sign in sign up
<<
Home , Danish language, Faroese language, Faroe Islanders

Farben: cyan magenta gelb schwarz

hjalmar p. petersen

The Dynamics of Faroese-Danish petersen petersen Contact The Dynamics of Faroese- Contact here are two official on the , Faroese and Danish; Faroese is the dominant language and Danish the first second language that children acquire. Contact Language Faroese-Danish of Dynamics The The question addressed in this book is what the exact transmission processes in this asymmetrical bilingual setting are. By combining van Coetsem’ notions of Recipient Language Agentivity and Source Language Agentivity with parts of Myers-Scotton’s and Jake’s frameworks, the author succeeds in explaining the lan- guage setting on the islands.

Universitätsverlag winter isbn 978-3-8253-5778-8 Heidelberg germanistische bibliothek

Herausgegeben von rolf bergmann und claudine moulin

Band 37 hjalmar p. petersen

The Dynamics of Faroese-Danish Language Contact

Universitätsverlag winter Heidelberg Bibliografische Information der Deutschen Nationalbibliothek Die Deutsche Nationalbibliothek verzeichnet diese Publikation in der Deutschen Nationalbibliografie; detaillierte bibliografische Daten sind im Internet über http://dnb.d-nb.de abrufbar.

Publiziert mit Unterstützung des Schweiterischen Nationalfonds zur Förderung der wissenschaftlichen Forschung.

Die vorliegende Arbeit wurde von der Philosophischen Fakultät der Universität Zürich im Wintersemester 2004 auf Antrag von Frau Prof. Dr. Elvira Glaser und Frau Prof. Dr. Claudine Moulin als Dissertation angenommen.

isbn 978-3-8253-5778-8

Dieses Werk einschließlich aller seiner Teile ist urheberrechtlich geschützt. Jede Verwertung außerhalb der engen Grenzen des Urheberrechtsgesetzes ist ohne Zustimmung des Verlages unzulässig und strafbar. Das gilt insbesondere für Vervielfältigungen, Über- setzungen, Mikroverfilmungen und die Einspeicherung und Verarbeitung in elektronischen Systemen. © 2o10 Universitätsverlag Winter GmbH Heidelberg Imprimé en Allemagne · Printed in Druck: Memminger MedienCentrum, 87700 Memmingen Gedruckt auf umweltfreundlichem, chlorfrei gebleichtem und alterungsbeständigem Papier

Den Verlag erreichen Sie im Internet unter: www.winter-verlag-hd.de Preface and acknowledgment

This book is part of the result of my research intoFaroese-Danish bilingualism and language contact.The research wasdone as part of theproject: K8 Variation in Bilingualism on the Faroe Islands,which was funded by the Deutsche Forschungsgemeinschaft,thatisThe German Research Council.The K8 project was oneofmany at the Research Centre 528: ,orSonderforschungsbereich538: Mehrsprachigkeit, as it is called in German. In short, we just callitthe SFB. Iwould firstand foremostlike to acknowledgeall of the informants that participated in the K8 project, andwho camefromdifferent parts of theFaroe Islands. Theseinformants spoke both Faroese andlater Faro-Danish,and gaveusvaluable data, which made it possible to study atopic that had never beenlooked at seriouslyinany depth before: bilingualism on the Faroe Islands. Iwould also like to thank . Braunmüllerand . Heycock for commentsonearlier drafts of this book,aswell as my colleague at theK8project, K. Kühl. Thanks to A. MacDougall for proof-reading themanuscript. Iwould then like to thankJ..Jónsson, Th.Eythórssonand Ei.Rögnvaldsson,aswellassome Icelandic friends fromway back, when Istudied in Reykjavík, for providing me with Icelandic examples. Thanks to . Ali-Laurilafor helping out withthe Finnish examples.Thanksalsotomycolleagues on the Faroe Islands at Fróðskaparsetur Føroya: A. Johansen,Ei. Weyhe, . íLon Jacobsen,K. Magnussen, M. Staksberg andZ.S.Hansen,for always beingwillingtoanswer different questionsand providingmewith copies of different articles that Icouldnot findelsewhere. J. íLon Jacobsen andZ.S.Hansen alsooffered to read through theFaroeseexamples, as did J. Thomsen.Thanks to the research assistants connectedwith theK8project in Hamburg andmycolleagues at the SFB, that is theResearch Centre 538:Multilingualism, or Sonderforschungsbereich Mehrsprachigkeit 538,and especiallytothose who offered with fruitful,constructivecomments to thepresentations thatwereconnected with the project. We had different workshops at theSFB,and it was duringone of these thatD. Winfordpresented van Coetsem’s model, theone Iuse in thisbook. It has proved to be a very usefulmodel thatmakes it possible formetocombine both what happens when the speak Faroeseand what happens whenthe same speakersspeak thespecial variantofDanishthatisusedand learned as afirstL2onthe islands,and which Ihave calledFaro-Danish in this bookand elsewhere.Thenthere arethe numerous peopleIhave asked for different judgments whendoing thegrammaticalityjudgmenttests as well as throughcorrespondence. These Iwould like to thank; no oneismentioned,and no one is forgotten. Iwould also like to thank . S. Jensenand . Theilgaardfor judgmentsonsome of theDAsentences.Theywerealways readytoanswerand didsoverycarefully.Iwould also like to thank Fjølrit, MentanargrunnurLandsinsand Vágakommuna forsupporting the printing.Having saidthis, all mistakes remain my own.

5

Contents

Preface andacknowledgment...... 5 Contents ...... 7 Abbreviations and symbols...... 11 1Introduction...... 15 2Ashortoverviewoflanguage contact andbilingualism ...... 19 3Anoverviewofthe history of the ...... 29 4Bilingualism on the FaroeIslandsand the social settings of Faroese...... 35 4.1. When didthe Faroe Islandersbecomebilingual?...... 35 4.2. When do Faroese children acquireDanish? ...... 40 4.3. Thesociallydominant language and thedomains of Faroeseand Danish ...... 40 4.4. The attitude towards Danish...... 42 4.5.Languageawarenessand purism ...... 44 5State-of-the-artreport ...... 49 6Methodology...... 55 7Some typological differences between Faroese andDanish...... 59 8Recipient language agentivity; Source language agentivityand neutralization...... 69 8.1. RL agentivity...... 75 8.2. SL agentivity ...... 79 8.3. Neutralization...... 80 8.4. Comparing with other models...... 85 9Recipientlanguageagentivity ...... 87 9.1. Phraseology. Cases of RL agentivity;multiple transfer...... 88 9.2. Lexicalborrowings ...... 90 9.3. The Borrowing Scale and Danish in Faroese ...... 91 9.3.1Borrowingofcompoundedverbs...... 97 9.4. Adaptationand imitation to Faroese morphology and ...... 99 9.5. ConversioninRecipient language agentivity...... 100 9.5.1. Deonticuse of mátti ‘may’...... 101 9.5.2. Future timereference...... 102 9.5.2.1. De-andative ...... 106 9.5.2.2. De-venitive...... 108 9.5.2.3. De-volitative...... 111 9.5.2.4. Blíva-‘become’future...... 114 9.5.2.5. Summarizing the future time references ...... 115 9.5.3. Pseudo-coordination...... 116 9.5.4. Thechange of argument structureinditransitive (triadic) verbs...... 120 9.5.5. From synthetic to analytic comparative andsuperlative...... 129 9.5.6. Existential sentences...... 134 9.5.7. between subjectand adjective-predicates...... 139 9.5.8. Adverb +verbinembeddedclauses ...... 142

