Acta Linguistica Lithuanica

LXI (2009), 31 – 40

The Segmentation of the Subjunctive Mood: Problems of Synchronic Morphology

Asta Balčiūnienė

Klaipėdos universitetas

This article deals with the morphological segmentation of the Lithuanian subjunctive mood forms from a synchronic point of view. The approaches advanced in earlier publications are discussed. The alternative is between disregarding diachronical facts and recognizing one affix -t-/-č- (*tj), and recognizing two suffixes -čia- and -tų/-tum, which entails accounting for the facts of allomorphy.

0. INTRODUCTORY REMARKS

In contemporary linguistics, many problems arise from unclear criteria for distinguishing diachronic from synchronic assessment of certain linguistic phenomena from the perspective of morphology. This issue has become particularly prominent at the level of the verbal system. The idea expressed in Jolanta Vaskelienė’s article (Vaskelienė 2006: 83–93)—that in synchronic morphology it is preferable to treat the stem-final element of the third person of verbs as an ending—sparked discussion (Ambrazas 2006: 109–110; Holvoet 2006: 112–166). This shows that distinguishing diachrony from synchrony still poses certain theoretical and methodological difficulties (at least in the verbal system)¹. The analysis of some other verbal forms is also problematic.

¹ If one maintains that, following the resegmentation of stems, the stem-final element coincided with the ending (and therefore synchronically may also be considered an ending), then not treating the third-person stem-final element as an ending contradicts the first claim. Holvoet’s view that it is incorrect not to treat the third-person verb ending as an ending, since in the noun system we treat the bare stem-final element as an ending (cf. žmonà, where -a < *-ā < *-eh₂-ø), shows that there is no unified system for assessing morphological phenomena (Holvoet, 2006: 112–115). Thus, this gives rise to both theoretical and methodological difficulties (an additional symbol for marking the stem-final element would even need to be introduced into the system of conventional morpheme symbols, and the distinction of the stem-final element would need to be taught in schools). Incidentally, the athematic verbs’

31

ASTA BALČIŪNIENĖ

morphological segmentation of forms². The morphological segmentation of the subjunctive mood varies in the linguistic literature, and the question of codifying its paradigm³ remains unresolved.

The aim of this article is to reveal problems concerning the identification and treatment of the subjunctive suffix or suffixes in contemporary morphology and to propose possible ways of resolving them.

1. REVIEW OF RESEARCH

The subjunctive mood has been examined in considerable detail in various works on morphology: its meaning and status within the system of mood categories have been investigated especially consistently (Holvoet, Judžentis 2004: 76–104; Paulauskienė 1979: 148–153; Paulauskienė 1994: 316–318; DLKG: 342–343; Holvoet 2002: 39–49; Holvoet 2007). Differences and similarities in the expression of moods in Lithuanian and English have also been studied (Usonienė 2004).

The 3rd person ending -ti is universally identified, although its functions and meaning are identical to those of the 3rd person ending in thematic conjugation. By contrast, for some reason there are still differing opinions regarding the overlap between the functions of the stem-final element and the ending.

²For example, the inconsistent inclusion of the supine in the verbal system raises doubts. In the Grammar of Contemporary Lithuanian (1994: 385; hereafter — DLKG), this verb form is described, whereas it is not mentioned in Aldona Paulauskienė’s Lithuanian Morphology (Paulauskienė 1994), although other dialectal facts are recorded. Dialects are an integral part of the Lithuanian language (they are used in non-public settings), and their preservation is safeguarded by the Law on the State Language, whose Article 19 states that “The State takes care of the prestige of correct Lithuanian, creates conditions for safeguarding language norms, personal names, place names, and dialects.” Therefore, a Lithuanian grammar that consistently presents facts of non-public language use would also be needed. The DLKG reflects dialectal facts, for example, mentioning the subjunctive forms suptau, suptai used by Dzūkian speakers (DLKG: 343), but does so inconsistently.

