DOCSLIB.ORG

Kilbury, Bontcheva. Some Empirical Aspects of Verb Inflection in Bulgarian and German

Total Page：16

File Type：pdf, Size：1020Kb

Download full-text PDF Read full-text
Kilbury, Bontcheva. Some Empirical Aspects of Verb Inflection in Bulgarian and German 1 Some Empirical Aspects of Verb Inflection in Bulgarian and German James Kilbury Katina Bontcheva Computerlinguistik Computerlinguistik Institut für Sprache und Information Institut für Sprache und Information Heinrich-Heine-Universität Düsseldorf Heinrich-Heine-Universität Düsseldorf Universitätsstr. 1 Universitätsstr. 1 D-40225 Düsseldorf, Germany D-40225 Düsseldorf, Germany [email protected] [email protected] Keywords: inflectional morphology, non- framework of Paradigm Function Mor- monotonic inheritance, DATR, empirical phology (cf. Stump 2001). evidence, language acquisition, Bulgarian In the analyses presented here we treat verb inflection, German verb inflection, Bulgarian and German verb inflection in abstract morphophonemic representation terms of NM and closely related frame- works. Inflectional types are represented as Abstract objects within a nonmonotonic inheritance The paper discusses analyses of Bulgarian hierarchy that provides an explicit account and German verb inflection that are based of regularities, subregularities, and excep- on nonmonotonic inheritance and encoded tions with highly constrained lexical en- in DATR. The analysis of Bulgarian also tries, default inflectional classes, and non- makes essential use of abstract morpho- productive classes. We try to avoid as phonemic representations. We argue that much as possible the use of virtual classes such analyses of inflectional morphology that do not reflect any concrete inflectional must reflect and be supported by empirical type, and, for Bulgarian, we employ ab- data on language acquisition and mental stract morphophonemic representations to lexical representations reported in psycho- simplify the description of the alternations linguistic studies. and morphotactics (for details on our use of such representations cf. Bontcheva & Kil- 1 Introduction bury 2003). Our analyses are encoded in Since the mid 1980’s representation lan- DATR. guages based on networks with multiple and nonmonotonic inheritance have been 2 Empirical basis of NM analyses employed in descriptions of inflectional While monotonic classifications of inflec- morphology. Typically, nodes denote gen- tional types or other objects can be deter- eralized paradigms or inflectional classes, ministically computed (cf. Petersen 2004), while the relations between inflectional nonmonotonic classifications in principle classes are captured by the hierarchical re- allow arbitrary structuring and require ex- lations between nodes. In 1989 DATR (cf. ternal specification as to what counts as Evans & Gazdar 1996) was introduced as a regular or general. Thus, NM-based ac- formalism for lexical knowledge represen- counts of inflectional morphology must tation that avoids certain problems such as specify further formal criteria and/or em- conflicts of multiple inheritance. Since then pirical evidence for particular analyses. it has become widely adopted in computa- Early accounts of NM and other tional linguistics. DATR-oriented work tended to neglect the Network Morphology (NM) utilizes empirical criteria for particular proposals DATR notation and principles of inheri- involving class hierarchies (cf. Corbett & tance networks in a linguistic theory of in- Fraser 1993), but recent studies seek to flectional morphology (cf. Corbett & Fraser provide empirical justification through cor- 1993, Corbett 2000). NM reflects a grow- pus investigation of frequencies (cf. Brown ing theoretical linguistic interest in inheri- et al. 2004). In particular, psycholinguistic tance-based morphology and is close to the data on first language acquisition has been advanced to support hierarchical analyses 2 of German noun inflection (cf. Cahill & never have a form like *muhl in analogy to Gazdar 1999, Clahsen et al. 2002). fuhr from fahren (Vaba). In this paper we couple psycholinguis- For geschrieben we expect the forms tic criteria with formal constraints on hier- *geschreiben (Vn/Vaba) and geschreibt archies in NM accounts of inflectional (V), but rufen would not have a form like classes: *geriefen (Vabb) in analogy to geschrieben • In accord with the studies on German from schreiben. mentioned above we require that im- Finally, beside gesungen the analysis mediate dominance (ID) relations be- predicts the three forms *gesangen (Vabb), tween nodes for inflectional classes *gesingen (Vn/ Vaba), and *gesingt (V). mirror the direction of overgeneraliza- All the predictions ultimately should be tion in language acquisition. confirmed in order to support the analysis. • Moreover, we extend the previous If only the form *gesingt is in fact attested, principle to the transitive closure of the then this would speak against the specific ID relation. hierarchical structuring proposed here. • Finally, we assume that overgeneraliza- Note also that our analysis excludes tion invariably involves analogy to a any paradigm not found in a real inflec- concrete inflectional type, which means tional type (e.g., rufen – rief – *geruft) and that hierarchies should include no accordingly involves no corresponding nodes for virtual classes not repre- nodes for virtual classes. This restriction is sented by real lexemes. fulfilled but not explicitly discussed in some recent descriptions of German noun The principles can be illustrated with a part inflection (cf. Cahill & Gazdar 1999, Kil- of our analysis of German verb inflection. bury 2001). We assume a linear ordering of nodes for Further details of this analysis and the the following classes with V as root: empirical evidence will be presented in an- • V for lachen ‘laugh’ other study. We break off the discussion • Vn for mahlen ‘grind’ here and go on to a parallel investigation of • Vaba for rufen ‘call’ Bulgarian morphology. • Vabb for schreiben ‘write’ • Vabc for singen ‘sing’. 3 Bulgarian verb inflection V is the regular inflection with preterite In this paper we describe the construction stem in –te (lachte) and perfect participle in of the synthetic forms that involve only –t (gelacht). Vn inherits everything from V inflection, i.e., the forms for present, aorist, except the form of the participle, which is and imperfect tense, the participles (present built with –n (gemahlen). active, aorist, imperfect, past passive), the Vaba introduces a distinct stem vowel verbal noun and adverb, and the simple in the preterite stem (rufen – rief – forms for imperative. gerufen), while Vabb has the same distinct Most Bulgarian verbs have forms for vowel in both preterite and the participle both perfective and imperfective aspect (schreiben – schrieb – geschrieben). Fi- although some verbs are imperfectiva tan- nally, Vabc introduces three distinct vowels tum, i.e., without a form for perfective as- (singen – sang – gesungen). pect. Quite often these forms have different Our hierarchy makes very explicit em- aspectual stems and belong to different pirical predictions that should be confirmed conjugations. Also, the perfective and the or refuted by psycholinguistic data. Over- imperfective stems take different sets of generalizations in language acquisition are nonfinite inflections. For this reason, al- invariably directed upward in the hierarchy, though they are forms of one and the same so that we find *gemahlt but never lexeme, the perfective and the imperfective *gelachen in data from children (cf. stems appear here as independent entries. Clahsen et al 2002). Traditionally, one speaks of three con- For the preterite rief from rufen we ex- jugations in Bulgarian. Two of them repre- pect regular *rufte (V), but mahlen should sent old inflectional types, inherited from Old Church Slavonic. They have numerous classes and even subclasses defined by the 3 relationship between the present-tense stem There are seven inflectional classes that and the aorist stem. Actually, these conju- belong to the 1st (e) conjugation and two to gations are virtual – no verb is defined as the 2nd (i) conjugation. Here the class- belonging to the 1st or 2nd conjugation specific features are stored that constitute without additional reference to the inflec- subregular exceptions to the defaults de- tional class. The third conjugation is new scribed in Va. Most classes have subclasses and emerged as the process of secondary with regard to stress and the type of base imperfectivization with the suffixes -а-(-я-) (i.e., vocalic or consonant). /-ва-/-ава-(-ява-)/-ува- became very pro- All lower nodes by default inherit the ductive. In