7 9.5.9. WH-constructions...... 145 9.5.9.1. Frominterrogative adverbstostranded wh-construction...... 145 9.5.9.2. KIND-questions...... 150 9.5.10 The indefinite ...... 152 9.5.11 Head +modifier and modifier +head in possessiveconstructions...... 156 9.5.12 Circumpositions ...... 160 9.5.13 Gender strategies...... 161 9.5.14 Innovative imitation...... 165 9.6. Codeswitchingorimitation as inclusioninFaroese...... 174 9.7. Whydiffusion in Faroese?...... 178 10 Source language agentivity: Faro-Danish...... 181 10.1. Singlyoccurringwords, embedded islands and nonceborrowings...... 182 10.2.Intra-sentential codeswitchingorimposition as inclusion [÷ nativized] ...... 183 10.2.1.Intra-sentential codeswitching: nouns...... 183 10.2.2. Self-repair:nouns...... 185 10.2.3.Intra-sentential codeswitching: verbs ...... 187 10.2.4. Self repair:verbs ...... 188 10.2.5.Intra-sentential codeswitching: Other content morphemes...... 189 10.2.6.Intra-sentential codeswitching: early system morphemes ...... 193 10.2.6.1. Determiners...... 195 10.3. Conjunctions...... 196 10.4. Embedded islands ...... 197 10.4.1 Embedded islands: self-repair...... 199 10.5. Summarizing the codeswitchingand embedded islanddata...... 200 10.6.Nonce borrowingsorimposition as integration [+ nativized]...... 200 10.6.1.Nonce borrowings: nouns...... 201 10.6.2.Nonce borrowings, nouns: self-repair...... 203 10.6.3. Nonce borrowings: verbs...... 203 10.6.4. Nonce borrowings: adjectives ...... 205 10.6.5.Nonce borrowings: adverbs...... 206 10.6.6.Adjustment rulesinnonce borrowings...... 207 10.7.Conversion as imposition in SL agentivity...... 209 10.7.1.Semantic conversion ...... 209 10.7.2. Syntactic conversion ...... 213 10.7.3. Collocates ...... 223 10.7.4.Morphologicalconversion ...... 225 10.8.GenderinFaro-Danish ...... 228 10.9. Mixed compounds...... 235 11. Imposition as approximation: Vowelvalues in Faro-Danish...... 237 12. Increasing English influence on Faroese ...... 239 13. Conclusion...... 247 References...... 249 Appendix...... 259 Some important numbers ...... 259 Theletter to HeðinBrú (around 1930)...... 259 Faro-Danishdata...... 260

8 Codeswitching,nouns ...... 260 Self-repair, nouns...... 263 Codeswitching, verbs...... 265 Self-repair, verbs...... 266 Codeswitching,othercontent morphemes ...... 266 Other morphemes:...... 270 Early systemmorphemes...... 272 Codeswitching, conjunction ...... 273 Embedded Islands ...... 274 Nonce borrowings ...... 278 Nonce borrowings, nouns: self repair...... 282 Nonce borrowings, verbs...... 282 Nonce borrowings, verbs: self-repair...... 284 Nonce borrowings, adjectives, pastparticiples...... 284 Nonceborrowings, adverbs ...... 286 Accommodation rules in nonceborrowings...... 287 Semantic conversion:...... 288 Gender...... 308 Mixed compounds ...... 312 ...... 314

9

Abbreviations and symbols

>becomes

11 DUR Durative EI Embedded Islands. EL Embedded Language. EN English EV2 Embedded V2 EXPL Expletive -.Feminine FA Faroese FAR-DANFaro-Danish FTRFuturetime reference 4-M modelThe four morpheme model GGrammar GE German GD Gøtudanskt GMCGermanic Gothic HHigh variant IO IndirectObject K8 The Hamburger Database on Faroese-Danishbilingualism. It consists of informal interviews with threegenerations,16-20, 40-50and 70+.The informants come fromdifferentparts of the Faroe Islands. IC Icelandic IE Indo-European IMPImperative IMPERF Imperfect INFInfinitive INFM Infinitivemarker IP Inflectional Phrase IScInsularScandinavian LLow variant L1 First language L2 Second language LA Latin LG MED-PASS Medio-Passive MidScMid Scandinavian MSc Mainland Scandinavian -m.Masculine ML Matrix Language NEUT Neutralization NO-AGRNoagreement Nom. Nominative

12 . Noun -n. Neuter NN NewNorwegian NP NounPhrase OD Old Danish ODS Ordbog over detdanskesprog =the largest Danish Dictionary availableat: http://ordnet.dk/ods/. OE OIr Old Irish ON OldNorse P. Person Part. Partitive Pl.Plural PP Prepositional Phrase PART Partitive PASS Passive PAST Past tense PH Phonology POS Positive PPPPast PP Preposition PRES Presenttense PROGProgressive PRT Particle redReduction REFL Reflexive RL Recipient (orreceiving) Language RLAg Recipient Language Agentivity RU Russian Sg Singular SL Source Language SLAg.Source Language Agentivity SLA Second LanguageAcquisition SP Spanish Standard DA Standard Danish SW Swedish SUFF Suffix SUP Supine SUPERL Superlative TECTransitive ExpletiveConstruction

TL2 Target Language 2 TRS Transitive suffix USubsetofV VVocabulary

13 Vf Functors V1 Primary vocabulary (contentives) V2 Secondary vocabulary (contentives) v. Verb VP Verb Phrase WFR Word FormationRules

14 1Introduction

Onecould well claim that theFaroe Islands today is one of the countries with the most consequent bilingualism, as thegrownupstoa high degree mastertwo official languages as spokenaswellas written medium. (Hagström 1984: 240)

The question to be addressed in this book is:what happens in the speech of asymmetrical bilinguals when language A, the dominantlanguage, andlanguage , the embedded language, arecloselyrelated? The data arefromFaroese,henceforthFA, and Faro-Danish, henceforth FAR-DAN, which is the special variant of DA that is used on the Faroe Islands as afirstsecond language.Inthisbilingual language situation,FAisalwaysthe dominant language,aswill become clear later when the notionofdominance is addressed. FA is aNorth Germaniclanguage. It is usually groupedtogether with Icelandic,IC, as an InsularScandinavianlanguage, ISc,asithas manyofthe same syntacticfeaturesasIC. Thisgroupingwill be addressed again in Ch.7.Some typological differencesbetween Faroese and Danish,but it should alreadybepointed out herethatthe division between IScand Mainland Scandinavian,MSc,cannotbeupheld, since FA allows both ISc syntax,and,as the resultoflanguage contact, MScsyntax (Barnes &Weyhe 1994). In addition to MSc thereare other loans fromDA, especially lexicalones, butalsoderivationalsuffixes,aswill be shownin9.3. The frameworkused is van Coetsem’s (2000) theory of thetransmission processin languagecontact, wherehefocuses on thenature of the(psycho-)linguistic processesof changethatreside in theminds of individual bilinguals.This Ihavecombined,partly, see especiallych. 10 on SL agentivity, withMyers-Scotton’s andJake’s4-M model, as this model predicts what can be transferred from thedominantlanguage, FA, to the non-dominant language,which is FAR-DAN in ourcontext. Iwillnot use theirdistinction between Matrixand Embedded Language, forreasonsthatwillbecome clear later. In asymmetrical bilingualismsituations as the one underconsideration, the agent speakersofthe dominant language, that is FA, imitate andborrow the unstableparts of grammar, secondaryvocabulary,syntax, modal anddiscourse markersfromthe source language, DA. The result is acomplication of FA, as will be shown in detail in ch.9. Complication is thentobeunderstoodasasystem “that needs [a] longerdescription” (Dahl 2009:42). In FAR-DAN, the agentspeakersimposethe stable parts,mainly the articulatory habits of the sourcelanguage, FA,ontothe receiving language, FAR-DAN, but also some bound morphemeslikeinfinitive,verb+particle, plural andpastparticiples. The reason is that