³Sergejus Temčinas raised this issue, arguing that the subjunctive paradigm should be codified in Standard Lithuanian: sg. 1st person būčiau, 2nd person būtum, 3rd person būtų; pl. 1st person būtume, 2nd person būtute, 3rd person būtų. This would also make the morphological analysis of this mood easier. This idea is indeed worth endorsing, but, in order to codify the paradigm, the frequency of use of related forms, such as būtute // būtumėte // būtumėt, in Standard Lithuanian needs to be investigated. According to data from the Corpus of Contemporary Lithuanian (hereafter — LKT), the most frequently used form is the 1st person plural būtume (total number of occurrences — 2022; cf. būtumėm — 88, būtumėme — 9, while the 2nd person plural būtumėte occurs 300 times; cf. būtumet — 166, būtute — 5. Therefore, on the basis of the frequency criterion, the codified 2nd person plural subjunctive form should most likely be būtumėte (a firm conclusion can be reached only after examining more material).

32

Tariamosios nuosakos skaida: sinchroninės morfologijos problemos

As for the basis of subjunctive formation, there is agreement from a synchronic perspective: it has an infinitive stem. However, there is no complete clarity regarding the derivational affix. Various studies of morphology record that the subjunctive has an infinitive stem⁴ and / “ending” (Jablonskis 1997: 83) / “inflectional suffix” (Valeckienė 1998: 362) / suffixes (DLKG: 342,) / “mood formatives” (Paulauskienė 1994: 316), “relatively stable markers” (a relatively stable marker) (Holvoet 2007: 35) / a suffix⁵ (LKG_II: 138–140). This variety of terms shows that the criteria for identifying the formation of this mood and distinguishing its suffix or suffixes (formative or formatives) are not uniform.

Of course, there is no basis for speaking of the endings -čiau, -čia, -tumei, -tum…, -tų, and so on (cf. Jablonskis 1997: 83). Jablonskis’s idea of derivation by endings for moods is unfounded if one holds that verbs have no ending-based derivation. Thus, this conception has been rejected in modern grammars.

Nevertheless, several questions arise concerning the endings mentioned above. First, do we distinguish one or several subjunctive suffixes? Second, which suffix or suffixes do we distinguish synchronically, and what are the boundaries of this morpheme or these morphemes?

1.1. Distinguishing suffix / suffixes

There are still many disagreements in current morphological studies concerning the subjunctive suffix / suffixes. What particularly complicates the evaluation of this morphological segmentation is that it is argued for especially tersely. A review of the main studies reveals, in essence, two positions: one distinguishes several subjunctive suffixes (their number varies), while the other posits a single suffix of the mood under discussion.

⁴ Of course, diachronic studies are not included here; these set out the historical aspect of the development of the subjunctive—the subjunctive was formed from the supine (supinum), “extended by the formant -b(i) and personal endings” (Zinkevičius 1981: 122–124; cf. also Dini 2000: 329; Kazlauskas 1968: 385–404; Stang 1966; Endzelynas 1957: 193–194, Holvoet 2003: 39–49, etc.).

⁵ In the textbook for higher education Morphology of the Lithuanian Language (Jakaitienė, Laigonaitė, Paulauskienė 1976: 150; hereafter LKM), it is stated rather abstractly that a formant is added to the infinitive stem, one “not so easy to distinguish from the perspective of the contemporary language—all forms are united by the presence of the element t,” along with personal endings.

33

ASTA BALČIŪNIENĖ

1.1.1. Several subjunctive suffixes are distinguished

Aldona Paulauskienė Lietuvių kalbos morfologijoje (Paulauskienė 1994: 316) nurodo tokią tariamosios nuosakos asmenavimo paradigmą:

Vienaskaita Daugiskaita 1. neš-čia-u 1. a) neš-tu-me b) neš-tumė-me 2. neš-tum- (ei) 2. neš-tu-te neš-tumė-te 3. neš-tų- – 3. neš-tų- – neš-tų- –

