Dėl sintaksinių ryšių skirstymo

ACTA LINGUISTICA LITHUANICA LII (2005), 157-166 Remarks on the classification of syntactic relations The above-mentioned thoughts were raised by the interesting commentary of Prof. Vytautas Ambrazos (ALL 52, 167-170) on the 2003 Syntax Relations Research. This review, which contains valuable polemical remarks, especially on the issue of the aforementioned publication Theory of Syntactic Relations, shows that the author’s goal —to stimulate theoretical discussions on the issues of Lithuanian grammar—- has been achieved. It is particularly pleasing that the discussion is opened by a significant authority in synchronic and diachronic Lithuanian language syntax, the author of many classical works, Prof. Vytautas Ambrazas. The professor himself has frequently participated in seminars of the Grammar Department of the Institute of Lithuanian Language, which have helped him to crystallize as a favourable, but selective and critical listener. It is to be hoped that an increasingly broader circle of researchers will be involved in the discussion. Not one syntax theory is perfect, not one is perfect, and in the Syntax Relations Studies the theory of syntax relations is presented, which for convenience I will abbreviate as:,,SRT theory“. | some of the problems associated with it Prof. Ambrazas rightly draws démesi. Hard to deny these problems, the worst bitty —withSweeping them under the carpet. However, the authors of the SRT concept decided to settle with them on the principle of the,,lesser evil*. Therefore, I would like to draw attention here to some of the greater evils that can be avoided by the SRT theory. The reviewer notes that the author of SRT uses very broad control and modification features. For example, the modifiers are considered both the circumstantial, the circumstantial, and the tactical modifiers (here it is necessary to clarify: the modifiers according to the SRT theory are considered only the free predicates that do not require a verb). The validity of this part of the remark is difficult to deny: for example, the »double syntactic relationship«, according to which the DLKG distinguishes between nominal parts of a verb and verb signs, is rejected. The problem is that this dual syntactic relationship leads to difficult to resolve confusion. Conjunction is defined by two features: it is connected to the verb and is combined with a factor (Broliai grizo links) or a direct adjective (Aš visada matau Jona links). This is how the case-mark differs from the circumstances, for example, in the sentence NudaZiau duris Zaliai. In this sentence the adverb Zaliai néra is combined with nothing, so the syntactic relationship is homogeneous. But why can't we say that the adverb Zaliai is joined to the noun duris as well as to the verb? After all, semantically speaking, the meaning is attributed to the door, not to the painting: Nudaziau duris Zaliai means ‘by painting I made the door Green’. Such a statement cannot be proved, but it cannot be refuted either. For the fusion is such a,,syntactic connection“, which is not visible — it does not manifest itself in morphological features. Thus, the statement that in the sentence I painted the door Green the word Green is linked by a double syntactic connection is just as true as the statement that it is linked by a single syntactic connection. 158 | REMARKS connection. The fact that the same statement and its prefix can be equally true is, of course, undesirable. Here it is necessary to immediately clarify: we do not see anything wrong in the fact that a syntactic type of relation without morphological significance is distinguished. In the SRT concept, syntactic relationships are differentiated without considering their morphological characteristics. The difficulty lies in the fact that purely morphological phenomena are raised in relation to syntactic dependencies. The relationship between morphological dependencies (when the variable form of one word is determined by the morphological significance of another word) can be conventionally called the type of syntactic relationship, but it is forgotten that the heuristic procedures are completely different, whether we are talking about morphological significance — harmonization, or about meaning without morphological significance — fusion. In the case of matching, a morphological connection is established and a conclusion is drawn about the existence of a syntactic connection: this is the basis for defining both the mutual connection and the double syntactic connection. In the absence of conjugation, the verb may be perceived as a verbally controlled, 0-word sign—as a Merge with the verb. In the case of fusion, the procedure is reversed: it is literally assumed that there must be a relation of syntactic dependence between the two words, and since this relation cannot be either control or coordination, it is assumed to be a fusion. For