15 these morphemes areearly in thesense of Myers-Scotton &Jake(2000), see ch. 10 for details. Thereisachapter specially devotedtothe articulation of FAR-DAN,whereitis shown that agent FA speakersusuallyaccommodate their pronunciation to DA,but withoutreaching aperfect DA pronunciation. We measured the vowels -e, -ø and - and found that thepronunciationinFAR-DANisanintermediate one, notpurelyFA, andnot purely DA (Petersen&Rakow 2010),for details see ch. 11. FAR-DAN is acase of Source language agentivity, SLAg,withthe imposition of FA linguistic material onto the receivinglanguage, whichisFAR-DAN.The result of SLAg is reduction, at least from thepoint of view of Faroese agent speakers.Reduction means that e.g. theFAagent speakersdonot speak FAR-DAN with the marked , the stød, and they do notuse many distinctions,whichare otherwise foundinDA, seech. 11. Fromtimetotime,theymake theirprocessing easier by using e.g. FA syntaxand even morphology, as will become clear in ch.10. It shouldbenoted that there is no reduction of standardDA. The Faroeseagent speaker of FAR-DAN simplifies his/her L2 for himself/herself.There is thusnochangeinthe standardDAsystem. Thesyntaxshows median stability in both RLAg and SLAg,where themain result of the transferenceisatwofoldconversion-process, with thecomplicationofFAand the reduction of FAR-DAN. An exampleofcomplication in RLAg is whenagent Faroesespeakersimitate different syntactic structuresofDA, as with futuretimereferencessuchaskoma ‘tocome’ as hann kemuratsiga ‘heisgoing to say’ (< DA hannkommertil at sige ‘he is going to say’),using this in FA side by side with the ‘proper’ future time reference construction fara at ‘go to’, hann fer at siga ‘he is goingtosay.’ This is what is meantbycomplication of the receiving language.Ittakes longertodescribethe standardFAsystemthanthe IC one when speaking about this particularconstruction. The same speakers occasionallyimposeFAsyntax onto FAR-DAN,like the supine jeg har prøvet at arbejdet ‘I have triedtowork’ instead of jeg harprøvet at arbejde ‘I have tried to work.’ In doingso, s/hesimplifies his/her FAR-DAN by covertlyusing FA syntax. The morphology andthe mental lexicon come from FA andDA, as does thesyntax, while the phonological system comes solely from FA,althoughwehave found an intermediate pronunciation in FAR-DAN (Petersen&Rakow2010). Regarding thesyntax, it is sometimesdifficult to saywhether the speakerusesFAor DA syntaxinRLAg,sincethe languages arelinearly equalinmanycases.But thenthereare otherreasons for presuming that there are twosyntacticsystems at play. If thiswerenot the case, oneshould findsuch specifically FA syntactic constructions like the hjá-possessive,double-definiteness, supine-attraction and FA genderinFAR-DAN,and thisisgenerallynot thecase. The hjá-possessive is seen in e.g. báturin hjá manninum,which lit. means: boat-the(N) with man-the(D)=‘theman’s boat.’ Supine attractionisseenineg havi prøvað at arbeitt ‘I have triedtowork’,where thesupine spreads from prøvað ‘tried’to arbeitt ‘work.’ Doubledefiniteness is the rule in FA, as in tann stóri maðurin (lit.: thebig man-the) =‘the bigman’opposite to DA den store mand ‘thebig man.’ If thespeakersused FA syntax whenspeakingFAR-DAN, we shouldexpect to find many of these features in FAR-DAN, but as Iwillshow, thisisnot the case, andthis shows thatthe agentspeakers use two syntacticsystems in their language processing. Reference is madetothe K8 database.This is the Hamburg Databaseon Faroese-Danish bilingualism. The work was doneinthe context of SFB 538, that is the

16 SonderforschungsbereichMehrsprachigkeit 538 or Research Center 538: Multilingualism in Hamburg, Germany, where manydifferentprojects on multilingualism are conducted. The K8 project consists of informal interviewswith three generations, 16-20, 40-50 and 70+. The informantscamefromdifferentpartsofthe FaroeIslands,and they spoke about different topics such as children'sgames,books they have readand so on.For details see Ch.6. Methodology. Of these, Ihaveonlyincludedthe youngest andthe oldest generation intothe present work primarilyfor time reasons. Thebook is organizedasfollows. First there is ashort overviewonlanguage contact andbilingualism.Thischapter mayseem trivial to readers confidentwith the topic anditis not intendedfor the specialist but ratherasanoverview forreaderswho arenot acquainted with language contact andbilingualism in general.Itisfollowed by an overview of the history of theFAlanguage, achapterthat is intended forreaders not familiar with FA so that s/he can obtainapictureofthe languagesituation. Chapter4offers amore detailed description of bilingualism on the FaroeIslands, whereitisshown thatthe process towardsthe high proficiency of FAR-DANfound on the Faroe Islands hasbeenagradual one.Inthatsamechapter, thereare afew words about thedomainofeach language, a section on when children acquireFAR-DAN,and onesection about the attitude towards DA. Here it becomes clearthatthe FaroeIslandsare aso-called open linguisticsociety with regard to borrowings, althoughthe purist movementmay trytodelay the changes in progress, c.f. 4.5. Languageawareness andpurism.Chapter 5isanovel reportshowingthat amazinglylittlescientificworkhas been done on Faroese-Danish language contact,even thoughithas existedfor centuries.This chapter is followed by themethodology chapter and achapter on the typological differences betweenFAand DA,afterwhich the theoretical frameworkisintroducedinchapter 8. Theactual data arethenpresented in chapters9and 10. In chapter 11, Ishowhow the agent FA speaker pronouncethe mid vowels -e, -ø and -o.Ch. 12 is about the increasing influence of English,EN, on FA, showing thatmanyyoung speakersare notbilingual, butrather trilingual,and in ch. 13 I have aconclusion andadiscussion. When on theFaroe Islands, youmay hear the term Gøtudanskt, GD, (lit.: street-Danish), or Danishfromthe village Gøta.GDisusedtorefer to DA,which is heavily influenced especially by FA pronunciation, anditissaidtooriginate from ateacher fromthe village of Gøta in (Poulsen 1993). He lived from1850to1930, and spoke DA withalot of Faroeisms. One example fromPoulsen (1993)is:

GØTUDANSKT: (1) a. De storefor flesen,dekan brække traver.

FAROESE: b. Teir stóru fyriflesini,teir kunnu bróta tráur. the big(N) for -the(D), they(N)can breakfishing rods(A) ‘The bigones[coalfish] outsidethe skerry canbreak fishing rods‘.

It is saidthatthe teacher spoke thiskindofDAeverywherehewent, and thisiswhy the first part of thecompound hasthe place name Gøta. Maybe thisisthe case, maybe not; still, it is agood story.

