Prijungiamasis ryšys sovietų okupacijos laikų lietuvių sintaksės darbuose

ARTŪRAS JUDŽENTIS Wróblewski Library of the Lithuanian Academy of Sciences doi.org/10.35321/bkalba.2022.95.08 SUBORDINATING RELATIONS IN LITHUANIAN SYNTAX WORKS FROM THE SOVIET OCCUPATION PERIOD KEYWORDS: subordination, coordination, government, adjoining, Vytautas Ambrazas, Jonas Balkevičius, Pranas Gailiūnas, Zelma Dumašiūtė, Česys Grenda, Bronius Kalinauskas, Vitas Labutis, Vytautas Sirtautas, Juozas Žiugžda. ANNOTATION The article surveys how methods of subordination were described in Lithuanian syntax works during the Soviet occupation, approximately from the mid-20th century to the end of the 1980s. It begins with Juozas Žiugžda’s Grammar of the Lithuanian Language (1945), and ends with the textbook for higher education by Vytautas Sirtautas and Česys Grenda, Syntax of the Lithuanian Language (1988). It examines the concepts and definitions of methods of subordination and their application in describing relations between words, word combinations, and sentence constituents. ABSTRACT The article analyses, how the types of syntactic relations are described in the works on Lithuanian syntax, during the Soviet occupation, during the years between the middle of the 20th century ant the end of eighties. The investigation opens with the Grammar of the Lithuanian language (1945) by Juozas Žiugžda and ends with manual for higher schools Syntax of the Lithuanian language (1988) by Vytautas Sirtautas and Česys Grenda. The concept and definition of subordination types and their application when describing relations between words, word groups and sentence parts are taken into account. INTRODUCTION This survey continues the study of the description of syntactic relations in Lithuanian linguistics, begun in the author’s earlier publications: The Syntactic Relation between Subject and Predicate: Changing Conceptions (Judžentis 2020) and Subordinating Relations in Lithuanian Syntax Works (First Half of the 20th Century) (Judžentis 2021). ARTŪRAS JUDŽENTIS. Prijungiamasis ryšys sovietų okupacijos laikų lietuvių sintaksės darbuose 133 An examination of Lithuanian syntax works from the first half of the 20th century has shown that they discuss only two types of subordination—agreement and government—although these are often not yet understood as varieties of subordination (or dependency) (Judžentis 2021). Agreement is mentioned in all Lithuanian syntax works of that period; they discuss agreement between the subject and predicate, and agreement of the attributive modifier with the word it modifies. Government is rarely understood as a syntactic relation. Most often, it is either not mentioned at all or is mentioned only in discussions of types of objects. Works from the period under review tend not to distinguish syntactic relations from their morphological expression (agreement in gender, person, and number; governed case forms). In addition, especially in the case of government, the syntactic nature of the relation between words is not distinguished from its semantic basis (case forms are required). This article surveys significant articles, school and university textbooks, monographs, and academic grammars from the period of Soviet occupation in order to explain how the system of three varieties of subordination—agreement, government, and adjacency—emerged and came to prevail to this day. The survey begins with Juozas Žiugžda’s Lietuvių kalbos gramatikos (1945) and concludes with Vytautas Sirtautas and Česys Grenda’s university textbook Lietuvių kalbos sintaksė (1988). After the Soviet Union occupied Lithuania and many Lithuanian linguists fled to the West, Lithuanian linguistics, like the nation as a whole, went through difficult times. The scholars who remained had to adapt to the conditions of life and thought imposed by the occupation, and to the tenets of the new ideology. Before long, the study of Lithuanian syntax, too, had to conform to the approach used to describe and explain syntactic phenomena in Russian. Paradoxically, the traditions of syntax scholarship in independent Lithuania were maintained in occupied Lithuania—and the subject’s continued presence in schools was thus sustained, as it were—by J. Žiugžda, one of the occupation regime’s ideologues. It took roughly two decades for Lithuanian linguistics to adopt Russian linguists’ definitions and classification of syntactic relations. Bronius Kalinauskas was perhaps the first to undertake this work, but Pranas Gailiūnas adapted it for secondary schools, and Jonas Balkevičius for universities. Thus, by the mid-1960s, a three-part classification of syntactic relations had become established in Lithuanian linguistics. It seems that B. Kalinauskas, drawing mostly on the work of Russian linguists, also helped popularize the study and description of word groups in Lithuanian linguistics. This approach was later taken up and developed primarily by Vitas Labutis and the academic grammars of Lithuanian. This topic remained current until 134 BENDRINĖ KALBA | 95 the end of the 20th century. Another approach to syntactic research and description, inherited from the period of independent Lithuania, focused on sentence constituents. It was developed further chiefly by Vytautas Sirtautas. Sentence constituents continue to be studied, and debated in Lithuanian linguistics, to this day. The author understands syntactic relations as a phenomenon at the level of syntax, grounded in semantic relations and most often expressed through morphological forms. Two types of subordination are distinguished: government (complementation) and modification. Government is a syntactic relation in which one word requires another because of its meaning. A relation in which no additional word is necessary to a given word either semantically or syntactically, but is nevertheless syntactically linked to it, is called modification. I propose that syntactic relations be strictly distinguished from their grammatical expression. Two types of such expression are distinguished: the selection of a case form and agreement. In the first case, the form of the dependent word is selected according to the syntactic and semantic properties of the head word. In the second, the form of one word is selected according to the morphological features of another semantically related word. The aim of this review is to trace the origins of some of the key concepts in 20th-century Lithuanian syntax and to examine how the conception of syntactic relations that remains predominant today came about. This aim also determines the tasks: to describe the syntactic relations distinguished in each work on syntax; to compare them with those described in earlier works (in terms of inheritance) and later works (in terms of possible influence); and to identify the presumed origins of their theoretical foundations. The review also seeks to examine how successfully Lithuanian linguistics in the mid-20th century managed to distinguish the levels of description and analysis associated with semantics, syntax, and morphology. The methods are standard for works of this kind. First, each work included in the review is discussed separately; it is then compared with earlier and later works to establish its place in the development of syntactic theory. Finally, the works of the period under study are considered from a broader perspective, and the characteristics of the development of syntactic thought and trends in the changing theoretical foundations are identified. In addition to the introduction, the article comprises the following sections: Continuity of the traditions of syntax scholarship in independent Lithuania and adoption of the Russian classification of syntactic relations, Syntactic relations as the structural basis of word groups, and Syntactic relations as a property of sentence constituents. The article concludes with a summary. ARTŪRAS JUDŽENTIS. Prijungiamasis ryšys sovietų okupacijos laikų lietuvių sintaksės darbuose 135 1. CONTINUITY OF THE TRADITIONS OF SYNTAX SCHOLARSHIP IN INDEPENDENT LITHUANIA AND ADOPTION OF THE RUSSIAN CLASSIFICATION OF SYNTACTIC RELATIONS In the 1945 edition of Juozas Žiugžda’s Grammar of the Lithuanian Language, intended for schools, syntactic (or word) relations—agreement and government—are described in several