example, in the sentence It quickly escaped it is assumed that the verb quickly is dependent on the verb, and this dependency relationship is called a fusion because it has no morphological indicator. We do not want to rule out this assumption. In the sentence above, the verb must be dependent on something quickly, and the assumption that it is dependent on the verb atbégo seems quite plausible. This assumption is also valid in the case of the sentence NudaZziau duris Zaliai. Also it is based on analyzing the sentence Always see Jonq linksmq. The adverb happy here depends on the verb see, and this is enough to find the place of Siam adverb in the structure of the sentence's syntactic hierarchy. To make the additional assumption that Sis is still dependent on the noun Jong is completely unnecessary. It is true that the form of its number and gender depends on the number and gender of the word Jong, but such morphological dependence can link even all the words in quite different sentences (e.g., On the table lay money. [I sat them j kiene], 0 in such case no one sees the need to establish the relationship of syntactic dependence between them. The worst thing is that by postulating a syntactic relationship between the words Jong and linksmq in the sentence I saw Jong linksmg on the basis of coordination, it is naturally assumed that a syntactic relationship also exists between the words door and Green in the sentence I painted the door Green. After all, everything here is similar both semantically and syntaxically; the only difference is in the language: either (variable) adjective, or (unvariable) adverb is used. But can such a difference be syntactic? Probably not, so given the fact that syntactic dependencies can be two-sided, i.e. one word can depend on several other words, it is natural to assume that the words door and green are connected by a fusion relationship. | but it can be answered that in this case (as opposed to, for example, in the sentence He quickly fled) a priori there is no basis for a syntactic connection. But then there was no such foundation and construction I saw Jonq funny @ case. In this construction we see only a morphological connection, which —as has already been observed—does not presuppose a syntactic connection. As we can see, the logical errors in the traditional syntactic relationship model arise due to the treatment of purely syntactic (fusion) and purely morphological (harmonization) functions on one level. I§ esmés néra grounds to draw combination and syntacsiniy relationship distribution scheme. Prof. Ambrazas was not silent about this in his commentary; However, the interrelationships and characteristics of the bilingual syntactic relationship are based on the false assumption that the combination is a syntactic relationship, REMARKS | 159 or assumes a syntactic relationship. If this assumption is abandoned, it becomes necessary to operate on the properties of interlinking and bilinear syntactic relationship, and then in the traditional syntactic relationship scheme only two properties remain: control and fusion. And these things should be defined exactly as they are in the SRT concept. According to Prof. Ambraz, the application of the control method only to one-way connections (leaving the interconnection and the two-way connection as separate types of connection) “permitted to realize the same author’s recognized need to define the functions of the sentence unit on the basis of syntactic connections”. In fact, the ability to distinguish sentence parts according to the form of the syntactic relationship is almost the same in both syntactic relationship models, both those proposed by Prof. Ambraz and those of the authors of SRT. The concept of SRT does not distinguish between the predicate, complement and binary predicate according to syntactic relationships. Instead, it distinguishes complements from circumstances and bit predicates from free-flows, which the DLKG iSdestytas model cannot. According to Prof. Ambraz's comment, an improved version is proposed, when control and fusion are no longer defined according to morphological criteria, in fact, it becomes possible to distinguish complements (controlled) and circumstances (merged) according to syntactic relationships. It is not possible to do this in the case of,,tarininiy pazyminiy“ and,,tarininiy noun part“. It is stated that in the sentence Jonas is happy the word linksmas is not a separate part of the sentence, because the connection bai has no independent meaning. The sentence part is recognized as,,tactical sign* in the sentence Jonas grizo links. And what will be the function of the word happy in the sentence John looks happy? | The DLKG gives two different answers to this question: p. 507508 the verb seem to be mentioned between non-verbal conjunctions, 0 p. 589 the sentence The teacher seemed angry is given as an example of a verbal noun. IS In this case, the