17

2Ashort overview of language contactand bilingualism

This chapter is trivialfor many readersacquainted withlanguagecontact andbilingualism. It is intendedfor readers who arenot allthatfamiliar with thetopic under consideration. No language contact situationiscontext free. Thenotionofcontext is not unimportant, as it is necessary to have agoodunderstanding of the history andsocial dynamics that exist between the languagesocieties in contact (Thomason 2001: 78, Winford 2003:28, Aikhenvald 2008:36), as extra-linguistic matters such as social factors, the intensityofcontact, speaker’s attitude towards the source language, purismand the lengthofcontact play astrong roleinexplainingwhy closely relatedlanguages like IC and FA, forexample, develop as differently as they havedone anddo. It is notpossibletopredict when acontact-induced change can or will happen or what kind of change will occur, as this is governed by probabilities, not possibilities, and because speakersare unpredictable, as pointed out by Thomason (2001: 61). Language contactoccursnaturally as areflexofhumans’social nature (van Coetsem 2000: 31), andone can claim, as does Schuchardt, that “ gibtkeinevölligungemischte Sprache” [there arenocompletely unmixed languages] (Schuchardt1884: 5). Differentfactorspromote bilingualism/multilingualism. One is aclose proximity between the languages in contact,asatthe Danish-German border, wheremanyspeak German, GE, and thespecial variant of DA which is calledDanish-German (Fredsted 1998, Kühl 2008). Another example from NorthGermanyisthe contact between Frisian andHighGermanand previously Low German, and thereare thousands of othersimilar examples, too many to mentionhere.Displacementpromotes bilingualism,e.g. the many immigrants to Europeand USA. Thethirdfactor that promotes bilingualism is colonialism, as on the FaroeIslands and , whereDAbecame theHigh, , variantafter the . Bloomfied (1933) wrote that bilingualism is the resultofthe additionofaperfectly learned foreign language to one’s own,but as pointedout by many others, e.g. Myers-Scotton (2006: 36), the criterion of speaking of twoormore languages with anative- like ability would rule out most bilinguals,asone of the twolanguagesisalways or will always be thedominantlanguage, even in balanced bilinguals, that is children wholearn languages Aand Bsimultaneously between theagesof0and 3. Childrenwho learn their secondlanguage between 3-4upuntil to puberty arechild secondlanguage learners, while those who learntheir L2 after puberty areadultL2learners (Klein 1986:15).This means thatthe FaroeIslanders arechild L2 learners, seech. 4. The question is whether abalancedspeaker can andwillremainsuch,ass/hewill almost certainlyuse onelanguageasthe dominant one later in life duetodifferentsocial circumstancessuchaswhere they endupinlife and where and in which domain they use each language,examples of domains beingunderstood as family,friends, religion, education andwork (Fishman 1972). This is illustrated verynicely by Hartmann Jacobsen’s (1984) MA thesis, whereshe describes how her twosons graduallybecome more andmore

19 dominantinFAafterinitiallybeing dominant in DA,simply becauseFAbecame the dominantlanguage after they moved from to theFaroeIslands. Abilingual maynot have thesamecommandoverthe four basiclanguage skills: listening, speaking, reading andwriting. Thebilingual may, for example, understand and speak language Bfairly well but not write it, or s/he may have goodcommand of the morphology andsyntax of B, but not of thephonology. Even if s/heisgrammatically competent in B, his or hercommunicative competencemay be halting. Thereisafluencycontinuuminlearning asecondlanguage. First, one has to take into consideration the age of acquisition,where pubertyisagoodcandidate foracut-offpoint. Klein (1986: 15) distinguishesbetween three acquisition periods.The first language acquisition period extendedbetween the ages of 0-3. This correspondstoMeisel’s(2007: 14ff)2L1 (two first languages), but Kleindiffers fromMeiselinhaving achild L2, thatis child secondlanguage, cL2, between the ages of 3-4and puberty, after which he has second language acquisition,SLA. Iwill adopt this classification in thisbook, simply because Meisel’sapproachimplies that e.g. aFaroese learner who starts withGerman, GE,atthe age of 14 should havethe same knowledge of GE as of DA,which s/he formallystarts withatthe ageof10/11.Thisisnot the case. Everyone speaks DA with ahigh level of proficiency, andthisisdefinitely notthe case with GE on the Faroe Islands,which some people begin speaking afterthe age of 14. Exposure obviously also playsarole here, as do language useand theneedtouse languageB. In addition to age, one must take into consideration thedomains in which the bilingual uses his/her languages.These are family,friends, religion, education andwork (Fishman 1972). Thenthere is thequestionofhow language Bislearned.One point is the aforementioned age;another,whether thereis/hasbeenany formalinstructione.g.ina class-roomsetting, howthis instruction hasbeen,and to whatextent thechild (or the adult) is exposedtothe second language.Another possibility in SLA is that the speaker learns language Bsimply by interacting with native speakers, andmay in this case reach a high levelofproficiency, especially withregard to speakingand pragmatics, but notalways in termsofgrammatical correctness. In the study of bilingualism, one language is alwaysthe dominant language;possible exceptions arespeakers whoare so-called balanced bilinguals,but as thecircumstances thesechildren live underchangelater in life, the result generallybeing that one of their languages will be the dominantone, as mentioned.Iunderstand dominance as reportedin Weinreich (1953: 75), where he saysthat“thedominantlanguage is the language the speaker hasgreater proficiency in.” In theconcrete setting, FA is alwaysthe dominant language,both in Recipient languageagentivity,where the agent speakersspeak FA, and Source languageagentivity,where theagentspeakers speakFaro-Danish,FAR-DAN. Iwill follow vanCoetsem(2000) and use the letters Aand Band underscoring, e.g. A, to indicate whichlanguage is dominant. Iwill also follow van Coetsemand use Recipient language, RL, andsource language, SL,and underscoring to show dominance.These terms willbeexplored andexplained in detail in ch.8. What may or can be transferredinlanguage contact? AccordingtoThomason (2001), everything. In away sheisright, butitisimportant to make adistinction between borrowing/imitation and what Thomason (2001: 75) calls TL2 (Target Language 2),thatisSource language agentivity.

20 In the case of aTL2or SLAg,learnerstransfer some features of theirnativelanguage into their version of theTL2.The difference between borrowingand TL2 is the same as the distinctionbetween Recipientlanguageagentivity and Source language agentivity in van Coetsem (2000). It is of some importance to keep these termsand distinctions in mind, as the mechanisms in RL and SL agentivityare complimentary. It is thus not the casethat everything can be transferred in anyway in language contact, as Iwillshowindetail in ch. 9and ch.10. Themainpoint is that whatistransferredfromDAtoFAiscomplimentary to what is transferred from FA to FAR-DAN by thesameagentspeakers. Atypical caseofRLagentivity is when Faroese agentspeakersborrow/imitateDA linguistic material and adopt it into FA.Inthis process of borrowing/imitation in the dominantlanguage, thestableparts of grammar,phonology and morphology, are left basically unaffected, while theunstable parts of thelanguage, meaning the non-basic vocabulary, and the somewhat stable parts of the language, includingits syntax, andmodal and discoursemarkers,are transferredfromthe source language, DA,tothe recipient language, FA.The opposite situation is TL2,Sourcelanguageagentivity, whereagent speakersofFAimpose FA linguisticmaterial onto FAR-DAN.FAR-DANis, as mentioned earlier, thespecial variant of DA that Faroemen use. In SL agentivity, the stable parts of the source language,FA, are imposed onto therecipient language, FAR-DAN. The resultisaDA thatisspoken with aFAaccent, an approximationofvowel values, seech. 11,withsomeFAmorphological features, butalmostnoFAvocabulary, with the exception of afew nonce borrowingsand code switches.Inaddition to this, the median parts of grammar, syntax,modal- and discourse markersare also imposed in SL agentivity, just as in RL agentivity. One principle in acontact situation,accordingtoAikhenvald (2008:28),is: “beas iconic withyourneighborasyou can.” In the concretecontactsituation,this manifests itself with thecomplicationofFA, as when the agent speaker borrows e.g.the argument structureofthe ditransitiveverb at give ‘to give’ into FA andusesaprepositional phrase insteadofanIO, DO construction. As such,s/heremains as faithful to DA as s/he can. Iconicity, so to say,inSLAgisobservedin, for example,the lack of stød.The agent FA speakers usepartsofthe FA consonant system, simplifyFAR-DAN andget rid of the markedglottal stop, which is quitedifficulttomaster.The receiving system thus becomes more natural, meaningless marked. An example of iconicity as acomplication/enrichment is e.g. the increasing useofprepositional phrases after at geva ‘to give’ instead of an indirect object,IO, or direct object, DO. Icallthiscomplicationorenrichment, as FA now allows both DO +PP+DOand IO, DO, as opposedtoIC, which hasonlyIO, DO after gefa ‘to give.’ The newemergingstructuresinRLAgmay be ‘spoiled’ by language planners/purists, who encouragespeakers to avoid the ‘new’ structures.These extra linguistic factors do play arole in acontact situation,asthey mayslowdownachange or preventitfrom happening. The problemindiscussing some of the FA data in ch.9isdefinitivelydetermining whichchanges are causedbyinherenttendencies, especiallydrift,and which changes are due contact. Another question is if extra-linguistic factors such as ageand sex play anyrole at all. Thequestion about language changeisofcourse not new, andithas occupied linguists for years. TheNeogrammarians claimed in the1870s that allsound changes are regular and

21 internallymotivated.This viewhas been extremelyinfluential, although therehavebeen others, such as that of Schuchardt,who wrote that:

Ichhabe behauptet,daß unterallen Fragenmit welchen dieheutigeSprachwissenschaftzu tun hat, keinevon größerer Wichtigkeit istals dieder Sprachmischung.