places. However, they are not called syntactic relations, but are regarded as means of expressing semantic relations between words. Apie derinimą kalbama skyrelyje Pagrindinių sakinio dalių derinimas (Žiugžda 1945: 14–15). Jame pasakyta, kad grynasis tarinys arba sudurtinio ir suvestinio tarinio jungtis su veiksniu, išreikštu vardininko linksniu, derinama asmeniu ir skaičiumi. Be to, vardinė sudurtinio tarinio dalis su tokiu veiksniu derinama skaičiumi ir linksniu, tačiau gimine jie gali būti ir nesuderinti, pvz., Šarka – paukštis. Svarbu, kad tarinio vardinės dalies derinimas su veiksniu niekuo neskiriamas nuo grynojo tarinio ar suvestinio ir sudurtinio tarinio jungties derinimo su veiksniu. Apie valdymą rašoma skyrelyje Žodžių valdymo būdai (Žiugžda 1945: 19–20). Pradedama nuo to, kad dažnai žodžiai valdo ne vieną linksnį, tačiau „iš tų valdymų galininko valdymas yra svarbiausias“, todėl toks valdymas vadinamas tiesioginiu. Jis yra „[V]isuomet būtinas todėl, kad to veiksmažodžio prasme be galininko visuomet bus lyg nebaigtinė, neaiški“. Kitų linksnių valdymas vadinamas šalutiniu. Jis yra „[N]e visuomet būtinas todėl, kad to veiksmažodžio prasme ir be minėtųjų linksnių, tik su galininko papildiniu, gali būti visai aiški, lyg baigtinė“. Tiesioginis valdymas nuo šalutinio skiriamas ne pagal linksnio žymėjimą, o pagal būtinumą. Štai būdvardis pilnas tiesiogiai valdo kilmininką, palankus – naudininką. Ir veiksmažodžiai, ir būdvardžiai tiesioginio valdymo gali ir visai neturėti, „[I]ie ir be papildinio esti pakankamai aiškūs“. Netiesioginiu valdymu vadinamas prielinksnis linksnių valdymas. Atkreipiamas dėmesys, kad valdyti gali ir kai kurie daiktavardžiai, skaitvardžiai, tam jiems nebūtina eiti tariniais. Trejopas valdymo skirstymas buvo lietuvių sintaksės mokslo naujiena. Jį perėmė Bronius Kalinauskas. Apie kokią nors trečią žodžių santykių atmainą šioje gramatikoje nekalbama. In the 1954 edition of Grammar of the Lithuanian Language, J. Žiugžda expanded and reorganized the section on government. He added information about relations between words, including agreement, and entitled it Relations between Words in a Sentence. Agreement and Government (Žiugžda 1954: 16–17). In the author’s view, semantic relations are expressed by word forms: in a sentence, words “are linked to one another by their meaning, and relations of meaning 136 BENDRINĖ KALBA | 95 in a sentence are expressed by their forms.” Two types of relations between words are distinguished: “There are two ways in which words relate to one another: agreement and government.” Thus, agreement and government are understood as ways of expressing semantic relations between words. Definitions of these types are provided. “Agreement is a relation between words in which one word (an adjective, numeral, pronoun, or participle) always takes the same form as another word (a noun).” In addition, agreement is also recognized between verbs (predicates) and subjects expressed by pronouns (but not by nouns). “Government in a sentence is a relation between words in which one word takes the form (case) required by another word’s meaning.” Verbs and some adjectives can govern. As in earlier editions of the grammar, direct, indirect, and prepositional government are distinguished. This classification and description of syntactic relations (more precisely, of ways of expressing semantic relations) also remained in the 1960 edition of the grammar (cf. Žiugžda 1960: 16–18). That same year, Bronius Kalinauskas’s article Word and Word Groups in a Sentence (1954) was published in a journal for teachers. It mentions three types of relations between words: agreement, government, and adjacency. The latter type is perhaps mentioned for the first time in Lithuanian linguistics—this is also noted by the author himself: “Adjacency had not been discussed in Lithuanian grammar textbooks until now” (Kalinauskas 1954: 18). On the basis of the Russian equivalent prišlijimas, primykanije, given alongside the name of this relation and the literature consulted, it may be supposed that the three-part model of subordinate relations was adopted in Lithuanian syntax studies from Russian linguistics. R. Vaskelaitė is considerably more cautious in her dissertation on models of syntactic relations in Lithuanian linguistics: she surmises that Russian syntax scholarship may merely have supported the identification of a third type of relation between words in Lithuanian (Vaskelaitė 2009: 69). B. Kalinauskas defines agreement as “the grammatical relation between two words when those words agree with one another in person, number, case, or gender” (Kalinauskas 1954: 18). The article’s author (or his authorities) understands the phenomenon as syntactic in nature—“a grammatical relation between words”—and associates it with the “agreement” of the meanings of certain morphological forms. The categories listed indicate that, in his view, not only an adjective and a noun can agree, but also a verb and a noun or pronoun. The author calls government a relation between words in which “the form of one word depends on another word” (Kalinauskas 1954: 18). As can be seen, this definition also relies on the morphological expression of the relation; the syntactic nature of the phenomenon is not yet discussed here. The definition of adjacency ARTŪRAS JUDŽENTIS. Prijungiamasis ryšys sovietų okupacijos laikų lietuvių sintaksės darbuose 137 is not provided in the article; instead, only its instances are listed (relations between verbs and adverbial words, sentence words and parenthetical words, forms of address, interjections, and negation), which, in R. Vaskelaitė’s view, would indicate that adjacency “is intended to encompass those words in a sentence that are not considered to be related to other words” (Vaskelaitė 2009: 69). Syntactic relations (called grammatical relations) are described in greater detail in Bronius Kalinauskas’s article The Word Group as an Object of Syntax (1961). They are first divided into subordinate and coordinate relations. Subordinate relations are then further divided into three types: agreement, government, and adjacency. Word agreement is defined as “the subordination of nouns, adjectives, numerals, pronouns, and participles to nouns, where the subordinated word acquires the forms of the word on which it depends” (Kalinauskas 1961: 31). As can be seen, only agreement between nouns and adjectival words is discussed. Although the words connected by this relation are mentioned, the relation itself is linked to their morphological expression (grammatical forms). Full agreement is distinguished (when all forms of the governing word are acquired) from partial agreement (when only some of the governing word’s forms are acquired). Word government is defined as “the subordination of nominal words to verbs, other nominal words, and certain adverbs, in which the subordinated word acquires the forms required by the meaning of the word on which it depends” (Kalinauskas 1961: 31). This definition also relies on the parts of speech of the words connected by the relation. The nature of the relation is conveyed through its expression, which is determined by the meaning of the head word (rather than its grammatical properties). The formal criterion in the definition is considerably too broad: government is expressed through the case of the dependent word, which is only one component of its form. As in J. Žiugžda’s grammars, three types of word government are distinguished: direct, indirect, and prepositional (or indirect). Direct government is defined as the subordination of nominal cases that, “by virtue of their meaning, are absolutely necessary to the governing word, are the most usual, and are used most often” (Kalinauskas 1961: 31–32). Accusative, genitive, dative, and instrumental cases can be governed directly. Some verbs can directly govern two cases (e.g., rašyti, klausti). Indirect government is the subordination of nominal cases “that are not essential to those governing words by virtue of their meaning, but are only secondary” (Kalinauskas 1961: 32). This definition contradicts the author’s general conception of government: after all, government is identified on the basis of the requirement imposed by the meaning of the main word in the phrase, which the form of the dependent word satisfies... A