contrast to the 1st and 2nd conju- features from higher nodes in the hierarchy. gations, it is not virtual and has no classes. However, the inheritance is nonmonotonic The verbs belonging to the 1st and 2nd and at each level the information inherited conjugations have two basic stems tradi- by default can be overridden (cf. the code tionally called present-tense stem and the below). aorist (infinitive) stem. The verbs that be- Bulgarian verbs have two aspectual long to the 3rd conjugation have only one stems described in different entries (cf. stem, which has no thematic vowel and is above). Each imperfective or perfective equal to the base (i.e., the aspectual stem). stem2 adds a set of inflectional elements The present-tense stem is used in the such as: following forms: present tense, imperfect tense, the simple forms for imperative, pre- • stem vowels for the temporal stems (sv sent active participle, imperfect participle, $tens, i.e., stem vowels for present, ao- and the verbal adverb. Actually, it is the rist, imperfect tense) imperfect-tense stem
Load more

Recommended publications
  • Greek Verb Aspect
    Greek Verb Aspect Paul Bell & William S. Annis Scholiastae.org∗ February 21, 2012 The technical literature concerning aspect is vast and difficult. The goal of this tutorial is to present, as gently as possible, a few more or less commonly held opinions about aspect. Although these opinions may be championed by one academic quarter and denied by another, at the very least they should shed some light on an abstruse matter. Introduction The word “aspect” has its roots in the Latin verb specere meaning “to look at.” Aspect is concerned with how we view a particular situation. Hence aspect is subjective – different people will view the same situation differently; the same person can view a situation differently at different times. There is little doubt that how we see things depends on our psychological state at the mo- ment of seeing. The ‘choice’ to bring some parts of a situation into close, foreground relief while relegating others to an almost non-descript background happens unconsciously. But for one who must describe a situation to others, this choice may indeed operate consciously and deliberately. Hence aspect concerns not only how one views a situation, but how he chooses to relate, to re-present, a situation. A Definition of Aspect But we still haven’t really said what aspect is. So here’s a working definition – aspect is the dis- closure of a situation from the perspective of internal temporal structure. To put it another way, when an author makes an aspectual choice in relating a situation, he is choosing to reveal or conceal the situation’s internal temporal structure.
    [Show full text]
  • Tense, Time, Aspect and the Ancient Greek Verb by Jerome Moran
    Tense, Time, Aspect and the Ancient Greek Verb by Jerome Moran early every – no, every – Greek The questions in the first sentence 3. In Greek the tense of a verb may Ngrammar and course book, even (‘deliberative’ questions, therefore in denote something different from or the most comprehensive (in English, the subjunctive) refer to present (or additional to the time at which the at any rate), gives a very skimpy, perhaps future) time. But one of act, event, occurrence, process, state perfunctory and unhelpful account — the verbs (εἴπωμεν) is in a past denoted by the verb is located. In insofar as it gives any account at all – of tense (aorist). The second sentence particular, it may denote something what ‘aspect’ is and how exactly it is refers to past time, but one of the called ‘aspect’. related to verb tense and time (which verbs (βούλοιτο) is in the present tend to be conflated). Most of the tense. 4. Whether the tense of a Greek verb books and articles on the subject of denotes time or/and aspect depends the aspect of the Greek verb are What is going on? The answer is in the first place on the mood of the accessible only to the professional something called ‘aspect’, and its verb (‘the form which a verb philologist, and can’t therefore be connection with tense and time. Just assumes in order to reflect the easily applied by non-specialists to the note for now a difference in the manner (modus) in which the speaker understanding of the actual usage of kind of things denoted by the verbs conceives the action’ (Woodcock)).