The morphemes -čia-, -tum-, -tu-, -tų-, -tumė- are also specified in another work by this author (Paulauskienė 1979: 149). It is stated there that “historically this division is not very precise, but synchronically no other analysis is possible.” This statement does not make clear whether several subjunctive suffixes -čia, -tum, -tu and -tumė are distinguished, or whether they are treated as variants of a single morpheme. Incidentally, few arguments are offered to explain why no other synchronic analysis is possible. One such argument would be Žulys’s position (Žulys 1974, 1988) that, essentially, there is no third person, and therefore -ų cannot be a personal ending in an impersonal form; however, this claim provides little support for such a principle of suffix segmentation. The athematic ending -ti (eš-ti) is treated as an ending, although it too represents an “impersonal form.” Of course, the argument that only relics of athematic conjugation remain in contemporary Lithuanian is indisputable, but both the form kósti and neša must be segmented and analyzed according to consistent principles. The structuralist view that the third person is essentially “not a person” should be qualified. Contemporary typological literature uses the so-called person-indexing hierarchy (cf. Croft 2003: 156, based on Greenberg 1966: 42–44), which reflects relative differences in markedness within the category of person encoded in verb forms. This hierarchy is presented as follows:

3rd person < 1st person < 2nd person

Thus, the 3rd person is the least marked member, and can often take a zero affix, but this does not mean that it is not a member of the person category. If we recognize the possibility of distinguishing a 3rd-person ending at all, then synchronically the distinctive subjunctive ending could be considered either -ų (ei-t-ų) or zero (ei-tų-ø).

It is difficult to answer the question of whether Paulauskienė distinguishes five separate suffixes or presents variants of a single suffix.

34

Tariamosios nuosakos skaidā: sinchroninės morfologijos problemos

A similar conception is presented in Adelė Valeckienė’s Functional Grammar of the Lithuanian Language, where three “inflectional suffixes” of the subjunctive mood are distinguished: -čia-, -tumė-, and -tų- (Valeckienė 1998: 362). No arguments are given for distinguishing these suffixes.

The claims made by the authors cited above also do not explain the relationships among these suffixes. If we treat the suffixes as variants of a single morpheme, it is entirely unclear what criteria would justify doing so. In distinguishing affixes, the phonetic criterion must be considered first (both vowel and consonant alternations among morpheme variants are possible, but they must be regular). Yet in the suffixes listed above, -tu, -čia, -tum, -tumė, it is impossible to discern regular vowel and consonant alternations. Therefore, there is no clear basis for treating them as variants of a single suffix. On the other hand, if we treat these suffixes as separate affixes, it is problematic to justify their status as markers of the subjunctive mood.

In DLKG (342), the number of subjunctive suffixes is reduced: two suffixal morphemes are distinguished, -čia- and -tum(ė)-, although the third-person suffix -tų, which coincides with the supine, is also noted. This interpretation is phonologically justified (-tum and -tų are both < *tun⁶) and entirely acceptable, except for the variant with ė -tumė-(-ė-; the appearance of -ė- here cannot be explained in any way, although we might interpret it as an extension). Nevertheless, what is most surprising is that the distinct suffixes -čia and -tum(ė)- are assigned to the same subjunctive category without any explanation as to why. If we assume that the subjunctive is formed with different suffixes, it would be problematic to understand the system of subjunctive formation, since it is obvious that a given grammatical category is expressed by stable affixes specific to that category⁷.

In Axel Holvoet’s monograph Mood and Modality in Baltic, it is argued that the subjunctive marker -tum- / -tų is attached to the infinitive stem, to which present and preterite endings are then added. Forms of the first person with -čiau are reviewed separately, as is the issue of Old Lithuanian -čia, -čio (Holvoet 2007: 35), but it remains unclear whether the author regards these endings as suffixes or also distinguishes endings within them.

⁶ Of course, the appearance of consonant -tum- m in place of n in the morpheme should also be explained (this is discussed in detail in Zinkevičius 1981: 123).

⁷ Only the category of evidentiality lacks affixes specific to that category (Holvoet 2004: 107–113). We may argue that the imperative mood, too, can be expressed by the suffix -k- (ei-k) and the prefix -te- (te-daro), but this situation is explained by not distinguishing the optative mood.