considerable number of verbs are listed below (e.g., eiti kelią, lūžti obuoliais, gulėti 138 BENDRINĖ KALBA | 95 hour) and nouns (e.g.: uncle’s tale, sowing rye, a man of learning, a silk scarf; a girl with light-colored braids) as examples of secondary government. Finally, prepositional, or indirect, government is the government of a case by means of a preposition. It is not the government of prepositional constructions, but specifically the government of cases through an intermediary—a preposition. Although it is acknowledged that the case of the governed word depends on the preposition... B. Kalinauskas’s article includes an interesting section on the relationship between prepositional and non-prepositional government, showing that some words can be governed in either way, e.g.: to die of hunger and to die from starvation. Adjoining of words is defined as “a way of connecting words based solely on semantic relations and unrelated to changes in forms” (Kalinauskas 1961: 34). The definition raises doubts about the syntactic nature of this type of relation. Full and partial adjoining are distinguished. Full adjoining includes cases in which either invariable or inflected words modify only one word in the sentence. Partial adjoining is defined as “the mutual relationship of words in which the same word is governed by, or agrees with, one word, while adjoining another word” (Kalinauskas 1961: 34). The following examples are given: Katrė got up and looked around; The stork flew back, happy; By the opening, on this side of the threshold, stands the mother. It is clear that what B. Kalinauskas calls partial adjoining was later regarded in Lithuanian linguistics simply as adjoinment (one of the double or twofold syntactic relations of the predicative attribute), or as semantic relations of specification (those of a specifying adverbial), and the concept of partial adjoining was abandoned. The article next discusses the coordination of words—a relation connecting “equal members of a phrase” (Kalinauskas 1961: 35). We see that even the definition uses a calque from Russian. A few years later, Bronius Kalinauskas’s article Žo-džiai ir žodžių junginiai sakinyje (1963) likewise distinguishes three ways of connecting words: agreement, government, and adjoining (Kalinauskas 1963: 5). In classifying and defining them, the author repeats the definitions in his earlier article and essentially follows the Russian grammar prepared by Evdokija Galkina-Fedoruk (Sovremennyj russkij jazyk. Sintaksis, Moskva, 1957, § 3). As in the earlier article, agreement is understood as a way of connecting words characteristic only of certain parts of speech, whereby words are attached to nouns and “the attached word acquires the form of the word to which it is attached”; full and partial agreement of words are distinguished (Kalinauskas 1963: 5). Agreement between a verb and a noun functioning as its subject is not mentioned. The definition of government likewise matches the one given in the earlier article (Kalinauskas 1963: 6). In B. Kalinauskas’s view, government can be expressed not only by verbs, but also by ARTŪRAS JUDŽENTIS. Prijungiamasis ryšys sovietų okupacijos laikų lietuvių sintaksės darbuose 139 nominals and some adverbs. Direct and indirect (i.e., prepositional) government are distinguished. As can be seen, the questionable category of secondary government, identified in the earlier article, has disappeared. In defining adjoining, the author likewise follows the earlier article, and adjoining is considered “the connection of words based solely on semantic relations and unrelated to changes in forms.” Not only invariable but also inflected words can be adjoined; however, “they do not change their form,” that is, their form is not determined by adjoining. Two types of adjoining are also distinguished: full and partial adjoining (Kalinauskas 1963: 9). In Zelma Dumašiūtė’s article Dėl sakinio dalių ir jų tarpusavio santykių (1963), published in the same year, syntactic relations are also taken into account in characterizing sentence constituents. In classifying them, the author drew on the work of Czech scholars (e.g., P. Mrazek). Three main types of “semantic relations between sentence constituents” are distinguished: predication, determination, and coordination (Dumašiūtė 1963: 237–248). The author associates coordination with coordinative relations, and predication and determination with subordinative relations. Predication is defined as “the relation of the predicate to the subject” (Dumašiūtė 1963: 242), while determination encompasses “object (complement), attributive (modifier), and adverbial relations” (Dumašiūtė 1963: 242–243). In cases of determination, one sentence constituent “logically expands another sentence constituent (one word expands another word) and thereby forms certain phrases” (Dumašiūtė 1963: 243). It is clear that semantic relations are not clearly distinguished from syntactic ones but are conflated (cf. the definition of determination). Z. Dumašiūtė classifies syntactic relations according to their morphological expression. In her words, some syntactic relations between words (government and agreement) are expressed by the morphological form of dependent words, which is required by the head word. Others (adjoinment) are unexpressed. Adjoinment is based solely on the semantic relation between words (the dependent word is connected to the head only by meaning), while its morphological form does not express the syntactic relation and “remains entirely free” (Dumašiūtė 1963: 243). Thus, expression through the dependent word’s morphological form is the principal basis for this classification of syntactic relations. The author concludes: “<...> the conditioning of the morphological form of certain words indicates only certain syntactic relations [government and agreement – A. J.] and therefore cannot be considered a criterion for all syntactic relations” (Dumašiūtė 1963: 243). Thus, the author also regards as syntactic those word relations that are not expressed by morphological means and are based solely on semantic relations between words. At the same time, in the higher-education textbook Jono Balkevičiaus Current Lithuanian Syntax (1963), published that year, three types of word depen- 140 BENDRINĖ KALBA | 95 The types of connection are classified similarly to B. Kalinauskas’s works. Adjoining is defined as unrelated to the morphological form of the dependent word, and even the term is the same: “Adjoining of words is a type of word connection based only on logical relations and unrelated to changes in form” (Balkevičius 1963: 35). Not only indeclinable but also inflected words can be adjoined. As in B. Kalinauskas’s works, full and partial adjoining are distinguished (Balkevičius 1963: 36). Agreement and government are associated with a change in the form of the dependent word. It is the nature of this connection that distinguishes agreement from government. Agreement of words is defined as “the connection of inflected words <...> to nouns, in which the connected word takes the forms of the word that connects it” (Balkevičius 1963: 28). It can be seen that J. Balkevičius regards only the relation of nouns to adjectival words as agreement. As in B. Kalinauskas’s works, full and partial agreement are distinguished (Balkevičius 1963: 28). Word government is defined as “the connection of nouns and other inflected words to verbs, nominal words, and adverbs, in which the connected word takes the form (case) required by the meaning of the word that connects it” (Balkevičius 1963: 29). This is merely a slightly refined definition by B. Kalinauskas. The semantic requirement is inseparably linked to the government of form (case). As in the works of J. Žiugžda and B. Kalinauskas, three types of government are distinguished—direct, indirect, and prepositional (Balkevičius 1963: 29). Indirect government is “the connection of nominal case forms that are not semantically essential to the governing word” (Balkevičius 1963: 29). Such government is characteristic not only of verbs (e.g.: Ten dirbau ilgesnį laiką; Mokiniai išvažiuavo atostogų), but also of nouns, which use indirect government to connect an attribute expressed by the genitive or instrumental case (e.g.: Subraškejo sausų eglių šakelės; Būriavosi vyrai šventiškais