    [Show full text]
  • The Aorist in Modern Armenian: Core Value and Contextual Meanings Anaid Donabedian-Demopoulos
    The aorist in Modern Armenian: core value and contextual meanings Anaid Donabedian-Demopoulos To cite this version: Anaid Donabedian-Demopoulos. The aorist in Modern Armenian: core value and contextual mean- ings. Zlatka Guentcheva. Aspectuality and Temporality. Descriptive and theoretical issues, John Benjamins, pp.375 - 412, 2016, 9789027267610. 10.1075/slcs.172.12don. halshs-01424956 HAL Id: halshs-01424956 https://halshs.archives-ouvertes.fr/halshs-01424956 Submitted on 6 Jan 2017 HAL is a multi-disciplinary open access L’archive ouverte pluridisciplinaire HAL, est archive for the deposit and dissemination of sci- destinée au dépôt et à la diffusion de documents entific research documents, whether they are pub- scientifiques de niveau recherche, publiés ou non, lished or not. The documents may come from émanant des établissements d’enseignement et de teaching and research institutions in France or recherche français ou étrangers, des laboratoires abroad, or from public or private research centers. publics ou privés. The Aorist in Modern Armenian: core value and contextual meanings, in Guentchéva, Zlatka (ed.), Aspectuality and Temporality. Descriptive and theoretical issues, John Benjamins, 2016, p. 375-411 (the published paper miss examples written in Armenian) The aorist in Modern Armenian: core values and contextual meanings Anaïd Donabédian (SeDyL, INALCO/USPC, CNRS UMR8202, IRD UMR135) Introduction Comparison between particular markers in different languages is always controversial, nevertheless linguists can identify in numerous languages a verb tense that can be described as aorist. Cross-linguistic differences exist, due to the diachrony of the markers in question and their position within the verbal system of a given language, but there are clearly a certain number of shared morphological, syntactic, semantic and/or pragmatic features.
    [Show full text]
  • The Indo-European K-Aorist
    Frederik Kortlandt, Leiden University, www.kortlandt.nl The Indo-European k-aorist Ten years ago I wrote (2007: 155f.): “It has been impossible to establish an original meaning for the alleged velar suffix in the root aorists fēcī and iēcī (cf. Untermann 1993). I therefore think that we have to look for a phonetic explanation. Since the -k- is limited to the singular in the Greek active aorist indicative, I am h inclined to regard fēc- as the phonetic reflex of monosyllabic *dhēk < *d eH1t, where *-k- may have been either an intrusive consonant after the laryngeal before the final *-t, like -p- in Latin emptus ‘bought’ or *-s- in Hittite ezta ‘he ate’ < *edto, or a remnant of the Indo-Uralic velar consonant from which the laryngeal developed, as in Finnish teke- ‘make’ (cf. Kortlandt 2002: 220). The present stems of faciō and iaciō support the former possibility. This would also account for Tocharian A tāk, B tāka ‘became’, h h which reflect *steH2t. In a similar vein I reconstruct *hēp < *g eH1b t and *sēp < *seH1pt for Oscan hipid, hipust ‘will hold’, sipus ‘knowing’. While Oscan hafiest ‘will hold’ is in accordance with the Latin, Celtic and Germanic evidence, Umbrian hab- suggests that h h h *g eH1b - yielded *g eb- with preglottalized *-b- at an early stage and that this root- final consonant was generalized in Italic. It appears that Latin capiō ‘take’ < *kH2p- adopted the -ē- of cēpī from its synonym apiō, ēpī, and that scabō, scābī ‘scratch’ reflects original *skebh- (cf. Schrijver 1991: 431).” As Untermann observes (1993: 463), the Latin stems fac- and iac- behave as roots of primary verbs, and the same holds for Sabellic and Venetic.