35

Asta Balčiūnienė

A review of these studies shows that, from a synchronic perspective, the problem of segmenting the first-person singular form is especially poorly argued (both with regard to the distinct suffix and to the distinction between suffix and ending). On what basis is the ending of this person identified as -u rather than -au? Little attention is also paid to the problem of distinguishing the second- and third-person singular forms: following the aforementioned DLGK conception, both forms formally lack endings—regardless of the fact that for the second-person form ei-tum we must posit a zero ending, while the third-person form ei-tų cannot in principle have one, since it consists of a pure supine stem. But then the question arises whether these forms are formally distinguished only by a phonetic variant of the suffix.

1.1.2. One subjunctive suffix is distinguished

There are grammar studies in which a single subjunctive suffix (formant) is distinguished synchronically: -t- / -č- (*t̑) _(LKGII: 138; LKM: 1976, 150). These works essentially adopt a unified position: what changes is considered an ending (see Paulauskienė 1994: 35–36 for further discussion), while what remains stable is a suffix: “grammatical suffixes differ from endings in that they are invariable” (ibid., 35). Of course, distinguishing endings unusual in the verb system, such as -ų (eit-ų) and -umėme (eit-umėme), is also problematic.

It is clear that opinions vary in morphological studies regarding the number of suffixes and the criteria for distinguishing them. This raises the questions of what accounts for this diversity of opinion, which method of morphological segmentation is more rational, and whether it is indeed rational to reject the single-suffix conception advanced in earlier works.

2. THE DIACHRONIC ASPECT

One of the most important reasons for the problematic morphological segmentation of the subjunctive is probably the modernization of the formation of this mood, which led to the emergence of a considerable number of atypical variants in Lithuanian dialects. Thus, the complex history of the formation of this mood has contributed to a certain degree of confusion in present-day morphology.

As we know, forms of the subjunctive with the ancient formant -b(i)- (ne nufidetumbiau AK 56₁₂ /chweftumbi Mž. 19_2-3; for a more detailed discussion, see Zinkevičius 1981: 122) are no longer found in contemporary Lithuanian. It is precisely because of the loss of this formant that, in synchronic morphology, the desiderative suffix -tų < *tum has to be regarded as a derivational affix of the subjunctive. This would pose no problems, since the desiderative is not a principal form of the verb; thus, it is not being claimed that the subjunctive is formed from the desiderative stem—

36

Tariamosios nuosakos skaida: sinchroninės morfologijos problemos

mood is formed from the desiderative stem. The historical aspect would not strongly contradict this, either, since, according to Holvoet, a form identical to the desiderative was adopted as the base form for the 1st- and 2nd-person conjugated forms of the subjunctive: “the 3rd person must have played an important part in the development of the forms in -tum from NON-FINITE to FINITE forms” (Holvoet 2003: 43). It would seem that forming the subjunctive from the infinitive stem with -tų / -tum resolves all the issues, since a zero ending can be discerned in the 3rd-person forms⁸, as in the 3rd-person future form ẽ-s-ø.

Nevertheless, the 1st-person form ending in -čiau has no phonetic connection with the aforementioned suffixes. Therefore, in the studies by DLKG, Valeckienė, Paulauskienė, and Holvoet, -čia- is treated as a separate suffix. Although the criteria for segmenting this suffix are not argued for in all the works, it is clear that they have been adopted from historical Lithuanian grammar. That is, they rely on diachronic evidence which states that “three variants of the suppletive 1 sg. form were used in 16th–17th-century texts: rašyčiau, rašyčio and rašyčia” (Zinkevičius 1981: 124).