drabužiais). The concept of indirect government was adopted from B. Kalinauskas’s earlier works. It can be seen that this classification is based on several criteria. First, government is based on a semantic requirement (the dependent word is not required by the head word). The morphological form of the governed word is decisive, as in the case of agreement, except that it is not identified (at least in part) with the form of the head word, as in agreement, but is selected. And third, the parts of speech capable of governing are listed (distinguished by meaning and expression). The innovations in Lithuanian syntax were finally incorporated, a decade after their publication in B. Kalinauskas’s article, into the textbook for secondary schools, Juozo Žiugždos Grammar of the Lithuanian Language. It was updated with the help of co-author Praną Gailiūną. The versions available to me ARTŪRAS JUDŽĖNTIS. Prijungiamasis ryšys sovietų okupacijos laikų lietuvių sintaksės darbuose 141 state on the title page of the 1969 edition that this was already the second edition of the textbook. The first appeared in 1964 and had been tested in schools by the Ministry of Education’s Institute for Scientific Research in Schools. The fourth subsection of the textbook’s very first chapter, Composition of speech in the sentence, is entitled Ways of combining words (Gailiūnas, Žiugžda 1969: 7–8), and it is there that syntactic relations, as we understand them, are discussed. Of course, the authors do not use the term syntactic relation, but speak of relations between words. In their view, these relations form the basis of word combinations: “<...> word combinations are formed by linking words according to meaning, and relations of meaning are often also expressed by the forms of words” (Gailiūnas, Žiugžda 1969: 7). Thus, as in earlier editions of the Grammar, it is held that the meaningful relations between words are expressed directly by word forms (though not always). Moreover, they are distinguished only at the level of word combinations. This is a clear influence of Russian grammatical scholarship of that period. Three (!) “ways of forming word combinations” are distinguished. The definition of agreement has been modified somewhat: “Agreement is a relation between words in a sentence in which the dependent word agrees with the head word in number, gender, and case, or in person” (Gailiūnas, Žiugžda 1969: 7). It relies on the concepts of the head and dependent words in a phrase, which were defined earlier. This makes it possible to avoid listing the parts of speech that agree in the definition (as had been done previously). The text goes on to specify which categories of the two words in the phrase agree, namely number, gender, and case or person. This is new in the definition. The categories listed show that the discussion concerns agreement of adjectives with nouns and agreement of verbs with pronouns and nouns. This is confirmed by the examples given. The definition presented required a note: “Only the third-person singular and plural endings are identical” (Gailiūnas, Žiugžda 1969: 7–8). The definition of government says nothing new, but has been adapted to the previously defined concepts of the head and dependent words in a phrase: “Government is a relation between words in a phrase in which the head word requires one case form or another” (Gailiūnas, Žiugžda 1969: 8). It is further shown that not only verbs and their non-finite forms can govern, but also “some adjectives, and sometimes nouns and numerals.” A completely new feature of this grammar is the identification of a third type of relation between words—adjunction: “Adjunction is a kind of relation between words in a combination in which the dependent word is linked to the head word only by meaning, and its ending (form) does not change.” The relation between a verb and an adverb is given as a typical example of adjunction. The section Agreement of the Main Sentence Parts 142 BENDRINĖ KALBA | 95 almost does not differ from the description in the 1945 edition; only the following was added: “Predicates expressed by an infinitive or the neuter gender do not agree with subjects (the infinitive and the neuter gender are invariable)” (Gailiūnas, Žiugžda 1969: 21). Thus, the school grammar adopted the system of three types of word relations only in 1964. This was probably connected with the theoretical positions of J. Žiugžda’s co-author P. Gailiūnas, who worked in the Ministry of Education at the time. The conception of syntactic relations essentially remained unchanged (they are ways of expressing semantic relations between words), but syntactic relations were linked to word combinations. Research on subordinate syntactic relations during the period under consideration can provisionally be divided into two approaches: in some works they are characterized as a property of sentence parts, while in others they are described as the basis of the structure of word combinations. By the end of the period, efforts to merge the two research approaches are apparent. 2. SYNTACTIC RELATIONS AS A PROPERTY OF SENTENCE PARTS Throughout the occupation, a prominent research approach linked syntactic relations to the conception of sentence parts. It continued and refined the studies of sentence parts inherited from independent Lithuania. The school grammars by J. Žiugžda, already discussed, belong to this approach. The most substantial contribution in this field was made by Vytautas Sirtautas. In Vytautas Sirtautas’s dissertation Veiksmažodinis tarinys dabartinėje lietuvių kalboje, defended in 1969 (published in 2019), only one type of syntactic relation is discussed—agreement. The same term is used both for the concord of adjectives with nouns and for the syntactic relation between subjects and predicate words. The agreement of the subject and predicate (both are considered the main parts of the sentence) is not examined in detail in V. Sirtautas’s dissertation; only at the end are some reflections presented, revealing “typical tendencies in the agreement of Lithuanian PSD [main parts of the sentence]” (Sirtautas 1969 [2019]: 325). In the fifth chapter of the dissertation, On the agreement of a verbal predicate, the author points out the unidirectional nature of the agreement of adjectives with nouns and contrasts it with the reciprocal agreement of the subject and predicate. Rather than defining what agreement is, V. Sirtautas first of all rejects what it is not. The following cases are not considered agreement: “when the idea of agreement is neither present nor possible, since one member of the agreement has no forms (I bang)” (Sirtautas 1969 [2019]: 316). Moreover, “[I]t is inaccurate to look for the idea of agreement in gender in most nouns, since such nouns do not inflect for gender; cf. The wolf is an animal” (Sirtautas 1969 [2019]: 316). On the other hand, the following are also not considered agreement, and ARTŪRAS JUDŽENTIS. Prijungiamasis ryšys sovietų okupacijos laikų lietuvių sintaksės darbuose 143 such cases “where a word, despite having the relevant forms, is for some reason not adapted to the forms of another word, e.g.: Uogos sveika. Žirgeliai sukinkyta” (Sirtautas 1969 [2019]: 317). The necessary nature of the relation between these sentence parts is emphasized several times: “<...> in expressing the relations between subject and predicate, agreement is an important, yet nonessential (optional), manifestation of those relations” (Sirtautas 1969 [2019]: 318, 325, 327). The factors determining agreement between principal sentence parts are identified as follows: “the composition of the subject (simple, compound, several homogeneous subjects), the type of predicate (verbal, nominal, adjectival), and also their valency and their ‘formality’ (having forms); to a lesser extent, word order and distance” (Sirtautas 1969 [2019]: 320). Of these, it is specifically the type of predicate that “essentially determines the differences in agreement” (Sirtautas 1969 [2019]: 322). Considering whose role—subject or predicate—is more important with respect to agreement, V. Sirtautas concludes that “it is precisely the subject that most often determines the forms of the predicate” (Sirtautas 1969 [2019]: 325). The role of agreement in language is indicated: “<...