    [Show full text]
  • Greek Tenses in John's Apocalypse
    CHAPTER 13 Greek Tenses in John’s Apocalypse: Issues in Verbal Aspect, Discourse Analysis, and Diachronic Change Buist M. Fanning This essay will concentrate on discourse functions for the Greek tenses in the Apocalypse of John. I will pursue this through a dialogue with and critique of David Mathewson’s views as presented in his 2010 monograph published by Brill and an earlier article in Novum Testamentum (2008) on this topic.1 Mathewson’s work is a reflection of a larger school of thought on nt Greek ver- bal usage that has been influenced greatly by Stanley Porter, and so this gives me an opportunity to interact with the views of a larger group of recent writers based on work that I have done on verbal aspect in nt Greek.2 The issue that triggers this discussion is what some have called the confu- sion of tenses or erratic shifting of tenses in John’s Apocalypse, just one area of the larger topic of John’s solecisms or unusual Greek grammatical expressions.3 What Mathewson argues for (following Porter) and what I will dispute in this paper is twofold: (1) that aspect alone is the focus of the ancient Greek tense forms; they do not in themselves express any temporal meaning; in regard to the Apocalypse if we take time out of the equation, we eliminate all of the supposed problems with shifting tense forms; (2) the main secondary effect of aspect is a certain function to reflect discourse prominence, that is, back- ground, foreground, and frontground events or features in a text.
    [Show full text]
  • Notes on Aorist Morphology
    Notes on Aorist Morphology William S. Annis Scholiastae.org∗ February 5, 2012 Traditional grammars of classical Greek enumerate two forms of the aorist. For beginners this terminology is extremely misleading: the second aorist contains two distinct conjugations. This article covers the formation of all types of aorist, with special attention on the athematic second aorist conjugation which few verbs take, but several of them happen to be common. Not Two, but Three Aorists The forms of Greek aorist are usually divided into two classes, the first and the second. The first aorist is pretty simple, but the second aorist actually holds two distinct systems of morphology. I want to point out that the difference between first and second aorists is only a difference in conjugation. The meanings and uses of all these aorists are the same, but I’m not going to cover that here. See Goodwin’s Syntax of the Moods and Tenses of the Greek Verb, or your favorite Greek grammar, for more about aorist syntax. In my verb charts I give the indicative active forms, indicate nu-movable with ”(ν)”, and al- ways include the dual forms. Beginners can probably skip the duals unless they are starting with Homer. The First Aorist This is taught as the regular form of the aorist. Like the future, a sigma is tacked onto the stem, so it sometimes called the sigmatic aorist. It is sometimes also called the weak aorist. Since it acts as a secondary (past) tense in the indicative, it has an augment: ἐ + λυ + σ- Onto this we tack on the endings.
    [Show full text]
  • Tense in Time: the Greek Perfect 1
    Gerš/Stechow Draft 8. January 2002: EVA-CARIN GER… AND ARNIM VON STECHOW TENSE IN TIME: THE GREEK PERFECT 1. Survey .................................................................................................................................1 2. On the Tense/Aspect/Aktionsart-architecture .......................................................................4 3. Chronology of Tense/Aspect Systems................................................................................10 3.1. Archaic Greek (700-500 BC) .....................................................................................11 3.2. Classical (500 Ð 300 BC)............................................................................................18 3.3. Postclassical and Greek-Roman (300 BC Ð 450 AD)..................................................29 3.4. Transitional and Byzantine/Mediaeval Period (300 Ð 1450 AD) .................................31 3.5. Modern Greek (From 1450 AD).................................................................................33 4. Conclusion.........................................................................................................................35 * 1. SURVEY The paper will deal with the diachronic development of the meaning and form of the Greek Perfect. The reason for focusing on this language is twofold: first, it has often been neglected in the modern linguistic literature about tense; secondly, in Greek, it is possible to observe (even without taking into account the meaning and form of the Perfect in Modern
    [Show full text]
  • The Origin of the Greek Pluperfect
    Princeton/Stanford Working Papers in Classics The Origin of the Greek Pluperfect Version 1.0 July 2007 Joshua T. Katz Princeton University Abstract: The origin of the pluperfect is the biggest remaining hole in our understanding of the Ancient Greek verbal system. This paper provides a novel unitary account of all four morphological types— alphathematic, athematic, thematic, and the anomalous Homeric form 3sg. ᾔδη (ēídē) ‘knew’—beginning with a “Jasanoff-type” reconstruction in Proto-Indo-European, an “imperfect of the perfect.” © Joshua T. Katz. [email protected] 2 The following paper has had a long history (see the first footnote). This version, which was composed as such in the first half of 2006, will be appearing in more or less the present form in volume 46 of the Viennese journal Die Sprache. It is dedicated with affection and respect to the great Indo-Europeanist Jay Jasanoff, who turned 65 in June 2007. *** for Jay Jasanoff on his 65th birthday The Oxford English Dictionary defines the rather sad word has-been as “One that has been but is no longer: a person or thing whose career or efficiency belongs to the past, or whose best days are over.” In view of my subject, I may perhaps be allowed to speculate on the meaning of the putative noun *had-been (as in, He’s not just a has-been; he’s a had-been!), surely an even sadder concept, did it but exist. When I first became interested in the Indo- European verb, thanks to Jay Jasanoff’s brilliant teaching, mentoring, and scholarship, the study of pluperfects was not only not a “had-been,” it was almost a blank slate.