second solution is preferable, because it is difficult to say that the verb appears to have no meaning, similar to be. According to the SRT concept, the binomial happy in the sentences He looked happy and He grew happy is associated with different syntactic connections — in one case, control, in another — modification. This explains why the adverb is binary in the first case (*It seemed to be an incomplete sentence- ), 0 in the second —no. This difference is quite analogous to the difference between complement and circumstance, which Prof. Ambrazas himself is more inclined to explain as a difference in syntactic connection. But in the case of,,tarininiy pazyminiy“ this difference is hidden under the same label,,double syntactic connection“. The fact that the double syntactic connection is better to be abandoned, because it leads to substantial confusion, has been pointed out earlier. Likewise, as with the two-way syntactic relationship, the relationship between the two is faulty. Therefore, it is doubtful whether the concept presented by DLKG allows to distinguish factors from complements according to syntactic relations. [only generative grammar, which is based on direct analysis and already at the first stage of division breaks the sentence into verb and verb-subject combinations, is able to do this. The authors of SRT do not use this method, therefore they acknowledge that the difference between the substantive and the adverb is not a difference in the type of syntactic relationship. The only syntactic difference between them concerns the position of jy in the obliqueness hierarchy—the other differences are morphosyntactic. In Prof. Ambrazo- ’s opinion, the interconnection of the two helps to avoid the term,,pre- -biotic circumstances*. This term is used in the SRT concept in an appendix, for example, at eZero in the sentence The house is by the lake. It must be admitted that the term itself is not very successful, but it can be replaced with a better one. One thing is the term, another is the essence of the thing. The essence is that the case of “necessary circumstances” is intermediate between control and modification. The sentence is locative, and the place noun obviously takes the position of the argument. The position of the argument is taken Anyway, it is good for complements, but to call the lake complement prevents a certain pronunciation anomaly: in reverse pronouncing the predicate should be formulated as “be near’, and this predicate opens the argument 160 | NOTES positions for object and place names. But the spatial relationship is not determined by the verb, but by the preposition used with the noun. Thus, by the meaning and the form of expression, the word “at the lake” resembles a circumstance (more generally —modifikatoriy), —so much so that in this case the modifier means not an additional, but the main predicate of the sentence. To be more precise, the predicate reSkia not the noun pats ežeras, but the preposition prie. But the usual rules of syntax splitting do not allow the sentence to split The house /is by /lake. It is accepted to divide differently: House /is /by the lake. Such an argument, syntactically realized as a modifier, is called a bit circumstance according to the principle "whatever you call it, it won't hurt". The fact that there may be an intermediate case between control and modification does not contradict the theory. The boundaries of language categories are generally blurred; Most importantly, the tools used by the researcher were bit astris. And the alternatives to the “probable circumstances”* are not much better. By stating that in the sentence The house is by the lake the word sequence by the lake is not a part of the sentence, but only a part of the part of the sentence, one simply avoids the answer to a tricky question. How to explain that the word sequence at eZero, which in one sentence (The house is by the lake) is,,nothing”, in another sentence (He built a house by the lake) becomes part of the sentence? This question can be answered by saying that in the sentence House is at eZero in fact is at eZero and is the true predicate; This is confirmed by the possibility of skipping the verb to be present tense form — a phenomenon to which Prof. Ambraz attaches great importance. SRT author’s decision to consider the binary predicate with the verb biti as controlled he commented as follows:,,- Then it must be assumed that in the sentence Knyga idomi nuliné conjunction (i.e., the indicator of time and sequence without lexical meaning) controls both nominatives as binary additions to the verb’s meaning- *. But if the zero connection is a problem for the SRT concept, it is also a problem for the DLKG and Prof. Ambraz concept. If the null connection cannot control, then it cannot go into the interconnection — in other words, the factor of the sentence Knyga jdomi cannot be syntactically defined. As for the meaning