> this is a clear manifestation of the correlation and mutual dependence of the principal sentence parts” (Sirtautas 1969 [2019]: 327). In Vytautas Sirtautas’s article Dėl derinimo sąvokos, published in 1972, no definition of agreement is provided either; attention is instead drawn to certain inconsistencies in its conception. The author proposes distinguishing the reciprocal adaptation of principal sentence parts from the agreement of an attribute with the word it modifies. The distinction is based not only on the freer or different nature of this adaptation (the person of adaptation), but also on the syntactic relation expressed: “<...> agreement, as a form of the subordinate relation between words, is wholly unsuitable for characterizing the syntactic relations of principal sentence parts, since the subject and predicate are connected by a predicative relation” (Sirtautas 1972: 109). Drawing on the works of Russian linguists, V. Sirtautas also proposes other terms for the reciprocal adaptation of principal sentence parts: parallelism, correspondence, or coordination (Sirtautas 1972: 109). In Vytautas Sirtautas’s article Dėl veiksnio ir tarinio santykių, published in 1976, on the basis of the differing tendencies and expression of predicate (or predicates) and attribute agreement with several homogeneous subjects, the author once again proposes treating the reciprocal dependence of subject and predicate as “higher-level syntactic relations (a predicative relation, ‘interdependence’) and not confusing it with the unilateral dependence of determiners (determination)” (Sirtautas 1976a: 80). In Vytautas Sirtautas’s teaching manual for higher education, published in 1976Sakinio dalys syntactic relations are described 144 BENDRINĖ KALBA | 95 [They] are discussed only to the extent required by the topic under examination. The section Problems of Syntax and Its Study presents a diagram entitled “Arrangement” of the Sentence (Sirtautas 1976b: 13), in which syntactic relations are divided into coordination, predication, subordination, and “double relations” (the latter include “detached sentence constituents,” specification, and apposition). Subordinate relations are further divided into three types: agreement, government, and adjunction. Government is divided into three subtypes: direct (I see him), indirect (is angry with him), and oblique (I write to him) government. Unfortunately, the diagram’s classification of syntactic relations is not discussed or explained in the booklet. A slightly revised version of the diagram is also included in the new edition of the booklet (Sirtautas 1981: 20), except that predication is replaced by linking, and specification by specifying sentence constituents. In the section Criteria and Methods for the Differentiation and Identification of Sentence Constituents, grammatical relations are also identified as one of the criteria for distinguishing sentence constituents. However, they are not discussed in detail; it is merely noted that “predicative, coordinative, or subordinate relations are characteristic of one or another sentence constituent; for example, adverbials are usually adjoined, while attributes agree” (Sirtautas 1976b: 20). It is further cautioned: “Here, too, only more characteristic tendencies are being indicated; there is no strict regularity” (Sirtautas 1976b: 20). The criterion of grammatical relations is described similarly in a later edition of the booklet (Sirtautas 1981: 27). In this section, some doubtful claims are also made when identifying the principal indicators of sentence-constituent position: that the subject and predicate are characterized by non-dependence (syntactic?), that “the subject (optionally) requires an attribute,” that the predicate governs adverbials, and that the object and adverbial may require an attribute (Sirtautas 1981: 33). Identifying additional indicators of the subject and predicate positions, V. Sirtautas mentions mutual agreement—coordination (Sirtautas 1976b: 24). In the new edition of the booklet, this passage is expanded: “The forms of the subject determine the forms of the predicate <...>. In this way we prove that the subject position is significant in the sentence. It is not merely the predicate’s ‘left actant’” (Sirtautas 1981: 33). In Vytautas Sirtautas’s teaching aid Agreement of the Principal Sentence Constituents (1978), subordinate relations are divided into agreement, government, and adjunction. The idea that agreement is not a type of syntactic relation but merely an indicator of such relations appears briefly in the section Agreement of S and P as an Indicator of Their Relation: “An important indicator of the relation between S and P is the agreement (coordination) of their forms” (Sirtautas 1978: 57). The same idea is evident in the concluding remarks: “Although agreement of the PSC [principal sentence constituents] is not an indispensable expression of the syntactic relation between S [subject] and P [predicate], nevertheless, when ARTŪRAS JUDŽĖNTIS. Prijungiamasis ryšys sovietų okupacijos laikų lietuvių sintaksės darbuose 145 ji egzistuoja, yra akivaizdi jų santykių manifestacija“ (Sirtautas 1978: 86). Vis dėlto ši mintis toliau nebuvo plėtojama. V. Sirtautas urged that agreement be distinguished from the subject–predicate relation: “The confusion is greater when agreement, as a form of subordination, is identified with the predicative relation. It would be wholly inappropriate to call both for this pupil and the sun is shining <...> agreement (‘a way of combining words’). The latter is not the same way of combining words, but linking” (Sirtautas 1978: 53–54). The differences between agreement and linking—now termed subordinate agreement and coordination—are summarized by V. Sirtautas in a table on page 78. The problem of distinguishing so-called weak government from adjunction was also recognized: “Government is difficult to distinguish from adjunction; between them ‘there is no sharp boundary’ <...>. These questions as a whole ‘still await more comprehensive investigation’” (Sirtautas 1978: 16). The problem became particularly apparent after reading foreign literature: “When the term adjunction is not used at all (and only government is used) [reference to Czech and German works], then, of course, this problem of differentiation does not arise. Perhaps it would be better to call intermediate cases adjunctive government or to seek other solutions” (Sirtautas 1978: 17). The author did not, however, venture to propose abandoning the distinction of adjunction. The failure to distinguish syntactic relations from the means by which they are expressed, conflating them into one, also led to the conclusions drawn. “S [subject] and P [predicate] are, one might say, of equal status. Although the predicate generally plays the greater role in the sentence <...>, a single S can sometimes form the grammatical center of a sentence; moreover, S often determines the forms of the predicate. There is no basis for treating the subject as dependent on P and demoting it to the role of other actants; indeed, one should not do so if one wishes to avoid distorting an objective analysis of sentence structure” (Sirtautas 1978: 84). As can be seen, the subject–predicate syntactic relation—a research topic pursued over many years—led the author’s thinking toward its exceptional status and an exaggerated elevation of the subject’s role in the sentence. It is doubtful that this can be regarded as “objective analysis,” especially since sentence structure is a phenomenon dependent on prior theoretical assumptions. The syntax textbook for higher education prepared by Vytautas Sirtautas and Česys Grenda, Syntax of the Lithuanian Language (1988), also distinguishes three ways of subordinating words, but first specifies the basis for distinguishing them: “According to the means by which one word is subordinated to another” (Sirtautas, Grenda 1988: 37). Unfortunately, this basis is not followed consistently in defining agreement, government, and adjunction; compare: “Agreement is a type of subordinate relation in which, as the form of the head word changes, the form of the dependent word changes accordingly” (Sirtautas, Grenda 1988: 38); “Government is a type of subordinate relation in which one word re- 146 BENDRINĖ KALBA | 95 kalauja kito žodžio atitinkamos formos“ (Sirtautas, Grenda 1988: 39); „Šliejimu laikomas toks prijungimo būdas, kur prie pagrindinio žodžio prijungiamas kita žodis be jokių gramatinių priemonių, tik pagal prasmę“ (Sirtautas, Grenda 1988: 41). Šiuose apibrėžimuose gramatikos lygmuo (žodžių formos) sumišęs su semantikos lygmeniu (reikalauja; pagal prasmę). Valdymas apibrėžtas per plačiai (reikalauja formos, o ne linksnio formos). A pair of more complex cases is then discussed, in which words in a sentence are linked by two kinds of syntactic relations. In the first case, among several words governed by certain verbs (consider, appoint, elect, etc.) (for example: elect him chairman, appoint her village elder, consider him bad, consider her bad), a “formal intermediate relation” is identified: the words agree in gender and number, which are determined by “the features present in the semantic structure of the noun (or nominal pronoun) occupying the object’s accusative position”—of course, in cases where the instrumental is a noun (or nominal pronoun) that has gender distinctions (Sirtautas, Grenda 1988: 42). In the second case, attention is drawn to the agreement of predicate attributes with the subject and their adjunction to the predicate. Furthermore, “the gender of the participle and half-participle here is determined not only by the grammatical properties (gender, number, case) of the noun (or nominal pronoun) functioning as the subject, but also by its semantic structure. For example, in the sentence That pig, without even saying thank you, went out, the noun pig means ‘male person’ and therefore determines the gender of the participle (without saying)” (Sirtautas, Grenda 1988: 42–43). The syntactic relation between subject and predicate is discussed separately and is termed predicative relations in the textbook. The authors even warn: “Sometimes the subject–predicate relation is needlessly identified with agreement, which is, of course, also one indicator of that relation. Moreover, agreement of the principal sentence constituents (correlation, mutual relation) is often confused with agreement of an attribute (a one-way relation)” (Sirtautas, Grenda 1988: 86). It is noted that “syntactic relations are not explicit (obvious); they can often be revealed only by indirect methods” (Sirtautas, Grenda 1988: 85–86). Another feature of subject–predicate relations is noted by comparing sentences with coordinated subjects, for example: Berries were bought. Mushrooms and berries were bought. Mushrooms, berries, and everything else was bought. Such sentences, the authors argue, “clearly show that the components of the subject determine the forms of the predicate (and which gender form is dominant). Thus, the predicate depends directly on the subject” (Sirtautas, Grenda 1988: 85–86). Recalling the correlation between subject and predicate forms in sentences such as Many factories had been profitable, Many factories had been profitable, “from which it is indisputably clear ARTŪRAS JUDŽENTIS. Prijungiamasis ryšys sovietų okupacijos laikų lietuvių sintaksės darbuose 147 the mutual connection of the main sentence parts”; it is concluded that “[I]t is inaccurate to regard the subject as merely a dependent (or ‘independent’) argument of the predicate” (Sirtautas, Grenda 1988: 85–86). 3. SYNTACTIC RELATIONS AS THE STRUCTURAL BASIS OF WORD GROUPS Interest in the topic of word groups entered Lithuanian linguistics under occupation mainly through Russian linguistics. Many concepts, terms, and classifications were transferred into Lithuanian linguistics from Russian grammar works, in particular through the articles by B. Kalinauskas (1954, 1961, and 1963) already surveyed in this article. In these works, syntactic relations were discussed as the structural basis of word groups. J. Balkevičius also briefly characterized word groups and classified them as simple and complex in Dabartinės lietuvių kalbos sintaksėje (1963). In the teaching aid prepared by Bronius Kalinauskas, Lietuvių kalbos žodžių junginių sintaksė (1972), the classification of syntactic relations was further refined. To the account of syntactic relations published in earlier works, the author added a further division of government “according to semantics, that is, according to the semantic relations of the governed word with the word governing it” (Kalinauskas 1972: 3). On this criterion, government is divided into “strong government, in which the case forms of certain nouns, alone or with prepositions, are necessary for governing words by virtue of their meaning, most usual and most frequently used, such that when the governing word is uttered, an association with that case or prepositional construction arises first” (Kalinauskas 1972: 3), and “weak government, in which, when the governing word is uttered, an association with some case or prepositional construction does not necessarily arise in our consciousness; that is, those cases or prepositional constructions are not necessary to those governing words by virtue of their meaning, but are merely secondary” (Kalinauskas 1972: 4). The division into strong and weak government is based on the semantic relations between words and their psychological perception. From the perspective of semantic relations, weak government is no longer government, since the dependent word is not required. This subtype of government was probably distinguished because of its externally similar expression, cf.: jaučiu vasarą and gražų vasarą (the accusative in the first case is grammatical, while in the second it is a semantic case). Apparently aware that he has introduced something new, the author adds the note: “Some linguists consider what we call weak government to be the attachment of case or prepositional constructions, since the governed words” 148 BENDRINĖ KALBA | 95 “no longer answer questions of case, but rather questions characteristic of adverbials or modifiers” (Kalinauskas 1972: 4). This shows that the semantic (syntactic) and formal criteria for classification are incompatible. Since the choice was made to classify according to semantics, the “semantic relations of the governed word with the word governing it” are similar in all cases (the governed word is governed), and classification on this basis is impossible. It follows that those “some linguists” who regarded cases of “weak government” as adjunction were right. The concept and classification of adjunction in the textbook differ little from those presented in the earlier article (Kalinauskas 1972: 4). Thus, in the works of the pioneers of the ternary classification of syntactic relations, we also find most of their finer distinctions (strong and weak, direct and indirect government, full and incomplete agreement, full and partial adjunction), most likely adopted from Russian linguistics and later giving rise to considerable controversy in Soviet syntactic works. In classifying syntactic relations, B. Kalinauskas relies on several criteria: meaning, morphological expression, and part-of-speech membership. The problem of distinguishing weak government (and, essentially, government in general) from adjunction remains unresolved, and is connected with reliance on the level of meaning. Vito Labučio mokomosios priemonės Žodžių junginių problemos (1976) skiriuje Žodžių ryšių reiškimo būdai skiriamos jau keturios žodžių ryšių rūšys: derinimas, valdymas, šliejimas ir formų koordinavimas. As can be seen, V. Labutis regards agreement and the coordination of forms as different types of syntactic relations. They differ primarily in the autonomy of the categories applied to the words being linked (the point is that the categories of adjectival words that agree are non-autonomous, whereas the corresponding verb categories are autonomous): “Agreement of words is the adaptation of the forms of words with non-autonomous categories to the forms of other words” (Labutis 1976: 29), whereas “[C]oordination of word forms is the possible (but not necessary) mutual adaptation of forms of words with autonomous categories” (Labutis 1976: 31). Moreover, the definitions show that agreement is a one-way dependent relation, whereas coordination of forms is mutual. The parentheses indicate that coordination of forms is a possible, but not necessary, syntactic relation. This qualification is not made about agreement; it is only stated that “[A]greement is not a complete assimilation of forms, but merely a selection according to the forms of the noun” (Labutis 1976: 29). V. Labutis perceptively captures the conditional nature of agreement, though only in terms of the role it plays in the sentence: “Agreement indicates which word the adjectival word is connected with, but says nothing about the content of the relation: what agrees ARTŪRAS JUDŽENTIS. Prijungiamasis ryšys sovietų okupacijos laikų lietuvių sintaksės darbuose 149 sis žodis gali būti pažyminys, vardinė tarinio dalis ir net aplinkybė“ (Labutis 1976: 29). Žodžių formų koordinavimo skiriamos dvi atmainos: ne tik veiksniu ir tariniu einančių žodžių, bet ir apoziciniu ryšiu susijusių daiktavardžių formų koordinavimas (Labutis 1976: 31). Valdymo apibrėžimas nedaug skiriasi nuo apibrėžimų, pateiktų kitų autorių: „Žodžių valdymas – tai daiktavardiškojo žodžio linksnio parinkimas tam tikrai sintaksinei ir semantinei funkcijai pagal kito žodžio reikšmę ar kokį kitokį semantinį elementą“ (Labutis 1976: 33). Tik sintaksinis ir semantinis žodžių ryšys gerai skiriamas nuo jo reiškimo linksnio forma. Nedaug skiriasi ir šliejimo apibrėžimas: „Žodžių šliejimas – tai žodžių ryšių reiškimo būdas, kur ryšį rodo tik siejamųjų žodžių reikšmės, nepadedamos jokios formos“ (Labutis 1976: 37). In the third volume of the Grammar of the Lithuanian Language, published that same year and devoted to syntax (1976), edited by Kazys Ulvydas, Vytautas Ambrazas, and Adelė Valeckienė, syntactic relations are likewise examined in the chapter on word combinations, as an essential property of their structure. Their more specific types are discussed in the analysis of subordinate word combinations. The introductory section of the chapter on word combinations, Syntactic Word Combinations, was written by A. Valeckienė. Three varieties of syntactic relations in subordinate word combinations are distinguished (as the text notes, most often): agreement, government, and adjoinment. Their classification “depends essentially on the lexical and grammatical properties of the head [of the word combination]” (LKG 1976: 11). Agreement is defined on the basis of the grammatical correspondence between the form of the dependent word and that of the head: “Agreement is the adjustment of the form of the dependent constituent to the form of the head constituent. Words that share the same grammatical categories—gender, number, and case—are joined by agreement” (LKG 1976: 11). The agreement of two nouns is also discussed: “Combinations of two nouns are formed by agreement, e.g., the writer Žemaitė, the son Petras. They agree in case, number, and often gender.” Nevertheless, the distinctive nature of agreement between nouns is acknowledged: “However, the agreement of two nouns is somewhat different from the agreement of an adjective with a noun. Two agreeing nouns are subordinated to each other only in meaning, while syntactically one could say that they do not depend on, but rather correlate with, each other: they can occur in the same position both together and independently, cf. The son Petras came and The son came, Petras came. For this reason, the relation between two agreeing nouns is sometimes called correlation and is closer to coordination than to subordination” (LKG 1976: 11). Here, the semantic subordination of the nouns is questionable, all the more so since they also have the same reference (the thing denoted); that is, 150 BENDRINĖ KALBA | 95 the same entity is named by different nouns at the level of meaning, from different perspectives (a kinship relation and a proper name assigned to the person in that relation). One may suppose that, from the speaker’s perspective, the set of referents identified by the kinship relation (sons) is smaller than the set named by the corresponding proper name (although more often, perhaps, the opposite is true). In this case, however, the proper name seems to denote not all possible referents bearing that name, but only those connected with the speaker, and thus specifies the reference of the first noun. A similar correlation is discerned between the indefinite pronouns something, someone, etc. (something foreign) and masculine forms of adjectives, as well as the nominative and accusative forms of the indefinite pronouns who, something, something or other, someone, everything, something, etc. and neuter adjectives (something new, something new): “The adjective is subordinated to the indefinite pronoun only semantically. It makes the broader concept expressed by the pronoun more specific. Syntactically, in combinations of this kind, the adjective is almost as independent as a noun. Here it resembles a substantivized adjective” (LKG 1976: 11–12). The claim of semantic subordination seems mistaken here. In terms of meaning, in the combinations something foreign and something new, the “heavier” constituent—and thus apparently the head—is the adjective with a specific meaning, not the pronoun with an indefinite meaning (and reference). Agreement as a syntactic relation seems to be contrasted with government and adjoinment (taken together). What the latter have in common is that “in most cases they can be distinguished by the form of the dependent constituent.” For “government is often considered a relation in which an oblique case of a noun or a prepositional construction attaches to the head constituent, whereas adjoinment is a relation in which an invariable form attaches to the head constituent” (LKG 1976: 12). Nevertheless, it is emphasized that “the morphological form of the dependent constituent does not always help to distinguish the syntactic relations between words precisely. For example, it rained this morning and it rained this morning express the same syntactic relations, although the morphological form of the dependent constituent differs. The same applies to I want sleep and I want to sleep” (LKG 1976: 12). Here we see a tendency to distinguish government from adjoinment solely on the basis of the form of the dependent word, although the author rejects this as a basis for distinguishing them. The examples given suggest that the semantic relations between the words in a combination are probably considered more important. Government is defined as a relation “in which the head constituent requires a particular form of the dependent constituent (the dependent constituent is formally determined by the head constituent)” (LKG 1976: 12). It is clear that the definition is based on formal dependence, or dependence of a word’s form—that is, a dependence marked by the word’s form, namely its case ARTŪRAS JUDŽENTIS. Prijungiamasis ryšys sovietų okupacijos laikų lietuvių sintaksės darbuose 151 or by a prepositional phrase. Dependence itself seems to be understood here syntactically rather than semantically (the relation is discussed in the chapter on word combinations). According to the strength of government, or the necessity dependent on meaning, obligatory (or strong) and non-obligatory (or weak) government are distinguished (LKG 1976: 12). However, it is immediately noted that the boundaries between them are unclear, and there are always transitional cases. Cases that are not essential to the lexical meaning of verbs, as well as prepositional phrases expressing quantity, separation, comparison, and so on, are weakly governed (LKG 1976: 13). In this respect, the grammar, like B. Kalinauskas’s work, takes a misguided path, yielding to the influence of formal expression (case). Complex, or gradual, government is also discussed, in which “the presence of one case with a verb seems to condition the presence of another case with the same verb,” e.g., carves a stick for a child (LKG 1976: 13–14). This is an erroneous account of syntactic relations and their expression: both nouns depend on the verb semantically and syntactically, and the form of one is not dependent on the other. Adjoinment is the relation in which “the dependent constituent attaches to the head without being required by it; it is formally undetermined and connected with the head constituent only logically” (LKG 1976: 12). The definition draws not only on the formal expression of the syntactic relation, but also on the semantic (logical) connection between the constituents of the combination. This is new in the definition of this relation. 