[‘I have maintained thatofall the questionscontemporary linguistics musttackle,none is of greaterimportance thanthatoflanguage mixing’].

(Schuchardt 1884: 3)

The view that all linguisticchange is internally based was notonlyaviewthatwas upheld by theNeogrammarians only, but was the dominant, perhapsonlyview of theStructuralists andisthe main assumption among Generative grammarians,all of whichdonot work with language variation and the individual speaker. Variationist work on language change, starting with especiallyWeinreich(1953)and Haugen’s study (1950) on language contact andfollowed by theinfluential works of Labov in the sixties, have focused on extra-linguistic andexternallinguisticfactorsastriggersof change. Labov’s research is areactiontoChomskyan formalism, and changes areusually explained as theresult of variation in speech between sexes, generations or classes, for example. It is importanttodifferentiate betweeninternal changes, external changes due to contact, andchanges thathave comeaboutasthe resultofextra-linguisticfactors such as age,sex,,attitude towardsasource language,purismand so forth. Contactasa trigger forachange was seen by linguistssuchasWeinreich (1953) to play asignificantrole, and research in contactlinguistics since then has shownthat contactinduced changesare by no means exceptional or rare, and are even very common (Thomason &Kaufman 1988, Thomason 2001 and referencestherein). Themainfocus of thisbookiscontact induced changes, as e.g. when aFAspeaker starts to usethe verb at koma ‘to come’for afuturetimereference,FTR,asin(2.1a)and (2.1b), wherethe latter is prediction-based. Theseare based on the DA construction, and a grammaticalization of koma ‘tocome’ is currently occurring in FA, as illustrated with the sentences in (2.1), which showthatthe verb koma ‘to come’ in thisspecific context haslost its lexical meaning. It is important to recognizecontact as thetrigger for thenew FTRin FA.

FAROESE: (2.1) a. Eg komiatsiga henni tað ímorgin. I(N) come(FTR) to tell(INF) her(D) it(A) in morning(A) ‘I am goingtotell herittomorrow.’ [DA base: jegkommer tilatsige ‘I come(FTR) to(PRT) to tell(INF)]

b. Tað kemur at regnaímorgin. it(N) comes(FTR)torain(INF)inmorning(A) ‘It is going to raintomorrow.’

22 The DA base is detkommertil at regne (lit.: it(N)comes(FTR) to(PRT) to(INFM) rain(INF)))))) ‘itisgoing to rain’. In geva klæðinitil Frelsunarherin ‘give theclothes to the Salvation Army’, which is emerginginFAatthe costofIO, DO,and the analytic comparative meira sjúkur ‘more sick’, which is spreadingatthe costofthe synthetic comparative sjúkari ‘sicker’, both are examples of whatSapir (1921:171-172) called adrift, which he defines(ibid)ascases in which related languages passthroughthe same or similar phases. For more on this problematictermand adiscussionthereof seeMalkien(1981). This is exactly what makes it sometimes difficulttowork with , GMC, as they generally show adrifttowardsanalyticity.Insaying difficult,Imeanthatthe very dichotomybetween languageinternal andlanguage external inducedchanges is not at all clear-cut,as“several factors areusually at workinorchestrating achange. To putit simply: in each case, thepossibility of multiple motivations muddiesthe waters” (Aikhenvald 2008: 9). By multiple motivations shemeans, Iguess, internal, external andextra-linguistic factors, andshe points out that “the more facilitatingfactors areatwork, the likelier is a form, or apattern to be established in acontact-affected language.” (Aikhenvald 2008:48); thisisher Mutual ReinforcementPrinciple. The phrasing “multiple motivationsmuddiesthe waters” is not unimportant in the present context, as Iwanttoavoid getting into an ‘either-or‘ mentality(Farrar &Jones 2002: 3) when presenting anddiscussing some of my data. That is, either thechangeis internal or it is external, andIcertainlywant to avoid thenot uncommonly expressedview that‘If-in-doubt-do-without’, meaningthe assumption that all changescan be explained as internal,and if we cannot identify these, then we must as alastresort try to establish some external causes or extra-linguisticfactorsfor thechanges. Generally, however,itisbetterto do withoutany external andextra linguisticexplanation(Farrar &Jones 2002: 4). Itakethe position that an external explanationalone is appropriate for aparticular change or in conjunctionwith an internalmotivation,see also Thomason&Kaufmann (1988:63). It is important to that theinternalchanges such as thechangeinargument structureofgeva ‘to give’fromIO, DO to DO +PP+DO andthe analyticcomparative are changes thatmightoriginally have beeninternally motivated. Theyhavebeenspedup, however, by language contact. As pointed outbyDahl(2001), the chancethatacertain morpheme or construction in alanguage will undergo aparticular kind of grammaticalization is rather small on the whole, but givenacontact situation,the probabilityofachangeincreases dramatically when aneighboring language is undergoing or has undergone theprocess in question. It is notamust, although it might seem so,totreatinternal andexternalchangesas mutually exclusive. Rather, andthisisthe view Itakeinthisbook, an internal change like the drifttowardsanalyticity in FA, is sped up by language contact. Additionally, it is historically incorrect,evennaïve, to assumethatthe hugeinfluence DA has on FA, e.g. the borrowing of ,suffixes,syntactic constructions andalot of vocabulary, does not haveany effect on the drift towards analyticity, even though the same drift happenedin DA ande.g. EN centuriesago. DA hasundergone achangetowardsanalyticity, andFAispresently changing in the same direction,while IC is notshowing thesame level of change.

23 Why is this so? Iwilloften useIC, and of courseOld Norse, ON, as akindofcontrol language(s) whendiscussingthe data. Thisisjustified,asIChas as arulepreserved ‘older’ syntactic structures,suchasthe mentioned general lack of drift towards analyticity with averbsuch as IC gefa ‘togive.’The reasonfor the change in FA is simply that FA has had intense contact withDA(andNorwegian, NO)for centuries (Sandøy 2005), andspeakersgenerally have apositive attitudetowardsDA. Thereis, in addition to this, alackofastrong purist tradition on theFaroe Islands as opposed to (Kvaran2007), andFAdoes nothave such alarge text-corpus as Iceland has.Mosttexts,e.g.novels and so on, are in DA, the dominantintruder, andthese facts facilitate the changes in FA. In addition to this, the Faroe Islanders have amuch higher proficiency in DA than theIcelanders. This is very nicelyillustrated by the following examplefrom the Danish newspaper Politiken,13/01/2010, accessed on thesame . TheIcelandicwriterAuður Jónsdóttir tells howshe came to Denmark in 2003. She stayed there until2006. Upon her arrival, she orderedsomething at ahot dog stand in broken DA,and thewoman in thehot dogstand suggestedthat she should go back to where shecamefrom. The same is extremelyunlikely to happen to aFaroe Islander, as s/he has amuch higher profiency in DA, as do older . J. G. Jónsson respondsinane-mailabout theIcelandersproficiency in DA by sayingthat: “Ég er ekki hissa áþví. Ungir Íslendingareru almenntmjöglélegir ídönsku og alls ekki jafngóðir og Færeyingar”, whichmeans: ‘Thatdoesnot surprise me at all.Young Icelanders’ proficiency in DA is bad, andisnot at all as good as theFaroe Islanders’. Anotherexample is from thenewspaperSosialurininJuly2010, wheretheywriteabout an Icelander,who hastostand court on theFaroe Islands. The language spoken in court is FA or DA,and the Icelander does not understand neither. Theextra-linguisticsettings of FA andICare nearlyasdifferent as can be,and my claim is thatthisinfluencesthe direction in whichthe languagesmove. Perhaps the increasingENbilingualism on IC will resultinsimilar changestothosethat FA has undergoneand is undergoing. Alanguagechangemight be complex,aspointed out by Aikhenvald,and it is important to keep in mindthatinternal, external andextra-linguistic factors allhave to be considered, and that there is often acomplexinterplay betweenthese (Farrar &Jones 2002:1,Papazian 2007:161). But how can one tell whether achange in FA has comeabout as theresult of DA influence or an inner change? Drift could apply to “a triad of resemblesinlanguagegrowth” (Malkien1981: 566):

(1)Those resultingfromcommondescendent of twoormore given languagesand traceable to an earlier common stage (recorded or reconstructed). (2) Those best explainedasinstances of independent if paralleldevelopment within the same , and (3) Those,obviously of atypologicalorder, whichcut across geneticallyunrelated languages (without necessarily amounting to universals).