of the conjunction, it is important to understand that whether we assign a lexical meaning to it or not is a matter of convention, determined by the consideration of the consistency of the theory. The conjunction biti does not always connect reference statements with predicative ZodZi (as in the example John is happy); Sometimes it also combines reference statements with reference statements, e.g. in the sentence John is the miisy teacher, where the miisy teacher is not a predicate, but a definite description, which denotes a person. In this case it expresses the identity of two said referents (‘The person, named John, and the man who teaches us, are one and the same person.’) The assumption that the conjunction has no meaning, in this case does not give good results, because the sentence cannot consist solely of referring sentences and not have predicative content. In other words, at least in some cases the compounds must be assigned a lexical need. In any case, lexical use has to be attributed to,.non-verbal connections* (DLKG 507-508), such as seem, pretend, etc., whose additions in the DLKG concept are lost in no man’s land somewhere between,- ,the verbal noun part“ and,,the verbal subjunctive“. Acknowledging that both biiti and atrodyti have a certain lexical meaning, it is possible to treat their additions equally as subordinates.] In such sentences as John is happy, this introduces a certain redundancy, because it must be acknowledged that the word /inksmas itself already has a predicative meaning. But the redundancy of various evils is probably the least, because no one will deny that the need for redundancy in the language is there. From a semantic point of view, the connection in the sentence John is happy is, of course, superfluous. But a syntactic expression that pretends to be it causes a much worse overstatement. 161 The search must be tested, and the confrontation of ideas and concepts will one day, no doubt, lead to the formulation of a better theory of syntactic relations than that formulated by the SRT. Axel Holvoet Received 2005 05 13 Lithuanian Language Institute P. VileiSio g. 5, 10308 Vilnius, Lithuania MAZMOZIS 7 We have known for a long time, we were taught, that the sonant n allophone 7 is soft only before k’ (this was investigated by prof. Aleksas Girdenio (Kalbotyra 48-49 (1), p. 165-167), presented in the author’s Lithuanian Language Dialect Zodyne (2001), This is the way it is pronounced throughout Lithuania (some southern highlands, called dziikai, due to the influence of Slavy jj change 7). Now in the journal Lituanistica (2004, 4 (20)) appeared (after research work of Vilnius Pedagogical University) Arimantas Raskinas, Gailius Raskinas, Astos Kazlauskienes article,,Lietuviu bendrinés Skekamosios kalbos garsy fondo kirimo principai* (p. 53-62). It states, probably according to prof. Antana Pakerj (as prof. A. Girdenis' quoted passage shows), that in common language ay’ks’t’ i, tan’kmé,, dy’ks’t’is (p. 56) should be pronounced, 0 usually everywhere lietuviy is pronounced anks'ti, tagkm’é:, d.yks'vis and so on. It is an unbelievable proposition and a negative thing (perhaps in that article the noun with the suffix -ybé priegaidés —tikimybé (p. 61) = probability should not be changed). IS VPU through the pupils injustice comes and 1 LKI: in the important publication Lietuviy kalbos tarmiy chrestomatija (2004) we find /uj'.kt',,lenkti* (SirvintiSkiai) (p. 162), an'ki'@du (Pakrazantis) (p. 215), dey'.kt' (p. 18 —theoretical part). It is even going to the book consultant Prof. habil. dr. A. Listen! There is a clear bé:k't' (p. 62). And this is SiauliSkiu- ... MAZMOZIS 8 Zemaiciy Naureje (all dounininkai, nauriniai dinininkai, vad. varni8kiai, Siauriniai donininkai —from Satigy j Siaure) iS ally briefy and tvirtagaliy finaly detracts kirti j Zodzio pradzia (universal detracting of the word). In those snippets it can also be that the conjunction is also given to the variable prefix of the word (this is observed by Jonas Jablonskis, editing Antanas Juška's zodyna). In the third volume of Lithuanian Linguistics (1960) (p. 62-63) it is written about this in the Zemaičiai — Kurvénu, Papilés, Kaunatavos, Raudény and other snekty, read about Kalnalis (P. Kniaikstos), Latikuvos (V. Grinaveckio), even in the west of the upstate Sakyna (A. Jonaitytes) Snekty similar phenomena, his own pronunciation used in the work: near Siupyliy near Gruzdzit. Many dialectologists, especially akad. Zigmas Zinkevicius, authors of the first Lithuanian dialect chrestomacy. The dialectologist Aleksas Girdenis and his student Dr. Lina Muriniene have now written extensively about this issue in the article,,Peculiarities of Kircio attraction in the Eastern Siaurés Zemaiciy dialect“ (Baltistica 36 (1), 2001, pp. 87-102). Already faded since the 1960s, the problem raised is discussed very clearly, with various encrypted examples, even the frequency is indicated,