4. ABIEJŲ KRYPČIŲ SULIEJIMO APRAIŠKOS Daugiausia Lietuvių kalbos ir literatūros instituto mokslininkų parašytoje ir V. Ambrazo redaguotoje vienatomėje lietuvių kalbos gramatikoje rusų kalba Grammatika litovskogo jazyka (1985) daug vietos skiriama žodžių junginiams aprašyti, tačiau joje daugiau dėmesio skirta, ypač aprašant vientisinio sakinio sandarą, ir sakinio dalims. Šioje gramatikoje skiriamos trys sintaksinių ryšių atmainos: prijungimas, tarpusavio sąsaja ir sujungimas (GLJ 1985: 417). Pagal sintaksinio ryšio tipą suskirstyti ir žodžių junginiai. Prijungimo atmainos skiriamos prijungiamųjų junginių skyriuje. Agreement is defined as the matching of the form of the dependent member of a phrase (upodoblenie) su pagrindinio nario forma (GLJ 1985: 422). Prie delio koordinacija su pagrindiniu žodžiu laikoma ypatinga derinimo atmaina, taip pat ir nežymimųjų įvardžių santykis su vyriškosios arba niekatrosios giminės būdvardžiais ir dalyviais (GLJ 1985: 422). Valdymu laikomas kitokios nei pagrindinio žodžio priklausomojo žodžio kaitomosios formos parinkimas, atsižvelgiant į pagrindinio žodžio leksines ir gramatines savybes (GLJ 1985: 422). 152 BENDRINĖ KALBA | 95 A distinction is made between prepositional and non-prepositional government, and strong and weak government. Government is considered strong when it expresses an obligatory (inter-word) relation, and weak when it expresses a non-obligatory, usually adverbial, relation (GLJ 1985: 422–423). The attributive genitive is also considered governed, e.g.: a working person, a forest bird (GLJ 1985: 423). Adjacency is the type of subordinating syntactic relation that is not expressed by inflectional forms of the dependent word (GLJ 1985: 423). Adverbial semantic relations are most often expressed by adjacency. Some verbs characteristically take strong adjacency, e.g.: began to work, I want to read, was here, seemed beautiful, behave manfully. Such word combinations express objective and completive semantic relations (GLJ 1985: 424). In the programmatic V. Ambrazas’s article Units of the Syntactic and Semantic Structure of the Lithuanian Sentence (1986), prepared as a synthesis of experience in writing a Russian-language grammar of Lithuanian, the most important aspects of the syntactic and semantic structure of the sentence are surveyed. In the introduction, the author very rightly emphasizes that in the work of researchers of Lithuanian syntax, “one should seek to define as clearly as possible and distinguish on a single basis at least the most important categories of the levels of form and semantics” (Ambrazas 1986: 4). Yet even in this quotation the article’s main shortcoming is apparent: in seeking to clarify some essential concepts of syntax and semantics used by Lithuanian linguists, the syntactic structure of language itself is constantly confused with the morphological level of expression. In the cited statement, instead of the level of “form” (i.e. morphology), the author most likely meant the syntactic level, or else syntax (like morphology) was understood as a level of linguistic expression. V. Ambrazas first divides syntactic relations into subordinating, coordinating, and mutual-dependence, or linkage relations (Ambrazas 1986: 7). The latter differ from the first two in being reciprocal. In typical cases, they are expressed through the coordination of forms (Ambrazas 1986: 7). The clearest example of such mutual dependence is said to be the relation of a verb to the nominative of a noun or pronoun (exactly!): this nominative form depends on the verb, but it “in turn determines the number form of the verb” (Ambrazas 1986: 7). This and other well-founded examples of mutual linkage (finite verb forms and a nominal constituent functioning as subject, the person forms of the verb to be functioning as a copula and the nominal forms associated with it, even indeclinable verb forms occupying the nominative position and verbs functioning as predicates) clearly conflate the syntactic relation of dependence with its morphological expression. V. Ambrazas knew that “the question of relations of mutual dependence” could be resolved differently; he surveys various approaches to such a solution (some of them tellingly show that what is called— ARTŪRAS JUDŽĖNTIS. Prijungiamasis ryšys sovietų okupacijos laikų lietuvių sintaksės darbuose 153 This “two-way” relation is between different levels), but the “Lithuanian” approach to resolving the issue was chosen because “the syntactic nature of the mutual relation between the members is fairly clear” (Ambrazas 1986: 10). The failure to distinguish the levels of syntax and morphology is also clearly apparent in the classification of subordination. Its (i.e., the syntactic phenomenon’s) varieties are distinguished according to their formal (i.e., morphological) expression. The conclusions state this even more clearly: “Syntactic relations are formally realized through agreement and coordination, government, and adjoinment” (Ambrazas 1986: 35). It would follow that the listed varieties of syntactic relations are morphological phenomena... On the basis of expression, three varieties of subordination that have become traditional are distinguished: agreement, government, and adjoinment (Ambrazas 1986: 7). Agreement is defined as “the adaptation of the subordinate word form to fit the head word according to a morphological category” (Ambrazas 1986: 8). The word adaptation in the definition itself does not express the nature of a syntactic phenomenon. The definition clearly indicates the morphosyntactic (i.e., syntactic in nature but expressed through morphological forms) character of the phenomenon, which in no way can be called purely syntactic. In V. Ambrazas’s view, the agreement of nouns with independent grammatical categories, called coordination, should not be an obstacle to treating it as a certain subtype of agreement (Ambrazas 1986: 8). Government is defined as “the selection of an inflected form of the subordinate word <...>, determined by the valency and grammatical properties of the head word” (Ambrazas 1986: 8). Again, the word selection hardly seems suitable for expressing the syntactic nature of the phenomenon, but the definition does not aim to remain at the level of syntax—instead, it boldly ventures into morphology (inflected form of a word), as though it and syntax constituted a single level of language structure and its analysis. Adjoinment is defined as “subordination not expressed by inflectional endings of words” (Ambrazas 1986: 8). This definition clearly distinguishes the levels of morphology and syntax. However, it says nothing about the syntactic nature of this type of relation. How does adjoinment differ from the mutual dependency relation between non-inflected verb forms and a verb functioning as a predicate (to live became tiresome)? The distinction of a subtype called “strong adjoinment” raises even more doubts. It includes the relations between “semantically incomplete” verbs and infinitives: stopped working, tried to return, can eat (Ambrazas 1986: 9). How does the syntactic nature of the relation between these words differ from government (cf.: I want dinner and I want to eat)? Nevertheless, this article succeeded in separating the (morpho)syntactic level of sentence analysis from the semantic one. However, in defining and classifying syntactic concepts themselves, their syntactic nature could not be clarified in many cases; the level of syntax was conflated with that of morphology. 154 BENDRINĖ KALBA | 95 CONCLUSION Initially, school grammar continued the syntactic tradition of independent Lithuania even under occupation and distinguished only two types of subordinate connection—agreement and government. Paradoxically, this tradition was continued in the works of J. Žiugžda, an ideologue of the Soviet occupation regime. From 1964, when P. Gailiūnas, who worked at the Ministry of Education, came to J. Žiugžda’s aid, a three-part model of classifying subordinate connections—agreement, government, and juxtaposition—already adopted from Russian linguistics in scholarly grammar was adapted for schools. The distinction of three types of government—direct, indirect, and oblique—is attributable to J. Žiugžda and was also most likely adopted from Russian linguistics. In the 1950s and early 1960s, B. