Vytautas Ambrazas. „Lietuvių kalbos istorinė sintaksė“

Minutes Linguistics Lithuanica LVI (2007), 117 - 144 Reviews Reviews VytauTas AMBRAZAS Lietuviy kalbos istoriné sintaksé Vilnius: Lietuviy kalbos institutas, 2006, 610 pp., ISBN 978-9955-704-18-8 Vytautas Ambraz Historical Syntax of Lithuanian Language is a synthesis of the author’s long-term research, the results of which have been published in many separate publications. It is not surprising that particular attention is paid to those things which the author has previously studied in detail. These are the most important issues of Lithuanian historical syntax, such as Word order, participle constructions, communicative constructions, development of passive. This book fills some gaps, discusses ir those things that the author had t. y. not previously ir devoted a separate monographic article. Of course, the wholeness of everything is an ideal to which we can ar only strive, even in large synthetic works we always recognize the traces of the author’s individual research and ir preferences. With these remarks I want to point out that I am not going to accuse the author that either aspect ir could have been discussed more extensively in the book, this or that thing was not mentioned in it. 8j ta When the time comes for a new synthesis, another author will write another work, also reflecting individual preferences. jo Po In addition to the main chapter on the principles of historical syntax, the book contains a chapter on Zodziy order ir her pokyéius. Po to There are three chapters, in which the i§ author discusses three types of sentences, the distinction of which reflects, in the author's opinion, the most important criteria for the classification of sentences from a historical perspective. Firstly, the du main types of verbZodiniy — actionZodiniy ir sentences are distinguished, — the third type consists of so-called »mixed sentences«, sentences with t. y. verbZodiniy su (primary ir secondary) predicates. Belonging to the Misriam type, e.g. sentence Sinus grizo healthy, connecting in a separate sentence Sanus grigo ir Sanus (ra) healthy features. Ar sure These sentences "is in the distant idea. kalby senoveéje susidare dél verbazodiniy vardazodiniy ir sakiniy saveikos* (p. 491) i8 Nobody really knows, ir but it's probably impossible to find out. But as a medium iSdéstymo criterion This thing, of course, doesn't need proof. In other sections, eilés are discussed nouns, verbs, prepositional i8 constructions, participle constructions, and conjunctives. ir ir su 117 "Reviews in Chinese greeting phrases". In a separate section are summarized the historical development of the dél lietuviy language syntax, the summary of the vokiSka book, bibliography and indexes. Meditation is always a conventional matter, often dictated by practical considerations or simply tradition, and cannot be the object of principled criticism. But the structure of the content often reflects the apparatus used in the work, which should already be critically evaluated. In the text of the book in question, there are many uncommon things that make reading difficult for an unprepared reader, especially for a person who has not attended Lithuanian school and has not learned Lithuanian school grammar (similarly, for example, the reading of academic Lithuanian grammar or contemporary Lithuanian grammar). What should such a reader do when confronted, say, with,,dvejybiniy linksniy* savoka? In any case, it seems that the noun itself is not sufficient: two nouns, two adjectives, two nouns, etc. are used in the sentence, e.g., varny ben porele sav pietums nusisauju (Donelaitis; p. 267). So probably ir sentences like ASoil hot wood (p. 133) are cases of a binary denominator? There are two nouns in this sentence, but it somehow does not satisfy the condition that it belongs to the two-word category. In fact, in the section on binary verbs (except for the constructions of a completely different kind mentioned in the introduction) the constructions with secondary predicates are discussed, e.g., Asch tawa Dukterj neradau Mergq BrB V Moz 22.17. Under the mysterious title of the “dual verb” discovering a su well-known category, the reader feels relief and continues to expect iSsamesniy ir Ziniy about antriniy predikativy Zyméjima. Zinome (Slavistics) There is a lot of literature ir ji dedicated to conditioning factors), that secondary predicates can be defined in different ways: by conjugation, by adjective prepositional construction ar (plg. The court found ji guilty; He / served as a bernas /uz (per) bernq) with the ikininkg. How the history of Lithuanian language rolled Six threes The ratio of birth to death? | §j We do not find a direct answer to