Note that point(2) is problematicasitstands, as it might be difficult to sayfor sure whether achangeisacaseofindependent or of parallel development withinthe same language family. Is achangetoaprepositional phrase like geva klæðini tilFrelsunarherin ‘give

24 the clothes to theSalvation Army’insteadofIO, DO or meirasjúkur ‘moresick’ insteadof sjúkari ‘sicker’the resultofanindependent drift towards analyticity in FA and DA only? Malkiel would treat thesechanges as theresult of drift, whereas Ithink thatcontact is accelerating the change,whichmight originally have been triggered by an inner change. It is often easytocite foreign influencewhenfocusing on . The word argi/ærgi ‘summerpasture’, which is used in place names, is found accordingtoMagnússon (1998) in FA andIC, andisaloanfromOld Irish, OIr, airghe.Asitisnot found in GMC, it is obviously aloan in FA.Another example is the FA phrase taðerótti ámær (lit.: there(N) is fear(N)onme(D))))=‘Iamafraid.’ This collocatedoesnot exist in any of theother North GMClanguages,but it does existinOIr tá eaglaorm (lit.: is fear on me) =‘Iamafraid’,and it is argued in Barnes &Weyhe(1994: 217)that it is borrowed fromCeltic. But things becomecomplicated when investigating structural features,especially syntactic transfer, as it shouldalwaysbeconsideredthat an internal motivation combines with an external motivation to produce achange. Thus it may be difficult to determine whether contact is responsible for thechange or not. One way of telling is to look at the language system as awhole,not just bitsofit, as a language is asystem with manysubsystems. In ch.9,section. 9.3.,Iintroduce the Borrowing Scale and illustrate it with examples of DA borrowings into FA which are not only lexical, but also phonological, morphological and later in ch. 9also syntacticborrowings. The point in bringing it up here is that as DA influence is deeplyembeddedinotherparts of grammar, vocabulary andeven language planning,contact is agoodcandidate fortriggering and/or speeding up the changesI discuss in ch.9.Thismeans that we cannot‘do without’ contact, contrarytoe.g. structuralist andgenerativists belief. Anotherway of seeingwhether achangehas come about as the resultofcontact is to keep in mindthatthe structuralfeatures must be shared by both thereceivinglanguageand the source language, as pointed out by Thomason (2001: 93). The shared linguisticmaterial doesnot need to be identicalinall respects,and it willoften not be, simply becauseagent speakersare innovative,asnumerous examples fromlanguage contact studies show (Matras, . 2009). Iwill give examplesofinnovative replication particularlyinch. 9, section 9.5.14. Innovativeimitation,where Imention exampleslike hvaðani frá (lit.: whence from) =‘where from’and meira sjúkari (lit.:moresicker) =‘moresick.’ Both structures show thatlanguage replication, as it is calledinHeine&Kuteva (2005), is notpure copying. The comparisonwith IC andONdoeshelp, as theformerhas nothad the same intensecontact with DA as FA has. As aconstructionlikethe de-venitive kemur at siga ‘is going to say’ is found in FA and DA but notinICorON, it is fairly certainthatweare dealingwith aDAloaninFA. There arecaseswherethe shared materialdoesnot even exist in DA, or IC for that matter,like theroot-imperative ikki tak! (lit.:not take(IMP))=‘don’ take!’(Petersen 2008c). This sentence structure is only acceptedbythe youngest speakers, and is presently spreadinginFA. We should be dealing withapurelyinner change, which is presently happening in FA, but thenone has to consider theincreasinginfluence of EN on FA;see ch. 12,where Iput forwardthe opinion that the root-imperative ikki tak ‘don’t take’ has come about as an innovativeimitation of theENimperative. The imitation is innovative, as FA does nothaveany do-support.

25 The forth point that must be taken into consideration in order to determine whether linguistic material has come aboutasthe result of language contact is that “theproposed interference features –were NOT presentinthe receiving languagebeforeitcameinto close contactwiththe source language” (Thomason 2001: 93-94). The aforementioned de-venitive construction is acaseinpoint, but thereare also other structures in which pre-existingstructures facilitate thechange. In the previously mentioned changeofthe argument structure of at geva ‘to give’,which now takes IO, DO andDO+PP+DO, and the changefromasyntheticcomparativeand superlativeto analytic comparativeand superlative, pre-existing structures within FA facilitate the changes. That is,DO+PP +DOexistsafter e.g. at senda ‘to send’ in FA (and IC), and analyticcomparatives arepresent with indeclinable adjectives like ótolandi ‘intolerable’ and compoundedadjectiveslike framsíggin ‘visionary.’ Thefifth pointthatThomason(2001: 94) mentions in explaining how to tell whether contactinducedchange has occurred is that we must provethatthe shared features were present in thesourcelanguage,beforetheycameintocontact with thereceivinglanguage. Butwhatcan be borrowedinlanguagecontact? Thomason(2001) says everything, as mentioned, and this is truewhen we look at the different parts of grammar.Phonology is borrowed, as well as morphology,syntax, modal and discourse markers andevenpragmatics. In SL agentivity, the speakersofthe source language impose especiallythe phonological habitsofthe SL onto theRL, and thisiswhy SLAlearners speakthe RL with aforeign accent. Thereasonisthat phonetic features are acquiredbetween theagesof0and 7. Myers-Scotton(2006: 341) mentions studies of immigrants who learnanL2uponthe arrival to aforeign country.Oftheseonlythe youngest children, those up to about age7, achieved anative-like or near-native-like pronunciation.This is then whyanasymmetrical bilingual will always speakhis/her non-dominantlanguage Bwithanaccent. Phonological borrowing is notruled out in RL agentivity, and can often be observedasgap-filling, as when thelong/a/instatur ‘state’, thelong/y/intypa ‘type’, andthe /au/in aula ‘meetinghall’ are borrowed into FA. These phonemes fill agap in FA phonology,as these long soundshaveonlycometobeasthe result of differentmergers and the development from ON to FA (Rischel 1967-1968, Snædal 1986). As such,the borrowed phonemes makethe system more natural. Loan morphology is borrowed in both SL andRLagentivity,but differently. The impositionofloanmorphology is more frequent in SL agentivity,while it is not as frequent in RL agentivity. Ifind e.g. verb +particle in codeswitchinginFAR-DAN,FAplural endings,and infinitives as well as past . Thesestructuresare easily imposedboth because FA is the dominant languageand becausethese are earlysystemmorphemes,as opposedtotense andagreement, which arelateinlanguageprocessing.For more on early system morphemes andlatesystem morphemes seech. 8. Easily separableformswithclear boundaries are more pronetobeing borrowedthan forms involving complex morphological alternations (Aikhenvald2008: 33). Evidencefrom RL agentivity (DA>FA)are the prefixes an-, be-and for-,and thesuffixes-agtigur, -arí, -heit, -ilsi (Wittkugel 2009). On the other hand,declinational and inflectionalborrowings arerare in RLAg, andthis is becausee.g. aplural-sis afused morphememeaning ‘nominative, accusative,plural, indefinite’ e.g.injeans in FA.This endingisextremely rare (Simonsen 2006,Sandøy&Petersen 2007).