    [Show full text]
  • Evidentiality and Mood: Grammatical Expressions of Epistemic Modality in Bulgarian
    Evidentiality and mood: Grammatical expressions of epistemic modality in Bulgarian DISSERTATION Presented in Partial Fulfillment of the Requirements o the Degree Doctor of Philosophy in the Graduate School of The Ohio State University By Anastasia Smirnova, M.A. Graduate Program in Linguistics The Ohio State University 2011 Dissertation Committee: Brian Joseph, co-advisor Judith Tonhauser, co-advisor Craige Roberts Copyright by Anastasia Smirnova 2011 ABSTRACT This dissertation is a case study of two grammatical categories, evidentiality and mood. I argue that evidentiality and mood are grammatical expressions of epistemic modality and have an epistemic modal component as part of their meanings. While the empirical foundation for this work is data from Bulgarian, my analysis has a number of empirical and theoretical consequences for the previous work on evidentiality and mood in the formal semantics literature. Evidentiality is traditionally analyzed as a grammatical category that encodes information sources (Aikhenvald 2004). I show that the Bulgarian evidential has richer meaning: not only does it express information source, but also it has a temporal and a modal component. With respect to the information source, the Bulgarian evidential is compatible with a variety of evidential meanings, i.e. direct, inferential, and reportative, as long as the speaker has concrete perceivable evidence (as opposed to evidence based on a mental activity). With respect to epistemic commitment, the construction has different felicity conditions depending on the context: the speaker must be committed to the truth of the proposition in the scope of the evidential in a direct/inferential evidential context, but not in a reportative context.
    [Show full text]
  • Errant Aorist Interpreters
    Grace Theological Journal 2.2 (Fall 1981) 205-26. [Copyright © 1981 Grace Theological Seminary; cited with permission; digitally prepared for use at Gordon College] ERRANT AORIST INTERPRETERS CHARLES R. SMITH The thesis of this essay is that exegesis and theology have been plagued by the tendency of Greek scholars and students to make their field of knowledge more esoteric, recondite, and occult than is actually the case. There is an innate human inclination to attempt to impress people with the hidden secrets which only the truly initiated can rightly understand or explain. Nowhere is this more evident than in the plethora of arcane labels assigned to the aorist tense in its supposed classifications and significations. Important theological dis- tinctions are often based on the tense and presented with all the authority that voice or pen can muster. It is here proposed that the aorist tense (like many other grammatical features) should be "de- mythologized" and simply recognized for what it is--the standard verbal aspect employed for naming or labeling an act or event. As such, apart from its indications of time relationships, it is exegetically insignificant: (1) It does not necessarily refer to past time; (2) It neither identifies nor views action as punctiliar; (3) It does not indicate once- for-all action; (4) It does not designate the kind of action; (5) It is not the opposite of a present, imperfect, or perfect; (6) It does not occur in classes or kinds; and, (7) It may describe any action or event. * * * THE ABUSED AORIST In 1972 Frank Stagg performed yeoman service in publishing an article titled "The Abused Aorist."1 A number of the illustrations referred to in the following discussion are taken from his article.
    [Show full text]
  • Aorist and Pseudo-Aorist for Svan Atelic Verbs — K. Tuite 0. Introduction. the Kartvelian Or South Caucasian Family Comprises
    Aorist and pseudo-aorist for Svan atelic verbs — K. Tuite 0. Introduction. The Kartvelian or South Caucasian family comprises three languages: Georgian, which has been attested in documents going back to the 5th century, Zan (or Laz-Mingrelian) and Svan. In the case of Zan and Svan, there is almost no documentation going back beyond the mid-19th century. There is a concensus among experts that Svan is the outlying member of the family, whose family tree would be as in this diagram: COMMON KARTVELIAN COMMON GEORGIAN-ZAN PREHISTORIC SVAN GEORGIAN ZAN MINGRELIAN LAZ Upper Bal Lower Bal Lashx Lent’ex (Upper Svan dialects) (Lower Svan dialects) In this paper I will look at one segment of the formal system in Svan — in particular, in the more conservative Upper Svan dialects — for coding aspect. In this introductory section I will go over the system of what Kartvelologists term “screeves” (sets of verb forms differing only in person and number), and then the classification of verbs by lexical aspect with which the morphological facts will be compared. 0.1. Screeves and series. In the morphology of verbs in each of the Kartvelian languages, one of the fundamental distinctions is between verb forms derived from the SERIES I OR “PRESENT-SERIES” stem, and forms derived from the SERIES II OR “AORIST-SERIES” stem. We will not go into all of the formal differences between these two stems for the various groups of verbs in each language, save to note that for most verbs the Series I stem contains a suffix (the “series marker” or “present/future stem formant”) which does not appear in Series II forms.
    [Show full text]
  • To R:Conjugation in Bulgarian, the Citation Form of a Verb Can Terminate in -A, -Ja Or -M
    E'aazap crcunm nuK u rumep amyp a e vycrc6uu a u e Etneapun ON SOME RECENT CHANGES IN BULGARIAN CONruGATION1 Stela Manova2 (Universität Wien, Institut fi.ir Slawistik, Universitätscampus AAKH stela.manova@univie. ac.at) The present paper focuses on two topics ttrat, evidently, have not been treated in recent literature on Bulgarian verb morphology (except in Manova 2006): 1) Since publication of the Bulgarian Academy Grammar (henceforth BAG) in 1983, some verbs have changed their conjugation, i.e. verbs originally belong- ing to the e-CONJ3 have moved to the r-CONl,i.g . mämrja,(I) murmur,, mämrei, etc. according to the BAG, but mämrja, mämrii now (cf. the New Spelling Diction- ary of Bulgarian, henceforth NSDB, published in2002); and 2) Since 1983, there has been a clear tendency toward preservation ofthe stress pattern of the present in the aorist, i.e. PRES pisa, piiei etc. and AoF.pßax, pfsa, etc. rather than pisax, pisö etc, with an end-stress (cf. BAG and NSDB). This paper examines the implications of the two changes above for the or- ganization of Bulgarian verb inflection. I contend that these changes are evidence for root-based phonological organization of Bulgarian verb inflection, i.e. the pa- per confradicts the classical stem-based classification of Bulgarian (and Slavic) verbs. The paper is structured as follows: section 1 establishes the number and phonological properties of the verbs with altered conjugations; section 2 treats the stress patterns ofthe present-day paradigm ofBulgarian verbs; section 3 discusses the implications of the changes accounted for in the previous two sections and sub- sequently defines a system of inflection class assignment rules; and in section 4 conclusions are drawn.
    [Show full text]