this question in the book, because, for example, the predicative joiner is discussed in a completely different chapter, where it is hidden under the no less meaningful title,,sentences with joiner“ (pp. 158-167). When $j opening the chapter, the reader will find it difficult to understand why Bread turned into kéin is ,,sayins su nyagininku“, o John rides a bicycle no longer. — This strange thing is finally explained only after studying the school grammar of Lithuanian language, in which the predicate of the conjunction (also ir su so-called,,non-linear connections“, such as to turn, to become, etc., the predicate ir is used) is not a separate part of the sentence, but only part of the Sudétian participle, o connected with the factor by a connection of coordination su (more precisely,,,interconnection“). About alignment in construction we essentially Liudinikais io (BrB Apd 5.32) Kalbeti 118 Reviews hard, so This predicative joiner finds itself in no man's land and, in order not to use a taboo-implying term, it is safer to hide behind a verb or other morphological category. uz This is how it happens ,,inagininku su sentences“ or_,,substantive sentences in the form of su adjectives“. How then to talk about what is not allowed to be talked about? ka The question of predicative conjugation or synonymy is discussed very briefly (pp. 165-167), but the author as if refuses the ir possibility based on his collected medZiaga about this something new to say, because ka already In Jablonski's grammar, the verb-conjugated predicate falls into j ,,dveyybiniy linksniy* category, one o sentence i§ action from this argument of the relationship su inagininku question there will not be even how to raise. Thus, both constructions live their own lives in the traditional Lithuanian grammar, and in principle they cannot meet. ir i8 O It is interesting that, for example, depictive secondary predicates, which in the present Lithuanian language are consistently combined with subject or object, in the old su Lithuanian language sometimes Signs of anxiety, e.g., to stop ChNT Apr 16.15 (now: not to walk naked); gime act ChNT Jn 9.20 (now: gimé blinds). Maybe it was just slavy kalby jtaka’, maybe o just the principles of linksniy usage were different than now? Mat Chylinskis combines the démenis in another ir construction, which should be recognized as complementary: neradau nes darba tawo pitno diewu po ChNT Apr 3.2. Triikstant specialiy research, the author ir may have i§ negléjo time i 81 the question, but at least the fact itself would be useful as a reference for other researchers in the future. ka But, in order to open a new perspective of research, it is necessary to abandon the school grammar categories and the ir ka i§ Modern theoretical linguistics has its own arsenal. That the author has read foreign literature, no doubt arises when looking at the bibliography of the book: i it, for example, quotes Johann Nichols' book on predicatives Zymejima (Nichols but 1981), there i§ received jZvalgas author reserves for himself, presenting readers the old good Jablonski's grammar memory. It is a pity, because such a position in fact blocks the way for new research. »Dviybiniy linksniy* savoka is faulty also in that it confuses in one verb essentially different syntactic requirements, without leaving the possibility of construction with ir a (deceptive or complementary) secondary j predicate i§ (O Joab leaves the rest of his men Mieste BrB I Kron 12.8; kq wadini mane gierq BrB Mt 19.17) different from tokiy Gia direktinés ponky language jtakos could not be, because Adverbs used in the Gdansk Bible in relevant places (nago, blind). 119 Review constructions as neger jra szmogui wienam biti VE 50,,, in which the predicate of the communicative sentence is combined at best only with the co-referential complement of the main sentence. Because such a sentence syntax8can be interpreted by bit or as neger(a) is [A man being alone], or how bad it is Zmogui, [pRo, biiti vienam] according to this, the dative will be treated as the subject ar of the communicative sentence, as the complement of ar the main sentence. The meaning of the sentence requires choosing the first interpretation, e.g. non est bonum esse hominem solum, ob Kcddv elvat tov &vOpwrov pdvoy. Todél pretend that sennam drunk jiikis B 73 ir neger jra szmogui wienam biti is the very ta syntactic structure, possible in school, but not in scientific grammar. This, by the way, is not only a matter of interpretation, because a false theoretical assumption leads to factual errors, which are difficult to correct later. Let's say, in the ancient writings we find we »dual suffix* in constructions where the secondary predicate is combined