26 But questionthenbecomes why FA pluralsoccur in FAR-DAN,thatisinSL agentivity,whileDAplurals are rarely seen in FA,inRLagentivity. Ithink the answertothisquestionisstraightforward:language dominance. It is easier to impose an early system morphemeontothe non-dominantlanguage than the other way around, wherethe morphology of the dominantlanguage blocks foreign declensional morphology. Aikhenvald (2008: 30) mentions that“borrowingapracticemay facilitate borrowing a set of linguisticexpressions which correlate with it.” As an example,she namesspeakers of languages in contact thatmay share culturalpractices likebuilding houses. An example fromFAisthe Danish trademonopoly andthe Danish school systemonthe Faroe Islands, which have givenrisetothe use of the Danish countingsystem, so that the colloquial language has ein-og-tjúgu (with no hyphen)=(lit.: one-and-twenty) fromDA en-og-tyve (with no hyphen)(lit.: one-and-twenty) =‘twenty one.’Ishould also mention that one of my informants from theK8projects says thatheremembers how an older man in his village sangthe multiplicationtableinDA. The practice of seamanshipcamefromEngland,asthe FaroeIslanders bought many sloops or smacks fromEngland towards the endofthe 1800s, andcoal-trawlers in the fifties, andthis gave rise to borrowingssuch as spinnigarður(in) ‘thespin-garden’fromEN spinnaker.The FA worddoesnot makeany sense at all. Theagent speakers have justtaken thefirst part spinn-and translated it as spinna ‘to spin.’ Phonetic transferencesfrom EN are seen in ship names, so that e.g. H.M.S Slater became Slatrið ‘the Gossip’, as /slt/ resembles FA /slaht/‘gossip.’ Loansfrom the coaltrawlers are e.g. at táttaí‘totighten’ from EN haultaught,thatis, the taught of EN is verysimilar to FA at tátta ‘totighten’ (Poulsen 1989). Other loansare hálvmáni ‘half moon’,which is from EN banana-link,where the shape has obviouslygiven rise to theFAnoun. Another loan is kameleyga ‘camel’s eye.’ It is abit difficult to explain exactly what it is,but themain point is that it is something that cannotget through alinkoraneye. TheENwordisKelly’seye,and there the agent speakers have used their knowledge of the . The sociolinguist settings must be taken into consideration, when evaluating a contact-inducedchange. Thehigher the degreeofknowledge in thesecond language,the higherthe degreeofborrowings,and thelonger thebilingual situation has existed historically, themore we expect to findborrowings of differentkinds.Thatis, theintensity of contact is important, as is proficiencyinlanguageBand speakers’ attitudestowardstheir secondlanguage. Agent speakers of FA haveahighlevel of proficiencyinDA, andthe bilingual situationhas lasted forcenturies,for details see ch.3and ch.4.Inaddition to that, speakersdonot typicallyhaveanegativeattitude towardsDanish. Jacobsen (2008: 292) shows, forexample,thatFaroeIslanders thinkthatthe situationasitis, with twoofficial languages, should not be alteredinany way, while more than 50% of theIcelanders think thatDAshouldnot be taughtinIcelandic schools. Indeed it is the case that Icelanders sometimes haveavery poor proficiencyinDAcompared to theFaroe Islanders. On the other hand,their proficiency in EN is better. Somelanguage societies aremoreopentowardsloans than others. Again, it is fruitful to compare the Faroe Islands,Iceland andDenmark. In Iceland, people areveryconservative regarding borrowings. In Denmark, borrowings are generallyaccepted,and many borrowings come into colloquialDAfromEN. Innovations have abetter chanceof spreadinginalanguage society with littleornoresistancetowards theinnovations. On the

27 Faroe Islands,speakerssay that theyprefer neologisms, butintheiractual speech they often useborrowedDanishwords (Gotved-Jacobsen &Nattestad-Steintún 1992, Jacobsen 2008). Puristshave tried for morethan acenturyto‘improve’FA, and they have succeeded to some degree whenwelookatwritten FA. By succeeded, Imeanthat some neologisms have found theirway into written FA,and thatitisjustrecentlythat writers ande.g. translators havestarted to usemore spokenFAintheirwork, acaseinpointbeing the conjunction vissi ‘if’ from DA hvis ‘if’, whichisusedinspoken FA, while um ‘if’ is therule in written FA,and not uncommoninspoken FA either. There areother socio-cultural parameters that promotetransference.Theyincludethe size of the community, interactionbetween rural and urban communities,marriage patterns, patternsoftrade andwarfare, the lifestyle of thespeakers,the division of labor between the sexes andbetween generations, social organization andthe kinship system, and religion and mythology.Giventhe Faroese-Danish situation, thesize of the communityisrelevant. TheFaroese languagesocietyissmall, andthereisnot enough economical recoursetotranslateeverything into FA. Hence,people read (andhear) alot in DA, andthis is why Ifinditimportant to stress role of the listener in alanguage change when discussingsome of thechangescurrently happening in FA. In acontact situation, thereceiving language may adopt certain patterns without losing the oldones, anditisnot uncommonthatolder andnewer constructions exist side by side forgenerations before theolder one disappears, if it ever does.Inthis context,Iwouldlike to pointout Kurolowicz’s 4th law of analogy, which says that when twoforms competefor one function, the newer form may takeoverthatfunction,while theolder form may become relegated to asubcategory of itsearlier function. The fact thatolderand newer forms can coexist is observedinFAwith e.g.the ditransitiveverbs, whichalternateswithDO+PP+DO, andthe regular comparative/superlative with-(a)ri vs. theanalytic comparative/superlative; cf. also Heine &Kuteva’s (2005: 71) notion of incipient categories, wheretheypoint out thatthese may be used optionally, andthattheyare not generally recognizedbysomespeakers(or grammarians) of the languageasdistinct entities of grammar, and areusually rejected by purists. To summarize: Ihavetouched uponfactorsthat may promotebilingualism and facilitate contact induced changes. In the present setting, Iwould again liketostress that it is important to keep in mind that thedistinction between internal,externaland extra-linguistic factors influencing achange are all but clear-cut. Further, Isupport the notionthatsomeofthe changesIwill discuss, especiallythe borrowingand conversion of DA syntacticstructures into FA,havecomeaboutasthe resultofaninnerchangewhich is sped up by languagecontact. It is unreasonabletoruleout languagecontact as afactor, or, to put it in thewords of Thomason (2001: 92), “ it is not historically realistic to assume that contact-induced change is responsible onlyfor changes that have never occurred elsewherethrough internal causation.” As we shall seeinch. 3and ch.4,Faroe Islanders hear andreadalot of DA, andthis is why Iwouldlike to stress the listener’s role in linguistic change.

