Pastabos

In Lithuania, piršys is also attested in the plural masculine form (1), even near Varniai (LKŽₓ 26 gives?piršis). Horses were generally cared for by men, who commonly say that it is piršys, piršims, tai piršys; the source of the singular masculine form is unknown. The book also contains quite a few stress errors: bruõžas = bruóžas, ruožas = ruõžas, rašliavà = rãšliava, Ežerūnà = Ėžerūna, tù lydžiu = tuõ lydžiu, skėtra = skė́tra (1) neuter, and skėtrà feminine (4), aržuolas = áržuolas, ž. puõkai = puo kai ž.

Vytautas Antanas Vitkauskas

Institute of the Lithuanian Language

P. Vileišio g. 5

LT-10308 Vilnius, Lietuva

Once Again on the Classification of Syntactic Relations

In 2006, the fourth, revised edition of The Grammar of Contemporary Lithuanian was published. Since the chief editor of this grammar recently acknowledged (Ambrazas 2005: 168) that in earlier editions of DLKG, “... the morphologized opposition between government and adjunction proved to be of little relevance to the study of the syntactic system of the Lithuanian language and its development,” we were prompted to see how types of syntactic dependency are defined in the new, revised edition. It turns out that this edition, too, continues to distinguish the same three relations: agreement, government, and adjunction. We will quote their definitions in turn:

“Agreement is the selection of the morphological form (case, number, gender) of a dependent word in accordance with the head word...” (DLKG⁴, p. 480);

“Government is the selection of a dependent word determined by the valency of the head word. The governed word is most often a case form of a noun or pronoun [...]. It may also be an indeclinable word form” (ibid., 480).

“Adjunction is the selection of a dependent word’s form according to its meaning, not determined by the valency of the head word” (ibid., 481).

For convenience, let us compare the definitions quoted here with those given in earlier editions. The definition of agreement has not changed in comparison with earlier editions, but the definitions of government and adjunction were previously phrased as follows:

“Government is the selection of the case form of a dependent word, determined by the valency of the head word.” (DLKG¹, 480).

104

Pastabos

“Adjoining is a type of subordination in which the dependent word’s inflectional ending is not expressed. In this respect, adjoining contrasts with agreement and government taken together.” (ibid., 481).

It is immediately apparent that the definitions given in the new edition are of different kinds: the definitions of agreement and adjoining concern the selection of the dependent word’s form, whereas the definition of government concerns the selection of the dependent word itself. Clearly, if government is not restricted to government of case alone (as in earlier editions of DLKG), selection must be understood as the choice of a part of speech and lexical content that is semantically appropriate: the verb norėti ‘want’, for example, opens a position for a propositional argument expressed by a noun (noriu miego ‘I want sleep’) or an infinitive (noriu miegoti ‘I want to sleep’); this position may, but need not, be occupied by a case-inflected word. Defining agreement according to similar principles would be impossible, so government is defined syntactically, while agreement is defined morphologically.

The definition of adjoining remains vague. In the construction noriu miegoti ‘I want to sleep’, the word miegoti ‘to sleep’ is selected according to the valency of the governing word, whereas in the construction varinė rankena ‘copper handle’, the varinė FORM of the word ‘copper’ is selected according to the head word. The question arises: in what sense is the FORM of the words perpus mažesnis, žvilgsnis atgal ‘half as small, a backward glance’ selected according to the meaning of the words perpus and atgal ‘half and back’? What form is meant here? After all, the words atgal ir perpus ‘back and half’ are invariable! Thus, the same situation arises that prompted the authors of DLKG to refine the definition of government: since the concept of adjoining also covers constructions of the type kelia vome pern ai, and kelia vome miškais, its definition should in fact be supplemented with the same words that were added to the definition of government (we give them in reverse order, reflecting the order of the examples on pp. 480–481): “An adjoining word is most often an invariable word form [...]. It may also be a case form of a noun or pronoun.” But then it becomes entirely unclear why in one case the dependent word itself is said to be selected, and in the other, the form of the dependent word.

Let us try applying the given definition of adjoining in practice: in the phrase ilgesingas žvilgsnis ‘a longing glance’, the morphological form of the word ilgesingas ‘longing’ is selected according to the head word, whereas in the phrase žvilgsnis atgal ‘a glance back’, the form of the word atgal ‘back’ is selected not according to the valency of the head word, nor, in general, according to the head word, but “according to its meaning.” What form of the word atgal is meant here—morphological or some other kind? Presumably morphological, since the authors mention no other kind (although only in the case of agreement do they specify that a morphological form is meant). However, this morphological form cannot be associated with the features of case, gender, and number, since the word atgal ‘back’ of course has none of these. By the way, if it did have them, it would most likely agree with the head word in those features

105

Pastabos

as well. So it remains unclear what form of the word atgal ‘back’ is selected other than according to the head word. Admittedly, for constructions such as atsisėdom pailsėti ‘we sat down to rest’, one could argue that this morphological form is the infinitive. But then it is unclear why adjoining is described differently from government. The definition of government adds that “the governing word is most often a case form of a noun or pronoun [...] It may also be an invariable word form.” In this context, “invariable” effectively means “not inflected for case.” But even if, in the case of adjoining, we adopted a broader notion of an “inflected word” than in the case of government, including also finite and non-finite verb forms, the definition of adjoining would still remain unclear, since nothing explains to the reader what other kind of form, besides a morphological one, the word atgal ‘back’ could have.

It is not difficult to guess why the definition of adjoining speaks of selecting the FORM of the dependent word rather than selecting the word itself. For if adjoining were defined as the selection of a dependent word not determined by the valency of the head word, then agreement would be a subtype of adjoining, and one would have to distinguish between agreeing and non-agreeing adjoining; or, better still, remove agreement from the list of subordination types and adopt the distinction between government and modification proposed in Lietuvių kalbos gramatikos darbuose (cf. Holvoet, Judžentis 2003, Holvoet, Judžentis 2005). This is avoided by speaking of the selection of a form. Whether this solution is better is open to question. For if the word atgal ‘back’ has some morphological form selected according to its meaning, then perhaps the word ilgesingas ‘longing’ also has such a form (in addition to the already mentioned features of case, gender, and number), in which case, according to this definition too, agreement is merely a subtype of adjoining. Since the morphological form of the word atgal postulated by the authors of DLKG is not visible to the naked eye, we can neither confirm nor reject this view.

In general, it cannot be said that the newly defined opposition between government and adjoining is less morphologized than the previous one. In fact, only the definition of government is less morphologized, while the definition of adjoining remains just as morphologized as before. Strictly speaking, the opposition has disappeared altogether, because the two definitions address entirely different things: in one case, the selection of a dependent word; in the other, the selection of the form of a dependent word. The opposition between adjoining and agreement has also essentially disappeared, because if the definition of agreement speaks of “the selection of a morphological form according to the head word,” this presumably means that the form is selected according to the form of the head word, not its valency (which is entirely irrelevant to agreement); for adjoining to contrast with agreement, it would have to be defined as selection of a form not according to

106

Pastabos

the form of the head word. However, the definition’s criterion of “selection of a form not according to the valency of the head word” does not in fact distinguish adjoining from agreement, because this criterion applies to both types of subordination.

Thus, although the definitions of subordination types have changed considerably compared with those in earlier editions of DLKG, the overall number of internal contradictions has remained more or less the same, or has even increased slightly. We will continue to follow closely the development of the concept of subordination used in this grammar, so important for university teaching. It will be interesting to see how the types of subordination are defined in future editions of Dabartinės lietuvių kalbos gramatikos, of which many more are undoubtedly yet to come.

Ambrazas, V. 2005: [Rec.] Axel Holvoet, Artūras Judžentis, red., Sintaksinių ryšių tyrimai, Vilnius: Lietuvių kalbos institutas, 2003. Acta Linguistica Lithuanica 52, 167–170.

Holvoet, A., Judžentis, A. 2003: Sintaksinių ryšių tipai. In: A. Holvoet, A. Judžentis, red., Sintaksinių ryšių tyrimai, Vilnius: Lietuvių kalbos institutas, 11–35.

Holvoet, A., Judžentis, A., 2005a: Sintaksinės priklausomybės tipai: papildymai ir patikslinimai. In: A. Holvoet, R. Mikulskas, red., Gramatinių funkcijų tyrimai, Vilnius: Lietuvių kalbos institutas, 11–38.

DLKG¹ = Vytautas Ambrazas, ed., Dabartinės lietuvių kalbos gramatika. Vilnius: Mokslo ir enciklopedijų leidykla, 1994.

DLKG⁴ = Vytautas Ambrazas, ed., Dabartinės lietuvių kalbos gramatika. Fourth revised edition. Vilnius: Mokslo ir enciklopedijų leidybos institutas, 2006.

Axel Holvoet

Institute of the Lithuanian Language

5 P. Vileišio Street

LT-10308 Vilnius, Lithuania

On the reviewof the Compendium of Lithuanian Dialects¹

It is clear that every book is alive only when it has a connection with its readers. Every author’s or editor’s wish is to hear as many comments as possible about the publication they have prepared. Compendium of Lithuanian Dialects (Vilnius: Institute of the Lithuanian Language, 2004, 328,

¹ Acta Linguistica Lithuanica 52, 170–174.

107