with the complement: Ghys ... bagged the tushes VEE any such 159, constructions, in which the predicate of the common sentence is combined with the object of the final part of the main su sentence, we do not find in the whole of the ancient Lithuanian language —everywhere Gia is present, e.g., sagens iemus nessibijoti /liepe ios linksmais biti BP 1 83.15. Kodél gi Gia is not used as a,,double-end word* (*lepe ios linksmus biti)? It is true that now common language requires the use of the final verb linksnj instead of inagininko (forcing them to be cautious), but the construction presented for current use is purely artificial, created (most likely) by Jablonski and based on the false assumption that the construction Gia is the same as, ta say, made them careful, namely ,,dual end''?. »Regular* (t. y. not supported by historical tradition, but probably corresponding to the linguistic recognition of the,,spirit of the language‘) construction forced them cautious bit was brought together su ,,right’’ her, as "dual ending" interpretation, todél probably not applySkas bitty question, why Bretkin does not use construction *liepe ios funny beiti, although it combines démenis in the construction God has promised us sweethearted beasts (BP Il 253,4). Ambrazas He does not raise this question, ir contenting himself with citing examples of,,dual utilitarianism”, e.g., it is not good to be one man. The construction was told to be quoted in a funny way to discuss simply ir no place, because the fact that it does not have the binary ending required by the present grammar, obvious to everyone, admits that it is a predicative joiner in communicatio ve bati, also cannot be, because it would mean to admit that the su construction is in fact quite different from $i 2Su This artificially created structure fought Leonard Dambritine, which Danger, believers must be more careful, ir here it is not written. j ... biti akylesmais (UrnéjZitite 2006: 7). Histori8kai Zidirint he was right, because ,,dual tail” This construction is based on misunderstandings. 120 The constructions of the »dual verb« in the review made them cautious, 0 then it must be admitted that the interpretation of the dual verb is not adequate for the construction God has sworn to us to be lovely and kind. What to do then? It's best to keep quiet. The following is an article by the author (Holvoet 2004), which draws attention to the disturbing constructions of the old Lithuanian language “liepe ios linksmus biti nebuvimo faktas sitilomas ai8kinimas, ir Ambrazo reasonably not quoted. In fact, the discussion with the author of the article was not necessary, if only because the problem he raised cannot be formulated within the framework of Jablonski's grammar, and therefore its questioning itself is probably not desirable. It is worth to say a few words about the principles of syntactic description applied in the book, because they determine many formulations, both the development of individual constructions and the general tendencies of the development of Lithuanian language syntax. The three types of syntactic relations accepted in the Lithuanian grammar tradition—control, coordination and fusion—are abandoned in this historical grammar, and replaced by two simpler types of control. jo — ir Assignm- — ent of casting (p. 20). By definition, it corresponds to Lietuviy in works of grammar (e.g., & 2005) adopted the designation of management and modification. Similarly, the author considers the coordination as a formal expression of syntactic dependence (p. 20). But it is strange that only coordination is considered to be a form of expression of such dependence: usually in linguistic literature coordination as a type of morphosyntactic dependence is mentioned together with morphosyntactic governance (English: government, French: rection, German: Rektion). However, morphosyntactic control (rection) is traditionally not distinguished from syntactic control (complementation) in Ambraz’s book. This, by the way, is the only thing that is unmodified transferred from the old syntactical system. The relation of the factor and the adverb,,interconnection“ is maintained (p. 21). This characteristic is taken from the i8 DLKG concept, according to which the verb is in a certain sense syntactically dependent on the factor, since it is combined with it. Since the factor in turn depends on the action word (the latter opens the position of the ne denominator factor), we get mutual dependence. Thus, when we abandon the combination as the third type of dependence, there is no,,interconnection- ”, the combination of the verb 0 with the factor is a classic example of the main demiurge. Since the bibliography mentions Johann's Nichols may 1986 article,,,main démens The author's own personality has to be known to him, even though he ignores it. In the arsenal of the old savoky there is also a word and a,,double dependency relationship”, which at the same time connects the tactical signs with both the tactical and the factor or complement. Traditionally $i is based on the combination of a predicate with a substantive or adjective, e.g. 121 Recenzie nau jj jaunq: since traditional conjugation is considered a type of syntactic connection, then in this construction the predicative adverb jaung must depend on the adverb ji (dél conjugation), but at the same time it must also depend on the taric verb (otherwise constructiony maéiau brolj linksmq maciau linksmq brolj ir difference can be syntactically explained). But by deleting the combination of syntactic relationships in the type list, there is no reason to talk about,,dual dependency*. In dependency grammar, double syntactic dependence is not recognized, not quite clear, as the author himself understands, for example, p. 491 we read:,,The second noun word in this type of sentences [i.e., sentences with adjectives — A. H.] has two-fold syntactic relations: it ja Melts to the first [substantive] as a certain prefix is also semantically associated ir with the verb*. But if the noun and the verb, as the author claims, have a semantic connection, then he probably no longer postulates a syntactic connection between them, so it is incomprehensible why the syntactic connection should be double? Let's assume that the author simply formulates incorrectly, and that along with semantic dependence, facts postulate syntactic dependence. But then we have different criteria for determining syntactic dependence: du (i) in one case (jj jaunq) i8 the morphosyntactical dependence of the words is inferred from their syntactic dependence, and (ii) in the other case (see jaunq) such an inference is made despite the fact that the mentioned semantic relationship between you and the words is apparently already considered as a sufficient basis for the — determination of the syntactic relationship. It should be understood that such a determination procedure and inconsistency is theoretically undesirable and in practice leads to unpredictable consequences. ir For it is hard to deny that, for example, in the sentence nudazé namq Green words namq ir The greens are connected by semantic connection, because the house is made Zalias, in ir many languages This connection is ir undergoing debugging. (latv. 100 mg/ day) If the alignment is a sufficient but not necessary condition for the establishment of a syntactic connection, then based on semantic connections it can be freely stated that the construction nudazé namq ir Greens have a double dependence. But then the whole system, postulating a syntactic difference between the construction circumstance su (time bottle stations) constructions ir with a nominal su value (time bottle stations), gritiva. These long digressions are necessary here in order to understand how to evaluate ir the na i§ the main thesis of the work in question — about 'strengthening governance'. The author relies on quotation i Antoine'o Come here. comparative grammar Access 1934) allowed: »ll there was no Indo-European en de ,,rection* of one word by another, en a in Latin for example; un autonomy word is 122 Reviews comme il y principle that commands structure phrase Indo-European“ du le la de la (p. 196, i8naga). It is not entirely clear whether Meillet meant syntactic or morphosyntactic dependence, although the term,,rection* has a long tradition. (t. y. Morphosyntactic) control. The author of the book in question, deciding on some formulations, has i8 in mind both morphosyntactic dependence: e.g., in the formulation,,Going with conjugated words prepositions control them, su t. y. lemia ju gramatine forma’ (p. 283). Unsyllable The word, apparently, cannot be controlled, so in the constructions from here, through where and so on, we probably have a fusion. But what then worth p. 20 the definitions provided? There is talk about syntaksinj requirement or absence, already jo o p. 283 turns out that it's a matter of morphology: if the dependent The word is inclined, we have control, if ne — or Fusion. But let’s assume that the definitions live their lives somewhere in a separate corner of the author’s conceptions have no use. ir Net taxes ir su such a clause is difficult to fully understand, the author has ka in mind when stating that esa ,,In the development of Lithuanian language there is a noticeable strengthening of the linksniy syntax dependence on verbs“ (p. 197) (enlarged hand) ~ A. H.). After all, verb belonging was not equal to word belonging, so the author Gia faktiSkai talks about rekcija, about complementacija. ne When reading the text it seems that the author himself has ir not decided what is the strength in the development of Lithuanian language. Because in specific cases there is a significant complementation:,,In general, in constructions with a time and place suffix, the same direction of development is observed: from the ta circumstance of the attachment to the verb, the direct complement is i controlled* (p. 206). This statement is based on such examples as I spent the night, Stare for a year, etc. Further apart from the special operative type, in which the duration involves the argument structure of the verb j (These verb types may be new, but the ir emergence of them is still a jy matter of ZodZiy verbs, only indirectly —syntaksés), o not to see what has changed. In construction miegovisq naktj jo Zodis naktj turbit ir still remains as buves ir ,,pri8lietas"? In any case, ar the transformation into a direct complement is, be without doubt, a syntactic ir jo process, postulating that we are talking about the emergence of a complementary relationship. In other places the choice of verb is also explicitly intended:,,With the weakening of the preposition of the partitive noun, the tendency to ir generalize either one or the other verb, i.e. the verb, in view of its action, becomes stronger. In this way, the semantic coordination of verbs is replaced by strong syntactic control.” su (p. 224). So controlled no longer Zodis, o funny. If the weakening of the partitive nobleman 123 The review and the final prepositions (remember brother: remember brother) summarize the final preposition (remember brother), this changes only the complement sign, but it is not possible to talk about the appearance of the complementation relationship. So what i8 is really changing? Gal ,,management“ means the consistent use of a certain verb to complement a certain lexical unit Zyméti? Or is it that the choice of a verb is not determined semantically? Good to see that suzZinojus. Amicus Plato, sed… No one doubts the importance of Antoine Meillet’s great work for the development of historical linguistics, but his syntactic statements are no longer worth quoting at the beginning of the twenty-first century. As you ir understand the above jo quoted statement, whether it is related to syntactic or morphosyntactic dependence, it is probably incorrect in both cases. First, if ,,control“ we perceive as syntactic dependence complementa- —- tion that according to — The Meillet reconstructed protospeech has different structural characteristics than living languages. These pro-language words were syntactically independent and did not have the dependency relationships found in modern languages. Of ju course, it is possible to make such a claim, but then we go back to the early comparatist j (Boppo ir Schleicher), when it was still widely accepted that languages may have different stages of structural change at different stages of development (in the periods of “childhood” and “old age”). ir From the neo-grammatical time implicit, from 0 XX a. on the other hand, the,,uniformitarian principle“ is explicitly observed (cf. Harris & According to 1994: 24-32), Campbell, linguistic interpretations must be the same regardless of whether the current language is studied or a ar hypothetical ancient protolanguage. ar Nothing that is not possible now should be postulated by studying the development of language in the past. The idea of a ir primitive, grammatically unformed language may never have been completely condemned, but at least ir Linguistics is ji currently ty under serious consideration. Such a pro-calve stage of development, when the modifications of complementation ir are of their own accord not applicable to it, is for us only a matter of semi-scientific speculation. So there's a second ,,management* concept in - morphosyntaksis. Bet ir ¢ia Meillet's statement seems highly doubtful. [at various stages of the development of a particular language we find a verbally-coded semantic ir means of eliminating structuririniy verbally-coded (e.g., the final of an object) in the case of a force, but we find the opposite in ir the case of a tendency. For example, the partitive designation of an object (i8 whose iSrutuliojo negation ir originator) in balty slavy languages ir greatly strengthened, most likely encouraged by abliative genitive syncretism ir (abliative is well demonstrated partityviniy constructiju Saltinis): this is clearly seen when compariir ng slavy balty su languages ir senovés graiky lo124 Reviews in tynu kalby texts. In the position of the predicate there is the nominative (final) Slavy and Balty language jsivedé jnagininka, which initially was associated with the i8 Signification of additional semantic su content (now the function of the predicative jnagininka is almost completely lost in both languages). In other words, in the Baltic and Slavic languages one syllable-coded semantic extension disappeared, but another appeared. Of course, the decrease of the semantic tension encoded by the vowel is also seen in those languages where the vowel is no longer present, such as the present Romance languages, but it is difficult to foresee the further spread of the idea of morphosyntactic »tectics«. Turning specifically to the thesis of the book in question, its evaluation is complicated by the fact that the author does not define,,verbal control* more precisely —as has already been said, his concept of verbal control is intertwined with the purely syntactic nature of complementary requirements. The examples given are extensive and show that the author does not distinguish syntactic control — (complementation) from morphosyntactic control (rection). ir If the above-mentioned thesis of the book should be understood as meaning that in the development of Lithuanian language syllable encodes semantic prepositions of ir subject, object, etc. In the case of the expression, it is reduced, so such a statement is theoretically meaningful, but probably false in fact. For example, the final and partitive noun prie8prieSa in Lithuanian language is still alive until Siol. In what sense has it been weakened and given way to,,government''? Of course, it is possible to find a single case when the alternance ir of the nobleman's last name was cancelled. According to the author, the origins of partitivity are with the action Zodiiais atminti, ùmirsti used as a pronoun (atmink mane BrB I 40,14), which in the present language is replaced by the adjective. But the competition ir of the nobleman's adjective is already noticeable with the Latin memini, the Greek juysvyoKw. Thus, this case is so old that probably already in the ancient Lithuanian language such a originator was a semantically unmotivated syntactic relic (the nature of partitivity can only be speculated), the reduction from the perspective of the Lithuanian language was jo merely the replacement of one semantically ir jo unmotivated iS type of »tection« by another (the principle of economy must have worked: instead of two erections, one type remained). ju Siuo poZiiiriu The origins of the meaning of ablativity are slightly different. About fear, avoid type verbs author rao:,,The originator, previously had the meaning of separation, over time is drawn to the semantic and syntactic field of attraction of the verb, loses its independent meaning and becomes a controllable complement verb* ir jo (p. 245). It is hard to believe that in the construction to fear the wolf the originator of the wolf is an addition to the existence of fear in the presence of,,management,, — such an opinion is a part of the primitive, syntactically unconnected, pro-cabbalistic to myth itself. Probably this ancestor in the construction to fear the wolf always Zyméjo addnj. The only question is, whether it (i.e., the slant) is now or /and in the old scripts controlled or not. Siuo 125 Reviews (eds.), Studies in Baltic and Indo-European linguistics in honor of William R. Schmalstieg. Amsterdam & Philadelphia, 75-90. Howvoer, A. 2004b: The category of evidence. In: A. Howvoer & L. Seménrene, red., Gramatiniy kategorijy tyrimai. Vilnius: LKI, 105-120, Hotvorr, A. 2007: Mood and Modality in Baltic. Kraków: Jagiellonian University Press. Forage, A. & Juzentis, A. 2005: Syntax dependency types* additions ir clarifications. In: A, Hot feed & R. Mikutskas, red., Gramatiniy funkcrju tyrimar, Vilnius: LKI, 11-38. Litvinow, W. P. 1989: Der Modus relativus of Baltic languages from typological Sight. Baltistica 25 (2), 146-155. LKG = K. Ulvydas, red., 1965-1976: Lietuviy kalbos gramatika. Vilnius: A thought. Meier, A. 1934. Introduction the a comparative study of Indo-European languages. 7° éd. Paris: Hachette. Moravesix, E. 1978. On the distribution of ergative and accusative patterns. Language 45, 233-279. Nicos, J. 1981: Predicate Nominals: A Partial Surface Syntax of Russian. Berkeley: University of California Press. Nicots, J. 1981: Head-marking and Dependent-marking Grammar. Language 62, 1, 56119. Supatant, M. 1998. Voice parameters. In: L. Kuukov & H. Varer, eds., Typology of Verbal Categories. Papers Presented to Vladimir Nedyalkov. Tiibingen: Niemeyer, 117-138. Look at you, R. 2006: Leonard Dambriiinui — 100 perfect. Native language 2006 February No. 2 (464), 3-7. Axel Holvoet Lithuanian Language Institute P. Vileisio g. 5, LT-10308 Vilnius, Lithuania 132