But diachronic studies do not answer all the questions concerning the morphological segmentation of the 1st-person form, either. The most important aspects that seem to indicate the unreliability of many hypotheses explaining the development of the 1st-person form are these: (1) whether, in such a recent mood paradigm (a paradigm characteristic only of the East Baltic languages), one can expect a suppletive form and a heteroclitic suffix, which presuppose a rather archaic situation (there is no other such variable means of distinguishing the 1st person in the Lithuanian verb system)⁹; (2) why the suffix is -čia- rather than -č- (< *ti̯), since the ending could also have been taken from the past tense¹⁰ (as we know, simple mood forms are not associated with a particular tense, i.e. they are not inflected for tense). The influence of past-tense forms would also be indicated by the 1st- and 2nd-person plural forms

⁸ Žulys laikosi pozicijos, kad nulinės galūnės skyrimas yra nelogiškas (Žulys 1983: 136–137), bet jei 3 a. įtraukiame į asmenavimo paradigmą, tai tokias formas, kaip eitu, turime vertinti kaip turinčias nulinę galūnę. Kitaip jos bus sutapatintos su nekaitomomis formomis, pavyzdžiui, auš-tant, auš-ti. Žulio siūlymas 3 a. asmens formas laikyti morfologiškai neskaidomomis (žr. ten pat) praktikoje taip ir neprigijo.

⁹ Chr. Stang’s view that this 1st-person form is supposedly an original noun form does little to explain how the form under discussion differs from other conjugated forms (Stang 1966: 252).

¹⁰ Incidentally, Zinkevičius (1981: 124; 1987: 222–223), Kazlauskas (1968: 401), and others also argue that a form of the būčiau type (established in standard Lithuanian) has a preterite ending.

37

ASTA BALČIŪNIENĖ

The variants with longer endings rašytum-ėme and rašytum-ėte, in which -ė- is probably also from the preterite, cf. also the longer 2nd person singular ending rašytum-ei (cf. the past-tense forms rašėme, rašėte, rašėi). Most likely, the paradigm of endings for these longer conditional forms was supplemented later and they were introduced according to the preterite formation pattern. This not only supports the segmentation of the 1st person ending -au, but also raises doubts about the validity of the suffix -tumė, indicated in some of the morphology works mentioned above.

If we acknowledge that the 1st person ending can be -au (in dialects, synchronically assessed—-a (eičia)), or even zero (eič’), then it is worth considering whether, synchronically, it would be clearer to distinguish a single conditional mood suffix -t- and its phonetic variant -č- (< *t̑).

The desiderative suffix -tų < *-tun, viewed from the standpoint of the present, could also be segmented, since synchronically we form the conditional from the infinitive, not from the desiderative (so why can we explain one thing synchronically but not the other?). In this case there are no links to the historical development of the conditional, but synchronic assessment may not coincide with diachronic assessment¹¹. If this approach is adopted, there would be no need to discern two different suffixes in the system of conditional formation, something that is difficult to explain even historically. All the more so from a synchronic perspective, it is difficult to imagine how two or more different suffixes could occur within the conjugation paradigm of the same grammatical category.

Of course, the argument that there are no endings such as -ų in the Lithuanian verb conjugation system is a strong one, but then the conjugation system has no heteroclitic suffixes either. Thus, synchronically distinguishing the conditional suffix -t-/-č- (< *t̑) would make the process of forming and recognizing the conditional simpler. Of course, this does not mean that we should abandon historical grammar. The proposal is simply to distinguish these levels more strictly.

Another approach, which does not strictly rule out diachronic assessment, could be to distinguish the two suffixes of this mood in the traditional way: -čia- and -tų/-tum-. In that case, one would need to explain what distinguishes the 2nd person singular and 3rd person forms, which is difficult to do synchronically. It seems that one would inevitably have to discern a zero ending in the form ei-tum, and the pure desiderative stem in the form ei-tų. Adhering to

¹¹ For example, we claim that the desiderative is formed from the infinitive stem, although historically the infinitive is more recent than the desiderative (see Zinkevičius 1981: 163-164). Thus, we could not say that the desiderative has an infinitive stem (from a diachronic perspective), even though historical grammar states that “desideratives <…> are formed from infinitive stems” (Zinkevičius 1981: 162).

38

Tariamosios nuosakos skaida: sinchroninės morfologijos problemos

from this perspective, to revise school textbooks by introducing a different segmentation of the conditional, since synchronic morphology should be based on consistent principles.

ABBREVIATIONS

DLKG — Grammar of Contemporary Lithuanian, ed. Ambrazas, V., Vilnius: Science and Encyclopaedia Publishing House, 1994.

LKG_II — Lithuanian Grammar, Vol. 2, ed. Ulvydas, K., Vilnius: Mintis, 1971.

LKM — Lithuanian Morphology, Jakaitienė, E., Laigonaitė, A., Paulauskienė, A., Vilnius: Mokslas, 1976.

LKT — http://donelaitis.vdu.lt/

LITERATŪRA

AMBRAZAS, V. 2006: Apie naujas veiksmažodžių 3 a. galūnės paieškas. Acta linguistica Lithuanica 55, 109–110.

CROFT, W. 2003: Typology and Universals, 2nd ed. Cambridge: Cambridge University Press.

DINI, P. U., 2000: Baltų kalbos. Lyginamoji istorija. Vilnius: Mokslo ir enciklopedijų leidybos institutas.

ENDZELYNAS, J. 1957: Baltų kalbų garsai ir formos. Vilnius: Pol सभा

GREENBERG, J. 1966: Language universals, with special reference to feature hierarchies. The Hague: Mouton.

HOLVOET, A. 2003: Notes of the development of the Lithuanian and Latvian conditional. Linguistica Baltica 10, 39–49.

HOLVOET A., JUDŽENTIS A. 2004: Nuosakos kategorijos struktūra. In HOLVOET A., SEMĖNIENĖ L., red., 2004, 77–104.

HOLVOET A., SEMĖNIENĖ L., red., 2004: Gramatinių kategorijų tyrimai. Vilnius: Lietuvių kalbos institutas.

39

ASTA BALČIŪNIENĖ

HOLVOET, A. 2006: Dėl galūnių ir kamiengalių. Acta Linguistica Lithuanica 55, 112–115.

HOLVOET, A. 2007: Mood and Modality in Baltic. Kraków: Wydawnictwo Uniwersytetu Jagiellońskiego.

JABLONSKIS, J. 1997: Lietuvių kalbos gramatika. Vilnius: Mokslo ir enciklopedijų leidykla.

KAZLAUSKAS, J. 1968: Lietuvių kalbos istorinė gramatika. Vilnius: Mintis.

PAULAUSKIENĖ, A. 1979: Gramatinės lietuvių kalbos veiksmažodžio kategorijos. Vilnius: Mokslas.

PAULAUSKIENĖ, A. 1994: Lietuvių kalbos morfologija. Paskaitos studentams. Vilnius: Mokslo ir enciklopedijų leidykla.

VASKELIENĖ, J. 2006: Dėl esamojo ir būtojo kartinio laiko dalyvių bei padalyvių darb ybos ir morfeminio skaidymo. Acta Linguistica Lithuanica 54, 83–94.

STANG, Chr. S. 1966: Vergleichende Grammatik der Baltischen Sprachen. Oslo etc.: Universitetsforlaget.

USONIENĖ, A. Modalumas anglų ir lietuvių kalbose: forma ir reikšmė. Vilnius: Vilniaus universiteto leidykla.

VALECKIENĖ, A. 1998: Funkcinė lietuvių kalbos gramatika. Vilnius: Mokslo ir enciklopedijų leidybos institutas.

ZINKEVIČIUS, Z. 1981: Lietuvių kalbos istorinė gramatika II. Vilnius: Mokslas.

ZINKEVIČIUS, Z. 1987: Lietuvių kalbos istorija II. Iki pirmųjų raštų. Vilnius: Mokslas.

ŽULYS, V. 1974: Trečiojo asmens nėra. Kalbotyra 25.1, 83–92.

ŽULYS, V. 1983: Dar dėl veiksmažodžio trečiojo asmens reikšmės. Baltistica 19.2, 133–144.

Asta Balčiūnienė

Klaipėdos universitetas

Baltų kalbotyros ir etnologijos katedra

Herkaus Manto g. 84, LT–92294 Klaipėda

[email protected]

40