokaali i̮ ) on venäjän aste, r:n edellä esiintyy sekundaarinen pitkä vokaali аа, oo, uu, ää, kielestä lähtöisin. Sama i̮ a-diftongi esiintyy myös livvin Kondušin ja öö, yy, ee, ii, esim. härgä ~ Drž. härg : häärät ʼhärkäʼ, kerda ~ Drž. varsinaiskarjalan Paatenen paikallismurteessa (Virtaranta 1973: 272). kerd : keerat ʼkertaʼ, kurgi ~ Drž. kurg : kuuret ʼkurkiʼ, märgä ~ Drž. Luultavasti i̮ a-diftongin i̮ -komponentti on kehittynyt jokaisessa näis­ märg : määrät ʼmärkäʼ, pirda ~ Drž. pird : piirat ʼpirtaʼ, šordua ~ Vsg. tä murteista itsenäisesti venäjän vaikutuksesta. Toisaalta on hyvinkin šordi̮ a­~ Drž. šordu : šooran ʼsortaaʼ. Vokaali on ensin pidentynyt ja sen mahdollista, että ilmiö on yhteistä alkuperää, koska Paatenesta tiede­ jälkeen geminaatta lyhentynyt, mikä on ilmeisesti venäjän vaikutuksen tään siirtyneen karjalaisia Tverin Karjalaan (Virtaranta 1961: 42–44). tulosta (Zlobina 1967: 98; Virtaranta 1972: 13, 17). Sekundaariset pit­ Tässä tapauksessa kyse voisi olla vanhemmasta venäläisvaikutuksesta. kät vokaalit kvalitatiivisen astevaihtelun alaisten konsonanttiyhtymien Djoržan alamurteessa tavataan väljenevien uo-, ua-, yö-difton­ rg : rr > r, rd : rr > r heikossa asteessa esiintyvät vain tverinkarjalassa gien redusoitumista, esim. Drž. kizuwmah ʼpelaamaanʼ, ei šuw ʼei (Bubrih ym. 1997: kartta 60), mikä todistaa, että ilmiö on hyvin nuori. saaʼ, istuwč ʼistuuduʼ, rebuw ʼkettuaʼ, kaluw ʼkalaaʼ (vrt. Tlm. Vsg. kizuamah, ei šua, istuoče, rebuo, kaluo). Luultavasti tämäkin ilmiö 2.4. Loppu- ja sisäheitto on venäjän fonetiikan vaikutusta. Redusoitumisen rinnalla väljene­ vät uo-, yö-, ie-diftongit (uo < *oo, yö < *öö, ie < *ee) ovat avartu­ Djoržan alamurteen tärkein äänneopillinen erityisominaisuus on neet muutamissa Djoržan alamurteessa seuraavalla tavalla: *uo > ua, sisä- ja loppuheitto, jotka ovat venäjän paikallismurteiden vaikutusta *yö > yä, *ie > iä etuvokaalisissa sanoissa ja *ie > ia takavokaalisissa (Virtaranta­ 1972: 12; Õispuu 1985: 181). sanoissa (Palmeos 1966: 3). Esimerkkejä tästä ovat Drž. tuadih ʼhe Djoržan alamurteen loppuheitossa vokaalit ja diftongien jälkikompo­ toivatʼ, myä ʼmeʼ, myän ʼminä myynʼ, vyä ʼvyöʼ, yä ʼyöʼ, viälˊ ʼvieläʼ, nentit heittyvät kaksi- ja monitavuisista sanoista, mutta joskus loppuvokaa­ biass ʼpaholainenʼ, pialuš ʼtyynyʼ, viaraš ʼvierasʼ (vrt. Tlm. Vsg. tuo- lin paikalla esiintyy redusoitunut e-vokaali (Palmeos 1966: 2–3; SKG: 6). dih, myö, myön, vyö, yö, vielä, biesa, pieluš, vieraš). Tämä kehitys Esimerkkejä loppuheitosta ovat Drž. a, ä: akk ʼakkaʼ, om ʼomaʼ, kylˊ ʼkyläʼ, on todennäköisesti omaperäinen, koska ua-, yä-, iä-, ia-diftongien miśś ʼmissäʼ; i: jog ʼjokiʼ, kiv ʼkiviʼ, kaikki šub ʼkaikkiʼ; e: ič ʼitseʼ, kolm rinnalla tavataan Djoržassakin uo-, yö-, ie-diftongeja. Samantapaista ʼkolmeʼ, naiń ʼnainenʼ; o, ö: kukk ʼkukkoʼ, reb ʼkettuʼ, ukk ~ ukke ʼukkoʼ, avartumista havaitaan myös Aunuksen Riipuškalassa ja suomen länsi­ dˊiad ʼvaariʼ, lötˊtˊ ʼsammakkoʼ; u, y: kowkk ʼkoukkuʼ, kylˊ ʼsaunaʼ, kydˊ murteissa (Rapola 1966: 359–361). Djoržan alamurteen diftongien ʼvävyʼ; diftongin jälkikomponentti: kevi ʼkevätʼ, šuappu ʼsaapasʼ (vrt. avartuminen on nähtävästi yhteydessä redusoitumiseen, ts. murteella Tlm. Vsg. akka, oma, kylä, missä; jogi, kivi, kaikki šubi; iče, kolme, naińe; voisi olla pyrkimys tasoittaa redusoitumista tai säilyttää diftongeja. kukko, rebo, ukko, dˊiedo, löttö; kowkku, kyly, kydy; keviä, šuappua).

234 235 IRINA NOVAK VENÄJÄN KIELEN VAIKUTUS TVERINKARJALAN MURTEIDEN ÄÄNNEJÄRJESTELMÄÄN

Loppuvokaalin kato on ominaista myös lyydin ja vepsän mur­ 3. Venäjän vaikutus tverinkarjalan teille. Loppuheitto kuuluu lyydin vanhoihin perusominaisuuksiin, ja alamurteiden konsonanttijärjestelmään ainakin lyydin loppu-i:n kato vertautuu vepsän edustukseen (Turunen 1950: 170, 253). Djoržan alamurteen loppuheitto on myöhäsyntyinen Venäjä on vaikuttanut monin tavoin jo muinaiskarjalan konsonantti­ ilmiö, koska viimeisten murteen vanhemman polven puhujien kielessä järjestelmän kehitykseen. Sen vaikutusta on k-, t-, p-konsonanttien sitä ei vielä esiintynyt. On hyvinkin mahdollista, että vokaalin heitty­ soinnillistuminen muualla kuin vienankarjalassa, č-affrikaatan synty­ minen sanan lopussa on venäjän murteiden vaikutusta, koska ilmiö on minen, uusien konsonanttiyhtymien ja kolmen konsonantin yhtymien tyypillinen venäjän kielen paikallismurteissa, vrt. ven. murt. ulits < käyttäminen, sananalkuisten konsonanttiyhtymien runsaus sekä liu­ ulitsa ʼkatuʼ, utr < utro ʼaamuʼ (SKG: 6; Õispuu 1985: 181). dentuminen (Kalima 1934; Beljakov 1949: 83–98; H. Leskinen 1964: Tolmatšussa ja Vesjegonskissa esiintyy yksittäisiä loppuheitto­ 99; 1998: 357, 378–380). Tämä vaikutus on jatkunut myös nykyisten tapauksia yhdyssanoissa, joissa tavallisesti määriteosa on kaksi­ karjalan murteiden erilliskehityksen aikana. tavuinen ja sen ensimmäinen tavu on pitkä. Tällaisissa yhdyssanoissa Tverinkarjalan konsonanttijärjestelmä on esitetty taulukossa 2. vokaalialkuisen perusosan edellä määriteosan loppuvokaali on kadon­ nut, joskus myös konsonanttialkuisen perusosan edellä. Esim. Tlm. heiń|aiga ʼheinäaikaʼ, heiń|arga ʼheinäarkiʼ, käź|aštie ʼkäsiastiaʼ, toiń|argi ʼtiistaiʼ, täm|piänä ʼtänäänʼ; Vsg. jumal|koda ʼikoninurkkaʼ,

peld|aitta ʼpeltoaittaʼ, šuol|vakka ʼsuola-astiaʼ, šiidˊ ʼsittenʼ. Saman­ bilabiaalinen labiodentaalinen dentaalinen alveolaarinen palataalinen velaarinen laista loppuheittoa tavataan karjalan kielen muissakin murteissa, vaik­ klusiilit soinniton p t tˊ k ka se onkin erilähtöinen kuin Djoržan loppuheitto. soinnillinen b d dˊ g Sisäheitto koskee painottoman ja joskus sivupainollisen ta­ frikatiivit sibilantit soinniton s ś š vun vokaaleja, jos niiden jälkeen seuraa umpitavu (Рalmeos 1966: soinnillinen z ź ž 3; SKG: 6). Esimerkkejä sisäheitosta ovat Drž. a, ä: hambhan spirantit soinniton f h obstruentit ʼhampaanʼ, puhthan ʼpuhtaanʼ, pelvhat ʼpellavatʼ, elˊmäh ʼelämäänʼ, soinnillinen v hyptäh ʼhe hyppäävätʼ; e: šiämnet ʼsiemenetʼ, elˊtäh ʼhe asuvatʼ; o, ö: affrikaatat soinniton c č šantah ʼhe sanovatʼ, hebźen ʼhevosenʼ, ruadmah ʼtyöskentelemäänʼ; soinnillinen dž u, y: kučtah ʼhe kutsuvatʼ, ikknan ʼikkunanʼ, vöntäh ʼhe makaavatʼ puolivokaalit w j (vrt. Tlm. Vsg. hambahan, puhtahan, pelvahat, elämäh, hypätäh; likvidat lateraalit l lˊ šiemenet, eletäh; šanotah, hebozen ~ heboźen, ruadomah; kučutah, tremulantit r ŕ resonantit ikkunan, vönytäh ~ venytäh). Palmeosin (1966: 4) mukaan sisäheitto nasaalit m n ń [ŋ] kehittyi Djoržan alamurteessa loppuheiton jälkeen, eli sanan lopussa syntynyt vokaalin heitto levisi toiseen tavuun saakka. Karjalan kie­ Taulukko 2. Tverinkarjalan konsonanttijärjestelmä. len murteissa painottoman tavun vokaali on kestoltaan erittäin lyhyt (E. Leskinen 1933: 95) ja sisäheitto voi olla näissä omaperäinen il­ Loppuheiton vuoksi Djoržan alamurteessa sanan lopussa voi esiintyä miö, ts. ne ovat juurtuneita pikapuhemuotoja. On myös mahdollista, mikä tahansa konsonantti ja konsonanttiyhtymiäkin, mutta Tolmatšun että sisäheiton leviäminen Djoržassa on venäjän vaikutusta. Djoržan ja Vesjegonskin alamurteissa sananloppuisina esiintyvät vain konso­ sisäheitto ei kuitenkaan ole vanha, koska sisäheittotapausten rinnalla nantit h, l, lˊ, n, ś, š, t. tavataan sisäheitottomia muotoja.

236 237 IRINA NOVAK VENÄJÄN KIELEN VAIKUTUS TVERINKARJALAN MURTEIDEN ÄÄNNEJÄRJESTELMÄÄN

Geminaattoina esiintyvät useimmat konsonantit, eivät kuiten­ 3.2. Sibilanttien distribuutio tverinkarjalassa kaan soinnilliset obstruentit, puolivokaalit ja h. Djoržan alamurteessa sananloppuinen geminaatta lyhenee säännönmukaisesti loppuvokaa­ Venäjän kieli on vaikuttanut monella tapaa karjala-aunuksen s:n kehi­ lin heiton vuoksi. tykseen (H. Leskinen 1998: 378). Virtarannan (1984: 260–261) mu­ Karjalan kielen kaikissa murteissa on lukuisia konsonantti­ kaan i-vokaalin edellä s on muuttunut š:ksi livvissä ja lyydissä, eikä yhtymiä. Sisäheiton tuloksena Djoržan alamurteessa esiintyy enem­ muinaiskarjalassa ollut konsonanttia š. Kuitenkin on hyvin mahdollis­ män sekä kahden että kolmen konsonantin yhtymiä kuin muissa kar­ ta, että tämä muutos on tapahtunut ennen muinaiskarjalan myöhäis­ jalan murteissa ja konsonanttiyhtymien frekvenssi on huomattavasti vaihetta, koska 1500-luvun ”verokirjan paikannimissä on usein z, š, ž korkeampi. Djoržassa tavataan myös kokonaisten tavujen katoamista, suomalaisen s:n paikalla” (Kalima 1934: 255). mikä johtuu pyrkimyksestä päästä eroon vokaaliheiton jälkeen muo­ Tverinkarjalan Tolmatšun ja Djoržan alamurteissa sekä Vesje­ dostuneista kolmen konsonantin yhtymistä. Tavallisesti tässä on kyse gonskin länsiryhmässä (Timoškino, Terpigora, Verhneje, Černjagino geminaattojen lyhenemisestä. Esim. Drž. pelhat ~ pelvhat ʼpellavatʼ, ja naapurikylät) s esiintyy vain i:n tai i-loppuisen diftongin jäljessä, tytˊren ʼtyttärenʼ, aštah ʼhe astuvatʼ (vrt. Tlm. Vsg. pelvahat, tyttären, kun taas š-konsonanttia käytetään kaikissa muissa asemissa sekä etu- aššutah). että takavokaalisissa sanoissa. Vesjegonskin itäryhmässä (Potšep, Dju­ dikovo, Mosejevskoje, Petrjajka, Benjakovo, Trufanovo, Pokryškino, 3.1. Konsonantit c ja f Ostolopovo ja naapurikylät) takavokaalisissa sanoissa esiintyy š, kun taas etuvokaalisissa sanoissa on s (Bubrih ym. 1997: kartat Venäjän kielestä ovat lähtöisin konsonantit c ja f, jotka esiintyvät mel­ 74–100). Poikkeuksena ovat Vesjegonskin itäryhmän itäisimmät kylät kein kaikissa karjalan kielen murteissa. Nämä konsonantit ovat tulleet (Čamerovo ja naapurikylät), joissa kaikissa asemissa esiintyy konso­ kieleen venäläisten lainojen mukana ja esiintyvät nimenomaan laina­ nantti s. Konsonanttien ž ja z distribuutio tverinkarjalan alamurteis­ sanoissa (Beljakov: 11). Todennäköisesti konsonantit c ja f ovat tulleet sa on samanlainen. Esimerkkejä s:n ja š:n esiintymisestä ovat Tlm. tverinkarjalaan ja samaan aikaan karjalan kielen muihin murteisiin Vsg. (länsiryhmä) Drž. šie ʼsinäʼ, šyön ʼsyönʼ; Tlm. Vsg. kakši ~ Drž. vasta vähän aikaa sitten. Aikaisemmin lainatuissa sanoissa venäjän c- kakš ʼkaksiʼ, kyläššä ~ Drž. kyläšš ʼkylässäʼ, kežä ~ Drž. kež ʼkesäʼ, affrikaatan paikalla on č-affrikaatta ja f:n paikalla esiintyy jokin muu kynži ~ Drž. kynž ʼkynsiʼ, istuo ~ Drž. istu ʼistuaʼ, tukista ~ Drž. tu- konsonantti (ks. Kalima 1952: 68–71), esim. bolˊńičča ~ Drž. bolˊńičč kist ʼhiuksistaʼ, kaččuośśa ʼkatsoessaʼ; vrt. Vsg. (itäryhmä) śie, syön, < ven. bolˊńica ʼsairaalaʼ, jäiččä ~ Drž. jäič < ven. jaico ʼmunaʼ, kakśi, kylässä, kezä, kynźi, ištuo, tukišta, kaččuošša; Vsg. (itäryhmän kolˊča ~ Drž. kolˊč < ven. kolˊco ʼsormusʼ, melˊlˊiččä ~ Drž. melˊlˊičč < itäosa) suola ʼsuola’, suossa ʼsuossa’. ven. melˊńica ʼmyllyʼ, ogurča ~ Drž. ogurč < ven. ogurec ʼkurkkuʼ; Paatenen paikallismurteissa on samanlainen sibilanttien distri­ škuappa ~ Drž. škuap < ven. škaf ʼkaappiʼ, Tlm. Vsg. sveŕeźi < ven. buutio kuin Vesjegonskin itäryhmässä. Virtaranta (1961: 42–44) ker­ ferjaź ʼvenäjäläisen kansallispuvun osaʼ, Hilˊippä < ven. Filˊipp, Jehro too: ”Perimätiedon mukaan osa asukkaista (Selkin kylästä) oli lähte­ < ven. Jefrem. Nuoremmissa lainoissa käytetään konsonantteja c ja nyt Tverin lääniin. – Selkin starikka Timo Tuurujev kertoi tavanneensa f. Esimerkkejä näistä ovat Tlm. Vsg. cirkulˊi < ven. cirkulˊ ʼharppiʼ, Pogreloin kylässä Selkin tapaan puhelevia karjalaisia, jotka olivat ker­ fanera < ven. fańera ʼvaneriʼ, forma < ven. forma ʼmuotoʼ, forśie < toneet esi-isäinsä siirtyneen sinne Selkistä.” Vesjegonskin piirissä on ven. forśitˊ ʼkoreillaʼ; Drž. cigan < ven. cigan ʼromaniʼ, finn < ven. finn edelleen Pogorelovo-kylä. Jos kyse on ollut juuri tästä kylästä, tämä ʼsuomalainenʼ. tieto sen asukkaiden lähtöpitäjästä selittäisi Vesjegonskin alamurteen sisäiset erot.

238 239 IRINA NOVAK VENÄJÄN KIELEN VAIKUTUS TVERINKARJALAN MURTEIDEN ÄÄNNEJÄRJESTELMÄÄN

3.3. kk-, tt-, pp-, čč-geminaattojen reduktio On myös tapauksia, joissa liudentuminen ei riipu konsonan­ l-, r-, n-konsonanttien jäljessä tin äänneympäristöstä. Liudentunutta konsonanttia voi seurata taka­ vokaali, ja liudennus voi esiintyä myös sanan lopussa. Esimerkkejä Tverinkarjalassa, samoin kuin Tihvinän, Porajärven ja Paatenen etelä­ liudennuksesta eri asemissa ovat malˊlˊa ~ Drž. malˊlˊ ʼmaljaʼ, ńaba ~ karjalaisissa murteissa sekä livvin ja lyydin murteissa, alkuperäisten Drž. ńab ʼnapaʼ, vihaźa ʼvihainenʼ; jäńiś ʼjänisʼ, valmiś ʼvalmisʼ. geminaattojen paikalla on lyhyt konsonantti l-, r-, n-konsonanttien jäl­ Säännönmukainen liudennus koskee dentaalikonsonantteja, jot­ jessä (Beljakov 1949: 98; Bubrih ym. 1997: kartat 54–56). Esimerkkejä ka esiintyvät ä-, ö-, y-vokaalien edessä ja lisäksi konsonanttiyhtymien reduktiosta ovat maltua ~ Drž. malta ʼymmärtääʼ, pertˊi ~ Drž. pertˊ alkukomponenttina muiden liudentuneiden konsonanttien ja j:n edes­ ʼpirttiʼ, pilkuo ~ Drž. pilku ʼpilkkoaʼ, šulku ~ Drž. šulk ʼsilkkiʼ, tartuo ~ sä. Esim. lˊötˊtˊö ~ Drž. lˊötˊtˊ ʼsammakkoʼ, tˊytˊtˊö ~ Drž. tˊytˊtˊ ʼtyttöʼ, Drž. tartu ʼtarttuaʼ, turki ʼturkkiʼ, venčä ~ Drž. veńč ʼvihkiäisetʼ, värči ~ lˊöwlˊy ~ Drž. lˊöwlˊ ʼlöylyʼ; elˊäńdˊä ~ Drž. elˊäńdˊ ʼelämäʼ, häńdˊä ~ Drž. värč ʼsäkkiʼ. Geminaattojen lyheneminen tässä asemassa on näh­ Drž. häńdˊ ʼhäntäʼ, maŕja ʼmarjaʼ, ńyblˊä ~ Drž. ńyblˊ ʼnappiʼ, Vsg. tävästi venäjän vaikutusta, sillä geminaatat eivät ole tyypillisiä venä­ śäŕgi ʼsärkiʼ. Pääpainollisten i:n ja e:n edessä esiintyvät tavallisesti jän perussanastolle. Lyheneminen on tapahtunut jo karjalan murteiden liudentuneet dentaalit paitsi r, esim. lˊeibä ~ Drž. lˊeib ʼleipäʼ, ńero ~ erilliskehityksen­ aikana, mihin viittaa soinnittomien klusiilien säilymi­ Drž. ńer ʼkyky, taito; äly, järkiʼ, tˊedŕi ~ Drž. tedŕ ʼteeriʼ, rebo ~ Drž. nen likvidan ja nasaalin jäljessä. On hyvinkin mahdollista, että käsi­ reb ʼkettuʼ. Dentaalikonsonanttien liudentuminen sanan sisällä riip­ teltävä ilmiö eteläkarjalan murteissa on samaa alkuperää kuin karjalan puu useimmiten siitä, onko sana etu- vai takavokaalinen, ts. esiin­ kielen muissa murteissa, mikä merkitsisi sen olevan vanhempaa perua. tyykö sanassa vokaalien etinen vai keskinen allofoni. Esim. kädˊeh ʼkäteenʼ, mäńdˊih ʼhe menivätʼ, vedˊehińe ʼvedenhaltijaʼ, hyö elˊetˊtˊih 3.4. Runsas liudentuminen ʼhe asuivatʼ, hyö jätˊetˊtˊih ʼhe jättivätʼ. Esimerkkejä liudentumattomis­ ta dentaalikonsonanteista takavokaalisessa ympäristössä ovat kačottih Terho Itkosen (1968: 81) mukaan venäjässä esiintyvä liudennus on il­ ʼhe katsoivatʼ, kattila ~ Drž. kattil ʼkattilaʼ, nagriś ʼnaurisʼ, tawdi ~ meisesti vaikuttanut karjalaan ”kaiken aikaa”. Karjalan liudennusta on Drž. tawd ʼtautiʼ. kaiken kaikkiaan pidetty venäjän vaikutuksena, ja se palautuu muinais­ Vesjegonskin alamurteessa liudennus on levinnyt Tolmatšun ja karjalaan, jossa se koski i:tä edeltäviä yksinäisdentaaleja (Itkonen 1968: Djoržan alamurteisiin verrattuna erittäin laajalle. Esim. Vsg. häńeh 100–101). Etuvokaalinen i on karjalassa voinut liudentaa myös sanan­ ʼhäneenʼ vrt. häneh; Vsg. kaŕielaźet naiźet ʼkarjalaiset naisetʼ vrt. alkuisia dentaaleja, alkuperäisiä dentaaligeminaattoja ja useita dentaalin Tlm. kaŕielazet naizet ~ Drž. kaŕilazet; Vsg. latˊtˊiet ʼlattiatʼ vrt. Tlm. sisältäviä konsonanttiyhtymiä. Liudennusta tavataan yhtä lailla i:n kuin Drž. lattiet ~ Drž. lattit; Vsg. mäńen ʼmenenʼ vrt. Tlm. Drž. mänen; muidenkin vokaalien jäljessä (H. Leskinen 1963: 298–299). Vsg. pagiźen ʼpuhunʼ vrt. Tlm. Drž. pagizen; Vsg. peŕeh ʼperheʼ vrt. Liudennus on levinnyt hyvin laajalle kaikissa tverinkarjalan ala­ Tlm. Drž. pereh; Vsg. raviźow ʼhän huutaaʼ vrt. Tlm. Drž. ravizov; murteissa. Sitä esiintyy sanansisäisten, sananalkuisten ja myös sanan­ Vsg. ŕebo ʼkettuʼ vrt. Tlm. rebo ~ Drž. reb; Vsg. ŕihma ʼrihmaʼ vrt. loppuisten dentaalikonsonanttien d:n, l:n, n:n, r:n, s:n, z:n, t:n yhteydessä, Tlm. rihma ~ Drž. rihm; Vsg. ŕiputtih ʼhe riippuivatʼ vrt. Tlm. ri- ja lisäartikulaationa vastaava palataalistuminen koskee myös labiaali- ja puttih ~ Drž. riptih; Vsg. šanotˊtˊih ʼhe sanoivatʼ vrt. Tlm. šanottih ~ velaarikonsonantteja p, b, k, g, m ja vähäisessä määrin konsonantteja Drž. šantih; Vsg. šińńe ~ śińńe ʼsinneʼ vrt. Tlm. Drž. šinne; Vsg. tytäŕ h ja v. Tässä tapauksessa liudennusta ei tavallisesti merkitä kirjoituksessa2. ʼtytärʼ vrt. Tlm. Drž. tytär.

2. Tverinkarjalaisten murteiden tekstilähteissä liudennuksen merkitsemistä ei ole takavokaalien ja j:n edessä sekä sanan lopussa. Näihin asemiin verrattuna ä-, ö-, y- yhdenmukaistettu. Tässä artikkelissa merkitään dentaalikonsonanttien liudentumis­ vokaalien edessä esiintyy aina liudentunut dentaalikonsonantti. Liudennus on mer­ ta, joka esiintyy i- ja e-vokaalien etisten allofonien, liudentuneiden konsonanttien, kitty suomalais-ugrilaisen tarkekirjoituksen mukaisesti.

240 241 IRINA NOVAK VENÄJÄN KIELEN VAIKUTUS TVERINKARJALAN MURTEIDEN ÄÄNNEJÄRJESTELMÄÄN

Loppuheiton tuloksena liudennus on saanut Djoržan alamurtees­ rammat ʼrampaʼ, pitˊkä- : pität ʼpitkäʼ, keške- : keššet ʼkeskiʼ, kisko- : sa merkitystä erottavan tehtävän. Esimerkiksi seuraavissa minimi­ kisson ʼkiskoaʼ, mušta- : muššat ʼmustaʼ, istu- : issun ʼistuaʼ, lˊehte- : pareissa liudennus erottaa toisistaan imperatiivin ja yksikön 3. per­ lˊehet ʼlehtiʼ. rg-, rd-konsonanttiyhtymien astevaihtelun heikossa as­ soonan imperfektin: Drž. män ʼmene’ vrt. mäń ʼhän meni’, ol ʼole’ vrt. teessa geminaatta rr on lyhentynyt sekundaaristen pitkien vokaalien olˊ ʼhän oli’, pan ʼpane’ vrt. pań ʼhän pani’, tul ʼtule’ vrt. tulˊ ʼhän tuli’. jäljessä (ks. esimerkit alaluvusta 2.3.). Todennäköisesti venäjän vai­ Minimiparissa tytˊtˊzen ʼtyttösenʼ vrt. tytˊtˊzeń ʼtyttöniʼ n on yksikön kutuksesta vokaali ensin pidentyi ja sen jälkeen geminaatta lyheni, genetiivin pääte ja liudentunut ń on yksikön ensimmäisen persoonan koska venäjän kielessä geminaatta ei voi olla pitkän vokaalin jäljessä omistusliite. (Zlobina 1967: 98; Virtaranta 1972: 13, 17). Kaikista tverinkarjalan alamurteista on löydettävissä paljon esi­ 3.5. Muutoksia astevaihtelujärjestelmässä merkkejä, jotka kertovat kvalitatiivisen astevaihtelun epäproduktiivi­ suudesta. Esim. Drž. čuwd : čuwdat < ven. čudo ʼihmeʼ, Tlm. diedo : Tverinkarjalassa kvantitatiivisessa astevaihtelussa geminaatat kk, tt, diedolla < ven. dˊed ʼvaariʼ, grib : gribat < ven. grib ʼtattiʼ, kruwg : pp, čč vaihtelevat yksinäiskonsonanttien kanssa (Novak 2014: 75–76), kruwgat < ven. krug ʼympyräʼ, luuga : luugat < ven. lug ʼniittyʼ, pluw- esim. harakka- : harakat3 ʼharakkaʼ, nuotta- : nuotat ʼnuottaʼ, pappi- : ga : pluwgat < ven. plug ʼauraʼ, Vsg. pruwda : pruwdat < ven. prud papit ʼpappiʼ, meččä- : mečät ʼmetsäʼ. Venäjän kielen vaikutusta on ʼlampiʼ, sadu : sadut < ven. sad ʼpuutarhaʼ; huaba : huabat ʼhaapaʼ, kvantitatiivisen astevaihtelun kk : k, tt : t, pp : p, čč : č puuttuminen itku : itkut ʼitkuʼ; Tlm. pada : padašša ~ puašša ʼpataʼ, ruaga : rua­ konsonanttiyhtymissä l-, r-, n-konsonanttien jäljessä, mikä puolestaan gat ~ ruat ʼraakaʼ, ruoda : ruodat ʼruotoʼ, Vsg. ruoga : ruogat ʼruokaʼ, on geminaattojen reduktion tulosta (ks. 3.3.). Esimerkkejä ovat turki- : taba : tabat ~ tavat ʼtapaʼ. Sama tendenssi on luonteenomainen myös turkit ʼturkkiʼ, pertˊi- : pertˊit ʼpirttiʼ, čirpi- : čirpit ʼsirppiʼ, värči- : toisille varsinaiskarjalan ja livvin murteille (Novak 2014: 100–101, värčit ʼsäkkiʼ. Kvantitatiivinen yksittäisgeminaattojen astevaihtelu on 116–117). Kvalitatiivinen astevaihtelu puuttuu venäläisistä laina­ hyvin produktiivinen venäläisissä lainasanoissa. Esim. kazakka- : ka- sanoista ja sama tendenssi leviää myös perussanastoon. zakat ’renki’, luukka- : luukat ’sipuli’, jäiččä- : jäičät ’muna’. Kvalitatiivisessa astevaihtelussa vokaalin jäljessä esiintyvät soinnilliset yksinäiskonsonantit vaihtelevat kadon, j:n tai v:n kanssa 4. Lopuksi (g : Ø ~ j ~ v, d : Ø ~ j ~ v, b : v) ja konsonantin jäljessä olevat klu­ siilit assimiloituvat edellisen konsonantin kaltaiseksi tai katoavat (lg : Karjalaiset ovat asuneet jo neljä vuosisataa Tverin alueella venäläis­ ll, rg : rr, ld : ll, rd : rr, nd : nn, mb : mm, tk : t, šk : šš, sk : ss, št : šš, ten keskuudessa ja puhuneet pitkään venäjää. Tverinkarjalaan on saa­ st : ss, ht : h) (Bubrih ym. 1997: kartat 58–68; Novak 2014: 78–99). tu erittäin paljon venäläisiä lainasanoja, minkä seurauksena karjalan Esimerkkejä soinnillisten yksinäiskonsonanttien astevaihtelusta ovat alkuperäinen äännejärjestelmä on muuttunut. Tämä selittää tverin­ jaga- : juan ʼjakaaʼ, igä- : ijät ʼikäʼ, joge- : jovet ʼjokiʼ; käde- : kiät karjalan alamurteiden foneettisia ominaisuuksia toisiinsa ja muihin ʼkäsiʼ, maido- : maijot ʼmaitoʼ, mado- : mavot ʼmatoʼ; rebo- : revot karjalan kielen murteisiin verrattuna. Toiset käsiteltävistä ilmiöistä ʼkettuʼ. Esimerkkejä konsonanttien jäljessä olevien klusiilien aste­ ovat vanhempia ja ulottuvat jopa muinaiskarjalan loppuvaiheeseen, vaihtelusta ovat jalga- : jallat ʼjalkaʼ, härgä- : härrät ʼhärkäʼ, peldo- : toiset ovat nuorempia ja ne ovat kehittyneet nähtävästi Tverissä. pellot ʼpeltoʼ, parda- : parrat ʼpartaʼ, lˊindu- : lˊinnut ʼlintuʼ, ramba- : Venäjän vaikutusta tverinkarjalan vokaalijärjestelmässä ovat pitkien vokaalien uu, yy ja u-, y-loppuisten diftongien jälkikompo­ 3. Esimerkeissä on esitetty sanan vahva-asteinen vartalo ja monikon nominatiivi tai indikatiivin preesensin yksikön 1. persoonan muoto, joissa esiintyy astevaihtelun nenttien redusoituminen ja sekundaaristen pitkien vokaalien syn­ heikko aste. tyminen. Konsonanttijärjestelmässä venäjän vaikutusta ovat f- ja

242 243 IRINA NOVAK VENÄJÄN KIELEN VAIKUTUS TVERINKARJALAN MURTEIDEN ÄÄNNEJÄRJESTELMÄÄN c-konsonanttifoneemien esiintyminen, geminaattojen reduktio lik­ Lähteet vidojen ja nasaalien jäljessä, runsas liudentuminen ja kvalitatiivisen Beljakov, A. A. 1949: Fonetika karelskogo dialekta s. Tolmatši Kalininskoi aste­vaihtelun epäproduktiivisuus. oblasti [Kalininin alueen karjalan kielen Tolmatšun murteen äänne­ Djoržan alamurteen äänneopillinen järjestelmä eroaa paljon oppi]. – Sovetskoe finno-ugrovedenie 5: 67–98. Tolmatšun ja Vesjegonskin alamurteesta. Djoržan alamurre kehittyi — Rezultaty vlijanija russkogo jazyka na karelskuju fonetiku i leksiku erillisenä pienenä saarekkeena, ja venäjän vaikutus oli siellä voimak­ [Venäjän vaikutuksen tuloksia karjalan kielen äänneoppiin ja sanas­ kaampi. Venäjän paikallismurteiden vaikutusta Djoržan alamurteessa toon. Käsikirjoitus]. Karjalan tiedekeskuksen tietoarkisto 15-1-20. ovat sisä- ja loppuheitto, joiden seurausta on konsonanttiyhtymien Bubrih, D. V., A. A. Beljakov & A. V. Punžina 1997: Karjalan kielen murre­ runsaus sanan sisällä, geminaattojen lyheneminen tietyissä asemissa kartasto. Dialektologičeskij atlas karelskogo jazyka. Toim. Leena sekä konsonanttiyhtymien ja muiden konsonanttien paitsi konsonant­ Sarvas. Venäjän tiedeakatemian Karjalan tiedekeskuksen kielen, kir­ tien t, h, s, l, r, n esiintyminen sanan lopussa. Puhujien voimakkaan jallisuuden ja historian instituutti & Kotimaisten kielten tutkimuskes­ assimilaation seurauksena alamurre on sammunut. kus. Kotimaisten kielten tutkimuskeskuksen julkaisuja 97. Helsinki: Suomalais-Ugrilainen Seura. Venäjän kielen äännejärjestelmä on myös vaikuttanut Vesje­ DKT = Õispuu, Jaan 1990: Djorža karjala tekstid. Tallinn: Tallinna Peda­ gonskin alamurteelle ominaisten piirteiden syntyyn. Vesjegonskin googiline Instituut. alamurteen ominaisuuksia on erittäin runsas liudentuminen. Lisäksi DKV = Õispuu, Jaan 1995: Djorža karjala vormisõnastik. Tallinn: Tallinna Vesjegonskin itäryhmässä verrattuna Tolmatšun ja Djoržan alamurtei­ Pedagoogikaülikool. siin sekä Vesjegonskin länsiryhmään tavataan i̮ a-diftongia ja erilaista Golovkin, A. N. 2001: Istorija Tverskoi Karelii [Tverin Karjalan historia]. sibilanttien distribuutiota. Vesjegonskin länsiryhmä on foneettisesti Tver: TšuDo. läheinen Tolmatšussa puhutulle kielimuodolle. — 2006: Tverin karjalaiset. Tver: Studia-S. Karjalan kieli ei ole lainannut venäjän äännejärjestelmää sellai­ Golovkina, Z. I. (toim.) 2001: Karielazet hormilla keššeššä. Tver: Triada. senaan, lukuun ottamatta konsonantteja f ja c, mutta se on ollut ve­ Gromova, L[judmila] G. 1997: Dvujazyčie kak uslovie i perspektiva näjän yleiskielen ja sen paikallismurteiden voimakkaan vaikutuksen sohranenija i razvitija jazyka i kultury tverskih karel [Kaksikielisyys tverinkarjalaisten kielen ja kulttuurin säilyttämisen ja kehittämisen alaisena. Tämän seurauksena karjalan kielen äännejärjestelmä on ehtona ja näkökulmana]. – A. N. Golovkin (toim.), Istorija i kultura muuttunut ja lähentynyt venäjän järjestelmää. Venäjän vaikutuksesta tverskih karel: perspektivy razvitija. Materialy meždunarodnoi kon­ ovat syntyneet tverinkarjalan tärkeimmät erikoispiirteet, joiden perus­ ferentsii. Tver: Tverskoi gosudarstvennyi universitet. 12–14. teella tverinkarjala jaetaan kolmeen alamurteeseen. Itkonen, Terho 1968: Itäsuomalaisen liudennuksen fonologinen paradoksi. – Fenno-Ugrica. Juhlakirja Lauri Postin kuusikymmenvuotispäiväksi 17.3.1968. Suomalais-Ugrilaisen Seuran toimituksia 145. Helsinki: Lyhenteet Suomalais-Ugrilainen Seura. Joki, Leena 2011: Tverinkarjalan aineistoista ja tutkimuksesta. – L. G. Drž. Djoržan alamurre Gromova (toim.), Tverskie karely: istorija, jazyk, kultura. Materialy Tlm. Tolmatšun alamurre Meždunarodnoi naučno-praktičeskoi konferentsii, posvjaštšennoi Vsg. Vesjegonskin alamurre 20-letiju obrazivanija Obštšestva kultury tverskih karel (Tver, 29-30 oktjabrja 2010 goda). Tver: Tver. gos. un-t. Kalima, Jalo 1934: Entisen Käkisalmen läänin alueen aikaisemmasta kieli­ muodosta. – Virittäjä 38: 254–256. — 1952: Slaavilaisperäinen sanastomme. Helsinki: Suomalaisen Kirjalli­ suuden Seura.

244 245 IRINA NOVAK VENÄJÄN KIELEN VAIKUTUS TVERINKARJALAN MURTEIDEN ÄÄNNEJÄRJESTELMÄÄN

KKS = Virtaranta, Pertti (toim.) 1968: Karjalan kielen sanakirja I. Lexica Rapola, Martti 1966: Suomen kielen äännehistorian luennot. Suomalaisen Societatis Fenno-Ugricae XVI. Helsinki: Suomalais-Ugrilainen Seura. Kirjallisuuden Seuran toimituksia 283. Helsinki: Suomalaisen Kirjal­ KPKS = Novak, I[rina] P. 2014: Rasšifrovki ekspeditsionnyh materialov A. V. lisuuden Seura. Punžinoi 1966–1973 gg. po Tverskoi Karelii [A. V. Punžinan kenttä­ SKG = Punžina, A. V. 2001: Slušaju karelski govor [Kuuntelen karjalaista aineistojen litteroinnit Tverin Karjalasta vuosilta 1966–1973]. – Kultura pakinaa]. : Periodika. povsednevnosti karelskoj semi. Petrozavodsk: Karelski nautšnyi tsentr. SKJ = Punžina, A. V. (toim.) 1994: Slovarʹ karelskogo jazyka: Tverskjie 84–154. govory [Karjalan kielen sanakirja: Tverinkarjalaiset murteet]. Petro­ Kujola, Joh. 1910: Äänneopillinen tutkimus Salmin murteesta. Helsinki: Suo­ zavodsk: Karelija. malaisen Kirjallisuuden Seura. Torikka, Marja 2004: Karjalan kielen murteet. Kotimaisten kielten tutki­ Kukk, Tiit 1989: Vesjegonskin karjalaisten tšastuškoja. Castrenianumin toimit­ muskeskuksen julkaisuja 129. Helsinki: Kotimaisten kielten tutkimus­ teita 32. Helsinki: Castrenianumin laitokset ja Suomalais-Ugrilainen Seura. keskus. [Viitattu 04.12.2015] Saatavissa: ‹http://scripta.kotus.fi/www/ Leskinen, Eino 1933: Tulemajärven murteen vokalismi. Helsinki: Suomalai­ verkkojulkaisut/julk129/karjalan_kielen_murteet.shtml› sen Kirjallisuuden Seura. Turunen, Aimo 1950: Lyydiläismurteiden äännehistoria II. Vokaalit. Suo­ Leskinen, Heikki 1963: Luoteis-Laatokan murteiden äännehistoria. I. Kon- malais-Ugrilaisen Seuran toimituksia XCIX. Helsinki: Suomalais- sonantit. Helsinki: Suomalaisen Kirjallisuuden Seura. Ugrilainen Seura. — 1964: Suomen itämurteet keskiajan ja uuden ajan taitteessa. – Virittäjä Virtaranta, Pertti 1961: Tverin karjalaisten entistä elämää. Helsinki: WSOY. 68: 97–115. Saatavissa: ‹https://journal.fi/virittaja/article/view/34719› — 1972: Die Dialekte des Karelischen. – Sovetskoe finno-ugrovedenie 8: 1–15. — 1998: Karjala ja karjalaiset kielentutkimuksen näkökulmasta. – Pekka — 1973: Über den olonetzischen Konduši-Dialekt. – Finnisch-ugrische Nevalainen & Hannes Sihvo (toim.), Karjala. Historia, kansa, kult- Forschungen XL: 259–277. tuuri. Suomalaisen Kirjallisuuden Seuran toimituksia 705. Helsinki: — 1984: Über das s im Karelischen. – Studien zur phonologischen Suomalaisen Kirjallisuuden Seura. 352–382. Beschreibung uralischer Sprachen. Budapest. 259–274. NKK = Rjagojev, Vladimir & Matti Jeskanen 1994: Näytteitä karjalan kie- Vserossijskaja perepis naselenija 2010. Natsionalnyi sostav naselenija po lestä. Joensuu–Petrozavodsk: Joensuun yliopiston monistuskeskus. subjektam Rossijskoi federatsii [Koko Venäjän kattava väestölas­ Novak, I[rina] P. 2014: Fonetičeskie markery tverskih dialektov karelskogo kenta 2010. Osa 1. Väestön lukumäärä ja jakauma. Taulukko 7 [MS jazyka [Tverinkarjalaisten murteiden äänneopillisia ominaisuuksia]. Excel-taulukko]]. Federalnaja služba gosudarstvennoi statistiki [Venä­ – Finno-ugorskie jazyki i kultury v sotsiokulturnom landšafte Rossii. jän federaation tilastovirasto], 2012. Moskova: ИИЦ ”Статистика Materialy V Vserossijskoi konferentsii finno-ugrovedov. Petrozavodsk: России”. [Viitattu 18.12.2015] Saatavissa: ‹http://www.gks.ru/free_ Karelski naučnyj tsentr RAN. 47–49. doc/new_site/population/demo/per-itog/tab7.xls› — 2016: Tverskie dialekty karelskogo jazyka: Fonetika. Fonologija. [Tve­ Zaikov, P[ekka] M. 2000: Glagol v karelskom jazyke [Karjalan kielen verbi]. rinkarjalaiset murteet: Fonetiikka. Fonologia.] Petrozavodsk: Karelski Petrozavodsk: PetrGU. naučnyj tsentr RAN. Zlobina, V. E. 1967: O nekotoryh zvukovyh izmenenijah v odnom iz govo­ ОKR = Makarov, G. N. 1963: Obraztsy karelskoi reči: Kalininskije govory rov karelskogo jazyka [Erään karjalan kielen paikallismurteen muu­ [Karjalan murrenäytteitä: Kalininin murre]. Moskva–Leningrad: tamista äänteenmuutoksista]. – Sovetskoe finno-ugrovedenie 2: 97–99. Izdatelstvo Akademii nauk SSSR. — 1971: Jazykovyje kontakty i izmenenija v grammatičeskoi strukture Orlova, V. G. (toim.) 1970: Obrazovanjie severnorusskogo narečija i sredne- jazykov [Kielikontaktit ja kielten kieliopillisen järjestelmän muutokset]. russkih govorov [Pohjoisvenäjän ja keskivenäjän murteiden muodos­ – Pribaltijsko-finskoe jazykoznanie: Voprosy vzaimodeistvija pribaltij- tuminen]. Moskva: Nauka. sko-finskih jazykov c inosistemnymi jazykami. Leningrad: Nauka. 33–37. Palmeos, Paula 1962: Karjala Valdai murrak. Emakeele Seltsi toimetised 5. Õispuu, Jaan 1985: Zur Synkope und Apokope in Verbparadigma der Djorža- Tallinn: Eesti NSV Teaduste Akadeemia. Mundart des Karelischen. – Sovetskoe finno-ugrovedenie 3: 171–181. — 1966: Über den Vokalismus der Zubcover Mundart des Karelischen. – Sovetskoe finno-ugrovedenie 1: 1–5. 246 247 IRINA NOVAK

Russian influence on the phonetics VESA KOIVISTO of the Tver dialects of Karelian Itä-Suomen yliopisto

Irina Novak

The Tver have lived in Tver for nearly 400 years. They are Rajakarjalaismurteiden th descendants of the Korela ethnos that emerged prior to the 13 century refleksiiviverbeistä1 on the Karelian Isthmus. The Russian–Swedish wars of the 16th and 17th centuries were the reason for their migration. Currently, there are 7 000 Tver Karelians. Their language remains alive and continues to Abstrakti Artikkeli käsittelee rajakarjalaismurteiden reflek­ be spoken in the villages of the Tver region. siiviverbejä. Rajakarjalaismurteet ovat karjalan kielen suomen­ The Tver dialect of Karelian is classified as a dialect of Kare­ vaikutteinen alamurteisto. Artikkelin aineistona on n. 120 lian Proper, and is especially similar to the South Karelian dialects in nauhatunnin laajuinen tekstimuotoinen Raja-Karjalan kor­ pus. Käsiteltäviä refleksiiviverbien tyyppejä ovat refleksiivi­ the . The Tver dialect of Karelian has three sub­ johdokset sekä refleksiivitaivutus. Molempia tavataan niin dialects: the Tolmachi and Vesyegonsk subdialects, and the Dyorzha karjalassa kuin suomessakin (refleksiivitaivutusta tosin vain subdialect, which is fundamentally different from the other two in suomen itämurteissa).­ Rajakarjalaismurteiden­ refleksiivitaivutus­ terms of phonology. The largest group of Tver Karelians speaks the osoittautuu samankaltaiseksi kuin karjalan saarekemurteissa Tolmachi dialect. (. tytär­karjalaisissa murteissa) Sisä-Venäjällä, joskin se on Tver Karelians are mostly bilingual. They are fluent speakers of paradigmaltaan vajaampi ja esiintymiseltään harvempi. Raja­ karjalaismurteiden­ refleksiivijohdoksissa­ puolestaan tavataan joi­ Russian and this is reflected in their language. The influence of Rus­ takin karjalalle vanhastaan tuntemattomia morfologisia tyyppejä, sian is strongest in the phonetic system: 1) vowels: replacement of the jotka ovat selvää suomen kielen vaikutusta. Artikkelin aineistossa vowels u, y with non-syllabic w in the original long vowels uu, yy and näkyy lisäksi rajakarjalais­ murteille­ ominainen karjalan kahden ending in u, y; phonematic differences of opening diph­ päämurteen, varsinaiskarjalan ja livvinkarjalan, sekoittuneisuus. thongs in Tver Karelian; secondary long vowels in front of r; apocope Avainsanat rajakarjalaismurteet, refleksiiviverbit, kielikontaktit and syncope mostly in the Dyorzha subdialect; 2) consonants: conso­ nants c and f in Russian loans; distribution of sibilants; reduction of the geminate consonants kk, tt, pp, čč when occurring before l, r, n; widespread palatalisation; changes in the system. Analysis of the phonetic system of the Tver Karelian dialects shows that Russian has influenced it considerably. The phonetics and morphology of the Tolmachi and Vesyegonsk subdialects are quite similar. The already extinct Dyorzha subdialect was influenced by Russian to a significant extent, which can be explained by its remote­ 1. Artikkeli on valmistunut Suomen Akatemian rahoittamissa Itä-Suomen yli­ ness relative to the other two subdialects. opiston tutkimushankkeissa FINKA (137479, Suomen ja karjalan rajalla: näkö- kulmia lähisukukieliin ja niiden murteisiin; ks. Palander, Zaikov & Uusitupa 2013: 365–366) ja KATVE (314848, Muutto ja kielellinen eriytyminen: karjala Tverissä ja Suomessa).

Multilingual­ Finnic. Language 248 contact and change. 249–300. Uralica Helsingiensia 14. Helsinki 2019. ‹https://doi.org/10.33341/uh.85038› VESA KOIVISTO RAJAKARJALAISMURTEIDEN REFLEKSIIVIVERBEISTÄ

1. Aluksi Karjalankieliset olivat uskonnoltaan pääosin ortodokseja ja suomen­ kieliset luterilaisia. Käsittelen tässä artikkelissa karjalan kielen ns. rajakarjalaisten mur­ Raja-Karjalan alue kuului alkuaan Novgorodille ja sittemmin teiden refleksiiviverbejä. Rajakarjalaismurteet ovat Suomessa puhut­ Venäjälle. Vuonna 1617 se liitettiin Stolbovan rauhanteon jälkeen tuja karjalan murteita, joissa on runsaasti suomen kielen vaikutusta Ruotsiin, mutta siirtyi 1721 jälleen takaisin Venäjän hallintaan osana ja siitä johtuvaa variaatiota (rajakarjalaismurteista ks. esim. Turunen ns. Vanhaa Suomea.3 Sittemmin Raja-Karjala kuului vuodesta 1812 1982; Koivisto 2018). Artikkelini aineistona toimivat Suomen kielen Suomen suuriruhtinaskuntaan ja toiseen maailmansotaan asti itse­ nauhoitearkiston (SKNA) pääasiassa 1960-luvulta peräisin olevat raja­ näiseen Suomeen. Sotien jälkeen alue luovutettiin Neuvostoliitolle karjalaismurteiden äänitteet, joista on Itä-Suomen yliopistossa koostettu ja sen väestö evakuoitiin Suomen puolelle, enimmäkseen Pohjois- n. 120 nauhatuntia käsittävä tekstikorpus (Raja-Karjalan korpus). Karjalaan, Pohjois-Savoon ja Oulun seudulle. Väestönsiirron jälkeen Tarkastelun kohteena olevat refleksiiviset verbit ilmaisevat perus­ raja­karjalaismurteet­ eivät enää ole Suomessa alueellinen kielimuoto, käytössään verbin toiminnan kohdistumista subjektitarkoitteeseen it­ joskin näiden murteiden puhujia – evakkoja tai heidän jälkeläisiään – seensä. Itämerensuomalaisissa kielissä refleksiiviverbeihin kuuluvat arvioidaan maassamme edelleen olevan useita tuhansia (ks. Jeskanen mm. refleksiiviset johdokset (kuten suomen peseytyä tai kääntyä) ja 2005: 278; Hämynen 2012: 266). refleksiivitaivutus (kuten livvinkarjalan heitämmös ’heittäydyn’ tai Karjalan kielen murteita ovat vienankarjala ja eteläkarjala, jot­ pezetöh ’peseydyt’; itämerensuomen refleksiiviverbeistä ks. tarkem­ ka yhdessä muodostavat varsinaiskarjalan murteen, sekä livvinkarjala min Koivisto 1995).2 Karjalassa esiintyy paljolti samoja refleksiivisiä eli aunukselaismurteet. Rajakarjalaiset murteet edustavat enimmältä johtimia ja johdostyyppejä kuin suomessakin. Refleksiivitaivutus­ osin eteläkarjalaa: tätä murretta puhuttiin Ilomantsissa, Korpiselässä, ta puolestaan tavataan vain itäisissä itämerensuomalaisissa kielissä, Suistamolla, Suojärvellä ja jonkin verran Impilahdellakin. Livvin­ kuten karjalassa ja suomen itämurteissa. Esitän tässä artikkelissa ai­ karjalan puhuma-aluetta Raja-Karjalassa olivat Salmin kunta sekä neistopohjaisen kuvauksen rajakarjalaismurteiden refleksiiviverbeistä osat Impilahtea, Suojärveä ja Suistamoa. Nimitys rajakarjalaismurteet tarkaten niiden yhteyksiä etenkin kontaktikieleen suomeen. on alueel­linen ja kattaa molemmat Raja-Karjalassa puhutut karjalan murteet. Kartasta 1 näkyvät karjalan kielen päämurteet sekä Raja- Karjalan alue vaakaviivoitettuna, kartasta 2 puolestaan Raja-Karjalan 2. Rajakarjalaismurteista pitäjät sekä tummennettuna Raja-Karjalan karjalankielinen alue. Rajakarjalaismurteiden kehitysvaiheille ovat olleet tyypillisiä Karjala on suomen kielen itäpuolinen naapurikieli ja sen lähin suku­ monenaikaiset ja monenlaiset kontaktit toisiin kieliin ja murteisiin. kieli. Karjalaa puhutaan pääasiassa Venäjällä, mutta myös Suomessa Kehityttyään muinaiskarjalaisen vaiheen jälkeen eroon toisistaan kar­ oli ennen toista maailmansotaa karjalankielinen alue, nimittäin Laa­ jalan ja suomen kielet ovat olleet Raja-Karjalassa kontakteissa pit­ tokan pohjoispuolinen Raja-Karjala, joka käsitti Salmin kihlakunnan kään – vuosisatoja. Raja-Karjala rajautui lännessä suurimmalta osin kunnat Korpiselän, Suistamon, Impilahden, Soanlahden, Suojärven suomen savolaismurteisiin, joita on 1600-luvulta lähtien puhuttu myös ja Salmin sekä nykyistä laajemman Ilomantsin kunnan itäosat. Raja- Raja-Karjalan puolella, mm. Korpiselän ja Suistamon länsiosissa. Ete­ Karjalassa puhuttiin karjalaa ja suomea, ja vuonna 1939 siellä oli kar­ lämpänä olivat karjalan kielinaapurina suomen kaakkoismurteet. (Ks. jalan kielen puhujia n. 30 000–40 000 (ks. Hämynen 2012: 251, 257). H. Leskinen 1998: 358–359, 362, 376; Uusitupa 2011: 23–24.)

2. Käytän tässä artikkelissa karjalan aunukselaismurteista nimitystä livvinkarja­ 3. Vanha Suomi oli se alue Kaakkois-Suomessa, jonka Ruotsi menetti Venäjälle la. – Itämerensuomalaisista (eli ims.) kielistä käytetään myös yhteisnimitystä itä­ vuosien 1721 ja 1743 rauhanteoissa. Vanhaan Suomeen kuuluivat Suomen tuolloisista meren­suomi (joka korostaa näiden kielten historiallista yhteenkuuluvuutta). kaupungeista Viipuri, Käkisalmi, Sortavala, Lappeenranta, Savonlinna ja Hamina.

250 251 VESA KOIVISTO RAJAKARJALAISMURTEIDEN REFLEKSIIVIVERBEISTÄ

Mittakaava 1:3 000 000 Paatene

Vienan- Porajärvi meri Mänty- VIENAN- selkä KARJALA Ilomantsi

Kemi

Suojärvi Munjärvi Korpi- ETELÄ- selkä KARJALA Soan- lahti Suistamo Säämäjärvi

Tule- Joensuu ma- Impilahti järvi Vieljärvi

Salmi Petroskoi Kotkat- Sortavala LIVVIN- Ääninen järvi LYYDI Vitele KARJALA

Käkisalmi Riipuš- Aunus Laatokka S y v ä r i kala Laatokka Nek- Tver Copyright Itä-Suomen yliopisto 2013 kula

Copyright Itä-Suomen yliopisto 2013 Kartta 2. Raja-Karjalan pitäjät. Raja-­Karjalan karjalankielinen alue on merkitty tummen- Kartta 1. Karjalan kielen päämurteet. Raja-Karjalan alue on merkitty vaaka­viivoi­tuksella. nuksella. 252 253 VESA KOIVISTO RAJAKARJALAISMURTEIDEN REFLEKSIIVIVERBEISTÄ

Raja-Karjalassa karjalan ja suomen kielen puhuma-alueet ovat kieleen syntyi ortodoksisen kirkon kautta sekä 1800-luvun loppu­ 1600-luvulta lähtien vähittäin limittyneet, kun suomenkielistä savo­ puolella Raja-­Karjalassa toimineissa venäjänkielisissä kouluissa. laista väestöä on muuttanut lännestä päin Pohjois- ja Raja-Karjalaan Kielisuhteet ovat vaihdelleet Raja-Karjalassa alueellisesti siten, (Kuujo 1963: 59–68).4 Kontaktit suomen kieleen näkyvät rajakarjalais­ että sen läntisissä pitäjissä Impilahdella, Suistamolla, Korpiselässä murteiden monissa kielenpiirteissä variaationa karjalaisen ja suoma­ ja Ilomantsissa suomen kielen vaikutus on ollut vahvempaa ja pitkä­ laisen edustuksen kesken. Raja-Karjalassa puhuttua karjalan kieltä aikaisempaa ja idiolektit suomalaisvärisempiä. Idempänä Suojärvellä voikin – samoin kuin sen eri pitäjänmurteita ja puhujien yksilöllisiä ja Salmissa taas karjalan kielen (ja samalla ortodoksisen uskonnon) kielimuotoja, idiolekteja – luonnehtia seka- tai siirtymämurteeksi asema oli vahvempi, ja suomenkielistä väestöä oli 1900-luvulle asti (esim. E. Leskinen 1934: VI; Nirvi 1961: 129; Ruoppila 1956: 12; varsin vähän. Raja-Karjalassa ei käytännössä solmittu seka-avio­ Uusitupa, Koivisto & Palander 2017). liittoja kahden uskontokunnan välillä, joten suomi (savolaismurre) ja Tuoreimman suomalaiskerrostuman rajakarjalaismurteissa muo­ karjala (rajakarjalaismurre) eivät yleensä kohdanneet ja limittyneet dostaa sotien jälkeen evakossa saatu savolaisvaikutus, jota voi olla samassa perheessä. Samoissa kylissä kylläkin asui niin ortodokseja vaikea erottaa murteeseen jo varhemmin tulleista savolaispiirteistä. kuin luterilaisiakin, ts. sekä karjalan- että suomenkielisiä. Suomen kirja- tai yleiskieli puolestaan on vaikuttanut rajakarjalais­ Heikki Ojansuu (1910: 10–11) on kuvannut suomen ja karjalan murteisiin hallinnon ja koulujen kielenä etenkin 1900-luvulla. Raja- suhdetta lingvistisenä jatkumona, jossa kielimuodot vaihtuvat vähit­ Karjalan yhteiskunnallinen ja kielellinen suomalaistuminen vauhdittui täin toisiin. Tällä jatkumolla rajakarjalaismurteet ovat muodostaneet Suomen itsenäistymisen jälkeen, kun raja itään päin sulkeutui ja maan sillan karjalan ja suomen kielten välillä. valtiollinen yhtenäistyminen eteni. Kaikki nämä tekijät ovat osaltaan Paitsi kielirajan välittäjäkieliä, rajakarjalaismurteet ovat olleet voimistaneet rajakarjalaismurteiden suomalaistumista (ks. Turunen myös kahden karjalan murteen, varsinaiskarjalan (sen etelämurteen) 1965: 27; 1982: 84–85). ja livvinkarjalan, välisiä siirtymämurteita. Eino Leskisen (1934: VI) Suomen kielen ohella toinen keskeinen kontaktikieli karjalan mukaan ”Suistamon ja Suojärven murteet ovat selvimmin karja­ koko puhuma-alueella on ollut venäjä, ja juuri venäjän kielen vai­ lan kielen aunukselaismurteiden ja varsinaisen karjalan kielen väli­ kutus on seikka, jonka suhteen karjalan kieli leimallisesti eroaa suo­ asteita”. Näiden kuntien itäisimmät tai lounaisimmat osat onkin luettu­ mesta. Muinaiskarjalan ja venäjän kontaktit ovat alkaneet jo näiden livvinkarjalan puhuma-alueeseen. Toisaalta rajakarjalaismurteet eroa­ kielten esimuotojen kesken n. 700-luvulta jKr. lähtien, ja myöhem­ vat suomenvaikutteisuutensa puolesta Venäjän puolella puhutusta liv­ pinä aikoina karjalan alueellisena kontaktikielenä ovat olleet luoteis­ vin- tai eteläkarjalasta (E. Leskinen 1934: V).5 Jos jokin rajakarjalais­­ venäläiset murteet (ks. Sarhimaa 1999: 13–14). Toisaalta rajakarjalais­­ murteiden piirre puuttuu muualta karjalasta, syynä tähän voi olla murteissa, joiden puhuma-alue kuului Ruotsiin n. 100 vuoden ajan suomen kielen vaikutus. Vastaavasti Venäjällä puhuttuun karjalaan on 1600–1700-luvuilla sekä 1800–1900-luvuilla autonomiseen Suomen venäjä vaikuttanut enemmän kuin Suomessa puhuttuun. suuriruhtinaskuntaan ja sittemmin itsenäiseen Suomeen, venäjän vai­ Rajakarjalaismurteet ovat siis olleet siirtymävyöhykettä kah­ kutus on ollut vähäisempää kuin muualla karjalassa. Tosin Suomen den kielen, karjalan ja suomen, välillä ja myös kahden murteen, itsenäistymiseen asti Raja-Karjalan alueelta voitiin pitää yllä tiiviitä livvinkarjalan­ ja eteläkarjalan, välillä. Etelä- ja livvinkarjalan välis­ kontakteja rajan yli Venäjän Karjalaan. Lisäksi kosketuksia venäjän tä rajaa on pidetty vahvimpana karjalan kielen sisäisenä murrerajana

5. Rajakarjalaismurteiden eroja Venäjän puolella puhuttuun karjalaan ei pystytä 4. Toisaalta jotkut tutkijat katsovat, että Impilahden, Suistamon ja Suojärven selvittämään Bubrihin & al. murrekartastosta (1997), sillä teos kuvaa Neuvostoliiton murteiden sekamurteisuus on jo vanhempaa perua, livvinkarjalan synnyn aikaista alueella 1930-luvulla puhuttua karjalaa eivätkä rajakarjalaismurteet ole mukana (Nirvi 1961; ks. Turunen 1982: 88). siinä.

254 255 VESA KOIVISTO RAJAKARJALAISMURTEIDEN REFLEKSIIVIVERBEISTÄ

(ks. Sarhimaa 1999: 22), mutta toisaalta Raja-Karjalassa tämä murre­ leksikaalista, semanttista ja kvantitatiivista tarkastelua. Karjalan ja raja ei kuitenkaan ollut jyrkkä. Siirtymä oli asteittainen esimerkiksi suomen kielten suuren samankaltaisuuden vuoksi karjalan käsittelys­ Suojärvellä, jonka eteläkarjala on ollut eriasteista livviläisväristä sä sovelletaan paljolti suomen kielen kuvauksen totunnaisia käsitteitä seka­murretta (ks. Genetz 1870: 206; Jeskanen 2011: 354, 357). ja kategorioita. Kaikissa rajakarjalaismurteissa on erotettavissa jonkinasteista Käsittelyn taustalla on väitöskirjani (Koivisto 1995), joka tarjoaa livvinkarjalan vaikutusta. Raja-Karjalan eteläkarjalassa voi havaita vertailutietoa refleksiiviverbikategorian edustuksesta niin karjalan eri livvin­karjalaisuuksia kaukana livvinkarjalan varsinaisesta puhuma-­ murteissa kuin muissakin itäisissä ims. kielissä. Tuossa tutkimuksessa alueestakin – ei tosin yksinomaisina, vaan vaihdellen varsinais­ ei karjalan kielestä kuitenkaan ole erikseen tarkasteltu rajakarjalaisia­ karjalan edustuksen rinnalla ja usein vain aivan satunnaisinakin murteita, vaan ne ovat siinä osana eteläkarjalan ja livvinkarjalan piirteinä. Tällainen on esimerkiksi livvinkarjalan h-loppuinen NUT- aineistoa. partisiippi, jota tavataan Raja-Karjalan länsiosissa asti Ilomant­ Keskeisimmät tässä artikkelissa tarkasteltavat refleksiiviset joh­ sia myöten (ks. Moshnikov 2014), esim. uskonuh, nähnyh, kuulluh dostyypit ovat varsinaiskarjalalle tyypilliset UtU- ~ UdU-johdokset (Korpiselkä). Liv­vinkarjalan vaikutus näkyy myös tässä artikkelissa ja livvinkarjalalle ominaiset vU-johdokset (ks. Markianova 1985: 21– käsiteltävässä rajakarjalaismurteiden refleksiiviverbistössä monin ta­ 23, 99–102; Koivisto 1995: 129–130, 133). voin. Rajakarjalaismurteiden­ livvinkarjalaisia piirteitä, niiden piirre­ Refleksiivisten verbijohdosten kanssa samaa ilmaisukenttää itäi­ kohtaista levinneisyyttä tai leviämisen yksityiskohtia ei kuitenkaan sissä ims. kielissä jakaa refleksiivitaivutus. Karjalan murteista reflek­ ole tarkemmin selvitelty. siivitaivutusta tavataan etenkin livvinkarjalassa, jossa se on käytän­ Raja-Karjala ei siis ole ollut kielellisesti homogeeninen, vaan nössä refleksiivijohdoksia yleisempi refleksiiviverbien laji (ks. Genetz kielten ja murteiden risteymäaluetta – niin historiansa eri vaiheissa 1884: 179), ja myös eteläkarjalassa. kuin siinä rajakarjalaisessa kielimuodossakin, jota tämän artikke­ Tarkastelen tässä artikkelissa refleksiiviverbejä rakennetyypeit­ lin aineisto synkronisesti edustaa. Kielikontaktien osapuolina Raja- täin: ensin luvussa 4 refleksiivijohdoksia (jotka ovat U-vartaloisia Karjalassa ovat olleet karjalan kieli sekä suomi ja venäjä jo edeltäjä­ verbejä) ja sitten luvussa 5 refleksiivitaivutusta. Refleksiivisten kielineen ja lisäksi karjalan kielen kaksi päämurretta varsinaiskarjala verbi-­ilmausten muita rakenteellisia tyyppejä itämerensuomessa ja livvinkarjala keskenään. Moniaineksisesta taustastaan johtuen raja­ ovat refleksiivipronominin objektinaan sisältävät ns. itse-rakenteet karjalaismurteet tarjoavatkin poikkeuksellisen kiintoisan selvittely­ (esim. heittää itsensä pitkäkseen) sekä implisiittiset refleksiivit (esim. kentän kontaktilingvistiikalle ja variaationtutkimukselle. heittää pitkäkseen) (joista ks. Koivisto 1995: 148–232). Nämä kaksi tyyppiä ovat kuitenkin karjalassa hyvin harvinaisia, niin myös rajakarjalaisaineistossani. 3. Tutkimuskohteesta Vertailukohtana karjalalle tässä artikkelissa toimivat suomen kie­ len (ja etenkin sen itämurteiden) refleksiiviverbit. Suomen ja karjalan 3.1. Tutkimustehtävä ja -menetelmä yhtenäinen edustus, esim. saman johdostyypin tai saman lekseemin esiintyminen kummassakin, kertoo joko ilmiön yhteisistä juurista tai Tässä artikkelissa tavoitteenani on luoda aineiston pohjalta kokonais­ niiden myöhemmästä keskinäisestä vaikutuksesta. Suomen kielen vai­ kuva rajakarjalaismurteiden refleksiiviverbistöstä sekä selvitellä sen kutus on tuottanut rajakarjalaismurteisiin sellaisia erityispiirteitä, jois­ erityispiirteitä murrealueen eri osissa ja etenkin suomen kielen vaiku­ sa ne eroavat muista karjalan murteista, ja tällaisia piirteitä pyrin tässä tusta siihen. Tutkimus on aineistolähtöinen ja sisältää morfologista, artikkelissakin tuomaan esiin.

256 257 VESA KOIVISTO RAJAKARJALAISMURTEIDEN REFLEKSIIVIVERBEISTÄ

3.2. Aineistosta Käytän tässä artikkelissa karjalan verbilekseemeistä KKS:n hakusanojen mukaista fonologista asua (joka noudattaa vienalais­ Tämän artikkelin pääaineistona toimii n. 120 nauhatunnin laajuinen kar­ murteita). Tämä esityskäytäntö koskee aineiston kaikkia lekseemejä, jalan kielen rajakarjalaismurteiden tekstikorpus (Raja-Karjalan korpus), myös sellaisia jotka eivät esiinny KKS:n hakusanana, samoin kuin joka on peräisin Suomen kielen nauhoitearkistosta (SKNA) ja litteroitu muita kielenaineksia, mm. johtimia (puhun esim. UtU- enkä UdU- Itä-Suomen yliopiston FINKA-hankkeessa (korpuksesta myös Palander johdoksista, vaikka rajakarjalaismurteissa esiintyy näitä kumpaakin et al. 2019 tässä julkaisussa). Aineisto on nauhoitettu pääosin 1960-lu­ johdinvarianttia). Käytän johdoslekseemin esitysmuotona yleensä sen vulla, jolloin puhujat olivat jo varsin iäkkäitä ja asuneet evakkoseuduil­ vokaalivartaloa (esim. oleutu-) enkä A-infinitiiviä (oleutuo). Aines­ laan parikymmentä vuotta. Litteroitua aineistoa on pitäjittäin seuraavat lähteistä poimitut käyttöesimerkit esitetään alkuperäisessä, litteroi­ tuntimäärät: Impilahti n. 19 tuntia, Korpiselkä n. 36 tuntia, Salmi n. 9 dussa asussaan. tuntia, Suistamo n. 17 tuntia, Suojärvi n. 10 tuntia, Ilomantsi n. 26 tun­ Sanakirjoissa karjalan u-, y-vartaloisten verbien hakusanat ovat tia. Käytän tästä korpuksesta nimitystä ”rajakarjalaisaineisto(ni)”. Näin uo-, yö-loppuisia, refleksiivitaivutuksen puolestaan yleensä kseh- laaja aineistokokonaisuus antaa uusia mahdollisuuksia rajakarjalais­­ tai ttšie-loppuisia (esim. heitteäkseh ~ heitteättšie), mutta osassa murteiden tutkimukselle, joka on tähän asti jäänyt vähäiseksi mm. juuri raja­karjalaismurteita myös VijA-loppuisia (esim. antaija, jaksaija, käyttökelpoisten aineistojen puutteen vuoksi.6 valaija; ks. myös luku 5.1. alaviite 24). Niinpä esimerkiksi verbistä Koska tässä tutkittava verbikategoria, refleksiiviverbit, ei ole jaksoa (’riisua’) johdettuja refleksiiviverbejä löytyy Karjalan kielen teksteissä kovin taajaan esiintyvä muotoryhmä, tarvitaan tutkimus­ sana­kirjasta (KKS) seuraavin hakusanoin: jaksoakseh ~ jaksoattšie ~ aineistoksi riittävän laaja tekstiaines. Olenkin poiminut kaikki reflek­ jaksaija ~ jaksautuo ~ jaksavuo (kaikki: ’riisuutua’). siivisten verbien esiintymät korpuksen viiden rajakarjalaispitäjän – muiden paitsi Ilomantsin – aineistosta (yhteensä lähes 100 tunnista). Ilomantsi ei ole mukana oletetusti suuren suomalaisuusasteensa vuok­ 4. Refleksiivijohdokset si (sen murteesta mm. puuttuu refleksiivitaivutus), minkä suhteen se poikkeaa muista rajakarjalaismurteista.7 Ilomantsin tietoja olen kui­ Refleksiivisellä johdoksella tarkoitetaan tässä sellaista johdosverbiä, tenkin tarvittaessa selvitellyt aineistosta pistokokein ja täsmähauin. jonka johtimella on semanttisesti refleksiivistä käyttöä mutta yleensä Toisen keskeisen aineslähteen muodostaa Karjalan kielen sanakirja muitakin semanttisia tehtäviä (ks. Koivisto 1995: 122). Näin määritel­ (KKS), josta olen kerännyt pitäjäkohtaista lisäaineistoa etenkin reflek­ tynä refleksiivisyys on verbikytköinen ja ims. kielissä johto-opillinen, siivitaivutuksesta. Lisäksi tukeudun muihin sellaisiin tutkimuksiin tai sananmuodostuksen ilmiö. Refleksiivisen merkityssisällön ilmaisi­ aineslähteisiin, jotka tarjoavat täydennystä ainespohjaan (mm. Genetz mena verbissä on erityinen morfeemi, refleksiivitunnus, joka voi olla 1870, 1884; E. Leskinen 1934; Punttila 1992, 1998; Ahtia 1938). johdin (esim. -UtU- tai -U-) tai refleksiivitaivutuksessa myös taivutus­ morfeemi. Refleksiivitunnuksen lisääminen verbikantaan tuottaa tyy­ 6. Litteroitu korpus käsittää otoksen SKNA:n rajakarjalaismurteiden nauhoitteista; pillisesti uuden verbilekseemin: se muuttaa transitiivisen kantaverbin kaikkiaan SKNA:ssa on näistä murteista äänitteitä yli 400 tuntia. Korpuksessa on mukana valtaosa Impilahden ja Korpiselän nauhoitteista, kun taas Salmin, intransitiiviseksi, ja samalla kantaverbin ilmaisema toiminta muuttuu Suistamon, Suojärven ja Ilomantsin litteroitu aineisto kattaa pienemmän osan näiden subjektitarkoitteeseen itseensä kohdistuvaksi (heittää + objekti > heit- murteiden äänitteistä SKNA:ssa (Soanlahdelta ei ole aineistoa). Aineiston pitäjittäiset täytyä ’heittää itsensä’). koko- ja kattavuuserot saattavat rajoittaa mahdollisuutta pitäjänmurteiden väliseen Toisaalta refleksiivitunnuksen sisältävällä verbirakenteella voi kvantitatiiviseen vertailuun. 7. Ilomantsin eteläkarjala on osoittautunut muita rajakarjalaismurteita suoma­ olla muunkinlaista käyttöä: toiseksi semanttiseksi päälajiksi reflek­ laisvärisemmäksi myös mm. venäläislainojensa valossa (Tavi 2015). siivisyyden rinnalle voidaan erottaa automatiivisuus, jossa lauseen

258 259 VESA KOIVISTO RAJAKARJALAISMURTEIDEN REFLEKSIIVIVERBEISTÄ subjektitarkoite ei ole samalla tavoin agentiivinen kuin semanttises­ kaikissa ims. kielissä (ks. Koivisto 1995: 123–124, 126).8 Kompleksi­ ti refleksiivisissä käyttöyhteyksissä, vaan esim. eloton (ks. Koivisto set refleksiivijohtimet taas ovat nuorempia, ja niitä tavataan harvem­ 1991: 48–93; ISK: § 334, 336). Samallakin verbillä voi olla sekä reflek­ missa itämerensuomalaisissa kielissä: esimerkiksi vepsässä ei esiinny siivistä että automatiivista käyttöä. Merkitystehtävien yleisyyssuhteet lainkaan suomen ja karjalan murteille tyypillisiä kompleksisia reflek­ vaihtelevat johtimittain: esim. johdinta -UtU- (esim. kääntä+yty-) siivijohtimia (ks. mts. 147). Lisäksi itäisille ims. kielille ominaiseen käytetään semanttisesti refleksiivisenä suhteessa enemmän kuin joh­ refleksiivitaivutukseen kuuluu taivutusainesten ohella myös eräitä dinta -U- (käänt+y-). Lisäksi muita karjalan refleksiiviverbien käyttö­ refleksiivisiä johdinaineksia (esim. kantasuomen *-δe- tai karjalan tapoja ovat esim. passiivi sekä inkoatiivisuus (eli toiminnan alkamisen -čče-), joita käsitellään refleksiivitaivutuksen yhteydessä luvussa 5. ilmaiseminen; ks. luku 4.1.4.). Refleksiivisyyden määritelmällisestä verbikytköisyydestä huo­ Itämerensuomalaisten kielten refleksiivijohdokset voidaan jakaa limatta U-aineksiset johtimet voivat liittyä myös nominikantoihin johtimiensa rakenteen perusteella kahteen päätyyppiin: yksinkertaisen (kuten monet muutkin ims. verbinjohtimet). Kantasanan sanaluokasta johtimen -U- sisältäviin, kuten suomen kuul+u-, siirt+y-, supist+u-, riippumatta U-verbijohdos on muutosta ilmaiseva: subjektitarkoittee­ sekä ns. kompleksisiin refleksiivijohdoksiin, joita edustavat suomessa seen kohdistuu muutos (tai sen vaikutus), jossa on tavalla tai toisella esim. johtimet -UtU- ja -VntU-, karjalassa myös -vU-, esim. pese+yty-, osallisena johdoksen kantasanan – verbin tai nominin – tarkoittama pila+antu-, pila+vu- (Itkonen 1969: 71–72; Koivisto 1995: 128–136). asia (ks. Räisänen 1988: 31; Koivisto 1995: 123). Kompleksisten johdostyyppien johtimet ovat U-loppuisia ja raken­ teeltaan yhdysjohtimia eli alkuaan usean morfologisen aineksen si­ 4.1. Kompleksiset refleksiivijohdokset sältäviä (yhdysjohtimista ks. Koivisto 2013: 274–275). Johdin -U- ja komplek­siset refleksiivijohtimet kytkeytyvät muodoltaan toisiinsa si­ Kompleksisia refleksiivijohdoksia esiintyy rajakarjalaisaineistossani käli, että -U- on tavallisesti mukana jälkimmäisten loppuaineksena. viittä eri morfologista tyyppiä (ja sporadisesti vielä kolmea muuta). Suomen murteissa kompleksisia refleksiivijohdoksia tunnetaan Päätyyppeinä ovat kaksi johdinta: -UtU- (tyyppi seka+utu-) ja -vU- runsaat 10 eri morfologista tyyppiä (ks. Koivisto 1995: 138), mutta kar­ (tyyppi seka+vu-). Aineistossa harvinaisempia ovat tyypit seka+u-, jalassa ei läheskään näin monta. Kompleksisten johdostyyppien verbit seka+atu- ja seka+antu-, joiden johtimet ovat vastaavasti -(V)U-, ovat vartaloltaan tavallisesti kolmi-, joskus nelitavuisiakin (pese+yty-, -VtU- ja -VntU-. (Lisäksi aineistossa esiintyvät satunnaisesti reflek­ pila+antu-, sijoitta+utu-), kun taas U-johdokset ovat tyypillisimmin siivijohtimet -hUntU-, -hU- ja -itU-.) Kaikki tässä mainitut rajakarja­ kaksitavuisia (kuul+u-, siirt+y-). Tosin U-johdoksissa on myös varta­ laismurteiden refleksiiviset johtimet ja johdostyypit tunnetaan myös loltaan kolmitavuisia, esim. asett+u- (< asetta-), kellist+y- (< kellistä-), suomen murteissa, ja kaikkien paitsi yhden niistä (-vU-) levikkialueet jotka sisältävät U-johtimen edellä toisen verbinjohtimen (näissä esi­ suomessa ja karjalassa rajautuvat suoraan toisiinsa. Karjalan kielen merkeissä -ttA- ja -stA-). (Suomen U-aineksisista verbijohdoksista ks. sanakirja tuntee kuitenkin näistä johdostyypeistä vain ensin mainitut myös Hakulinen 1979: 269–271, 292–293; Räisänen 1988; ISK: § 333, kaksi päätyyppiä (-UtU- ja -vU-). 339–349.) Lisäksi kompleksisiin refleksiivijohdoksiin voidaan lukea Tarkastelen seuraavaksi rajakarjalaisaineistostani ensin komplek­ (V)CehtA-johtiminen verbityyppi (esim. ajelehta-, keäntelehtä-), jota sisia refleksiivijohdoksia ja sitten U-johdoksia, jälkimmäisten mu­ tavataan karjalassa sekä suomen murteissa (ks. luku 4.1.5.). kana myös vartaloltaan kolmitavuisia, johdinyhtymän sisältäviä Refleksiivijohdokset sisältävät siis yleensä vartalonsa lopussa U-johdoksia (esim. asett+u-, opast+u-, imelt+y-, kumart+u-) sekä vokaalin u, y, joka sellaisenaankin voi olla johdin, ja niiden yhteis­ tU- ja pU-johdoksia (esim. jeä+ty-, nuor+tu-, juo+pu-). Vertailen nimitykseksi sopii ”U-vartaloiset” tai ”U-aineksiset johdokset”. 8. Ekspressiivisiin kantoihin liittyessään johdin -U- taas on merkitykseltään Verbin­johdin -U- on vanhaa suomalais-ugrilaista perua, ja se tunnetaan lähinnä kontinuatiivinen, esim. kirsku-, välkky- (ks. Kulonen 2010).

260 261 VESA KOIVISTO RAJAKARJALAISMURTEIDEN REFLEKSIIVIVERBEISTÄ rajakarjalaisaineiston verbejä ja johdostyyppejä niiden esiintymiseen Rajakarjalaisaineistooni sisältyy 108 UtU-johdoslekseemiä.9 suomessa ja toisaalta muualla karjalan kielessä. Vertailussa muuhun Johtimen asuna on astevaihtelun vahvassa asteessa -UdU- tai -UtU- ja karjalaan hyödynnän Karjalan kielen sanakirjaa. Suomen osalta ver­ heikossa asteessa -UvU-. Esim. tailu tapahtuu yleiskieleen ja osin murteisiinkin. (1) – – vettä sih vain otettii pezeyvyttii dä, kauhazel piäh lykittii pe- huuhtouvuttii dai hyvä. (Korpiselkä, SKNA 3039:1a) 4.1.1. UtU-johdokset (2) – – ńygie tukkiloi tahi ropsiloi ja, sillä keinol se, avauduu ruhka. – – no siid ku ruuhka purgaudu. (Suistamo, SKNA 3044:1b) Johtimen -UtU- (~ -UdU-) sisältäviä refleksiivijohdoksia tavataan (3) tuola sanoo kolkuttii meijjän ovvii ka minä vaan kävin sanomaa karjalassa ja suomessa. Karjalassa UtU-johdostyyppi on ominainen jotta, raivauvu pois pihas ei meil‿oo mitää tiäl tekemistä. (Impi­ vienalaismurteille ja eteläkarjalalle; livvinkarjalassa sitä vastaa vU- lahti, SKNA 6820:1a) johtiminen tyyppi. Lisäksi UtU-johdoksia on tavattu lyydin poh­ joismurteissa eteläkarjalan vaikutuksesta. (Koivisto 1993: 199–200; Aineiston UtU-verbeistä verbikantaisia on 81 (kuten esimerkkien 1–3 1995: 130, 133, 147.) peseyty-, huuhtoutu-, avautu-, purkautu-, raivautu-); niistä on varta­ Suomen murteissa UtU-johdostyyppi on levikiltään varsin laaja- loltaan nelitavuisia seitsemän (mm. asettautu-, kellistäyty-). Nomini­ alainen: sitä on tavattu laajalti itämurteissa ja Pohjois-Suomen murteis­ kantaisia UtU-verbejä on puolestaan 23 erilaista (esim. hätäyty- ’hä­ sa sekä jonkin verran länsimurteissakin (ks. tarkemmin Koivisto 1995: tääntyä’, kipeyty-, vierautu- ’vieraantua, erota’). Muutama verbi 129–130). Lähtöjään UtU-tyyppi on itäinen, ja suomen itämurteissa­ voidaan katsoa niin verbi- kuin nominikantaiseksikin (mm. häpeyty- sen kaikkein vahvinta esiintymisaluetta on Pohjois-Karjala, missä se ’hävetä’, kokoutu-, likautu-, pilautu-, vastautu-; kantasanan sanaluokan on esiintynyt jokseenkin yksinomaisena kompleksisena refleksiivi­ kaksitulkintaisuudesta ks. Koivisto 1991: 17–18; 2013: 302–303).10 johdostyyppinä, kun taas lännempänä ja pohjoisempana UtU-tyyppi Rajakarjalaisaineiston UtU-verbien kantasanojen sanaluokka­ on limittynyt toisten refleksiivisten johdostyyppien kanssa. Levik­ jakauma on samansuuntainen kuin suomen yleiskielessä, jonka UtU-joh­ kinsä perusteella UtU-johdin ja -johdostyyppi palautuvat karjalan ja doksista on verbikantaisia n. 300, nominikantaisia n. 150 (joista substan­ suomen itämurteiden yhteiseen kantakieleen, muinaiskarjalaan, mut­ tiivikantaisia­ runsaat 100 ja adjektiivikantaisia n. 50) ja kantanaan muun ta eivät sitä kauemmas. (Mts. 139.) Tyyppi ei siis ole rajakarjalais­­ nominivartalon sisältäviä kymmenkunta lekseemiä (ISK: § 343). murteissa myöhäinen suomalaislaina vaan vanhempaa perua. Johdin Pitäjittäin UtU-johdoslekseemejä esiintyy aineistossani seu­ -UtU- takautuu varhempaan asuun *-δUtU- (ks. Räisänen 1967: 102). raavat määrät: Impilahti 12, Korpiselkä 67, Salmi 4, Suistamo 35, UtU-verbijohdosten muodostusperiaatteet ovat karjalassa ja Suojärvi 21. Huomionarvoista on, että UtU-johdoksia tavataan myös suomessa samat. Johdin liittyy niin verbi- kuin nominikantaan­ livvinmurteisen Salmin aineistossa, vaikka KKS:ssa Salmista ei ole ai­ kin. Verbikanta voi olla vartaloltaan kaksi- tai kolmitavuinen, jol­ noatakaan UtU-verbin esiintymätietoa. (UtU- ja vU-johdosten pitäjä­ loin UtU-johdoksen vartalo on vastaavasti kolmi- tai nelitavuinen kohtaisesta vaihtelusta rajakarjalaismurteissa ks. luku 4.1.3.) (jaksa+utu-, sotke+utu-, vetä+yty-; lujitta+utu-, kellistä+yty-, Rajakarjalaisaineiston UtU-johdoksista suuri osa tunnetaan suo­ varusta+utu-). Nominikanta sen sijaan on UtU-johdoksessa aina messakin: 75 verbiä aineiston 108 verbistä esiintyy myös suomen kaksitavuinen ja johdoksen vartalo siten kolmitavuinen (hike+yty-, yleiskielessä. Jotkin karjalan UtU-verbeistä tosin esiintyvät suomessa lapse+utu-, lika+utu-, noke+utu-). UtU-johdokset ovat yleensä kanta­sanoiltaan läpinäkyviä ja merkitykseltään säännöllisiä. (Suo­ 9. Verbien esiintymien määrä on suurempi kuin verbien lekseemimäärä, sillä samasta lekseemistä voi olla aineistossa useita esiintymiä. men osalta ks. Koivisto 1991: 17; ISK: § 343; karjalan osalta ks. 10. Kantasanaltaan epäselviä, mutta korrelaatillisia ovat esim. verbit liipiyty- Markianova 1985: 22, 99–102.) ’laueta’ ja lymeyty- ’mennä piiloon’ (vrt. liipaista; lymy, lymytä).

262 263 VESA KOIVISTO RAJAKARJALAISMURTEIDEN REFLEKSIIVIVERBEISTÄ tavallisemmin VntU-johtimisena varianttina, kuten hätäyty-, ilmautu-, keskinäistä vastaavuutta ja osin rajakarjalaismurteiden myöhäisem­ loukkautu-, pilautu-, siunautu-, vierautu- (vrt. suomessa hätäänty-, män suomalaistumisen seurausta. ilmaantu- jne.). Verrattaessa rajakarjalaismurteita muuhun karjalan kieleen toimii Suomesta (myös murteista) puuttuvia rajakarjalaisaineiston vertailuaineistona Karjalan kielen sanakirja. Valtaosa rajakarjalais­ UtU-johdoksia ovat mm. verbikantaiset loatiutu- ’sukeutua, mais­ aineistoni UtU-johdoksista esiintyy hakusanoina KKS:ssa: 74 lek­ tua’ (čuajuj juonta rubee luatiutumaa; Korpiselkä, SKNA 8539:2a), seemiä 108:sta (eli 68,5 %). Lisäksi joillakin aineiston verbeistä malttautu- ’malttaa (mielensä)’, pyrkiyty- ’pyrkiä; pyytää päästä (jon­ on KKS:ssa vastineena fonologisesti läheinen synonyymi, esim. nekin)’ (toiset yöks pyrkiyvytää; Korpiselkä, SKNA 4384:2a) sekä hautuutu- (Suistamo, SKNA 0657:1a) vrt. hautoutu- (KKS); liseyty- nominikantaiset­ hiileyty- ’hiiltyä, palaa hiileksi’ (sit tervan kusta (Suojärvi, SKNA 2756:1a) vrt. lisäyty- (KKS). Ne rajakarjalaismurtei­ hiileydyi äijä; Suistamo, SKNA 0657:1a), häpeyty- ’hävetä’, lapseutu- den verbit, jotka puuttuvat KKS:sta, saattavat olla suomalaisperäisiä. ’saada (paljon) lapsia’, liivautu- ’tulla limaiseksi’, rauhoutu- ’rauhoit­ Tällaisia ovat aineistossa esimerkiksi verbit loukkautu- (aineistossa 1 tua’, tukkuutu- ’ker(ään)tyä; tulla kootuksi’. Nämä verbijohdokset esiintymä), pukeutu- (2 esiintymää) ja riisuutu- (1 esiintymä), joiden ovat paljolti karjalan kielelle ominaisia, mutta niiden kantasanat tun­ asemesta muualla karjalassa ovat käytössä verbit sattautu- (aineistos­ netaan suomessakin. Joitakin tällaisia suomen yleiskielestä puuttu­ sa 1 esiintymä), šuoriutu- (’pukeutua’; ei esiinny aineistossa) ja jak- via rajakarjalaisia verbejä tavataan suomen murteissa, mahdollisesti sautu- (’riisuutua’; 2 esiintymää), sekä verbit käyttäyty- (1 esiintymä) muina(kin) kuin UtU-johtimisina variantteina: esim. johdos kapautu- ja suoriutu- (’suoriutua’; 2 esiintymää). esiintyy Sotkamossa UtU-johtimisena, mutta muutoin enimmäkseen Selvien suomalaislainojen lisäksi rajakarjalaisaineistossa on sel­ jonkin toisen kompleksisen refleksiivijohtimen sisältävänä (esim. laisia UtU-johdoksia, jotka eivät esiinny KKS:ssa ja vaikuttavat siten ka(p)pa+antu-, ka(p)pa+atu-, ka(p)pa+u- ym.; SMSA). karjalalle vierailta, mutta joille kuitenkin on karjalassa sopiva kanta­ Edellä mainittujen rajakarjalaisten johdosten kantasana tunne­ sana: esim. roapeutu- (Suistamo, SKNA 0657:1b) ~ roapiutu- (Salmi, taan myös suomessa, vaikka itse johdos suomesta puuttuukin. Sen SKNA 1312:1a) ’kulua, virttyä’ (< roapie ’raapia, kynsiä’), siunautu- sijaan seuraavissa tapauksissa UtU-johdos sekä sen kantasana (tässä (Suistamo, SKNA 0658:1b) ’tulla siunatuksi’ (< siunata), kellistäyty- merkityksessään) esiintyvät pelkästään karjalassa: jaksautu- ’riisuu­ (Korpiselkä, SKNA 4378:1b) (< kellisteä). tua’ (< jaksoa ’riisua’ : jaksa-), kotkautu- ’juuttua kiinni’ (< kotkata Vartaloltaan nelitavuisista UtU-johdoksista (esim. huvittautu-, ’kääntää puun läpi lyödyn naulan pää takaisin puuhun’), liččautu- valmistautu-), joita aineistossa on seitsemän, vain yksi esiintyy ’tunkeutua, ahtautua’ (< ličata ’painaa, tunkea’), puuttautu- ’tar­ KKS:ssa hakusanana (varustautu-). Kaikkien näiden kantaverbinä on rautua’ (< puuttoa ’tarttua’), sattautu- ’loukkaantua’ (= sattoakseh), A-vartaloinen johdosverbi. šiileyty- ’siitä, lisääntyä’ (< šiilie id.), tapautu- ’pyrkiä’ (< tavata Karjalalle ominainen tapausryhmä ovat le-verbeistä johdetut ’tavoittaa’ ym.). Sama koskee sellaisia venäläiseen lainaverbiin UtU-johdokset, joissa johdinainesta on -leUtU- tai -liUtU-. Tällaisia pohjautuvia johdoksia kuin jeäpiyty- ’ilmestyä, näyttäytyä’, rotiu- verbejä ei viiden pitäjänmurteen rajakarjalaisaineistossani ole, mut­ tu- ’syntyä, valmistua’ tai votriutu- ’puhua sievistellen’; tosin ver­ ta Ilomantsin aineistosta löytyi yksi tapaus: siit ei suuteliuvu (SKNA bistä rotiutu- on suomesta pari esiintymätietoa Pohjois-Karjalasta 13014:1b; KKS:ssa ei ole tätä suuteliutuo- ~ suuteleutuo-verbiä). (SMSA), mutta ne ovat siellä varmastikin karjalan kielen vaikutus­ KKS:ssa on leUtU- ~ liUtU-verbejä kaikkiaan 124 hakusanaa (esim. ta. Myöskään merkitykseltään inkoatiivisia verbijohdoksia (ks. luku haikotteliutu-, muutteliutu-); niiden levikkialueena on varsinaiskarjala, 4.1.4.) ei yleensä esiinny suomen kielessä. tarkemmin ottaen vienan- ja eteläkarjalan läntiset osat. Suomen kieles­ Kaikkiaan rajakarjalaismurteiden UtU-verbien vastaavuus suo­ sä tällaisia le-verbikantaisia UtU-verbejä on tavattu Pohjois-Karjalan meen on suuri. Yhtäläisyys on osin koko karjalan kielen ja suomen (SMSA: 12 lekseemiä, esim. kuonisteleutu-, työnteleyty-) ja Kainuun

264 265 VESA KOIVISTO RAJAKARJALAISMURTEIDEN REFLEKSIIVIVERBEISTÄ murteista (SMSA: 20 lekseemiä, esim. veteleyty-, virotteleutu-), mutta vU-johdin voi UtU-johtimen tavoin liittyä niin verbi- kuin nomini­ ei muualta. Näihin itäisimpiin itämurteisiin rajoittuvan levikkinsä pe­ kantaankin, ei kuitenkaan kolmitavuisiin kantoihin, joten vU-johdok­ rusteella tyyppi vaikuttaa suomessa selvästi karjalaislainalta.11 sissa ei ole vartaloltaan nelitavuisia verbejä (*venyttävy-, *varustavu-; vrt. venyttäyty-, varustautu-). Aineiston vU-johdoksista on verbikan­ 4.1.2. vU-johdokset taisia 30 (mm. esimerkkien 4–6 avavu-, kerävy-, hankavu-) ja nomini­ kantaisia 13 (esim. kipevy- ’kipeytyä’, nokevu- ’nokeentua’, uravu- Livvinkarjalassa esiintyy kompleksisena refleksiivijohtimena -vU-, ’tulla hulluksi’). Kantasanojen sanaluokkajakauma on näin ollen esim. ava+vu- ’avautua’, pese+vy- ’peseytyä’, tunke+vu- ’tunkeutua’ samansuuntainen kuin UtU-johdoksilla: verbikantaisia vU-lekseemejä (jälkimmäiset livvinkarjalaisessa asussaan pezevy-, tungevu-). vU- on noin kolme kertaa nominikantaisten määrä. johtimen levikki ei ulotu KKS:n sana-aineksessa eteläkarjalaan, mut­ Aineistoni 43:sta vU-verbistä on Karjalan kielen sanakirjassa ta tämän artikkelin rajakarjalaisaineistossa se kuitenkin yltää jossain hakusanoina 36 kpl eli 83,7 %. Osuus on suurempi kuin aineiston määrin myös eteläkarjalan puolelle.12 Lisäksi vU-johdoksia on tavat­ UtU-johdoksilla, joista 68,5 % esiintyy KKS:n hakusanoina. KKS:sta tu lyydissä, lähinnä sen keskimurteessa livvinkarjalan vaikutuksesta puuttuvat rajakarjalaisaineiston vU-johdokset korjavu-, pestavu-, (Turunen 1946: 140). riisuvu-, vihkivy- ’tulla vihityksi, mennä naimisiin’ sekä nominikan­ Suomen murteissa vU-johdostyyppiä vastaa asultaan tyyppi seka- taiset kiimevy-, kouluvu-, loittovu-. Näistä tosin verbillä kiimevy- ’tulla vu- ~ seka(u- (johtimena -vU- ~ -(U-), jonka levikki on länsimurteinen: kiimaan’ on KKS:ssa synonyyminen lähivastine kiimavu- sekä ver­ lounaismurteiden itäryhmä ja etelähämäläiset murteet sekä näiden vä­ billä korjavu- vastaava UtU-variantti korjautu- ’kohentua; sulkeutua, limurteet. Tällä suomen vU-johdostyypillä ei ole areaalista yhteyttä salpautua’. Levikkinsä ja merkityksensä perusteella verbi korjautu- livvinkarjalan vU-tyyppiin, vaan ne ovat kehittyneet kumpikin omalla (vienalaismurteet sekä Korpiselkä ja Suistamo) ~ korjavu- (Impilahti) tahollaan yhteisestä lähtökohdasta, historiallisesta johtimesta *-δU-, vaikuttaa suomalaisperäiseltä. Jälkimmäisessä on UtU-johdin vaihtu­ johon palautuu myös suomen pohjoisille ja itäisille murteille ominai­ nut vU-johtimeksi, ja suomen UtU-johdoksiin perustunevat muutkin nen johdostyyppi sek(k)au- (josta lisää luvussa 4.1.5.). (Ks. Koivisto tässä mainitut KKS:sta puuttuvat vU-verbit: pestavu-, riisuvu-, vih- 1995: 132–133, 137.) kivy-, kouluvu-, loittovu-, joiden kaikkien kantasana on suomea eikä Rajakarjalaisaineistossani on esiintymiä 43:sta vU-johtimisesta karjalaa.13 verbistä, esim. 4.1.3. UtU- ja vU-johdosten vaihtelu rajakarjalaismurteissa (4) – – ne olˊ umbiläηget muudu gu rindumee sai siduo vai se hevon ryndäs ei avavunnuh. (Salmi, SKNA 3047:2a) Karjalan UtU- ja vU-johdostyypit näyttäytyvät Karjalan kielen sa- (5) myö kerävyimmö yhteh taloih täs da – – kai kerävyimmö sih nakirjassa levikiltään jokseenkin komplementaarisina siten, että täyzi mökki meidy akkoidu. (Suojärvi, SKNA 1010:1b) vU-tyyppi on ominainen livvinkarjalalle ja UtU-tyyppi varsinais­ (6) tää lenkkiharččovoi se ei kestä. ne niäd lenkit hangavuu kivilöis karjalalle (etelä- ja vienankarjalalle). Kuitenkin omassa aineistos­ – –. (Suistamo, SKNA 3044:1b) sani tyypit limittyvät: vU-johdoksia esiintyy kaikissa Raja-Karjalan

13. Lisäksi muutamalla rajakarjalaisaineiston vU-verbillä on suomenmukainen 11. Myös eteläkarjalan saarekemurteissa Sisä-Venäjällä esiintyy le-verbikantaisia merkitys, joka poikkeaa sen merkityksestä muualla karjalassa: esim. verbin hätävy- refleksiiviverbejä (KKS:ssa myös liettšie-loppuisina hakusanoina). käyttöesiintymistä osan merkitys on suomen tapaan ’hätääntyä’ (esim. ei pie 12. Myös jo 1800-luvun loppupuolen murrenäytteestä on todettu vU-johdostyypin hädävyö ei ensinkää; Suistamo, SKNA 0658:1a) eikä karjalan mukaisesti ’joutua levinneen Suojärven ”varsinaiseen”, eteläkarjalaiseen murteeseen (Jeskanen 2011: hätään, puutteeseen’ (esim. leiväs hädävyin ’minulle tuli puute leivästä’; Säämäjärvi, 354). KKS) tai ’turvautua’ (esim. hädävyi miuh ’turvautui minuun’; Suojärvi, KKS).

266 267 VESA KOIVISTO RAJAKARJALAISMURTEIDEN REFLEKSIIVIVERBEISTÄ pitäjänmurteissa (myös eteläkarjalaa edustavissa), ja toisaalta kaik­ (ks. KKS I: XXXI), jotka ovat murteeltaan varsin livvinkarjalaisia kuu­ kialla tavataan vU-johdosten rinnalla UtU-johdoksia (myös livvin­ lumatta kuitenkaan livvinkarjalan varsinaiseen puhuma-alueeseen.14 karjalaisessa Salmissa). Livvinkarjalainen vU-johdostyyppi esiintyy siis myös Raja- Tyypit -UtU- ja -vU- esiintyvät siis rinnakkain koko rajakarjalais­ Karjalan eteläkarjalaismurteisessa osassa, mutta sen levikki ei KKS:n murteiden alueella. Aineistossani UtU-johdostyyppi on yleisempi perusteella näytä ulottuvan Venäjän puolella puhuttuun eteläkarjalaan kolmessa pitäjässä ja vU-tyyppi kahdessa. Lekseemeittäiset luku­ (esim. Porajärvelle tai Repolaan). Tämän yhden johdostyypin levikin määräsuhteet­ ja prosenttiosuudet ovat pitäjittäin seuraavat (UtU- suhteen koko rajakarjalaisalue näyttäytyy siten yhtenäisenä ja livvin­ verbejä – vU-verbejä): Impilahti 13 – 15 (45 % – 55 %), Korpiselkä karjalaisempana kuin Venäjällä puhuttu eteläkarjala. 47 – 7 (89 % – 11 %), Salmi 4 – 7 (36 % – 64 %), Suistamo 35 – 13 Raja-Karjalan eteläkarjalaisessa osassa kuitenkin vallitsee UtU- (73 % – 27 %), Suojärvi 21 – 11 (64 % – 36 %). vU-johdokset ovat johdostyyppi UtU- ja vU-tyyppien rinnakkaiselosta huolimatta. Osa siis enemmistönä Salmissa ja Impilahdella, ja suurin osuus niillä on rajakarjalaismurteiden UtU-verbistöstä on myös omaleimaista muu­ odotuksenmukaisesti livvinmurteisessa Salmissa (64 %). (Aineistos­ hun karjalaan nähden. Aineistossani on varsin paljon KKS:sta puut­ sani UtU-verbien osuus Salmissa on siis 36 %, kun taas KKS:ssa UtU- tuvia UtU-johdoksia, joiden sijalla KKS:ssa on hakusanana vastaava verbeistä ei ole Salmista tietoja lainkaan.) Muualla Raja-Karjalassa – vU-johdos: 7 verbikantaista (7a) ja 8 nominikantaista (7b) UtU-ver­ Korpiselässä, Suistamolla ja Suojärvellä – vU-tyyppi on UtU-tyyppiä biä, esim. harvinaisempi, mutta missään rajakarjalaispitäjässä livvinkarjalaista (7) (a) huuhtoutu- ’huuhtoutua’ (Korpiselkä, SKNA 3039:1a; Suistamo, vU-tyyppiä ei voi pitää vain satunnaisena: vU-verbien esiintymiä on SKNA 1884:1a) vrt. KKS: huuhtovu- kaikissa pitäjänmurteissa huomionarvoinen määrä ja niitä esiintyy kiskoutu- (Suojärvi, SKNA 3086:1a) vrt. KKS: kiskovu- usealla eri puhujalla. Aineistorajaukseni ulkopuolelle jääneestä Ilo­ mantsista tarkistin erikseen vU-tyypin mahdollisen esiintymisen: ai­ pensoutu- (meččäruis se pensoudu; Suistamo, SKNA 0656:1a) neistosta ei löytynyt ainoatakaan esiintymää, joten Ilomantsi poikkeaa vrt. KKS: pensavu- ’kertyä paljon’ tässä suhteessa Raja-Karjalan muista pitäjänmurteista. puuttautu- (Suistamo, SKNA 3044:1b) vrt. KKS: puuttavu- Silti myös KKS:ssa vU- ja UtU-verbityyppien limittyminen raja­ ’tulla liitetyksi, tarrautua’ (< puutta- ’tarttua’) karjalaismurteissa näkyy jossain määrin: vU-johdoksista on jonkin šiileyty- (Suistamo, SKNA 0658:1a) vrt. KKS: šiilevy- ’siitä, verran tietoja myös eteläkarjalaismurteiden puolelta. Kolmeentois­ lisääntyä’ (~ šiiliekseh) ta KKS:n vU-hakusanaan sisältyy esiintymätieto Suistamolta (esim. (b) häpeyty- (Suojärvi, SKNA 3086:1a) vrt. KKS: häpevy- hikevy-, littšavu-, rikkavu-, torevu-), joskin verrattomasti useam­ ’hävetä’ pia esiintymätietoja Suistamolta on UtU-johdoksista: peräti 164:ssä toteutu- (Suistamo, SKNA 0656:1b) vrt. KKS: totevu- ’käydä KKS:n UtU-hakusanassa. Omassa aineistossani vU-verbiesiintymiä toteen’ on myös Korpiselästä (6 lekseemiä), mutta KKS:ssa ei lainkaan (KKS:n UtU-verbihakusanoista on 68:ssa esiintymätieto Korpiseläs­ tukkuutu- (Korpiselkä, SKNA 4383:1b) vrt. KKS: tukkuvu- tä). Impilahden sana-ainesta KKS taas sisältää ylipäätään varsin vä­ ’ker(ään)tyä; tulla kootuksi’ hän: Impilahdesta­ on esiintymätieto yhdessä vU- (istuvu-) ja kahdessa UtU-johdoshakusanassa (oleutu- ’tottua, kotiutua’, terstautu- ’kylläs­ 14. KKS:n sana-artikkeleista eivät näy esiintymätietojen keruukylät, joten ei voida tyä’). Suojärvellä taas vU-tyyppi näyttää KKS:n aineistossa paljon tietää tarkemmin, mistä päin pitäjää tieto vU-johdoksesta kulloinkin on. KKS:n Suojärven aineksen kuitenkin tiedetään olevan Raja-Karjalan alueen eteläkarjalaista yleisemmältä kuin omassa (tosin suppeassa) aineistossani; kuva voi murretta, eikä se ole peräisin esim. Hautavaaran tai Hyrsylän livvinmurteisista johtua siitä, että KKS:n informantit ovat olleet Suojärven itäkylistä kylistä (Jeskanen 2011: 357).

268 269 VESA KOIVISTO RAJAKARJALAISMURTEIDEN REFLEKSIIVIVERBEISTÄ

Esimerkkien 7 UtU-verbit ovat todennäköisesti syntyneet rajakar­ verbin refleksiivinen käyttö esimerkissä 1).16 Aineiston automatiivi­ jalaismurteessa alkuperäisempien livvinkarjalaisten vU-johdosten sissa UtU-verbeissä on myös passiivimuotojen tapaisia, joissa tekijäk­ pohjalta: ne ovat käytännössä ”samoja” kuin livvinkarjalaiset vU- si tulkitaan henkilö: – – luvala sivottii luda. – – ńii hyväšti kui sidoudu varianttinsa, mutta vU-johdin on korvautunut eteläkarjalaisella UtU- se luda da – – (Korpiselkä, SKNA 4370:1b) (joku sitoo); se tuota hei- johtimella (vrt. myös luvun 3.1.2 lopussa mainitut KKS:sta puuttuvat nikko se ei tuota kyntäyvy hyvin (Korpiselkä, SKNA 5137:1b) (joku vU-johdokset pestavu-, riisuvu-, vihkivy-, kouluvu-, loittovu-). kyntää). UtU- ja vU-johdostyypit ovat siis rinnakkaisia ja vaihtelevat koko vU-verbeissä puolestaan aineiston 30:stä verbikantaisesta joh­ rajakarjalaismurteiden alueella. Tällaista lekseemikohtaista vaihtelua doksesta kahdeksalla on semanttisesti refleksiivistä käyttöä (27 % voi esiintyä samassa idiolektissakin, esim. verbeistä) ja automatiivista käyttöä yhdellätoista (37 % verbeistä). Jälkimmäisistä passiivimaisia ovat verbit reššivy- ’tulla päätetyksi, (8) (a) – – jotta tetri pilautuu – –. (Suistamo, SKNA 0656:1b) ratkaistuksi’ (< reššie ’päättää, ratkaista’) ja vihkivy- ’tulla vihityksi, (b) jott‿ei – – hebosen häntä pilavu – –. (Suistamo, SKNA 0656:1b) mennä naimisiin’. (9) (a) ta sih olˊi loppujen lopuksi sih tabautunnuh [’pyrkinyt’] Verbikantaisten vU-verbien merkityslajien yleisyyssuhteet ovat sih – –. (Suistamo, SKNA 0656:1b) rajakarjalaisaineistossani käänteiset UtU-verbeihin verrattuna: vU- (b) kai [muurahaiset] sinne liipin keskikohal ne tabavuttih johdoksissa refleksiiviskäyttöisiä on 27 % lekseemeistä, mutta UtU- [’joutuivat, juuttuivat’] – –. (Suistamo, SKNA 0657:1a) johdoksissa 64 %. Näitä refleksiivisen merkityksen osuuksia voidaan (10) (a) kai liččaudu [’tunkeutui’] sih ni – –. (Suojärvi, SKNA 3095:1a) verrata koko karjalan kielen kattavaan refleksiivijohdosaineistoon (ks. (b) ei täh mahtih ei tiälä midä nyt pietäh g‿ei taloloih liččavutt Koivisto 1995: 249–251): siinä refleksiiviskäyttöisten lekseemien yksih. talol‿ei mänty. (Suojärvi, SKNA 3095:1a) osuus on vU-johdoksissa 47 % ja UtU-johdoksissa 66 % (mts. 249). Vertailen tässä rajakarjalaismurteiden UtU- ja vU-johdoksia toi­ UtU-johdosten refleksiiviskäyttöisten verbien osuus on siis jokseen­ siinsa myös semanttiselta kannalta. Aineistoni 81:stä verbikantaisesta kin yhtäläinen koko karjalan kattavassa aineistossa ja toisaalta tämän UtU-johdoksesta esiintyy semanttisesti refleksiivisessä käytössä 52 artikkelin rajakarjalaisaineistossa. Rajakarjalaismurteiden vU-johdos­ lekseemiä ja automatiivisessa käytössä on 24 (näistä refleksiivi­verbien ten merkitysjakauma sen sijaan poikkeaa koko karjalan vastaavasta. päämerkityksistä ks. Koivisto 1991). Edellisiä on 64 % ja jälkimmäisiä Tähän eroon ei ole osoitettavissa mitään selvää syytä. 30 % kaikista johdostyyppinsä verbeistä.15 Merkitykseltään läpeensä refleksiivinen UtU-johdosten alaryhmä ovat vartaloltaan nelitavuiset 4.1.4. Inkoatiiviset UtU- ja vU-johdokset verbit (esim. lujittautu-, varustautu-). Automatiivisia ovat esim. seuraavat UtU-verbien käyttöyhtey­ Karjalan verbikantaisten UtU- ja vU-johdosten sellainen semanttinen det: nyt‿on terva jo hauduutunu (Suistamo, SKNA 0657:1a); – – ńi erityisryhmä, jota ei tavata suomessa, ovat inkoatiiviset, toiminnan alka­ silmää sattu piä. siidä mäńi mussaks ja pahaks mäńi ylen sih rikkout mista ilmaisevat verbijohdokset.17 Seuraavissa raja­karjalais­esimerkeissä sih men (Korpiselkä, SKNA 4370:1a). Lisäksi refleksiiviskäyttöisistä 11a–f niitä ovat verbit kaččeutu- ’ruveta kiistelemään’, toreutu- ~ UtU-verbeistä seitsemällä on aineistossa myös automatiivista käyttöä, torevu- ’ruveta kiistelemään, tappelemaan’, eläyty- ’asettua asumaan, kuten verbillä peseyty-: ne [= pellavan siemenkodat], kevol oldii siinä alkaa elää’, šireyty- ’(auringosta:) ruveta paistamaan’, säräyty- ’ruveta muudama vuorokaus, ja sid levitimmä gu vihma sattu ni sit kiireel vapisemaan’ ja vareutu- ’alkaa pelätä, pelästyä, säikähtää’. levitimmä jotta ne pezeydyy (Impilahti, SKNA 7481:1a) (vrt. saman 16. Jollain muullakin tämän artikkelin aineistossa pelkästään refleksiivisenä 15. Muutama verbi jää selvän refleksiivisen tai automatiivisen merkitysryhmän esiintyvällä johdoksella voi olla myös automatiivinen käyttömahdollisuus. ulkopuolelle, esim. istautu-, kaivautu-, liitäyty-. 17. Suomessa inkoatiivisten ryhmään voidaan tosin lukea UtU-johdos istuutu-.

270 271 VESA KOIVISTO RAJAKARJALAISMURTEIDEN REFLEKSIIVIVERBEISTÄ

(11) (a) erähäŋ kerran siel‿olˊ, muškat kačeuvuttu. viinua onnakko vU-verbistä; jälkimmäiset ovat rähevy- ’alkaa rähistä’ (< rähise-) mistä lie suatu ja sit, juuvvessa siel, kačeuvuttih kiistele- (Korpiselkä, SKNA 4378:1a) sekä esimerkin 11b torevu-. Lisäksi mäh. toreu(uttih. (Korpiselkä, SKNA 8539:2a) esiintyy Ilomantsissa eläyty-verbin variantti eleyty-: esim. eleyvvytää (b) a sit viinoa ku juodii ka sielpä torevuttiiki. torevuttii ku (SKNA 0509:1e), eleyvyttii (SKNA 5315:2b). Inkoatiivisilla reflek­ veiččilöin ker ruvettii hyppimää – –. (Suojärvi, SKNA 3086:1a) siivijohdoksilla ei ole rinnallaan samakantaisia refleksiivitaivutus­ (c) ([haastattelija:] Mindäh hyö [= tytöt] niin rakkahat tule- vastineita – toisin kuin karjalan refleksiivijohdoksilla yleensä. ­Kar mah oldih?) en tˊii ne talolois tykättii ja taloloih mändih ei, jalaan tämä suomessa tuntematon refleksiivijohdosten käyttötapa on nygyjäh tuada ei niiŋgää taloloih mändäne, van siidä ne selitetty saadun venäjästä, jossa sen vastineena ovat za-prefiksilliset taloh elläyvyttii van talo-. (Korpiselkä, SKNA 4378:1a) verbit (ks. Lehtinen 1985: 67–68). (Inkoatiivisuudesta ks. myös Räi­ (d) no kačommo huondeksel päivät sireydy gä kupperoittaa da sänen 1985: 3; Laakso 1990: 18, 97–98.) läht astumaa. (Suojärvi, SKNA 3095:1a) (e) käet säräydy sih da, – –, häbeydyy mužikka ga, kui sit 4.1.5. Rajakarjalaismurteiden marginaaliset mänöö. (Suojärvi, SKNA 3086:1a) refleksiivijohdostyypit (f) opettaja, kalamaŋ galabie da nous sanoo usson ussoj ja – –, livisti kogo, huonehis, vareutu. Maksima, tappo sem Yleisten UtU- ja vU-johdostyyppien lisäksi rajakarjalaismurteissa ta­ mavon. (Korpiselkä, SKNA 8539:2a) vataan muitakin kompleksisiin refleksiivijohdoksiin luettavia verbi­ tyyppejä. Näitä ovat (v)U-, VtU-, VntU- ja (V)CehtA-johtimiset verbit. Inkoatiivisten verbien kantana tai korrelaattina on usein ise- tai AjA-johtiminen ekspressiivinen kontinuatiiviverbi (Lehtinen 1985: Tyyppi -(V)U- (< *-δU-) 69–70). Näiden oma johdinaines -ise- ~ -AjA- on kuitenkin kadon­ nut inkoatiivijohdoksesta, jonka johdin -UtU- ~ -vU- kiinnittyy verbin Rajakarjalaisaineistooni sisältyy johdoksia, joiden vartalo on asultaan konsonanttiloppuiseen vartaloainekseen (jonka perään tulee siteeksi tyyppiä avau-, heittäy-, pesey-. Tässä johdostyypissä johdinaines -U- e tai A): kačče+utu- (< kač+ise- ’kinailla, riidellä’) (11a), šire+yty- liittyy kantavartalon loppuvokaaliin, joka säilyy. Sen ja U:n välissä ei (< šir+ise-’paistaa kirkkaasti, helottaa’) (11d), säräyty- (< sär+äjä- ole tavunrajaa, ja tyyppi on kuin UtU-johdostyyppi ilman tU-ainesta ’vapista; järistä’) (11e). Myös kaksitavuinen verbi on mahdollinen (ava+u-, heittä+y- vrt. ava+utu-, heittä+yty-). Tyypin nimityksenä on inkoatiivijohdoksen kanta: aineistossani mm. A(A)-vartalokantaiset tässä (V)U-johdokset (vrt. U-johdoksiin tyyppiä heitt+y-, pain+u-). elä+yty- (< elä-) (11c), tore+utu- ~ tore+vu- (< toraa- ’tapella, rii­ Suomen murteissa (V)U-johdostyyppi on laaja-alainen: se on le­ dellä’) (11a–b), vare+utu- (< varaa- ’pelätä’) (11f) ja oleutu- ’tottua, vikiltään leimallisesti pohjoissuomalainen, mutta sitä tavataan myös eri alkaa viihtyä, kotiutua’ (< ole-) (Suojärvi, SKNA 2755:1b).18 puolilla itämurteita (ks. Koivisto 1995: 133). (V)U-tyypin johdin palau­ Rajakarjalaismurteet ovat refleksiiviverbien inkoatiivisen käy­ tuu ainekseen *-δU-, ja sen voidaan katsoa edustavan kompleksisten tön suhteen muun karjalan mukaisia. Merkitykseltään inkoatiivisia refleksiivijohdosten ”alkutyyppiä” (ks. Räisänen 1967: 102–104; Koi­ UtU- tai vU-verbejä esiintyy karjalan kaikissa päämurteissa (ks. Leh­ visto 1995: 137). Sama johdinaines *-δU- on myös livvinkarjalan vU- tinen 1985: 63, 70–72). Rajakarjalaisaineistossani niistä on esiintymiä johtimen lähtökohtana, mutta (V)U-tyypin selvänä erona livvinkarjalan­ vain Suojärveltä ja Korpiselästä: kuudesta UtU-verbistä ja kahdesta vU-tyyppiin on tavunrajattomuus kantavartalon ja johtimen välillä. Pohjoisempana etelä- ja vienankarjalassa *δU-johtimen jatkajaa taas ei 18. Rajakarjalaisaineistoni kahdeksasta inkoatiivijohdoksesta kuusi esiintyy KKS:ssa. Lisäksi aineiston UtU-verbeillä sireyty- (11d) ja vareutu- (11f) on KKS:ssa tavata itsenäisenä refleksiivijohtimena lainkaan (Koivisto 1995: 139); samamerkityksinen vU-vastine širevy-, varavu-. näissä murteissa vallitsee kompleksinen UtU-johdostyyppi.

272 273 VESA KOIVISTO RAJAKARJALAISMURTEIDEN REFLEKSIIVIVERBEISTÄ

Rajakarjalaismurteiden aineistossa esiintyy kahdeksan (V)U- on johtimen t:n edellä kantaverbin vartalovokaalin pidentymä (mil­ johdosta, esim. tä osin tyyppi muistuttaa kohta käsiteltäviä VntU-johdoksia, esim. kerä+änty-). VtU-johdokset ovat vartaloltaan aina kolmitavuisia. (12) (a) puolel toizel avau ta umbee mändii. (Korpiselkä, SKNA 3039:1a) VtU-verbien aineistoesiintymiä ovat esim. seuraavat: (b) Haisu-Miitrei. se ku haisu, pahala. eigä häi käyny saunaa eigä pezeyny eigä. (Impilahti, SKNA 7440:1a) (13) (a) olˊi hyvä hevońi vain niät kum pakkases tulˊi seisoa ńi, jalat (c) täss‿on ukko ei äijäl läägärlöi oo turvau. (Suistamo, SKNA pilaatu hevosella. (Korpiselkä, SKNA 5137:1a) 3044:1b) (b) ku parih mäńin [= menin naimisiin] ja lapseevuin [’sain lapsia’] ńi eihän siid eńeä, tanssissa käyvä. (Suistamo, SKNA Esiintymiä (V)U-tyypistä ei ole monta, mutta kuitenkin neljästä eri pi­ 1883:2a) täjänmurteesta: Korpiselästä neljä lekseemiä, Suistamolta kolme sekä Suojärveltä ja Impilahdesta kummastakin yksi. Kun esiintymät ovat VtU-tyyppiä tavataan myös suomen savolaismurteissa, joissa jotkin lisäksi peräisin kahdeksalta eri puhujalta, kyse ei näytä olevan mis­ VtU-verbit voisi ulkoasunsa puolesta hahmottaa myös UtU-tyypin dif­ tään puhuja- tai pitäjäkohtaisesta erikoisuudesta vaan Raja-Karjalan tonginreduktioiseksi variantiksi (avautu- > avaotu- ~ avaatu-). Kui­ alueen kielimuotoon vakiintuneesti kuuluneesta, joskin harvinaisesta tenkin VtU-johdokset ovat suomessa oma morfologinen tyyppinsä (ks. refleksiivijohdostyypistä. Koivisto 1995: 136) ja sama pätee karjalaan, jossa diftonginreduktiota KKS ei tunne (V)U-johtimisia johdoksia, joten niitä ei ilmei­ ei tunneta lainkaan. sestikään esiinny karjalassa muualla kuin rajakarjalaismurteissa. VtU-johdostyyppi esiintyy karjalassa vain rajakarjalaismurteissa Raja-Karjalan naapurustossa suomen murteissa (V)U-tyyppiä on ta­ (sitä edustavia verbejä ei ole KKS:ssa19 eikä Koiviston 1995 karjalan vattu Pohjois-Karjalan murteiden pohjoislaidalla (missä kuitenkin aineistossa), mikä viittaa lainautumiseen näiden suomenkielisistä vallitseva refleksiivijohdosten tyyppi ovat UtU-johdokset; Koivisto naapuri­murteista. Suomen murteissa VtU-tyyppiä tavataan Savon­ 1995: 133) sekä sporadisesti kaakkoismurteiden länsilaidalla (Muoto- linnan seudun ja Kainuun murteissa sekä Pohjois-Karjalan murteiden opin arkistossa on tietoja mm. Lumivaarasta ja Sortavalasta). Vaikka eteläosissa ja niihin rajautuvissa kaakkoismurteissa (Koivisto 1995: (V)U-tyypin levikki suomen murteissa ei täysin osukaan rajakarjalais­ 136), joista kaksi jälkimmäistä murteistoa sijaitsevat juuri Raja-Kar­ murteita vasten vaan jää pikemminkin kielirajan etelä- ja pohjois­ jalan naapureina. VtU-tyypin suomalaisperäisyyteen Raja-Karjalassa laidoille, on mahdollista ja todennäköistäkin, että (V)U-tyyppi on viittaa sekin, että johdosten pitkä vokaaliaines ei edusta karjalan mur­ Raja-­Karjalassa suomen murteiden vaikutusta. Toisaalta se voisi myös teille muutoin tyypillistä pitkien vokaalien äännekehitystä (esim. aa > edustaa alueella ennen yleisemmin esiintynyttä, sittemmin kadonnutta oa ~ ua; ää > eä ~ iä). refleksiivi­johdos­tyyppiä (johtimena *-δU-), joka on oletettu myöhäis­ kantasuomalaiseksi (ks. Räisänen 1967: 102) ja joka olisi etenkin Poh­ jois-Karjalan murteiden vaikutuksesta voinut korvautua UtU-tyypillä. 19. Mahdollinen VtU-johdos tosin on KKS:n verbi soriitu- ’tulla roskaiseksi’ Tyyppi -VtU- (= soreutuo, sorivuo): pertˊi soŕiidu (Tihvinä; KKS). Toisaalta tämä verbi voisi olla myös sporadista itU-johdostyyppiä (joka tyyppinä on niin ikään Raja-Karjalassa suomalaisperäinen). Toinen rajakarjalaisaineiston itU-johdosesiintymä on haudoiduu Lisäksi rajakarjalaismurteissa esiintyy VtU-aineksisia refleksiivi­ (Suistamo, SKNA 0658:1b); mahdollisia itU-verbejä ovat myös i-vartalokantaiset johdoksia. Aineistossani on esiintymiä kahdestatoista lekseemis­ pyrkiity- ja rotiitu-, jotka tässä on kuitenkin laskettu yleisemmin esiintyvään VtU- tyyppiin. Pidentyneen vokaalin ollessa uu, yy (eli UUtU-vartaloisissa tapauksissa: tä: esim. ava+atu-, kerä+äty-, pyrki+ity-, viska+atu-. Tässä UtU- esim. venyyty-, poakkuutu-) tyypit -UtU- ja -VtU- lankeavat yhteen, jolloin johdoksia muistuttavassa kompleksisessa refleksiivijohdostyypissä frekventimpi UtU-tyyppi on ensisijainen tulkintavaihtoehto.

274 275 VESA KOIVISTO RAJAKARJALAISMURTEIDEN REFLEKSIIVIVERBEISTÄ

Rajakarjalaisaineistoni VtU-johdokset ovat enimmälti verbi­ lainaa suomesta. Samoin kuin VtU-tyypissä, ei VntU-tyypissäkään kantaisia ja merkitykseltään refleksiivisiä: avaatu-, kerääty-, purkaatu-, pitkä vokaali sovi karjalan äännekehitykseen. rotiitu-, viskaatu-. Seuraaville on nominikantaisuuskin mahdollinen tul­ Suomessa VntU-tyyppi on leimallisesti länsimurteinen, vaikka­ kinta: kihlaatu-, kuivaatu-, pilaatu- ja vastaatu- sekä kouleetu- (< koulia ~ kin sitä tavataan myös itämurteissa, ei kuitenkaan Pohjois-Karjalan ja koulu?). Nominikantainen on lapseetu- ’saada (paljon) lapsia’. Kainuun murteiden itä- tai pohjoisosissa (Koivisto 1995: 128–129). Aineiston kahdentoista VtU-johdoksen esiintymät jakautuvat Tyypin puuttuminen juuri Pohjois-Karjalan murteesta ei siis tue ole­ pitäjittäin seuraavasti: Impilahdesta yksi lekseemi, Korpiselästä kuu­ tusta sen lainautumisesta rajakarjalaisiin naapurimurteisiin juuri sieltä. si, Salmista kaksi, Suistamolta neljä ja Suojärveltä yksi. Huomion­ Myös suomen yleiskieli on mahdollinen lainalähde, samoin evakko­ arvoista on jälleen se, että esiintymät hajautuvat usean pitäjänmurteen seutujen savo sotien jälkeen. Kaikki rajakarjalaisaineistosta tavatut ja usean puhujan kesken. Kuten (V)U-tyypissä, tässäkään ei siis ole VntU-lekseemit esiintyvät suomen yleiskielessä ja vastaavat myös kyse paikallisesta tai idiolektikohtaisesta erikoisuudesta, vaan alueen merkitykseltään niitä. murteessa laajemmin, joskin harvaan esiintyvästä johdostyypistä. Tyyppi -(V)CehtA- Tyyppi -VntU- Erilaisten U-vartaloisten verbijohdosten lisäksi refleksiivijohdok­ Rajakarjalaisaineistossa esiintyy myös VntU-johtimisia verbejä (tyyp­ siin voidaan lukea johtimen -(V)CehtA- sisältävä verbityyppi, esim. pi sekaantua, sotkeentua, ärsyyntyä), joita tavataan myös suomessa, ajelehta-, muuttelehta-, höperehtä- (johtimessa C on konsonantti l, r niin murteissa kuin yleiskielessäkin (ks. ISK: § 344). VntU-johdosten tai k). Tyyppiä tavataan koko karjalan kielen alueella ja myös suo­ esiintymiä rajakarjalaisaineistossa ovat: men murteissa. Suomessa se ei kuitenkaan kuulu yleiskieleen, jossa sen vastineena ovat (V)Cehti-vartaloiset verbit (ajelehti-, muuttelehti-, (14) (a) sitki meijjän nuaburin minun velˊ- tuata, dˊiädˊöm poika höperehti-). Tätä i-vartaloista (V)Cehti-tyyppiä puolestaan ei tavata pojal‿olˊ nenkä juur kauppa tuota, hää olˊi sairas. tuola karjalassa. jo olˊi nii huono jotta – – kuoleendu kuvil ihan. (Impilahti, Suomen murteissa A-vartaloista (V)CehtA-johdostyyppiä tava­ SKNA 6818:2a) taan laajalti: itämurteissa sekä länsimurteiden puolella kaikissa pohja­ (b) – – yhtenä päivänä kokouvutah kaikki hyvälˊistö enzin ńiät laismurteissa Etelä-Pohjanmaalta Perä-Pohjolaan (SMSA). Murteiden tietysti parempiozazet. – – kokoontuu kaikki. (Suistamo, (V)CehtA-tyyppi on myös leksikaalisesti paljon monimuotoisempi SKNA 1625:2b) kuin yleiskielen (ja murteiden) (V)Cehti-tyyppi, mm. lekseemejä on VntU-tyypin verbien esiintymiä on aineistossa seitsemän: hajaantu-, enemmän. Suomen murteiden lisäksi (V)CehtA-tyyppiä tavataan kai­ keräänty-, kokoontu-, kuoleentu-, likaantu-, loukkaantu- ja pilaantu- kissa karjalan päämurteissa (myös Raja-Karjalassa). (verbikantaisia muut paitsi hajaantu-). Esiintymiä on Impilahdesta Rajakarjalaisaineistossani esiintyy 18 (V)CehtA-johdosleksee­ kolmesta lekseemistä, Korpiselästä kahdesta ja Suistamolta kolmesta. miä, joista 11 on hakusanana KKS:ssa. Osassa KKS:n sana-artikke­ (Lisäksi olen löytänyt pari tyypin esiintymää Ilomantsin aineistosta.) leista on esiintymätieto rajakarjalaismurteistakin. Kaikkiaan KKS:ssa Jälleen tässäkin on Raja-Karjalassa harvaan esiintyvä johdostyyppi (V)CehtA-verbejä on hakusanoina 361. Suomen murteiden sana-arkis­ silti levikiltään siellä suhteellisen kattava. tossa (SMSA) niitä on arviolta lähemmäs tuhat lekseemiä. Edellä käsiteltyjen VtU- ja (V)U-tyyppien tavoin VntU-johdok­ Leksikaaliselta koostumukseltaan (V)CehtA-tyyppi on niin suo­ siakaan ei tunneta muualta karjalan kielestä (niistä ei ole hakusanoja messa kuin karjalassakin toisenlainen kuin varsinaiset kompleksiset KKS:ssa), joten VntU-tyyppi on rajakarjalaismurteissa selvästikin refleksiivijohdostyypit:(V)CehtA -verbeistä valtaosa on ekspressiivisiä

276 277 VESA KOIVISTO RAJAKARJALAISMURTEIDEN REFLEKSIIVIVERBEISTÄ muodosteita, jotka perustuvat fonologiseen variaatioon, esim. raja­ 1985: 295; Räisänen 1988: 28.) Niillä on usein saman vartaloaineksen karjalaismurteiden hieverehtä- ’liikahtaa hiukan’, räimäkehtä- ’heilua, sisältäviä korrelaatteja (esim. suomen heilu- – heilahta- – heilautta-), toikkaroida’. Kantaverbillisiä, merkitykseltään refleksiivisiä muodos­ jolloin U-aines hahmottuu herkemmin johtimeksi. Lisäksi on sellai­ teita on näihin verrattuna vähän, esim. heittelehtä- (< heittä-), murte- sia kaksitavuisia U-vartaloisia verbejä kuten karjalan ja suomen astu-, lehta- (< murta-). kiehu-, kučču- ~ kutsu-, kysy- tai sattu-, jotka ovat lähinnä johtamat­ tomia perussanoja, vaikka ne taustaltaan voivatkin olla johdoksia (ks. 4.2. U-johtimiset refleksiivijohdokset Räisänen 1988: 24; Koivisto 1995: 241–242). Kaksitavuiset U-verbivartalot ovat Raja-Karjalan eri pitäjän­ Lyhyen U-verbinjohtimen sisältäviä karjalan verbikantaisia johdoksia murteissa kantasanojensa sanaluokan suhteen (kantana substantiivi ovat esim. heitt+y- ’heittyä’, kast+u- ’kastua’, keänt+y- ’kääntyä’, tai verbi tai ei kantasanaa) yhdenmukainen ryhmä: verbikantaisia on näk+y- ’näkyä’, tunt+u- ’tuntua’, asett+u- ’asettua’, nominikantaisia aineistossa pitäjittäin 51–61 %, nominikantaisia 14–18 % (paitsi Sal­ taas esim. kyps+y- ’kypsyä’, must+u- ’mustua’. U-johdostyyppi on missa vähemmän) ja ekspressiivisiä tapauksia 25–38 % verbeistä. vanha, karjalalle ja suomelle yhteinen ja muodostusehdoiltaan niis­ Kaksitavuisten U-johdosten lisäksi rajakarjalaismurteissa tava­ sä yhtäläinen. Johdin ‑U- liittyy kaksitavuiseen (vahva-asteiseen) taan useampitavuisia U-johdoksia. Ne ovat sanavartaloltaan kolmi­ lyhyeen vokaaliin päättyvään verbi- tai nominivartaloon tai moni­ tavuisia ja (i)ttU-, (i)stU-, ntU-, rtU-, ltU- tai htU-loppuisia; näistä tavuiseen A-loppuiseen verbivartaloon. Vartalovokaali jää pois joh­ on (i)ttU- ja (i)stU-verbeissä myös vartaloltaan nelitavuisia. Tämäkin timen edeltä (käänt+y-, kuul+u-, must+u-, asett+u-). Verbikantaiset verbityyppi on karjalassa samanlainen kuin suomessa (ks. ISK: § 347). U-verbit ovat yleensä merkitykseltään automatiivisia, harvemmin Näitä monitavuisia CtU-verbejä esiintyy rajakarjalaisaineistossani 98 refleksiivisiä, kun taas UtU-verbit ovat leimallisemmin refleksiivisiä, lekseemiä, pitäjittäin seuraavasti: Impilahti 34, Korpiselkä 40, Salmi vrt. esim. kast+u- ~ kasta+utu-, murt+u- ~ murt+autu-, järjest+y- ~ 20, Suistamo 42, Suojärvi 22. Lisäksi karjalassa tavataan vartaloltaan järjestä+yty-, pelast+u- ~ pelasta+utu-. (ISK: § 340.) U-johdos­tyyppi (i)ččU-loppuisia verbejä (havaičču- ’herätä’, valičču- ’tulla valituksi, käsittää karjalassa ja suomessa jokseenkin samat verbit, ja ne ovat valikoitua’). käytössä yleistä perussanastoa. Kunkin rajakarjalaispitäjän aineistossa ovat verbikantaiset Aineistoni kussakin pitäjänmurteessa on vartaloltaan kaksitavui­ CtU-verbit (esim. pelast+u- < pelasta-) yleisempiä (osuus pitäjittäin sia verbikantaisia U-johdoksia 14–28 lekseemiä ja kantasanattomia 45–63 %) kuin nominikantaiset tai kantasanattomat. Kantasanattomia (mahdollisesti korrelaatillisia) ekspressiivisiä U-verbejä 9–17 leksee­ ovat esim. kimpastu- ’kimmahtaa; kompastua’, ramistu- ’rapistua, miä pitäjää kohden (Salmia lukuunottamatta). Nominikantaisia kaksi­ raihnaantua’, öntästy- ’kompastua’. tavuisia U-verbejä on yhdestä pitäjästä yleensä 5–10 lekseemiä, mutta CtU-verbien johtimen rakenne on (i +) konsonantti C + tU (eli Salmista on vain yksi nominikantainen tapaus eli 4 % sen kaikista -(i)CtU-). Mahdollisia konsonantteja tU-aineksen edellä ovat t, s, n, kaksitavuisista U-johdosverbeistä. Yhteensä U-johdoksia on raja­ r, l ja h. Rajakarjalaisaineistossa esiintyy CtU-johtimesta eniten va­ karjalaisaineistossani 55 lekseemiä. rianttia -stU- (58 lekseemiä) ja seuraavina ovat -ttU- (19 lekseemiä) Kantasanattomia vartaloltaan kaksitavuisia U-verbejä ovat esim. ja -htU- (10 lekseemiä); muut variantit ovat harvinaisempia. Myös suomen haukku-, huoju-, kirku-, läikky-, pärsky- tai karjalan nuuru- KKS:ssa stU-verbit ovat selvästi yleisempiä kuin ttU-verbit (-stU- ’kitua, aristella’, risu- ’lohkeilla, murtua, särkyä’, vanku- ’parkua, n. 1200, -ttU- n. 650 verbiä), samoin suomen yleiskielessä, jossa ulista, mankua’. Tällaiset U-verbit ovat merkitykseltään yleensä (i)stU-vartaloisia verbejä on nelisensataa ja (i)ttU-vartaloisia pari­ kontinuatiivisia eli jatkuvaa tapahtumaa ilmaisevia ja fonologisel­ sataa sanakirjalekseemiä, muunvartaloisia selvästi vähemmän (ISK: ta asultaan ja motivaatioltaan ekspressiivisiä. (Kulonen-Korhonen § 347).

278 279 VESA KOIVISTO RAJAKARJALAISMURTEIDEN REFLEKSIIVIVERBEISTÄ

Monitavuiset CtU-vartalot ovat usein tulkittavissa vastaavan 4.3. Yhteenvetoa rajakarjalaismurteiden CtA-vartaloisen verbin U-johdoksiksi, esim. asett+u- < asetta-, refleksiivijohdoksista alent+u- < alenta-, kumart+u- < kumarta-, hivelt+y- ’nyrjäh­ tää’ < hiveltä- ’nyrjäyttää’, unoht+u- < unohta-, valičč+u- < (e-var­ Rajakarjalaismurteiden refleksiivijohdoksisto poikkeaa morfologisten talo) valičče-. Johdin -U- ja sitä edeltävä Ct-aines muodostavat johdostyyppien määrän suhteen muusta karjalasta. Karjalan komp­ siis yhdessä johdinten yhtymän (-stU-, -ntU- jne.). Toisaalta kanta­ leksiset refleksiivijohtimet ovat -UtU- (varsinaiskarjalassa) ja -vU- verbin puuttuessa johdoksen kantasanaksi voi hahmottua kaksi­ (livvinkarjalassa), joiden levikkialueet ovat jokseenkin komplemen­ tavuinen nomini (adjektiivi tai substantiivi) ja johtimeksi -(i)CtU-, taariset (Koivisto 1995: 147). Kuitenkin koko rajakarjalais­murteiden esim. heiko+stu- ’heiketä’ (< heikko), laho+stu- ’lahota’ (< laho), alueella (pois lukien Ilomantsi) UtU- ja vU-tyypit esiintyvät rin­ polv+istu- ’polvistua; mennä mutkalle’ (< polvi), tai konsonantti­ nakkain, ja lisäksi niitä täydentää koko joukko suomalaisperäisiä loppuisen kantasanan tapauksessa -tU-, esim. viisas+tu- (< viisas), refleksiivi­johdostyyppejä (ks. luku 4.1.5.). Mikään aineistoni pitäjän­ humal+tu- (< humala) (ks. Koivisto 1995: 242). murteista ei tässä suhteessa erotu joukosta, joskin UtU- ja vU-tyyppi­ Lyhyitä, muita kuin kompleksisten ryhmään luettavia refleksiivi­ en keskinäiset yleisyyssuhteet vaihtelevat pitäjittäin. johtimia ovat myös johtimet -tU- ja -pU-. Johdin -tU- liittyy yleen­ UtU- ja vU-tyyppejä harvinaisemmat (V)U-, VtU-, VntU- ja itU- sä nominin konsonanttiloppuiseen vartaloon: joko yksitavuiseen, johdostyypit eivät ole karjalaisia, vaan ne ovat Raja-Karjalassa selväs­ esim. hiil+tu- (< hiili : hiile-) ’hiiltyä’, laih+tu- (< laiha) ’laihtua’, ti suomen kielen vaikutusta. Ne eivät myöskään ole aineistossani kos­ neyr+ty- (< neyrä) ’nöyrtyä’, nuor+tu- (< nuori : nuore-) ’nuortua’, kaan puhujan yksinomainen kompleksinen refleksiivijohdostyyppi.­ vart+tu- (< varsi : varte-) ’kasvaa; odottaa’, tai kaksitavuiseen, esim. Ylipäätään rajakarjalaismurteille on luonteenomaista refleksiivisten hämär+ty- (< hämärä) ’hämärtyä’, kevät+ty- (< kevät-) ’laihtua, hei­ johdostyyppien vaihtelu samassakin idiolektissa: esim. Suistamo ketä keväällä’, sairas+tu- (< sairas) ’sairastua’. Kantasanan vokaali­ kokouvutah ~ kokoontuu (SKNA 1625:2b), pilautuu ~ ei pilavu ~ vartalo menettää siis johdoksessa loppuvokaalinsa; tämä on yleensä pilaantuu ~ pilaatunu (SKNA 0656:1b). e, mutta joskus A vaikka A-vartaloisen kantasanan taivutusmuodoissa Harvinaisten (V)U-, VtU-, VntU- ja itU-johdostyyppien si­ ei tällaista konsonanttivartaloa esiintyisikään (esim. laih-, hämär-). nänsä vähäisehköt esiintymät jakaantuvat Raja-Karjalan alueella Joskus johdin -tU- liittyy yksitavuiseen vokaaliloppuiseen nominiin, usean pitäjänmurteen ja usean puhujan kesken. Niiden esiintymi­ esim. jeä+ty- ’jäätyä’ (< jeä ’jää’), peä+ty- ’päätyä’ (< peä ’pää’). nen Raja-Karjalan alueella vaikuttaa näin ollen tasaiselta ja vakiin­ Nominikantaisia tU-verbejä aineistossani on 13 vartaloltaan tuneelta. Tyypeistä yksikään ei näytä olevan puhujilla niin tuore, kaksi­tavuista ja 8 kolmitavuista. Verbikantaisia tU-johdoksista ovat että se olisi vasta evakkoaikana uuden asuinseudun murteesta saa­ (e)le- ja Oi(čče)-verbeistä johdetut, jollaisia on aineistossani vain tua vaikutusta. kaksi lekseemiä: palel+tu- ja ahavoi+tu-. Rajakarjalaisaineiston tU- Inkoatiivista eli toiminnan alkamista ilmaisevaa käyttöä (ks. johdokset ovat suurimmalta osin samoja kuin suomessa (jonka tU- luku 4.1.4.) esiintyy rajakarjalaisaineistossani vain UtU- ja vU-joh­ verbeistä ks. ISK: § 341–342; Räisänen 1988: 32–38, 43–54). Verbin­ dostyyppien verbeillä. Tämä käyttötapa on todettu venäjän vaiku­ johdin -pU- on harvinaisempi. Se esiintyy joihinkin yksitavuisiin tukseksi (ks. Lehtinen 1985: 72–73), ja siinä karjala eroaa selvästi vartaloihin liittyneenä, esim. juo+pu-, syö+py- (ks. Koivisto 1995: suomesta, jossa tällaista inkoatiivista käyttöä ei tavata. Inkoatiivisuu­ 126). Rajakarjalaisaineistossa tällaisia johdoslekseemejä on 8. den esiintymättömyys aineiston harvinaisemmissa, suomalaisperäi­ sissä johdostyypeissä vahvistaa osaltaan käsitystä näiden johdos­ tyyppien suomalaislähtöisyydestä. (Huomionarvoista on toisaalta myös se, että refleksiivitaivutuksella, joka tyypillisesti on karjalassa

280 281 VESA KOIVISTO RAJAKARJALAISMURTEIDEN REFLEKSIIVIVERBEISTÄ refleksiivijohdoksille vaihtoehtoinen ilmaisukeino, ei tavata inkoatii­ 5. Refleksiivitaivutus vista käyttöä millään karjalan murrealueella.) Karjalan kielelle ovat ominaisia runsaat leksikaaliset lainat ve­ 5.1. Refleksiivitaivutus itämerensuomalaisissa kielissä näjästä. Rajakarjalaisaineistossani kuitenkin esiintyy kantasanaltaan ja karjalassa venäläisperäisiä UtU- tai vU-verbijohdoksia hyvin vähän, vain seu­ raavat: jeäpiyty- (Korpiselkä, SKNA 7495:2a, 8539:2a; vrt. KKS: Refleksiivitaivutus on verbien kieliopillinen kategoria, jota tavataan jeäviyty-) ’näyttäytyä, ilmestyä’ (< jeävi-), molivu- (Suistamo, SKNA itäisissä itämerensuomalaisissa kielissä sekä vatjassa (Koivisto 1995: 3044:1a) ’rukoilla’ (< moli-), reššivy- (Suojärvi, SKNA 3095:1a) ’tulla 57). Määritelmän mukaan refleksiivitaivutuksella tarkoitetaan sarjaa päätetyksi, ratkaistuksi’ (< rešši-), uravu- (Impilahti, SKNA 7440:1b) (semanttisesti refleksiivisiä) verbinmuotoja, jotka eroavat vastaavista ’tulla hulluksi’ (< uroa ~ urai ’hullu’). Suurin osa näiden kantasanois­ normaalitaivutuksen verbinmuodoista persoonapäätteidensä tai mui­ ta on verbejä: kaksitavuisia i-vartaloisia venäläislainoja. den suffiksiensa puolesta (Posti 1980: 111). Refleksiivinen taivutus­ Rajakarjalaisaineistossani leksikaalisia lainoja suomesta puoles­ muoto sisältää normaalitaivutuksen tunnusten sijaan tai lisänä reflek­ taan ovat (joko kokonaisina johdoksina tai vain kantaverbiltään) esim. siivismerkityksisen morfeemin eli refleksiivitunnuksen (tai useamman seuraavat: korjautu- (Suojärvi, SKNA 2755:1a), pestavu- (Suistamo, sellaisen). Refleksiivitaivutus edustuu normaalin verbintaivutuksen SKNA 0658:1a), riisuutu- ~ riisuvu- (Suistamo, SKNA 3043:1a), tel- rinnalla omana taivutusmuotojen sarjanaan. Normaalitaivutuksen läyty- ~ tellävy- (Korpiselkä, SKNA 4378:1a), uskoutu- (Impilahti, muodoista poiketen refleksiivitaivutuksen muodot ovat yleensä se­ SKNA 1274:1a), vihkivy- (Suojärvi, SKNA 3095:1a). Näitä verbi­ manttisesti refleksiivisiä, ja ne vastaavat merkitykseltäänU -vartaloisia johdoksia tavataan karjalassa vain rajakarjalaismurteissa ja UtU-va­ refleksiivijohdoksia (lähinnä UtU-johtimisia).20 Refleksiivimuoto siis riantteja mahdollisesti myös vienalaismurteissa. vU-johtimisissa ta­ eroaa paitsi muodoltaan myös merkitykseltään vastaavasta normaali­ pauksissa johdin on rajakarjalaista perua, mutta sen kantaverbi sekä taivutuksen muodosta.21 malli johdoksen muodostukselle ovat peräisin suomesta. Lekseemei­ Karjalan refleksiivitaivutuksen morfologia vaihtelee murteittain nä suomalaislähtöisiksi voidaan katsoa myös kaikki harvinaisempien paljon. Karjalan päämurteista vain livvinkarjalassa sekä eteläkarjalaa VtU-, VntU- ja (V)U-johdostyyppien verbit (joista ks. luku 4.1.5.). edustavissa saarekemurteissa refleksiivitaivutus on koko murrealueella­ Lekseemi voi olla suomalaislähtöinen joko merkityksineen kaik­ suhteellisen yhtenäinen. (Yleiskuva karjalan refleksiivitaivutuksesta ks. kineen tai vain fonologiselta tai morfologiselta asultaan. Esimerkik­ Koivisto 1995: 82–84, jossa vienalaismurteita edustavat paradigmat si verbi pelastu- esiintyy rajakarjalaisaineistossani kaikissa kolmen 1–3, eteläkarjalaa paradigmat 4–9 ja livvinkarjalaa paradigma 10.) eri pitäjänmurteen esiintymissään suomen mukaisessa fonologisessa Seuraavassa asetelmassa on näyte karjalan refleksiivitaivutukses­ asussa, kun taas etelä- ja livvinkarjalassa sen vartaloon kuuluu ge­ ta Ahtian (1938: 120–125) Karjalan kieliopin mukaisena, jonka kuvaus­ minaatta ll (sekä ai-diftongi): pella(i)stu- (ks. KKS s.v. pelastuo ja perustuu tältä osin lähinnä eteläkarjalaiseen Suojärven murteeseen. pelastoa). 20. Esimerkiksi seuraavassa karjalan tapauksessa refleksiivitaivutus andaiččov( ) ja UtU-johdos (andauduv) esiintyvät samassa merkityksessä (’sopii hyvin, maistuu’): ”süömine andaiččov, kui ombelov (Kid.), nügöin uńi andauduv (Suoj.)” (Genetz 1884: 36–37). 21. Merkitysero ei tosin koske aivan kaikkia refleksiivimuotoja: mm. AhtA- ja le- johtimiset verbit saattavat (ainakin tietyissä muodoissaan ja tietyillä murrealueilla) taipua refleksiivitaivutuksen mukaisesti, vaikka niiden merkitys on tällöin sama kuin ilman refleksiivitunnustakin (Koivisto 1995: 244–245), esim.kummitella+kseh ’kummitella’, hypähti+h ’hypähti’.

282 283 VESA KOIVISTO RAJAKARJALAISMURTEIDEN REFLEKSIIVIVERBEISTÄ

Verbinä on heitteä ’heittää’, jonka refleksiivisten taivutusmuotojen Verbin refleksiivinen paradigma esitetään sanakirjoissa omana leksee­ merkityksenä on ’heittäytyä’. Refleksiivitaivutus kattaa periaatteessa­ minään: KKS:ssa niiden hakusanat ovat tavallisimmin kseh-loppui­ samat taivutusmuodot (sananmuototyypit) kuin normaalitaivutuskin,­ sia, esim. heitteäkseh (’heittäytyä’), ja Sisä-Venäjän tytärkarjalaisten joskin refleksiiviset paradigmat ovat käytännössä usein vajaita murteiden osalta ttšie-loppuisia: heitteättšie (id.). Lisäksi KKS:ssa on normaalitaivutuksen­ paradigmaan nähden. (Tässä näyteparadigmassa pieni joukko hakumuodossaan VijA-loppuisia Impilahden, Suistamon on mukana keskeisimpiä verbinmuotoja, ei kaikkia.) ja Salmin rajakarjalaismurteiden refleksiivitaivutteisia verbejä, esim. jaksaija ’riisuutua’, šuoreija ’pukeutua’, viruttaija ’loikoa’ (näiden REFLEKSIIVITAIVUTUS NORMAALITAIVUTUS paradigmamuodot voivat silti olla samanasuisia kuin vastaavassa kseh- tai ttšie-loppuisessa hakusanassa). AKT IND PRS Refleksiivitaivutuksen vanhimpia muotoaineksia on 3. persoo­ Y1 heitämmös heitän nan imperfektin refleksiivitunnus, taivutussuffiksi *-hen, joka esiintyy Y2 heitättös heität malliparadigmamme muodossa heitti+h ’heittäytyi’ (< *heitti+hen) Y3 heitäh(es) heittäy sekä preesensiin levinneenä myös muodossa heitä+h ’heittäytyy’. M1 heitämmökseh heitämmö Alkuperäinen preesensin 3. persoonan refleksiivitunnus on *-ksen, M2 heitättökseh heitättö joka tavataan esim. vienalaismurteiden muodossa pane+kše ’paneu­ M3/PASS heitetähes ~ heitäh(es) heitetäh tuu’. Refleksiivinen taivutussuffiksi *-hen (~ *-sen) on Laanestin AKT IND IMPF (1975: 156–158) mukaan alkuaan pronominaalinen aines ja historial­ Y1 heitimmös heitin lisessa yhteydessä samanasuiseen possessiivisuffiksiin sekä persoona­ Y2 heitittös heitit pronominiin hän. Y3 heittih(es) heitti Toinen refleksiivitaivutuksen perusaines on johdin *-δe-, joka M3/PASS heitettihes heitettih esiintyy mm. paradigman kielteisissä muodoissa (en, et… heittä+i NEG ’en, et… heittäydy’ < *heittä+δe+k) sekä yksikön 2. persoonan impe­ kielt. PRS (konnegatiivi) en, et… heittäi en, et… heitä ratiivissa (heittä+i ’heittäydy’ < *heittä+δe+k). Nämä ovat yksikön 3. kielt. IMPF en, et… heittänyhes en, et… heittänyh persoonan indikatiivin preesensmuotojen ohella refleksiivitaivutuksen­ kielt. PASS PRS ei heitetähes ei heitetä alkuperäisimpiä muotoja (refleksiivitaivutuksen historiasta ks. Pos­ kielt. PASS IMPF ei heitettyhes ei heitetty ti 1980; Koivisto 1989a, 1995: 120–121). Lisäksi refleksiivitaivu­ tuksen johtimia ovat -tte- ja -čče-, jotka osallistuvat astevaihteluun IMPT (-tte- : -te-; -čče- : -če-) ja joista jälkimmäinen on kehittynyt edel­ Y2 heittäi ~ heittäjje heitä lisestä.22 Rajakarjalaismurteissa tavataan lähinnä vain čče-johdinta, M2 heittäkkeäkseh ~ heittäkkeä mutta aineistossani on myös muutama tte-aineksinen muoto (ks. heitteättökseh luku 5.2.1.). Näitä johtimia edeltää sananmuodossa yleensä joh­ INF din *-δe-, esim. pezi+e+če+mmä ’peseydymme’ (Koivisto 1995: A-INF heitteäkseh heitteä 88) (< *pese+δe+tte+mmä), šuori+e+te+t ’pukeudut’ (mts. 87) Asetelma 1. Näyte karjalan refleksiivitaivutuksesta, heitteä ’heittää’. (< *suori+δe+tte+t). Johdin *-δe- toteutuu karjalan refleksiivimuo­ doissa tavallisesti diftongin jälkikomponenttina: heittä+i ’heittäydy’, 22. Affrikaattaa merkitään tässä artikkelissa č:llä muutoin paitsi KKS(A):sta peräisin olevassa aineksessa tš:llä.

284 285 VESA KOIVISTO RAJAKARJALAISMURTEIDEN REFLEKSIIVIVERBEISTÄ ei laske+i ’ei laskeudu’, älä azetta+i ’älä asettaudu’; šuori+e+te+t 1932: 77), mon. 3. p./pass. preesens ripuštuatšetah ’hirttäytyvät’, im­ ’pukeudut’, murdo+a+te+h ’murtautuu’, valu+a+tše+mma ’huuh­ perfekti rikkuotšettih ’rikkoutuivat’ (KKS), A-inf. ńägietšie ’nähdä toudumme’, laški+e+tšo+u ’laskeutuu’, rikku+o+tšo+u ’rikkoutuu’ toisensa’ (KKS), NUT-partis. särgietšen ’särkynyt’ (KKSA). Myös (KKS s.v.). yksikön 3. persoonan muodoissa čče-johdokset ovat selvänä enem­ Vähimmillään yksittäisen ims. kielen tai murteen refleksiivi­ mistönä, esim. preesens kˊiskuotšou ’kiskoutuu’ (KKS), šuoŕieččoo taivutukseen kuuluvat vain edellä mainitut kulmakivimuodot, joiden ’pukeutui’ (Rjagojev 1980: 117), imperfekti muuttuattši ’muutt(aut) suffikseina ovat *-hen, *-ksen ja johdin *-δe-, eikä refleksiivinen pa­ ui’, peittiätši ’peittäytyi, piiloutui’ (KKS), joskin myös alkuperäisiä h- radigma ole laajentunut niistä muihin taivutusmuotoihin. Tälle tasalle loppuisia taivutusmuotoja tavataan imperfektissä, esim. rodˊih ’syntyi; refleksiivitaivutus on jäänyt esimerkiksi karjalan vienalaismurteissa, tuli’ (Palmeos 1962: 44), šeizattih ’seisattui’, ulloštih ’ulosti’ (KKSA), kun taas mm. livvinkarjalassa ja vepsässä paradigma on täydentynyt myös pitkä-i:llisinä, kuten lˊiikahtiih ’liikahti’ (KKS), ja preesensis­ erilaisia taivutustunnuksia sisältävillä persoonamuodoilla. sä sekä h- että ks-loppuisia muotoja: skokahtah ’hypähtää’ (Rjagojev Ims. kielten refleksiivitaivutukseen kuuluu siis paitsi taivutus­ 1980: 88), venyttääkš ’venyttäytyy’ (KKSA). Lisäksi imperfektissä on muotoja myös sellaisia johdoksia, jotka toimivat osana refleksiivistä sekamuotoja, joissa on sekä johdin että refleksiivinen taivutusaines-h : taivutusparadigmaa. Näiden johtimina ovat karjalassa historiallinen heitˊtˊiätšiih ’heittäytyi’, laškietših ’laskeutui’ (KKS).24 *-δe- sekä -tte- ja -čče-.23 Johdinten levinneisyys paradigman sisällä Tytärkarjalaiset murteet ovat kehittyneet Raja-Karjalassa vaihtelee kielittäin ja murteittain. Osassa karjalan murteita johtimilla 1600-luvulla puhuttujen murteiden pohjalta. Refleksiiviparadigman -čče- ja -tte- on taipumus muodostaa kokonaisia persoonamuotojen samankaltaisuus kaikissa tytärkarjalan alamurteissa (Tver, Tihvinä sarjoja (johdosmuoto 1:sen ja 2:sen persoonan lisäksi 3:nnessakin), ja Valdai) viittaa siihen, että paradigman johdostumiskehitys on ollut joka kehitys on edennyt pisimmälle tytärkarjalaisissa murteissa ja nä­ Raja-Karjalassa meneillään jo silloin, kun karjalan puhujia on muut­ kyy myös rajakarjalaisaineistossani. Samankaltainen kehitys on lyy­ tanut sieltä uusille asuinsijoille. (Rajakarjalaismurteiden refleksiivi­ din murteissa ollut taivutustunnuslähtöisellä refleksiivijohtimella -ze- taivutuksen johdosmuodoista ks. luku 5.2.1.) (< *-sen) (ks. Koivisto 1993). čče-ainesta hyödynnetään karjalassa paitsi refleksiiviaineksena Tytärkarjalaisten murteiden refleksiivitaivutuksessa siis vallitsee myös lainattaessa venäjän verbivartaloita karjalaan. Tässä käytössä čče-johtiminen paradigma: johdoksia esiintyy kaikissa persoonissa -čče- ei ole tehtävältään varsinaisesti johdin vaan toimii verbivartaloa (mukaan lukien passiivimuodot) ja myös infiniittisissä muodoissa. muodostavana elementtinä, jolla venäjän Vj-loppuinen verbivartalo Seuraavassa on joitakin näytteitä tällaisista muodoista poimittuina li­ sovitetaan karjalaan, esim. torgui-čč-i ’kaupitsi’, gulˊai-č-i-n ’käve­ sensiaatintyössäni (1989b) esimerkkeinä esitetyistä n. 130:sta tytärkar­ lin’. Tosin čče-ainesta ei tule niihin muotoihin, joissa taivutussuffik­ jalaisten murteiden refleksiivimuodosta: yks. 1. p. preesens laškiečen si on konsonanttialkuinen: torgui-mah ’kaupitsemaan’, pruaznui-ja ’laskeudun’, imperfekti muuttuatšiin ’muut(taud)uin’ (KKS), yks. 2. p. ’juhlia’ (poikkeuksena preesensmuodot: niissä -čče- on mukana, esim. preesens valuačet ’valaudut’ (Virtaranta 1967: 52), mon. 1. p. preesens pruaznui-če-mmo ’juhlimme’). (Pugh 1992: 90–91; ks. myös Markia­ ńäǵiečemmä ’näemme toisemme’, imperfekti jakšuattšima ’riisuu­ nova 1985: 78.) Venäläislainoissa čče-aines ei siis ulotu yhtä syvälle duimme’ (KKS), kielt. preesens et tungieče ’et tunkeudu’ (E. Leskinen paradigmaan kuin refleksiivitaivutuksessa, ja niissä tytärkarjalaiset murteet edustavat samaa kantaa muun karjalan kanssa. 23. Koiviston (1995) morfologisessa kuvauksessa refleksiivitaivutuksen kolmen joht­imen symboleina ovat TE (= *-δe-), TTE (= -tte-), CCE (= -čče- ~ -ttše-) 24. Johdinaines -čče- on itse asiassa sama kuin (i)tse-aines suomen merkitse- (mts. 65–66; aineksen alleviivaus tarkoittaa, että kyseessä on johdin eikä tai­vutus­ tai urakoitse-tyyppisissä verbeissä. Näillä verbeillä on tosin konsonanttivartalo, aines). Taivutusainesten symboleita kyseisessä kuvauksessa ovat vastaavasti 1. per­ esim. suomen merkit+tiin, karjalan merkit+tih, jollaisia karjalan čče-refleksiivi­ soo­nan M, 2. persoonan T ja 3. persoonan H ja KS (mts. 60–64). paradigmassa taas ei ole.

286 287 VESA KOIVISTO RAJAKARJALAISMURTEIDEN REFLEKSIIVIVERBEISTÄ

Karjalassa čče-johdinaineksella on siis kahdenlaista käyttöä: se tiedot ovat Suojärven itäkylistä (ks. Genetz 1870: 205–206; tosin Ge­ toimii osassa murteita refleksiivitunnuksena (liittyneenä *δe-johtimen netz ei mts. 230–232 mainitse tarkemmin, minkä kylien murteeseen perään) ja toisaalta venäläisten lainaverbien mukauttaja­suffiksina. tämä refleksiivitaivutuksen­ kuvaus perustuu).26 Karjalan kielen sanakirjassa ero kahdenlaisten čče-verbien – Rajakarjalaisaineisto täydentää aiempia tietoja rajakarjalais­ refleksiivi­verbien ja venäläislainojen – kesken näkyy yleensä siinä, murteiden refleksiivitaivutuksesta olennaisesti, joskin kuvausta hait­ että (ei-refleksiivisten) lainaverbien (kuten venčaija) hakumuoto taa epävarmojen, tulkinnallisten muotojen suuri määrä (etenkin Korpi­ (A-infinitiivi) on VijA-loppuinen, kun taas tytärkarjalan refleksiivis­ selän ja Suistamon aineistoissa). Refleksiiviset taivutusmuodot eivät ten čče-verbien hakusanat ovat ttšie-loppuisia.25 Muiden murteiden myöskään ole teksteissä siinä määrin frekventtejä, että suurestakaan refleksiivi­taivutushakusanat ovat kseh-loppuisia. aineistosta kertyisi riittävästi muotoja täysin kattavaan kuvaukseen tai selvien alueellisten erojen esittämiseen. Nyt litteroitu aineisto ei ole 5.2. Refleksiivitaivutus rajakarjalaismurteissa ylipäätään tarpeeksi laaja esim. variaation kvantitatiiviseen tarkaste­ luun. Impilahden, Korpiselän ja Suistamon aineiston täydentämiseksi Rajakarjalaismurteista etenkään Impilahden, Suistamon ja Korpiselän­ olen hyödyntänyt Karjalan kielen sanakirjaa sekä pitäjänmurteittain refleksiivitaivutuksesta ei ole tähän asti ollut kovin tarkkaa kuvaa. muitakin lisälähteitä (joista ks. tarkemmin jäljempää).27 Näitä pitäjänmurteita on ylipäätäänkin dokumentoitu (esim. KKS:ssa) vähemmän kuin Suojärven ja Salmin murteita. Toisaalta juuri kahden 5.2.1. Suistamo ja Impilahti viimeksi mainitun pitäjänmurteen refleksiivitaivutuksen tarkastelu jää tässä artikkelissa muutaman yleisen huomion varaan siitä syys­ Refleksiivitaivutus on morfologisesti samankaltainen Suistamon ja tä, että niiden litteroitu aineisto on suppeahko ja sisältää vain vähän Impi­lahden murteissa. Siinä on yhtäläisyyksiä niin karjalan kuin suo­ refleksiivitaivutusesiintymiä. menkin murteiden refleksiivitaivutuksen kanssa. Suomen mukaisia Rajakarjalaismurteiden refleksiivitaivutus edustaa yleisesti ot­ ovat muodot, joissa refleksiivitunnuksena on johdin-i- (< *-δe-), esim. taen eteläkarjalaismurteille ominaista jossain määrin johdospitoista Impilahti jaksa-i-jja ’riisuutua’ (E. Leskinen 1934: 95), laske-i-jjessa paradigmaa (ks. Koivisto 1995: 90–91; Genetz 1880: 234–238). Joh­ ’laskeutuessa’ (Punttila 1992: 21), Suistamo viruta-i-ja ’virua’ (vrt. doksen asuisia refleksiivimuotoja tavataan myös livvinkarjalaisessa suomen pesse-i-jjä, huuhto-i-jjessa, huvittelle-i-jjessa; ks. Koivisto Salmissa, esim. istu+o+če-kkah ’istuutukoon’ (Genetz 1884: 182), 1995: 90–91). jonka refleksiivitaivutus kuitenkin muutoin on paljolti livvinkarjalan Rajakarjalaisaineistossani on Suistamon refleksiivitaivutuksesta mukainen. Raja-Karjalan pitäjistä Suojärven refleksiivitaivutus on 29 esiintymää, jotka edustavat 24:ää eri lekseemiä, ja Impilahdesta jokseenkin samanlainen kuin Ahtian malliparadigmassa edellä (ase­ 34 refleksiivitaivutusesiintymää (ja lisäksi 3 epävarmaa) 30:stä eri telma 1) ja kaikkiaan hyvin samankaltainen kuin livvinkarjalassa (ks. lekseemistä. Koivisto 1995: 82–84, jossa Suojärven murretta edustaa paradigma 9 ja Nauhoista litteroitua aineistoa täydentävänä lähteenä toimii livvinkarjalaa­ paradigma 10). Koiviston (mp.) esittämän paradigman KKS, jossa on Impilahden murteesta esiintymätieto kahdeksassa 25. Kun myös venäläisten lainaverbien hakusanat ovat VijA-loppuisia, ne on osattava erottaa Impilahden, Suistamon ja Salmin VijA-loppuisista refleksiiviverbeistä. 26. Erona Ahtian paradigmaan (ks. luku 5.1.) ovat Suojärvellä refleksiiviaineksen Esimerkiksi Impilahden karjalan sanakirjassa (Punttila 1998) refleksiiviverbejä TE+TTE+H sisältävät yksikön 3. persoonan preesensmuodot, esim. heitte+ä+te+h ovat mm. kiändäijä ’kääntyä’, pezeijä ’peseytyä’, ravahtaija ’kiirehtää, rientää’ ’heittäytyy’, laski+e+te+h ’laskeutuu’, issuttelˊi+e+te+h ’istuutuu vähän väliä’ ja varustaija ’varustautua, valmistautua’, venäläislainoja taas esim. duumaija (KKS s.v.) (Koivisto 1995: 84). ’tuumata, ajatella’, igruija ’ilvehtiä, leikkiä, pelehtiä’, voruija ’olla vallaton, telmiä; 27. Aineistopohjaisen tarkastelun ulkopuolelle tässä artikkelissa jäävästä Ilo­mant­ varastella’, fatieroija ’asua’ ja plotnikoija ’veistää, olla kirvesmiehenä’. sista ei ole KKS:ssa lainkaan refleksiivitaivutustietoja.

288 289 VESA KOIVISTO RAJAKARJALAISMURTEIDEN REFLEKSIIVIVERBEISTÄ refleksiivitaivutuslekseemin sana-artikkelissa (näiden hakusana on mukitta+i+čč+i ’kaatui’, varusta+i+čč+i ’varustautui’ (SKNA AijA-loppuinen, esim. jaksaija, murtaija, hölmistäijä). Näistä verbeis­ 0656:1b). (Ks. AhtA-johtimisista verbeistä erikseen jäljempää.) tä seitsemän ei esiinny rajakarjalaisaineistossani. Mainittujen kahden johtimen yhtymää *-δe- + -čče- (TE+CCE) Myös Genetzillä (1884) on joitakin refleksiivitaivutustietoja tavataan myös tytärkarjalaisissa murteissa, joissa muotojen vokaa­ Impilahdesta (merkitty ”Impil.” tai ”Kid.”); näistä neljästä kolme ei listo kuitenkin poikkeaa Impilahden ja Suistamon rajakarjalaismur­ esiinny omassa aineistossani. Lisäksi tekstikokoelman Karjalan kie- teista: čče-johtimen edellä vastaa äänneasua ei (ei rodˊeiče), joka on len näytteitä II Impilahden murteen näyte (E. Leskinen 1934: 92–101) suomen murteiden mukainen, tytärkarjalassa ie (ei rodˊieče); samoin sisältää 15 refleksiivitaivutusesiintymää (kahdeksasta eri verbistä). rajakarjalan oi (istoičen) – tytärkarjalan uo (istuočen); rajakarjalan Laajempi litteroitu kielennäyte on Punttila 1992, joka on litteroitu ai (virutaittšimmo) – tytärkarjalan oa ~ ua (viruttuattši; KKS); raja­ SKNA:n kahdesta nauhatunnista; siinä on viisi refleksiivitaivutus­ karjalan äi (vedäiččöö) – tytärkarjalan eä ~ iä (v́ edˊiätšöu; KKS). Ra­ esiintymää (kolmesta eri verbistä). Edelleen Impilahden murteen sa­ jakarjalaismurteista Suojärvellä ja Salmissa nämä diftongit esiinty­ nakirja (Punttila 1998) sisältää 28 refleksiivitaivutuslekseemiä, joista vät sen sijaan samassa säännönmukaisessa karjalaisessa asussa kuin 12:ta ei tavata missään muussa lähteessä. Kaikkiaan Impilahdesta on tytärkarjalassa. kertynyt esiintymätietoja rajakarjalaisaineistostani sekä tässä maini­ Toisena erona tytär- ja rajakarjalaisten murteiden välillä on tuista lisälähteistä 32 refleksiivitaivutuslekseemistä. se, että tytärkarjalaisissa murteissa johtimisto *-δe- + -čče- esiintyy Suistamon refleksiivitaivutuksesta puolestaan on KKS:ssa esiin­ kaikissa refleksiivimuodoissa läpi paradigman, kun taas Impilahdel­ tymätietoja 18:ssa hakusanassa (joista kymmenen on kseh- ja kah­ la ja Suistamolla on osassa paradigman muotoja johtimena pelkkä deksan VijA-loppuisia, esim. lyymisteäkseh, veteäkseh; istoija, vi- -i- (< *-δe-), esim. Impilahti jaksa+i+jja ’riisuutua’ (E. Leskinen ruttaija). Lisäksi Suistamolta on näyte Karjalan kielen näytteitä II 1934: 95), puikkelda+i+t+i+i ’puikkelehdittiin’ (SKNA 6818:2a), ei -tekstikokoelmassa (E. Leskinen 1934: 103–112). Kaikkiaan Suista­ lask+i+t+a ’ei laskeuduta’ (SKNA 6836:2a), Suistamo viruta+i+ja molta on rajakarjalaisaineistossani sekä lisälähteissä tietoja yhteensä ’loikoa’ (KKSA), anta+i+maa ’antautumaan’ (SKNA 0656:1b), rou- 42 refleksiivitaivutuslekseemistä. va oli vierel isto+i+nuh ’istuutunut’ (SKNA 3043:2a). Nämä reflek­ Impilahden ja Suistamon refleksiiviparadigma rakentuu paljolti siivimuodot ovat samanlaisia kuin maantieteellisesti läheisissä suo­ johtimille. Tyypillinen on johdinaines -čče- (CCE), jota edeltää sanan­ men kaakkoismurteissa (esim. pesse+i+jjä, lasse+i+jes; ks. Koivisto muodoissa johdin *-δe- (TE) (joka edustuu verbivartalon loppuvokaalia­ 1995: 90–91). Tytärkarjalaisissa murteissa refleksiiviverbien (*δe- +) seuraavana diftongin jälkivokaalina), esim. Impilahti­ ista+i+če+n čče-johdin on siis levittäytynyt koko refleksiiviparadigmaan, kun taas ’istuudun’ (SKNA 6855:1b), virutta+i+čče+t ’asetut makaamaan’ rajakarjalaismurteissa kehitys ei ole yhtä pitkällä, joskin sama tendens­ (SKNA 7440:1a,b), kiändä+i+č+i+n ’käännyin’ (SKNA 6855:1b), si näkyy aineistossa hyvin. Muualta karjalan puhuma-alueelta vastaa­ Suistamo isto+i+tše+n ’istuudun’, šuore+i+tše+n ’pukeudun’ (KKS), vanlaista paradigman johdostumiskehitystä ei tunneta (ks. Anttikoski viruta+i+ttš+immo (KKS s.v. viruttaija), käperdä+i+č+in ’käperryin’ 2003: 32), ja čče-johtimen leviäminen refleksiiviparadigmaan on alun (SKNA 0656:1b). Johdosmuotoja on myös yks. 3. persoonassa: Impi­ perin selvästikin ollut rajakarjalainen kielenpiirre. Paikoin varsinais­ lahti vedä+i+čč+öö ’vetäytyy’ (SKNA 7440:1a), anda+i+ččo+v ’so­ karjalassa, etenkin eteläkarjalan murteissa, tavataan čče-johtimen pii’ (Genetz 1884: 36), heittä+i+čč-i ’heittäytyi’ (SKNA 7440:1a,b), varhempaa kehitysastetta johdinta -(t)te- (TTE; ks. Koivisto 1995: Suistamo tunnusta+i+ttšo+o ’ilmaisee tuntevansa’ (KKS s.v. tunnustai- 65–66, 82–84), joka esiintyy rajakarjalaisaineistossakin muutamas­ ja), viändä+i+ttšö+ö ’vääntyy’ (KKS s.v. veäntäijä), vuorota+i+ttš+i sa muodossa, mutta on täällä harvinainen čče-johtimeen verrattuna: ’(sade) lakkasi’ (KKS s.v. vuorottaija), kyrčistä+i+čč+i ’kyyristyi’, Impilahti šmuttˊi+e+tto+o (TE+TTE) ’pelleilee’ (SKNA 7439:2a,b),

290 291 VESA KOIVISTO RAJAKARJALAISMURTEIDEN REFLEKSIIVIVERBEISTÄ

Suistamo jotta ei röngähtä+tt+is tuleh kogo hauta (SKNA 0657:1a), Jonkinlaisia sekamuotoja, joissa yhdistyvät refleksiivinenU -joh­ Suojärvi midäbä siit roi+te+h, ei nimidä ei roi+te (SKNA 2268:2a).28 dinaines sekä refleksiivitaivutus, ovat Suistamon käpertä+i+čč+y+y Impilahdella ja Suistamolla vallitseva refleksiivitaivutuksen joh­ ([orava] mänöö ihan semmoseh paikkah käpertäiččyy erittäin dinaines on siis *δe- ja čče-johdinten yhtymä (TE+CCE). Sen ohella ńärepuus; SKNA 0656:1b) ja Impilahden taiv+u+i+d+i+i (niin ku paradigmassa tavataan edelleen refleksiivistä *hen-taivutustunnusta käydii lammas ottamaa ńi, siinä se taivuidii; SKNA 6821:1a) ja mah­ alkusijoillaan yksikön 3. persoonan imperfektimuodoissa: Impilahti dollisesti purista+u+d+i ’puristautui’ (SKNA 6853:1b). roi+h ’syntyi, tuli’ (E. Leskinen 1934: 92, 95, 101), loppi+h ’loppui’ Rajakarjalaismurteissa voivat esiintyä refleksiivitaivutteisina – ku­ (SKNA 6853:1a), virutti+h ’viruttautui, asettui makaamaan’ (SKNA ten muuallakin karjalassa – myös AhtA-johdosverbit (ks. Genetz 1870: 6855:1a), Suistamo loppi+h ’loppui’ (SKNA 3043:1a), roi+h ’tuli’ 231). Tällöin refleksiivitunnus ei muuta AhtA-verbin merkitystä, eikä (SKNA 3043:1a), vedi+h ’vetäytyi’ (KKS s.v. veteäkseh) (myös AhtA- refleksiivitunnuksen esiintyminen ole myöskään säännöllistä. ­ Imper johtimisissa verbeissä, joista ks. jäljempää). Joskus loppu-h:ta vastaa fektissä esiintyy alkuperäinen yksikön 3. persoonan taivutustunnus -h pidentynyt vokaali: Impilahti laški+i ’laskeutui’ (SKNA 6836:2a), (< *-hen): esim. Impilahti napsahti+h ’napsahti’ (SKNA 6855:1a), Suis­ ta leibähä loppi+i ’ja leipähän loppui’ (SKNA 6855:1a), Suistamo tamo no miulen, ńaba ńulahti+h ’luiskahti’ (SKNA 0657:1b), sit siel aina pyrähti+i (SKNA 0656:1b). Lisäksi Suistamolta on tieto persoonapäät­ pyrähti+i huonteksil tetrikarja (SKNA 0656:1b), samoin preesensissä, teestä -hes preesensissä: suoh upotta+hes ’suohon upottautuu, uppoaa’ esim. Impilahti kurahta+h, murahta+h (E. Leskinen 1934: 99), Suistamo (Nenonen & Rajamo 1955: 328). Aineslähteiden valossa kuitenkin jorahta+h, ähkähtä+h (SKNA 1625:2b).29 Muilta osin AhtA-verbeissäkin čče-johdin näyttää levittäytyneen vahvasti myös 3:nteen persoonaan on vallitsevana čče-paradigma (TE+CCE): Impilahti hengähtä+i+če+n taivutustunnuksen sijalle (niin imperfektissä kuin preesensissäkin; ’hengähdän’ (SKNA 7481:1b), röngähtä+i+č+i+mmö ’karjahdimme, ks. esimerkkejä edellä). – KKS:ssa on Suistamolta myös yksittäinen rääkäisimme’ (SKNA 6836:2a), lomahta+i+tš+i+n ’kaaduin, romah­ monikon 1. persoonan refleksiivimuoto, joka sisältää taivutussuffiksin din’ (Punttila 1998 s.v.), Suistamo ryöpsähtä+i+č+i+t sinne, järvee -mOš (*δe-johtimen jäljessä, siis TE+M): tšiepusta+i+moš ’tarrau­ ’ryöpsähdit’ (SKNA 0658:1a). Toisaalta čče-johdosparadigma on voinut dumme, kierrymme (toisiimme)’ (vrt. yleensä monikon 1. persoonassa edetä yksikön 3. persoonaankin: Impilahti ravahta+i+ttšo+o ’havah­ on TE+CCE-johtiminen muoto: viruta+i+ttš+i+mmo) (KKS). tuu’ (E. Leskinen 1934: 94), hairahta+i+čč+is ’hairahtuisi, erehtyisi’ Rajakarjalaisaineistossani Suistamon erikoisuutena ovat d:lliset (SKNA 6826:2a). Aineistossani Impilahden refleksiivitaivutteisissa passiivin imperfektin muodot, joita yleiskarjalaiseen tapaan käytetään AhtA-verbeissä vallitsee čče-paradigma yleisemmin (useita esiintymiä) myös monikon 3. persoonan muotoina. Näissä muodoissa odotuksen­ kuin Suistamon (josta vain yksi esiintymä). Mm. kielteisessä preesen­ mukaisen t:n sijalla esiintyvää d:tä edeltää refleksiivijohdin-i- (= TE) sissä ja A-infinitiivissä AhtA-verbien muodoissa esiintyy johdin *-δe- (< *-δe-): esim. Suistamo putši+i+dih ’puolustautuivat’ (Kalkkinen (TE): Impilahti­ veri ei tˊipahta+i ’veri ei tipu’ (SKNA 7440:1a,b), 1987: 44), jaksa+i+dih ’riisuutuivat’ (SKNA 1625:1a), pala+i+dii hitškahta+i+ja ’hiiskahtaa’ (E. Leskinen 1934: 95). ’palattiin’ (”palauduttiin”) (SKNA 0658:1a), varusta+i+dii ’varus­ tauduttiin’ (SKNA 0658:1a), jiäksi+i+dih ’näyttäytyivät’ (SKNA 5.2.2. Korpiselkä 0658:1b), hengästä+i+dii ’hengästyivät’ (SKNA 0658:1b) (vrt. vas­ taava t:llinen muoto puikkelda+i+tii ’puikkelehtivat’; Impilahti, Rajakarjalaisaineistossani on Korpiselän refleksiivitaivutuksesta 34 SKNA 6818:2a). Tällaisia muotoja ei löydy KKS:sta. esiintymää (nämä jokseenkin varmoja ja lisäksi 11 epävarmaa). Esiin­ tymiä on 11:stä eri lekseemistä, mutta Korpiselän refleksiivitaivutus­ 28. Genetz (1870: 233) esittää Suojärveltä yksikön kolmannen persoonan myön­ teiset preesensmuodot roih ~ roite ~ rodˊiete (ks. Ahtia 1938: 98). (Vrt. alaviite 23, 29. AhtA-verbin merkitys on usein sillä tavalla kontekstuaalinen, että sitä on vai­ jonka mukainen vastaava muoto olisi rodˊieteh.) kea täsmällisesti ilmaista.

292 293 VESA KOIVISTO RAJAKARJALAISMURTEIDEN REFLEKSIIVIVERBEISTÄ aineisto keskittyy vahvasti yhteen lekseemiin, verbiin rotiekseh ’syn­ 5.3. Yhteenvetoa rajakarjalaismurteiden tyä, tulla joksikin’, josta on 22 esiintymää eli yli kaksi kolmasosaa refleksiivitaivutuksesta Korpiselän kaikista refleksiivitaivutusesiintymistä. Aineistossani varsin moni Korpiselän refleksiivitaivutusesiinty­ Rajakarjalaismurteiden refleksiivitaivutuksessa on eroa pitäjänmurteiden mistä on jotenkin epäselvä tai tulkinnallinen; niissä on sekä muodon välillä – toisin kuin refleksiivijohdoksissa. Suistamo ja Impilahti muo­ että merkityksen epävarmuutta. Monilla refleksiivimuodon näköisillä dostavat rajakarjalaismurteiden refleksiivitaivutuksen yhtenäisen ydin­ (usein h- tai i-loppuisilla) verbinmuodoilla on lauseessa esim. objek­ alueen, jolla on selvät yhtäläisyydet tytärkarjalaisiin murteisiin. Molem­ ti, vaikka objektillisuus (verbin transitiivisuus) on refleksiivitaivutus­ missa paradigmaan ovat yleistyneet *δe- + čče-johtimiset muodot. Erona muodolle epäodotuksenmukaista ja poikkeuksellista (ks. kuitenkin näihin murteisiin on Raja-Karjalassa kuitenkin vajaampi paradigma sekä Koivisto 1995: 313–314, jossa esitellään karjalan objektillisia reflek­ refleksiivitaivutuksen selvästi vähäisempi käyttö, joka johtunee raja­ siivimuotoja 18 lekseemistä, valtaosin ele-johdoksista). karjalaismurteisiin­ tulleesta suomen kielen vaikutuksesta. Sama vaikutus Korpiselän refleksiivitaivutuksesta ei ole aiempaa dokumentoi­ selittää myös refleksiivitaivutuksen esiintymättömyyden Ilomantsissa. tua tietoa. KKS:ssa on yksi ainoa refleksiivitaivutushakusana, jossa on esiintymätieto Korpiselästä (s.v. alkoakseh), mutta sekin on sanan­ parresta. Muutoin Korpiselän murre on KKS:n aineistossa UtU-johdos­ 6. Lopuksi ten kannalla. Ainoasta kirjoitetusta kielennäytteestä (E. Leskinen 1934: 139–141) ei löydy refleksiivitaivutusesiintymiä, eikä muita lähteitä ole. Rajakarjalaismurteissa refleksiivijohdokset ovat käytössä yleisempi reflek­ Aineistossani Korpiselän refleksiivitaivutuksesta on eniten yk­ siivityyppi kuin refleksiivitaivutus. Refleksiivitaivutusta tavataan näissä sikön 3. persoonan muotojen esiintymiä. Yleensä ne ovat imperfekti­ murteissa harvemmin kuin muualla etelä- tai livvinkarjalan murteissa (ja muotoja, joissa refleksiivitunnuksena on-h (< *-hen): kiärildi+h ’kää­ Ilomantsissa ei lainkaan), mitä voidaan pitää suomen kielen vaikutuksena. riytyi’ (SKNA 3040:1a), loppi+h ’loppui’ (SKNA 3040:2a), karhu Rajakarjalaismurteiden refleksiiviverbeissä on paljon suomalai­ siitä röhkähti+h (SKNA 4383:1a), kun otti+he eukon ’kun otti itsel­ sia lainoja ja suomen kielen vaikutusta. Tietyt harvemmin esiintyvät leen eukon’ (SKNA 3041:1a,b). kompleksiset­ refleksiivijohdostyypit ovat kokonaisuudessaan, johdos­ Aineiston valtaverbistä rotiekseh yksikön 3. persoonan imperfekti­ tyyppeinä, suomalaislainaa (luku 4.1.5.). Näiden harvinaisten tyyppien muoto ei kuitenkaan ole h-loppuinen, vaan siinä tavataan i-vokaalin pi­ havaitseminen ja niiden rajakarjalaisen levikin selvittäminen on vaatinut dentymää: rodii ~ rodˊii ~ rotˊii. Lisäksi imperfektimuotona on rodie riittävän laajan aineiston. Suomen kielen vaikutuksen paljastamiseen on (yhdellä puhujalla jopa yksinomaisena). Tämän jälkimmäisen imper­ tarvittu myös vertailutietoa Venäjällä puhutusta karjalan kielestä. fektimuodon asulle ei löydy vastinetta muualta karjalasta. KKS:n ai­ Tuoreemman suomalaisvaikutuksen alla rajakarjalaismurteiden neksessa muoto rodˊie on aina konnegatiivi (ei rodˊie ’ei synny, ei tule’). refleksiiviverbistössä on (muinais)karjalainen pohja. Leksikaaliselta Korpiselän refleksiivitaivutus vaikuttaa labiililta ja reliktimäisel­ koostumukseltaan rajakarjalaismurteiden refleksiiviverbistö on suu­ tä. Sen naapurimurteesta Ilomantsista ei ole rajakarjalaisaineistossani relta osin karjalan mukainen. Venäjästä lainattuja refleksiiviverbejä­ tai eikä KKS:ssa esiintymiä refleksiivitaivutuksesta. Myöskään naapu­ venäjän mallin mukaista refleksiiviverbien käyttöä rajakarjalaismur­ ri- ja kontaktikielessä suomen Pohjois-Karjalan murteissa refleksiivi­ teissa kuitenkin esiintyy vähemmän kuin muualla karjalassa. Aineis­ taivutusta ei tavata. tossa esiintyviä suomesta poikkeavia venäjän mukaisia refleksiivi­ verbien merkityslajeja ovat passiivinen (reššivyö ’tulla päätetyksi, ratkaistuksi’), resiprookkinen­ (näkieččie ’nähdä toisensa, tavata’) ja datiivis-refleksiivinen (ottihe ’otti itselleen’) (ks. Koivisto 1995: 48).

294 295 VESA KOIVISTO RAJAKARJALAISMURTEIDEN REFLEKSIIVIVERBEISTÄ

Rajakarjalaismurteissa refleksiivijohdosten inkoatiivinen käyttö Hakulinen, Lauri 1979: Suomen kielen rakenne ja kehitys. Neljäs, korjattu ja (luku 4.1.4.) on sellainen karjalan kielen piirre, joka on suomelle jok­ lisätty painos. Helsinki: Otava. seenkin vieras. Rajakarjalaismurteiden refleksiiviverbeistä erottuvat Hämynen, Tapio 2012: Changes in the linguistic identity of the borderland selvästi niin niiden karjalainen kuin myöhempi suomalainenkin kom­ Karelians in Finland up to the year 2009. – Tapio Hämynen & Alek­ sander Paskov (toim.), Nation split by the border. Changes in the ethnic ponentti. Lisäksi on yhteisiin muinaiskarjalaisiin juuriin palautuvaa identity, religion and language of the Karelians from 1809 to 2009. vanhempaa ainesta (esim. U-johdoksissa). [Joensuu]: UPEF. 246–271. Livvinkarjalaista vU-johdostyyppiä tavataan koko Raja-Karja­ ISK = Hakulinen & al. 2004. lassa, ja vastaavasti varsinaiskarjalaista UtU-johdostyyppiä esiintyy Itkonen, Terho 1969: Muoto-opin keruuopas. Laatinut Terho Itkonen apu­ myös Raja-Karjalan livvinkarjalaisissa murteissa. Rajakarjalaismur­ naan Heikki Leskinen, Heikki Paunonen ja Tapani Lehtinen. Tieto­lipas teet edustavat etelä- ja livvinkarjalan leikkausaluetta, jossa kummankin 59. Helsinki: SKS. piirteet kohtaavat. Ylipäätään Raja-Karjalassa on sellaisia livvinkarja­ Jeskanen, Matti 2005: Karjalan kieli ja karjalankieliset Suomessa. – Marjatta Palander & Anne-Maria Nupponen (toim.), Monenlaiset karjalaiset. laisia piirteitä, jotka ulottuvat yli alueen koko eteläkarjalaismurteisen­ Suomen karjalaisten kielellinen identiteetti. Studia Carelica Humanis­ osan. Tässä saattaa näkyä rajakarjalaismurteiden yleinen kehityshis­ tica 20. Joensuu. 214–285. toria ja Raja-Karjalan asutushistoria, joskin historiallisten kontakti­ — 2011: Karjalan kieli Suojärvellä. – Tapio Hämynen (toim.), Omal mual, tilanteiden ja asutuksen liikkeiden heijastumat nykymurteisiin ovat vierahal mual. Suojärven historia III. Nurmes: Suojärven Pitäjäseura. kaikkiaan haastavaa selviteltävää. 352–357. Kalkkinen, Aune (toim.) 1987: Tälleh siel elettih. Aleksanteri Saralan tallentamaa muistitietoa Suistamon Koitosta. Turku: Karjalaisen kulttuurin säätiö. Lähteet KKS = Karjalan kielen sanakirja 1–6. Päätoimittajat Pertti Virtaranta & Ahtia, E. V. 1938: Karjalan kielioppi. Äänne- ja sanaoppi. Suojärvi. Raija Koponen. Lexica Societatis Fenno-Ugricae XVI, 1–6. Kotimais­ Anttikoski, Esa 2003: The problem of dialectal differences in the creation ten kielten tutkimuskeskuksen julkaisuja 25. Helsinki 1968, 1974, 1983, of a unified Karelian literary language: the experience of the 1930s. – 1993, 1997, 2005. Esa Anttikoski (toim.), Developing written Karelian. Papers from the KKSA = Karjalan kielen sana-arkisto. Joensuu: Itä-Suomen yliopisto. Karelian session of the 11th International Conference on Methods of Koivisto, Vesa 1989a: Itämerensuomen refleksiivitaivutuksen ja mediaalitai­ Dialectology. Studies in Languages 38. Joensuu: University of Joensuu, vutuksen suhteesta. – Virittäjä 93: 102–110. Saatavissa: ‹https://journal. Faculty of Humanities. 29–36. fi/virittaja/article/view/38271› Bubrih, D. V., A. A. Beljakov & A. V. Punžina 1997: Karjalan kielen mur- — 1989b: Itämerensuomen refleksiiviverbien semantiikkaa: karjalan ja suo- rekartasto. Dialektologitšeski atlas karelskogo jazyka. Toim. Leena men itämurteiden refleksiivitaivutus. Lisensiaatintyö. Helsingin yliopisto. Sarvas. Venäjän tiedeakatemian Karjalan tiedekeskuksen kielen, kir­ — 1991: Suomen verbikantaisten UtU-verbijohdosten semantiikkaa. jallisuuden ja historian instituutti & Kotimaisten kielten tutkimuskes­ Suomi 161. Helsinki: SKS. kus. Kotimaisten kielten tutkimuskeskuksen julkaisuja 97. Helsinki: — 1993: Lyydin ze-refleksiivijohdosten synnyn ongelma. – Virittäjä 97: Suomalais-Ugrilainen Seura. 187–202. Saatavissa: ‹https://journal.fi/virittaja/article/view/38573› Genetz, Arvid 1870: Kertomus Suojärven pitäjäästä ja matkustuksistani — 1995: Itämerensuomen refleksiivit. SKST 622. Helsinki: SKS. siellä v. 1867. – Suomi II: 8: 201–275. — 2013: Suomen sanojen rakenne. Suomi 202. Helsinki: SKS. — 1880: Tutkimus Venäjän Karjalan kielestä. Kielennäytteitä, sanakirja ja — 2018: Border Karelian dialects – a diffuse variety of Karelian. – Marjatta kielioppi. – Suomi II: 14. Palander, Helka Riionheimo & Vesa Koivisto (eds), On the Border of Lan- — 1884: Tutkimus Aunuksen kielestä. – Suomi II: 17. guage and Dialect. Studia Fennica Linguistica 21. Helsinki: Finnish Lite­ Hakulinen, Auli, Maria Vilkuna, Riitta Korhonen, Vesa Koivisto, Tarja rature Society. 56–84. Saatavissa: ‹http://dx.doi.org/10.21435/sflin.21› Riitta Heinonen & Irja Alho 2004: Iso suomen kielioppi. SKST 950. Kulonen, Ulla-Maija 2010: Fonesteemit ja sananmuodostus. Suomen konti- Helsinki: SKS. nuatiivisten U-verbijohdosten historiaa. Suomi 197. Helsinki: SKS. 296 297 VESA KOIVISTO RAJAKARJALAISMURTEIDEN REFLEKSIIVIVERBEISTÄ

Kulonen-Korhonen, Ulla 1985: Deverbaalisten U-verbijohdosten semantiik­ Pugh, Stefan M. 1992: Baltic Finnic verbal derivation and the assimilation of kaa. – Virittäjä 89: 290–309. Saatavissa: ‹https://journal.fi/virittaja/ Russian lexemes. – Linguistica Uralica 2/1992: 89–101. article/view/37949› Punttila, Matti 1992: Impilahden karjalaa. Castrenianumin toimitteita 41. Koti­ Kuujo, Erkki 1963: Raja-Karjala Ruotsin vallan aikana. Joensuu: Karjalai­ maisten kielten tutkimuskeskuksen julkaisuja 64. Helsinki: Yliopistopaino. sen Kulttuurin Edistämissäätiö. — 1998: Impilahden karjalan sanakirja. Lexica Societatis Fenno-Ugricae Laakso, Johanna 1990: Translatiivinen verbinjohdin NE itämerensuomalai- XXVII. Helsinki: Suomalais-Ugrilainen Seura & Kotimaisten kielten sissa kielissä. SUST 204. Helsinki. tutkimuskeskus. Laanest, Arvo 1975: Sissejuhatus läänemeresoome keeltesse. Tallinn: Eesti Pyöli, Raija 1996: Venäläistyvä aunuksenkarjala. Kielenulkoiset ja -sisäiset NSV Teaduste Akadeemia Keele ja Kirjanduse Instituut. indikaattorit kielenvaihtotilanteessa. Joensuun yliopiston humanistisia Lehtinen, Tapani 1985: Venäjän verbisemantiikkaa itämerensuomessa: kar­ julkaisuja 18. Joensuu: Joensuun yliopisto. jalan ja vepsän inkoatiiviverbit. – Sananjalka 27: 63–83. Saatavissa: — 2012: Livvinkarjalan kielioppi. Helsinki: Karjalan Kielen Seura. ‹https://doi.org/10.30673/sja.86480› Rjagojev, V. D. 1980: Obraztsy karelskoi retši (tihvinski govor sobstvenno Leskinen, Eino 1932: Karjalan kielen näytteitä I. Tverin ja Novgorodin kar- karelskogo dialekta). Leningrad: Nauka. jalaa. SKST 193. Helsinki: SKS. Ruoppila, Veikko 1956: Etelä-Karjalan murreopas. Helsinki: Otava. — 1934: Karjalan kielen näytteitä II. Aunuksen ja Raja-Karjalan mur- Räisänen, Alpo 1967: Kainuun murteiden sekautua ja sekauta -tyyppiset teita. SKST 193. Helsinki: SKS. verbit. – Virittäjä 71: 99–107. Saatavissa: ‹https://journal.fi/virittaja/ Leskinen, Heikki 1998: Karjala ja karjalaiset kielentutkimuksen näkökul­ article/view/35232› masta. – Pekka Nevalainen & Hannes Sihvo (toim.), Karjala. Historia, — 1985: Suomen kielen ne- ja ntu-, nty-johtimiset translatiiviverbit. – Virit- kansa, kulttuuri. Helsinki: SKS. 352–382. täjä 89: 1–32. Saatavissa: ‹https://journal.fi/virittaja/article/view/37908› Markianova, L. F. 1985: Glagolnoje slovoobrazovanije v karelskom jazyke. — 1988: Suomen kielen u-johtimiset verbit. Suomi 141. Helsinki: SKS. Petrozavodsk: Karelija. Sarhimaa, Anneli 1999: Syntactic transfer, contact-induced change, and the Moshnikov, Ilia 2014: NUT-partisiipin variaatio Ilomantsin rajakarjalais- evolution of bilingual mixed codes. Focus on Karelian-Russian langu- murteessa. Pro gradu -tutkielma. Suomen kieli. Itä-Suomen yliopisto. age alternation. Studia Fennica Linguistica 9. Helsinki: SKS. Saatavissa: ‹http://urn.fi/urn:nbn:fi:uef-20140271› SKNA = Suomen kielen nauhoitearkisto. Helsinki: Kotimaisten kielten keskus. Nenonen, Väinö & Mikael Rajamo (toim.) 1955: Suistamo. Muistelmia ja SMSA = Suomen murteiden sana-arkisto. Helsinki: Kotimaisten kielten keskus. kuvia kotiseudustamme. Pieksämäki: Suistamon pitäjänseura. Tavi, Susanna 2015: Rajakarjalaismurteiden venäläiset lainasanat. Pro gradu Nirvi, R. E. 1961: Inkeroismurteiden asema. – Kalevalaseuran vuosikirja 41: -tutkielma. Suomen kieli. Itä-Suomen yliopisto. Saatavissa: ‹http://urn. 99–132. fi/urn:nbn:fi:uef-20160039› Ojansuu, Heikki 1910: Karjalan (ja Aunuksen) kielestä. – Iivo Härkönen Turunen, Aimo 1946: Lyydiläismurteiden äännehistoria I. Konsonantit. (toim.), Karjalan kirja. Porvoo: WSOY. 10–16. SUST 89. Helsinki. Palander, Marjatta, Helka Riionheimo, Hannu Kemppanen & Jukka Mäkisalo — 1965: Suojärven murre. – Lauri Pelkonen (toim.), Suojärvi I. Pieksä­ 2019: Kielikorpuksia Suomen itärajalta. – Sofia Björklöf & Santra Jantu­ mäki: Suosäätiö. 21–38. nen (toim.), Multi­lingual Finnic. Language contact and change. Uralica — 1973: Raja-Karjalan murteet ja vepsän kieli. – Hannes Sihvo (toim.), Karjala. Helsingiensia 14. Helsinki: Suomalais-Ugrilainen Seura. 425–438. Idän ja lännen silta. Kalevalaseuran vuosikirja 53. Porvoo: WSOY. 83–94. Saatavissa: ‹https://doi.org/10.33341/uh.85045› — 1982: Raja-Karjalan murteet. – Karjala 2. Karjalan maisema ja luonto. Palander, Marjatta, Pekka Zaikov & Milla Uusitupa 2013: Karjalan kielen 65–89. tutkimusta ja opetusta kahden puolen rajaa. – Pekka Suutari (toim.), Uusitupa, Milla 2011: Avoimet persoonaviittaukset rajakarjalaismurteissa. Karjala-kuvaa rakentamassa. SKST 1389. Helsinki: SKS. 358–385. Pro gradu -tutkielma. Suomen kieli. Itä-Suomen yliopisto. Saatavissa: Palmeos, P. 1962: Karjala Valdai murrak. Emakeele Seltsi toimetised 5. Tal­ ‹http://urn.fi/urn:nbn:fi:uef-20110443› linn: Eesti NSV Teaduste Akadeemia. Uusitupa, Milla, Vesa Koivisto & Marjatta Palander 2017: Raja-Karjalan mur­ Posti, Lauri 1980: The origin and development of the reflexive conjugation teet ja raja-alueiden kielimuotojen nimitykset. – Virittäjä 121: 67–106. in the Finnic languages. – Osmo Ikola (toim.), Congressus Quintum Saatavissa: ‹https://journal.fi/virittaja/article/view/53121› Internationalis Fenno-Ugristarum I. Turku. 111–144. Virtaranta, Pertti 1967: Lähisukukielten lukemisto. SKST 287. Helsinki: SKS. 298 299 VESA KOIVISTO

On reflexive verbs in Border Karelian dialects

Vesa Koivisto

This article deals with reflexive verbs in Border Karelian dialects, a variety of the spoken in eastern Finland before World War II. Border Karelian dialects represent two main dialects of Karelian, South and Olonec Karelian, which also are spoken in Rus­ sia. Border Karelia was ceded to the in 1944 and its en­ tire population was evacuated to the remaining territories of Finland. This article is based on a linguistic corpus representative of Border II Karelian dialects as present in audio recordings made in the 1960s and Reports 1970s. Border Karelia was an area of diverse language contacts. Rus­ sian has been a contact language for Karelian for centuries. In Border Karelia, there were also contacts between the Karelian and Finnish languages since the 17th century. Border Karelian dialects can be de­ scribed as a Finnicized variant of Karelian. In this article, Border Ka­ relian dialects are compared to the Finnish language and to the Kare­ lian dialects spoken in Russia. This article aims to describe the extent of Finnish influence on Border Karelian reflexive verbs. Morphologically, a is derived from a transitive non-reflexive base verb. In Karelian, there are two main types of re­ flexives: derivatives and the reflexive conjugation. In Border Karelian, mostly reflexive derivatives are used, and the reflexive conjugation is used to a lesser extent than in the Karelian dialects spoken in Russia. In Border Karelian, there are also less verbal loans from Russian. Reflexives in Border Karelian mainly represent two derivative types: the types containing the morpheme -UdU- ~ -UtU- and the morpheme -vU-. In Karelian, in general, the distribution of these two derivative types is areally complementary: -UdU- ~ -UtU- in South (and North) Karelian and -vU- in Olonec Karelian. In Border Karelian, their distribution, however, overlaps. There are also some morpho­ logical derivative types in Border Karelian that have been loaned from Finnish and are not used elsewhere in Karelian.

300 FEDOR ROZHANSKIY University of Tartu & Institute for Linguistic Studies RAS

ELENA MARKUS University of Tartu & Institute of Linguistics RAS

A new resource for Finnic languages: The outcomes of the Ingrian documentation project

Abstract The report introduces a new digital resource on minor Finnic languages. This resource is the main outcome of the project “Documentation of Ingrian: collecting and analyzing fieldwork data and digitizing legacy materials” carried outby Fedor Rozhanskiy and Elena Markus at the University of Tartu in 2011–2013. The collected materials cover several minor Finnic languages with a special focus on varieties spoken in Western Ingria: the Soikkola, Lower Luga, and Heva dialects of Ingrian, the Lower Luga varieties of Votic, and Ingrian Finnish. The re­ source contains (a) legacy recordings of different genres made by previous researchers in 1968–2012; (b) new audio and video materials recorded mostly in 2011–2013 by the project partici­ pants; (c) transcriptions and translations into Russian and English synchronized with sound and video using the ELAN software. Altogether the resource presents 510 hours of audio recordings, 21 hours of video recordings, and 15 hours of ELAN annota­ tions. All media files in the resource are provided with detailed metadata specifying the place and time of the recording, sociolin­ guistic data about the speaker, the contents of the recording, and the access rights. The resource is available on the websites of the Endangered Languages Archive (London, UK) and the Archive of Estonian Dialects and Kindred Languages of the University of Tartu (Estonia).

Multilingual­ Finnic. Language contact and change. 303–326. Uralica Helsingiensia 14. Helsinki 2019. ‹https://doi.org/10.33341/uh.85039› FEDOR ROZHANSKIY & ELENA MARKUS A NEW RESOURCE FOR FINNIC LANGUAGES: THE OUTCOMES OF THE INGRIAN DOCUMENTATION PROJECT

1. Introduction stored in institutional archives, are fully digitized, and are already provided with metadata. Many recordings that we managed to col­ This paper reports on the project “Documentation of Ingrian: collect­ lect were stored on magnetic media (cassettes or tapes) and risked ing and analyzing fieldwork data and digitizing legacy materials” car­ being lost or damaged. Most of the collections were not provided ried out by Fedor Rozhanskiy and Elena Markus at the University of with metadata and therefore had low research value (the information Tartu in 2011–2013. The project was financed by the Endangered Lan­ on the contents, time, and place of the recording is essential for most guages Documentation Programme housed at the School of Oriental tasks). and African Studies of the University of London. Recording of new field materials was aimed not only at expand­ The main project outcome is a new digital resource on minor ing the amount of data but also at filling gaps in previously collected Finnic languages. Initially, the work was focused on the documenta­ materials. In particular, one of our tasks was to provide high quality tion and description of the Ingrian language, but in the course of the recordings that can be used for various types of linguistic work includ­ project the scope was broadened to include other Finnic varieties spo­ ing experimental phonetic research. Also, we tried to fill the genre ken in the area, the first of which were the Votic language and Ingrian gaps and record both spontaneous speech and elicitations. Legacy Finnish. The amount of collected data exceeded all expectations and collections often present samples of spontaneous speech with a nar­ for this reason language documentation became the main project focus row focus on a particular topic, and prove a rather limited data source while descriptive tasks were partially postponed for future work. The for many grammatical and phonetic questions. Additionally, we did resulting resource contains materials on minor Finnic languages with video linguistic documentation that was totally missing in previous a special focus on varieties spoken in Western Ingria. collections. This paper is organized in the following way. Section 2 lists the Summing up, the main goals of the project were the following: main goals and tasks of the project. Section 3 gives some informa­ tion on Ingrian and related languages. Section 4 addresses the pro­ 1) To track down, digitize and prepare for archiving potentially ject methodology. Section 5 describes the resulting data resource: the endangered collections of audio recordings; collections, organization of metadata, archiving institutions, etc. Sec­ 2) To make contemporary field recordings of different genres: nar­ tion 6 discusses further prospective work and research tasks based on ratives on various subjects, dialogues, elicitations of grammar, the collected materials. and phonetic questionnaires; 3) To provide video linguistic documentation; 4) To create a corpus of transcribed and translated recordings syn­ 2. Project goals chronized with sound (and video); 5) To provide a data basis for compiling dictionaries, , The main goal of the project was to create a language resource with and annotated text collections. various types of data. A major part of this resource consists of audio recordings. Two main sources of materials were 1) legacy recordings made by previous researchers and 2) new field materials recorded in the course of the project by the project participants. The task of collecting legacy materials was aimed both at pre­ serving the highly valuable data and making them accessible for re­ searchers. For obvious reasons we left out legacy recordings that are

304 305 FEDOR ROZHANSKIY & ELENA MARKUS A NEW RESOURCE FOR FINNIC LANGUAGES: THE OUTCOMES OF THE INGRIAN DOCUMENTATION PROJECT

3. Ingrian and related languages From a linguistic point of view, the Ingrian language is valu­ able not only as contributing to the general picture of the history and Ingrian belongs to the northern group of Finnic languages. At pre­ development of the Finnic family, but also as possessing some typo­ sent it comprises two dialects, Soikkola and Lower Luga, which differ logically rare features. In particular, the Soikkola dialect demonstrates significantly from each other.1 Lower Luga Ingrian is a convergent a ternary contrast of consonant lengths, while the Lower Luga dialect variety with a high degree of subdialectal variation (Rozhanskiy & has voiceless vowel phonemes. Markus 2013b, 2014a) spoken along the lower course of the Luga Ingrian is perhaps the least studied Finnic language. Porkka’s River. Soikkola Ingrian is spoken on the Soikkola peninsula. Both (1885) grammatical description and a school grammar by Junus (1936) dialects are located in the region of the Leningrad oblast are obviously outdated and incomprehensive, and do not correspond (Russian Federation). to modern standards of language description. Works by Laanest (1966, By our estimation, the total number of Ingrian speakers is now 1978, 1986) mostly address the historical development of Ingrian and less than 20 people. The average age of the speakers is more than dialect variation. There is no synchronic description of Ingrian that 80 years. In the majority of cases, middle-aged Ingrians either do not would reflect the contemporary state of the language. It is therefore know the language at all, or their language competence is limited to essential to document the language while it is still spoken and thus a few sentences on everyday topics. Children neither speak, nor un­ create a basis for future language description. derstand Ingrian. All speakers are bilingual, and for most of them the Ingrian lexicography is in a slightly better condition. There is main language of communication is Russian. Until recently, some flu­ a dictionary by Nirvi (1971) that represents all Ingrian dialects, al­ ent speakers used their native language in everyday communication, though the data on Lower Luga Ingrian are very limited2. There is also mostly to talk with neighbors or relatives, but by now communication a dictionary of the already extinct Heva dialect by Laanest (1997). in Ingrian is almost gone. In addition to the Ingrian language, the two languages that con­ The Ingrian written variety was introduced in the 1930’s. School stitute a considerable part of the audio materials in the resource are education in Ingrian lasted for several years, but in the beginning of the Votic language and the Ingrian dialects of Finnish (see Table 1 in 1938, Ingrian was banned from schools and teachers were repressed. Section 5 for a full list of languages represented). At present, the language prestige of Ingrian is very low. Most people The Votic language is on the verge of extinction. No more than see no benefit in studying Ingrian and only few native Ingrians show five people can currently be considered fluent speakers. They repre­ any interest in their ancestors’ language. Ingrian language courses are sent the Liivtšülä-Luuditsa and the Jõgõperä varieties of the Western taught by Nikita Diachkov at the Ingrian museum in Vistino, but these Votic dialect3. Both varieties are located in the Kingisepp region. The courses do not involve fluent speakers. Most speakers are too old and Votic language has never had a written variety. In the 1930s Votic have health problems, so they cannot participate in the courses. Some children attended schools in Ingrian together with their Ingrian neigh­ speakers are very skeptical about the new learners. Skepticism comes bors. The Jõgõperä and Liivtšülä-Luuditsa villages had a mixed Votic- either from a monopolistic attitude (“We have suffered but preserved Ingrian population. For that reason both varieties experienced Ingrian the language, but where have you been all that time? Now let us die peacefully with our language.”), or from the idea that new learners 2. A highly specific group of southern varieties of Lower Luga Ingrian is not make many mistakes and are not able to speak a “pure language” (see represented at all, see Muslimov 2005 regarding Ingrian dialectology. more details in Rozhanskiy & Markus 2013a: 270, 294). 3. It is not certain whether there are any fluent speakers of the Kukkuzi variety, a mixed Votic-Ingrian variety (see Suhonen 1985; Markus & Rozhanskiy 2012) traditionally listed as a Votic dialect. The last fluent speaker that we worked with 1. Two other Ingrian dialects – Heva and Oredeži – are already extinct. passed away several years ago.

306 307 FEDOR ROZHANSKIY & ELENA MARKUS A NEW RESOURCE FOR FINNIC LANGUAGES: THE OUTCOMES OF THE INGRIAN DOCUMENTATION PROJECT contact influence, but the degree of influence differs depending both 4. Project methodology on the variety and the concrete speaker. The language was described in five grammars (Ahlqvist 1856; Ariste 19684; Agranat 2007; Tsvet­ An audio corpus of legacy recordings was compiled in the following way. kov 2008 [the original manuscript from 1922]; Markus & Rozhanskiy 2017), and four dictionaries (Posti 19805; Kettunen 1986; Tsvetkov 1) Agreements were made between the University of Tartu and the 1995 [compiled in the 1920’s]; Grünberg 20136). owners of the original legacy collections. The agreements speci­ Ingrian Finnish comprises a large heterogeneous group of va­ fied the conditions for processing the collections and the details rieties spread across the entire territory of Ingria (from the Narva of archiving. River up to Karelia). The information on the Ingrian Finnish dialects 2) Four out of six collections were stored on magnetic media and is very scarce; the published works address either some particular had to be digitized. This work was done partially by the own­ aspects or some concrete varieties (Galahova 1974, 2000; Leppik ers of the collections (collections from Ilya Nikolaev and the 1975; Kirpu 1989; Lehto 1996; Kokko 2007; Riionheimo 2007; Institute of Language, Literature, and History of the Karelian Muslimov 2009, 2014; and others), and there is no comprehensive Research Centre) and partially by Fedor Rozhanskiy and Elena description of Ingrian Finnish. Although at the end of the 19th cen­ Markus (collections from Mehmet Muslimov and Enn Ernits). tury, the number of Ingrian Finnish speakers in Ingria was around All the recordings were digitized as WAV files with a sampling 130 000 (Musaev 2004: 26), at present it is no more than 500 na­ rate 44,1 kHz, 16 bit, mono. tive speakers7. Most of these speakers were born in the 1930s; their 3) The resulting files corresponded to the original side of the tape or average age approaches 80 (Kuznetsova et al. 2015: 24–25). These cassette and often contained several working sessions with dif­ dialects have been influenced by Standard Finnish (the degree of ferent speakers. In order to provide proper metadata, such files influence depends greatly on the region and particular speaker). In were cut into smaller files to separate the sessions and named in the middle of the 1930s, the number of Finnish schools in the re­ a uniform way (see Section 5.2. on naming conventions). gion was quite impressive (Musaev 2004: 248 indicates more than 4) Finally, metadata files were provided for each recording (see the 300 schools), but the language of instruction was Standard Finnish details in 5.2.1.). The relevant details about the recordings were (Kuznetsova et al. 2015: 26). obtained either from the original owners or from the recording itself. In the latter case, the project participants listened through the recordings in order to find the necessary information.

New field materials were recorded in the course of the project from Ingrian and Votic speakers. These sessions were aimed at collect­ ing data not represented in legacy collections and providing a better quality of the data. In particular: 4. The grammar (Ariste 1968) is the English translation of Ariste 1948 written originally in Estonian. 5. Posti 1980 is the dictionary of the Kukkuzi variety. 6. Grünberg 2013 is the second edition. The first edition in seven volumes was published in 1990–2011 (1st–4th volumes edited by Elna Adler and Merle Leppik, 5th–7th volumes edited by Silja Grünberg). 7. It is difficult to estimate the number of Ingrian Finnish speakers in other regions (Karelia, Estonia, and Finland).

308 309 FEDOR ROZHANSKIY & ELENA MARKUS A NEW RESOURCE FOR FINNIC LANGUAGES: THE OUTCOMES OF THE INGRIAN DOCUMENTATION PROJECT

1) The recordings represent the contemporary state of the two lan­ Language Audio (hrs) Video (hrs) guages8. Ingrian 302.7 15.2 2) Apart from the narratives that are traditionally considered the Votic 83.7 6.2 main type of language data by Finno-Ugric researchers, we Ingrian Finnish 51.0 recorded many hours of elicitations. These included nominal Veps 9.0 and verbal paradigms and specially designed phonetic question­ Karelian 2.6 naires. Many of the recorded narratives were transcribed and South Estonian 1.0 Estonian 0.2 translated with the help of language consultants9. Mixed varieties 59.3 3) All the sessions were recorded with high quality equipment pro­ TOTAL 509.5 21.4 viding modern recording standards: audio files in WAV format Table 1. The amount of data (in hours) for the six languages represented in the with a sampling rate 48 kHz, 16 bit, stereo. resource. Under mixed varieties we list (a) instances when speakers of different 4) Video documentation was done for both the Ingrian and Votic varieties are involved in one session; (b) sessions with speakers of mixed varieties languages. (these are mostly various mixtures of Ingrian, Votic, and Ingrian Finnish).

All recorded data were provided with detailed metadata files. A part of the recorded narratives and grammar sessions was tran­ 5.1. Resource contents scribed and translated with the help of language consultants. The tran­ scriptions were time aligned with media files (either audio or video This section provides details about the contents of each collection. together with the extracted waveform) in the ELAN software. 5.1.1. Collection compiled by Ilya Nikolaev

5. Project results The materials collected by Ilya Nikolaev (Saint Petersburg, Russian Federation) comprise about 20 hours of Ingrian recordings. A major The main result of the project is the language resource on Ingrian part of the data represents Soikkola Ingrian; a few recordings rep­ and other Finnic languages. The resource consists of six collections. resent the Lower Luga dialect. The recordings were made in 1996– Five collections come from private owners: Ilya Nikolaev, Enn Ernits, 2002 on a mini cassette recorder. Most of the recording sessions are Natalia Kuznetsova, Mehmet Muslimov, Fedor Rozhanskiy & Elena the interviews of Nikolaev with the speakers on different topics (life Markus. One collection comes from the Institute of Language, Litera­ in the village, deportation to Finland, fishing, traditional festivities, ture, and History of the Karelian Research Centre in Petrozavodsk. the Ingrian language, personal biographies, and other topics). Several All collections were prepared for archiving by Fedor Rozhanskiy and recordings contain elicitations of basic grammar and vocabulary. Elena Markus. The materials were digitized by Ilya Nikolaev; metadata were Table 1 shows the amount of data (audio and video) for each provided by Ilya Nikolaev, Fedor Rozhanskiy, and Elena Markus. language represented in the resource. 8. Although both Votic and Ingrian languages are on the verge of extinction, they 5.1.2. Collection compiled by Enn Ernits demonstrate development processes typical of fully alive languages (see Rozhanskiy & Markus 2014b) and should not be treated as degraded. The collection compiled by Enn Ernits (Tartu, Estonia) contains about 9. We prefer the term “language consultant” to the synonymic “language informant” (see, e.g., Mosel 2012: 76 on the specific connotations for both terms). 23 hours of recordings of Finnic languages: Ingrian (Soikkola and

310 311 FEDOR ROZHANSKIY & ELENA MARKUS A NEW RESOURCE FOR FINNIC LANGUAGES: THE OUTCOMES OF THE INGRIAN DOCUMENTATION PROJECT

Lower Luga), Votic, Veps, and Karelian10. The recordings (1971–1986) Muslimov visited almost all the villages in the Lower Luga area that were made on a reel-to-reel recorder and later on a cassette recorder. had remnants of Finnic population in the beginning of the 21st century, Most of the recordings are conversations between the interviewers and and worked with most of the speakers. This is probably the largest col­ speakers on various topics, the main focus being folk medicine, folk lection of recordings of the Lower Luga varieties, and it is extremely astronomy, and folk music. The interviewers are Enn Ernits, Tiiu Er­ valuable for the study of language contact, dialectology, and the his­ nits, and occasionally other Estonian researchers, among them Paul tory and culture of the region. Unfortunately, the quality of the record­ Ariste. The interviews are done in the speaker’s language. ings is not always good. The materials were digitized by Fedor Rozhanskiy; metadata The materials were digitized by Fedor Rozhanskiy; metadata were provided by Enn Ernits, Fedor Rozhanskiy, and Elena Markus. were provided by Mehmet Muslimov, Natalia Kuznetsova, Fedor Rozhanskiy, and Elena Markus. 5.1.3. Collection compiled by Natalia Kuznetsova 5.1.5. Collection from the Institute of Language, Literature, The collection compiled by Natalia Kuznetsova (Saint Petersburg, and History of the Karelian Research Centre, Russian Russian Federation) contains about 30 hours of Ingrian recordings Academy of Sciences (Petrozavodsk, Russian Federation) (Soikkola and Lower Luga dialects). Additionally, there is one record­ ing session with a speaker of the Kukkuzi variety. The recordings were This collection contains more than 65 hours of recordings made by made in 2008–2012 on a digital recorder. Most of the data are elicita­ folklore researchers (Eino Kiuru, Elina Kylmäsuu, and others) in the tions of nominal morphology and phonetic questionnaires. Some of late 1960s–1970s. A major part of the collection represents the Ingrian the recordings are interviews in Ingrian and Russian. language (Soikkola, Lower Luga, and the now extinct Heva dialects). Metadata were provided by Natalia Kuznetsova, Fedor Rozhan­ There are also some recordings of Ingrian Finnish and several Votic skiy, and Elena Markus. recordings. The materials were recorded using a reel-to-reel (some­ times cassette) recorder and were digitized by the Institute under the 5.1.4. Collection compiled by Mehmet Muslimov agreement made in the course of the project. The collection covers various genres of Ingrian folklore including songs, laments, riddles, The collection by Mehmet Muslimov (Saint Petersburg, Russian Fed­ folk rhymes, and proverbs. There are also interviews with the speakers eration) contains more than 245 hours of recordings made in the Low­ about traditional ceremonies and beliefs. er Luga area. Muslimov recorded minor Finnic languages spoken in Metadata were provided by the Institute, Ilya Nikolaev, Fedor the area, including the Lower Luga dialect of Ingrian, the Lower Luga Rozhanskiy, and Elena Markus. varieties of Votic, the Kukkuzi variety, and the Lower Luga varieties of Ingrian Finnish. There are also a few recordings of the Veps, South 5.1.6. Collection compiled by Elena Markus Estonian, and Estonian (Siberian variety) languages made in the Low­ and Fedor Rozhanskiy er Luga region. The recordings were made in 2000–2005 on a cassette recorder. The materials include talks on various topics (mostly festivi­ The collection compiled by Elena Markus and Fedor Rozhanskiy (Tar­ ties, food, school education, local languages, and folk linguistics), and tu, Estonia) contains about 120 hours of Ingrian (Soikkola and Lower elicitations of dialect vocabulary, paradigms, and simple sentences. Luga dialects) and Votic (Lower Luga varieties) recordings. A major part of the collection was recorded in 2011–2013 (the earliest record­ 10. Veps and Karelian data were recorded in Boksitogorsk and Prionezhsk regions. ings are from 2006). There are three types of data in this collection:

312 313 FEDOR ROZHANSKIY & ELENA MARKUS A NEW RESOURCE FOR FINNIC LANGUAGES: THE OUTCOMES OF THE INGRIAN DOCUMENTATION PROJECT

• audio recordings, 5.2. Representation of the data and • video recordings, structure of metadata files • ELAN annotations (audio and video recordings aligned with transcription and translation into Russian and English). In the resulting resource, the files are organized into bundles. Each bundle consists of files with the same name but different extensions Most of the audio materials were made with high-quality record­ that correspond to different data types. In the simplest variant, a bundle ing equipment (digital recorders with external electret microphones, consists of two files: an audio file plus a metadata file, e.g., sampling rate 48 kHz). • ING_MAR_PAJA_OM110727.WAV (sound) The collection includes both samples of spontaneous speech • ING_MAR_PAJA_OM110727.RTF (metadata) (narratives and dialogues), and elicitations (phonetic and grammar In more complex cases, a bundle also contains a video file and an questionnaires, nominal and verbal paradigms). ELAN annotation, e.g., Figure 1 plots the six collections from the point of view of lan­ • ING_MAR_SKVORTSAD_EN120622.MPG (video) guages represented and the time of the recording. • ING_MAR_SKVORTSAD_EN120622.WAV (sound) • ING_MAR_SKVORTSAD_EN120622.EAF (ELAN transcription) • ING_MAR_SKVORTSAD_EN120622.RTF (metadata) File naming conventions are slightly different for each collection. A name always begins with the language code and the collection code, e.g., ING_ NIK_VEN_SEV110502, FIN_KRC_01706A_SEPPANEN, where ING stands for Ingrian, FIN stands for Finnish, NIK and KRC denote the owners of the collection (Ilya Nikolaev and the Institute of Language, Literature, and History of the Karelian Research Centre, respectively). The remaining part of the name is a unique identifier that refers to the original recording on magnetic tapes and usually contains a reference to the native speaker. Each file in the resource is provided with a metadata file. A meta­ data file is a table in theRTF format11, see Table 2. The metadata are presented in five subsections: 1) File (lists the file properties); 2) Recording (provides details about the recording, including the collector, recording date, processing stages, etc.); 3) Language consultant (provides sociolinguistic data about the speaker: the name, date, and place of birth of the speaker and his parents, the spoken variety, etc.); Figure 1. Languages and periods of data collection. 4) Contents (briefly describes the contents of the recording); 5) Access (specifies the owner of the recording and access rights). 11. The archive of the University of Tartu has a special way of representing metadata: the metadata are given in special forms accompanying each audio file.

314 315 FEDOR ROZHANSKIY & ELENA MARKUS A NEW RESOURCE FOR FINNIC LANGUAGES: THE OUTCOMES OF THE INGRIAN DOCUMENTATION PROJECT

File Access File name ING_MAR_TORMI_ Rights to process the files Fedor Rozhanskiy, Elena Markus VV110729 The owner of the recording Fedor Rozhanskiy, Elena Markus File type Sound Location of original carrier N/A File properties WAV, 48000 Hz, 16-bit, stereo Access for listening No limitations Duration of sound 00:02:19 Access for copying No limitations File size 26.869.894 bytes Comments Full or partial publication requires the Processing completed owner’s permit ([email protected]) Recording Collector’s name Fedor Rozhanskiy, Table 2. A sample of a metadata file. Elena Markus Date of collection 2011-07-29 Recorded with Recorder: Edirol R-09HR, Additionally, all individual metadata files are combined in an Microphone: Edirol -15 Excel spreadsheet. The spreadsheet contains: Recording transcribed (by whom, when) No Recording digitized (by whom, when) N/A a) the main worksheet with a full list of files and their detailed Recording annotated (by whom, when) N/A characteristics (one line for one recording corresponds to one Recording prepared for archiving Fedor Rozhanskiy, bundle); (by whom, when) Elena Markus, March 2012 b) a worksheet with the list of all the recorded speakers and basic Other file processing (by whom, when) sociolinguistic data; Language consultant c) a list of all mentioned settlements with parallel Russian and Language Ingrian Finno-Ugric names and location. Dialect Soikkola (South) Consultant name Petrova Valentina Vasilyevna The main worksheet lists the same metadata as the RTF files. Each line Place of collection Slobodka [Säätinä] in the sheet corresponds to one individual RTF file. The spreadsheet Consultant’s date of birth 1937 allows scrolling through the whole dataset and applying all Excel pos­ Consultant’s place of birth Krasnaja Gorka [Yhimägi] sibilities to sort, filter, and search the data. Consultant’s parents (mother: native Ingrian, Smenkovo [Otsave]; The worksheet with the list of speakers provides basic sociolin­ language, place of birth; father: native Ingrian, Krasnaja Gorka guistic data necessary for language analysis: the name, date, and place language, place of birth) [Yhimägi], KIA in WWII of birth, the native variety, place of recording, the native language and Comments place of birth of the speaker’s parents. Contents The third worksheet lists all settlements mentioned in the meta­ Genre Narrative data and specifies the official Russian name, the original names in the Detailed contents Story about fishing local languages, the geographical coordinates, and the administrative Intermediary language Ingrian, Russian region. Other comments Noise – thunder Interviewer: Elena Markus

316 317 FEDOR ROZHANSKIY & ELENA MARKUS A NEW RESOURCE FOR FINNIC LANGUAGES: THE OUTCOMES OF THE INGRIAN DOCUMENTATION PROJECT 5.3. Archiving 5.4. Access rights and restrictions

The language resource created in the course of the project has been Under the agreements made with the owners of the original collec­ archived at four institutions: tions, the owners preserved full rights on their materials. The project participants were granted non-exclusive rights to process the record­ • the Endangered Languages Archive (London, UK), ings (for example, split and rename the files, provide metadata). • the University of Tartu (Estonia), A major part of the resource is open access for scientific pur­ • the Institute of Language, Literature, and History of the Karelian poses. The conditions and a few restrictions on the data usage are the Research Centre (Petrozavodsk, Russian Federation), following: • the Ingrian museum in the village of Vistino (the Kingisepp region of the Leningrad oblast, Russian Federation). • listening of the recordings does not require the owner’s permit; • copying of the materials can be free or can require the owner’s Each institution has some specific archiving conditions, so the permit. Each metadata file specifies access/usage conditions and amount of data deposited and the access conditions differ, see Table 3. restrictions, the owner of the particular recording, and the email address, which can be used for contacting the owner; Endangered University of Karelian Research Ingrian • full or partial publication of any materials requires the owner’s Languages Archive Tartu Centre Museum permit. Location London, Tartu, Petrozavodsk, Vistino, UK Estonia RF RF The archives with online access (ELAR and UT) have some Online access yes yes no no specific terms of use. ELAR has four access categories: free access, Audio files (WAV), hrs 510 510 510 240 access for researchers, access for the members of the language com­ Video files (mpg, mp4), hrs 21 – 21 – munity, and access for subscribers (for more details see http://www. ELAN annotation (eaf), hrs 15 – 12 – elar-archive.org/using-elar/access-protocol.php). A major part of our resource allows free access. A small part of the materials is currently Table 3. Archiving institutions. accessible only for subscribers, because the corresponding materials are being prepared for publication. The Endangered Languages Archive and the University of Tartu The UT archive has two access conditions: free access or access provide online access to the materials via the addresses ‹http://elar. for authorized users. Most of the materials from our resource are free­ soas.ac.uk/deposit/0147› and ‹http://www.murre.ut.ee/arhiiv/otsi.php›. ly accessible. Respecting the wishes of the owners, access to certain parts of the data requires authorization.

318 319 FEDOR ROZHANSKIY & ELENA MARKUS A NEW RESOURCE FOR FINNIC LANGUAGES: THE OUTCOMES OF THE INGRIAN DOCUMENTATION PROJECT 6. Further work References

The materials accumulated in the course of the Ingrian documentation Agranat, Tatjana B. 2007: Zapadnyj dialekt vodskogo jazyka [Western project open wide possibilities for further descriptive work. The main dialect of Votic]. Mitteilungen der Societas Uralo-Altaica, Heft 26. tasks intended to be carried out by the authors of this paper are the Moskva–Groningen. following: Ahlqvist, August 1856: Wotisk grammatik jemte språkprof och ordförteckning [Votic grammar with language samples and vocabulary]. Helsingfors. Ariste, Paul 1948: Vadja keele grammatika [A grammar of the Votic lan­ 1) Compiling a concise morphological dictionary of Soikkola Ingrian guage]. Tartu: Teaduslik Kirjandus. (around 2 000 entries). A major part of the materials used in the — 1968: A grammar of the Votic language. Indiana University publica­ dictionary was collected during the documentation project. The tions, Uralic and Altaic series vol. 68. Bloomington–the Hague: Indi­ first (electronic) version of the dictionary has been placed at ana University. ‹http://ingrian.org› (the electronic dictionary was designed by Fedor Galahova, Lidia Ja. 1974: Osnovnye osobennosti konsonantizma v finskix Rozhanskiy during the course of the project financed by the Kone govorax Leningradskoj oblasti [The main characteristics of consonan­ Foundation in 2015–2018). This software builds inflectional para­ tal inventory in Finnish varieties of the Leningrad oblast. PhD thesis]. digms and provides illustrative audio material for the dictionary Sankt-Peterburg: Leningradskij gosudarstvennyj universitet. entries. Preparation of the paper version of the dictionary is in pro­ — 2000: Čeredovanie stupenej soglasnyx v osnove slova v finskix govo­ gress. A similar dictionary is later planned for Lower Luga Ingrian. rax Leningradskoj oblasti [Grade alternation in Finnish varieties of 2) Compiling a corpus of transcribed and annotated recordings. the Leningrad oblast]. – Kafedra finno-ugorskoj filologii: Izbrannye trudy k 75-letiju kafedry. Sankt-Peterburg: Izdatelʹstvo Sankt-Peter­ Within the documentation project, several hours of Ingrian nar­ burgskogo universiteta. 115–133. ratives and dialogues were recorded, transcribed, translated, and Grünberg, Silja (ed.) 2013: Vadja keele sõnaraamat [Dictionary of Votic]. 2., aligned with the audio track in ELAN. The annotation work will täiendatud ja parandatud trükk. Tallinn: Eesti Keele Instituut & Eesti be continued in order to enlarge the corpus. Keele Sihtasutus. 3) Research on Ingrian prosody. As mentioned above, Ingrian has Junus, Väino I. 1936: Ižoran keelen grammatikka. Morfologia opettaijaa a non-trivial system of quantity relations. Based on the phonetic vart [A grammar of the Ingrian language. Morphology for teachers]. questionnaires recorded during the documentation project we Leningrad–Moskva: Ucpedgiz. intend to thoroughly investigate segmental and suprasegmental Kettunen, Lauri 1986: Vatjan kielen Mahun murteen sanasto [Vocabulary characteristics of the Soikkola Ingrian dialect. of the Votic Mahu dialect]. Castrenianumin toimitteita 27. Helsinki: 4) Compiling a grammar of Soikkola Ingrian based on the data on Castrenianumin laitokset & Suomalais-Ugrilainen Seura. phonetics, morphology, and syntax collected during the project. Kirpu, Lilia 1989: O nekotoryx fonetičeskix osobennostjax markovskogo govora finskogo jazyka Leningradskoj oblasti [On some phonetic characteristics of the Markovskij variety of the Ingrian Finnish of the Along with the descriptive tasks listed above, we intend to con­ Lenin­grad Oblast]. – Fenno-Ugristica 15: 80–87. tinue documentation and further expand the resource. In particular, we Kokko, Ossi 2007: Inkerinsuomen pirstaleisuus. Eräiden sijojen kehi- are planning to include more of our Ingrian and Votic field recordings. tys murteen yksilöllistymisen kuvastajana [Scattered Ingrian At the moment, our collection of Votic recordings comprises about Finnish. The development of selected cases as reflectors of the 260 hours, and the Ingrian collection12 approaches 950 hours. individualization of a dialect. PhD thesis]. University of Joensuu Pub­ lications in the Humanities 48. Joensuu. Available at: ‹http://urn.fi/ 12. The Ingrian collection includes recordings made by Fedor Rozhanskiy, Elena Markus, Natalia Kuznetsova, and some other participants of the Ingrian field trips. URN:ISBN:978-952-219-036-9›

320 321 FEDOR ROZHANSKIY & ELENA MARKUS A NEW RESOURCE FOR FINNIC LANGUAGES: THE OUTCOMES OF THE INGRIAN DOCUMENTATION PROJECT

Kuznetsova, Natalia, Elena Markus & Mehmed Muslimov 2015: Finno- — 2009: K klassifikacii finskix dialektov Ingermanlandii [On classifica­ Ugric minorities of Ingria: the current sociolinguistic situation and tion of the Finnish dialects in Ingria]. – Sergej Myznikov & Igorʹ Brod­ its background. – Heiko Marten, Michael Rieβler, Janne Saarikivi & skij (eds), Voprosy uralistiki 2009. Naučnyj alʹmanax. Sankt-Peterburg: Reetta Toivanen (eds), Cultural and linguistic minorities in the Russian Nauka. 179–204. Federation and the European Union. Multilingual Education 13: Com­ — 2014: Zametki o moloskovickom ingermanlandskom dialekte [Notes parative studies on equality and diversity. Berlin: Springer. 127–167. on the Moloskovitskij variety of Ingrian Finnish]. – Valentin F. Vydrin Laanest, Arvo 1966: Ižorskie dialekty. Lingvogeografičeskoe issledovanie & Natalia V. Kuznetsova (eds), From Bikin to Banbaluma, from the [Ingrian dialects. A linguistic-geographical study]. Tallinn: Akade­ Varangians to the Greeks. Field-inspired essays in honour of Elena V. mija nauk Estonskoj SSR. Perekhvalskaya. Sankt-Peterburg: Nestor-Istorija. 277–287. — 1978: Istoričeskaja fonetika i morfologija ižorskogo jazyka [Histori­ Nirvi, Ruben Erik 1971: Inkeroismurteiden sanakirja [Dictionary of Ingrian cal phonetics and morphology of Ingrian. PhD thesis]. Tallinn: Institut dialects]. Lexica Societatis Fenno-Ugricae XVIII. Helsinki: Suoma­ jazyka i literatury. lais-Ugrilainen Seura. — 1986: Isuri keele ajalooline foneetika ja morfoloogia [Historical pho­ Porkka, Volmari 1885: Über den ingrischen Dialekt mit Berücksichtigung netics and morphology of Ingrian]. Tallinn: Valgus. der übrigen finnisch-ingermanländischen Dialekte [On the Ingrian — 1997: Isuri keele Hevaha murde sõnastik [Vocabulary of the Ingrian dialect with respect to other Ingrian Finnish dialects]. Helsingfors: Heva dialect]. Tallinn: Eesti Keele Instituut. J. C. Frenckell & Sohn. Lehto, Manja Irmeli 1996: Ingrian Finnish: Dialect Preservation and Posti, Lauri 1980: Vatjan kielen Kukkosin murteen sanakirja [Dictionary of Change. [PhD thesis.] Acta Universitatis Upsaliensis. Studia Uralica the Votic Kukkuzi dialect]. Ainekset kerännyt Lauri Posti. Painokun­ Upsaliensia 23. Uppsala: Uppsala University. toon toimittanut Seppo Suhonen Lauri Postin avustamana. Lexica Leppik, Merle 1975: Ingerisoome Kurgola murde fonoloogilise süsteemi Societatis Fenno-Ugricae XIX, Kotimaisten kielten tutkimuskeskuk­ kujunemine [The formation of the phonological system of the Ingrian sen julkaisuja 8. Helsinki: Suomalais-Ugrilainen Seura & Kotimaisten Finnish Kurgola dialect]. Tallinn: Eesti NSV Teaduste Akadeemia. kielten tutkimuskeskus. Markus, Elena & Fedor Rozhanskiy 2012: Votic or Ingrian: new evidence Riionheimo, Helka 2007: Muutoksen monet juuret. Oman ja vieraan ris- on the Kukkuzi variety. – Finnisch-Ugrische Mitteilungen 35: 77–95. teytyminen Viron inkerinsuomalaisten imperfektinmuodostuksessa Markus, Elena B. & Fedor I. Rozhanskiy 2017 [2011]: Sovremennyj vod- [Multiple roots of change. Mixing native and borrowed influence in skij jazyk. Teksty i grammatičeskij očerk. 2-e izdanije, ispravlennoje i the past tense formation by Ingrian Finns. PhD thesis]. Suomalaisen dopolnennoje [Contemporary Votic language. Texts and grammar. 2nd Kirjallisuuden Seuran Toimituksia 1107. Helsinki: Suomalaisen Kir­ edition]. Sankt-Peterburg: Nestor-Istorija. Available at: ‹http://ingrian. jallisuuden Seura. org/Votic-grammar/› Rozhanskiy, Fedor I. & Elena B. Markus 2013a: Ižora Sojkinskogo poluos­ Mosel, Ulrike 2012: Morphosyntactic analysis in the field: a guide to the trova: fragmenty sociolingvističeskogo analiza [Soikkola Ingrians: a guides. – Nicholas Thieberger (ed.), The Oxford Handbook of Linguis- sociolinguistic study]. – Acta linguistica Petropolitana. Transactions tic Fieldwork. Oxford University Press. 72–89. of the Institute for Linguistic Studies. Vol. IX, part 3. Sankt-Peterburg: Musaev, Vadim I. 2004: Političeskaja istorija Ingermanlandii v konce XIX– Nauka. 261–298. XX veke [Political history of Ingria at the end of the 19th–20th century]. — 2013b: O statuse nižnelužskogo dialekta ižorskogo jazyka sredi rodst­ 2nd ed. Sankt-Peterburg: Nestor-Istorija. vennyx idiomov [On the status of Lower Luga Ingrian among related Muslimov, Mehmed Z. 2005: Jazykovye kontakty v Zapadnoj Ingermanlan- varieties]. – Tatjana B. Agranat, Olga A. Kazakevich & Egor V. dii (nižnee tečenie reki Lugi) [Language contacts in Ingria (the lower Kashkin (eds), Lingvističeskij bespredel – 2. Sbornik naučnyx trudov k course of the Luga River). PhD thesis]. Sankt-Peterburg: Institute for jubileju A. I. Kuznecovoj. Moskva: Izdatelʹstvo Moskovskogo Univer­ Linguistic Studies of the Russian Academy of Sciences. siteta. 219–232.

322 323 FEDOR ROZHANSKIY & ELENA MARKUS A NEW RESOURCE FOR FINNIC LANGUAGES: THE OUTCOMES OF THE INGRIAN DOCUMENTATION PROJECT

Rozhanskiy, Fedor & Elena Markus 2014a: Lower Luga Ingrian as a con­ Результаты проекта по документации vergent language. – FINKA Symposium: On the Border of Language ижорского языка: новый интернет-ресурс and Dialect. University of Eastern Finland, Joensuu, 4–6 June, 2014. по прибалтийско-финским языкам Joensuu. [Abstracts.] 36–37. Rozhanskiy, Fedor I. & Elena B. Markus 2014b: Dinamika morfologii vod­ Федор Рожанский & Елена Маркус skogo jazyka s načala XX veka [Dynamics in the morphology of Votic since the beginning of the 20th century]. – Finno-ugorskie jazyki i Статья посвящена описанию нового интернет-ресурса, который kultury v sociokulturnom landšafte Rossii. Materialy V Vserossijskoj konferencii finno-ugrovedov. Petrozavodsk, 25–28 ijunja 2014 g. Petro­ стал главным результатом работы над проектом по документации zavodsk: Karelskij naučnyj centr RAN. 118–121. ижорского языка. Проект выполнялся в 2011–2013 годах в Тарту­ Rozhanskiy, Fedor & Elena Markus 2019: Ingrian morphologial dictionary. ском университете Ф. Рожанским и Е. Маркус при финансовой Electronic dictionary. Available at: ‹http://ingrian.org/Ingrian-dictionary/› поддержке Программы по документации языков под угрозой ис­ Suhonen, Seppo 1985: Wotisch oder Ingrisch? [Votic or Ingrian?] – Wolfgang чезновения (Лондон, Великобритания). Исходно проект был на­ Veenker (ed.), Dialectologia Uralica: Materialen des ersten Interna- целен в равной степени на документацию и описание ижорского tionalen Symposions zur Dialektologie der uralischen Sprachen 4.–7. языка, однако, впоследствии на первый план вышли задачи по September 1984 in Hamburg. Veröffentlichungen der Societas Uralo- документации. При этом производился сбор материала не только Altaica. Band 20. Wiesbaden: Harrassowitz. 139–148. по ижорскому, но и по соседствующим малым языкам (прежде Tsvetkov, Dmitri 1995: Vatjan kielen Joenperän murteen sanasto [Vocabu­ всего, водскому). Двумя основными направлениями проекта ста­ lary of the Votic Jõgõperä dialect]. Toimittanut, käänteissanaston ja ли: (а) оцифровка и архивация записей ижорского и других малых hakemiston laatinut Johanna Laakso. Lexica Societatis Fenno-Ugricae прибалтийско-финских языков, сделанных предшествующими XXV, Kotimaisten kielten tutkimuskeskuksen julkaisuja 79. Helsinki: Suomalais-Ugrilainen Seura & Kotimaisten kielten tutkimuskeskus. исследователями; (б) сбор нового полевого материала в соответ­ — 2008: Vadja keele grammatika [Grammar of the Votic language] ствии с современными стандартами документации. (Эсимейнʹ ваддя чээле грамаатикк. Первая грамматика водьского К архивации было подготовлено пять коллекций предше­ языка, 1922). Jüri Viikberg (ed.). Tallinn: Eesti Keele Sihtasutus. ствующих исследователей: коллекция Ильи Николаева (около 20 часов записей ижорской речи и некоторое количество элицитаций, по большей части сойкинский диалект, 1996–2002 гг.); коллек­ ция Эна Эрнитса (23 часа записей сойкинского и нижнелужского ижорского, водского, карельского и вепсского языков, 1971–1986 гг.); коллекция Натальи Кузнецовой (около 30 часов записей сой­ кинского и нижнелужского ижорского, а также куровицкого идио­ ма, в основном анкеты по фонетике и грамматике, 2008–2012 гг.); коллекция Мехмеда Муслимова (245 часов записей нижнелужско­ го ижорского, водского, куровицкого идиома, ингерманландского диалекта финского, а также других языков, интервью на разные темы и анкеты по грамматике и диалектологии, 2000–2005 гг.); коллекция Института языка, литературы и истории Карельского научного центра (более 65 часов записей сойкинского, нижне­ лужского и хэваского ижорского, ингерманландского финского и

324 325 FEDOR ROZHANSKIY & ELENA MARKUS

водского, содержащих фольклорные тексты и интервью на тему LIINA LINDSTRÖM, PÄRTEL LIPPUS & обрядов и ритуалов, 1968–1977 гг.). TUULI TUISK Шестую коллекцию составили полевые материалы исполни­ University of Tartu телей проекта, включающие в себя около 120 часов аудиозаписей ижорского и водского языка (образцы спонтанной речи и анкеты по грамматике и фонетике), а также видеозаписи и аннотации в программе ELAN (транскрипция, сопровождаемая русским и ан­ The online database of глийским переводами). Основная часть записей сделана в 2011– the University of Tartu 2013 гг. Для всех файлов были подготовлены подробные метадан­ Archives of Estonian Dialects ные, содержащие информацию о медиафайле, месте и времени and Kindred Languages and записи, носителе языка, содержании записи и правах доступа. В общей сложности ресурс содержит 510 часов аудиозапи­ the Corpus of Estonian Dialects сей, 21 час видеозаписи и 15 часов аннотаций в ELAN. Ресурс размещен на сайте Архива языков под угрозой ис­ чезновения (Лондон, Великобритания) и на сайте Архива эстон­ 1. Introduction ских диалектов и родственных языков (Тарту, Эстония). Также копии ресурса переданы в архив Института языка, литературы и This paper introduces the online database of the University of Tartu истории Карельского научного центра (Петрозаводск, Россия) и в Archives of Estonian Dialects and Kindred Languages (AEDKL; in Ижорский музей (Вистино, Ленинградская область, Россия). Estonian, Tartu Ülikooli eesti murrete ja sugulaskeelte arhiiv), which is freely accessible and open to researchers at ‹https://murdearhiiv. ut.ee/›, and as an independent part of the archives, also the Cor­ pus of Estonian Dialects (CED; in Estonian, Eesti murrete korpus, ‹https://www.keel.ut.ee/et/keelekogud/murdekorpus›). Both sources have been developed at the University of Tartu. These sources are re­ lated, as the recordings and transcribed texts of the CED are held in the archives, while the materials of the AEDKL are used in the CED. The first half of this paper introduces the history and materials held in the AEDKL and how they can be used online. The second half of the paper gives an overview of the dialect corpus. This paper describes the state of the archives as of December 2019.

Multilingual­ Finnic. Language 326 contact and change. 327–350. Uralica Helsingiensia 14. Helsinki 2019. ‹https://doi.org/10.33341/uh.85040› LIINA LINDSTRÖM, PÄRTEL LIPPUS & TUULI TUISK THE ONLINE DATABASE OF THE UNIVERSITY OF TARTU ARCHIVES ...

2. The Archives of Estonian Dialects In 1947, the Institute of Language and Literature at the Estonian and Kindred Languages (AEDKL) SSR Academy of Sciences (Eesti NSV Teaduste Akadeemia Keele ja Kirjanduse Instituut1) was established. The Mother Tongue Society The archives are a collection of Finno-Ugric linguistic materials and along with its archives was taken over by the Estonian SSR Academy are located at the University of Tartu. The collection contains four of Sciences (see EF 1997: 10, Rätsep 2003: 169). This was part of a types of materials: 1) sound recordings of Estonian dialects and other policy to organize research into institutes under the umbrella of the Uralic languages; 2) unpublished manuscripts, including student Academy of Sciences with the university instead focusing on teach­ coursework and theses defended at the Institute of Estonian and Gen­ ing. At first, the two institutions were both located in Tartu and used eral Linguistics, fieldwork diaries, transcriptions and written notes on the common archives, but in 1952 the institute together with the ar­ Estonian and Finno-Ugric languages; 3) photos from fieldwork expe­ chives moved to Tallinn. In order to carry on with research in Tartu, ditions and linguistic events; 4) video recordings. the university had to build a new collection of linguistic data. The The organization of the archives began in 2000 with the digitiza­ basis of the current archives is formed from the seminar papers and tion of the sound recordings of Estonian dialects and the creation of the MA theses of the 1920s and 1930s which the institute returned to the Corpus of Estonian Dialects. The work then moved on to digitizing the university. Step by step more materials were added with students mak­ recordings of other Finno-Ugric languages, scanning the written materi­ ing handwritten copies of manuscripts of the Mother Tongue Society als, and organizing the metadata into an online database. By now, most collection as well as through the process of collecting new data on of the materials are digitized and accessible through the online database. annual fieldwork expeditions. A new era dawned in 1957, when the The majority of the recordings are of Estonian dialects but another sig­ first battery-powered tape recorder was obtained and the first sound nificant portion contains recordings of other Uralic languages. recordings were made. Since then, fieldwork expeditions have been organized every summer (Kingisepp 1967). Therefore, the amount of 2.1. History and background materials held in the archives increases every year.

The history of the archives dates back to the post- period 2.2. Sound recordings early in the history of the Republic of Estonia. In 1920, a year after the University of Tartu was opened as an Estonian language university, The archives consist of about 2 800 hours of sound recordings. The the Mother Tongue Society (Emakeele Selts) was established. The So­ majority of the recordings are of Estonian dialects with the remainder ciety initiated systematic work in collecting Estonian dialect materials composed of recordings of other Finnic languages (Livonian, Votic, and conducting research into these dialects (Rätsep 2003, Erelt 2010). Ingrian, Veps, Karelian, Olonets Karelian, Lude, and Ingrian Finn­ The collection containing mostly manuscripts but also sound record­ ish) and Finno-Ugric (Inari Saami, Erzya, Moksha, Komi, Udmurt, ings on wax cylinders was organized as the Archives of the Estonian Khanty, Hungarian) and Samoyedic (Kamas, Nenets) languages (see Language and Kindred Languages and due to the fact that the Mother Table 1). Tongue Society was related to the University of Tartu, the archives were stored in the rooms of the university phonetics lab. Near the end of World War II when Tartu was heavily bombed, most of the archives were evacuated, but the collection of wax cylinders was destroyed in a fire (Ahven 1955). After the war the manuscripts were returned to Tartu. 1. Beginning in 1993: Eesti Keele Instituut, the Institute of the Estonian Language.

328 329 LIINA LINDSTRÖM, PÄRTEL LIPPUS & TUULI TUISK THE ONLINE DATABASE OF THE UNIVERSITY OF TARTU ARCHIVES ...

Family Language Total duration (h:mm:ss) Series Content Total duration (h) Finnic Estonian 1 866:00:12 F reel-to-reel tape and cassette recordings of 1 044 Finnic Ingrian 390:18:37 Estonian dialects (this series also contains some recordings of other Finno-Ugric languages) Finnic Votic 192:26:58 SU reel-to-reel tape and cassette recordings of other 133 Finnic Livonian 145:25:27 Finno-Ugric languages Finnic Ingrian Finnish 64:37:21 digital sound recordings 695 Finnic Veps 103:27:21 EMH copies of recordings of Estonian dialects from the 396 Institute of the Estonian Language Finnic Karelian 34:44:02 SUHK Finno-Ugric languages from the Institute of the 77 Finnic Olonets Karelian 10:12:28 Estonian Language Finnic Lude 36:41:30 IHF collection of Ingrian sound recordings 480 Finno-Ugric Khanty 8:11:24 (see Rozhanskiy & Markus 2019 in this volume) Finno-Ugric Inari Saami 7:16:23 Table 2. Series of sound recordings in the AEDKL. Finno-Ugric Erzya 6:16:04 Finno-Ugric Udmurt 5:17:03 In the online database the metadata include information about Finno-Ugric Komi 3:39:52 the participants, recording time and place, content of the recording, Finno-Ugric Hungarian 2:28:44 and also technical details, e.g., the type of recording equipment and resolution of digital files. Finno-Ugric Moksha 0:24:22 Samoyedic Kamas 11:00:38 2.3. Manuscripts Samoyedic Nenets 0:16:23 There are a total of 396 000 pages of written manuscripts in the ar­ Table 1. Uralic languages represented in the archives. chives (~ 270 000 pages are digitally available). Written materials Depending on the period to which the recordings date, they are found include student coursework and theses defended at the Institute of Es­ on reel-to-reel tapes, cassettes, or digital media. The first reel-to-reel tonian and General Linguistics, fieldwork diaries, transcriptions and tape recordings date back to 1959; between 1980 and 2000 most of the written notes on Estonian and other Finno-Ugric languages. The earli­ recordings are on cassettes, and digital recordings have been made from est written material (a description of the Coastal dialect) dates back 2000. Most of the reel-to-reel tapes and cassettes have been digitized. If to 1910. There are 16 series of materials in the archive (see Table 3). undocumented fieldwork materials from earlier periods are donated to Some of the manuscripts were lost during the fire in the univer­ the archives, these recordings are digitized and added to the database. sity main building in 1965. The most well-preserved manuscripts are The main series that contain sound recordings are presented in Table 2. the student papers on Estonian dialects and other Finno-Ugric lan­ The sound recordings contain mainly interviews on different sub­ guages. There are BA and MA theses on the Estonian language dating jects, such as biographic facts, descriptions of ethnographic household from 1946 or later and other Finno-Ugric languages dating from 1956 labor, customs, everyday life, traditional events, etc. There are also or later defended at the Institute of Estonian and General Linguistics linguistic questionnaires (e.g., phonetic questionnaires, different word of the University of Tartu (~ 215 000 pages). lists), folk songs and tales, lectures and seminars held at the university.

330 331 LIINA LINDSTRÖM, PÄRTEL LIPPUS & TUULI TUISK THE ONLINE DATABASE OF THE UNIVERSITY OF TARTU ARCHIVES ...

Series Content digitized (these include recordings made during fieldwork and at vari­ C various topics (including professors’ lecture notes and teaching ous university events). There are 73 hours of video recordings and in materials, phonetic data: palatograms and sonograms, etc.) the new version of the database viewing these videos is integrated into D BA and MA theses on Estonian language defended at the Institute the online archive system (see Section 2.5 for update information). of Estonian and General Linguistics at the University of Tartu (since 1946) 2.5. Using the AEDKL H student papers on sound systems of Estonian dialects K fieldwork questionnaires for collecting data on Estonian dialects The online user interface of the AEDKL database was launched in L student papers on Estonian lexicology 2012. It is available at ‹https://murdearhiiv.ut.ee/›. The metadata are LFS Livonian folklore collection by Oskar Loorits (copies from the arranged into a relational MySQL database, keeping the speaker in­ Estonian Literary Museum) formation and the different media information in separate cross-linked M student papers on Estonian morphology tables. In this way, even if there is more than one recording and/or text MKT phonetic transcriptions of texts used in the Corpus of Estonian transcription from the same consultant, then there is only a single data­ Dialects base entry containing the consultant’s personal information (name, MT Mihkel Toomse manuscript of Estonian dialects date of birth, etc.) linked to all the recording and manuscript entries P fieldwork diaries from Estonian dialect expeditions where (s)he is participating. S student papers on various topics, mainly from the 1920s and 1930s In 2019, updating of the online database was finished and the SUD BA and MA theses on Finno-Ugric languages defended at the user interface was renewed. The whole database was relocated and the University of Tartu Institute of Estonian and General Linguistics software was upgraded. As a result, viewing the videos is integrated SUKD seminar papers and theses on Finno-Ugric languages from 1956– on the online archive system, searching results appear on the map, 2011 sound files are presented withHTML5 player, etc. SUPP fieldwork diaries from Finno-Ugric language expeditions The user should note that only the user interface has an English T transcriptions of Estonian dialects translation, which means that only the field names are translated. The Tx dialectology exams database is monolingual with most entries in Estonian. Instructions for Y student papers describing particular Estonian dialect areas searching in the database and a small Estonian-English dictionary are Table 3. Written materials found in the archives. presented on the homepage (https://murdearhiiv.ut.ee/abi.php?t=otsi).

2.4. Photos and videos 2.5.1. Simple search

The collection holds about 3 000 photos from fieldwork expeditions A simple search in the database is carried out on all fields of the data­ and linguistic events (e.g., conferences and seminars). Photos are di­ base. The results give all database entries containing the search term vided into two series based on media type: paper and digital photos. within any field. For example, if one searches for “kala” (‘fish’), one There are around 1 300 paper photos that are digitized, and digitiza­ gets over 900 matches (see Figure 1), where the search term can either tion is still in progress. Around 1 700 digital photos are from recent be found within the topic or the name of the speaker. Of course, this years of fieldwork and different linguistic events. search also returns irrelevant matches, for example where a field con­ Video recordings are from fieldwork conducted during re­ tains the word “foneetikalabor” (‘phonetics lab’). In this case it would cent years. Also, old film rolls from the 1970s and 1980s have been be better to use the detailed search option.

332 333 THE ONLINE DATABASE OF THE UNIVERSITY OF TARTU ARCHIVES ...

If the search term consists of multiple words, it is split by its spaces and all individual words are searched within all fields of the database separately. For example, if one searches for all the work in­ volving Prof. Paul Ariste, one simply searches for “Paul Ariste”. This will give over 400 matches where either there was a field containing the string “Paul Ariste” or there was one field containing “Paul” and another containing “Ariste”.

2.5.2. Detailed search

In case the simple search gives too many irrelevant results, the more detailed search provides another option. First of all, one has to se­ lect whether to search within audio or video recordings, photos, or manuscripts. Secondly, it is necessary to know a little about the da­ tabase structure to select the database fields to search. In the detailed search option, the search is carried out only on the specified fields. For example, if one wants to find all the recordings of Votic made by Prof. Ariste in the year 1975, one has to 1) select “Search audio track” and enter 2) Language: “vadja” 3) Recording time: “1975” to “1975”, 4) First name: “Paul”, 5) Last name: “Ariste” (see Figure 2).

2.5.3. User access

The AEDKL user interface has two types of database users (see Table 4). An anonymous user without a user account has limited ac­ cess to the database while the authorized users have full access to the archives. Information for obtaining a user account can be found at the AEDKL website ‹https://murdearhiiv.ut.ee/abi.php?t=otsi#parool›.

Anonymous user Authorized user Limited access Can see all database entries Can listen to recordings online Can download .wav files Can see reduced images Can download full size images Can see the name and village of Can see the full personal info of Figure 1. Example of results of a simple search with the search term “kala” (‘fish’). the consultants the consultant Table 4. AEDKL user rights.

334 335 THE ONLINE DATABASE OF THE UNIVERSITY OF TARTU ARCHIVES ...

3. The Corpus of Estonian Dialects

The Corpus of Estonian Dialects (CED) is a collection of electronic data containing authentic dialect texts from all Estonian dialects. This project was initiated in 1998 by the University of Tartu in cooperation with the Institute of the Estonian Language. Its main aim is to provide 1) access to a carefully chosen collection of accurately transcribed dia­ lect materials. These materials allow one to compare the phonological and grammatical features of Estonian dialects. An approximately equal amount of data is provided for each Es­ tonian dialect in the corpus; in addition to Estonian dialects, data from Votic and Livonian also are included. The CED consists of • sound recordings, • transcribed texts utilizing Finno-Ugric phonetic transcription, • dialect texts in simplified transcription, • morphologically annotated texts, • syntactically parsed texts, • a database containing information about consultants and record­ ings. Figure 3 illustrates the workflow of the CED: first, the sound record­ ings are phonetically transcribed, later these texts are converted to a simplified transcription and are morphologically annotated. After the 2) morphological annotation, they can be syntactically parsed. During this process, relevant information about speakers and recordings is 3) added to the metadata database.

Transcribing sound Simplified Morphological 4) recordings in Finno-Ugric transcrip­

annotation phonetic transcription tion 5)

Figure 2. Example of a detailed search within audio recordings of Votic recorded by Paul Ariste in 1975. The numbers 1–5 illustrate the steps explained in the text. Adding annotated Automatic Adding metadata texts to the online syntactic to the database database analysis

Figure 3. Workflow of the Corpus of Estonian Dialects.

336 337 LIINA LINDSTRÖM, PÄRTEL LIPPUS & TUULI TUISK THE ONLINE DATABASE OF THE UNIVERSITY OF TARTU ARCHIVES ...

3.1. Materials No. of recordings fromfrom different different decades decades 250250 The corpus is based on the materials of the AEDKL and on the materi­ als of the Institute of the Estonian Language, which has a large col­ lection of dialect recordings that are also transcribed in Finno-Ugric 200200 phonetic transcription. For more information about this collection, see Ermus et al. 2019 in this volume. The interviewed consultants were chosen on the basis of their social background: they are typically elderly people with little formal 150150 education, who have lived their entire lives in the same place in the countryside, and whose parents have these same characteristics. In the past, the dialect research tradition of Estonia has considered such con­ 100100 sultants to be good representatives of the old local dialect. Typically, the recordings contain traditional dialect interviews. In these a linguist interviews the consultant in surroundings familiar to 5050 him or her (the consultant’s home or backyard). The topics of the in­ terviews include life in earlier times, folk traditions, traditional work, the speaker’s biography, etc. 00 The materials in Votic and Livonian, however, differ from Es­ 1938 1956-19591956–1959 1960-19691960–1969 1970-19791970–1979 1980-19891980–1989 1990-20101990–2010 unknownunknown tonian dialects since these originate from more heterogeneous sourc­ Figure 4. Number of recordings from different decades. es. In addition to sound recordings, there are also earlier published texts (on Votic, Elna Adler: Vadjalaste endisajast (1960), Paul Ariste: Vadja­lane kätkist kalmuni (1974) and Vadja muistendeid (1977); on 3.3. Phonetic transcription Livonian, E. N. Setälä: Näytteitä liivin kielestä (1953)). In the first stage, the recordings have been transcribed using the Finno- 3.2. Sound recordings Ugric phonetic transcription, utilizing Unicode fonts and a virtual key­ board that allows for easily combining characters with (SUT, The oldest recordings of Estonian dialects date back to 1938 but the compiled by Esko Oja). The aim has been to transcribe the texts as ac­ majority of the interviews were recorded during the 1960s and 1970s curately as possible; the features of spontaneous speech (e.g., discourse (see Figure 4). Although older materials were recorded in the studio particles, corrections, repetitions), which traditionally have not been by Paul Ariste in 1938, the nature of the interviews is the same com­ considered important in dialect research, have been added to the texts. pared with later recordings. The speech of the interviewer has been transcribed as well. The Finno-Ugric phonetic transcription has been a standard for transcribing dialect texts for a long time in traditional Estonian dia­ lectology and most of the older transcriptions are available only in Finno-Ugric phonetic transcription. Since we have preferred the re­ cordings that were already transcribed by earlier researchers as much

338 339 LIINA LINDSTRÖM, PÄRTEL LIPPUS & TUULI TUISK THE ONLINE DATABASE OF THE UNIVERSITY OF TARTU ARCHIVES ... as possible, we also have preserved the transcription of these materi­ 3.5. Morphological annotation als. However, the Finno-Ugric phonetic transcription is problematic for many reasons: 1) it is not an international standard and therefore it As the speech of Estonian dialect varieties shows remarkable varia­ is hard to follow for many users; 2) it requires the use of specific non- tion on all linguistic levels, the morphological annotation cannot be standard fonts, a large number of diacritical marks and combinations done automatically. In the CED, it is done semi-manually, using the of different diacritics for the same character; 3) it is redundant for program Liivike (compiled by Külli Prillop). This program automati­ modes of linguistic analysis other than phonetics and phonology (e.g., cally creates an XML file which is later uploaded to the online corpus for analyzing dialect syntax). search engine ‹www.murre.ut.ee/mkweb›. The XML files can also be To solve the problem, we automatically convert phonetically downloaded and used independently from the online database. transcribed texts into a simplified transcription for futher use. For every word the following fields have been tagged: Sound recordings and transcriptions are stored at the AEDKL and can be searched via the AEDKL online database (Search: Manu­ • Word (sõne). The original form of the token as it occurs in the scripts; Series = MKT; the links to sound recordings can be found in text (in simplified transcription), e.g.,t’s’ibõrdõl’l’i ‘fidget’ (past the manuscript description field “Related materials”). sg 3), `vaesõq ‘poor’ (pl nominative), sääl ‘there’. • Keyword (märksõna). The keyword (lemma) as it occurs in the 3.4. Dialect texts in simplified transcription literary language, e.g., hüva ‘good’ (using standard orthography without vowel harmony). If the word occurs only in dialects, the All phonetically transcribed texts have been converted into a simpli­ dialect form is used as the keyword, e.g., tsiberdelema ‘to fidget’. fied transcription. There are some differences in the simplified tran­ If the same stem with the same meaning exists in standard Esto­ scriptions as compared to the standard ortography; for example, the nian, the standard Estonian word has been given as the keyword, geminate voiceless consonants in Quantity 2 words have been writ­ e.g., vaene ‘poor’, seal ‘there’. For Estonian verbs, the keyword ten with two letters (kattus ‘roof’, kadakkad ‘junipers’, in standard (lemma) form is the 2nd infinitive (e.g., õppima ‘to learn’), for Estonian ortography correspondingly: katus, kadakad), and the acute Votic and Livonian, the 1st infinitive (e.g., Votic: õppia ‘to learn’, accent inserted before the word helps to differentiate Quantity 3 words Livonian: oppõ ‘to learn’) is used. from Quantity 2 (the acute is added only to Quantity 3 words: `katta • Meaning (tähendus). This field is used only when the meaning ‘to cover’, `kas’si ‘cat, sg partitive’). of the word is different from the literary language or when there In the case of an ambiguously long quantity degree, the symbol * is no equivalent word in the literary language, e.g., tsiberdelema has been used (mainly in the Northeastern Coastal dialect group where ‘siplema’ (‘to fidget’). For Votic and Livonian, the field is always there is no distinction between Quantity 2 and Quantity 3, e.g., *kassi filled. The meaning is provided in Estonian. ‘cat’ in genitive or partitive). • Word class (sõnaliik). In the dialect corpus, the main aim of iden­ Palatalization is marked with an (e.g., palk ‘salary’ tifying word class has been to use a classification which would vs. pal’k ‘log’ ). be sufficiently understandable and detailed for researchers work­ A more detailed overview of simplified transcription principles ing with particular dialects, while also clear-cut enough for those can be found in Lindström 2015. who do the morphological tagging. Words are divided into 24 word classes according to their morphological inflections, syn­ tactic characteristics, and semantics. The classification is based on the system of word classes presented in Estonian grammars

340 341 LIINA LINDSTRÖM, PÄRTEL LIPPUS & TUULI TUISK THE ONLINE DATABASE OF THE UNIVERSITY OF TARTU ARCHIVES ...

(EKG I: 14–41). However, since the language in the dialect cor­ Word class Abbrev. Example pus is ­spoken, more subclasses are distinguished than in tra­ Noun (substantive) S kas’s ‘cat’, hommik ‘morning’ ditional grammars, referring to phenomena related to spoken Proper name H Jüri, Pärnumaa language use (discourse particles, communicative words). The Verb V ostma ‘buy’ word classes used in the CED are represented in Table 5. For more details on the issue of word classes in the dialect corpus see Auxiliary Va olema ‘be’ in compound tenses (oli olnud ‘have been’, oli tehtud ‘was done’) Lindström et al. 2006. Adverb Adv täna ‘today’, kiiresti ‘quickly’ • Morphological information (vorm). Morphological informa­ tion has been added to inflected words (nouns, verbs, pronouns, Verb particles Adva particles belonging to verbs, e.g., välja (mõtlema) ‘(think) out’, ära (jooksma) ‘(run) adjectives, numerals, etc.). away’ Modal adverb ModAdv ilmselt ‘probably’ * The headers of the morphologically annotated XML files include Numerals geographic information (longitude, latitude), as well as some addi­ tional data about the recording (year of the recording, names of the Cardinals Nump kaks ‘two’, viis ‘five’ recorders), and speakers (age, gender, year and place of birth, educa­ Ordinals Numj teine ‘second’, viies ‘fifth’ tion), if available. The recordings in mp3 format are linked with the Adjective A vana ‘old’, kole ‘ugly’ annotation. Pro-words All the texts are in XML format (UTF8) and can be used as an Pronoun ProS see ‘this, it’, too ‘that’, tema ‘(s)he’, mina ‘I’ independent source for different studies. For this purpose, a number of Proadjective ProA niisuke, sihuke ‘such’ scripts for R and Python have been developed (by Kristel Uiboaed), Proadverb ProAdv siin ‘here’, seal ‘there’, siis ‘then’ which enable one to search for various information, to manipulate data, to collect and handle frequency data, and to visualize the results Pronumeral ProNum mitu ‘several, many’ on maps. (See Uiboaed & Kyröläinen 2015, visualization options can Adpositions be seen also in Lindström et al. 2019 in this volume.) Postposition Post (maja) taga ‘behind the house’ The online database of the morphologically annotated texts is Preposition Pre pärast (hommiku­sööki) ‘after breakfast’ available at ‹www.murre.ut.ee/mkweb›; it is freely accessible and up­ Discourse particle Par noh, jah, no, oi dated regularly. The user interface of the search engine is only avail­ Communicative word Suht aitäh ‘thanks’, palun ‘please’, tere ‘hi’ able in Estonian. In order to search within the corpus, the following fields are relevant:Märksõna (keyword), Sõne (word), Sõnaliik (word Onomatopoetic word Ono mürts, pauh class), Vorm (morphological information), Tähendus (meaning), Keel Interrogative word Intr kas ‘whether’, kes ‘who’, mis ‘what’, kus (language), Murre (dialect), Murrak (parish, sub-dialect), Aasta (year ‘where’ of the recording), Vanus (age of the consultant), Sugu (gender of the Conjunction Konj ja ‘and’, et ‘that’ consultant). Negation word Mn ei, mitte ‘not’ Search results are presented as a table (see Figure 5). The results Comparative word Ms kõige (ilusam) ‘the most (beautiful)’ can be sorted by columns. All search results can be downloaded either Interjection Intj oh, oi ‘oh’ * in CSV or Excel format (buttons Lae CSV, Lae Excel). The results Table 5. Word classes in the Corpus of Estonian Dialects. (* = Not used for Estonian can be viewed in context and the original recording can be played by dialects.)

342 343 LIINA LINDSTRÖM, PÄRTEL LIPPUS & TUULI TUISK THE ONLINE DATABASE OF THE UNIVERSITY OF TARTU ARCHIVES ...

Dialect No. of words in phonetic No. of morphologically transcription annotated words Coastal 141 390 99 372 Northeastern 83 185 63 173 Insular 317 040 220 747 Western 369 716 265 489 Mid 347 981 249 065 Eastern 66 180 50 570 Mulgi 83 335 65 432 Tartu 113 351 81 101 Võru 167 312 112 700 Seto 126 156 85 114 South Estonian enclaves in 24 268 0 Latvia (Lutsi & Leivu) Estonian dialects, Total 1 839 914 1 292 763 Table 6. Number of phonetically transcribed words and morphologically anno- tated words in the CED (as of December 2019).

Figure 5. Search for the in the CED online search engine.

clicking on the button . The website has a simple Google map ap­ plication to visualize the results on a map (button Näita kaardil). Adding the original recordings to annotated texts started in 2014. By December 2019, 38% of morphologically annotated files (178 out of 467) have a sound recording available in the online database. The number of texts included in the CED has continued to increase but still there are “white spaces” without data on the map of Estonian parishes (see also maps in Lindström et al. 2019 in this volume). Table 6 shows the content currently found in the CED. In Map 1, the data points (villages where the materials of the corpus are collected) are shown. Map 1. Data points (villages) of the corpus materials (as of December 2019). Map data ©2015 Google, Maarja-Liisa Pilvik.

344 345 LIINA LINDSTRÖM, PÄRTEL LIPPUS & TUULI TUISK THE ONLINE DATABASE OF THE UNIVERSITY OF TARTU ARCHIVES ...

The data from Votic and Livonian come from more heterogene­ 4. Conclusions ous sources than the Estonian data (i.e., from spoken interviews and earlier published texts). Currently, the online database contains 34 331 The Archives of Estonian Dialects and Kindred Languages and the morphologically annotated words from Votic and 60 321 annotated Corpus of Estonian Dialects form a collection of materials in Estonian words from Livonian. and related languages that are freely available online. The AEDKL The morphologically annotated texts form the core of the corpus, consists of materials that have been systematically collected in field­ as they enable one to conduct different analyses on Estonian dialects work trips by various researchers over a long period. Collecting and as well as on Votic and Livonian. It is also the input for syntactic digitizing the data will go on and researchers continue to be encour­ parsing. aged to deposit their unarchived fieldwork materials in the AEDKL. The CED offers a balanced data set from the “golden age” of 3.6. Syntactically parsed texts Estonian dialectology, i.e., materials recorded mainly in the 1960s and 1970s. The sound recordings are annotated on various linguistic lev­ Syntactic parsing of the CED has been done automatically by using els: phonetic transcription, morphological annotation, and syntactic the parser developed by Kaili Müürisep. The parser is based on the parsing are provided. The corpus enables one to apply various meth­ Constraint Grammar framework and it was first developed for writ­ ods that are used in corpus linguistics and corpus-based dialectology, ten texts in Standard Estonian, later adapted for spoken Estonian and thereby opening up new horizons in the study of certain aspects of Estonian dialects. (Lindström & Müürisep 2009.) For dialect texts, it Estonian dialects such as dialect syntax. uses morphological annotation as an input; the rules of the parser were adapted to the morphological annotation and XML used in the CED. Importantly, clause boundaries also had to be added by the parser, as Acknowledgements they are missing in other layers of the corpus. (For more details, see Lindström & Müürisep 2009.) As a result, only part of the corpus is We would like to thank Eva Liina Asu for the first proofreading of this syntactically parsed (approx. 650 000 words). For searching within paper. The work of the AEDKL and CED is supported by the national parsed texts by combining syntactic and morphological information, a program “Estonian Language and Cultural Memory” (project “Database Python script is available. of Estonian Dialects and Kindred Languages II”), the national program It must be noted that after syntactic parsing was carried out on “Collections of Humanities and Natural Sciences”, and the European the CED in 2009–2010, both the technical basis of the parser and also Regional Development Fund (The Centre of Excellence in Estonian the XML of the corpus have changed, which means that the syntac­ Studies). tic parsing requires technical improvements in order to continue this work.

346 347 LIINA LINDSTRÖM, PÄRTEL LIPPUS & TUULI TUISK THE ONLINE DATABASE OF THE UNIVERSITY OF TARTU ARCHIVES ...

References musi eesti murrete korpuse põhjal [Questions on word categorization explored using the Corpus of Estonian Dialects]. – E[llen] Niit (ed.), Adler, Elna 1968: Vadjalaste endisajast I. Idavadja murdetekste [The Votes Keele ehe. Tartu Ülikooli eesti keele õppetooli toimetised 30. Tartu: in former times I. Dialect texts in eastern Votic]. Ed. M[erle] Leppik. Tartu Ülikool. 154–167. Tallinn: Eesti NSV Teaduste Akadeemia Keele ja Kirjanduse Instituut. Lindström, Liina & Kaili Müürisep 2009: Parsing corpus of Estonian dia­ Ahven, Heino 1955: Emakeele Seltsi tegevusest sõjajärgseil aastail (1945– lects. – E. Bick, K. Hagen, K. Müürisep & T. Trosterud (eds), Pro- ceedings of the NODALIDA 2009 workshop Constraint Grammar and 1954) [Activities of the Mother Tongue Society during the post-war robust parsing, Odense, Denmark; 14.05.2009. NEALT Proceedings years (1945–1954)]. – Emakeele Seltsi aastaraamat I: 5–26. Series. Tartu: Tartu University Library. Ariste, Paul 1974: Vadjalane kätkist kalmuni [Vote from cradle to grave]. Lindström, Liina, Maarja-Liisa Pilvik, Mirjam Ruutma & Kristel Uiboaed Emakeele Seltsi toimetised 10. Ed. T[iit]-R[ein] Viitso. Tallinn: Eesti 2019: On the use of perfect and pluperfect in Estonian dialects: Fre­ NSV Teaduste Akadeemia. quency and language contacts. – Sofia Björklöf & Santra Jantunen — 1977: Vadja muistendeid [Votic stories]. Emakeele Seltsi toimetised 12. (eds), Multi­lingual Finnic. Language contact and change. Uralica Hel­ Tallinn: Eesti NSV Teaduste Akadeemia Emakeele Selts, Valgus. singiensia 14. Helsinki: Finno-Ugrian Society. 155–193. Available EF 1997 = Eesti filoloogia poolsajand Teaduste Akadeemias [A half cen­ at: ‹https://doi.org/10.33341/uh.85035› tury of Estonian philology at the Academy of Sciences]. Tallinn: Eesti Rätsep, Huno 2003: Tartu Ülikooli eesti keele arhiivi saamisloost ja saa­ Keele Instituut. tusest [On the origin and destiny of the Estonian language archive at EKG I = Erelt, Mati, Reet Kasik, Helle Metslang, Henno Rajandi, Kristiina the University of Tartu]. – 200 aastat eesti keele ülikooliõpet. Juubeli- Ross, Henn Saari, Kaja Tael & Silvi Vare 1995: Eesti keele gram- kogumik. Tartu Ülikooli eesti keele õppetooli toimetised 25. 153–170. matika I. Morfoloogia. Sõnamoodustus [Estonian grammar I. Mor­ Rozhanskiy, Fedor & Elena Markus 2019: A new resource for Finnic lan­ phology. Word formation]. Tallinn: Eesti Teaduste Akadeemia Eesti guages: The outcomes of the Ingrian documentation project. – Sofia Keele Instituut. Björklöf & Santra Jantunen (eds), Multi­lingual Finnic. Language con- Erelt, Mati 2010: 90 years of the Mother Tongue Society. – Linguistica Ura- tact and change. Uralica Helsingiensia 14. Helsinki: Finno-Ugrian lica 46 (2): 144–153. Society. 303–326. Available at: ‹https://doi.org/10.33341/uh.85039› Ermus, Liis, Mari-Liis Kalvik & Tiina Laansalu 2019: The Archive of Uiboaed, Kristel & Aki-Juhani Kyröläinen 2015: Keeleteaduslike andmete Estonian Dialects and Finno-Ugric Languages at the Institute of the ruumilisi visualiseerimisvõimalusi [Spatial visualizing possibilities Estonian Language. – Sofia Björklöf & Santra Jantunen (eds), Multi­ for linguistic information]. – Eesti Rakenduslingvistika Ühingu aasta­ lingual Finnic. Language contact and change. Uralica Hel­singiensia raamat 11: 281–295. 14. Helsinki: Finno-Ugrian Society. 351–366. Available at: ‹https:// doi.org/10.33341/uh.85041› Kingisepp, Valve-Liivi 1967: Tartu Riikliku Ülikooli eesti keele kateedri murdearhiiv [The dialect archive of the Department of the Estonian language at the State University of Tartu]. – Kodumurre 8. Tallinn: Eesti NSV Teaduste Akadeemia, Emakeele Selts. 12–17. Lindström, Liina 2015: Ülevaade eesti murrete korpusest [Overview of the Corpus of Estonian Dialects]. Unpublished manuscript. Available at: ‹http://www.keel.ut.ee/sites/default/files/www_ut/emk_teejuht2015.pdf› Lindström, Liina, Liisi Bakhoff, Mari-Liis Kalvik, Anneliis Klaus, Rutt Läänemets, Mari Mets, Ellen Niit, Karl Pajusalu, Pire Teras, Kristel Uiboaed, Ann Veismann & Eva Velsker 2006: Sõnaliigituse küsi­

348 349 LIINA LINDSTRÖM, PÄRTEL LIPPUS & TUULI TUISK

Tartu Ülikooli eesti murrete ja sugulaskeelte LIIS ERMUS, MARI-LIIS KALVIK & digitaalne arhiiv ja Eesti murrete korpus TIINA LAANSALU Institute of the Estonian Language Liina Lindström, Pärtel Lippus & Tuuli Tuisk

Tartu Ülikooli eesti murrete ja sugulaskeelte arhiivi on koonda­ tud helisalvestised eesti murretest ja sugulaskeeltest, käsikirjalised­ The Archive of Estonian Dialects materjalid,­ keele kogumise ja keeleteadusega seotud fotod ning and Finno-Ugric Languages at the video­salvestused. Arhiiv sisaldab nelja tüüpi materjali: 1) helisal­ vestisi eesti murretest ja sugulaskeeltest (alates 1950. aastatest); 2) Institute of the Estonian Language käsi­kirjalisi materjale, sh 1920.–1980. aastatel eesti keele kateedris tehtud kursuse- ja seminaritöid ning lõputöid alates 1940. aastatest Abstract This report gives an overview of the materials in kuni tänapäevani,­ murdetekstide transkriptsioone ja kuuldelisi kirja­ the Archive of Estonian Dialects and Finno-Ugric Languages panekuid, murdepäevikuid jms; 3) keele kogumise ja keeleteadusega (AEDFUL) at the Institute of the Estonian Language (IEL). The seotud fotosid; 4) videosalvestusi. Arhiivi kogusid on digitaliseeritud AEDFUL holds the world’s largest collection of Estonian dialect alates 2000. aastast. Praeguseks on arhiivis u 2800 tundi helisalves­ examples as well as many other materials on other Finno-Ugric tusi, käsi­kirju u 396 000 lehekülge ja fotosid ligi 3000. Arhiivi maht languages. Materials in the AEDFUL have been collected by re­ kasvab iga-aastaste välitööde ja murdepraktika materjalide ning lõpu­ searchers from the IEL and the Mother Tongue Society during the 20th century. All the Estonian dialect areas as well as all of tööde lisandudes. Aastal 2012 valmis arhiivi veebipõhine andmebaas, the Finnic languages are represented in written and/or recorded mis asub aadressil ‹https://murdearhiiv.ut.ee/› ning on avatud kõigile form. Especially large amounts of language materials have been huvilistele ja uurijatele. collected for Livonian, Ingrian, and Votic. At the beginning of the Arhiivi eraldiseisev osa on Eesti murrete korpus – elektrooni­ 21st century when active collecting work ended, a new era began line andmekogu, mis sisaldab morfoloogiliselt märgendatud murde­ focused on digitization and making these materials publicly avail­ tekste ja on kasutatav iseseisvana. Selle põhieesmärk on teha hoo­ able. At present, electronic databases and dictionaries are avail­ able via the Internet. likalt valitud ja täpselt litereeritud materjalid kõigist eesti murretest uurijatele elektrooniliselt kättesaadavaks. Lisaks on murdekorpuse tekstid märksõnastatud ja morfoloogiliselt märgendatud. Märgenda­ tud korpus paikneb aadressil ‹www.murre.ut.ee/mkweb› ning sisaldab 1 241 233 tekstisõna eesti murretest, 34 331 vadja keelest ja 60 321 sõna liivi keelest. Korpuse põhjal on võimalik uurida võrdlevalt eesti murdeid nii häälikulisel, morfoloogilisel kui süntaktilisel tasandil ning võrrelda eesti murrete andmeid liivi ja vadja keelega.

Multilingual­ Finnic. Language 350 contact and change. 351–366. Uralica Helsingiensia 14. Helsinki 2019. ‹https://doi.org/10.33341/uh.85041› LIIS ERMUS, MARI-LIIS KALVIK & TIINA LAANSALU THE ARCHIVE OF ESTONIAN DIALECTS AND FINNO-UGRIC LANGUAGES AT THE INSTITUTE OF THE ESTONIAN LANGUAGE 1. Brief overview 2. Composition of the AEDFUL collections

The language collections of the Institute of the Estonian Language 2.1. Manuscripts (IEL; in Estonian, Eesti Keele Instituut, EKI) are located in the Ar­ chive of Estonian Dialects and Finno-Ugric Languages (AEDFUL; in The manuscript collection consists of scientific overviews, dialect tran­ Estonian, Eesti murrete ja soome-ugri keelte arhiiv, EMSUKA). scriptions, expedition diaries as well as dialect and ethnographical over­ In 1947, the Institute of Language and Literature of the Academy views written by researchers of the IEL and members of the MTS. The of Sciences of the Estonian SSR was established in Tallinn. In the same manuscripts are mostly descriptions or overviews of Estonian dialects year, a dialect archive was established based on the collections of the or of specific phenomena within dialects; there are also some works on Mother Tongue Society (MTS; in Estonian, Emakeele Selts) and the other Finno-Ugric languages (dissertations, scientific overviews, expe­ University of Tartu. These materials were collected during 1920–1946 dition diaries). There are altogether about 153 000 pages of manuscripts. and moved to Tallinn with the establishment of the new archive. In The main series are: 1947, systematic collection work was begun by the researchers of the • Dialect texts in Finno-Ugric phonetic transcription: series Department of Estonian Dialects and the Department of Finno-Ugric ES MT and KKI MT Languages of the IEL and the amount of archive materials increased. • Expedition diaries: series ES P and KKI P Later collections from different departments of IEL were incorporated • Overviews: series AES, ES, KKI (with subseries), SU, KT into the Institute’s archive: the Archive of Standard Estonian (Est. Ees- ti kirjakeele arhiiv; collected in the Department of Lexicology during Series marked with ES, AES, and KT contain works of MTS members 1955–2003), the Archive of Etymological Data (Est. Etümoloogiline (also works made under support or supervision of the MTS); series kartoteek; collecting started in the Department of Finno-Ugric Lan­ marked with KKI mostly contain works of IEL researchers. There are guages in 1970), the collection of the materials of related languages, also subseries under the KKI series containing copies of some dialect- etc. During the 1970s and 1980s the students of Tallinn Pedagogical related studies (bachelor’s and master’s theses) from the University of Institute (from 2005 to present Tallinn University) also gathered dia­ Tartu Archives of Estonian Dialects and Kindred Languages. lect materials during expeditions (mostly to the islands in western Es­ There are over 53 000 pages of dialect transcriptions in Finno- tonia and to Tartumaa in southern Estonia). Active collection work by Ugric phonetic transcription (FUPT), collected between 1916 and the IEL (for the archive) ended in the 1990s and most slip collections 1998. Older transcriptions were done by ear in real time during the were completed in 2002. By the turn of the century, the AEDFUL had consultation, but from 1956 onward most of the transcriptions are become the largest Estonian dialect archive in the world. based on sound recordings. Recording-based transcriptions are writ­ The main subcollections of the AEDFUL are the Archive of Es­ ten in full Finno-Ugric phonetic transcription, transcriptions done by tonian Dialects (Est. Eesti murrete arhiiv; 5.5 million slips with dia­ ear in real time are usually written in a simplified form. lect vocabulary and sentence examples and 2 345 volumes / 153 000 Expedition diaries (544 volumes with almost 11 000 pages) are pages, both co-owned by the IEL and the Mother Tongue Society), from 1921 to 1987 and mostly derive from dialect-collecting expe­ the Archive of Recordings of Estonian Dialects and Finno-Ugric Lan­ ditions; however, a small amount of this material is from Estonians guages (Est. Eesti murrete ja soome-ugri keelte heliarhiiv; 4 355 hours outside Estonia (e.g., in exile and in colonies) and from Finno-Ugric of original recordings), and the slip collection of dialect vocabulary of expeditions. These diaries also offer information about the places vis­ Finno-Ugric languages (281 500 slips). ited, the language consultants (sometimes with photos), and an over­ view of the local environment and circumstances.

352 353 LIIS ERMUS, MARI-LIIS KALVIK & TIINA LAANSALU THE ARCHIVE OF ESTONIAN DIALECTS AND FINNO-UGRIC LANGUAGES AT THE INSTITUTE OF THE ESTONIAN LANGUAGE A series of overviews (886 volumes, 56 000 pages) are written can be marked either on the front or back side of the slip. Data on slips by researchers of the IEL and members of the MTS, some univer­ are written by hand. sity dissertations are also included in these series. These works are The main dialect archives are usually organized by Estonian his­ mostly devoted to different aspects of Estonian dialects – phonetics, torical parishes excluding the Archive of Conceptual Lexis (in Estonian, morphology, syntax, and lexicon. In many texts written especially by Mõisteline sõnavarakogu) and the Aggregated Alphabetical Archive of MTS correspondents, aspects of everyday life, local traditions, and Dialect Vocabulary (Est. Murdesõnavara koondkartoteek). Most other ethnographical materials are reflected. slip archives are organized by the alphabetical order of the headwords. Works on other Finno-Ugric languages include the dissertations Dialect archives also include some smaller collections (around a few of the researchers of the Department of Finno-Ugric Languages of the thousand slips) like those of Jaan Jõgever, the Estonian Literature Soci­ IEL (for example, Arvo Laanest, Elna Pajusalu (Adler), Paul Kokla), ety, Elmar Wrager, Herman Sergo, the Archive of Grammatical Content, there are also some expedition diaries (e.g., the diaries of Arvo Laa­ etc. Dialect slips are mostly written in FUPT, except those of MTS cor­ nest’s Ingrian expeditions), grammatical overviews, manuscripts of respondents that are in simplified transcription. dictionaries, etc. Most written materials are about Ingrian and Kare­ The core of the dialect slip archives is Wiedemann’s Archive (Est. lian. Works are mostly in Estonian or Russian. Wiedemanni sõnavarakogu). The materials in this archive were col­ lected by MTS correspondents on the basis of F. J. Wiedemann’s Ehst- 2.2. Slip and card archives nisch-deutsches Wörterbuch (1869) during the period between 1922 The main slip archives at the AEDFUL are: and 1949. The basic archive contains 715 000 slips from 46 parishes from all dialect groups, additionally also material from the Lutsi and • Wiedemann’s Archive (in Estonian, Wiedemanni sõnavarakogu) Kraasna linguistic enclaves (located in Latvia and Russia, respective­ • Archive of Conceptual Lexis (Mõisteline sõnavarakogu) ly). The number of slips from one parish varies from a few thousand to • MTS Correspondents’ Archive of Dialect Vocabulary (Murde- 60 000. The biggest collections are from Hargla, Karksi, Kihelkonna, korrespondentide sõnavarakogu) Kodavere, Lüganuse, Martna, and Muhu parishes (each of them con­ • Additional Archive (Täiendav sõnavarakogu) tains over 20 000 slips). Wiedemann’s Archive also contains a sub­ • Aggregated Alphabetical Archive of Dialect Vocabulary (Murde­ archive of additional vocabulary (about 90 000 slips) composed of sõnavara koondkartoteek) lesser-known words that are not presented in Wiedemann’s dictionary. • Place Names Archive (Kohanimekartoteek) The Archive of Conceptual Lexis contains 150 000 slips arranged • Archive of Persons’ and Animals’ Names (Isiku- ja looma­nimede according to 48 conceptual areas. The idea of collecting material by kogu) conceptual areas came from Andrus Saareste. The material was gath­ • Archive of Finno-Ugric Languages (Soome-ugri keelte arhiiv) ered in the 1920s and 1930s and later was also published in Saareste’s • Etymological Archive (Etümoloogiline kartoteek) four-volume Eesti keele mõisteline sõnaraamat, Conceptual diction­ • Archive of Jargon (Argoosõnavara kogu) ary of Estonian (Stockholm, 1958–1968). • Archive of Standard Estonian (Eesti kirjakeele arhiiv) The Place Names Archive (Est. Kohanimekartoteek) contains • Archive of Terminology (Terminoloogiakartoteek) 650 000 text slips and small maps or charts of Estonian parishes and A typical slip collection at the AEDFUL consists of a number of paper towns (e.g., Tallinn, Valga) and also a small amount on Livonian slips with standard measurements of ca. 10 x 8 cm. Usually, the follow­ (60 slips) and Votic (957 slips) place names. The material is collected by ing types of data are found on the front side of the slips: dialect words MTS correspondents and IEL researchers. The Place Names Archive also in transcription, major morphological forms, and usage example(s) in contains the personal collections of Endel Varep and Gustav Vilbaste. sentence(s). Information about the collector and the time of collection Varep’s collection (over 111 000 sheets) is about settlement history; it 354 355 LIIS ERMUS, MARI-LIIS KALVIK & TIINA LAANSALU THE ARCHIVE OF ESTONIAN DIALECTS AND FINNO-UGRIC LANGUAGES AT THE INSTITUTE OF THE ESTONIAN LANGUAGE contains information about manors and villages. Vilbaste’s collection The second largest Finno-Ugric slip archive is the Ingrian archive. contains personal notes and excerpts from publications about certain It contains over 60 000 slips devoted to Ingrian vocabulary and grammar. parishes (Harju-Jaani, Jõelähtme, Kose, Kuusalu) and islands in north­ Most of the materials were gathered by Arvo Laanest on his expeditions ern Estonia. The material from the Place Names Archive is the basis for during 1960–1980. The vocabulary is partially divided into subarchives Eesti kohanimeraamat, the Dictionary of Estonian place names (KNR). by dialect area. The largest dialect vocabulary collection is from the Other larger dialect slip archives are the MTS Correspondents’ Hevaha dialect (almost 9 000 slips); collections from Oredež, Soikkola, Archive of Dialect Vocabulary (Est. Murdekorrespondentide sõna- and Lower Luga contain fewer than 2 000 slips each. The collection of varakogu; 400 000 slips, collected from 1940) and the so-called Addi­ Hevaha dialect material was used to compile Isuri keele Hevaha murde tional Archive (Est. Täiendav sõnavarakogu; 300 000 slips, collected sõnastik, Vocabulary of the Ingrian Hevaha dialect (Laanest 1997). between 1947 and 2002 by IEL researchers with the purpose of sup­ There are also about 10 000 slips of Karelian vocabulary collected plementing the existing collections of the MTS). by Konstantin Manzhin in the archive. In addition, Aime Kährik has com­ The Aggregated Alphabetical Archive of Dialect Vocabulary piled a systematic collection of southern Veps verbs (about 2 500 slips). (Est. Murdesõnavara koondkartoteek) gathers the copies of all basic Conceptual lexis collections of Finno-Ugric languages include parish-based dialect slip archives into an alphabetically arranged ar­ two collections: a collection of Votic and Ingrian plant names, col­ chive where material is organized according to headwords. Most of the lected by Gustav Vilbaste (over 4 000 slips), and a collection of Votic hand-copy work was done during 1951–1963. This type of slip catalog maritime vocabulary, collected by Tiit-Rein Viitso (over 2 000 slips). serves as a user-friendly base for both lexicographical and scientific The Etymological Archive (Est. Etümoloogiline kartoteek) was work. Based on the materials in this archive, Väike murdesõnastik, established in the 1970s and its intention was to gather data from all Dialectological dictionary of Estonian (VMS) I–II and Eesti murrete scientific publications about the etymologies of Estonian words. This sõnaraamat, Dictionary of Estonian dialects (EMS), and also many archive contains over 50 000 cards with etymological data; it was used other subdialect dictionaries have been compiled. The Aggregated Ar­ to compile Eesti etümoloogiasõnaraamat, Etymological dictionary of chive contains over 2.7 million slips and it is still actively used. Estonian (Metsmägi et al. 2012) and currently is being used to com­ The Archive of Finno-Ugric Languages (Est. Soome-ugri keelte pile a new, academic etymological dictionary, as well. arhiiv) contains over 281 500 slips and consists of archives of Votic, The Archive of Jargon (Est. Argoosõnavara kogu) contains about Ingrian, Karelian, and Veps. The largest of these is the Votic collection 50 000 slips, collected during 1964–2002 mostly by high school and (228 000 slips). university students. The majority of the collected material represents 24 000 slips of Votic vocabulary was inherited from the MTS language use of high school and university students; however, the jar­ when the Department of Finno-Ugric Languages was established. In gon of soldiers, prisoners, and some common language use from dif­ the 1950s, the IEL started its own collection. 204 000 slips were added ferent parts of Estonia is also found in this archive. based on the expeditions of the researchers of the Department of Finno- The addition of material to the Archive of Standard Estonian (Est. Ugric Languages. An especially large number of expeditions took place Eesti kirjakeele arhiiv) continued from 1955 until 2000. This archive during the years 1959–1965. Particularly important subcollections are contains approximately 4.3 million slips with vocabulary examples the collections gathered by Paul Ariste (44 100 slips) and Dmitri Tsvet­ mostly from Estonian literature. It was also the source used for com­ kov (18 500 slips of Jõgõperä vocabulary, collected in 1920). piling the Explanatory dictionary of Standard Estonian (first edition as Material from the Votic archive was used to compile the Dictio­ EKKS, 1988–2007, second edition as EKSS 2009). nary of Votic (VKS, 1959–2011), later published as a one‑volume dic­ The Archive of Terminology (Est. Terminoloogiakartoteek) con­ tionary (ed. Grünberg 2013). From 2019, the dictionary is also avail­ tains about 576 000 slips. It has been used for compiling various ter­ able at ‹http://www.eki.ee/dict/vadja/›. minology dictionaries. 356 357 LIIS ERMUS, MARI-LIIS KALVIK & TIINA LAANSALU THE ARCHIVE OF ESTONIAN DIALECTS AND FINNO-UGRIC LANGUAGES AT THE INSTITUTE OF THE ESTONIAN LANGUAGE

2.3. Sound archive Swedes living in Ukraine (in Gammalsvenskby). This subarchive con­ tains a total of 280 hours of recordings: from Siberia 130 h, Georgia The sound archive was established in 1956 with the aim of gather­ () 48 h, Russian Far-East 35 h, USA 28 h, 20 h, the ing tape recordings from all Estonian dialects and Finno-Ugric lan­ European part of Russia 15 h, Finland 3 h, Ukraine 1 h, and Gammal­ guages. Recordings were made during the IEL’s expeditions during svenskby 13 h. Materials from Russia and Georgia were recorded be­ 1950–1990. A typical dialect recording contains at least 30 minutes tween 1968 and 1992 and those from the USA and Canada in 1992; of monologue spoken by a previously selected speaker (typically an Estonians in Finland were recorded in 1965, Ukrainian recordings were older person with limited formal schooling whose family has been made in 2001, and recordings in Gammalsvenskby in 1982. The material living exclusively in that particular historical parish). The archive also recorded in areas within the former Russian Empire represent language contains material from the Estonian diaspora as well as contemporary use at the end of the 19th and the beginning of 20th century. In the USA speech. Systematic recording by the IEL ended in the 1990s as well and Canada, the speakers were Second World War refugees and their as other systematic collecting of dialect materials; however, sporadic descendants. There is no separate series for recordings of the Estonian recording continued until the beginning of the 2000s. A selection of diaspora; these recordings are filed either in the EMH or SUHK series. examples of the recordings is available for listening on the website of There are about 1 140 hours of recordings from 17 other Finno- the IEL’s sound archive ‹http://eki.ee/murded/fonoteek/›. Ugric languages: Livonian, Votic, Ingrian, Veps, Finnish, Karelian, Altogether there are about 4 360 hours of recordings. Most of Ingrian Finnish, Saami (all Saami languages together), Komi, Erzya, them are on reel-to-reel tapes and about 25% are also on cassettes. The Moksha, Mari, Mansi, Udmurt, Khanty, Selkup, and Nenets. Finnic main series are: language recordings make up 74% of all Finno-Ugric recordings. The largest amount of recordings is from Livonian (321 hours), followed • EMH(K): recordings of Estonian dialects (both on reel-to-reel by Votic (186 h), Ingrian (123 h), Veps (85 h), Finnish (53 h), Karelian tapes and cassettes) (50 h), and Ingrian Finnish (3 h). The majority of recordings were • SUH: Finno-Ugric languages on reel-to-reel tapes made during field expeditions in the 1960s, 1970s, and to a lesser ex­ • SUHK: Finno-Ugric languages on cassettes tent also in the 1980s. There are also recordings of Votic, Ter Saami, The 2 900 hours of recordings of Estonian dialects contain material from Livonian, and a small amount of Veps and Karelian that was recorded all Estonian historical parishes (except Paide parish), including the Leivu in the sound studio of the IEL. The most productive collectors of mate­ and Lutsi linguistic enclaves in Latvia. There are language examples from rial in related languages were Tiit-Rein Viitso and Arvo Laanest. Tiit- minorities in Estonia: Russian from Iisaku parish (8 hours) and Swedish Rein Viitso recorded most of the Livonian material; Arvo Laanest was from Vormsi, Noarootsi, and Naissaare (55 hours), recorded during 1956– the main collector of Ingrian but has also recorded Finnish and Votic. 2001. The oldest recordings of Estonian dialects are from 1938 (made by During the 1960s and 1970s, most of the original recordings Paul Ariste, originals on vinyl disks); the AEDFUL has restored copies from the EMH, SUH, and SUHK series were also copied onto back- on reel-to-reel tapes (made in 1986). There are on average 27 hours of re­ up tapes as well as tapes available for use by those utilizing the archive corded material from each parish, with the largest amounts from Kuusalu (about 3 900 hours); back-up tapes and tapes intended for general use parish (166 h), Vastseliina parish (162 h), and Setu parish (150 h). are all reel-to-reel tapes. Recordings from the Estonian diaspora include the following The sound archive also holds the ES and Ü series, which con­ countries or regions: Russia (Siberia as well as the Far-East and Euro­ tain common speech from the second half of the 20th century: mostly pean parts of Russia), Georgia (Abkhazia), USA, Canada, Finland, and speeches from language meetings, conferences and seminars, spoken Ukraine. In addition there are recordings from descendants of Estonian memoirs, etc.

358 359 LIIS ERMUS, MARI-LIIS KALVIK & TIINA LAANSALU THE ARCHIVE OF ESTONIAN DIALECTS AND FINNO-UGRIC LANGUAGES AT THE INSTITUTE OF THE ESTONIAN LANGUAGE

3. Digitization and access to digital materials sometimes include additional diary-like materials about the collection work itself or speaker interviews. Files are in groups of 50–200 files 3.1. Digitizing archive materials that are browsable 10 files at a time; it is also possible to look at the back sides of slips. The website of the Archive of Conceptual Lexis is Digitization of the AEDFUL’s sound archive subcollections began in available at ‹http://heli.eki.ee/moisteline/›. 1992 when the first sound recordings were copied onto DAT-cassettes. The Place Names Archive is divided by parish and then by sub­ From 1992 to 1999 about 335 hours of original tapes were copied to archive (the guide to the subarchives is given on the website). Within DAT format. Since 1999, digitized copies are stored both on the server each subarchive, files are arranged alphabetically by headword and are and on CD/DVD disks. Almost all original tapes are available in digital organized into groups of 50–200 files, which are browsable 10 files at form as are the back-up tapes of the Finno-Ugric series. a time; it is also possible to look at the back sides of slips. The website Active digitizing of slip collections began in 2011 when the of the Place Names Archive has a search engine. Indexing of slips IEL acquired a fast document scanner. At the beginning of 2019, the for the search engine database is currently in progress. As of October Archive of Conceptual Lexis, the Place Names Archive, and Wiede­ 2019, approximately 50% of all slips have been indexed: those from mann’s Archive have been digitized. Elmar Wrager’s and Herman Ambla to Pilistvere parishes, and those pertaining to Livonian and Sergo’s Hiiumaa vocabulary collections are also digitized. Votic place names. The search engine does not cover Endel Varep’s Digitization of manuscripts became large-scale some years later personal place names archive. The website of the Place Names Ar­ due to the following reasons. First, as sound recordings are less durable, chive is available at ‹http://heli.eki.ee/kohanimed/›. their digitization has been the most urgent. At first, paper materials were Digitization of Wiedemann’s Archive was completed in 2016. scanned only upon request. The second reason was the lack of a proper Digital files will be accessible by the end of 2019 at ‹http://emsuka. flat scanner. Acquiring a high-speed document scanner put the focus on eki.ee›. digitizing slip collections. From 2010 to 2013 around 1 000 pages of Since 2015, data from the AEDFUL have been accessible at manuscripts were scanned per year. In 2014, the IEL got a new high- ‹http://emsuka.eki.ee›. The addition of further data and digital files is in speed flat scanner and scanning rates went up to about 5 000 pages per progress. Since the beginning of 2017, the metadata for the manuscript year. From 2016 to 2018 about 20 000 pages were scanned every year. series and sound series have been inserted into the system and are avail­ As of October 2019, over 72 000 pages (over 1 800 volumes) have been able for searching. Manuscript files are in PDF or PDF/A format, sound digitized. All digitized files will be added to the electronic database. files are in MP3 (for unregistered users) and WAV (for registered users) format. By the beginning of June 2019, about 4 400 sound objects and 3.2. Electronic databases over 1 800 PDF objects had been added to the system and are available along with their metadata. Currently, digitized materials of the Archive of Conceptual Lexis and the Place Names Archive are accessible online. The associated web 3.3. Electronic dictionaries pages are in Estonian. On these web pages, the slips are divided as they are in the original archives. One of the first uses of the AEDFUL’s materials has been tocom­ The Archive of Conceptual Lexis is first divided by concept pile and edit dialect dictionaries. Most of these have been compiled areas. In one concept area, files are organized according to their loca­ and edited using the IEL’s electronic environment EELex. The IEL’s tion in the catalogs: first by drawers, then by parishes. Files are organ­ policy is to make dictionaries available online. By June 2019, nine of ized alphabetically or by individual collections. Individual collections such electronic dictionaries have been made accessible.

360 361 LIIS ERMUS, MARI-LIIS KALVIK & TIINA LAANSALU THE ARCHIVE OF ESTONIAN DIALECTS AND FINNO-UGRIC LANGUAGES AT THE INSTITUTE OF THE ESTONIAN LANGUAGE

Väike murdesõnastik, Dialectological dictionary of Estonian Kihnu sõnaraamat, Dictionary of the Kihnu dialect (Leas et al. (VMS), was published in two volumes 1982–1989. It contains 73 397 2016), was published in the end of 2016. It presents the vocabulary of the dialect words, which are or will be presented in the EMS (see be­ small Kihnu island and is available at ‹http://www.eki.ee/dict/kihnu/›. low). It contains headwords, their geographical distribution, and basic Kuusalu rannakeele sõnaraamat, Dictionary of the Coast dialect meanings for each entry. The website ‹http://www.eki.ee/dict/vms/› of Kuusalu (Norvik & Kendra 2019), presents dialect vocabulary from contains an application that shows the geographical distribution of any parts of the historical parish of Kuusalu: the coastal area from Kaber­ given word on a map of Estonia. neeme to Salmistu and the peninsulas of Juminda and Pärispea. The Eesti murrete sõnaraamat, Dictionary of Estonian dialects (EMS), dictionary is available at ‹http://www.eki.ee/dict/kuusalu/›. is the largest and most complete dictionary of Estonian dialects. It has Eesti etümoloogiasõnaraamat, Etymological dictionary of Esto­ been published as a series of separate subvolumes (each about 200 pages) nian (Metsmägi et al. 2012), is based on the word stems presented in since 1994; in November 2019, the 30th sub­volume was published. By Eesti õigekeelsussõnaraamat ÕS 2006, The Dictionary of Standard the end of 2019, the electronic dictionary ‹http://www.eki.ee/dict/ems/› Estonian. This dictionary is based on the Etymological Archive of the contains material from subvolumes 1–30. IEL. The dictionary compares Estonian words and their meanings with Vanapärase Võru murde sõnaraamat, Dictionary of the old Võru their Finnic and Finno-Ugric cognates. The Etymological dictionary dialect (Käsi 2011), presents the old dialect vocabulary from the par­ of Estonian is available at ‹http://www.eki.ee/dict/ety/›. ishes of Rõuge, Vastseliina, and Setu. The dictionary is available as a Eesti kohanimeraamat, the Dictionary of Estonian place names PDF file at ‹http://www.eki.ee/dict/Voru_murde_sonaraamat.pdf›. (EKNR), is the first compendious work on Estonian place names. Mulgi sõnastik, Vocabulary of the Mulgi dialect (Laande & There are over 6 200 place name articles in the dictionary, containing Todesk 2013), gathers dialect vocabulary from all historical parishes etymological, geographical, and historical information. The dictionary of Mulgimaa: Tarvastu, Paistu, Halliste, Karksi, and Helme. It is avail­ is available at ‹https://www.eki.ee/dict/knr/›. able at ‹http://www.eki.ee/dict/mulgi/›. Hiiu sõnaraamat, Dictionary of the Hiiumaa dialect (Kokla 3.4. Corpus of Estonian dialects 2015), gathers dialect vocabulary from all historical parishes of Hiiu­ maa: Emmaste, Käina, Reigi, and Pühalepa. It is available at ‹http:// The Corpus of Estonian dialects is compiled in collaboration between www.eki.ee/dict/hiiu/›. the IEL and the Institute of Estonian and General Linguistics at the Idamurde sõnastik, Vocabulary of the Eastern dialect (Pall University of Tartu. It was established in 1998 at the University of 2016), is the new edition of Valdek Pall’s Idamurde sõnastik, Vo­ Tartu. The corpus contains phonetically transcribed texts (both in cabulary of the Eastern dialect from 1994 (Pall 1994). It gathers dia­ FUPT and simplified transcription) as well as sound files from all Es­ lect vocabulary from the historical parishes where the Eastern dialect tonian dialect areas. The corpus also contains morphologically and was spoken: Iisaku (southern part), Torma, Kodavere (northern part), syntactically tagged texts. In addition to Estonian dialects, there is also Maarja-Magdaleena, Palamuse, Äksi, and Laiuse. It is available at material from the Leivu and Lutsi dialect areas in Latvia as well as ‹http://www.eki.ee/dict/ida/›. materials in Votic and Livonian. The website of the Corpus of Esto­ Seto sõnastik, Vocabulary of the Setu dialect (Käsi 2016) is based nian dialects is available at ‹http://www.murre.ut.ee/murdekorpus/›. It on the Dictionary of the old Võru dialect (Käsi 2011). It gathers vo­ is possible to use a search engine to search the contents of the website cabulary from Setu parish, mostly from the eastern part of the parish. ‹http://www.murre.ut.ee/mkweb/›. For more information about the It is available at ‹http://www.eki.ee/dict/setosonastik/›. corpus, see Lindström et al. (2019) in this volume.

362 363 LIIS ERMUS, MARI-LIIS KALVIK & TIINA LAANSALU THE ARCHIVE OF ESTONIAN DIALECTS AND FINNO-UGRIC LANGUAGES AT THE INSTITUTE OF THE ESTONIAN LANGUAGE

References Leas, Reene, Reti Könninge, Silvi Murulauk & Ellen Niit 2016: Kihnu sõnaraamat [Dictionary of the Kihnu dialect]. Karl Pajusalu & Jüri EKKS = Eesti kirjakeele seletussõnaraamat [Explanatory dictionary of Viikberg (eds). Tallinn: Eesti Keele Instituut, Kihnu Kultuuri Instituut, Standard Estonian] 1–7. 1988–2007. Rudolf Karelson, Valve Kullus Tartu Ülikooli eesti ja üldkeele­teaduse instituut & Eesti Keele Sihtasu­ (Põlma), Erich Raiet, Mai Tiits, Tiia Valdre & Leidi Veskis (eds). Tal­ tus. Available at: ‹http://www.eki.ee/dict/kihnu› linn: Eesti Keele Instituut. Lindström, Liina, Pärtel Lippus & Tuuli Tuisk 2019: The online database EKNR = Eesti kohanimeraamat [Dictionary of Estonian place names]. 2016. of the University of Tartu Archives of Estonian Dialects and Kindred Marja Kallasmaa & Peeter Päll (eds). Tallinn: Eesti Keele Instituut, Languages and the Corpus of Estonian Dialects. – Sofia Björklöf & Võru Instituut, Eesti Kirjandusmuuseum & Eesti Keele Sihtasutus. Santra Jantunen (eds), Multi­lingual Finnic. Language contact and Available at: ‹https://www.eki.ee/dict/knr/› change. Uralica Helsingiensia 14. Helsinki: Finno-Ugrian Society. EKSS = Eesti keele seletav sõnaraamat [Explanatory dictionary of Esto­ 327–350. Available at: ‹https://doi.org/10.33341/uh.85040› nian] 1–6. 2009. Margit Langemets, Mai Tiits, Tiia Valdre, Leidi Metsmägi, Iris, Meeli Sedrik & Sven-Erik Soosaar 2012: Eesti etümo­ Veskis, Ülle Viks & Piret Voll (eds). Tallinn: Eesti Keele Instituut & loogiasõnaraamat [Etymological dictionary of Estonian]. Tallinn: Eesti Keele Sihtasutus. Available at: ‹http://www.eki.ee/dict/ekss/› Eesti Keele Instituut & Eesti Keele Sihtasutus. Available at: ‹http:// EMS = Eesti murrete sõnaraamat [Dictionary of Estonian dialects] 1–30 www.eki.ee/dict/ety/› (a–osatama). 1994–2019. Anu Haak, Evi Juhkam, Mari-Liis Kalvik, Norvik, Piret & Heli Kendra 2019: Kuusalu rannakeele sõnaraamat [Dic­ Mari Kendla, Tiina Laansalu, Varje Lonn, Helmi Neetar, Ellen Niit, tionary of Kuusalu coastal language]. Vilja Oja (ed.). Tallinn: Eesti Piret Norvik, Vilja Oja, Valdek Pall, Eevi Ross, Aldi Sepp, Mari-Epp Keele Instituut, Juminda Poolsaare Selts & Emakeele Sihtasutus. Tirkkonen & Jüri Viikberg (eds). Tallinn: Eesti Keele Instituut & Eesti Available at: ‹http://www.eki.ee/dict/kuusalu/› Keele Sihtasutus. Available at: ‹http://www.eki.ee/dict/ems/› ÕS 2006 = Eesti õigekeelsussõnaraamat ÕS 2006 [Dictionary of Standard Grünberg, Silja (ed.) 2013: Vadja keele sõnaraamat [Dictionary of Votic]. 2., Estonian]. Tiiu Erelt, Tiina Leemets, Sirje Mäearu & Maire Raadik täiendatud ja parandatud trükk. Tallinn: Eesti Keele Instituut & Eesti (eds). Tallinn: Eesti Keele Sihtasutus. Keele Sihtasutus. Available at: ‹http://www.eki.ee/dict/vadja/› Pall, Valdek 1994: Idamurde sõnastik [Vocabulary of the Eastern dialect]. Käsi, Inge 2011: Vanapärase Võru murde sõnaraamat [Dictionary of the old Tallinn: Eesti Keele Instituut. Võru dialect]. Helmi Neetar (ed.). Tallinn: Eesti Keele Instituut & Eesti — 2016: Idamurde sõnastik [Vocabulary of the Eastern dialect]. 2., Keele Sihtasutus. Available at: ‹http://www.eki.ee/dict/Voru_murde_ täiendatud ja parandatud trükk. Ellen Niit (ed.). Tallinn: Eesti Keele sonaraamat.pdf› Instituut & Eesti Keele Sihtasutus. Available at: ‹http://www.eki.ee/ — 2016: Seto sõnastik [Vocabulary of the Seto dialect]. Maeve Leivo, dict/ida/› Nele Reimann & Karl Pajusalu (eds). Tallinn: Eesti Keele Instituut. Saareste, Andrus 1958–1968: Eesti keele mõisteline sõnaraamat [Concep­ Available at: ‹http://www.eki.ee/dict/setosonastik/› tual dictionary of Estonian] 1–4. Stockholm: Vaba Eesti. Kokla, Paul 2015: Hiiu sõnaraamat [Dictionary of the Hiiumaa dialect]. VKS = Vadja keele sõnaraamat [Dictionary of Votic] 1–7. 1990–2011. Elna Meeli Sedrik (ed.). Kärdla: Eesti Keele Instituut, MTÜ Hiiu Öko & Adler, Merle Leppik & Silja Grünberg (eds). Tallinn: Eesti Keele Eesti Keele Siht­asutus. Available at: ‹http://www.eki.ee/dict/hiiu/› Instituut. Laande, Alli & Triin Todesk 2013: Mulgi sõnastik [Vocabulary of the Mulgi VMS = Pall, Valdek (ed.) 1982, 1989. Väike murdesõnastik [Dialectologi­ dialect]. Karl Pajusalu & Urmas Sutrop (eds). Abja-Paluoja–Tallinn– cal dictionary of Estonian] I–II. Tallinn: Valgus. Available at: ‹http:// Tartu: Mulgi Kultuuri Instituut, Eesti Keele Instituut, Tartu Ülikooli www.eki.ee/dict/vms› eesti ja üldkeeleteaduse instituut & Eesti Keele Sihtasutus. Available Wiedemann, Ferdinand Johann 1869: Ehstnisch-deutsches Wörterbuch. St.- at: ‹http://www.eki.ee/dict/mulgi/› Petersburg: Buchdruckerei der Kaiserlichen Akademie der Wissen­ Laanest, Arvo 1997: Isuri keele Hevaha murde sõnastik [Vocabulary of the schaften. Ingrian Hevaha dialect]. Tallinn: Eesti Keele Instituut.

364 365 LIIS ERMUS, MARI-LIIS KALVIK & TIINA LAANSALU

Eesti murrete ja soome-ugri keelte arhiiv ANNELI SARHIMAA Eesti Keele Instituudis University of Mainz

Liis Ermus, Mari-Liis Kalvik & Tiina Laansalu

Eesti murrete ja soome-ugri keelte arhiiv (EMSUKA) koondab endas Finnic data sets in the Eesti Keele Instituudi (varasema Keele ja Kirjanduse Instituudi) uuri­ ELDIAdata databank jate ja Emakeele Seltsi liikmete poolt kogutud keelematerjale nii eesti murrete kui ka soome-ugri keelte kohta. EMSUKA sisaldab käsikirju,­ sedelkartoteeke ja helisalvestisi. EMSUKA murdearhiiv, mis on maa­ ilma suurim eesti murdekeelt käsitlev kogu, on olnud paljude murde­ 1. Introduction sõnastike ja teaduslike murdeülevaadete aluseks. Ka soome-ugri keel­ te materjalide põhjal on koostatud nii sõnastikke kui ka teaduslikke The ELDIAdata is a digital databank containing all empirical materi­ ülevaateid. als collected during the EU FP-7 research project ELDIA (European Arhiivis on tallel ka etnograafilisi kirjeldusi ning fotosid. Heli­ Language Diversity for All). ELDIA was an international, interdis­ arhiivis leidub salvestisi kogu Eesti murdealalt ja samuti pea kõigist ciplinary project coordinated by the University of Mainz, , teistest soome-ugri keeltest, sealhulgas kõigist läänemeresoome keel­ and conducted by experts in applied linguistics and sociolinguistics, test ning ka Eestist välja rännanute keelest. law, social studies, and statistics in 2010–2013. The overarching aim Suur osa EMSUKA materjale on digiteeritud ning praegu on töös was to contribute to a better understanding of how local, national, and nende veebi kaudu kättesaadavaks tegemine, et kõik soovijad neid va­ international (vehicular) languages interact in contemporary Europe, balt interneti teel kasutada võiksid. Samuti saab Eesti Keele Instituudi and to enhance the reconceptualisation, re-evaluation, and promotion veebilehe kaudu kasutada järjest suuremat hulka EMSUKA materjali­ of individual and societal multilingualism. del põhinevaid murdesõnastikke. The empirical data were collected in eight countries from speak­ ers of thirteen Finno-Ugric minority languages, ten of which belong to the Finnic group: Meänkieli, Kven, Finnish in Sweden, Estonian in Finland, Estonian in Germany, Karelian in Finland, Karelian in Rus­ sia, Veps, Võro, and Seto. In addition to the primary study popula­ tions, that is, the minority groups whose languages were at stake, a control group representing all other citizens of the countries where the investigated minorities live was surveyed and interviewed, as well. The most concrete goal of ELDIA was to create a European Lan­ guage Vitality Barometer (EuLaViBar). The barometer is a tool for mea­ suring the level of language vitality and it helps identify areas within which the assessed language requires more societal support. At the most general level, it serves as a testable model for assessing not only the vital­ ity but also the social significance of different languages in multilingual

Multilingual­ Finnic. Language 366 contact and change. 367–377. Uralica Helsingiensia 14. Helsinki 2019. ‹https://doi.org/10.33341/uh.85042› ANNELI SARHIMAA FINNIC DATA SETS IN THE ELDIADATA DATABANK communities. The barometer also can be applied to other languages be­ Number of question- Number of question- Response rate Number of Interview material yond those in the Finno-Ugric family and those spoken in Europe. naires distributed naires returned in % interviews (hours:minutes) The results of ELDIA have been published in case-specific reports, Target Control Target Control Target Control in an overview report (Laakso et al. 2013), and as a toolkit supervising Case study group group group* group* group group Indiv. Group Indiv. Group Meänkieli end users in creating and using the barometer (Spiliopoulou Åkermark 941 554 58.87 7 8 07:15 13:11 in Sweden et al. 2013). The reports and the toolkit are available online on the 1 000 227 22.7 Finnish 1 000 369 36.9 – – – – project website. A book placing the project results in wider academic in Sweden and language-political contexts appeared in 2016 (Laakso et al. 2016). Kven in 1 500 1 000 85 107 5.67 10.7 8 8 04:45 10:10 Veps in Russia 301 299 99.34 7 6 05:51 05:31 Karelian 302 302 100 2. An overview of the Finnic data sets 301 296 98.34 6 6 04:16 07:17 in Russia Karelian 1 034 356 34.43 8 8 08:33 12:13 The ELDIAdata databank contains all the statistical data collected in Finland 800 146 18.25 with a large-scale questionnaire survey as well as all interviews car­ Estonian in 800 170 21.25 8 8 09:58 11:18 ried out with speakers of the investigated minority languages and with Finland Estonian in representatives of the control groups. The databank consists of two 420 none 71 – 16.9 – 8 3 13:03 05:33 major parts: the minority language target group database and the con­ Germany Seto in Estonia 418 294 70.33 8 8 06:38 06:20 trol group database. Table 1 summarizes the Finnic data sets. 1 000 363 36.3 As Table 1 indicates, joint control group survey questionnaires Võro in Estonia 409 296 72.37 8 8 05:54 07:56 were used for minorities that were studied in the same country, that is, In total 7 124 4 102 2 790 1 145 51.44 37.59 68 63 65:12 80:04 for Meänkieli and Finnish in Sweden, for North Saami and Kven in Norway, for Karelian and Estonian in Finland, for Karelian and Veps in Table 1. The Finnic data sets within the ELDIAdata databank Russia, and for Seto and Võro in Estonia. The data sets from Sweden and Germany are not complete. The case study concerning Finnish in Sweden had to be given up due to organisational and other problems in 3.1. Data collection tools 2011, and so only survey data were collected there. In Germany, con­ trasting a couple of thousand Estonians who live scattered all over the The new data were collected using the following tools: country with the 82 million other inhabitants would have been utterly senseless, and so no control group data were collected there. • a unified survey questionnaire designed for the minorities; • a unified survey questionnaire designed for the control groups; • a semi-structured matrix for focus group interviews with minor­ 3. How the data were collected ity stakeholder and speaker groups; • a semi-structured matrix for focus group interviews with selected The empirical data collection in ELDIA aimed at accumulating new representatives of the control groups consisting of politicians, information on the investigated Finno-Ugric minority languages in a authorities, and representatives of media; systematic manner for the purposes of developing the barometer. An­ • a semi-structured matrix customised to be case-specific for inter­ other aim was to fill gaps in the existing research with the help of a views with individual speakers of the minority language at issue. multilingual corpus containing statistical and interview data. 368 369 ANNELI SARHIMAA FINNIC DATA SETS IN THE ELDIADATA DATABANK

The questionnaire survey among the investigated minority communi­ research gaps that had been identified in ELDIA desk research. The origi­ ties served as the main means for collecting new information on the nal plan was to conduct eight individual interviews with members of the current use and the current state of the investigated minority languag­ minority, eight focus group interviews with members of the minority, es. The survey sought to provide a broad and general insight into the and three focus group interviews with representatives of society at large. state and the use of the investigated languages in a way that also would The eight individual interviews and the eight focus group interviews facilitate generalizable comparisons between the minority groups. with members of the minorities were carried out in five age groups (18– The minority language questionnaire was structured around the 29, 30–49, 50–65, and over 65). In order to obtain a wider perspective following main topics: on the current parental generation, the 30–49 age group was divided into two groups with a group of women and a group of men interviewed sepa­ Part A. Personal background information on the respondent rately. According to the project design, the three minority focus groups Part B. Background information on the respondent’s language use should have covered (i) minority politicians and civil servants belonging Part C. The respondent’s language skills (minority language, majority to the investigated minority group, (ii) minority activists and (iii) repre­ language, other languages) sentatives of the minority media. However, in several ELDIA case studies Part D. The respondent’s use of minority, majority, and other it only was possible to create a separate group composed of activists while language(s) in different domains the other two groups had to be combined. In some case studies it was only Part E. Language attitudes and the respondent’s desire to use languages possible to find people for the age-based focus groups while the other Part F. Public language use vs. private language use focus group interviews had to be omitted completely. Part G. Consumption and active use of languages by the respondent in The original data collection design included two focus groups different media involving representatives of the control group: one for politicians and There were 63 questions in the minority questionnaire, but as there civil servants dealing with (minority-)language matters, and one for was a high number of sub-questions, the total number of questions representatives of the majority-language media. The case study focus­ was well over 300. The statistical datasets included in the databank ing on Estonian in Germany did not involve any control group, there­ contain only variables derived from the closed questions of the survey fore, these interviews were not conducted. In some case studies such questionnaire; the variables total exactly 340. as Hungarian in Austria and for Kven and North Saami in Norway, The control group survey contained the same primary topics as there were problems in getting enough interviewees, especially for the the minority language survey, with the exception that Part F “Public focus group interview with politicians and civil servants. language use vs. private use” was omitted. The contents of the control All focus group interviews followed a joint thematic interview group questionnaire were largely the same as those of the minority template that was modified to meet the case-specific needs. The main group questionnaire; however, some questions not relevant for major­ themes were the interviewee’s bilingualism or multilingualism, their ity language speakers were omitted and some questions were formu­ use of different languages in everyday life and views on bilingualism lated differently. The control group survey questionnaire included 47 or multilingualism as being either an asset or a problem. The inter­ questions. Again, due to the large number of sub-questions, the vari­ viewees were also asked about their perception of the term ‘minority’ ables in the control group statistical data set total 280. and their opinion of the role of language in an individual’s identity; The ELDIA interview design aimed primarily at completing the sur­ for example, whether knowing the minority language is necessary for vey data with in-depth qualitative insights into the themes covered by the identification as a member of a given minority. Yet another issue was survey. The interviews also offered the researchers the possibility of col­ the perception of the investigated minority and its language by the lecting information that would shed some new light on the case-specific society, the role of the investigated minority language in the society at

370 371 ANNELI SARHIMAA FINNIC DATA SETS IN THE ELDIADATA DATABANK issue as well as the role of societal diversity in general. Special atten­ study to assure the maximal comparability of the obtained data. Ultimate­ tion was also drawn to mapping the interviewees’ opinions concerning ly, random sampling could be applied in most cases in one form or anoth­ the responsibility of society in maintaining and revitalising the inves­ er. Yet, in the case studies concerned with Estonian in Germany, in those tigated minority language and on their views concerning the future of concerning the minority language groups in Russian Karelia and Norway the minority language in a ten-year perspective. as well as in the case study on Meänkieli speakers in Sweden, the obtained The individual interviews were conducted with one male and one data show a clear bias in favour of language activists as respondents. female per age group. The interviews were designed to collect in-depth In an ideal case, all study populations should have been divid­ information on four thematic fields: mother tongue; other languages; ed equally by gender and four rough age categories (18–29, 30–49, attitudes towards multilingualism; and languages and modernisation. 50–64, and 65+) with as much variation as possible in terms of the The interview format was semi-structured in the sense that the top­ demographic background of the respondents. In those case studies in ics were predefined but it was left to the interviewer to formulate the which the sampling frame was an official register, it was possible to questions in a way that will be experienced as maximally “natural” in obtain the intended division by gender and age, while in the majority the interview situation and to ask further questions as suitable in the of case studies this ideal had to be eased, as no official registers were communication situation. available for sampling. According to the initial interview design, the interviewees in the age cohort focus groups should have been selected 3.2. Principles for identifying survey mainly from those who had returned the survey questionnaire; the idea respondents and interviewees was to ask in the questionnaire whether the respondent will allow a further contact and is willing to give an interview. In practice, howev­ As explained earlier, concurrent sample surveys were conducted in er, this recommendation could be followed in only three case studies eight countries to obtain information on the target populations (minor­ (Seto, Võro, Meänkieli), and even in these only partially: in all case ity language groups) and the corresponding majority populations (con­ studies at least a part of the interviewees had not participated in the trol groups) in all other countries except Germany. Originally, the idea survey. One of the reasons for these deviations from the pre-planned was to obtain about 300 responses from each survey population. As procedure was that although there were survey respondents who had the non-response rates are high in mail surveys, the sample size had to announced their willingness to be contacted for an interview, they be inflated by the anticipated non-response. Not even these measures mostly belonged to the two oldest age groups. Furthermore, in most were always sufficient: as can be read in Table 1, in some cases the cases the participants of the focus group interviews were selected by final number of respondents remained clearly below the target. the responsible research teams, usually with the help of the minority Ideally, the sampling design should have been a true random sam­ organisations, and the selection was based on researchers’ scholarly ple in each country and study population. However, the project team was knowledge of the local circumstances and networks within the minor­ aware that in some countries there are no proper ways to carry out real ity and the majority communities. In some case studies, Facebook and probability sampling methods among the study populations. Those prob­ Myspace were used to complement the selection of interviewees. lems arise either from the lack of availability of comprehensive sampling When compiling the minority focus groups, participants with at frames like population registers, or legislation on disclosure control. All least a receptive (“passive”) command of the minority language were researchers involved in sample selection and data collection co-operated preferred. The control group interviewees were selected from people closely with the project statistician Kari Djerf (University of Helsinki) engaged in decision-making bodies, an additional criterion was that and with Karl Pajusalu (University of Tartu) who was in charge for the they had shown some interest – positive or negative – to matters con­ fieldwork in ELDIA. The best possible solution was sought for each case cerning the investigated minorities.

372 373 ANNELI SARHIMAA FINNIC DATA SETS IN THE ELDIADATA DATABANK

3.3. Case-specific deviations in capable of using Karelian or Veps. As screening was not characteristic the modes of data collection of any other case study, the Karelian and the Veps data are not strictly comparable with the other data sets. After the respondent screening, According to the original project plan, the new data were to be gath­ the actual data collection was conducted using interviewer-aided self- ered in late 2010 to early 2011 using mail surveys addressed to ran­ completion, that is, fieldworkers helped the respondents fill in the domly selected respondents, and by semi-structured focus group and questionnaires; for example, by providing explanations of what was individual interviews which should have taken place during the spring meant by the questions and by translating parts of the questionnaire of 2011. In practice, however, the data collection modes and the time­ when needed. In Russia, the data collection took place in March 2011. frames were only followed as planned in the case studies involving In Germany, the search for potential Estonian speakers was initi­ Hungarian in Austria and Slovenia and the two case studies in Finland, ated already in spring 2010, but recruiting participants proved to be namely, Estonian and Karelian. very difficult. One reason for this was that as Estonians in Germany In Estonia and Finland, the data were collected by mail surveys constitute a very small group, guaranteeing the anonymity of the re­ between January and March 2011. In Estonia, however, the question­ spondents posed a considerable challenge. As a result, many individu­ naire surveys among Võro and Seto speakers were not carried out as als did not want to participate as they feared being identified by others a mail survey but the Estonian research team thought it best to change later. Ultimately, the respondents in the case study of Estonian in Ger­ the survey mode to interviewer supported self-completion; the control many consisted of members of Estonian associations who volunteered group survey, however, was conducted as a mail survey. to participate and of members of a selection of Facebook groups who In Norway, the mail survey addressed to the Kvens was launched reacted to an open call inviting people to participate in a mail survey. as planned in October 2010, but due to problems in acquiring a proper The data collection among Estonians in Germany was conducted be­ sampling frame, the survey could only start after a maximally repre­ tween November 2010 and May 2011. sentative sample had been obtained from the Norwegian population register. The survey was finally conducted between April and June 2011; however, due to the extremely low response rate, an addition­ 4. Digital formats of the data sets, al web survey was conducted in summer 2011. The web survey also data processing, and editing yielded a very low response rate and for this reason the web responses are not included in the Kven data set. All the data collected by the surveys and tape- and video-recorded in In Sweden, the mail surveys among Meänkieli speakers and the interviews were processed into a digital format. The interviews Finnish speakers were delayed due to organisational problems at the were transcribed employing a rough transcription system and are in­ University of Stockholm. The empirical data were collected there be­ cluded in the database as Transcriber files. For project-internal purpos­ tween February and May 2011. es, the interview data also were analysed using ELAN; the encoded The most divergent data collection modes were adopted in Russia ELAN files were not included in the database but remained for per­ and in Germany. In the Republic of Karelia, the data collection ended sonal use by members of the case-specific research teams. up being a fairly heterogeneous combination of different sampling The survey data sets include only the results of the closed ques­ procedures. Speakers of Veps and Karelian were first screened by lo­ tions, since only these could provide numerical values automatically. cal researchers from larger groups of people; in practice this meant The entire ELDIAdata minority-language database contains in total that the research teams made sure that those who were selected as par­ 3 388 individual records; the language-specific data sets include 340 ticipants in the survey as well as those to be interviewed, actually were variables each. The entire control group database contains in total

374 375 ANNELI SARHIMAA FINNIC DATA SETS IN THE ELDIADATA DATABANK

1 460 records from seven countries; each country-specific data set 5. Availability of the ELDIA data and the covers 280 variables. conditions for using these in further research All values in all data sets have been checked by statisticians. In order to make it possible to scan the responses directly into individ­ The ELDIAdata Database is managed by the University of Mainz and ual data sets, the survey questionnaires were typeset in a coordinated a Board of Administration consisting of representatives of the research manner. The scanning process was decentralised among various par­ institutions that participated in ELDIA. The database is available for ticipating organisations while the optical character recognition (OCR) research purposes only and only per a written request addressed to the of data sets took place at the University of Vienna. In the databank the Board. The EuLaViBar Toolkit (Spiliopoulou Åkermark et al. 2013), survey-questionnaire data are stored in a format which allows for com­ containing a revised version of the ELDIA questionnaire, templates puter-based processing with a wide variety of statistical applications. and explanations of the statistical method, can be downloaded from The basic data sets and analysis were programmed using the the PHAIDRA repository (https://phaidra.univie.ac.at/o:301101). SAS software. Later on, the data sets were transformed into the SPSS data format as well. Although the coordination of the digitalisation process of the survey data by the ELDIA teams in Mainz and Vienna References was mostly successful, some problems occurred due to technical dif­ ficulties with the optical character recognition (OCR) software and Laakso, Johanna, Anneli Sarhimaa, Sia Spiliopoulou Åkermark & Reetta due to inconsistencies with the questionnaire contents. Hence, after Toivanen 2013: Summary of the Research Project ELDIA (European the questionnaires were scanned (or entered into the required form Language Diversity for All). Abridged version of the original English- manually), a substantial effort had to be taken to edit all data sets language report. Available at: ‹https://fedora.phaidra.univie.ac.at/ fedora/get/o:304813/bdef:Content/get› as uniformly as possible for comparative analyses. The editing was — 2016: Towards openly multilingual policies and practices. Assessing conducted by the ELDIA statistics team in Helsinki. The first edition minority language maintenance across Europe. Linguistic Diversity round was completed in September–October 2011, the second round and Language Rights 11. Bristol: Multilingual Matters. which provided the final data files now included in the ELDIAdata Spiliopoulou Åkermark, Sia, Johanna Laakso, Anneli Sarhimaa, Reetta was conducted in June–July 2013. Toivanen, Eva Kühhirt & Kari Djerf 2013: EuLaViBar Toolkit. Wien: The data editing process revealed a few technical errors in the ELDIA. Available at: ‹https://phaidra.univie.ac.at/o:301101› comparability of the questionnaires and hence in the comparability of the survey data across all data sets. Some of the corrections have to be taken into account when using certain parts of the Kven and the Meänkieli data sets and the control group data sets from the Russian Federation and Estonia. These data sets are provided with explana­ tions and instructions for future users.

376 377 NINA ZAICEVA Karelian Research Centre of the Russian Academy of Sciences

Veps language heritage in Karelia1

1. Introduction

Veps people live in Russia on the border of the Republic of Karelia, in Leningrad region and Vologda region. Based on the most recent Russian population census in 2010 (Vserossijskaja perepis nasele­ nija 2010), 5 936 people consider themselves Veps, of which 3 423 live in Karelia, 1 380 live in Leningrad region, and 412 live in Vo­ logda region. The decline in the Veps population and Veps language skills has been an urgent matter for researchers, society, and state authorities. In 2000, a Russian Federation government decree listed the Veps people in the Russian Federation Joint List of Indigenous Minorities as a community under special state protection (see Stro­ galschikova 2014: 237, 245). The Veps language also was included in the Russian Federation Red List of Languages (Krasnaja kniga 1994: 21–22). These efforts were intended to stimulate Veps lan­ guage revitalization and cultural preservation as well as restoring active bilingualism in areas where Veps language teaching in schools would be helpful. In the Republic of Karelia, several books in Veps have been published: textbooks for schools and universities, phrase books, bilingual dictionaries, and spelling dictionaries (see Zaiceva 2006: 119–135). The Veps media consist of the newspaper Kodima, the magazine Kipinä for children (in Veps, Kibin), and weekly TV and radio programs.

1. The study was carried out under state order (Project No. АААА-А18- 118012490344-5).

Multilingual­ Finnic. Language contact and change. 379–400. Uralica Helsingiensia 14. Helsinki 2019. ‹https://doi.org/10.33341/uh.85043› NINA ZAICEVA VEPS LANGUAGE HERITAGE IN KARELIA

Publishing in Veps was possible due to the gathering of Veps 2. Veps recordings in the phonogram language material and language research. Special mention should be archive of the Institute of History, made of the valuable Veps dialect dictionary by Maria Zaiceva and Literature, and Language at the Maria Mullonen (1972), which is one of the best works by Russian Karelian Research Centre of the RAS linguists in the field of Veps language research. The knowledge and insight provided by the dictionary into Veps grammar and its lexicon The phonogram archive of the Institute of History, Literature, and allowed for the development of richer materials for language revi­ Language at the Karelian Research Centre of the RAS has significant talization representative of the way contemporary Veps is used by recorded material in Veps dialects, which also includes information its speakers. about the culture and the history of the Veps nation. The phonogram The first expeditions to collect linguistic and ethnographic archive has over 400 hours of Veps tape recordings. Currently, this is material in Veps villages were made quite late. Finnish research­ one of the richest and most unique collections of Veps dialect samples. ers were the first to visit Veps villages in the second half of the The dialogue and monologue texts in this collection are on different 19th century during the period characterized by the Neogrammar­ topics relating to Veps ethnography, culture, daily life, and folklore. A ian movement. As a result of these first expeditions by Finnish re­ small part of this material has been transcribed and used in research. searchers, linguists today have dialect samples from the end of the The Karelian Research Centre of the RAS first began carrying out 19th century and from the beginning of the 20th century. This fact expeditions to the Veps villages in the beginning of the 1950s when is extremely important for a nation that does not have old written language researchers, ethnographers, and folklore researchers started records. to conduct many expeditions to the Veps-speaking areas. Expeditions A wider interest in the Veps language and its cultural heritage were also carried out to areas with place names indicating that Veps developed at the beginning of the 20th century. The most effective had once been spoken there. The recordings from these expeditions work on Veps language research was carried out in Karelia at the are preserved in the phonogram archive of the Institute. Institute of History, Literature, and Language at the Karelian Re­ The Veps material in the phonogram archive was digitized and search Centre of the Russian Academy of Sciences (RAS). Follow­ has an electronic registration, which identifies the tape number, the ing expeditions to Veps-speaking regions during the beginning of year and place of recording, the consultant, the researcher, the topic the second half of the 20th century, researchers began using these of conversation, and the genre of the recording. Using this electronic language data in their own studies, and the language was presented registration, the author of this report was able to make a map of the on the maps of an atlas, Atlas Linguarum Europae (1975–2007). The expedition destinations (Map 1) and to compile a table listing the ex­ archives of the Institute have a significant amount of Veps language peditions (see Appendix 1 at the end of this report). material. These materials include dialect recordings on magnetic The researchers indicated the local center when making their tapes, written dialect samples, manuscripts, and material in the mod­ lists; however, each center includes many small localities. For exam­ ern Standard Veps language published in books, newspapers, maga­ ple, the place name Voilahta (Veps Voilaht) covers three villages: Mar­ zines, and electronically. kovo (Markutan), Virino (Virahtan), and Pustoshka (Pust). Therefore, the number of localities visited during the expeditions greatly exceeds the number of place names in Appendix 1. The researchers often made many trips to the same regions conducting 130 expeditions in total. The dialect material is represented heterogeneously: the largest portion of the material was recorded in central localities, while less

380 381 VEPS LANGUAGE HERITAGE IN KARELIA

Northern Veps dialect Him = Himdeg, Kl = Kaleig, Kas = Kaskezoja, Ms = Matvejan Selʹg, Mec = Mecantaga, Mäg = Mägi, Oger = Ogerišt, Tž = Toižeg, Š = Šokš, Št = Šoutarv, Veh = Vehkoja. Middle Veps dialect Eastern subdialects Kj = Kuja, Šušt = Ošt, Pnd = Pondal, Pk = Püutkask, Päž = Päžarʹ, Sär = Särggärʹ, Šim = Šimgärʹ, Vl = Voilaht, Vär = Väräsärʹ. Middle Veps dialect Western subdialects Har = Haragl, Čik = Čikl, Jä = Järved, Kar = Karhil, Kor = Korbal, Korv = Korval, Kkz = Kukkaz, Kurb = Kurb, Ladv = Ladv, Mg = Mäggärʹ, Nem = Nemž, Nir = Nirgl, Noid = Noidal, Nor = Norj, Nür = Nürgl, Pec = Pecoil, Peč = Pečlʹ, Reb = Rebagʹ, Šon = Šondjal, Vil = Vilʹhäl. Southern Veps dialect Ars = Arskahtʹ, Bm = Bölämägi, Bor = Bor, Čid = Čidoi, Fm = Fedramägi, Krl = Kortlaht, Mai = Maigär, Lah = Laht, Mm = Maksimägi, Pož = Požarišš, Sod = Sodjärv, Šid = Šidjärv, Tim = Timoo, Vg = Vaagedjärv.

material was recorded in peripheral areas. For example, many record­ ings were made in Pondala (Veps Pondal; Vologda region, Babayevo district). This choice can be easily explained: this dialect is the base for the Veps dialect dictionary (M. Zaiceva & M. Mullonen 1972). A large amount of material also was recorded in Ladva (Ladv) and Ozyora (Järved) in Leningrad region, Podporozhye district; in Sheltorezo (Šoutarv) in the Republic of Karelia; and in Radogotchshi (Arskahtʹ), Bobrozero (Maigär), and Beloye ozero (Vaagedjärv) in Leningrad re­ gion, Boktogorsk district. The phonogram archive has unique mate­ rial recorded in Veps localities, which are no longer inhabited. This is the case for the villages of former Shimozero rural council (Šimgärʹ; Vologda region, Vytegra district), which was rehoused in the 1960s2. The materials include, for example, recordings of Veps from Kukkas village (Kukkaz, tape 102; recorded in Shapsha (Šapš), Lenin­ grad region, Lopeynopolsky district) made by one of the first Veps language researchers, M. M. Hämäläinen. Kukkas village is described by E. A. Tunkelo as completely Russified in his 1946 monograph (Tunkelo 1946: 5). M. M. Hämäläinen, however, was able to find con­ sultants and record their speech in 1959. The recordings are extremely

2. E. A. Tunkelo (1946: 6) mentioned Shimozero village with a population of 4 936 of which 100% spoke the Veps language in the 1930s. Map 1. Expedition destinations in the Veps areas of the researchers from the Institute of History, Literature, and Language at the Karelian Research Centre of 383 the RAS (1952–2012). NINA ZAICEVA VEPS LANGUAGE HERITAGE IN KARELIA interesting phonetically, due to the emergence of long vowels and ex­ ‘There were collective farms in our villages in the 1930s. We had ceptional short vowels. 3 collective farms. The collective farm in Markovo (Markutan) The Kukkas subdialect features long vowels, apart from the was called Krasny Chukhar. We did not know the word Veps. We other Veps dialects. In addition, rising diphthongs are represented in were called Chukhars. The collective farm’s name was Chukhar, an unusual way, for example luda ‘to make a basic fabric; to stack’ even Krasny (‘Red’) Chukhar. The collective farm in Virino (Vi- (compare the usual Veps loda), lüda ‘hit’ (comp. löda), mühä ‘late’ rahtan) was called Krasny Partizan (‘Red partisan’), the farm in Pustoshka (Pust) was called Trinadtsataya Godovshchina (‘The (comp. möha), etc. These phonetic facts have not been mentioned in 13th century’). Kuya (Kuja) village is 6 versts from here, there the past. Few examples from the materials of the archive show these were several collective farms: Krasny Pakhar (‘Red plowman’), phenomena (transliterated by the author of the report): Bolshevik, Krasnoarmeyets (‘Red Army soldier’). There were collective farms near Pondala (Pondal). But one collective farm — Ka sä pagižed lüdikš? ‘Do you speak Veps?’ was called in Veps: Usʹ Elo (‘New life’).’ — Ka maltan! ‘Yes, I do!’ — Sanu minei erasid sanoid. Kut om ‘položitʹ ceno v zarod (stog)’? The self-designation of Veps people speaking different Veps dialects ‘Say some words. How to say: ‘to put the hay to a stack’?’ is different. As seen in the example, the southern Veps call themselves — Luda hein sabrha. — Kut sanutas – ‘derevoʹ, ‘šestʹ?’ ‘How to say ‘tree’, ‘six’?’ bepslaažed or čuharid, whereas the northern Veps call themselves lü- — Puu, kuuzʹ. dinikad and the Veps of Vologda region call themselves ičemoi rahvaz ‘(our) own nation’. Verbs about speaking are also different from each Mina paginan kuulʹen ‘I hear some speech’ other: the southern Veps could say pagišta počudski ‘to speak in a Luu ‘bone’ way’, the northern Veps say pagišta lidikš ‘to speak Veps’, and Pidäb juda ‘One has to drink’ the Vologda Veps say pagišta ičemoi kartte ‘to speak in (our) own Pidäb tuda leibad ‘One has to bring some bread’ Lüda ‘hit’, süda ‘eat’, müda ‘sell’ way’. Hän tuli kodihe mühä ‘He came home late’

Another example from the materials, a text recorded in 1960 by M. M. 3. Veps materials in the manuscripts Hämäläinen from Grigoriy Bolshakov, born in 1914 (Voylakhta (Voi­ archive of the Institute of History, laht) village, Vologda region, Babayevo district, record 103), shows Literature, and Language at the how the southern Veps designate themselves: Karelian Research Centre of the RAS

Triccatįjiš voziš miiden derüuniš organizęihes kolhozat. Kuumo The Karelian Research Centre has a manuscript archive, as well. kolhozat oli. Markutanas kolhos kucįhe Krasnįi Čuharʹ. Silęi The dialect collection, which consists of some transcribed manu­ mö em tunnuhu nečida vajeht ’veps’. A meitʹ kucuiba čuharikš. I scripts from the collection of Veps narratives, Obrazcy vepsskoj reči kolhos-ki mugįine oli Čuharʹ, Krasnį vüu. Virahtanas oli kolhos (M. Zaiceva & M. Mullonen 1969), is not very large. The card index Krasnįi Partizan, Pustou Trinaccataja Godovššina. Kujahasai of the Veps dialect dictionary published in 1972 by Maria Zaiceva and tägoupei kusʹ virstat, sigou mugažo kolhozat oliba: oli Krasnį paharʹ, Bolʹševik, Krasnoarmejec oli. Pondlou-ki kolhozat Maria Mullonen is located in the archive of the Linguistics Section organizęihe. No sigou oli üksʹ kolhos miiden kartho nazįvajihe: at the Institute. The card index contains handwritten cards in alpha­ nece kolhos Usʹ Elo. betical order with Veps words presented along with their meaning and their use in the dialects together with examples.

384 385 NINA ZAICEVA VEPS LANGUAGE HERITAGE IN KARELIA

The Veps materials in the manuscript archive mostly consist of Babayevo district of Vologda region was collected during the expedi­ various types of questionnaires. These include the questionnaires used tions. It was determined that in 1981 there were 4 087 Veps in villages in compiling the Comparative Onomasiological Dictionary of the Kare­ in Leningrad region (318 according to the 1979 census) and 773 in lian, Veps, and Saami languages, Sopostavitelno-onomasiologičeskij Vologda region (41 according to the 1979 census). This meant that the slovarʹ (2007), which was prepared by the linguists of the Institute. 1979 census did not count approximately 4 500 Veps. Information on The questionnaire included 1 500 lexical questions. The dialect ma­ inhabitants was collected in 81 larger and smaller villages. Material terial used for the dictionary was recorded in the field at the end of was also collected in Prionezhye district in the Republic of Karelia the 1980s and it presents the state of the dialects at that time. For (see Strogalschikova 1989: 27–42 for more details). the Veps material, the dictionary questionnaires were filled out in 6 The material gathered in 1981 became the basis for ethnoso­ Veps villages: Northern Veps from Sheltozero (Šoutarv) and Kaskes­ ciological research carried out in 1983. Gender, age, education, and ruchey (Kaskezoja); Middle Veps from Oshta (Ošt; includes material nationality were taken into consideration when deciding on a repre­ from Shimozero (Šimgärʹ) as well), Voylakhta (Voilaht), and Ozyora sentative sample of interviewees. 400 Veps from Prionezhye district (Järved); and Southern Veps from Sidorovo (Sodjärv). in Karelia and 400 Veps from Leningrad and Vologda regions were The dictionary in general is a linguistic-geographic handbook of interviewed. The 1983 expedition lasted from June to December and the aforementioned languages; its materials show lexical similarities interviews were conducted in all villages except the smallest ones. and differences among languages and dialects. In compiling the dic­ The interviews were based on a questionnaire created by Zinaida tionary, a conceptual-thematic method was used along with the con­ Strogalschikova who used the questionnaires of V. V. Pimenov, L. S. ceptual register of Russian so that use of the large amount of mate­ Khristolyubova, and G. N. Belorukova as a base. Some questions were rial it contained would be easy for dialect research (of the Finnic and borrowed from the 1979 survey papers of the Karelian sociologists Saami languages presented in the dictionary). Y. I. Klementyev and A. A. Kozhanov. The questionnaire consisted The sociological questionnaires include a considerable amount of 118 questions and it included some open questions, for example, of Veps material. The material collected by ethnosociologist Zinaida on opinions regarding the need for a Veps written language, the ne­ Strogalschikova between 1981–1983 forms a significant part of the ar­ cessity of Veps language study in schools, and reasons for linguis­ chive. Two population censuses of the Soviet Union in 1970 and 1979 tic and ethnic assimilation. As one of the outcomes of the interviews, showed an almost complete disappearance of Veps people from their 92% of those interviewed in the Leningrad and Vologda regions con­ traditionally inhabited locations in Leningrad and Vologda regions. firmed their nationality as Veps, 6% called themselves Russians (half Many of the expeditions led by Zinaida Strogalschikova3 had the goal of these respondents had one Russian parent), and 2% were not able of providing an alternative census of the Veps population. This mate­ to choose between Veps and Russian as their nationality. The 1989 rial was drawn from the data contained in the rural household registers census showed that there were 4 500 more Veps in the Leningrad and of the villages inhabited by Veps people. These registers included basic Vologda regions. Also, tape recordings of significant folklore mate­ information about the inhabitants: gender, age, place and time of birth, rial were made in the expeditions and transferred to the phonogram relation to the householder, marital status, place of work, employment, archive of the Institute of History, Literature, and Language at the Ka­ education, nationality, and duration of residency in a particular vil­ relian Research Centre of the RAS. lage. Information on the population of four of Leningrad region’s dis­ The research material of the archives of the Institute of History, tricts (Podporozhye, Lodeynopolsky, Boksitogorsk, Tikhvin) and of Literature, and Language served as a foundation for further research. This material was important in the revitalization of Veps at the end of 3. The author of the report also participated in these expeditions. the 1980s.

386 387 NINA ZAICEVA VEPS LANGUAGE HERITAGE IN KARELIA

4. Veps dialect samples published 5. Veps dialect dictionary and atlases

There are a few dialect samples published in Russia. Compared to the A great deal of material can be found in the Veps dialect dictionary samples published in Finland, these include newer texts, as the expedi­ and atlases. The Veps dialect dictionary, Slovarʹ vepsskogo jazyka tions and language research resulting in these samples was conducted (M. Zaiceva & M. Mullonen 1972), includes materials from almost more recently. all Veps localities. The authors made multiple expeditions to almost The collection of Veps narratives, Obrazcy vepsskoj reči, pub­ all locations within the Veps-speaking area. There they recorded a lished by Maria Zaiceva and Maria Mullonen (1969), includes Middle significant amount of material on Veps spiritual and material culture and Southern Veps texts, but no Northern Veps texts, which is a gap in as well as Veps folklore. Recordings made by the authors are pre­ Russian science. This collection was published as a result of work on sented along with rich information on the consultants. Aleksandra the Veps dialect dictionary. Kalinina, a consultant with an excellent knowledge of Veps language Middle Veps texts recorded in Ladva (Ladv) village in Pod­ and folklife culture, was discovered by the authors in Pondala (Ba­ porozhye district of Leningrad region, transcribed by Maria Mullonen, bayevo district, Vologda region). For this reason, this dialect was were published in the collection Jazyk i narod (Irma Mullonen 2002). used as the base dialect for the Veps dialect dictionary. The lexis of These texts include conversations on ethnographic topics with village 31 localities (18 Middle Veps, 7 Southern Veps, and 6 Northern Veps) inhabitants of various ages. was used in the dictionary. The dictionary includes subdialects of the Veps folk tales can be found in the collection Vepsän rahvhan Middle Veps dialect from the villages of Čikl, Enarʹv, Järved, Karhil, sarnad by Nina Onegina and Maria Zaiceva (1996). This collection Korbal, Kuja, Ladv, Nirgl, Noidal, Pecoil, Pondal, Püutkask, Päžarʹ, utilizes the recordings of the Institute’s phonogram archive and it in­ Šimgärʹ, Särggärʹ, Vilʹhäl, Voilaht, and Väräsärʹ; of the Southern Veps cludes texts from Prionezhye district in the Republic of Karelia as well dialect from Arskahtʹ, Čaigl, Kortlaht, Maigärʹ, Sodjärv, Šidjärv, and as from the Leningrad and Vologda regions. In other words, it covers Vaagedjärv; and the Northern Veps dialect from Kaleig, Kaskezoja, all Veps dialects: the Northern Veps of Šoutarv, Mecantaga, Šokš, Ka- Mecantaga, Pervakat, Šokš, and Šoutarv. A large proportion of the leig, Matvejanselʹg, Tihoništ; the Middle Veps of Ladv, Järved, Pecoil, recordings was made in the remote and easternmost Veps villages of Pondal, Kuja; and the Southern Veps of Šidjärv and Sodjärv. Pondal, Kuja, and Voilaht in Babayevo district of Vologda region. The The collection Käte-ške käbedaks kägoihudeks, edited by Nina dictionary also utilizes recordings from villages, which have ceased to Zaiceva and Olga Zhukova (2012), introduces the largely unknown folk­ exist since then: Šimgärʹ, Väräsärʹ, Särggärʹ, and Kortlaht. Each word lore genre of Veps lamentation. The editors transcribed material on almost is presented together with its variants in other dialects. Veps was used all lamentation types (wedding, funeral, occasional lamentations) from the as a written language in the 1930s and in the beginning of the 1990s phonogram archive of the Institute of History, Literature, and Language. when the Veps written tradition was revitalized. This dictionary served The book includes detailed information on the texts, consultants as well as as a foundation for the Veps orthography and various dictionaries and the place and time of each recording. The number of localities providing textbooks. the recorded material for this book is large: 19 Veps districts (see the list Veps dialect material is also presented in linguistic-geographic of localities in Zaiceva & Zhukova (eds) 2012: 207–208). This large num­ studies. The first atlas that included Veps dialect material was the ber of localities makes possible examination of dialect characteristics. In Atlas Linguarum Europae (ALE 1975–2007), which included Veps addition, the book includes musical notation from S. V. Kosyreva of the language material on separate maps. The material for the atlas was Petrozavodsk Conservatory for 25 lamentation songs. This genre of Veps collected by Maria Zaiceva and Nina Zaiceva in six Veps localities folklore is presented for the first time with this amount of material. (Šoutarv, Kaskezoja, Ošt, Voilaht, Järved, Sodjärv). The questionnaires

388 389 NINA ZAICEVA VEPS LANGUAGE HERITAGE IN KARELIA completed in the 1980s are kept in the manuscript archive of the Insti­ 6. Veps language corpus and tute of History, Literature, and Language at the Karelian Research material in the media Centre of the RAS. Veps dialect material is described quite completely in the three- Currently, dialect material and material written in the modern Veps lan­ volume Atlas Linguarum Fennicarum (ALFE 2004–2010) prepared guage is available online as the Veps language corpus (Korpus vepsskogo by an international group of authors from Finland, Estonia, and the jazyka), see ‹vepsian.krc.karelia.ru›. The idea to make a Veps language Republic of Karelia in Russia. In preparing the atlas, the work of corpus originated in 2010. The goal of this project was a straightforward completing the questionnaires was done in seven Veps localities one: to make a corpus that includes a small amount of original oral and (Northern Veps: Šoutarv, Kaskezoja; Middle Veps: Järved, Voilaht, written texts and to put these online. A lemma is defined as a Veps writ­ Šimgärʹ; Southern Veps: Sodjärv, Kortlaht). It should be noted that ten language word and is used as a reference for the dialect forms. the material was collected in Kortlaht village by Nina Zaiceva and This electronic resource is open to all users. The texts are in five Irma Mullonen at the end of the 1980s from the last three inhabitants subcorpuses allowing one to become acquainted with dialect differ­ who left the village relatively soon after the expedition. This village ences and with materials in the modern written language: is not inhabited anymore. 1) Dialect text subcorpus (parallel with translations in Russian) The Atlas of the Veps language (LAVJa) was published in 2019. 2) Veps lamentation corpus (parallel with translations in Russian) The idea of a Veps atlas originates in the 1940s when work on the Kare­ 3) Veps folk tale corpus (parallel with translations in Russian) lian language atlas (Bubrih, Beljakov & Punžina 1997), performed 4) Bible texts (modern written language texts) under the guidance of D. V. Bubrikh, a professor and a corresponding 5) Modern written language text corpus (texts from the newspaper member of the USSR Academy of Sciences, was being carried out. The Kodima, the magazine Kipinä, and other literary texts). work on a linguistic atlas of the Veps language began in 2012 with the support of the Russian Foundation for Humanities under the guidance In addition to the the texts, the corpus also includes an electronic dic­ of project manager Nina Zaiceva. The dialect material was collected in tionary. Lexical elements are marked with their part of speech and 2012–2013 using a new questionnaire4 (see Voprosnik 2013) based on have Russian and English translations. At the end of 2014, the corpus some of the ideas incorporated into the questionnaire from the 1940s. and the dictionary included over one thousand texts, about 800 biblio­ A large portion of the new questionnaire (266 questions of 395) consists graphic sources, and over 10 000 lemmas and word forms. of lexis questions. The material for the atlas was collected from 30 lo­ The corpus includes various texts in the modern Veps written calities where good informants could be found. In addition, published language created mostly due to the efforts of researchers from the Re­ and archival sources from a total of 75 localities were used, of which 17 public of Karelia. The texts from the newspaper Kodima, many of are Northern Veps, 40 Middle Veps, and 18 Southern Veps. It is neces­ the texts from the Children’s Bible and the New Testament, and the sary for this work to be the largest ever on dialect material in order for literary texts are written after 1992. These show changes in the writ­ it to expand on the topic of the dialects, to characterize the criteria for ten language, changes in its lexis, and the development of the rules of the formation of the Veps dialect regions, and to analyze, if possible, the Veps grammar, stylistics, and spelling. formation and development of the dialects. Periodika, the publisher of the newspaper Kodima, the children’s magazine Kipinä, and the literary almanac Verez tullei, keeps all the text files in its archive. In addition, the publishing house places the 4. Nina Zaiceva, Irma Mullonen, researchers from Petrozavodsk, and Sergei Myznikov and I. V. Brodsky, researchers from Saint Petersburg, participated in the material from the newspaper and magazine online, thereby enabling work on the new questionnaire. the wider public to have access to it.

390 391 NINA ZAICEVA VEPS LANGUAGE HERITAGE IN KARELIA

References vodsk: Rossijskaja akademija nauk. Karelʹskij naučnyj centr. Institut jazyka, literatury i istorii. ALE = Atlas Linguarum Europae. 1975–2007. Assen: Van Gorcum & Roma: Strogalschikova, Z[inaida] I. 1989: Ob ètnodemografičeskix tendencijax, Poligrafico. socialʹno-èkonomičeskom i kulʹturnom razvitii vepsskoj narodnosti ALFE = Atlas Linguarum Fennicarum 1–3. Itämerensuomalainen kielikartasto. [On the ethnodemographic tendencies, socioeconomic and cultural Läänemeresoome keeleatlas. Ostseefinnischer Sprachatlas. Lingvističe- development of the Veps people]. V. V. Pimenov, Z. I. Strogalschikova skij atlas pribaltijsko-finskix jazykov. 2004–2010. Helsinki: Suomalaisen & Ju. Ju. Surxasko (eds), Problemy istorii i kulʹtury vepsskoj narod- Kirjallisuuden Seura & Kotimaisten kielten tutkimuskeskus. Available at: nosti. Petrozavodsk: Karelʹskij filial AN SSSR. ‹https://avaa.tdata.fi/web/avaa/-/kotus-kielikartastot› — 2014: Vepsy. Očerki istorii i kulʹtury. [The Veps. Essays on history and Bubrih, D[mitri] V., A[leksandr] A. Beljakov & A[leksandra] V. Punžina culture.] Sankt-Peterburg. 1997: Dialektologičeskij atlas karelʹskogo jazyka. Karjalan kielen Tunkelo, E. A. 1946: Vepsän kielen äännehistoria [Sound history of the murrekartasto. [Atlas of the Karelian language.] Ed. Leena Sarvas. Veps language]. Suomalaisen Kirjallisuuden Seuran Toimituksia 228. Institut jazyka, literatury i istorii Karelʹskogo naučnogo centra RAN & Helsinki: Suomalaisen Kirjallisuuden Seura. Naučno-issledovatelʹskij centr jazykov Finljandii. Kotimaisten kielten Verez tullei. [Annual literature almanac.] Petroskoi: Periodika. tutkimuskeskuksen julkaisuja 97. Helsinki: Société Finno-Ougrienne. Voprosnik 2013 = Voprosnik “Lingvističeskogo atlasa vepsskogo jazyka” Kipinä. [Children’s magazine.] Petroskoi: Periodika. 2011–. Available at: [Questionnaire “Linguistic atlas of the Veps language”]. N. G. ‹http://kipina.rkperiodika.ru› Zaiceva & S. A. Myznikov (eds), Vepsskie arealʹnye issledovanija. Kodima. [Newspaper.] Petroskoi: Periodika. 1993–. Available at: ‹http://kodima. Petrozavodsk: Karelʹskij naučnyj centr RAN. 7–45. rkperiodika.ru› Vserossijskaja perepis naselenija 2010. Natsionalnyi sostav nasele­ Korpus vepsskogo jazyka [Veps language corpus]. Available at: ‹http://vepsian. nija po subjektam Rossijskoi federatsii [The Russian census of krc.karelia.ru/text/› 2010. Part 1. Population number and distribution. Table 7 [MS Krasnaja kniga 1994 = Krasnaja kniga jazykov narodov Rossii. Ènciklo­ Excel table]]. Federalnaja služba gosudarstvennoi statistiki [Sta­ pedičeskij slovarʹ-spravočnik. [The Red Book of the languages of the tistical office of the Russian Federation], 2012. Moscow: ИИЦ peoples of Russia.] Ed. V. P. Neroznak. 1994. Moskva: Academia. “Статистика России”. [Accessed 15 Sept 2015] Available at: LAVJa = Lingvističeskij atlas vepsskogo jazyka [Atlas of the Veps language]. ‹http://www.gks.ru/free_doc/new_site/population/demo/per-itog/tab7.xls› 2019. N[ina] G. Zaiceva (ed.). Sankt-Peterburg: Federalnyj issledova­ Zaiceva, N[ina] G. 2006: Mladopisʹmennyj jazyk vepsov: periody i perspek­ telʹskij centr. “Karelʹskij naučnyj centr Rossijskoj akademii nauk”. tivy razvitija [Veps as a newly written language: periods and perspec­ Institut jazyka, literatury i istorii & Nestor-Istorija. tives for development]. Sovremennaja nauka o vepsax: dostiženija i per- Mullonen, I[rma] I. 2002: Obrazcy reči žitelej sela Ladva Podporožskogo spektivy (pamjati N. I. Bogdanova). Ed. I. Ju. Vinokurova. Petrozavodsk: rajona Leningradskoj oblasti [Examples of the speech of the inhabit­ Rossijskaja akademija nauk. Karelʹskij naučnyj centr. 119–177. ants of the village of Ladva in Podporože disrict in Leningrad region]. Zaiceva, M[aria] I. & M[aria] I. Mullonen 1969: Obrazcy vepsskoj reči A. Gerd, M. Savijärvi & T. de Graaf (eds), Jazyk i narod. Teksty i [Samples of Veps dialects]. Akademija Nauk SSSR. Karelʹskij filial. kommentarii razgovornoj reči pribaltijsko-finskix jazykov i dialektov Leningrad: Nauka. russkogo jazyka na Severo-zapade Rossii. Sankt-Peterburg. 55–96. — 1972: Slovarʹ vepsskogo jazyka [Dictionary of the Veps language]. Onegina, N[ina] F. & M[aria] I. Zaiceva (eds) 1996: Vepsän rahvhan sarnad. Akademija Nauk SSSR. Karelʹskij filial. Institut jazyka, literatury i Vepsskie narodnye skazki. [Veps folk tales.] Pamjatniki folʹklora Kare­ istorii. Leningrad: Nauka. lii. Petrozavodsk: Karelija. Zaiceva, N[ina] G. & O[lga] Ju. Zhukova (eds) 2012: Käte-ške käbedaks Sopostavitelno-onomasiologičeskij slovarʹ = Sopostavitelno-onomasio­ kägoihudeks. ‘Obernisʹ-ka miloj kukušečkojʹ. Vepsskie pričitanija. logičeskij slovarʹ dialektov karelʹskogo, vepsskogo, saamskogo jazykov [Turn into a sweet cuckoo. Veps lamentations.] Petrozavodsk: Jumin­ [Comparative onomasiological dictionary of the Karelian, Veps, and keko & Karelʹskij naučnyj centr. Rossijskaja akademija nauk. Institut Saami languages]. Ju. S. Eliseev, N. G. Zaiceva (eds). 2007. Petroza­ jazyka, literatury i istorii. 392 393 NINA ZAICEVA VEPS LANGUAGE HERITAGE IN KARELIA

Appendix I. Expeditions to the Veps-speaking Place name in Veps Place name in Region/district of Year Researcher areas conducted by researchers from the Institute and abbreviation Russian the place where the of the of History, Literature, and Language at the Karelian recording was made recording Research Centre of the RAS (1952–2012). Šoutarv (Št) Шелтозеро — «— 1952, Evseev, V. Ja., 1959, Hämäläinen, M. M., 1962, Mullonen, M. I., 1964, Zaiceva, M. I., Northern Veps dialect 1976, Zaiceva, N. G., 1980, Mullonen, I. I. Place name in Veps Place name in Region/district of Year Researcher 1981, and abbreviation Russian the place where the of the 1983, recording was made recording 1989, Himdeg (Him) Гимрека Leningrad 1962, Vinokurova, I. Ju. 1991 region, 1983 Vehkoja (Veh) Вехручей — «— 1989 Mullonen, I. I., Podporozhye Strogalschikova, Z. I. district Kaleig (Kl) Рыбрека Republic 1983, Strogalschikova, Z. I., of Karelia, 1991 Vinokurova, I. Ju. Middle Veps dialect Eastern subdialects Prionezhsky district Place name in Veps Place name in Region/district of Year Researcher Kaskezoja Каскесручей — «— 1970, Mullonen, I. I., and abbreviation Russian the place where the of the (Kas) 1983 Zaiceva, N. G., recording was made recording Vinokurova, I. Ju., Kuja (Kj) Куя Vologda 1962, Mitrofanova, A. A., Strogalschikova, Z. I. region, 1963 Zaiceva, M. I., Matvejan Матвеева — «— 1983, Strogalschikova, Z. I., Babayevo Mullonen, M. I., Selʹg (Ms) Сельга 1991 Vinokurova, I. Ju. district Zaiceva, N. G. Mecantaga Залесье — «— 1983 Strogalschikova, Z. I. Ošt (Šušt) Ошта Vologda 1962, Mitrofanova, A. A., (Mec) region, 1970, Zaiceva, M. I., Mägi (Mäg) Горнее — «— 1991 Vinokurova, I. Ju. Vytegra 1974, Mullonen, M. I., Шелтозеро district 1975, Zaiceva, N. G., 1976, Mullonen, I. I. Ogerišt (Oger) Огеришта — «— 1983 Strogalschikova, Z. I. 1987 Toižeg (Tž) Другая Река — «— 1980, Onegina, N. F., Pondal (Pnd) Пондала Vologda 1957, Bogdanov, N. I., 1983, Strogalschikova, Z. I., region, 1958, Zaiceva, M. I., 1991 Vinokurova, I. Ju. Babayevo 1961, Mullonen, M. I., Šokš (Š) Шокша — «— 1964, Hämäläinen, M. M., district 1962, Mitrofanova, A. A., 1983 Onegina, N. F., 1963, Zaiceva, N. G., Strogalschikova, Z. I., 1965, Mullonen, I. I., Vinokurova, I. Ju. 1981, Vinokurova I. Ju., 1983 Strogalschikova, Z. I., Onegina, N. F.

394 395 NINA ZAICEVA VEPS LANGUAGE HERITAGE IN KARELIA

Place name in Veps Place name in Region/district of Year Researcher Middle Veps dialect Western subdialects and abbreviation Russian the place where the of the recording was made recording Place name in Veps Place name in Region/district of Year Researcher Püutkask (Pk) Пелкаска Vologda 1962 Mitrofanova, A. A. and abbreviation Russian the place where the of the region, recording was made recording Vytegra Haragl (Har) Харагиничи Leningrad 1986 Zaiceva, N. G., district region, Tikh­ Mullonen, I. I., Päžarʹ (Päž) Пяжозеро Vologda 1983 Strogalschikova, Z. I. vin district Vinokurova, I. Ju. region, Čikl (Čik) Чикозеро Leningrad 1959, Hämäläinen, M. M., Babayevo region, Podpo­ 1965 Zaiceva, M. I., district rozhye district Mullonen, M. I. Särggärʹ (Sär) Сяргозеро Vologda 1962 Mitrofanova, A. A. region, Järved (Jä) Озера — «— 1961, Zaiceva, M. I., Vytegra 1980, Mullonen, M. I., district 1983, Zaiceva, N. G., Šimgärʹ (Šim) Шимозеро — «— 1970, Mullonen, M. I., 1985 Mullonen, I. I., 19915 Zaiceva, N. G. Vinokurova, I. Ju., Strogalschikova, Z. I. Voilaht (Vl) Войлахта Vologda 1960, Hämäläinen, M. M., Karhil (Kar) Каргиничи — «— 1958, Hämäläinen, M. M. region, 1962, Mitrofanova, A. A., 1959 Babayevo 1963, Mullonen, M. I., district 1972, Zaiceva, M., I., Korbal (Kor) Корбиничи Leningrad 1962, Mitrofanova, A. A., 1975, Väizinen, T. I., region, 1986 Zaiceva, N. G., 1981, Onegina, N. F., Tikhvin Mullonen I. I., 1983 Zaiceva, N. G., district Vinokurova, I. Ju. Strogalschikova, Z. I. Korval (Korv) Корвала — «— 1972 Zaiceva, M. I. 8 Väräsärʹ (Vär) Кривозеро Vologda 19686, Mullonen, M. I., Kukkaz (Kkz) Куккас Leningrad 1959 Hämäläinen, M. M. region, 19707 Zaiceva, M. I. region, Lodey­ Vytegra nopolsky­ district district Kurb (Kurb) Курба Leningrad 1989 Mullonen I. I., region, Podpo­ Vinokurova, I. Ju., rozhye district Kuznetsova, V. P. Ladv (Ladv) Ладва — «— 1961, Mullonen M. I., 1968, Zaiceva, M. I., 1980, Onegina, N. F., 1981, Zaiceva, N. G., 1983, Mullonen I. I., 1989, Vinokurova, I. Ju., 1991 Strogalschikova, Z. I., 5. Recorded in Oshta (Ošt), in Gimreka (Himdeg), in Voznesenye (Voznesenj). Kuznetsova, V. P., 6. Recorded in Nemzha (Nemž). Lukina, N. 7. Recorded in Oshta (Ošt). 8. Recorded in Shapsha (Šapš). 396 397 NINA ZAICEVA VEPS LANGUAGE HERITAGE IN KARELIA

Place name in Veps Place name in Region/district of Year Researcher Southern Veps dialect and abbreviation Russian the place where the of the recording was made recording Place name in Veps Place name in Region/district of Year Researcher Mäggärʹ (Mg) Мягозеро — «— 1980, Zaiceva, N. G., and abbreviation Russian the place where the of the 1983, Mullonen I. I., recording was made recording 1985, Vinokurova, I. Ju., Arskahtʹ (Ars) Радогощь Leningrad 1969, Zaiceva, M. I., 1991 Strogalschikova, Z. I. region, 1983, Mullonen, M. I., Nemž (Nem) Немжа — «— 1980, Mullonen I. I., Boksitogorsk 2012 Strogalschikova, Z. I., 1990 Vinokurova, I. Ju. district Zaiceva, N. G., Nirgl (Nir) Ниргиничи — «— 1958 Hämäläinen, M. M. Mullonen I. I. Noidal (Noid) Нойдала Leningrad 1972 Zaiceva, M. I. Bölämägi (Bm) Белячиха — «— 1969 Zaiceva, M. I., region, Tikh­ Mullonen, M. I. vin district Bor (Bor) Бор (Крас­ — «— 1983, Strogalschikova, Z. I., Norj (Nor) Норгина Leningrad 1983 Mullonen I. I., ный Бор) 1985 Vinokurova, I. Ju. region, Podpo­ Strogalschikova, Z. I. Čidoi (Čid) Чидово — «— 1972 Zaiceva, M. I. rozhye district Fedramägi Федорова — «— 1969 Zaiceva, M. I., Nürgl (Nür) Нюрговичи Leningrad 1986 Zaiceva, N. G., (Fm) Гора Mullonen, M. I. region, Tikh­ Mullonen I. I., Kortlaht (Krl) Кортлахта — «— 1966 Zaiceva, M. I., vin district Vinokurova, I. Ju. Mullonen, M. I. Pecoil (Pec) Пелдуши Leningrad 1968, Zaiceva, M. I., Maigär (Mai) Боброзеро — «— 1969, Zaiceva, M. I., region, 1980, Mullonen, M. I., 1983 Mullonen, M. I., Podporozhye 1991 Zaiceva, N. G., Vinokurova, I. Ju., district Mullonen I. I., Zaiceva, N. G., Vinokurova, I. Ju. Mullonen I. I., Pečlʹ (Peč) Печеницы Leningrad 1958, Hämäläinen, M. M. Strogalschikova, Z. I. region, Lodey­ 1959 Laht (Lah) Лахта — «— 1958, Bogdanov, N. I., nopolsky­ 1985 Zaiceva, M. I., district Vinokurova, I. Ju., Rebagʹ (Reb) Ребов Leningrad 1972 Zaiceva, M. I. Zaiceva, N. G., Конец region, Tikh­ Mullonen I. I. vin district Maksimägi Максимова — «— 1969, Zaiceva, M. I., Šondjal (Šon) Шондовичи Leningrad 1985, Zaiceva, N. G., (Mm) Гора 1991 Mullonen, M. I., region, Podpo­ 1991 Mullonen I. I., Zaiceva, N. G. rozhye district Vinokurova, I. Ju. Požarišš (Pož) Пожарище — «— 1966 Zaiceva, M. I., Vilʹhäl (Vil) Ярославичи — «— 1968, Zaiceva, M. I., Mullonen, M. I., 1980, Mullonen, M. I., Sodjärv (Sod) Сидорово — «— 1958, Bogdanov, N. I., 1983, Zaiceva, N. G., 1964, Zaiceva, M. I., 1985 Mullonen I. I., 1983, Mullonen, M. I., Vinokurova, I. Ju., 1985, Strogalschikova, Z. I., Strogalschikova, Z. I. 2012 Zaiceva, N. G., Mullonen I. I.

398 399 NINA ZAICEVA

Place name in Veps Place name in Region/district of Year Researcher TONI SUUTARI & ULRIIKKA PUURA and abbreviation Russian the place where the of the Kotimaisten kielten keskus recording was made recording Šidjärv (Šid) Прокушево — «— 1958, Bogdanov, N. I., 1985 Vinokurova, I. Ju., Zaiceva, N. G., Kotimaisten kielten keskuksen Mullonen I. I. Timoo (Tim) Саньков — «— 1983 Zaiceva, N. G., itämerensuomalaiset aineistot Бор Mullonen I. I. Vaagedjärv Белое озеро — «— 1966, Zaiceva, M. I., (Vg) 1983, Mullonen, M. I., 1985 Strogalschikova, Z. I., 1. Johdanto Zaiceva, N. G., Mullonen I. I. Kotimaisten kielten keskuksella eli Kotuksella on mittavat yli sadan vuoden aikana kootut arkistot. Murteita ja nimistöä alettiin kerätä sys­ temaattisesti 1800-luvun loppupuolella, ja murteiden laaja tallentami­ nen äänitteiksi alkoi avokelanauhatekniikan yleistyttyä 1950-luvun lopussa. Kokoelmista ja niiden kartuttamisesta vastasivat eri säätiöt ja organisaatiot, tunnetuimpana Sanakirjasäätiö, kunnes vuonna 1976 perustettiin Kotimaisten kielten tutkimuskeskus. Arkistot koottiin tuolloin yhteen, samoin aineistoihin liittyvät sanakirja- ja tutkimus­ hankkeet sekä kielenhuoltotyö. Kotuksen arkistojen laajuus on nykyisin noin 2,2 hyllykilo­ metriä. Suurin osa arkistotiloissa säilytteillä olevista kokoelmista on noin A6-kokoisia käsin tai koneella kirjoitettuja sana- ja nimilippuja. Sähköiset aineistot puolestaan koostuvat yhteensä satojen miljoonien sanojen laajuisista korpuksista, miljoonista kuvamuotoon skannatuis­ ta arkistolipuista ja muista dokumenteista sekä noin 17 000 tunnista digitoituja tai suoraan digitaaliseen muotoon kerättyjä av-tallenteita. Kaikkiaan äänitteitä ja videoita on noin 24 000 tuntia. Kotus soveltaa arkistoihin ja aineistoihin avoimuuden periaatetta. Arkistojen kokoelmat ovat pääosin asiakkaiden vapaasti käytettävissä, ja verkossa aineistoja tarjotaan käyttöön mahdollisimman avoimesti ja maksutta. Toki on myös sellaisia aineistoja, joiden käyttöä joudutaan rajoittamaan henkilösuojan tai tekijänoikeuden perusteella.

Multilingual­ Finnic. Language 400 contact and change. 401–423. Uralica Helsingiensia 14. Helsinki 2019. ‹https://doi.org/10.33341/uh.85044› TONI SUUTARI & ULRIIKKA PUURA KOTIMAISTEN KIELTEN KESKUKSEN ITÄMERENSUOMALAISET AINEISTOT

2. Murteita ja nimistöä synonyymien tuntemus on tietojen hakemisen kannalta olennaista. Syno­ nyymikortistoa voi käyttää vapaasti verkossa osoitteessa ‹https://www. Suomen murteiden sana-arkistoa (SMSA) alettiin kartuttaa 1900-lu­ kotus.fi/aineistot/sana-aineistot/suomen_murteiden_sana-arkisto/ vun alussa Suomen murteiden sanakirjan toimitustyötä silmällä pitäen. kortistot_ja_luettelot›. Keruutyön taustalla on E. N. Setälän sanakirjaohjelma vuodelta 1896. Nimiarkisto sai alkunsa vuonna 1915, kun useat tieteelliset seu­ Keruista huolehti alkuun Suomalaisen Kirjallisuuden Seura, vuosina rat perustivat paikannimitoimikunnan. Tuolloin alkoi suomenkielisen­ 1916−1924 Suomen kielen sanakirjaosakeyhtiö ja vuosina 1924−1976 paikannimistön systemaattinen keruu. Suomen ruotsinkielistä paikan­ Sanakirjasäätiö, jolta kokoelmat siirtyivät Kotukselle. nimistöä oli alettu kerätä jo 1800-luvun puolella, ja asiakirjanimien­ Murresanastoa ovat tallettaneet koulutettujen kerääjien ohella poiminnan oli aloittanut tieteellisten seurain paikannimikomitea­ merkittävissä määrin myös maallikkokirjeenvaihtajat, jotka ovat vas­ vuonna 1909. Saamenkielisiä nimiä alettiin kerätä 1930-luvulla, ja tanneet lehdissä, erityisesti vuosina 1927–1990 julkaistussa Sanasta- 1950-luvulla paikannimien keruun rinnalle tulivat myös henkilön­ jassa, esitettyihin kyselyihin. Myös murteenkeruukilpailuja on järjes­ nimien ja muiden nimien keruu. tetty. Suomen murteiden sanakirjan tavoitteena on 1900-luvun alun Alkuperäisiä tai niihin rinnastettavia, pitäjittäin järjestettyjä kielimuodon kuvaus, joten arkiston sana-aineistoa ei enää kartuteta paikan­nimilippuja Suomesta ja lähialueilta on noin 2,7 miljoonaa, vaan nykymurteita tallennetaan muihin kokoelmiin. Painettuna murre­ ja kokoelmiin liittyviä keruukarttoja on noin 10 000. Lisäksi kopio­ sanakirjasta on ilmestynyt kahdeksan osaa, ja nykyisin sanakirja on lipuista on muodostettu koko Suomen kattava kaikki nimet samassa siirtynyt ilmestymään verkossa. Sanakirjassa arkistoon kerätyt tiedot aakkostuksessa sisältävä yleiskokoelma. Paikannimikokoelmat ovat ovat toimitettuina helposti ja tiiviisti käytettävissä osin painettuna, pääosin kielitieteellisen koulutuksen saaneiden tutkijoiden ja opiske­ osin osoitteessa ‹http://kaino.kotus.fi/sms/› – jossa vuonna 2019 on lijoiden kentältä tallentamaa suullista perimätietoa. Nimen lisäksi on saatavissa sanaväli a–lysmä. siis tallennettu nimiin ja nimettyihin paikkoihin liittyvää oheistietoa. Murrearkiston pääkokoelmassa, jossa kaikki sanaliput ovat Jonkin verran tietoja on kerätty nimikilpailujenkin avulla. aakkos­järjestyksessä, on reilut 8 miljoonaa sanalippua n. 350 000 eri Nimiarkiston digitalisointia viimeistellään parhaillaan. Kaikki sanasta. Pelkästään murteissa esiintyvän sanaston lisäksi mukana on paikannimiliput, keruukartat ja muutkin Nimiarkiston nimikokoelmat runsaasti myös murteille ja yleiskielelle yhteistä sanastoa. Arkistossa on skannattu kuvatiedostoiksi. Paikannimilipuista on siirretty tieto­ on myös sanaston erilliskokoelmia ja muita murteisiin liittyviä doku­ kantamuotoon paikannimen hakuasu, paikannusviite, keruupitäjä, mentteja. Aiemmin murrearkistoon kuului myös noin 1,5 miljoonan kerääjä, keruuvuosi sekä paikan laji. Sijainti on tarkennettu keruu­ sananparren kokoelma, jonka ylioppilasosakunnat ovat keräyttäneet karttojen avulla pistekoordinaattitasolle. Lisäksi niiden nimien osalta, 1930-luvulla. Kotus luovutti sananparsikokoelman vuonna 2016 jotka ovat samoja kuin Maanmittauslaitoksen Paikannimirekisterissä,­ Kansallis­arkistolle, joka on digitoinut kokoelman ja avannut sen tietokantaan on lisätty Paikannimirekisterin paikannimi-id. Näin syn­ verkossa vapaasti käytettäväksi osoitteessa ‹http://digi.narc.fi/digi/ tynyt paikkatietoaineisto julkaistiin avoimena datana käyttöön vuo­ dosearch.ka?sartun=385077.KA›. den 2017 lopussa osoitteessa ‹www.nimiarkisto.fi›. Sähköisen nimi­ Murrearkiston sanalippuja ei ole ainakaan toistaiseksi digitoitu arkiston julkaisu juhlisti osaltaan satavuotiasta­ Suomea. Digitaalisesta lukuun ottamatta osaa valokuvia ja piirroksia sisältävistä sanalipuista.­ nimiarkistosta­ muodostuu kansallinen tietovaranto, sillä samaan koko­ Sen sijaan kerääjä- ja vastaajakortistot ovat käytettävissä myös sähköi­ naisuuteen yhdistetään ajan myötä myös Suomen ruotsin- ja saamen­ sinä. Hiljan on skannattu kuvamuotoon myös laaja synonyymikortisto,­ kieliset paikannimikokoelmat. jonka avulla voi selvittää, mitä eri ilmauksia­ murteissa käytetään. Ar­ Digitaalisessa arkistossa ei ole kyse vain olemassa olevan tiedon kistohan on järjestetty nimenomaan hakusanoittain eikä käsitteittäin, eli käytettävyyden parantamisesta, vaan sähköinen toimintaympäristö

402 403 TONI SUUTARI & ULRIIKKA PUURA KOTIMAISTEN KIELTEN KESKUKSEN ITÄMERENSUOMALAISET AINEISTOT tarjoaa tärkeitä mahdollisuuksia tietojen rikastamiseen ja täydentä­ 3. Karjalaa monessa muodossa miseen. Kansalaiset voivat tuottaa tietoa siitä, ovatko arkistoon tal­ lennetut nimet vielä käytössä, onko nimi korvautunut uudella, onko Karjalan kielen sanakirja ilmestyi kuusiosaisena vuosina 1968–2005 nimettyä paikkaa enää olemassakaan tai mitä kokonaan uutta nimistöä Kotimaisten kielten (tutkimus)keskuksen ja Suomalais-Ugrilaisen on syntynyt. Samoin voidaan täydentää paikkoihin ja nimiin liitty­ Seuran yhteistyönä. Vuonna 2009 sanakirja julkaistiin kokonaisuu­ vää taustatietoa sekä muuntaa nimiin liittyvää tietoa kuvatiedostoista dessaan verkossa. Sanakirja perustuu Karjalan kielen sana-arkiston tietokantamuotoon. aineistoon, joka koostuu reilusta puolesta miljoonasta sanalipusta. Nimiarkistossa on aineistoja myös Suomen nykyisen ja luovute­ Pääosa aineistosta on yhtenä aakkosiin järjestettynä kokoelmana. tun alueen ulkopuolelta. Aunuksen ja Vienan alueelta on pitäjittäiset Erikseen ovat Juho Kujolan tverinkarjalan sekä Valdain ja Tihvinän paikannimikokoelmat sekä erilliskokoelmina suomalais-karjalaiset karjalaismurteiden sanaliput sekä Helmi ja Pertti Virtarannan kerää­ ja venäläiset kunnan- ja kylännimet sekä tiedot talojen ja asukkaiden mät vienankarjalan ja tverinkarjalan sanaliput. Vanhin aineisto on lukumääristä vuodelta 1905. Itä-Karjalan yleiskokoelma käsittää Vie­ 1800-luvun lopulta ja perustuu Arvid Genetzin muistiinpanoihin sekä nan ja Aunuksen nimet. Inkerin alueelta on pitäjittäiset paikannimi­ K. F. Karjalaisen keruisiin. E. V. Ahtia on kerännyt arkiston tiedoista kokoelmat sekä yleiskokoelma. Arkangelin aluearkistosta kopioituna lähes puolet vuosina 1898–1936. Sodan jälkeen tietoja keräsivät Raja- on käytössä Vienan Karjalan ja Vienanmeren paikannimikokoelma Karjalan karjalaisilta ja Suomeen siirtyneiltä Itä-Karjalan pakolaisilta (myös sähköisenä muttei verkossa). Kokoelmiin kuuluvat myös Inke­ Eino Leskinen, R. E. Nirvi ja Aimo Turunen. rin Pähkinälinnan läänin ja Karjalan asiakirjanimikoelmat. Kaikki karjalan sanalippukokoelmat on skannattu kuvamuotoon, Viron paikannimistöstä on Lauri Kettusen laatima 1500-sivui­ ja aakkosten alkupäästä nimiliput on linkitetty verkkosanakirjaan. nen luettelo, joka perustuu M. J. Eisenin Tartossa säilytettävään ko­ Koko skannattu aineisto on verkossa ladattavissa osoitteessa ‹https:// koelmaan, ja luettelon pohjalta tehdyt nimiliput. Luettelo on digitoitu www.kotus.fi/aineistot/sana-aineistot/karjalan_kielen_sana-arkisto›. mutta se ei ole vielä verkon kautta saatavilla. Kokoelmissa on myös Kuvamuotoisia sanalippuja linkitetään vähitellen sanakirjan sana- Pohjois-Viron talonnimien luettelo vuodelta 1887. Niin ikään vepsä­ artikkeleihin. Alkuperäiset sanaliput on deponoitu vuonna 2016 Itä- läisistä paikannimistä on aakkosellinen kokoelma (kerääjinä Viljo Suomen yliopistolle karjalan kielen ja kulttuurin oppiaineen haltuun. Nissilä 1942–1944 ja Jussi Rainio 1967). Nimistötietoja sisältyy myös Karjalan kielen sanaston suurkerääjä E. V. Ahtia laati myös kie­ moniin Kotukseen arkistoituihin henkilöarkistoihin, joista seuraavas­ lioppeja. Karjalan kielioppi (äänne- ja sanaoppi) julkaistiin vuonna sa on nostettu esiin keskeisimmät. Heikki Ojansuun henkilöarkisto 1938. Lauseoppi ja johto-oppi jäivät aikanaan julkaisematta mutta säi­ sisältää paikannimiä ja asiakirjapoimintoja Suomesta, Virosta (mm. lyivät arkistossa. Lauseopin alkuperäinen käsikirjoitus on digitoitu ja Fr. Kuhlbarsin Viron paikannimien kokoelma) ja liiviläisalueelta. käytettävissä Kotuksen verkkosivuilla (Ahtia, ei vuosilukua). Lisäksi Viljo Nissilän henkilöarkistossa on paikannimiin liityviä asiakirja­ verkkosivuilla ovat Karjalan Kielen Seuran tuottama konekirjoitettu poimintoja, muistiinpanoja ja nimiluetteloita mm. vepsäläisestä versio sekä lauseopista että johto-opista, samoin näköisversio paine­ asutuksesta. Nissilän aineistoihin kuuluu myös laaja ns. Päämajan tusta äänne- ja sanaopista (Ahtia 1938, 2014a, 2014b). paikannimitoimikunnan­ Itä-Karjalan kokoelma. O. A. Louhelaisen arkistossa on suomalaista ja karjalaista sekä Inkerin nimistöä, erityi­ sesti historiallisia tietoja 1500–1800-luvuilta.

404 405 TONI SUUTARI & ULRIIKKA PUURA KOTIMAISTEN KIELTEN KESKUKSEN ITÄMERENSUOMALAISET AINEISTOT

4. Muiden pienten itämerensuomalaisten kielten sana-aineistoja

Vepsän verkkosanasto osoitteessa ‹http://kaino.kotus.fi/sanat/vepsa/› sisältää Lauri Kettusen laatiman kenttämuistiinpanoihin perustuvan sanakirjan käsikirjoituksen aineiston. Tiedot on siirretty tietokantaan sellaisenaan, mutta pitäjänlyhenteet on korvattu pitäjien koko nimillä käytön helpottamiseksi. Kettunen teki kaksi matkaa vepsäläiskyliin.­ Vuosina 1917–1918 hän liikkui yksinomaan etelävepsäläisten alueel­ la, mutta vuonna 1934 hän eteni etelävepsäläisistä kylistä keski­ vepsäläisen alueen kautta äänis- eli pohjoisvepsäläisiin kyliin. Ensim­ mäisen matkansa hän teki yksin, toisella mukana olivat myös Lauri Posti ja Paavo Siro. Sanastoa Kettunen keräsi kummallakin matkalla. Ensimmäisellä matkalla hän myös äänitti vepsää, ja tämäkin aineisto on Kotuksessa arkistoituna ja digitoituna. Kotuksessa on säilytteillä laaja Reino Peltolan vepsän sana­ lippu­kokoelma, 58 lippaallista, jossa mukana on myös Setälän, Postin, Kettusen ja Tunkelon keräämää aineistoa. Kokoelmiin kuu­ luu myös inkeroismurretta (lähinnä Soikkolasta ja Laukaansuusta), vähän Gunvor Löppösen ja R. E. Nirvin keräämää vatjaa sekä Tar­ ton yliopiston kokoelmasta kopioitua etelävirolaista aineistoa. Li­ säksi Eesti Keele Instituutin viron murteiden kokoelmista on vanhat mikro­filmi­kopiot. Sähköisessä muodossa ovat lyydin, liivin ja vatjan käänteissanastot, jotka on aikanaan julkaistu painettuina (Elomaa & Laakso toim. 1986; Laakso toim. 1988, 1989). Nämä pyritään saa­ maan myös verkon kautta vapaasti käytettäviksi. Syksyllä 2016 Kotuksessa alettiin toimittaa hieman suurempien­ itämerensuomalaisten nykykielten kaksikielistä sanakirjaa: Viro– suomi-­sanakirja ilmestyi Suomen ja Viron 100-vuotisitsenäisyys­ päivien kunniaksi tammikuussa 2019. Sanakirjaprojekti on virolaisen Eesti Keele Instituutin aloittama ja organisoima yhteistyöhanke. Sana­ kirja on vapaasti käytettävissä Eesti Keele Instituutin verkkosivuilla. Vuonna 2003 ilmestynyt Suomi–viro-suursanakirja puolestaan saatiin verkkoon vapaasti käytettäväksi vuonna 2017.

Kuva 1. E. V. Ahtian karjalan kielen sanalippu. Kuva Kotuksen kokoelmista.

406 407 TONI SUUTARI & ULRIIKKA PUURA KOTIMAISTEN KIELTEN KESKUKSEN ITÄMERENSUOMALAISET AINEISTOT

5. Kielikartastot (117.3) äänneasujen eri merkitysryhmät; ’lehtipuiden rungossa vir­ taavan nesteen’ lisäksi MAHLA voi tarkoittaa itämerensuomalaisissa Itämerensuomalainen kielikartasto (IMSK) eli Atlas Linguarum kielissä ’nilaa’, ’verta’ tai ’räkää’. Fennicarum (ALFE) on syntynyt suomalais-virolais-karjalaisena Kartaston kartat jakautuvat seuraaviin päätyyppeihin: yhteis­työhankkeena ja ilmestynyt painettuna kolmessa osassa vuo­ 1) Onomasiologiset eli leksikaaliset kartat kuvaavat eri kielten tun­ sina 2004, 2007 ja 2010. Toimittamisesta ovat vastanneet Kotimais­ teman yhteisen käsitteen nimityksiä. Onomasiologinen kartta­ ten kielten (tutkimus)keskus, Eesti Keele Instituut ja Venäjän Tiede­ tyyppi on kartastossa selvästi yleisin. akatemian Karjalan tiedekeskuksen Kielen, kirjallisuuden ja historian­ 2) Formatiivikartat kuvaavat kielenilmiön, erityisesti sanan muo­ instituutti. Tietokantamuotoiseksi aineistoksi kartasto on muokattu don levikkiä. Formatiivilla tarkoitetaan sanan äänteellisen hah­ Kotimaisten kielten keskuksessa Avoin data ja paikkatieto -projekteis­ mon yleistystä, ”etymologista sanaa”, esimerkiksi LÄMMIN- sa vuosina 2013 ja 2014. Aineisto on avoimena datana käytettävissä formatiivi on yleistys sanoista lämmi(n), lämmyn, lämi, lämmi CSC:n ylläpitämässä AVAA-palvelussa osoitteessa ‹http://avaa.tdata. jne. Formatiivit on pääsääntöisesti kirjoitettu versaalikirjaimin. fi/web/avaa/-/kotus-kielikartastot›. Tarjolla on käyttöliittymä, mut­ Pelkän formatiivin levikin lisäksi voidaan kuvata myös forma­ ta lisäksi aineiston voi ladata csv- ja shape-muodoissa tai hyödyntää tiivin kattamien sanojen äänne- ja muotopiirteiden levikkejä ja kyselyrajapintapalvelua. semantiikkaa. Usein karttoihin on yhdistetty kahden tai useam­ Itämerensuomalainen kielikartasto keskittyy pääasiassa sanasto­ man, tavallisesti samaa merkitsevän, formatiivin tiedot. ilmiöihin, ja sen tavoitteena on kuvata itämerensuomalaisten kielten 3) Semanttiset kartat havainnollistavat sanan eri merkitysten levik­ keskinäisiä suhteita sekä kontakteja naapurikieliin. Kielenilmiöiden kiä. Esimerkiksi formatiivilla HAKO (suomen hako, viron haga) levikit kuvastavat itämerensuomalaisten kielten jakautumista nykyi­ on seuraavia merkityksiä: ’havupuun lehvä’, ’maatunut tai veteen siksi tytärkieliksi. Sanasto liittyy kiinteästi myös kielenulkoisiin kult­ kaatunut puu’, ’oksa’, ’neulanen’. Jos kyseessä on semanttinen tuuri-ilmiöihin, joten kartasto palvelee myös kulttuurimaantiedettä. kartta, tämä on usein kerrottu kartan otsikkotiedoissa. Kartoissa kuvattavat tiedot on kerätty 317 kysymystä sisältävän 4) Motiivi- eli miellekartat kuvaavat käsitteen taustalla olevaa kyselysarjan perusteella. Aihepiirit kattavat painetun kartaston joh­ yhteistä motiivia. Esimerkiksi perunan nimitys voi perustua dannon sanoin ”koko alkukantaisen inhimillisen elämän alan”. Kysy­ ”päärynän” (peruna < muinaisruotsi, vrt. nykyruotsin päron myksistä valtaosa on onomasiologisia, eli ne kartoittavat tietyn käsit­ ’päärynä’, jord|päron ’maapäärynä = peruna’), ”omenan” teen nimityksiä, kuten ilmasuuntien tai viikonpäivien nimityksiä. Osin (maaomena), ”nauriin” (na˒ggõrz) tai ”munan” (maamuna) nimi­ kartoitetaan alkuperältään saman sanan alueittain vaihtelevia muotoja tykseen. Jos kyseessä on motiivikartta, tämä on usein kerrottu ja merkityksiä. Kartasto sisältää monen tyyppisiä karttoja, ja kaikkiin kartan otsikkotiedoissa. karttoihin liittyy yksityiskohtainen nelikielinen kommenttiosuus, joka selittää kartoissa kuvattavia ilmiöitä ja niiden variaatiota. (Verkko­ Edellä kuvattujen karttatyyppien lisäksi painetun kartaston johdan­ sovellus sisältää vain karttojen sisältämän tiedon, mutta sovelluksen nossa mainitaan frekvenssi- eli tilastokartat sekä ideaalityyppi- eli re­ karttatasolta on ladattavissa painetun kartaston 2. ja 3. osasta pdf:nä konstruktiokartat, mutta näitä karttatyyppejä esiintyy vain vähän. kartta ja siihen liittyvä seliteosa. 1. osan pdf:t on tarkoitus liittää mu­ Kartasto kuvaa 1900-luvun alun kielimuotoa. Tiedot on kerätty kaan tulevaisuudessa.) 259 perushavaintopisteestä eli kartoituspisteestä, jotka jakautuvat kie­ Usein kartoituksen tulos on jaettu useaksi kartaksi. Esimerkiksi littäin seuraavasti: suomi 185, karjala 29, vepsä 7, inkeroinen 3, vatja MAHLA-formatiivi kuvataan kolmessa kartassa, joista kaksi ensim­ 3, viro 30 ja liivi 3. Tietoja on hankittu sekä arkistoista että kenttä­ mäistä (117.1 ja 117.2) esittelee eri äänneasut (mahla ym.) ja kolmas keruin. Suomen ja viron tiedot perustuvat yksinomaan arkistojen

408 409 TONI SUUTARI & ULRIIKKA PUURA KOTIMAISTEN KIELTEN KESKUKSEN ITÄMERENSUOMALAISET AINEISTOT kokoelmiin, ja karjalan ja vepsän tiedot ovat lähes kokonaan kenttä­ 6. Äänitteitä kaikista itämerensuomalaisista keruun tulosta. kielistä Sähköinen kartasto noudattaa painetun kartaston hierarkiaa ja kuvaustapaa, mistä aiheutuu muun muassa tietojen toisteisuutta. Käy­ Suomen kielen nauhoitearkiston kokoelmissa on suomen murteiden tännössä tämä näkyy esimerkiksi kartoissa, jotka kuvaavat ’muikun’ ja puhekielen lisäksi paljon myös suomen sukukielten äänitteitä. Par­ (Coregonus albula) nimityksiä. Ensimmäisessä kartassa esitetään haiten on edustettuna karjala (n. 1 500 tuntia). Runsaasti on kerätty eri nimitysten levikit formatiivitasolla (muikku, rääpys, maiva, sil- myös eri saamen kieliä, viroa ja unkaria. Aivan vähäisiä eivät ole mui­ li) ja seuraavissa kartoissa kuvataan muikun ja rääpyksen eri muo­ denkaan sukukielten kokoelmat. tojen levikit, jolloin MUIKKU-levikki (256.1) on sama kuin muikku Valtaosa nauhoitearkiston äänitteistä sisältää haastatteluja ja muje muotojen yhteislevikki (256.2) ja RÄÄPYS-levikki on sama ja kerrontaa, mutta erityisesti sukukielisissä äänitteissä on myös kuin rääpys, rääppö ~ rääpöi, rääbis, rääpus ja rääpyskä -muotojen kansan­musiikkia. Tiedot kielenoppaasta, haastattelijasta ja tallennus­ yhteis­levikki (256.3). tilanteesta on tallennettu kokoelmatietokantaan, mutta tietoja tallen­ Samassa palvelussa IMSK:in kanssa on tarjolla tietokantamuotoi­­ teiden sisällöistä ei juuri ole. Nauhoitearkiston kokoelmatietokanta sena myös Lauri Kettusen murrekartasto, joka on ilmestynyt alun on käytettävissä erillisen liittymän kautta Kotuksen verkkosivuilla perin Suomalaisen Kirjallisuuden Seuran kustantaman Suomen mur- osoitteessa ‹http://kaino.kotus.fi/naark/›. Tarkoituksenmukaista olisi teet -teossarjan osana. I osassa (1930) Kettunen julkaisi keräämiään yhdistää tiedot Kotuksen yleiseen aineistotietokantaan, minkä jäl­ murrenäytteitä, II osassa (1930) hän kuvasi suomen murrealueet, ja III keen myös av-aineistojen tiedot näkyvät Finna-hakupalvelun kautta. osa (1940) on kahteen niteeseen jakautuva murrekartasto. Osa III A Henkilö­suoja tosin asettaa rajoituksia sille, missä laajuudessa tietoja sisältää 213 karttaa ja osa III B karttoihin kuuluvat yksityiskohtaiset voidaan avoimessa verkossa näyttää. selitykset. Litterointeja on tehty suomen murteiden äänitteistä noin tuhan­ Kettusen karttojen sisältö on siirretty tietokantamuotoon 1990-lu­ nesta tunnista, mutta sukukielten äänitteistä litterointeja on varsin vun lopussa Kanadassa Yorkin yliopistossa Sheila Embletonin ja Eric S. vähän. Usein aineiston valinnassa onkin tukeuduttava äänitteiden Wheelerin johtamassa Finnish Dialect Atlas Project -hankkeessa. Kotus kuuntelemiseen. Tutkijoiden julkaisuista, keruu- ja matkakertomuk­ sai aineiston vuonna 2000. Aineisto tarkastettiin ja muokattiin nykyi­ sista ja muistiinpanoista voi myös löytyä arvokasta tietoa aineiston seen asuunsa vuodesta 2009 alkaen Urho Määtän johtamassa monitie­ käytön tueksi. Esimerkiksi Pertti Virtarannan julkaisuissa ja syste­ teisessä BEDLAN-tutkimushankkeessa (Biological Evolution and the maattisissa muistiinpanoissa on paljon taustatietoa informanteista ja Diversification of Languages Project) yhteistyössä Kotuksen kanssa. äänitystilanteista. Kettusen kartasto kuvaa suomen murteiden äänne- ja muoto- Nauhoitearkiston perustaja Pertti Virtaranta itse oli arkiston ahke­ opillisia piirteitä. Kartat on jaoteltu äänneopillisin perustein. Pää­ rimpia kerääjiä. Hän oli erityisen kiinnostunut suomesta ja karjalasta, jako on konsonantismi (kartat 1–140) ja vokalismi (kartat 141–213). mutta nauhoitti entisen Neuvostoliiton alueelle suuntautuneilla lukui­ Karttojen otsikointi on Kettusen kartaston sisällysluettelon mukainen. silla retkillään myös lyydiä, viroa, vepsää, inkeroista, vatjaa ja liiviä Verkkosovelluksesta on ladattavissa myös painetun kartaston kartat (keruumääräjärjestyksessä). Karjalaa Virtaranta tallensi 1950-luvun pdf-tiedostoina. loppupuolelta alkaen lähes vuosittain aina 1980-luvun lopulle saakka. Kotus on ollut mukana myös Atlas Linguarum Europaen (ALE) Usein hänen mukanaan keruuretkillä kulki puoliso Helmi Virtaranta toimitustyössä. Aiemmin Kotuksessa säilytteillä olleet materiaalit on ja lisäksi suomalainen kollega tai paikallinen asiantuntija, kuten esi­ siirretty Turun yliopistoon lukuun ottamatta toimitustyön tueksi koot­ merkiksi itäkarjalainen kielentutkija Grigori Makarov. Pertti Virta­ tua Euroopan kielikartaston sanalipustoa. rannan ja Väinö Kaukosen kesällä 1968 äänittämistä karjalankielisistä

410 411 TONI SUUTARI & ULRIIKKA PUURA KOTIMAISTEN KIELTEN KESKUKSEN ITÄMERENSUOMALAISET AINEISTOT

vatjaa, liiviä ja unkaria 1914–1932) ja A. O. Väisäsen (mordvaa 1914, viroa 1922) äänityksiä. Osa kokoelmista on vielä tarkemmin käymättä läpi ja tunnistamatta. Monet sukukielten lieriökokoelmat siirtyivät kerääjiltä Suoma­ lais-Ugrilaiselle Seuralle ja sitä kautta vuonna 1959 perustetulle Suo­ men kielen nauhoitearkistolle, missä osa lieriöistä kopioitiin avokela­ nauhoille 1960-luvulla. Kaikki lieriöt jäivät Vironkadulle Helsingin yliopiston fonetiikan laitoksen ullakolle, kun nauhoitearkisto muutti vuonna 1988 Sörnäisiin. Lieriöt ”löydettiin” reilut kymmenen vuotta sitten, minkä jälkeen ne järjestettiin, luetteloitiin ja pakattiin vanerisiin reikäpehmustein varustettuihin laatikkoihin. Vuonna 2008, kun Ko­ tus muutti Kaisaniemeen, lieriöt pääsivät muiden tallenteiden kanssa asianmukaisiin ja vakioilmastoituihin arkistotiloihin. Kotuksen muut­ taessa Hakaniemeen vuonna 2016 av-aineistojen säilytys ulkoistettiin, koska nykyisessä kiinteistössä ei ole tarvittavia erikoistiloja. Vuonna 2009 kaikki toistokelpoiset lieriöt digitoitiin nauhoite­ arkiston 50. toimintavuoden kunniaksi. Arvokas lisä kokoelmiin saa­ tiin samana vuonna, kun Kansallisarkisto luovutti E. N. Setälän lieriöt (liiviä 1912, vepsää 1916) Kotukselle. Myös nämä lieriöt on digitoitu. E. N. Setälä tallensi vepsää vuonna 1916 yhdessä A. O. Väisä­ sen kanssa. Tältä retkeltä on säilynyt eri arkistoissa paljon materiaa­ lia. Väisäsen valokuvat ovat Museoviraston kuva-arkistossa. Lisäksi Väisänen on luetteloinut aineiston ja litteroinut sen sisältämät lau­ lut. Muistiinpanoista käyvät ilmi pitäjät, esiintyjät ja heidän ikänsä. Käsin kirjoitettu luettelo ja joitakin litterointeja sisältävä muistikirja Kuva 2. Pertti Virtaranta haastattelemassa karjalaisnaista. Kuva Museoviraston ovat Kansallisarkistossa (kopio Kotuksessa). Konekirjoitettu luettelo kokoelmista. sekä puhtaaksikirjoitetut ja tarkastetut laulujen tekstilitteraatiot ovat puolestaan Suomalaisen Kirjallisuuden Seuran arkistossa Väisäsen ai­ nauhoista (n. 90 tuntia) on laadittu sisällysluettelot, jotka tutkija voi neiston osana (näistäkin kopiot Kotuksessa). Lieriöt sisältävät lauluja, saada digitaalisina käyttöönsä. Kerronnan lisäksi aineisto sisältää hää- runoja ja itkuvirsiä, paimensoittoa, eläinten kutsuntaa, huhuilua sekä ja itkuvirsiä, joikuja, kehto- ja muita lauluja. kanteleen- ja harmonikansoittoa. Nauhoitearkistossa on myös Väisä­ Nauhoitearkiston vanhimmat äänitteet ovat vahalieriötallenteita sen haastattelu, jossa hän ikämiehenä muistelee muun muassa vepsän 1900-luvun alkuvuosikymmeniltä. Kokoelmissa on noin 950 lieriötä, maille suuntautunutta keruuretkeä. muun muassa K. F. Karjalaisen (hantia 1900–1902), G. J. Ramstedtin Lauri Kettunen suuntasi vepsän puhuma-alueille pian Setälän (kalmukkia 1903), Artturi Kanniston (mansia 1905–1906), Yrjö Wich­ jälkeen vuonna 1917–1918. Vuonna 1918 hän äänitti etelävepsäläisis­ mannin (maria 1906), Kai Donnerin (samojedia 1912–1914), T. I. It­ sä kylissä muun muassa satuja, tarinoita ja lauluja. Kuurinmaan liivi­ kosen (koltansaamea 1913), Lauri Kettusen (suomen murteita, vepsää, läiskylissä Kettunen vieraili toistakymmentä kertaa. Vuosina 1920

412 413 TONI SUUTARI & ULRIIKKA PUURA KOTIMAISTEN KIELTEN KESKUKSEN ITÄMERENSUOMALAISET AINEISTOT

Kuva 3. Sakari Pietarila digitoimassa Kotuksen vahalieriöitä vuonna 2009. Valo­ kuvaaja Toni Suutari. Kuva Kotuksen kokoelmista. Kuva 4. E. N. Setälä vuonna 1916 Korvoilan kylässä vepsäläisten luona haastattele- massa 11-vuotiasta Mark-poikaa. Taustalla näkyy äänittämisessä käytetty parlografi ja vahalieriöitä. Valokuvaaja A. O. Väisänen. Kuva Museoviraston kokoelmista.

ja 1923 hän äänitti satuja ja tarinoita. Vuodelta 1927 Kettuselta on 7. Etymologiset aineistot yksi Helsingissä tehty liivinkielinen äänitys, ja lisäksi on muutama tallennusajan­ ja ‑paikan suhteen tunnistamattomaksi jäänyt äänitys. Etymologinen arkisto sisältää suomen sanaston historiaa sekä kieli­ Vatjaa Kettunen tallensi vuosina 1915, 1923 ja 1932, satuja, lauluja, tieteellisesti että asia- ja kulttuurihistoriallisesti selvittävää aineistoa. muistelmia ja muuta vastaavaa. Kaikkiaan Kettusen eri kielistä ää­ Aineistoa on käytetty Suomen kielen etymologisen sanakirjan ja Suo- nittämiä lieriöitä on säilynyt lähes kaksisataa. Kettunen on julkaissut men sanojen alkuperä ‑teoksen toimitustyössä. Teosten ilmestyttyä osan äänittämästään aineistosta teksteinä, ja lisäksi hänen kirjoina il­ etymologista arkistoa ei ole enää kartutettu. Erillisenä kokoel­mana mestyneet matkakertomuksensa sisältävät paljon tietoa kielenoppaista on suomen etäsukukielten sellaisten sanojen etymologioita, joille ei ja keruutilanteista.­ Kotuksen verkkosivuilla osoitteessa ‹http://www. ole vastineita itämerensuomessa. Sähköisenä aineistona on käytös­ kotus.fi/aineistot/puhutun_kielen_aineistot/sukukielten_vahalierio­ sä Etymo­loginen viitetietokanta, joka nimensä mukaisesti sisäl­ kokoelmat› on näytteitä ja kattavat kirjallisuusviittein varustetut luette­ tää viittauksia sanojen etymologioita käsittelevään kirjallisuuteen lot Kettusen tallentamista vepsän, liivin ja vatjan vahalieriöäänitteistä. (http://kaino.kotus.fi/sanat/evita/).

414 415 TONI SUUTARI & ULRIIKKA PUURA KOTIMAISTEN KIELTEN KESKUKSEN ITÄMERENSUOMALAISET AINEISTOT

Saamelaiskielten etymologinen tietokanta Álgu osoitteessa • Eino Leskisen henkilöarkistossa on karjalan kieltä koskevia ‹http://kaino.kotus.fi/algu/› sisältää n. 88 000 saamelaiskielten sanaa lehtiartikkeleita, luetteloita, sanastokartunnan luetteloita sekä etymologisine tietoineen. Eniten tietoja on tutkituimmasta ja puhu­ Karjalan kielen sanakirjan kokouspöytäkirjoja. tuimmasta kielestä pohjoissaamesta (n. 35 000 sanaa). Tietokantaan • Pertti Virtarannan henkilöarkisto on hyvin laaja ja sisältää muun on koottu suurin osa tähän asti ilmestyneestä saamelaiskieliä koske­ muassa hänen yksityiskohtaisia muistiinpanojaan keruumatkoilta. vasta etymologisesta tiedosta. Aineistoa on paljon monesta itämerensuomalaisesta kielestä. Etymologiset tiedot näkyvät tietokannassa sanojen välisinä suh­ teina, jotka kertovat esimerkiksi sanavastaavuuksista tai lainautumi­ Lähisukukansoihin ja -kieliin liittyvää aineistoa on myös muissa sesta. Tällä hetkellä tietokannassa on lähes 0,5 miljoonaa suhdetietoa.­ erilliskokoelmissa.­ Vaikkapa Suomen murteiden sana-arkiston Tietokanta antaa tietoa myös etymologisesta kirjallisuudesta, sillä käsi­kirjoitusten joukosta löytyy Laina Syrjäsen kirjoitus Vatjalaisten suhteen lähdetieto on aina näkyvissä. leivän valmistuksesta ja Paavo Siron Vatjalaisten teurastustapoja. Kir­ Kotuksessa on toimitettu etymologinen sanakirja Suomen sanojen jallisen aineiston lisäksi löytyy valokuviakin, muiden muassa Reino alkuperä, joka ilmestyi vuosina 1992, 1995 ja 2000. Sanakirjasta tehty Peltolan kuvia Karjalan Šungusta vuodelta 1942. Erityisesti nauhoite­ sähköinen versio on ollut tutkijoiden käytettävissä, ja tavoitteena on, että arkiston aineistoihin liittyviä valokuvia on myös Museo­viraston kuva- jossakin vaiheessa sanakirja voitaisiin avata laajemmin käytettäväksi. arkistossa, esimerkiksi kuvia Setälän Liivinmaan matkalta vuodelta 1912 sekä Setälän ja A. O. Väisäsen vepsäläisalueen matkalta vuo­ delta 1916. Myös Pertti Virtarannan laaja kenttäkeruiden yhteydessä 8. Variaa karttunut valokuva-arkisto on luovutettu Museovirastolle.

Erilaisia pieniä itämerensuomalaisten kielten kokoelmia, kenttä­ muistiinpanoja,­ käsikirjoituksia ja vastaavia dokumentteja sisältyy 9. Lopuksi moniin Kotuksessa säilytettäviin tutkijoiden henkilöarkistoihin. Suo­ men kieleen liittyviä tietoja on eniten, mutta muihin itämerensuoma­ Edellä esillä olleiden aineistojen lisäksi Kotuksessa on paljon erilai­ laisiin kieliin liittyviä tietoja löytyy ainakin (kaikkia ei ole yksityis­ sia suomen kielen aineistoja, esimerkiksi Nykysuomen sana-arkisto, kohtaisesti luetteloitu) seuraavista henkilöarkistoista: Vanhan kirjasuomen arkisto, 1800-luvun suomen kokoelmat ja mo­ nia sähköisiä tekstikorpuksia. Kotuksen aineistoihin voi perehtyä • Elvi Erämetsän henkilöarkistossa on kasvinnimitietoja vat­ pääpiirteissään Kotuksen verkkosivuilla ja Finna-hakupalvelussa jasta, karjalasta, liivistä, virosta sekä muistiinpanoja Eesti Rahva (http://www.kotus.fi/aineistot, https://www.finna.fi/), mutta kiinnostu­ Muuseumin kasviväritiedoista, myös Vilho Setälän muistiin­ neen kannattaa vierailla arkistossa, tutustua kokoelmiin ja keskustella panoja vatjasta. Erämetsän käsikirjoitus kasvivärjäyksestä on henkilökunnan kanssa, sillä luettelointitiedot eivät ole kattavia. Henkilö­ digitoitu ja saatavissa arkistojen asiakaspalvelun kautta. kunnalla on hiljaista tietoa, joka voi auttaa tiedon lähteille, ja varsinkin • Helmi Helmisen henkilöarkistossa on Karjalaa, Inkeriä ja moniin sisällöltään epäyhtenäisiin­ erilliskokoelmiin ja henkilöarkistoi­ Viroa koskevaa aineistoa, muistiinpanoja ja kuvia Karjalasta hin voi sisältyä materiaalia, joka on aineistoa läpikäymällä tutkijan itse sekä tietoja inkerikoista ja vatjalaisista. löydettävä. Tutkijoilta ja muilta käyttäjiltä saatava tieto kokoelmien si­ • Jalo Kaliman henkilöarkistossa on karjalaisia paikannimiä sällöstä otetaan mielellään vastaan ja se voidaan liittää täydentämään ja sekä sana- ja nimitietoja suomalais-ugrilaisten kansojen asuma- tarkentamaan luettelointia. Myös digitointiehdotuksia kannattaa esittää. alueilta Venäjältä. Vuorovaikutus arkiston ja asiakkaiden välillä on kaikille hyödyksi.

416 417 TONI SUUTARI & ULRIIKKA PUURA KOTIMAISTEN KIELTEN KESKUKSEN ITÄMERENSUOMALAISET AINEISTOT

Lähteet

Koonnut Sofia Björklöf (Helsingin yliopisto & Kotimaisten kielten keskus) Ahtia, E. V. 1938: Karjalan kielioppi. Äänne- ja sanaoppi. Suojärvi: Kar­ jalan kansalaisseura. Karjalan Kielen Seura julkaissut näköispainok­ sena 2009 ja antanut pdf-version julkaistavaksi Kotuksen aineisto­ palvelu Kainossa 2014. Saatavissa: ‹http://kaino.kotus.fi/digitointi/pdf/ Ahtian_sana-_ja_aanneoppi.pdf› Ahtia, Edvard V. 2014a: Karjalan kielioppi II. Johto-oppi. Konehkirjutti Sirpa Hentula, skanniruičči, toimitti i lad’d’ai Martti Penttonen, ilmah­ piästäi Karjalan Kielen Seura r.y. 2014. Saatavissa: ‹http://kaino.kotus. fi/digitointi/pdf/Ahtian_johto-oppi.pdf› — 2014b: Karjalan kielioppi III. Lauseoppi. Skanniruičči Kotus, koneh­ kirjutti Tuomo Salonen, toimitti i lad’d’ai Martti Penttonen, ilmahpiäs­ täi Karjalan Kielen Seura r.y. 2014. Saatavissa: ‹http://kaino.kotus.fi/ digitointi/pdf/Ahtian_lauseoppi.pdf› — Karjalan kielen lauseoppi. Käsikirjoitus. Säilytteillä Kotimaisten kielten keskuksessa. Saatavissa: ‹http://kaino.kotus.fi/digitointi/pdf/ Ahtia_Karjalan_kielen_lauseoppi.pdf› ALE = Atlas Linguarum Europae. 1975–2007. Assen: Van Gorcum & Roma: Poligrafico. ALFE = Atlas Linguarum Fennicarum 1–3. Itämerensuomalainen kielikartasto. Läänemeresoome keeleatlas. Ostseefinnischer Sprachatlas. Lingvističe­ skij atlas pribaltijsko-finskix jazykov. 2004–2010. Helsinki: SKS & Kotimaisten kielten tutkimuskeskus. Saatavissa: ‹https://avaa.tdata.fi/ web/avaa/-/kotus-kielikartastot› Álgu. Saamelaiskielten etymologinen tietokanta. Helsinki: Kotimaisten kielten keskus. Saatavissa: ‹http://kaino.kotus.fi/algu/› Digitaalinen Nimiarkisto. Helsinki: Kotimaisten kielten keskus. Saatavissa: ‹https://nimiarkisto.fi/› Eesti-soome sõnaraamat. Viro–suomi-sanakirja 2019. Verkkosanakirja. Pää­toim. Heikki Hurtta & Margit Langemets. Tallinn–Helsinki: Eesti Keele Instituut & Kotimaisten kielten keskus. Saatavissa: ‹http://www. eki.ee/dict/efi/› Elomaa, Jarmo & Johanna Laakso (toim.) 1986: Lyydiläismurteiden kään- teissanasto. LSFU IX:2, KKTKJ 44. Helsinki: Suomalais-Ugrilainen Kuva 5. Karjalaisnainen Reino Peltolan kuvaamana Šungun kylässä vuonna 1942. Seura & KKTK. Kuva Kotuksen kokoelmista. Etymologinen arkisto. Helsinki: Kotimaisten kielten keskus.

418 419 TONI SUUTARI & ULRIIKKA PUURA KOTIMAISTEN KIELTEN KESKUKSEN ITÄMERENSUOMALAISET AINEISTOT

Etymologinen viitetietokanta. Viitteitä vuodesta 1966 2000-luvun alkuun SMSA = Suomen murteiden sana-arkisto. Helsinki: Kotimaisten kielten mennessä ilmestyneeseen kirjallisuuteen. Helsinki: Kotimaisten kiel­ keskus. (Katso: ‹http://www.kotus.fi/aineistot/sana-aineistot/suomen_ ten keskus. Saatavissa: ‹http://kaino.kotus.fi/sanat/evita/› murteiden_sana-arkisto›) Euroopan kielikartaston sanalipusto. Helsinki: Kotimaisten kielten keskus. Suomen kielen etymologinen sanakirja I–VII. Toivonen, Y. H., Erkki Henkilöarkistot. Helsinki: Kotimaisten kielten keskus. Tarkempia tietoja Itkonen, Aulis J. Joki & Reino Peltola 1955–1981. LSFU XII. Sana­ henkilöarkistojen sisällöistä saatavissa: ‹http://www.kotus.fi/aineistot/ hakemiston (VII) koostaneet Satu Tanner ja Marita Cronstedt. Hel­ tietoa_aineistoista/henkiloarkistot› sinki: Suomalais-Ugrilainen Seura. Karjalan kielen sana-arkisto. Digitoidut sanaliput (alkuperäinen aineisto Suomen kielen nauhoitearkisto. Helsinki: Kotimaisten kielten keskus. siirretty helmikuussa 2016 Joensuuhun Itä-Suomen yliopiston yhtey­ Ko­koel­ma­tietokanta saatavissa: ‹http://kaino.kotus.fi/naark/› teen). Helsinki: Kotimaisten kielten keskus. ‹https://www.kotus.fi/ Suomen murteiden sanakirja 1–8 (a–kurvottaa). 1985–2008. KKTKJ 36. aineistot/sana-aineistot/karjalan_kielen_sana-arkisto› Helsinki: Kotimaisten kielten tutkimuskeskus. Karjalan kielen sanakirja 1–6. 1968–2005. Toim. Pertti Virtaranta (1–3), Suomen murteiden sanakirja (a–lysmä). 2012–. Kotimaisten kielten keskuk­ Raija Koponen (4–6). LSFU XVI, 3–6 myös KKTKJ 25. Helsinki: sen verkkojulkaisuja 30. Helsinki: Kotimaisten kielten keskus. Päivi­ Suomalais-Ugrilainen Seura & Kotimaisten kielten tutkimuskeskus. tetty 22.5.2019. Saatavissa: ‹http://kaino.kotus.fi/sms› Karjalan kielen verkkosanakirja. Toim. Marja Torikka, vastaava toimittaja Suomen sanojen alkuperä. Etymologinen sanakirja 1–3. 1992–2000. Päätoi­ 8.9.2010 lähtien Leena Joki. Verkkosovellus Jari Vihtari. Kotimaisten mittajat Erkki Itkonen & Ulla-Maija Kulonen. SKST 556, KKTKJ 62. kielten tutkimuskeskuksen verkkojulkaisuja 18. Helsinki: Kotimais­ Helsinki: SKS & Kotimaisten kielten tutkimuskeskus. ten kielten tutkimuskeskus. Julkaistu 15.10.2009, päivitetty 22.2.2019. Suomi–viro-suursanakirja. Soome-eesti suursõnaraamat 2003. Päätoim. ‹http://kaino.kotus.fi/kks› Valdek Pall. Tallinn–Helsinki: Eesti Keele Instituut & Kotimaisten Kettunen, Lauri 1930a: Suomen murteet I. Murrenäytteitä. SKST 188. Hel­ kielten keskus. Saatavissa: ‹http://www.eki.ee/dict/fie/› sinki: SKS. Vepsän verkkosanasto. 2007. Helsinki: Kotimaisten kielten tutkimuskeskus. — 1930b: Suomen murteet II. Murrealueet. SKST 188. Helsinki: SKS. Saatavissa: ‹http://kaino.kotus.fi/sanat/vepsa/› — 1940: Suomen murteet III. A. Murrekartasto. B. Selityksiä murre- kartastoon. SKST 188. Helsinki: SKS. Saatavissa: ‹http://avaa.tdata. fi/web/avaa/-/kotus-kielikartastot›. Kartat pdf-tiedostoina saatavissa: ‹http://kettunen.fnhost.org/?i=1› Laakso, Johanna (toim.) 1988: Rückläufiges Wörterbuch des Livischen. Anhand des Livischen Wörterbuches von Lauri Kettunen, herausgege­ ben von Johanna Laakso. LSFU V:2, KKTKJ 47. Helsinki: Suomalais- Ugrilainen Seura & Kotimaisten kielten tutkimuskeskus. — 1989: Vatjan käänteissanasto. LSFU XXII, KKTKJ 49. Helsinki: Suo­ malais-Ugrilainen Seura & Kotimaisten kielten tutkimuskeskus. Nimiarkisto. Helsinki: Kotimaisten kielten keskus. Kokoelmatietokanta saatavissa: ‹http://kaino.kotus.fi/nimikokoelmat/?a=etusivu› (Katso: ‹https://www.kotus.fi/aineistot/nimiaineistot›) Sananparsikokoelmat. 2016–. Kansallisarkiston digitaaliarkisto. Saatavissa: ‹http://digi.narc.fi/digi/dosearch.ka?sartun=385077.KA›. 25 kunnan sananparret korpusmuodossa saatavissa: ‹http://kaino.kotus.fi/korpus/ sp/meta/sp_coll_rdf.xml›

420 421 TONI SUUTARI & ULRIIKKA PUURA KOTIMAISTEN KIELTEN KESKUKSEN ITÄMERENSUOMALAISET AINEISTOT

Finnic materials at the Institute for the oldest recordings – wax cylinders containing, for example, Esto­ the Languages of Finland nian, Veps, Votic, and Livonian – are from the beginning of the 20th century. Toni Suutari & Ulriikka Puura There are over 550 000 card-files in the Karelian Language Lexi­ cal Archive. All the card-files are digitized and can be found on the The Institute for the Languages of Finland is a central expert organiza­ Institute’s websites. The Dictionary of the Karelian language (1968– tion specializing in languages. Its core areas are language planning and 2005) is also available online. Other dictionaries compiled by the dictionary compilation, together with related research. The Institute’s Institute are also now being published on the Internet, e.g., the Dic­ tasks also include maintaining and developing linguistic archives and tionary of Finnish Dialects, and digital versions are being created of collections. These dictionaries are based on large scientific collections books originally published in paper format. accumulated over decades, and the collections of the Institute are also Other works available online include, e.g., the Online Vocabulary widely used by researchers, students, and the general public. of Veps, based on the manuscript of Lauri Kettunen, the Finnic Lan­ The archives of the Institute cover approximately 2 200 shelf me­ guage Atlas (ALFE = Atlas Linguarum Fennicarum), the Etymologi­ ters. There are millions of card-files and other documents and about cal Reference Database, and the Etymological Database of the Saami 24 000 hours of audiovisual recordings. Some of the collections have Languages (Álgu). There are also many lexical collections available been digitized, and corpora have been created directly in digital form. in paper format, which cover the Finnic languages, e.g., the card-files The digital collections include hundreds of millions of words and ap­ of Veps vocabulary collected by Reino Peltola. proximately 17 000 hours of audiovisual recordings that have been The focus of the Institute’s data services is on providing archives, digitized or collected directly in a digital format. collections, dictionaries, and other databases for public use as freely as The archives and collections of the Institute contain a great deal possible. Users will be involved in collecting and developing data by of lexical and onomastic data, but there are also many smaller collec­ increasing interaction and by finding modern ways of crowdsourcing. tions containing, for example, field notes of researchers and photo­ graphs. The main collection of the Lexical Archive and the Dictionary of Finnish dialects comprises more than 8 million items of data. The Names Archive currently contains about 2.7 million original or com­ parable place name card-files covering Finland and its neighboring territories, all itemized by parish. The cartographic collection contains about 10 000 maps connected to the place name collections. The place name collections of the Names Archive have been converted into a digital archive. The digitization project (started in 2014) was one of the initiatives celebrating Finland’s centenary in 2017. The Audio Recording Archive holds almost 24 000 hours of au­ dio and video recordings (about 70% is digitized), mainly Finnish dia­ lects and contemporary Spoken Finnish, but also a significant number of languages related to Finnish (especially the Saami languages and Karelian). Recordings have been systematically made since 1959, and

422 423 MARJATTA PALANDER, HELKA RIIONHEIMO, HANNU KEMPPANEN & JUKKA MÄKISALO Itä-Suomen yliopisto

Kielikorpuksia Suomen itärajalta

1. Johdanto

Itä-Suomen yliopistossa on pitkät perinteet raja-alueiden ja erityi­ sesti Suomen itärajan monitieteisessä tutkimuksessa. Yliopiston ny­ kyisessä tutkimusstrategiassa yksi kansainvälisistä huipputason tut­ kimusalueista on Rajat, liikkuvuus ja kulttuurien kohtaaminen, joka on edustettuna paitsi yhteiskunta- ja historiatieteissä myös kieli- ja käännöstieteissä, mm. suomen ja venäjän kielessä sekä karjalan kie­ len ja kulttuurin oppiaineessa. Näissä kieliaineissa on viime vuosina työstetty kolmea digitaalista korpusta (Raja-Karjalan korpus, Inkerin- suomen korpus ja Karjalan suomen korpus), jotka ovat jo nyt useiden tutkijoiden käytössä. Rajakarjalais- ja inkeriläismurteiden aineistot edustavat 1900-luvun jälkipuoliskolla tallennettua vanhan polven puhekieltä, kun taas Petroskoin suomen aineisto koostuu 2000-luvun mediakielestä. Näitä korpuksia yhdistää se, että niiden edustamat kielimuo­ dot ovat olleet omalla alueellaan vähemmistökieliä ja niihin on tul­ lut runsaasti kontaktivaikutusta seudun valtakielestä. Inkerinsuomi ja Petroskoin (yleiskielinen) suomi ovat siirtolaiskielimuotoja, jotka ovat syntyneet siirtolaisryhmien muuttaessa uudelle kielialueelle: inkerinsuomalaisten esi-isät siirtyivät 1600-luvulla Kannakselta ja Savosta Inkerinmaalle ja Petroskoin suomalaiset 1930-luvulla Suo­ mesta ja Yhdysvalloista Neuvosto-Karjalaan. Molemmissa suomen muodoissa venäjän kielen vaikutus on selvää. Rajakarjalaismurteita taas on alun perin puhuttu Suomen ja Venäjän välisellä raja-alueella, jossa karjalan kieli on saanut vaikutusta sekä venäjästä että suomesta.

Multilingual­ Finnic. Language contact and change. 425–438. Uralica Helsingiensia 14. Helsinki 2019. ‹https://doi.org/10.33341/uh.85045› MARJATTA PALANDER, HELKA RIIONHEIMO, KIELIKORPUKSIA SUOMEN ITÄRAJALTA HANNU KEMPPANEN & JUKKA MÄKISALO

Inkerinsuomen korpus ja Raja-Karjalan korpus ovat puolestaan siinä 2. Raja-Karjalan korpus suhteessa samanlaisia, että molemmissa on näkyvissä kaksi kielikon­ taktien kerrosta. Osa kieltenvälisestä vaikutuksesta on pitkäaikaista ja Raja-Karjalan korpus perustuu Kotimaisten kielten keskuksessa säi­ peräisin tilanteesta, jossa eri kielten puhujia on asunut samoilla seu­ lytteillä oleviin suomen kielen nauhoitearkiston murreäänitteisiin, duilla, Inkerinmaalla ja Raja-Karjalan alueella. Osa aineistoissa ilme­ jotka on tallennettu pääosin 1960-luvulla ja digitoitu vuosina 2009– nevästä vaikutuksesta on kuitenkin tuoretta ja idiolektikohtaista ja on 2011. Äänitteet edustavat luovutetun Raja-Karjalan pitäjien murteita, syntynyt haastateltujen ihmisten henkilökohtaisten kokemusten kautta joiden tutkimus on tähän saakka ollut vähäistä: 1800-luvun lopulla ja heidän asetuttuaan asumaan läheistä sukukieltä puhuvaan uuteen kie­ 1900-luvulla on ilmestynyt yksittäisiä, lähinnä äänneopillisia kuvauk­ liyhteisöön. Inkerinsuomalaisista informanteista osa oli haastatteluti­ sia (Genetz 1870; Kujola 1910; Turunen 1965, 1973, 1982), joiden lanteessa asunut noin 50 vuotta Virossa. Rajakarjalaiset oli jatkosodan perusteella saa yleiskuvan siitä, millaisia murre-eroja alueella on ollut jälkeen asutettu muualle Suomeen suomenkielisille alueille, ja haas­ ennen viime sotia. Rajakarjalaismurteet eivät kuitenkaan ole mukana tattelujen aikaan he olivat olleet Suomessa noin 20–30 vuotta. esim. Bubrihin, Beljakovin ja Punžinan Karjalan kielen murrekartas- Korpustyön tavoitteena on yhtäältä parantaa aiemmin, jo tossa (1997), ja yksityiskohtainen tieto idiolektien välisestä variaati­ 1960-luvulta lähtien koottujen puhekielisten aineistojen käytettävyyt­ osta on puuttunut kokonaan. tä ja toisaalta tarjota Karjalan suomen lehtiteksteistä yhtenäinen tutki­ Raja-Karjalan korpukseen kuuluu yhteensä noin 120 tuntia musaineisto, jollaista ei ennestään ole olemassa. Korpusten laadinta litteroitua vanhan ikäpolven (70–90-vuotiaiden) haastattelupuhetta hyödyttää myös kieliteknologista tutkimusta: korpustyössä voidaan seuraavista pitäjistä: Ilomantsi, Korpiselkä, Suistamo, Suojärvi, Im­ testata kieliteknologisten työkalujen soveltuvuutta sekakieliseen ja pilahti ja Salmi. Murrenäytteiden valintaperusteena on pidetty sitä, murteelliseen puhekielen aineistoon. Tavoitteena on, että korpukset että ne edustavat äidinkieleltään karjalankielisten puhujien murretta. tarjoaisivat monipuolisia mahdollisuuksia sekä perinteisiin että uu­ Haastateltavat ovat siis syntyneet luovutetun Raja-Karjalan alueel­ dentyyppisiin murteita ja kontaktivarieteetteja koskeviin tutkimusai­ la, mutta heidät on asutettu siirtolaisina toisen maailmansodan jäl­ heisiin. Suomalaisessa dialektologiassa on toistaiseksi ollut niukasti keen nykyisen Suomen rajojen sisäpuolelle, enimmäkseen Pohjois- käytettävissä varsinaisia korpuksia, jotka mahdollistaisivat esimer­ Karjalaan ja Pohjois-Savoon. Erityisesti Ilomantsin, Korpiselän ja kiksi yhtenäisistä teksteistä tehtävät automaattiset hakutoiminnot. En­ Impilahden kielenoppaiden puhekieli on saanut paljon vaikutteita simmäinen korpuksen muotoon koostettu murreaineisto on Lauseopin suomen savolaismurteista, kun taas Suistamon, Suojärven ja Salmin arkiston kokoelma. Se käsittää noin 130 tuntia Kotimaisten kielten murteiden puhujilla karjalan kieli on säilynyt paremmin. keskuksen Suomen kielen nauhoitearkiston litteroituja murrehaastat­ Rajakarjalaismurteiden korpus luotiin Raja-Karjalan kielikon­ teluja. Kotimaisten kielten keskuksessa on lisäksi saatu valmiiksi Suo- taktien tutkimusta varten. Valtaosa äänitteistä litteroitiin puolikar­ men kielen näytteitä -sarjan litteroitujen tekstien (100 t) yhdistämi­ keaa transkriptiota hieman karkeammalla tarkekirjoituksella vuosina nen ääninäytteisiin. Itä-Suomen yliopiston Raja-Karjalan korpuksen 2009–2011 Itä-Suomen yliopistossa opiskelijavoimin, Karjalaisen ja Inkerinsuomen korpuksen lopullinen tavoite on vastaavanlainen: Kulttuurin Edistämissäätiön rahoituksella. Kun Suomen Akatemian puheäänen kohdistaminen tarvittavalla tarkkuudella litteraatiotekstiin. rahoittama nelivuotinen tutkimushanke FINKA (Suomen ja karjalan rajalla: näkökulmia lähisukukieliin ja niiden murteisiin) perustet­ tiin 2011, aloitettiin litterointien tarkistustyö ja aineiston täyden­ täminen erityisesti Raja-Karjalan itäisimpien murteiden litteroin­ neilla. FINKA-hankkeen lisäksi korpustyötä on rahoitettu Koneen Säätiön apurahalla, joka myönnettiin vuosiksi 2013‒2015 Suomen

426 427 MARJATTA PALANDER, HELKA RIIONHEIMO, KIELIKORPUKSIA SUOMEN ITÄRAJALTA HANNU KEMPPANEN & JUKKA MÄKISALO itäpuolisten lähialueiden kielikorpukset (SILK) -hankkeelle, sekä 3. Inkerinsuomen korpus Karjalaisen Kulttuurin Edistämissäätiön apurahalla v. 2016–2017 (SILK 2). Aineisto on nyt lopullisessa koossaan, ja se on kokonaan Inkerinsuomen korpus perustuu Joensuun yliopiston tutkimushank­ tarkistettu. Korpuksen koko on noin 850 000 sanaa. Korpusta käyte­ keessa tehtyihin ja nykyisin Itä-Suomen yliopistossa säilytteillä oleviin tään FINKA-hankkeessa alkaneeseen, erityisesti suomen ja karjalan murreäänitteisiin, jotka on tallennettu 1990-luvulla. Kielentallennus­ sekamurteiden morfologisten, morfosyntaktisten ja foneettis-fono­ ta tehtiin hankkeessa Inkerinsuomalaisten kieliolot ja inkerin­suomen logisten ilmiöiden tutkimukseen. Aineistosta on valmistunut kaksi nykytila, jota johtivat suomen kielen professori Ilkka Savijärvi ja ve­ väitös­kirjatutkimusta (Kok 2016, Uusitupa 2017), siitä on valmis­ näjän kielen professori Muusa Savijärvi. Hankkeen lähtökohtana oli teilla viisi väitöskirjatutkimusta (Laura Arantola, Natalia Giloeva, inkeriläismurteiden kohtalo 1990-luvun uudessa yhteiskunnallises­ Henna Massinen, Ilja Moshnikov ja Susanna Tavi) ja tehtynä on sa tilanteessa Neuvostoliiton romahtamisen jälkeen ja vähemmistö­ useita muita tutkimuksia (Palander & Riionheimo 2018, Uusitupa, emansipaation alkuvaiheissa (hankkeen taustoista ks. esim. Riion­ Koivisto & Palander 2017, Palander, Riionheimo & Koivisto 2018, heimo 2007: 21–22). Nauhoiteaineistoa kerättiin Virossa kolmella Koivisto 2018). paikkakunnalla (Tartto, Pärnu ja Järvamaan maakunta) yhteensä noin Korpus on nykyvaiheessaan digitaalisessa muodossa oleva teksti­ 60 tuntia ja Venäjällä inkerinsuomen alkuperäisellä puhuma-alueella kokoelma, jossa jokaisen nauhoitteen litterointi on omana Word-tiedos­ Inkerinmaalla viidessä pitäjässä (Toksova, Keltto, Skuoritsa, Kupanit­ tonaan sekä Unicode-muotoisena tekstitiedostona.1 Tekstitiedostot on sa ja Narvusin Kurkolanniemi) yhteensä noin 125 tuntia. Aineistosta nyt kohdistettu äänitiedostoihin. Ääninäytteet on yhdistetty teksteihin on tähän mennessä julkaistu kolme kielennäytekokoelmaa (Riion­ Praat-ohjelmaa käyttäen ns. puoliautomaattisella nimikointimenetel­ heimo & Kivisalu 1994, Savijärvi ym. 1996, Kokko ym. 2003). mällä. Korpuksen valmistuttua ääni- ja tekstitiedostot ovat käytettävis­ Inkerin suomalaismurteista on olemassa jonkin verran aikaisem­ sä rinnakkain, jolloin tutkija voi seurata litteroitua tekstiä ja autenttista paa tutkimusta, vaikkakin se on ollut huomattavasti vähäisempää ja murrepuhetta samanaikaisesti. Äänen ja tekstin yhdistäminen palvelee epäsystemaattisempaa kuin Suomen alueella puhuttujen murteiden esimerkiksi lausepainon hyödyntämistä, sillä lausepainolla sekä äänen­ tutkimus. Ennen 1990-lukua tutkimus on tapahtunut samoin menetel­ sävyn ja -voimakkuuden vaihtelulla on merkitystä puheen syntaktisen min ja päämäärin kuin muidenkin suomen murteiden: päähuomio on rakenteen ja lausesemantiikan tutkimuksessa. Tekstin ja äänen yhdistä­ ollut äännehistoriassa, ja morfologiaa on kuvattu vain vähän, syntaksia vä korpus antaa mahdollisuuksia myös uudenlaisten tutkimuskysymys­ tuskin lainkaan. Tutkimuskohteena ovat olleet vanhat inkerinsuomen­ ten kehittelyyn: miten suomen ja karjalan kohtaaminen on vaikuttanut murteet sellaisina kuin niitä puhuttiin ennen sotia alkuperäisillä­ asu­ rajakarjalaismurteiden intonaatioon tai rytmiin (esim. lyhyttä ensi tavua ma-alueilla tiiviissä suomenkelisissä yhteisöissä, ja aineistoa on koot­ seuraavan toisen tavun lyhyen vokaalin kestoon; vrt. savolaiseen ns. tu joko 1900-luvun alussa tai sotien jälkeen iäkkäiltä kielenoppailta.­ puolipitkään vokaaliin: talò : talòssa). (Inkerin­suomen aiemmasta tutkimuksesta tarkemmin esim. Kokko 2007: 25–27, Riionheimo 2007: 20–21.) 1900-luvun loppu­puolella puhuttu inkerinsuomi on kuitenkin aivan toisenlainen kielimuoto kuin vanhat paikallismurteet. Inkerinsuomalaiset ovat Neuvosto­ liiton karkotusten ja toisen maailmansodan tapahtumien vuoksi elä­ neet pääasiassa­ alkuperäisen kotiseutunsa ulkopuolella, hajallaan eri alueilla ja eri maissa ja kaikkialla pienenä vähemmistönä toisen­ 1. Korpuksen metatiedot ovat META-SHARE-tietokannassa osoitteessa ‹http:// kielisen enemmistön keskuudessa. Kieleen ovat vaikuttaneet erilaiset meta-share.csc.fi/repository/browse/the-corpus-of-border-karelia/f2fdd49­caac211e3 90f0005056be118eda6c88241c1440678c85b11488d58ae0/›. muutosvoimat kuin perinteisiin murteisiin, esimerkiksi kielenvaihto

428 429 MARJATTA PALANDER, HELKA RIIONHEIMO, KIELIKORPUKSIA SUOMEN ITÄRAJALTA HANNU KEMPPANEN & JUKKA MÄKISALO enemmistökieleen, monikielisyys ja äidinkielen hiipuminen. Inkerin­ passiivimuotojen kohtaloa inkerinsuomen ja viron kielikontaktissa sii­ suomen tutkimus on tämän myötä suuntautunut kohti uudenlaisia hen, mitä suomen passiiville tapahtuu kääntämisessä, ja siihen, miten näkö­kulmia, kuten kielikontaktien ja kielen attrition tutkimusta. suomen passiivi vaikuttaa englannin kielen passiivin omaksumiseen. Joensuun yliopiston inkerinsuomen tutkimushankkeessa aineis­ Nämä julkaisut osoittavat, että inkerinsuomen aineistolla on edelleen tonkeruun tavoitteena oli hankkia yleiskuva inkerinsuomen silloisesta runsaasti annettavaa tutkimukselle ja että aineistoa voidaan lähestyä tilanteesta ja kielimuodon moninaisesta vaihtelusta, ja siksi haastatel­ uusista näkökulmista. tavina on ollut monenlaisia kielenpuhujia: sekä murteensa säilyttänei­ Inkerinsuomen digitaalisen korpuksen työstäminen on tätä kir­ tä että sellaisia yksilöitä, joilla alkuperäinen murre on muuttunut joko joitettaessa loppusuoralla. Työ aloitettiin kirjoittamalla eri aikoina suomen yleiskielen tai vieraan kielen (venäjän tai viron) vaikutukses­ tehtyjä alkuperäisiä (osin käsin kirjoitettuja) litteraatioita sähköiseen ta. Suurin osa haastateltavista edustaa tuolloista vanhinta ikäpolvea muotoon tekstinkäsittelyohjelmalla. Aineiston nauhoitteet digitoitiin eli ikäluokkaa, joka oli ehtinyt omaksua Inkerinmaalla suomen äidin­ jo 2000-luvun alussa Joensuun yliopiston suomen kielen aineisto­ kielekseen ennen toisen maailmansodan aikaista ja jälkeistä kansallis­ kokoelmien laajemman digitoinnin yhteydessä, mutta muu korpus­ ta hajaannusta. Aineisto on ainutlaatuinen dokumentti erään Suomen työ tuli mahdolliseksi keväällä 2013 Koneen Säätiön rahoittaman valtion ulkopuolella puhutun suomen murteen kohtalosta: sosio­ CROSSLING-hankkeen puitteissa. CROSSLING- ja SILK-hankkei­ poliittisista syistä aiheutuneesta kielenvaihdosta ja sen seurauksista­ den mahdollistamien tutkimusapulaisen työkuukausien aikana kaik­ puhujien äidinkieleen (inkerinsuomeen). Koska inkerinsuomalaisten­ ki olemassa olevat litteraatiot (noin 92 tuntia) kirjoitettiin tietokone­ yhteis­kunnallinen asema oli Neuvostoliitossa­ suhteellisen saman­ muotoon Word-tiedostoiksi. Litteraatioita olivat tehneet useat eri lainen sekä Inkerinmaan alueella että Virossa, aineiston kaksi osaa litteroijat eri aikoina eri tarkoituksiin, ja puhtaaksikirjoittamisen ai­ mahdollistavat vertailun kahden erilaisen kielikontaktitilanteen välil­ kana käytettyä tarkemerkistöä yksinkertaistettiin jonkin verran. Kor­ lä: Virossa valtakielenä on ollut läheinen sukukieli viro, Inkerinmaalla pustyö on viimeistelty FIN-CLARINin rahoituksella kesällä 2019. taas suomesta typologisesti paljon poikkeava venäjän kieli. Viimeisessä vaiheessa eri tarkkuusasteilla tehdyt litteraatiot on kar­ Inkeriläismurteista on ilmestynyt kaikkiaan neljä väitöskirjaa keistettu yhdenmukaisiksi­ ja muunnettu txt-tiedostoiksi. Aineistot on (Lehto 1996, Kokko 2007, Riionheimo 2007 ja Mononen 2013), joista luovutettu Kielipankkiin2, ja korpuksen viimeistely siellä aloitetaan Kokon ja Riionheimon tutkimukset perustuvat Itä-Suomen yliopiston syksyllä 2019. Inkerinsuomen korpus julkaistaan tekstikorpuksena, Inkerinsuomen korpukseen. Samasta aineistosta on lisäksi valmistunut jossa myös äänitiedostot ovat tutkijoiden­ saatavilla. Tekstin ja äänen pro gradu -tutkielmia (tuorein Surakka 2011). Helka Riionheimo on kohdistaminen on hidas ja suuritöinen urakka, johon ei inkerinsuomen julkaissut väitöskirjansa jälkeen useita inkerinsuomen aineistoa hyö­ korpuksen osalta ole ryhdytty. dyntäviä artikkeleita sekä suomenkielisissä että englannin­kielisissä julkaisuissa tai kokoomateoksissa. Uusin vaihe aineiston hyödyntämi­ sessä on Riionheimon ja Maria Frickin yhteistyö, jossa Viron inkerin­ suomalaisten aineistoa on verrattu 1990-luvun jälkeen Viroon muutta­ neiden suomalaissiirtolaisten käyttämään kieleen (Frick & Riion­heimo 2013, Riionheimo & Frick 2014). Tämä yhteistyö on osoittanut myös sen, että haastatteluformaatista huolimatta inkerinsuomen aineisto on kiinnostava tutkimuskohde myös vuorovaikutus(sosio)lingvis­ 2. Tiedot korpuksesta on jo liitetty FIN-CLARINin META-SHARE-tieto­ tiikan näkökulmasta. Lisäksi Riionheimolta on ilmestynyt Virittä- kantaan: ‹http://meta-share.csc.fi/repository/browse/the-corpus-of-ingrian-finnish/ jä-lehdessä yhteisartikkeli (Riionheimo ym. 2014), jossa verrataan 0bed3e04aacb11e390f0005056be118e57c9201eecd4428a9e86b7ac323f8ea8/›.

430 431 MARJATTA PALANDER, HELKA RIIONHEIMO, KIELIKORPUKSIA SUOMEN ITÄRAJALTA HANNU KEMPPANEN & JUKKA MÄKISALO

4. Karjalan suomen korpus sanaa Karjalan Sanomat -sanomalehden 2000-luvulla ilmestyneistä teksteistä. Kukin aineiston artikkeli on annotoitu käännöstieteellisen Suomalaisia ja suomea puhuvaa väestöä on Karjalan tasavallassa ja ja kielikontaktien tutkimuksen kannalta relevantilla tavalla, josta käy Petroskoissa asunut koko 1900-luvun ajan. Suomen itsenäistymisen ilmi artikkelin syntytapa, toisin sanoen, onko kyseessä käännetty vai jälkeen suomalaisia muutti paljolti poliittisista syistä Neuvostoliit­ alun perin suomeksi tuotettu aineisto. Analyysissa on hyödynnetty toon, ja suomalaisten määrä tasavallassa nousi tasaisesti vuoden 1926 myös venäjänkielisiä lähdetekstejä. Korpus3 on nyt Unicode-muotoon väestölaskennan 2 544:stä vuoden 1959 väestölaskennan 27 829:ään koodattuina tekstitiedostoina, ja se mahdollistaa seuraavassa vaihees­ (Vsesojuznaja perepis naselenija 1926 goda, Vsesojuznaja perepis na­ sa tekstin morfologis-syntaktisen annotoinnin. Hankkeen käyttöoike­ selenija 1959 goda). Sen jälkeen suomalaisten osuus on taas tasaisesti us on ensi vaiheessa sekä Petroskoin valtiollisella yliopistolla että Itä- vähentynyt, niin että vuoden 2010 väestölaskennan mukaan suoma­ Suomen yliopistolla. laisia on koko tasavallassa 8 577 (Vserossijskaja perepis naselenija Jukka Mäkisalo, Hannu Kemppanen ja Anna Saikonen (2016) 2010). Petroskoissa suomen kielellä ilmestyy sanomalehti Karjalan ovat esitelleet Karjalan suomen korpusta ja ensimmäisiä siitä tehtyjä Sanomat ja kulttuuriaikakauslehti Carelia, ja kaupungissa toimii mm. korpusanalyyseja käännöstieteellisessä MikaEL-julkaisussa. Käänne­ suomalainen teatteri. tyn ja ei-käännetyn kieliaineiston vertailu vähemmistökielen näkö­ Karjalan suomen korpus on koottu Venäjän Karjalassa Petros­ kulmasta on kyseenalaistanut aiempia väittämiä näille kielimuodoille koissa ilmestyvän Karjalan Sanomat -sanomalehden teksteistä kahta tyypillisistä piirteistä. tarkoitusta varten: Karjalan suomen ja käännetyn suomen variaation tutkimukseen. Perusteena kyseisen aineiston valinnalle on se, että Karjalan suomea ei ole toistaiseksi koottu elektroniseksi korpukseksi. Kiitokset Mediatekstien voidaan katsoa edustavan normia luovaa osaa kielen­ käytöstä. Karjalan suomen korpuksen edustama kielimuoto on lähellä Korpushankkeemme ovat saaneet taloudellista tukea seuraavilta ra­ suomen yleiskieltä. hoittajilta: Suomen Akatemialta (137479; Raja-Karjalan korpus), Karjalan suomen korpuksen perustana ovat digitaaliset aineistot, Koneen Säätiöltä (40-5091; Raja-Karjalan, Inkerinsuomen ja Kar­ joiden kokoaminen aloitettiin Suomen Akatemian ja Venäjän huma­ jalan suomen korpukset), Karjalaisen Kulttuurin Edistämissäätiöltä nistisen tiedesäätiön vuosina 2009–2011 rahoittamassa tutkimushank­ (Raja-Karjalan ja Inkerinsuomen korpukset) sekä FIN-CLARINilta keessa Venäjästä suomeksi ja suomesta venäjäksi: kääntäminen mo- (Inkerinsuomen korpus). Kiitämme lämpimästi hankkeille osoitetusta nikulttuurisessa yhteisössä. Hanke toteutettiin yhteistyönä Joensuun tuesta. yliopiston (nykyisen Itä-Suomen yliopiston) humanistisen osaston ja Petroskoin valtiollisen yliopiston suomen kielen ja kirjallisuuden laitoksen kanssa. Yhtenä hankkeen osa-alueena oli koota kirjoitettua Venäjän Karjalan suomea elektroniseksi korpukseksi, jonka pohjalta voisi analysoida vähemmistökielen erityispiirteitä kääntämisen ja kie­ likontaktien näkökulmasta sekä verrata aineistoa vastaavaan Suomes­ sa tuotettuun suomenkieliseen materiaaliin. Karjalan suomen tekstikorpus saatiin valmiiksi Koneen Säätiön 3. Korpuksen metatiedot ovat FIN-CLARINin META-SHARE-tieto­ kannas­sa osoitteessa ‹http://meta-share.csc.fi/repository/browse/the-karelian-­ rahoittaman SILK-hankkeen aikana vuosina 2013–2014 yhteistyössä finnish-newspaper-corpus/80fa56f0454e11e49821005056be118e6a793­e3276­ Petroskoin valtiollisen yliopiston kanssa. Korpus sisältää noin 600 000 d84c95b8d9cf6ff7d867c8/›.

432 433 MARJATTA PALANDER, HELKA RIIONHEIMO, KIELIKORPUKSIA SUOMEN ITÄRAJALTA HANNU KEMPPANEN & JUKKA MÄKISALO

Lähteet maisten kielten ja kirjallisuuksien laitos. Saatavissa: ‹http://urn.fi/ URN:ISBN:978-952-10-8657-1› Bubrih, D. V., A. A. Beljakov & A. V. Punžina 1997: Karjalan kielen murre­ Mäkisalo, Jukka, Hannu Kemppanen & Anna Saikonen 2016: Karjalan Sano- kartasto. Dialektologičeskij atlas karelskogo jazyka [mukana tverin­ mat -korpus. Petroskoin (käännös)suomen piirteitä. MikaEL. Kääntä­ karjalan murrekartat]. Toim. Leena Sarvas. Venäjän tiede­akatemian misen ja tulkkauksen tutkimuksen symposiumin verkkojulkaisu, vol. 9. Karjalan tiedekeskuksen kielen, kirjallisuuden ja historian insti­ Saatavissa: ‹http://www.sktl.fi/liitto/seminaarit/mikael-verkkojulkaisu/› tuutti & Kotimaisten kielten tutkimuskeskus. Kotimaisten kielten tut­ Palander, Marjatta & Helka Riionheimo 2018: How is Karelian recalled and kimuskeskuksen julkaisuja 97. Helsinki: Suomalais-Ugrilainen Seura. imitated by Finns with Border Karelian roots? – Marjatta Palander, Frick, Maria & Helka Riionheimo 2013: Bilingual voicing: A study of code- Helka Riionheimo & Vesa Koivisto (eds), On the Border of Language switching in the reported speech of Finnish immigrants in Estonia. – and Dialect. Studia Fennica Linguistica 21. Helsinki: Finnish Litera­ Multilingua 32:5: 565–599. ture Society. 85–122. Saatavissa: ‹http://dx.doi.org/10.21435/sflin.21› Genetz, Arvid 1870: Kertomus Suojärven pitäjäästä ja matkustuksistani Palander, Marjatta, Helka Riionheimo & Vesa Koivisto (eds) 2018: On siellä v. 1867. – Suomi II: 8. Kirjoituksia isän-maallisista aineista. Hel­ the Border of Language and Dialect. Studia Fennica Linguistica sinki: SKS. 21. Helsinki: Finnish Literature Society. Saatavissa: ‹http://dx.doi. Koivisto, Vesa 2018: Border Karelian dialects – a diffuse variety of Kare­ org/10.21435/sflin.21› lian. – Marjatta Palander, Helka Riionheimo & Vesa Koivisto (eds), On Riionheimo, Helka 2007: Muutoksen monet juuret. Oman ja vieraan risteytymi- the Border of Language and Dialect. Studia Fennica Linguistica 21. nen Viron inkerinsuomalaisten imperfektinmuodostuksessa. [Väitös­kirja.] Helsinki: Finnish Literature Society. 56–84. Saatavissa: ‹http://dx.doi. Suomalaisen Kirjallisuuden Seuran toimituksia 1107. Helsinki: SKS. org/10.21435/sflin.21› Riionheimo, Helka & Maria Frick 2014: Emergence of Finnish-Estonian bilin­ Kok, Maria 2016: Varjon kieliopillistuminen. Itse-sanan paradigman gual constructions in two contact settings. – Sociolinguistic Studies 8:3: rakenne ja merkityksenkehitys itäisessä itämerensuomessa. [Väitös­ 409–447. kirja.] Publications of the University of Eastern Finland. Dissertations Riionheimo, Helka & Krista Kivisalu (toim.) 1994: Inkeriläiskertomuksia. in Education, Humanities, and Theology 83. Joensuu: Itä-Suomen yli­ Studia Carelica Humanistica 4. Joensuu: Joensuun yliopisto. opisto. Saatavissa: ‹http://urn.fi/URN:ISBN:978-952-61-2064-5› Riionheimo, Helka, Leena Kolehmainen & Lea Meriläinen 2014: Suomen Kokko, Ossi 2007: Inkerinsuomen pirstaleisuus. Eräiden sijojen kehitys passiivi kontaktissa. Kieltenvälisiä kytköksiä migraatiossa, toisen kie­ murteen yksilöllistymisen kuvastajana. Joensuun yliopiston humanis­ len omaksumisessa ja kääntämisessä. – Virittäjä 118: 334–371. Saata­ tisia julkaisuja 48. Joensuu: Joensuun yliopisto. Saatavissa: ‹http://urn. vissa: ‹https://journal.fi/virittaja/article/view/9249› fi/URN:ISBN:978-952-219-036-9› Savijärvi, Ilkka, Muusa Savijärvi & Janne Heikkinen (toim.) 1996: Vot, Kokko, Ossi, Ilkka Savijärvi & Muusa Savijärvi (toim.) 2003: Ennev vanha- ihminen tahtoo kotimaalle. Länsi-Inkerin kieltä ja kohtaloita. Studia sii – Pohjois-Inkerin kieltä ja kohtaloita. Studia Carelica Humanistica Carelica Humanistica 8. Joensuu: Joensuun yliopisto. 18. Joensuu: Joensuun yliopisto. Surakka, Anne 2011: Yleistävän yksikön 2. persoonan käyttö inkerinsuo- Kujola, Joh. 1910: Äänneopillinen tutkimus Salmin murteesta. Eripainos messa. Pro gradu -tutkielma. Itä-Suomen yliopisto, suomen kieli. Saa­ Suomi-kirjasta. Helsinki: SKS. tavissa: ‹http://urn.fi/urn:nbn:fi:uef-20110436› Lehto, Manja Irmeli 1996: Ingrian Finnish: Dialect preservation and Turunen, Aimo 1965: Suojärven murre. – Lauri Pelkonen (toim.), Suojärvi I. change. [Väitöskirja.] Acta Universaliensis Upsaliensis. Studia Uralica Kajaani: Suo-säätiö. 21–38. Upsaliensia 23. Uppsala: Uppsala University. — 1973: Raja-Karjalan murteet ja vepsän kieli. – Hannes Sihvo (toim.), Mononen, Kaarina 2013: Inkerinsuomalaisten suomen kielen käyttö Kalevalaseuran vuosikirja 53. Helsinki: SKS. 83–94. Pietarissa ja sen lähialueella. [Väitöskirja.] Helsinki: Helsin­ — 1982: Raja-Karjalan murteet. – Karjala 2. Karjalan maisema ja luonto. gin yliopiston suomen kielen, suomalais-ugrilaisten ja pohjois­ Hämeenlinna: Karisto. 65–89.

434 435 MARJATTA PALANDER, HELKA RIIONHEIMO, KIELIKORPUKSIA SUOMEN ITÄRAJALTA HANNU KEMPPANEN & JUKKA MÄKISALO

Uusitupa, Milla 2017: Rajakarjalaismurteiden avoimet persoonaviittaukset. Language corpora from the eastern border of Finland [ Väitösk i r ja.] P ubl icat ion s of t he Un ive r sit y of Ea st e r n Fi n la nd. Disse r t a­ tions in Education, Humanities, and Theology 117. Joensuu: Itä-Suomen Marjatta Palander, Helka Riionheimo, Hannu Kemppanen & yliopisto. Saatavissa: ‹http://urn.fi/URN:ISBN:978-952-61-2646-3› Jukka Mäkisalo Uusitupa, Milla, Vesa Koivisto & Marjatta Palander 2017: Raja-Karjalan murteet ja raja-alueiden kielimuotojen nimitykset. – Virittäjä 121: This report presents three language corpora on language varieties used 67–106. Saatavissa: ‹https://journal.fi/virittaja/article/view/53121› on both sides of the eastern border of Finland: the Corpus of Bor­ Vserossijskaja perepis naselenija 2010. Natsionalnyi sostav naselenija po subjektam Rossijskoi federatsii [Koko Venäjän kattava väestölaskenta der Karelia, the Corpus of Ingrian Finnish, and the Karelian Finnish 2010. Osa 1. Väestön lukumäärä ja jakauma. Taulukko 7 [MS Excel- Newspaper Corpus. These digital corpora have been (and are in the taulukko]]. Federalnaja služba gosudarstvennoi statistiki [Venäjän process of being) compiled on the subjects of the Finnish language, federaation tilastovirasto], 2012. Moskova: ИИЦ ”Статистика Рос­ , and Karelian language and culture at the Univer­ сии”. [Viitattu 18.12.2015] Saatavissa: ‹http://www.gks.ru/free_doc/ sity of Eastern Finland. The varieties in question have been small mi­ new_site/population/demo/per-itog/tab7.xls› nority languages in their respective areas, and thus, the corpora offer Vsesojuznaja perepis naselenija 1926 goda. Natsionalnyi sostav naselenija perspectives into the fates of minority languages and cross-linguistic po regionam RSSR [Koko Neuvosto-Venäjän kattava väestölaskenta influence from the dominant languages. 1926. Väestön kansallisuus ja jakautuminen maaseutu- tai kaupunki­ The Corpus of Border Karelia consists of 120 hours of tran­ asukkaisiin. Karjalan ASSR]. Демоскоп Weekly. [Viitattu 18.12.2015] scribed dialect samples from Karelian speakers who were born in the Saatavissa: ‹http://demoscope.ru/weekly/ssp/rus_nac_26.php?reg=53› Border Karelia parishes of Ilomantsi, Korpiselkä, Suistamo, Suo­ — 1959 goda. Natsionalnyi sostav naselenija po regionam Rossii [Koko järvi, Impilahti, and Salmi. The recordings were conducted mainly Neuvostoliiton kattava väestölaskenta 1959. Väestön kansallisuus Venäjän alueilla. Karjalan ASSR]. Демоскоп Weekly. [Viitattu in the 1960s in Finland in the places where the inhabitants of Border 18.12.2015] Saatavissa: ‹http://demoscope.ru/weekly/ssp/rus_nac_59. Karelia were resettled after World War II (i.e., when Finland ceded php?reg=81› the Border Karelia region to the Soviet Union). The samples were transcribed in 2009–2014 and the Unicode texts and sound files are currently being aligned. The interviews that form the basis of the Corpus of Ingrian Finn­ ish were recorded in the 1990s in two locations: in Estonia and in Rus­ sia (in the area named Ingria, surrounding St. Petersburg). The corpus consists of 125 hours of recordings from Ingria (the parishes of Tok­ sova, Keltto, Skuoritsa, Kupanitsa, and Kurkolanniemi) and about 60 hours of recordings made in Estonia (in the towns of Tartu and Pärnu and the Järvamaa district). Parts of this data have been transcribed dur­ ing the last two decades by several transcribers. At present, we are in the process of transforming these miscellaneous (partly hand-written) texts into digital form in order to build them into a text corpus in the Unicode format. In the future, it will be possible to align the text and sound files as well.

436 437 MARJATTA PALANDER, HELKA RIIONHEIMO, HANNU KEMPPANEN & JUKKA MÄKISALO

The Karelian Finnish Newspaper Corpus is a text corpus which ULDIS BALODIS comprises written texts that have been published in the Karjalan Sano- University of Latvia Livonian Institute mat newspaper in Petrozavodsk (in the Karelian Republic of the Rus­ sian Federation) in the 2000s. The corpus was compiled in 2009–2014 in cooperation with the State University of Petrozavodsk. It contains about 600 000 words, and all the texts have been annotated so that we Expeditions among the Lutsi know whether the text was translated from Russian or originally writ­ Estonians and the design of ten in Finnish. Work on this corpus has now been completed and it has been handed over to the FinClarin database. Language Learning Materials Ultimately, the Corpus of Border Karelia and the Corpus of Ingrian Finnish will also be included in FinClarin. The metadata Abstract The Lutsi Estonians (Lutsis) are a historically South of these corpora have already been included in the META-SHARE Estonian-speaking minority that has inhabited a network of more database. than 50 villages in the historical rural parishes of Pilda, Nirza, Brigi, and Mērdzene surrounding the city of Ludza, in the Lat­ gale region of eastern Latvia, for at least three to four centuries. Between 2013 and 2016, I received funding from the Kone Foun­ dation to document the present state of the Lutsis and to write a Lutsi language primer for Latvian speakers. The first part of this paper gives an overview of previous work on Lutsi followed by a description of the present state of the population of Lutsi descend­ ants as I found it during the period of my Kone-funded research. The second part of this paper describes the Lutsi language primer, beginner’s grammar reference, and dictionary, which were the other main products of this research and discusses plans for their future use. The first extensive documentation of the Lutsis, their culture, and language was undertaken by researcher Oskar Kallas in 1893. Relatively substantial subsequent documentation of Lut­ si was carried out over subsequent decades as the Lutsi population continued to be assimilated primarily into the Latvian speakers of their home region. The last fluent speaker died in 2006 and the last known passive partial speaker died in 2014. Presently, some fragmentary knowledge of Lutsi survives among descend­ ants; however, the Lutsi Estonians today have shifted entirely to using Latvian, and less frequently also Russian, as their primary language.

Multilingual­ Finnic. Language 438 contact and change. 439–478. Uralica Helsingiensia 14. Helsinki 2019. ‹https://doi.org/10.33341/uh.85046› ULDIS BALODIS EXPEDITIONS AMONG THE LUTSI ESTONIANS AND THE DESIGN OF LANGUAGE LEARNING MATERIALS 1. Introduction 2. On names This paper describes some of the findings of my Kone-funded research The terms “Lutsi Estonians” and “Lutsis” are used interchangeably in during the period between 2013 and 2016 when I was funded to docu­ this report. “Estonian” is also used to refer to this community unless ment the current state of the Lutsi Estonians, one of the three South Esto­ otherwise indicated to be related to Estonia specifically. The reason for nian language island communities and one of two that are located within this is that the Lutsis themselves identify the ethnicity of their Lutsi- present-day Latvia. The Lutsis were first documented extensively in speaking relations – or in some cases their own ethnicity – as Esto­ 1893 by researcher Oskar Kallas1. During the course of the 20th century, nian. While perhaps one could ponder whether successful language the Lutsis and their language would be subject to relatively extensive revitalization would bring about the emergence of a unique Lutsi iden­ though infrequent documentation with many years and even decades tity, such an identity has never really existed in the past. Therefore, I separating major documentation efforts (see Section 5). In addition, this am cautious about claiming one here for the Lutsis. paper describes the other portion of my Kone-funded work, which is the One may wonder then why I do not simply always refer to the design and writing of a Lutsi language primer intended to be used for Lutsis consistently as Estonians throughout this paper. This is because, reacquainting Lutsi descendants and other residents of Latvia with the from the perspective of an exterior observer, the Lutsis clearly do form Lutsi language, culture, and people as well as the beginner’s grammar a unique community, which historically has spoken a unique variety of reference and dictionary, which were also products of this work. An ad­ South Estonian and – especially presently – whose members, unlike ditional product of my Kone-funded work not discussed in this paper, Estonians in Estonia, see themselves as people connected with Latgale is an informative website about the Lutsis, the Lutsi language, and my and with Latvia as a whole rather than with Estonia. research (located at: www.lutsimaa.lv), which is available in Latvian, In conversation with various Lutsi descendants, I have not fully English, and Estonian. This paper gives new facts, which I unearthed gauged the nature of their own view on their identity. In one very par­ during the course of my work and also tells some of the stories I en­ ticular case, a Lutsi woman I interviewed did tell me that she consid­ countered during this same work, which are relevant to characterizing ered herself to be Estonian rather than Latvian. However, during the the present situation of the community of Lutsi descendants. course of this work I did not find other Lutsi descendants making such Section 2 of this paper discusses naming conventions used for the firm statements and self-identifying as Estonians. Lutsis in English, Latvian, and Lutsi; section 3 discusses the history of In Latvian, the Lutsis are referred to as Ludzas igauņi ‘Ludza changing administrative boundaries in the Lutsi home region and how Estonians’, igauņi ‘Estonians’, or occasionally as luci ‘Lutsis’. This this relates to describing the location of different Lutsi communities; final term luci or derived forms such as luciski ‘in Lutsi’ are some­ section 4 describes the fieldwork in the historical Lutsi region in east­ what rarely used, but do appear to a limited extent in some recent and ern Latvia, which I undertook for this research; section 5 describes the forthcoming works such as the collection of translated Lutsi stories linguistic background of the Lutsis and the history of Lutsi documen­ Ludzas igauņu pasakas (‘Lutsi stories’; Godiņš 2015) and the volume tation; section 6 describes the present state of the Lutsi language; sec­ on Lutsi history and origins Ludzas igauņi: Zemes dieva tauta (‘The tion 7 discusses the possibility of Estonian or Lutsi habitation in other Lutsis: People of the Earth God’; Korjus 2017). villages beyond those visited by Kallas; section 8 discusses my Lutsi In Lutsi, the Lutsis have referred to themselves as ēstläzeq ‘Es­ primer, its layout, and the orthography I use for it; section 9 provides tonians’ or mārahvas ‘country folk’. Similarly, they have called their a short overall summary of the subjects in this paper. language ēstu kīļ ‘Estonian language’ or mākīļ ‘country language’. Presently, as mentioned above, the Latvian-speaking Lutsi descend­ ants refer (in Latvian) to their ancestors as igauņi ‘Estonians’ and to 1. Kallas published a report of his work as the book Lutsi maarahvas in 1894. their ancestors’ language as igauņu valoda ‘Estonian language’. 440 441 ULDIS BALODIS EXPEDITIONS AMONG THE LUTSI ESTONIANS AND THE DESIGN OF LANGUAGE LEARNING MATERIALS

3. Changing administrative boundaries both maps. Map 2 also shows most of the villages visited by Kallas as well as the location of other villages (e.g., Špegi, Zurzi) mentioned The change in administrative divisions in Latvia from the pre-WWII elsewhere in this report. years to the present creates a potentially confusing situation in dis­ cussing where the Lutsi villages are located. Prior to WWII and the Soviet occupation, Latvia had two types of administrative divisions. 4. Fieldwork The top-level division apriņķis ‘district’ was composed of a lower- level division pagasts ‘rural parish’. During the Soviet occupation, the I conducted six trips to Ludza and the surrounding rural areas during administrative divisions of Latvia changed many times. The apriņķis the period of my Kone funding. Additionally, I traveled to the Ludza and pagasts divisions were eliminated at one point and ultimately re­ area one time prior to this funding in August 2012 when I first met the placed with a top-level division called rajons ‘district’. The duties of final person with passive knowledge of Lutsi, Antonīna Nikonova, at the lower-level administrative division were in some sense shared be­ her home in Pilda, Latvia. tween collective farms and the ciema padome ‘village soviet’ (also The purpose of my fieldwork was to meet with and interview Lutsi referred to just as ciems ‘village’) (Šķiņķis 1999: 69–72). descendants and document their knowledge of Lutsi language and cul­ The boundaries of the administrative divisions of the Soviet pe­ ture as well as their own family history, in order to better understand­ riod did not at all match those of the apriņķis or pagasts boundaries of the interrelationship between various Lutsi families. Through my ex­ the pre-WWII years. In 1990, with the end of the Soviet occupation, isting contacts, most notably Professor Karl Pajusalu of the University the rajons division remained, but the ciema padome/ciems division of Tartu and Hannes Korjus, a researcher of Latvian Estonians, I first was renamed pagasts without any accompanying boundary changes began by meeting and interviewing the Lutsi descendants of whom (Šķiņķis 1999: 74). This meant that there were pre-WWII and post- they were aware. In addition, my first research trip was conducted occupation pagasts divisions with the same name, but totally different in conjunction with Indrek Jääts and Maido Selgmäe of the Estonian boundaries (e.g., Pilda pagasts, Nirza pagasts, Mērdzene pagasts, Bri­ National Museum (Eesti Rahva Muuseum). Jääts and Selgmäe were gi/Briģi pagasts). In speaking about the location of the Lutsi villages, in the process of making a documentary film about the Lutsis (Kadu- this creates a real problem, because some of these villages are found in nud hõim: Lugu Lutsi maarahvast. Lost Tribe: A Story about the Lutsi pagasts divisions of different names in pre-WWII Latvia as opposed Estonians, released in 2014) and so through them I also met with and to present-day Latvia. interviewed a number of Lutsi descendants. More recently Latvia underwent a further administrative reform, After this initial trip, my fieldwork would be composed of meet­ culminating in 2009 with the elimination of both the rajons and pa- ings with individual descendants in Ludza, other particular villages, or gasts division. These were replaced with a single type of adminis­ other cities in the greater Ludza area. During the course of my field re­ trative unit novads ‘rural municipality’. The pagasts division still re­ search, I traveled to as many of the 53 villages noted by Oskar Kallas mains relevant, however, as it is used for the purposes of describing as having Estonian-speaking inhabitants as I could. In the end I visited locations and still appears on published regional maps. To avoid con­ 40 of these villages or sites where these villages had once stood, and fusion, in this paper all locations are given with respect to the location in subsequent years I have visited the remainder of these villages and of the villages within their pre-WWII parishes. Maps 1 and 2 show a village sites. I spoke with village inhabitants and in this way also was comparison between administrative divisions in the Ludza area before able to determine the level at which knowledge of Lutsi origins still WWII (left) and currently (right). Cities of regional importance and persists in the actual villages where the Lutsis had lived in Kallas’ other communities historically important to the Lutsis are shown on time.

442 443 ULDIS BALODIS EXPEDITIONS AMONG THE LUTSI ESTONIANS AND THE DESIGN OF LANGUAGE LEARNING MATERIALS

Maps 1 (left) & 2 (above). Comparison between pre-WWII (left) and current adminis- trative divisions (above) in the Lutsi region. (Top-level administrative division names shown in bold italics, lower-level pagasts division names shown in plain italics, towns significant to the Lutsis are shown on the map with/without names. Smaller dots on Map 2 mark Lutsi villages.) Maps by Uldis Balodis (utilizing information from Turlajs 2012: 46–47, 74–75 and RAPLM/PRD 2007).

444 445 ULDIS BALODIS EXPEDITIONS AMONG THE LUTSI ESTONIANS AND THE DESIGN OF LANGUAGE LEARNING MATERIALS 5. Background

The Lutsis are one of three groups of South Estonian speakers re­ ferred to collectively as the “South Estonian language islands”. The Lutsis were located in villages spanning the pre-WWII rural parishes of Mērdzene (formerly, Mihalova), Pilda, Nirza, and Brigi (formerly, Janovole). The Leivus were a language island located in northeastern Latvia centered on the towns of Ilzene and Lejasciems near the city of Alūksne. A third language island, Kraasna, was located near the Lutsis but across the present-day border in Russia in a group of villages near the city of Krasnogorodsk. None of these South Estonian language varieties is spoken any longer (Kallas 1894, 1903; Niilus 1937; Korjus 2001; Pajusalu 2009). Map 3 shows the approximate location of the three language islands relative to present-day boundaries. Documentation of the Lutsis occurred during four primary peri­ ods and one other more general period which followed. These include (1) Oskar Kallas’ trip through the Lutsi region in 1893, (2) Heikki Ojansuu’s written documentation in 1911 of the speech of inhabitants of Pilda rural parish and to a considerably lesser extent Nirza rural parish, (3) the written documentation of Lutsi during Latvia’s interwar independence most notably by Paulopriit Voolaine and August Sang, and (4) the documentation of Lutsi by linguists from Estonia (Salme Nigol, Salme Tanning, Elna Adler, Aino Valmet) sometimes in con­ junction with Voolaine, which occurred primarily in the 1960s and 1970s. The documentation from the 1960s and 1970s is also the main Map 3. The South Estonian language islands and dialect areas (Iva & Pajusalu 2004). source of audio recordings of Lutsi. The fifth more general period of Lutsi documentation followed this documentation in the 1960s. Voo­ laine continued to travel to the Lutsi region and meeting with Lutsi due to the lack of speakers. The very first documentation of the Lutsi families he knew until his death in 1985. Other researchers includ­ community, carried out by Oskar Kallas in 1893, gives us the only in­ ing Hannes Korjus and Karl Pajusalu have continued to document the formation we have about the Lutsis of Mērdzene rural parish. Already present state of the Lutsi community primarily in the period follow­ at the time that Kallas visited the villages of Mērdzene, the Lutsis ing 2000. Most recently, my Kone-funded research has continued this there were linguistically quite assimilated. In his documentation of tradition and has documented the current state of the Lutsi community Lutsi throughout the Lutsi-speaking region, Kallas wrote down some as well as the last remnants of language knowledge which still exist short word lists along with a fairly extensive collection of folk songs, presently among descendants of this community. riddles, and other examples of folklore. Those from Mērdzene are the Each period of research documented knowledge which later pe­ only information available on the Lutsi spoken by the inhabitants of riods of research either did not focus on or were unable to focus on this region.

446 447 ULDIS BALODIS EXPEDITIONS AMONG THE LUTSI ESTONIANS AND THE DESIGN OF LANGUAGE LEARNING MATERIALS

Likewise, Ojansuu’s fieldwork in 1911 gives us examples of Lut­ si spoken in villages in the Pilda rural parish for which later language documentation does not exist. Ojansuu wrote down word lists, stories, songs, and what appear to be transcriptions of other utterances spoken to him by his consultants. This information gives us the only record of speech from such villages as Raibakozi (Kirivä-kidze külä, in Lutsi) and Belamoiki (Belomoikino) in Pilda rural parish, for which no later records exist. Documentation following Ojansuu focused on the villages of Lielie Tjapši (Sūre Tsäpsiq or Jāni külä, in Lutsi), Mazie Tjapši (Väiku Tsäpsiq), and Škirpāni (Kirbu külä) in Pilda rural parish and the vil­ lages of Greči (Grēki), Ščastļivi (Tati külä), and Barisi (Mägize külä) in Nirza rural parish. The documentation from the third and fourth pe­ riods identified above would be the richest in terms of linguistic mate­ rial. A large part of this documentation is composed of fairy tales or narratives focusing on experiences from the consultants’ own lives. It provides an extensive record of Lutsi connected speech and, as stated above, work later in this period provides the only audio documentation of Lutsi. In addition, Voolaine and Sang wrote several unpublished pi­ lot studies of subjects focusing on aspects of Lutsi phonology and mor­ phology, which are freely available at the University of Tartu Archives of Estonian Dialects and Kindred Languages (Tartu Ülikooli eesti mur- rete ja sugulaskeelte arhiiv) at ‹http://www.murre.ut.ee/arhiiv/›2. Speakers of Lutsi have steadily decreased in number since their Figure 1. Antonīna Nikonova’s and Nikolājs Nikonovs’ house in Lielie Tjapši (left), first estimation by Oskar Kallas. Kallas estimated a population of ap­ likely the last place where Lutsi was spoken as a language of everyday communi- cation. Photo: Uldis Balodis. proximately 800 Lutsi speakers in 1893. Early in the twentieth cen­ tury, Heikki Ojansuu (1912) and Villem Grünthal (1912) gave an estimate of about 200 Lutsi speakers. Paulopriit Voolaine (1925) es­ most a short phrase in Lutsi. At this time, there appear to be no remain­ timated approximately 120 Lutsi speakers and August Sang (1936) ing people with Lutsi speaking or comprehension abilities. estimated approximately 30–40 Lutsi speakers. The last mostly flu­ The origin of the Lutsis has long been a matter for discussion ent speaker of Lutsi, Nikolājs Nikonovs, died in 2006. His widow, without clear agreement. The similarity of Lutsi to modern-day Seto Antonīna Nikonova, who died in 2014, could not hold a conversation argues against the Lutsis being a relic ancient population that would in Lutsi, but did appear to have a great deal of passive knowledge and have lived in the Ludza area for more than a few centuries (Karl Paju­ was able to remember short phrases and individual words. Currently, salu, p.c.). Two commonly stated theories regarding Lutsi origins are Lutsi descendants remember a few individual words, a greeting, or at that the Lutsis either are (1) the descendants of South Estonian-speak­ 2. This archive is described elsewhere in this volume (see Lindström, Lippus & ing Catholic peasants who fled southern Estonia in the seventeenth Tuisk 2019). century during the period of Swedish rule to avoid forced conversion

448 449 ULDIS BALODIS EXPEDITIONS AMONG THE LUTSI ESTONIANS AND THE DESIGN OF LANGUAGE LEARNING MATERIALS to or (2) refugees fleeing southern Estonia in the eight­ eenth century as a result of the . Lutsi last names are not necessarily of Estonian derivation. Some Lutsi last names, such as Mekšs and Germs/Germovs, may indeed de­ rive from an Estonian source – in this case Mäks and Härm, respec­ tively. Other last names, such as Jarošenko, Nikonovs, and Buls/Buļs/ Bulis3 do not seem to have an Estonian source.

6. Current state of the Lutsi language

As stated previously, the last known fluent speaker of Lutsi died in 2006 with the last knowledgeable passive speaker (the last speaker’s widow) dying in 2014. Lutsi descendants do not appear to be concentrated in any particular area within the Ludza region. Many Lutsi descendants still live in Ludza, other larger towns in this region, or also in some of the villages documented by Kallas as having Estonian-speaking in­ Figure 2. Antonīna Nikonova and Nikolājs Nikonovs. Photo: Nikonovs family photo habitants (a list of these villages appears in Kallas 1894: 13–15). Just album. as other residents of Latvia, Lutsi descendants are to be found in the Latvian capital city of Rīga, other larger cities across Latvia, or abroad. Among Lutsi descendants in their middle years, language loss Documentation of Lutsi before WWII suggests that active cent­ occurred for the most part 2–3 generations ago. Thus, for these people ers of use for Lutsi were the villages of Lielie Tjapši and Škirpāni their grandparents or even great-grandparents were the last generation in Pilda rural parish, the cluster of villages Greči, Ščastļivi, Barisi in who they remember speaking Lutsi. The main exception is the village Nirza rural parish, and Abricki (Dunduri, in Lutsi) in Brigi rural par­ of Lielie Tjapši where active use of Lutsi as an everyday language ish. However, as I state below, it is likely that several other villages, continued in the Nikonovs family until the 1980s4. The Nikonovs fam­ such as Vorkaļi (Vārkali) in Pilda rural parish, also still saw some day- ily were, to my knowledge, the last family to maintain active use of to-day use of Lutsi in individual families, but for one reason or another Lutsi and I discuss this in further detail later in this section. the speech of these speakers was not documented during this time. During the course of my Kone-funded research, Lutsi passed 3. This last name has several variants. from having a proficient passive partial speaker to only having a few 4. I also recorded other accounts of language use during this same period. One of individuals with fairly sparse knowledge. The last known person the Lutsi descendants I interviewed, originally from Lielie Tjapši but now living in a different village (Horoševa), recalled in the 1970s when he worked as a truck driver within the group of Lutsi descendants to have had an actual degree once having to drive a delivery to Tartu. This man does not speak Estonian and so of proficiency in Lutsi was Antonīna Nikonova née Strumpe (1949– decided that he would bring an old Estonian-speaking (i.e., Lutsi-speaking) man 2014). Mrs. Nikonova was the widow of Nikolājs Nikonovs (1944– with him, so that he could communicate with local people in Estonian. However, 2006) who was the last person reasonably fluent in the Lutsi language. when they arrived in Tartu, they quickly found that the Estonian (i.e., Lutsi) that this old man spoke was so different from the Estonian spoken by people in Tartu, that no Both Mr. and Mrs. Nikonovs were from Lielie Tjapši village in Pilda one could understand him. rural parish. They both were relatives of some of the more famous

450 451 ULDIS BALODIS EXPEDITIONS AMONG THE LUTSI ESTONIANS AND THE DESIGN OF LANGUAGE LEARNING MATERIALS

in interacting with more proficient speakers of Estonian and South Estonian, she would understand individual phrases said to her. She also would volunteer individual words, such as the names of animals or short phrases, such as “I love you” told as part of a longer story in Latvian or Russian. I do not know of and have not found any other individuals who have the level of proficiency in Lutsi that Mrs. Nikonova still had, therefore with her passing it seems logical to conclude that there is most likely no one remaining who can understand Lutsi at the level that she could. In the course of my field work trips revisiting the villages that Kallas noted as having Estonian-speaking inhabitants in 1893, I found that in some places there were still individuals who knew a few words of Lutsi. In Ļukati (Lukodi, in Lutsi) village in Nirza rural parish there were two women, one who was older middle-aged, one who was elderly5, who both still remembered a greeting in Lutsi (i.e., tereq, tereq). In Ludza I met a very elderly woman in her 90s whose family Figure 3. Antoņina Nikonova (Nikolājs Nikonovs’ grandmother) and Paulopriit Voo­ was from Abricki village in Brigi rural parish. She, however, grew up laine. Photo: Nikonovs family photo album. in a non-Lutsi-speaking village, Voiti. She remembered a few Lutsi numerals, though after the numeral ‘four’ she mixed them with Ger­ man. In Boldači (Paldatsi, in Lutsi) village in Pilda rural parish, I met Lutsi speakers that linguists had worked with in the 20th century. Mr. a man in later middle years whose family was originally from Mazie Nikonovs’ grandmother was Antoņina Nikonova and his great-grand­ Tjapši village and who also remembered the tereq, tereq greeting. mother was Tekla Jarošenko. Mrs. Nikonova (Mr. Nikonovs’ widow) In my first year of Kone-funded research, I met an older Lutsi was more distantly related to Tekla Jarošenko. descendant, Helēna Kravale who since has passed away. Ms. Kravale’s From stories told by Antonīna of her husband’s grandmoth­ Lutsi relations come from her mother’s side of the family. Her mater­ er Antoņina, one gets the impression that Antoņina Nikonova née nal grandparents’ last name was Mekšs, which is one of the last names Jarošenko (1898–1983) was a driving force behind the continued associated with the Lutsis. Her maternal grandfather had been a Lutsi use of Lutsi in Lielie Tjapši. She would speak Lutsi to her grand­ speaker and would speak Lutsi with his sister. She recalled that her son Nikolājs and she would encourage other Lutsi residents of Lielie maternal grandmother, who was Latvian (i.e., Latgalian), had actually Tjapši to speak Lutsi. Another story I was told in the course of one learned Lutsi, but that her maternal grandparents had spoken Latvian of my interviews was that Antoņina would sometimes greet strangers (i.e., Latgalian) to their children. Ms. Kravale is pictured in Figure 4 who she would encounter in her home village by speaking Estonian with her niece and fellow Lutsi descendant, Līga Kondrāte, of Ludza. (i.e., Lutsi) to them. Her death in 1983, also likely marked the end of Ms. Kravale was born in Potorova village though the roots of active use of Lutsi. her Lutsi family are in Vorkaļi village in Pilda rural parish. She knew I first met Antonīna Nikonova in the late summer of 2012 and 5. The elderly woman told me that her family came from Barisi, which, as stated after the beginning of my Kone grant in April of 2013, met with elsewhere in this paper, was one of the villages characterized by active use of Lutsi her repeatedly over the course of the following year. I observed that before WWII.