28 3Anoverview of the historyofthe Faroese language

This chapterisincluded for thesakeofcompletenesssothat the foreign reader can getan idea of thelanguagesituation on the FaroeIslands. It is notachapter on the historical development of FA showing thenumerous lexical, phonological andmorphological changes from ON,nor does it concern itself with contact-induced changes,the chapteris an outline of the sociolinguistic settingofFAwithanemphasis on thestatus of the language andafew words on thegrowing text-corpus. The Faroe Islands were firstinhabited around800 AC (Arge 1997, Edwards 2005). The settlersarrived directlyfromNorwayorindirectlyvia the NordiccoloniesinEngland, Scotland and Ireland,and thisgaverisetoalinguistic melting pot, just as in Iceland. There are some Celtic loanwordsinFAasinIC. Thesecame with Celticslaves and mistresses,and/ortheyentered the Norselanguage of thesettlers in thecolonies in England, Scotland andIreland.Inaddition to theloanwords, it hasbeenarguedthata phrase like taðerótti ámær (lit.: there(N) is fear(N) on me(D)))) =‘Iamafraid’ is borrowed into FA fromCeltic, c.f. Old Irish tá eagla orm (lit.: is fearon-me)=‘I am afraid.’This phrase does notexistinany other North GMClanguage, only FA,hence the hypothesis thatitisborrowedfromCeltic(Barnes &Weyhe 1994: 217). Among the Celtic loanwords are,for example, kjalllámur ‘thelefthand’ (Gael. lamh chearr)and perhaps dunna ‘duck’ (Gael. tunnag)(Matras, Chr. 1965). Some contact with EN existedlater, mainlythroughsailors who people came into contact withintrade situations; thus, thereare loanwords such as húkur < hook and peia < pay from aroundthe 15th to 16th century.Among the EN loans fromaround1600 to 1800 are fittur ‘nice’

29 more specifically Low Germans, who were allowed to trade directlywith theFaroe Islands in 1361.In1529,amerchantfromHamburg wasgiven half of the islands as afiefdom (Debes 1995, Sandøy 2005:1929). Thedominance of /Bergenlasted until 1620, after which trade andadministration ties shiftedtoCopenhagen(Sandøy2005: 1930). This is when DA influence reallybegan, although (some) people must havehad, if nothing else, apassiveknowledge of DA before 1620,simply because DA was used in churches andby the administration,asDAhad become theofficial language on the FaroeIslands with the Reformation (1538). It is during this period that the basis of bilingualism was laid (Hagström2002: 102).The situation was oneofdiglossia with DA as the High (H) variant andFAasthe Low (L)variant. FA continued to be aspokendialect up until the middleof the ,when aFAorthographywas createdand people started to gain ageneral awareness of thelanguageand cultureasthe resultofLateRomanismand changes in the socialand economical structureofthe islands, which hadbythen moved fromafeudal to a fishingsociety. This in turn gave rise to thegrowth of thenew bourgeoisieand resultedin national pride. The creation of anew was notenoughtochangethe status of thelanguage fromLtoH,asalmost no written code existed. DA was still the language of administration,the church and, beginningin1846, the schools. It is only after the Second World Warthatthe written standardbegan to have anysignificant effect, as the FA corpus startedtogrow considerably and FA obtained thestatusofthe main language, on equal footing with DA,accordingtothe Heimastýrislóg (‘the Home RuleAct’) in 1948; more about this later. Virtuallynotexts are available from between 1400 until thelate 1700s.Thereare some letters from1400, andafew other documentswhich were editedbyJakobsen (1907). Some letters datefromthe 16th century andtheywereedited by A. C. Evensen in Savntil føroyinga søgu í16. øld (‘ACollection of theHistoryofthe Faroe Islanders in the16th century’).Then thereare the Jarðarbøkur,the manorial court rolls frombetween 1584 and1750. The court rolls arewritten in DA, but as the Danish writers reproducedFaroese place andproper names, the Jarðarbøkur gives some information about certain historicalchanges in FA (Hamre1944). The Løgtingsbøkur from1615 arebooks fromthe Local Parliament; these are also in DA,but as with the Jarðarbøkur,they contain some attempts to reproduceFAplace and proper names. There is atranscript fromaround 1600 of the Seyðabrævið (theSheep letter), originallywritten in 1298 by Gabriel Mitens, wholived on theFaroeIslands between 1584-1620 (Hamre1944: 8). Then thereare descriptions of the islands from1600 in Ferøers beskrifvelser by ThomasTarnovius (1669) and Færoæ et Færoa reserata by Lucas Debes from1673. These arewritten in DA, buttheydocontainsome,though little, FA. Generallyspeaking,writtenrecords arescanty from1400 until the latter part of 1700, when JensChristianSvabo collected ballads and compiled several dictionary manuscripts. These were only published in the 20th century, however, and hencedid not have any impact on the corpus at thetime. In 1811,Rasmus K. Rask wrote abrief descriptionofFAinhis book Vejledningtil det Islandskeeller gamle Nordiske Sprog (‘Guide to the Icelandic or the oldScandinavian Languages’),wherehementions FA and says that manychanges have happened,sothat the language is ‘plattere’ that is, ‘more vulgar.’ In 1822, Christian Lyngbye editedthe ballads of Sjúrður,which corresponds to the Germanic episode about the Dragon-Slayer in his book Færøiske Qvæder om Sigurd Fofnersbane og hansÆt(‘FaroeseBallads

30 about Sigurd theDragon-Slayerand his Relatives’). Theyearafter, atranslationofMatthew appeared,and in 1832 aversionofthe Icelandic Færeyinga (Saga of theFaroe Islanders). Withregard to the bilingual situation, 1845 was an importantyear in which a governmental decree stated that “the language of introduction as well as ‘themother ’ of the children shouldbeDanish” (Hagström2005: 1750). Theyearafter, Danish schools were introducedonthe islands. Fromthis point on,one can speculate thatpeople gradually moved from being passive bilingualstomore activebilinguals. In the Provisorisk Reglement,the governmentaldecreefrom the28th of May, 1845,itis said thatteachersshoulduse FA when it wasconsiderednecessary,inthe sensethatit allowed the childrentobetterunderstand the material. Theschool experiment, in which Danish authorities tried to introduce Danish schools on the Faroe Islands, failed anditwas abandoned in 1854. Therewas asecond attempt in 1872, where Danish schools were introduced on the Faroe Islanders. Theteaching languagewas supposed to be DA, butone can assume that some spoke FA, even though all teaching materialswereinDA. Theperiodbetween 1846 and1888 was characterizedbyacertain optimismregarding thefutureofFA. Writers were no longer as pessimistic as Svabo, who was certain that FA would not survive. Beginning in the mid-1800s, otherchangesoccurred in thesociety. In 1856,the Danish monopoly was abandonedand free trade was introduced. Anotherimportantchangeinthe late 1800s was the arrivalofthe sloops or smacks.The first one came in 1872, withatotal of six two yearslater; in 1882 therewere22, and in 1890there were 14 sloops; in 1906 therewereasmany as 128. The islands nowchanged rapidly from afeudal society to a fishing, industrializedsociety, and thechangewas reflectedinthe growthofthe population. At the time theislands were first inhabited, there had been around3000 to 4000 people. By the first censusin1801, therewere 5265. In 1860,the population was 8922;in1880, eight yearsafter the first sloop,there were 11,220;in190115,230 andin191118,000.Today, there are around49,000 peopleonthe FaroeIslandsand around20,000 in Denmark. Crew was abandoned in 1865; poor people gained the opportunitytoearn money, have families andbuild their own homes, which resultedinanincrease in population. Thegrowth of the new industry was reflected in theincrease in thenumberof bank accounts.In1900, there were 293,600kr. in 1155accounts andin1906478,500 kr. in 1527 accounts, for example. An industrialfishing society requires awrittenstandard, one that, with time, had to and has replacedDAtolarge extent.The standard FA orthographyinits more or less final version came in 1896; there wasalso an earlierversion from1846. It was composed by the priest V.U.Hammershaimb. The orthography is basedonthe etymological principle, which causedand still causesmany problems andspellingerrors. When people say,assome of my informants and those in an investigation conducted by Ólavsstovu (2007) did, that they knowDAbetterthanFA, Isuspect thattheymeanthatitiseasier to write the DA orthography, which is indeedthe case formanyFaroe Islanders. The period between 1846-1888 also marksthe birth of Faroese poetry, which started with the fosturlandssangir,the patriotic songs, in the1870s.Before this time,some200 balladsexisted -and still exist -which people sangand dancedto, typical of the oral culture with DA as the written mediuminthis case.Among the themes areKing Charlemagneand his paladins andthe episode about Sigurd the Dragon-Slayer.

31