452 453 ULDIS BALODIS EXPEDITIONS AMONG THE LUTSI ESTONIANS AND THE DESIGN OF LANGUAGE LEARNING MATERIALS

course of my work who possess this kind of memory of individual words or perhaps even short phrases. However, it seems unlikely that there are any unknown speakers of Lutsi remaining just based on the complete absence of anyone with even passive proficiency other than Mrs. Nikonova who passed away in 2014. One other interesting observation to make is that Lutsi knowl­ edge likely persisted in villages beyond those where Lutsi was docu­ mented in the 20th century. The fact that individuals with whom I spoke had memories of grandparents with Lutsi knowledge in other villages (Sokāni, Vorkaļi) and that I found people with some very basic knowl­ edge in yet other villages (Ļukati) suggests that prior to WWII there may well have been somewhat proficient speakers or at least people with some passive Lutsi knowledge living in these villages. We have very few data points with which to confirm this6. So we cannot know exactly how the retreat of Lutsi knowledge specifically proceeded in the Pilda and Nirza rural parishes. Figure 4. Lutsi descendants Līga Kondrāte (left) and Helēna Kravale (right). Photo: Mērdzene Estonian was already fairly far gone by the time Kallas Uldis Balodis. documented it in 1893. In visiting this region and its villages, I found little remembrance of Estonian heritage aside from a few specific in­ dividuals. One such individual is Ilmārs Silkāns (2011) who recalls in a single word of Lutsi, but the memory of this one word demonstrates an article that he himself wrote that in his youth a local minister had the way in which some of the still existing fragmentary knowledge of talked about Estonian heritage in this area. In visiting the villages of Lutsi has survived among Lutsi descendants. The word Ms. Kravale Mērdzene, I found signs of Lutsi roots mainly in the presence of the remembered was suzi ‘wolf’. She knew it because she remembered last names such as Buls/Buļs/Bulis noted by Kallas (1894: 33) as being her mother telling a story of once walking home as a child and being among those characteristic of the Lutsis or Lipskis, which is similar to followed by what seemed like a large dog. When she returned home the last name Lipskina (masculine form: Lipskins) of one of Kallas’ she had told her parents about this dog and she remembered them consultants in the Mērdzene rural parish (Kallas 1894: 75). However, speaking about it with each other excitedly in Estonian (i.e., Lutsi). there was no sign of any knowledge of Lutsi in the Mērdzene vil­ Her father’s use of the word suzi had stood out to her (as apparently lages – not even on the level seen among some Lutsi descendants in the “dog” was in fact a wolf) and so for this reason this one word of Pilda and Nirza rural municipalities. Lutsi remained known in her family at least up to family members in their early middle years (40s) today. The person with the most knowledge today is most likely a wom­ an I met in Barisi village in Nirza rural parish. She could still count to ten in Lutsi and remembered a sentence in Lutsi that was the begin­ 6. It should be noted that Ojansuu records Lutsi language data from Vorkaļi village as part of his 1911 documentation of the Lutsis and their language. So, at the very ning of a fairy tale that her grandfather would tell. It seems entirely least it can be confirmed that Lutsi was still spoken there in the early twentieth possible that there are more people beyond those that I found in the century.

454 455 ULDIS BALODIS EXPEDITIONS AMONG THE LUTSI ESTONIANS AND THE DESIGN OF LANGUAGE LEARNING MATERIALS 7. Other Estonians – possible other Lutsis eating lunch one day with the people working at the Ludza Handicrafts Center (Ludzas Amatnieku Centrs), one of the women there told me In the course of this research, I have encountered intriguing signs that that her husband’s family had Estonian ancestry and it happened to be there may have been other Estonian speakers in the Ludza region beyond that her husband’s family name was Ulass. Speaking further with one those living in the villages visited by Kallas. The uncertainty arises from of this woman’s husband’s relatives, it turned out that indeed she re­ not knowing the source of these people or anything about the language membered that Estonian had been spoken by her Ulass grandparents. they spoke. Did they migrate at the same time as the rest of the known Just as in the accounts of a number of Lutsi families from known Lutsi population? Are they more recent arrivals from Estonia who Lutsi villages, these Ulass great-grandparents had used their Estonian quickly assimilated? Or are they the final remnants of even older migra­ as a “secret” language using it as a means to have private conversa­ tions of Estonians predating the more well-known Lutsi community? tions in front of their children. Their home village had been Zeltiņi, which is directly next to Cibla. Interestingly, or perhaps not surpris­ 7.1. Cibla ingly, Zeltiņi is also the village that my own Ulass relative from Cibla I found one such possible group of Estonians following a lead from a identified as her family’s ancestral village. relative of mine, Irēna Okuņeva7, who lived in the village of Feliciano­ Thus, it appeared from these conversations, that there is evidence va. This has a personal connection for me, as part of my own family, that there was at least one family of Estonian speakers living in Zeltiņi through my paternal grandfather, comes from this region and his family in the 20th century. Yet, it is important to state that as interesting as this name, Soikans, has been given in some lists (e.g., Latkovskis 1968: 94), newfound information is, there is much that is unknown about this situ­ as a name of Finno-Ugric origin. However, this has always been mys­ ation. Given the close proximity of Cibla to the other known Lutsi areas terious to me, as in my own family no memory of Estonian origins ex­ and the interaction of Ulass people with known Lutsi families, it seems ists. Additionally, I have not been able to find evidence for an Estonian possible that the Estonians of Zeltiņi were part of the same population origin for “Soikans” beyond the possible Estonian etymologies given as the Lutsis – just a portion unknown and not noticed by previous re­ by Latkovskis in his monograph8. Therefore, one of the directions I pur­ searchers. In speaking with this woman, she told me that her family also sued in these years was attempting to determine whether my own rela­ had links with Zurzi village in Pilda rural parish. While, Zurzi also has tives from the Soikans family are indeed of Estonian descent. My rela­ not been listed among the villages with Lutsi habitation, it is extremely tive in Felicianova is from a parallel family line (last name: Ulass). Our close to other villages, such as Lielie Tjapši, with known Lutsi habita­ last common ancestor was in my great-great-grandparents’ generation. tion. Thus, it is also possible that her family did not have an extended Mrs. Okuņeva herself had no knowledge of Estonian ancestry history in Zeltiņi village in Cibla, but instead were mainly from Zurzi. in our shared family, but over time I noticed that the Ulass last name What is at the same time puzzling and fascinating is that if indeed continued to appear in Lutsi contexts. (The exact same thing can be this woman’s memory is correct and her grandparents were Estonian said for my own family’s last name – Soikans.) For example, I found or South Estonian speakers, then in pre-WWII Cibla there were still at that a member of the Ulass family was married to a known Lutsi fam­ least a couple of speakers of this language completely unknown and un­ ily from Greči village in the Nirza rural parish. By coincidence, while noticed by researchers. Of course, one problem with this account is that we have no other corroborative evidence and only one person’s account. 7. Mrs. Okuņeva was alive throughout the course of my Kone-funded research, but Another problem is that the other side of this woman’s family had the passed away in early 2017. last name Matulis/Mutulis, which is known to have Lutsi connections in 8. Latkovskis (1968: 94) gives soik ‘silence, peace’ and soikuma ‘to become calm’ as Brigi rural parish nearby. Thus, it could be that she was confusing which possible Estonian etymologies, but then also sojka ‘jay’ as a possible Russian etymology for Soikans. No further information is provided as to how these etymologies were ar­ side of her family had Estonian (i.e., Lutsi) roots. Still, these possible rived at or what background, aside from surface similarity, links them to “Soikans”. connections are intriguing and worthy of further investigation. 456 457 ULDIS BALODIS EXPEDITIONS AMONG THE LUTSI ESTONIANS AND THE DESIGN OF LANGUAGE LEARNING MATERIALS

The story of my own family’s connection is similar but even 8. Lutsi Primer and Dictionary more tenuous. With the evidence of Estonian origin for the Ulass fam­ ily and knowing that the Ulass and Soikans families stem from the The design and writing of the Lutsi primer was the other major portion same line (and at least in the case of the Soikans family, only one fam­ of my Kone-funded work during these last years. As discussed above, ily is known to have this last name), it suggests that the Soikans family Lutsi no longer has an active speech community; however, there is too could have some type of Estonian origins if the Estonian origins of still a relatively sizable group of descendants aware of their “Esto­ the Ulass family are further corroborated. nian” roots9. Ultimately, I settled on creating a primer, which would teach readers some basics of Lutsi grammar, while also including sec­ 7.2. Špegi and Franapole tions describing the Lutsis, their history, the researchers who have studied them, historical and modern photographs of particular villages In the Cibla area, but further to the west towards Ludza, there was as well as samples of Lutsi from these villages given in the original another possible connection to Estonian speakers. An individual with and translated into Latvian. A primer containing language basics as the last name Buļs who I interviewed in Rēzekne, discussed his fam­ well as specific information on communities in the Ludza region with ily tree with me. In this family tree, he said that he believed one part which Lutsi descendants are familiar seemed to be the best way to re­ of his family, with the last name Katkēvičs had Estonian speakers. He acquaint Lutsi descendants with their ancestral language and to raise traced this part of his family to the village of Špegi. When I visited interest in their ancestry. Špegi village, I spoke with people living there and they indicated that In the future I plan to create a Lutsi language course, so that de­ the Katkēvičs family no longer lived in the village, but had moved to a scendants and others interested in the Lutsi people will be able to learn different village nearby called Franapole. I was not able to track down some Lutsi at a rudimentary level. As a first step towards this future a further connection in Franapole, but in visiting the village cemetery goal and as a supplement to the primer, I also created a beginner’s I found a mix of known Lutsi last names such as Buls/Buļs/Bulis with grammar reference for Lutsi in the course of my Kone-funded work. other names including Katkēvičs. Once again no clear proof that Lutsi I created this resource so that there would be an option for readers to or other Estonian speakers lived in this village, but somewhat sugges­ take the next step and learn how the Lutsi language actually functions tive of their presence there nevertheless. Further investigation of this and is structured. The topics in the beginner’s grammar reference will Katkēvičs connection may yield additional information. also serve as the basis for the language course I plan on writing in the future. Finally, during my last year of Kone-funded work, I cre­ 7.3. Other areas ated a learner’s dictionary, which includes approximately 780 words translated into a number of languages. This is discussed in more detail Lutsis and people of Lutsi descent also surely lived in other villages in below. The words in the dictionary were chosen based on the 100 most this region for reasons such as marrying into another non-Lutsi fam­ common words in Estonian and words, which typically are found in ily. So, for example, the elderly Lutsi descendant I met in Ludza (dis­ beginning language courses, as it seemed that words from these two cussed above in Section 6) who could still count to 4 in Lutsi, was groups would be the most useful to learners in beginning to construct born and grew up in a non-Lutsi village – Voiti – though her family simple sentences. The dictionary also includes conjugation tables for came from a known Lutsi village in Brigi rural parish – Abricki. Simi­ 71 verbs. larly, today there are Lutsi descendants living in other villages in this region simply because they have come to own a home there, but not due to old family roots in these villages. 9. “Estonian” is how Lutsi descendants typically refer to their Lutsi origins.

458 459 ULDIS BALODIS

The target audience for the primer is first the Lutsi descendants themselves and second the larger Latvian-speaking public at large. The intended goal is at the minimum to acquaint descendants and others with the Lutsi language and the Lutsis as an ethnic group within Lat­ via; however, the hope is that the primer, beginner’s grammar reference, and dictionary will lay the groundwork not only for a future language course, but also for eventual Lutsi language revival efforts.

8.1. Orthography

Lutsi has no consistent history of being used as a written language. Its only use as a written language seems to have been in the form of a few letters written in Lutsi by Paulopriit Voolaine and perhaps other Lutsi community members10 to Oskar Kallas, who at that time was serving as the ambassador of the Republic of Estonia to the . I found these letters in the archives of the Estonian Cultural Society. One such letter11 is reproduced in Figure 5. Voolaine employed an orthography similar to that used for Esto­ nian. Some additional sounds and features are marked in Voolaine’s Lutsi orthography, which are not marked in the standard Estonian or­ thography. These include, palatalization, which Voolaine writes using a straight apostrophe <′> after the palatalized letter, and the glottal stop written using a curved apostrophe <’>. Long vowels and gemi­ nate consonants are written, as in Estonian, using a sequence of two identical letters. The close-mid back unrounded vowel in initial sylla­ bles is written, as in Estonian, with the character <õ>. The retracted-e Figure 5. Letter written by Lielie Tjapši Lutsi residents and Paulopriit Voolaine to vowel (transcribed in the Finno-Ugric Phonetic Alphabet as ) is Oskar Kallas in 1934. Source: Estonian Literary Museum (Eesti Kirjandusmuuseum). also written with <õ>, as in Võro. The vowels written in Estonian as <ä>, <ö>, <ü> are written the same in Voolaine’s Lutsi orthography. As my orthography would serve a different population than that using Voolaine’s orthography, I chose not to adopt Voolaine’s system 10. The letters seem to be written in several different types of handwriting, so likely it was not only Paulopriit Voolaine writing them. However, it seems logical for the Lutsi primer. The population that Voolaine studied was either to presume that Voolaine was involved with encouraging and/or helping Lutsi still Lutsi-speaking or, if not, had fluent Lutsi speakers available for community members to write these letters to Kallas in their own language. modeling speech. The current target population is Latvian-speaking 11. The text of the letter translates as: “In Jānikülä (Lielie Tjapši), Sunday 28. I. and there are no Lutsi speakers available any longer for modeling 1934. Dear Mr. O. Kallas! A great thanks for the Bible, which we received today and started to read. There is much that gives knowledge (to us). The elder Lutsis:” proper pronunciation. Therefore, I decided to model my orthography (Thank you to Karl Pajusalu for this translation.) on that used for Latvian and specifically for the High Latvian dialect

460 461 ULDIS BALODIS EXPEDITIONS AMONG THE LUTSI ESTONIANS AND THE DESIGN OF LANGUAGE LEARNING MATERIALS

(i.e., Latgalian), which has a separate literary tradition from the rest Voolaine Balodis of Latvian and which is native to the region where Lutsi was spoken. ä ä In other words, Latgalian is the type of Latvian that most Lutsi speak­ ers either speak themselves or would at least be familiar with. This ö ö is convenient, because while Latvian does not have a vowel equiva­ ü ü lent to Estonian <õ>, Latgalian does have a similar-sounding vowel. õ (in initial syllables) y This vowel, written in Latgalian with is a close central unrounded vowel. Much as in Lutsi, consonants in Latgalian are palatalized be­ õ (in non-initial syllables) e fore front vowels with palatalization only being written explicitly for ’ (glottal stop) q palatalized consonants occurring in other positions. In Latgalian, long vowels are marked with a macron, as in Latvian. Palatalization marked with ′ Palatalization marked with comma under after palatalized consonant palatalized sound (e.g., ņ, ŗ), unmarked Thus, the Lutsi orthography I designed is a hybrid of the Latga­ (e.g., n′, r′), unmarked before before front vowels except where lian orthography and the orthography used for Võro. The intent was front vowels. palatalization is also marked explicitly. also to make an orthography that could be read as easily as possible by Latvian speakers and, while not necessarily resulting in a perfect Long vowels are written with Long vowels are written with a macron a double vowel (e.g., aa, üü). (e.g., ā, ǖ). rendition of Lutsi as once spoken by actual speakers, a reader’s pro­ nunciation would come fairly close to this. Again, the main issue at Geminate consonants are Geminate consonants are written with two hand is the lack of Lutsi speakers and so readability and ease of use written with two consonants consonants (e.g., pp, tt). (e.g., pp, tt). for non-speakers seeking to learn some Lutsi was the most important criterion I felt it was necessary to keep in mind. Table 1. Comparison of the Lutsi orthographies used by Voolaine and that de- In the Lutsi orthography I use in the primer, the close-mid back signed by myself for the Lutsi primer. unrounded vowel in initial syllables is written with , while retract­ ed-e found in non-initial syllables is written with . As retracted- e sounds quite similar to , this seemed a logical step to take to simplify the Lutsi orthography for readers of the primer and potential to use for early stages of language revitalization. I may revisit the language learners. Palatalization is not marked before front vowels. marking of half-long vowels in the future if language revitalization Palatalization is written explicitly in other environments, and also pre­ efforts begin to be successful. ceding , using a comma under the palatalized letter as in Latvian The following compares the text of Voolaine’s letter in his or­ and Latgalian. Long vowels are written with a macron, as in Latvian thography and mine: and Latgalian. Geminate consonants are written by doubling the con­ sonant. The glottal stop is written, as in Võro, with . A comparison Voolaine’s orthography: My Lutsi orthography: between Voolaine’s orthography and mine is shown in Table 1. Jaanikülähn, pühäpäiv 28. I. 1934. Jānikülähn, pühäpäiv 28. I. 1934. In an earlier version of my orthography (described in Balodis Hüä Ezänd O. Kallas! Hüä Ezänd O. Kallas! 2015), I also marked half-long vowels with a macron. This is also Suur ait′umma üle piiblia, miä Sūr aițumma üle pīblia, miä saim done in the Livonian orthography and at the time I felt it would re­ saim täämbä kätte ne naksi lugõ­ tǟmbä kätte ne naksi lugema! Um sult in a more accurate Lutsi pronunciation. However, ultimately this ma! Um pal′l′ö, miä and oppuist. pallö, miä and oppuist. seemed to yield a Lutsi spelling that seemed possibly too complicated Vanõmba’ maamihe: Vanembaq māmihe:

462 463 ULDIS BALODIS EXPEDITIONS AMONG THE LUTSI ESTONIANS AND THE DESIGN OF LANGUAGE LEARNING MATERIALS

8.2. Design of the primer and dictionary In the second half of the primer, I juxtapose historical and mod­ ern photographs of particular Lutsi villages, give information about This section describes the design of the Lutsi primer, beginner’s those villages and their inhabitants, and also include text samples from grammar reference, and dictionary. The Lutsi primer is divided into the villages included in the primer from earlier documentation work. two main sections. The first section provides information on the Lutsi This was intended to show readers how Lutsi varied from region to language and teaches the Lutsi alphabet and rudimentary grammar region and village to village. Some of this material was quite old as, using example sentences. The second section gives information on for example, the Lutsi spoken in the villages of Mērdzene pagasts is particular Lutsi villages, their inhabitants, and also examples of Lutsi only known to us from the documentation of Oskar Kallas in 1893. from each pagasts in order to show how Lutsi varied from region to When possible, each text example is presented as it would be written region. in the practical Lutsi orthography I developed in the course of this The first part of the primer begins with a description of the Lutsi work. This is done so that readers who have learned how to use the orthography and other introductory material. The Lutsi alphabet then Lutsi orthography can read each example out loud in order to hear is presented with language examples structured in the following way. approximately how it sounded and to make the Lutsi of each village Two facing pages are devoted to each letter of the Lutsi alphabet. more tangible to them. When available, the Lutsi is presented along­ For each letter a word is selected which either begins with that let­ side its translation in literary Estonian and then my own translation of ter or contains that letter, as not all sounds occur at the beginning the literary Estonian into Latvian. of words. On the right facing page, the Lutsi word is given with its Figures 7–9 show examples from the second half of the primer. Latvian translation. Five sentence examples using this word are given The villages discussed in this section are grouped according to the below. As the nominative, genitive, and partitive cases are the three historical pre-WWII pagasts in which they are located. As described most important cases for forming simple expressions in Lutsi and as earlier in this report, while the pagasts administrative division existed knowledge of a word’s genitive case form is necessary for forming both before and after the Soviet occupation of Latvia, the actual bound­ most other singular cases and the nominative plural, the first three ex­ aries of these units were completely different in these two periods. To amples for each word present the nominative, genitive, and partitive my knowledge, there exist no maps overlaying these boundaries on singular case forms in short sentence examples with Latvian transla­ top of each other. As all Lutsi documentation and its related discus­ tions. The other two sentence examples show a different case form in sion is referenced using the old pre-occupation pagasts divisions, I use, either a locative case or one of the other oblique cases such as continued this practice in my primer. However, so that present-day the comitative. readers more acquainted with the modern-day pagasts divisions could On the left facing page, a photograph is presented, which is re­ understand where these villages are located, I drew maps, which show lated to the word in the sentence examples. Primarily, I used either where present-day pagasts division boundaries fall within the bounda­ pre-WWII photographs showing Lutsi people in their villages or my ries of the historical pagasts divisions of Pilda, Nirza, and Mērdzene. own photographs of these communities from the last years. In the ex­ The maps also show the location of each village discussed in that sec­ ample shown in Figure 6, the word ahi ‘oven’ is presented with five tion. Figure 7 shows the map I created for the historical Pilda pagasts examples below it and a picture of a Lutsi woman baking bread in at the beginning of that section. an oven. The photograph is a historical photo of Lutsi speaker Tekla The example pages from this part of the primer are shown for Jarošenko baking bread in her home in Lielie Tjapši taken by V. Nii­ Lielie Tjapši village. Figure 8 shows the introductory information on lus in 1936 and is from the archive of the Estonian National Museum the village with historical pictures from the Estonian National Mu­ (Eesti Rahva Muuseum). seum archive on the right. Figure 9 shows present-day photographs

464 465 Aa

ahi 'krāsns' Ahi um pallav. Krāsns ir karsta.

Vanaimä saiz aho man. Vecmāmiņa stāv pie krāsns. Vanaimä näge ahja. Vecmāmiņa redz krāsni. Vanaimäl um ahi. Vecmāmiņai ir krāsns. Lēbä um ahohn. Maize ir krāsnī. Figure 6. Example pages from the language section of the Lutsi primer.

466 467 ULDIS BALODIS EXPEDITIONS AMONG THE LUTSI ESTONIANS AND THE DESIGN OF LANGUAGE LEARNING MATERIALS

of the village taken by me during the course of my fieldwork and then a description of the Nikonovs family on the right. As detailed ear­ Pildas pagasts lier in this report, the Nikonovs family was the last family to actively speak Lutsi. The published version of the grammar will also contain Lutsi language examples with translations into Latvian from sources like the Eesti murded volume on the Estonian language islands (Mets et al. 2014). As stated previously, in addition to writing the primer, I also wrote a beginner’s grammar reference for Lutsi, which is intended as a supplementary resource for readers who wish to learn more about Lutsi grammar and who are interested in attempting to form more complex sentences. This is a “beginner’s” reference in the sense that a full reference grammar would be more extensive, include more ex­ amples, and would show in detail the differences in Lutsi between villages. This grammar reference gives basic information with short explanations and sentence examples for a wide variety of topics, but is not as in-depth or as detailed as a full reference grammar would be and does not discuss Lutsi dialect differences. The grammar reference was created primarily based on my analysis of the Lutsi texts in Eesti murded IX (Mets et al. 2014). See Table 2 on page 474. The Lutsi dictionary, which I created in the final year of my Kone- funded work, is structured as a word list and contains approximately 780 words. As the dictionary is a word list, it was possible to translate each word into several languages, thereby increasing the scope of the potential audience for the dictionary. The languages (or written tradi­ Mūsdienu pagasti Ciemi tions, in the case of Latgalian) used for the translations include Lat­

1 - Pildas pagasts a - Pylda (Pilda) vian, Latgalian, English, Estonian, German and Livonian. Latvian and 2 - Kaunatas pagasts b - Sūre Tsäpsiq / Jānikülä (LIelie Tjapši) Latgalian were included as these are the languages or written forms, 3 - Pureņu pagasts c - Kirbu külä (Škirpāni) which now constitute the primary language of Lutsi descendants. Eng­ 4 - Isnaudas pagasts 5 - Ņukšu pagasts lish and German were included to give this work a point of connection 6 - Nirzas pagasts beyond Latvia and allowing others to also learn about Lutsi. Estonian was included as it is a close relative of Lutsi and also due to the fact that Estonia is the historical point of origin for the Lutsis. Livonian Mūsdienu pagastu robežas rādītas pelēkā krāsā vēsturiskā pirmskara Pildas pagasta robežās. was included as it and Lutsi are Finnic languages spoken in Latvia, its Figure 7. Map of the historical Pilda pagasts with present-day (mūsdienu pagasti, inclusion in the dictionary allows readers to compare these two lan­ in Latvian) boundaries shown with the location of villages (ciemi) discussed in this guages and implicitly draws a connection between these two nations, section. Lutsi village names are followed by Latvian in parentheses. which form part of Latvia’s Finnic heritage.

468 469 Sūre Tsäpsiq / Jānikülä Lielie (un Mazie) Tjapši Lielie Tjapši toreiz

Meikula Jarošenko mājas Antons Jarošenko Lielie Tjapši, Pildas pag., L. Tjapši, Pildas pag., Foto: V. Niilus, 1936, ERM 754:106 Foto: A. Sang, 1936, ERM 756:22

Nikonovu ģimenes mājas Lielo Tjapšu ciemā. Foto: U. Balodis, 2013

Lielo Tjapšu ciems ir pēdējā vieta, kur ikdienā vēl varēja dzirdēt sarunas Ludzas igauņu valodā. Ludzas igauņu valodas pēdējais runātājs, Nikolajs Nikonovs, nomira 2006. gadā, taču aktīvas sarunas Ludzas igauņu valodā Tjapšu ciemā būtu bijis vēl iespējams šad tad dzirdēt līdz 1980. gadu vidum. Nikolaja vecmāmiņa Antoņina Nikonova bija ļoti prasmīga un dedzīga Ludzas igauņu valodas runātāja līdz savam mūža galam 1983. gadā. Skats uz Mazajiem Tjapšiem, kādreiz bijis igauņu ciems, tagad pārlatviskots Pats nekad nesatiku Antoņinu un arī ne Nikolaju, taču satiku Nikolaja atraitni Antonīnu Pildas pag., Foto: A. Sang, 1936 Nikonovu, kas nemācēja brīvi runāt Ludzas igauņu valodā, bet gan varēja daudz vārdu un īsus ERM 756:61 teikumiņus pateikt. Man prātā vienmēr paliks no mūsu sarunām, kā viņa teica – stāstot par savu Vecā tipa krāsns Rudzu pļauja. u Mazie Tjapši?, Pildas pag., vectēvu – Ludzas igauņu valodā “maq sinnu sali” jeb “es tevi mīlu”. Šos vārdus viņš viņai bija L. Tjapši, Pildas pag., Foto: A. Sang, 1936, ERM 756:53 Foto: A. Sang, 1936, ERM 756:83 teicis un vēlāk skatoties Heikki Ojansū piezīmes pierakstītas 1911. gadā redzēju šo pašu teikumu Ludzas igauņu valodā. No valodnieciskās transkripcijas bija skaidrs, ka Antonīnas izruna – par spīti tam, ka viņa brīvi nevarēja sarunāties Ludzas igauņu valodā – bija ļoti pareiza. Citu stāsti arī liecināja par Ludzas igauņu valodas aktīvo lietošanu Lielajos Tjapšos vēl samērā nesen. Jānis Germans tagad Lielajos Tjapšos vairs nedzīvo, bet ir cēlies no šī ciema. Viņš strādājis par šoferi un atceras, ka viens viņa draugs – arī šoferis – 1970. gados vedis kravu uz Tartu Igaunijā. Igauniski neesot pratis un tāpēc esot paņēmis līdzi no Lielajiem Tjapšiem vecu vīru, kas runājis Ludzas igauņu valodā. Šoferis neesot zinājis, cik atšķirīgas igauņu valoda runāta Ludzas apkārtnē un tā, ko runā Igaunijā, patiesībā ir. Kad abi aizbraukuši uz Tartu, vecais vīrs ir centies sarunāties ar vietējiem cilvēkiem Ludzas igauņu valodā, bet neviens neesot viņu varējis saprast. Šis stāsts liecina ne tikai par to, ka 1970. gados bija vēl valodas pratēji Lielajos Tjapšos, bet arī to, cik atšķirīga ir Ludzas igauņu valoda no literārās igauņu valodas.

Tekla Jarošenko ar savu meitu Antoņinu Maizes cepšana Meikula Jarošenko mājās L. Tjapši, Pildas pag., L. Tjapši, Pildas pag., Foto: A. Sang, 1936 Foto: A. Sang, 1936, ERM 756:13 ERM 756:19

Figure 8. Introduction to the section on Lielie Tjapši village with historical images of this Lutsi village and its inhabitants. Lielie Tjapši tagad Nikonovu ģimene: Pēdējie Ludzas igauņu valodas runātāji

Ceļš uz L. Tjapšiem no Pildas Pļava Foto: U. Balodis, 2013 Foto: U. Balodis, 2013

Antonīna Nikonova (1949-2014) ģimenes mājās Pēdējie Ludzas igauņu valodas runātāji: Nikolajs Lielajos Tjapšos. Antonīna saprata un mācēja Nikonovs ar savu vecmāmiņu Antoņinu Nikonovu. pateikt daudz vārdu Ludzas igauņu valodā. Foto: Nikonovu ģimenes foto albums Foto: U. Balodis, 2013

Vecas ēkas L. Tjapšos Vecas ēkas L. Tjapšos Foto: U. Balodis, 2013 Foto: U. Balodis, 2013

Antoņina Nikonova (1898-1983) ar Pauloprītu Volaini Antonīna Nikonova ar vīru Nikolaju Nikonovu (1944-2006) Foto: Nikonovu ģimenes foto albums Foto: Nikonovu ģimenes foto albums

1972. gadā, Guna Pence satikās ar Antoņinu Nikonovu un pierakstīja vairākas interesantas viņas atmiņas. Antoņina bija visu mūžu dzīvojusi Tjapšos. Viņas dēls prata ne tikai Ludzas igauņu valodu, bet arī literāro igauņu valodu, jo bija mācījies Igaunijā. Pence rakstīja, ka Antoņinas mājas bijusi kādreiz kā “pulcēšanās centrs” Ludzas igauņiem: “te svinēti svētki, notikušas sastapšanās ar pētniekiem no Igaunijas.” Pence arī rakstīja, ka Antoņina bijusi vienīgā, kas vairs nodevusi Ludzas Vecas ēkas L. Tjapšos Nikonovu ģimenes mājas Foto: U. Balodis, 2013 Foto: U. Balodis, 2013 igauņu valodu tālākām paaudzēm (Pence 1972:7-8). Līdzīgi kā citu atmiņās no pārējiem ciemiem, kas apskatīti šajā ābecē, Antoņina atcerējās, ka agrāk igauņi precējušies tikai viens ar otru. Viņa gan nezināja kāpēc Ludzas igauņi nonākuši Ludzas apkaimē kaut gan teica, ka “Tjapšos ienākušas 4 igauņu ģimenes (un sākušas te dzīvot) no Germu [ciema] puses. Agrāk te [Tjapšos] bijis mežs. (Pence 1972:12). Pauloprītam Volainem turpinājās draudzība ar Nikonovu ģimeni līdz mūža galam. Par šo liecina fotogrāfijas Nikonovu ģimenes albumos (dažas no kurām ir redzamas šajā lappusē). Pence arī pierakstīja Antoņinas atmiņas par Volaines braucieniem, kas atbilst tam, ko arī pats dzirdēju vēl pat šodien no vecāka gada gājuma Ludzas igauņu pēctečiem. Antoņina atcerējās, kā Volaine bija dzīvojis pie viņas vairākas vasaras, bet arī bija palicis pie citiem Ludzas igauņiem. Ziemassvētkos viņš dāvanas bija vedis no igauņiem Igaunijā. Pence raksta, ka svinības arī rīkotas pie Antoņinas un, ka “[r]īkoti lieli svētki, kur katram mazajam vajadzējis zināt kaut ko igauniski noskaitīt, tad arī dotas dāvanas.” (Pence 1972:12) Nikonovu ģimenes mājas Nikonovu ģimenes mājas Foto: U. Balodis, 2013 Foto: U. Balodis, 2013 Figure 9. Lielie Tjapši in the present day and information on the Lutsi-speaking Nikonovs family. ULDIS BALODIS EXPEDITIONS AMONG THE LUTSI ESTONIANS AND THE DESIGN OF LANGUAGE LEARNING MATERIALS

Chapter Topics 8.3. Future and implementation 1 Nominative case; genitive case; introduction to adjectives; olema ‘to be’ in present tense; With the completion of the Lutsi primer and following further editing ībä ‘very’; aitüma ‘thank you’ and review, I am seeking a publisher for it. I may also make it freely available on my Lutsi website (www.lutsimaa.lv). I hope also to trans­ 2 Present tense verb conjugation; late the Lutsi primer into English in the near future in order to make question words the information within it available to as wide of an audience as pos­ 3 Inner and outer locative cases, showing possession with the sible. Ultimately, I would like to use it for language classes, but further ; indirect object constructions with the ; funding and planning will be required to realize this plan. present tense negation; more question words 4 Partitive case; numerals; more on present tense verb conjugation 9. Conclusion 5 Imperfect tense; ordinal numerals; introduction to subordinate clauses This paper describes some of my primary observations and accompany­ 6 Total and partial objects; ; ; imperfect ing stories characterizing the present state of the community of Lutsi tense negation; introduction to postpositions/prepositions; descendants in eastern Latvia (i.e., Latgale), the background of research demonstratives into the Lutsis as well as gives information about the language/culture 7 ; ; primer, beginner’s grammar reference, and dictionary I have written and vyima ‘to be able to’; imperative; prohibitive the practical Lutsi orthography I have designed during the course of 8 Conditional; comparative and superlative my Kone-funded research into the Lutsis in the period of 2013 to 2016. 9 tense; pluperfect tense; ma/da verb infinitives; vaja During this work I found that presently the Lutsi language is no longer ‘have to, need (to)’; introduction to plural case forms actively spoken; however, I also found that Lutsi descendants are fre­ quently aware of their heritage and so this may create fertile ground for Table 2. Lutsi beginner’s grammar reference chapter topics. attempting a language revitalization effort for Lutsi in the future.

The back of the dictionary contains conjugation tables for 71 verbs found in the dictionary. These tables contain the affirmative and Acknowledgements negative present, imperfect, conditional, and imperative forms as well as the ma and da infinitive andnuq participle forms. While Lutsi con­ I would like to give my sincerest and most heartfelt thanks to the Kone jugation (and declension) types are no doubt similar to those of other Foundation for making possible the work described in this report and South Estonian varieties, since these have not yet been fully described, for funding my postdoctoral research on the Lutsis between 2013 and I included the conjugation tables so that readers and students could use 2016. I would also like to warmly thank Karl Pajusalu, Riho Grünthal, them as a reference in forming their own sentences utilizing the verbs Hannes Korjus, Līga Kondrāte, Aija Andersone, Antonīna Nikonova, in the dictionary. Indrek Jääts, Maido Selgmäe, and the many members of the commu­ nity of Lutsi descendants with which I met and worked during the course of this research as well as the many others whom I spoke with and who supported my work during these years.

474 475 ULDIS BALODIS EXPEDITIONS AMONG THE LUTSI ESTONIANS AND THE DESIGN OF LANGUAGE LEARNING MATERIALS

References lects IX. South Estonian language island texts.] Tallinn: Eesti Keele Instituut & Tartu Ülikool. Balodis, Uldis 2014: Lutsimaa: Land of the Ludza Estonians. Available at: Niilus, Valter 1937: Valimik leivu murdetekste. Choix de textes dialectaux ‹http://www.lutsimaa.lv/› leivu [A selection of Leivu dialect texts]. Akadeemilise Emakeele — 2015: Writing down Lutsi: Creating an orthography for a South Esto­ Seltsi toimetised 31. Tartu: Akadeemilise Emakeele Seltsi kirjastus. nian variety of Latgale. – Valoda: Nozīme un forma 6. Valodas sistēma Ojansuu, Heikki 1912: Virolaiset siirtokunnat lättiläisalueella, niiden lähtö­ un lietojums. Rīga: LU Akadēmiskais apgāds. 55–67. paikka ja -aika [Estonian settlements in the Latvian region, their ori­ Godiņš, Guntars 2015: Ludzas igauņu pasakas [Lutsi stories]. Rīga: Apgāds gin and time]. – Suomalainen Tiedeakademia. Esitelmät ja pöytäkirjat. Mansards. 7–26. Grunthal, Villem 1912: Eesti asundused lätlaste seas [Estonian settlements Pajusalu, Karl 2009: Estonians of Latgale. – Valodas Austrumlatvijā: Pētī- among the Latvians]. – Eesti Kirjandus: 369–373. juma Dati un Rezultāti. Languages in Eastern Latvia: Data and Results Iva, Sulev & Karl Pajusalu 2004: The Võro Language: Historical Develop­ of Survey. Via Latgalica, Supplement to Journal of the Humanities. ment and Present Situation. – Language Policy and Sociolinguistics Rēzekne, Latvia: Rēzeknes Augstskola. 171–181. I: “Regional Languages in the New Europe”. (International Scientific RAPLM/PRD (Reģionālās attīstības un pašvaldību lietu ministrija, Conference. Rēzekne Augstskola, Latvija; 20–23 May 2004.) Rēzekne: Pašvaldību reformu departaments [Ministry of Regional Development Rēzekne Augstskolas Izdevniecība. 58–63. and Self-Government, Department of Self-Government Reform]). Jääts, Indrek & Maido Selgmäe 2014: Kadunud hõim: Lugu Lutsi maarahvast. (2007). Administratīvi teritoriālā iedalījuma projekts [Administra­ Lost Tribe: A Story about the Lutsi Estonians. [Documentary film.] Eesti tive territorial division project]. [Accessed 4 April 2017.] Available at: Rahva Muuseum. ‹http://www.cirkulis.lv/wp-content/gallery/novadu_karte.jpg› Kallas, Oskar 1894: Lutsi maarahvas [Lutsi country folk]. Helsinki: Suoma­ Sang, August 1936: Lutsi maarahvas a. 1936 [Lutsi country folk in 1936]. – laisen Kirjallisuuden Seura. Eesti Kirjandus: 399–409. — 1903: Kraasna maarahvas [Kraasna country folk]. Helsinki: Suoma­ Silkāns, Ilmārs (2011, January 11): Soidi, kivki, raibakozi (pastāstīšu par laisen Kirjallisuuden Seura. Latgales igauņiem) [Soidi, kivki, raibakozi. (I’ll tell you about the Korjus, Hannes 2001: Ludzas rajona igauņi 2001. gadā [The Estonians of Estonians of Latgale)]. – Ezerzeme: Aglonas, Dagdas un Krāslavas Ludza District in 2001]. Latviešu folkloras krātuves arhīvs, LKF Nr. novada avīze. Available at: ‹http://www.ezerzeme.lv/index.php?action= 2116. 6. tabula. show&type=news&id=11028› — 2017: Ludzas igauņi: Zemes dieva tauta [The Lutsis: People of the Šķiņķis, Pēteris 1999: Teritorijas un teritoriālās stratēģijas Latvijā [Territo­ Earth God]. Rīga: Lauku Avīze. ries and territorial strategies in Latvia. PhD Dissertation]. University Latkovskis, Leonards 1968: Latgaļu uzvōrdi, palames un dzymtas [Latga­ of Latvia. lian last names, nicknames, and families/clans]. München: Latgaļu Turlajs, Jānis (ed.) 2012: Latvijas vēstures atlants [Historical atlas of Latvia]. izdevnīceiba. Rīga: Karšu izdevniecība “Jāņa sēta”. Lindström, Liina, Pärtel Lippus & Tuuli Tuisk 2019: The online database of Voolaine, Paulopriit 1925: Lutsi maarahvas 1925. a. [Lutsi country folk in the University of Tartu Archives of Estonian Dialects and Kindred Lan­ 1925]. – Eesti Kirjandus: 372–379. guages and the Corpus of Estonian Dialects. – Sofia Björklöf & Santra Jantunen (eds), Multi­lingual Finnic. Language contact and change. Uralica Helsingiensia 14. Helsinki: Finno-Ugrian Society. 327–350. Available at: ‹https://doi.org/10.33341/uh.85040› Mets, Mari, Anu Haak, Triin Iva, Grethe Juhkason, Mervi Kalmus, Miina Norvik, Karl Pajusalu, Pire Teras, Tuuli Tuisk & Lembit Vaba 2014: Eesti murded IX. Lõunaeesti keelesaarte tekstid. [Estonian dia­

476 477 ULDIS BALODIS

Pētniecības ekspedīcijas pie Ludzas igauņiem SOFIA BJÖRKLÖF un valodas apguves materiālu izveide Helsingin yliopisto

Uldis Balodis

1 Ludzas igauņi, jeb luci, ir vēsturiski dienvidigauniski runājoša ko­ Vepsän mailla Herran vuonna 2014 piena, kas vismaz trīs līdz četrus gadsimtus apdzīvojusi apmēram 50 ciemus pirmskara Pildas, Nirzas, Brigu un Mērdzenes pagastos Koneen Säätiön kieliohjelman rahoittaman hankkeemme Itämeren- Ludzas apkaimē. No 2013. līdz 2016. gadam, kā Kone fonda stipen­ suomalaisten kielten muutos monikielisessä ympäristössä – jonka tu­ diāts, doku­mentēju Ludzas igauņu kopienas pašreizējos apstākļus un loksiin käsillä oleva julkaisukin kuuluu – tavoitteisiin kuului jatkaa uzrakstīju Ludzas igauņu valodas ābeci latviešu valodā. Raksta pir­ Helsingin yliopiston tutkijayhteisön vepsän kielen puhujien keskuu­ majā daļā aprakstu Ludzas igauņu pētniecības vēsturi, kā arī Ludzas teen tekemiä kenttätyöretkiä (näistä ks. Puura 2006). Teimme näis­ igauņu kopienas un to pēcteču stāvokli mūsdienās. Raksta otrajā daļā tä lähtökohdista kesä–heinäkuussa 2014 matkan vepsäläisalueille. aprakstu Ludzas igauņu ābeci, vārdnīcu un gramatikas ievadu, kurus Hankkeemme vetäjä, itämerensuomalaisten kielten professori Riho sastādīju savas pētniecības gaitā kā arī aprakstu ieceres to turpmākajai Grünthal ja Helsingissä opintonsa aloittanut ja sittemmin oululais­ izmantošanai. Igauņu pētnieks Oskars Kallass (Oskar Kallas) bija pir­ tunut Heini Karjalainen ajoivat vepsäläisalueelle Petroskoin kautta, mais, kas 19. gadsimtā plaši dokumentēja Ludzas igauņus, to kultūru jossa Grünthal oli ollut konferenssin kutsupuhujana. Minä matkustin un valodu. Citi pētnieki visai plaši dokumentējuši Ludzas igauņu valo­ muiden sitoumusteni vuoksi yksin rautateitse Pietarin kautta Babaje­ du desmitgadēs pēc Kallasa pētniecības ekspedīcijas. Pēdējais Ludzas voon, josta muu seurue minut haki. Helsingistä vuokratun auton taka­ igauņu valodas runātājs nomira 2006. gadā un pēdējais cilvēks ar pla­ kontti oli täynnä vepsänkielisiä julkaisuja puhujille jaettaviksi. šām Ludzas igauņu valodas zināšanām aizgāja mūžībā 2014. gadā. Matkamme suuntautui ensinnä Vologdan läänin Babajevon kun­ Mūsdienās daži Ludzas igauņu pēcteči vēl atceras atsevišķus vārdus taan, jossa kukaan meistä ei ollut aiemmin käynyt. Kylät, joissa Vo­ un īsus teikumus Ludzas igauņu valodā, taču ikdienā šīs kopienas vai­ logdassa vierailimme, olivat Päžarʹ sekä kaikkein itäisimmät vepsä­ rākums lieto latviešu valodu (atsevišķi pēcteči arī krievu valodu) kā läis­kylät Pondal ja Voilaht. Näissä kylissä matkaseurueeseemme savu galveno sarunvalodu. kuului myös Irina Sotnikova, vepsäläisen kuukausittain ilmestyvän Kodima-lehden päätoimittaja ja Petroskoissa opiskellut syntyperäinen pohjoisvepsäläinen. Hän oli tuonut vepsäläisille vepsänkielisiä lehtiä, ja koska hän halusi edistää lasten lukutaitoa, hän järjestikin Päžarʹissa lukusession paikallisille lapsille. Sotnikova myös kirjoitti vepsäksi matkastamme jutun toimittamaansa lehteen (Sotnikova 2014). Loppu­ ajan vietimme Leningradin läänissä Podporožjen kunnassa Ladvan kylässä, josta kävimme myös Mäggärʹven kylässä. Kylien sijainti on nähtävissä vepsän puhuma-alueen kartasta, joka tässä julkaisussa on painettu Heini Karjalaisen artikkeliin (Karjalainen 2019: 59). 1. Kiitän Riho Grünthalia, Heini Karjalaista, Ulriikka Puuraa ja Saarni Laitista raporttini kommentoinnista.

Multilingual­ Finnic. Language 478 contact and change. 479–493. Uralica Helsingiensia 14. Helsinki 2019. ‹https://doi.org/10.33341/uh.85047› SOFIA BJÖRKLÖF VEPSÄN MAILLA HERRAN VUONNA 2014

Kielitieteellisestä työstä ja kielenoppaista indefiniittipronominien järjestelmää käsittelevässä artikkeliväitös­ kirjassaan (Karjalainen 2016). Riho Grünthal jatkoi aiempia tutkimuksiaan keräämällä sopivilta Minä olin matkalla oppimassa. Hankkeessamme ennen minua kielenoppailta­ itselleen mieluista aihetta, verbiparadigmoja, sekä var­ reilun puoli vuotta työskennellyt ja useilla aikaisemmilla Vepsän-­ misti nominien paradigmoja, muutoin mitä milloinkin sai. Oli kiin­ matkoilla kenttätöitä tehnyt Ulriikka Puura väitteli sittemmin vepsä­ nostavaa huomata, että oppaat eivät kuitenkaan millään tuottaneet läisestä kieliyhteisöstä (Puura 2019), mutta omat tutkimuskohteeni perfektimuotoja,­ vaan perfektin funktiossa käytettiin imperfektiä. eivät ainakaan vielä ole liittyneet vepsään saati itse kerättyyn kenttä­ Grünthal on julkaissut tältä ja muilta matkoiltaan keräämäänsä aineis­ aineistoon. Kuunteluoppilaana kirjoitin muistiinpanoja korva tarkkana toa Vepsän kieliopissa (Grünthal 2015). Hauskimpiin haastatteluihin ja äänitin kaiken siinä missä muutkin. Joitakin haastatteluja videoin. kuuluivat ne, jotka tehtiin erään keski-ikäisen pariskunnan kanssa. Puhutusta vepsästä kiinni saaminen vei oman aikansa, sillä vepsän Heistä toinen, Aleksander Mednikov, oli kotoisin Pondalasta, toinen, kielen kursseilla opitun lisäksi venäjän taitoa olisi selvästi tarvinnut Pondalan museonjohtaja Vera Mednikova, puolestaan Voi­lahdesta. alkeiskurssia laajemmin. Lähes kaikilla puhujilla tapahtuva toistuva Heillä oli tapana hymy suupielessä mainita, että toisen antama muoto koodinvaihto venäjään aiheutti minulle päänvaivaa. Koodinvaihtoa oli vallankin väärä ja kertoa, kuinka sama asia heidän omassa koti­ tapahtui jatkuvasti siitäkin huolimatta, että puhujat tiesivät kieli­ kylässään sanottiin. Nämä kielenoppaat olivatkin poikkeuksellisen rajoitteestani. Tämä johtui todennäköisesti siitä, että puhujilla, jotka tietoisia kielestään. Vanhin kielenoppaamme, Maria Triškina, oli eivät ole olleet tekemisissä standardivepsän kanssa, ei välttämättä vuonna 1927 syntynyt Päžarʹin Jušmäellä samaan taloon, pertʹhe, jos­ omassa idiolektissaan ole olemassa kategorioita ”vepsä” ja ”venäjä”.2 sa hän edelleen nyt jo kylänsä vanhimpana eli. Hän oli mennyt kou­ Onpa tunnustettava, että matkamme alussa venäjän aksentilla lausut­ luun vuonna 1934, jolloin vepsän kielellä alettiin opettaa, ja osasi lu­ tuihin vepsänkielisiin tervetulotoivotuksiin tulin koko päivän matkus­ kea ja kirjoittaa vepsää sekä pystyi edelleen laskemaan vepsäksi. Se tamisesta väsyneenä vastanneeksi, etten puhu venäjää. on hyvin harvinaista, sillä sujuvankin vepsän puhujat tapaavat vaihtaa Kuuntelemiseen harjaantuessa korvaan tarttui kuitenkin leksi­ venäjän koodiin juuri numeraaleja käytettäessä. kaalisiakin eroja: Voilahdessa parillakin kielenoppaalla esiintyi hond Heini Karjalainen, jonka väitöskirjan tekoa Koneen Säätiö ra­ ’huono’, kun kylävepsässä tavallisempi adjektiivi on huba. Päžarʹissa hoitti neljän vuoden ajan, keräsi aineistoa väitöskirjaa varten. Karja­ samanmerkityksisen sanan komparatiivi oli honomb. laisen kenttätyö liittyi ensisijaisesti hänen laatimaansa indefiniitti­ Kaikki majoittajamme toimivat myös kielenoppainamme. pronomineja käsittelevään kyselykaavakkeistoon, nk. elisitaatiotestiin, Päžarʹissa meidät majoittivat opettaja Olga Smirnovan vanhemmat jota hän kävi läpi yhdessä kielenoppaiden kanssa. Karjalainen luki Nina Fomina ja Nikolai Fomin. Olga Nikolajevna itse kiersi muka­ kielenoppaille esimerkkilauseensa ja mahdolliset pronominit, joil­ namme näyttämässä meille paikkoja, tutustuttamassa ihmisiin ja aut­ la lauseet voisi täydentää. Tiivis yhteistyö olikin tarpeen, sillä välil­ tamassa haastatteluissakin. Kotona meidät ruokittiin muun muassa eri­ lä oppaille oli aihetta selittää mitä heiltä toivottiin. Yhtäältä latina­ laisilla itse paistetuilla herkullisilla piirakoilla, joita oli aina runsaasti. laisten kirjainten lukeminen ei olisi kaikilta sujunut, toisaalta aivan Niitä riitti myös Pondalassa, jossa Vera Mednikova oli järjestänyt kaikki kielenoppaat, joiden kanssa kyselyä tehtiin, eivät välttämättä emännäksemme Galina Serdtsovan. Hän asuu perinteistä vepsäläis­ heti oikein tavoittaneet kyselyn ideaa. Haastateltavien kanssa pääs­ taloaan kesäisin ja viettää talvet kaupungissa. Kaikissa paikoissa kai­ tiin kuitenkin aina yhteisymmärrykseen, ja he antoivat keskustelujen killa aterioilla tarjottiin ruokajuomana teetä ja perusruoka-aineksiin aikana paljon hyviä kommentteja eri vaihtoehdoista. Näiden haas­ kuului keitetty tattari, jossa rasvaa ei säästelty. tattelujen tuloksia on sekä tässä julkaisussa Karjalaisen artikkelissa Borrowing morphology (Karjalainen 2019) että hänen vepsän kielen 2. Kiitos Ulriikka Puuralle tästä huomiosta.

480 481 VEPSÄN MAILLA HERRAN VUONNA 2014

Kuva 1. Fominien pihapiiri Päzarʹissa. Kuva: Sofia Björklöf.

Ladvaan menimme matkakumppanieni vanhojen ystävien Maria Abramovan ja hänen poikansa Nikolai Abramovin luokse (kuva 2). Sekä äiti että poika ovat tulleet tunnetuiksi vepsäläisaktivisteina. Nikolai Viktorovitš, vepsän literaturan klassik, on suosittu runoili­ ja. Ensimmäinen hänen runokokoelmistaan, saman ikäisenä julkais­ tu Koumekümne koume (Abramov 1994), on samalla ensimmäinen vepsän­kielinen kaunokirjallinen teos. Maria Aleksejevna puolestaan on laulanut ja lausunut omia runojaan vepsäläisten tilaisuuksissa ja jaksoipa hän niitä meillekin lausua. Grünthalia sydämellinen Maria Aleksejevna kutsui pojakseen, Karjalaista tyttärekseen, ja minustakin taisi sellainen tulla. Emäntämme opetti Karjalaista ja minua paista­ maan piirakoita, pirgoid, joiden taikina kaulittiin ohueksi, keskelle lisättiin sokeria, ja ennen paistamista taitettu piirakka vielä voideltiin itse linnunsulasta tehdyllä siveltimellä (kuvat 3 ja 4). – Tammikuussa 2016 saimme suruviestin Kolja Abramovin menehtymisestä. Olen kii­ Kuva 2. Maria Abramova ja Nikolai Abramov. Kuva: Sofia Björklöf. tollinen, että sain tutustua häneen.

482 483 SOFIA BJÖRKLÖF VEPSÄN MAILLA HERRAN VUONNA 2014

Paikallisia kuriositeetteja

Matkalla riitti kommelluksiakin. Mieliimme lienee parhaiten jää­ nyt, kuinka professori Grünthal yritti keskustella sonnin kanssa sil­ lä tuloksella, että jouduimme miltei eläimen hyökkäyksen kohteeksi. Vähäinen karja kulkee vielä Päžarʹin kylänraitilla ajettuna ja niityil­ lä vapaina. Sievistä lehmistä ja vasikoista ei juuri harmia ole, mutta kulkureitillämme oleili pari sonnia lehmineen, joita ensinmainittujen tietysti oli suojeltava. Eräänä päivänä kulkiessamme kylän toiseen päähän haastatteluja tekemään sonni ja lehmät sattuivat olemaan rei­ tillämme. Sonneista ei arvaa kovin läheltä ohi mennä, joten jäimme odottelemaan niiden liikkumista tieltämme. Toinen sonni alkoi vai­ kuttaa meistä hieman kiinnostuneelta, ja oli virhe sanoa sille matalalla miehen äänellä ”muu”. Tästäkös sonni pahastui ja kukapa tietää, minkä solvauksen professori sille tuli esittäneeksi. Eläimen käytös muuttui ja se alkoi lähestyä meitä. Paikka oli täysin aukea: ainoa pelastuksemme oli vain muutaman kymmenen metrin päässä sijainnut luhistumaisil­ Kuva 3. Piirakat ja linnunsulkasivellin. Kuva: Sofia Björklöf. laan oleva katos, jonka suojiin hakeuduimme. Suurempaa hyökkäystä barrikadimme ei olisi kestänyt, sillä sen päälle kiipeäminen olisi ollut paitsi turhaa myös vaarallista. Tämän kohtaamisen jälkeen kuljimme kyläteillä kirjaimelliset kättäpidemmät, karahkat, käsissämme. Elämä vepsäläiskylissä on edelleen monin osin kuin suomalai­ sella maaseudulla ennen viime sotia. Heinä kuivatetaan edelleen haa­ sioissa ja nostetaan sen jälkeen heinäsuoviin (kuva 5), joita molem­ pia näimme Päžarʹissa useita. Pyykkiä pestään ja huuhdotaan joissa (kuva 6) ja kylpeminen tapahtuu luonnollisesti saunoissa, jotka pe­ rinteisimmillään sijaitsevat rivissä joen rannalla (kuva 7). Jos külbet on kovin kuumaksi lämmitetty, ei parane nautiskella löylyistä turhan pitkään ennen kuin aloittaa pitkien hiusten pesemisen, mikäli haluaa välttyä huonovointisuudelta, sillä – kuten suomalaiset hyvin tietävät – vanhassa rakennustyypissä peseytyminen tapahtuu löylyhuoneessa.­ Pesuvesi lämmitetään kiukaan yhteyteen rakennetussa padassa. Kylä­ kaupassa ostosten summa lasketaan nopein sormin helmitaululla (kuva 8). Päžarʹissa Fominin perheen venevajassa taas oli isännän it­ sensä rakentama soutuvene (kuva 9) ja verkkojen kohot olivat koivun­ tuohta (kuva 10). Kuva 4. Piirakanpaistajat Sofia Björklöf, Maria Abramova ja Heini Karjalainen.Kuva: Nikolai Abramov. 484 485 Kuva 5 (vas. ylh.): Heinäsuova ja matala haasia Päzarʹissa. Kuva 6 (vas. alh.): Pyykin huuhtomista Pondalassa. Kuva 7 (oik. ylh.): Saunat Ladvassa. Kuva 8 (oik. alh.): Vepsän­kielistä asiointia Pondalan kaupassa. Helmitaulu etualalla, asiakkaana Vera Mednikova. Kuvat: Sofia Björklöf. VEPSÄN MAILLA HERRAN VUONNA 2014

Kuva 10. Verkkojen koivuntuohikohot. Kuva: Sofia Björklöf.

Vanhana kansatieteen harrastajana ja opiskelijana jaksan edel­ leen innostua pienten museoiden aineellisen kulttuurin kokoelmista. Sanastontutkijalle on myös mukavaa, että esineiden nimet on merkit­ ty itämerensuomalaisella kielellä, vaikka ne sinänsä tuttuja ovatkin. Matkamme aikana kävimme Päžarʹin, Pondalan ja Kurban vepsäläis­ museoissa. Päžarʹissa eksentrinen museonjohtaja Vladimir Triškin, aiemmin mainitun kielenoppaan poika ja ilmiselvä taiteilijaluonne, halusi tarjota meille opastetun vepsänkielisen kierroksen rakkaassa museossaan. Äänitinkin saamamme opastuksen. Halusimme kaikki hankkia kesäasuntoihimme Vepsänmaalla sa­ moin kuin Venäjän maaseudulla yleisemminkin käytetyt, käsienpesuun tarkoitetut metalliset vesiastiat, mutta emme onnistuneet niitä sinänsä kaikenlaista myyvistä kyläkaupoista löytämään. Muovisia olisi ollut tar­ jolla, mutta niistä emme viehättyneet. Mainittiinpa, että jostakin vanhasta

Kuva 9. Soutuvene ja sen veistäjä Nikolai Fomin. Kuva: Sofia Björklöf. 489 SOFIA BJÖRKLÖF VEPSÄN MAILLA HERRAN VUONNA 2014 talosta pitäisi saada, jos metallista mielisi. Lienimmekö muotivirtauksen henkisiä edelläkävijöitä, sillä vuonna 2016 hieman erilaisella mekanis­ milla toimivia metalliversioita oli tarjolla Suomen tavarataloissa. Olin viisi vuotta aikaisemmin ollut suomalais-ugrilaisen kielen­ tutkimuksen määräaikaisen professorin Janne Saarikiven johtamal­ la kenttätyökurssilla Tverin Karjalassa, lukenut ja kuullut matka­ kertomuksia. Silti en välillä voinut olla ihmettelemättä, miten paljon voi Venäjän maaseudulla (huonokuntoista) tietä riittää. Kartalla lyhyeh­ költäkin näyttävän hiekkatien ajaminen voi kestää pitkään, sillä joitakin teitä ajavat jatkuvasti raskaat puutavaraa kuljettavat rekat ja tiestö on kauttaaltaan suurilla, syvillä kuopilla. Ajo tällaisella tiellä onkin tark­ kaa puuhaa, kun kuoppia pitää parhaansa mukaan kierrellä. Alkuun vain tunnin ajomatka keinutti tottumatonta vielä seuraavanakin aamuna kuin olisi merenkäynnissä ollut. Kolmen ajokortillisen seurue helpottaa myös matkantekoa huomattavasti, kun ajamista voi olla päivässä yhtä mittaa helposti kahdeksan tuntia. Vaikka taisinpa minä jonkin kerran sellaisen rupeaman innostua kerrallakin ajamaan. Matkapuhelinverkossakaan ei maaseudulla ole kehumista, ja kotiin soitetaan niin, että kiivetään kaik­ kein korkeimmille paikoille ja vielä niillä sijaitseviin pylväisiin tai tika­ puille (kuva 11). Internet taas ei tule Vepsänmaalla mieleenkään.

Kenttätyön käytäntöä ja tekniikkaa

Kantapään kautta tunnetusti oppii, tällaisellakin matkalla. Sen tie­ sin, että nauhuriin on oltava mukana paljon uusia paristoja. Äänitystä aloitettaessa laitteessa olevien paristojen pitää mieluiten olla uudet ja äänittäviä nauhureita on hyvä olla useita; videokameran ääniraita ei korvaa erillistä nauhuria, vaikka onkin oiva lisä. Suomalais-ugrilaisen oppiaineemme kenttätyömatkoilla ei ollut videoitu järjestelmällises­ Kuva 11. Riho Grünthal Mäggärʹvellä kenttää etsimässä. Kuvassa etuvasemmalla myös aura. Kuva: Sofia Björklöf. ti, ja videokameraa käyttäessäni tulin havainneeksi, että siihenkin pitäisi olla mukana useampia akkuja. Myös muistikortteja on oltava useita, sillä sen varaan ei missään nimessä pidä laskea, että ehtii tyh­ pienille dvd-levyille. Tietokone myös tarvitsee säännöllisesti virtaa, jentää korttia kannettavalle tietokoneelle. Tyhjentäminen ei kannata jota kenttäolosuhteissa ei aina ole saatavilla. Videokameran kolmi­ siksikään, että on sitä parempi, mitä useammalla tallennusvälineellä jalka on välttämätön, jotta kuva ei tärise ja kuvaaja pystyy käyttämään tiedostot ovat tallessa. Videotiedostot vaativat lisäksi paljon tallennus­ myös nauhuria sekä tekemään muistiinpanoja, mutta varusteisiimme tilaa, eikä niitä kannata kotioloissakaan turhaan yrittää tallentaa liian se onneksi kuuluikin. Mutta vasta vuonna 2016 järjestetyssä Koneen

490 491 SOFIA BJÖRKLÖF VEPSÄN MAILLA HERRAN VUONNA 2014

Säätiön uralilaisten kielten dokumentaatiota, kuvaamista ja revitali­ Kirjallisuutta saatiota käsitelleessä kesäkoulussa opin Michael Rießleriltä, että Mik­ ki Hiiri -mallisen nauhurin mikrofonien vaahtomuovisuojukset ovat Abramov, Nikolai 1994: Koumekümne koume. Runokirj. Petroskoi: Karjala. itse asiassa ulkokäyttöön tarkoitetut ja ne pitää poistaa sisällä nau­ Grünthal, Riho 2015: Vepsän kielioppi. Apuneuvoja suomalais-ugrilaisten hoitettaessa. Haastatteluja tehdessä kannattaa myös muistaa kirjoittaa kielten opintoja varten XVII. Helsinki: Suomalais-Ugrilainen Seura. muistiinpanoja nauhoittamisesta huolimatta, sillä nauhoituksista on Karjalainen, Heini 2016: Yhteisöt kontaktissa, systeemit muutoksessa. Vepsän kielen indefiniittipronominien järjestelmä. [Väitöskirja.] myöhemmin huomattavasti helpompi etsiä haluamiaan asioita, kun Acta Universitatis Ouluensis B Humaniora 142. Oulu: Oulun yli­ niiden metatiedot ja sisältöä on jo paperilla. Esimerkiksi paradigmat opisto. Artikkeliväitöskirjan johdanto-osa saatavissa: ‹http://urn.fi/ kannattaa niitä kyseltäessä kirjoittaa saman tien kokonaan. Samoin urn:isbn:9789526212234› tuttu tapa tehdä nauhoituksen alkuun sanallinen muistiinpano sen — 2019: Borrowing morphology: The influence of Russian on the Veps osanottajista, paikasta ja päivämäärästä on edelleen tarpeen. system of indefinite pronouns. – Sofia Björklöf & Santra Jantunen Kaikenlainen dokumentoiminen, siis valokuvaus ja päiväkirjan (eds), Multi­lingual Finnic. Language contact and change. Uralica Hel­ pitäminen kenttätyöretkellä ovat mitä suositeltavimpia harrasteita. singiensia 14. Helsinki: Finno-Ugrian Society. 55–87. Saatavissa: Mitä enemmän jaksaa kirjoittaa tilanteista ja tunnelmista heti sama­ ‹https://doi.org/10.33341/uh.85033› na päivänä, sen enemmän päiväkirjasta on jälkeenpäin iloa ja hyötyä Kettunen, Lauri 1945: Tieteen matkamiehenä. Kaksitoista ensimmäistä ret- tutkimuksessakin. Asiat, jotka ajankohtaisina ja juuri tapahduttuaan keä 1907–1918. Helsinki: Werner Söderström osakeyhtiö. tuntuvat aivan yksinkertaisilta tai itsestäänselviltä, unohtuvat kuiten­ Puura, Ulriikka 2006: Vepsäläinen kielikylpy. – Alkukoti 8: 38–41. [Hel­ kin yllättävän helposti ja nopeasti. Matkapäiväkirjaa pitikin jo itäme­ singin yliopiston ylioppilaskunnan sukukansavaliokunnan lehti.] Saa­ tavissa: ‹http://www.helsinki.fi/hyy/skv/alkukoti/alkukoti8www.pdf› rensuomen tutkimuksen saralla valtavan työn tehnyt Lauri Kettunen, — 2019: Sinä iče oled vepsläine, voib sanuda, ka? Vepsäläisyyden raken- jonka ensimmäisestä Vepsän-matkasta vuosina 1917–1918 voi lukea tuminen ja 2000-luvun vepsän kieli. [Väitöskirja.] Helsinki: Helsingin hänen muistelmateoksestaan Tieteen matkamiehenä (Kettunen 1945: yliopisto. Artikkeliväitöskirjan johdanto-osa saatavissa: ‹http://urn.fi/ 272–407). E. N. Setälän kanssa vuotta aiemmin vepsäläisalueen läpi URN:ISBN:978-951-51-4878-0› matkannut etnomusikologi A. O. Väisänen on puolestaan julkaissut Sotnikova, Irina 2014: Matk vepsläižidennoks Vologdan agjaha. – Kodima, sanomalehdessä sarjan lyhyitä matkakertomuksia (Väisänen 1916). Reduku 2014 №10 (259): 4–5. Saatavissa: ‹http://kodima.rkperiodika. ru/wp-content/uploads/2015/02/Kodima_10_2014.pdf› Väisänen, O[tto] 1916: Vepsän mailla. Matkamuistoja I–V. – Uusi Suometar Matkan jälkeen 19.11., 21.11., 29.11., 7.12. ja ?.12.1916. [Saatavissa Helsingin yliopiston pääkirjastosta kopiosidoksena.] Kotiin palattuaan vastuullinen tutkija luonnollisesti järjestää ja arkistoi kenttäsaaliinsa. Aineistosta myöhemmin syntyviä julkaisuja on hyvä aikanaan viedä alkulähteilleen: näin tutkija pystyy antamaan jotakin takaisin puhujayhteisölle, jolta itse sai materiaalia tutkimuksiinsa. Ja­ kamalla julkaisuja kielenpuhujille voi paitsi kertoa tutkimustuloksis­ taan, myös osoittaa arvostustaan ja tukea kielellistä identiteettiä. Jois­ sain tapauksissa pystytään myös tuottamaan esimerkiksi yleistajuista oppimateriaalia puhujayhteisön aktiiviseen käyttöön. Näitä seikkoja on osattu alkaa ajatella vasta viime aikoina.

492 493 List of contributors

Balodis, Uldis Novak, Irina (University of Latvia (Karelian Research Centre of Livonian Institute) the Russian Academy of Sciences) ‹[email protected]› ‹[email protected]› Björklöf, Sofia Palander, Marjatta (University of Helsinki) (University of Eastern Finland) ‹[email protected]› ‹[email protected]› Ermus, Liis Pilvik, Maarja-Liisa (Institute of the (University of Tartu) Estonian Language) ‹[email protected]› ‹[email protected]› Puura, Ulriikka Jantunen, Santra (Institute for the Languages of Finland) (University of Helsinki) ‹[email protected]› ‹[email protected]› Riionheimo, Helka Kalvik, Mari-Liis (University of Eastern Finland) (Institute of the ‹[email protected]› Estonian Language) Rozhanskiy, Fedor ‹[email protected]› (University of Tartu & Karjalainen, Heini Institute for Linguistic Studies RAS) (University of Oulu) ‹[email protected]› ‹[email protected]› Ruutma, Mirjam Kemppanen, Hannu (University of Tartu) (University of Eastern Finland) ‹[email protected]› ‹[email protected]› Sarhimaa, Anneli Koivisto, Vesa (University of Mainz) (University of Eastern Finland) ‹[email protected]› ‹[email protected]› Söder, Torbjörn Laansalu, Tiina (Uppsala University) (Institute of the ‹[email protected]› Estonian Language) Suutari, Toni ‹[email protected]› (Institute for the Languages of Finland) Lindström, Liina ‹[email protected]› (University of Tartu) Tuisk, Tuuli ‹[email protected]› (University of Tartu) Lippus, Pärtel ‹[email protected]› (University of Tartu) Uiboaed, Kristel ‹[email protected]› (University of Tartu) Mäkisalo, Jukka ‹[email protected]› (University of Eastern Finland) Zaiceva, Nina ‹[email protected]› (Karelian Research Centre of Markus, Elena the Russian Academy of Sciences) (University of Tartu & ‹[email protected]› Institute of Linguistics RAS) ‹[email protected]› 494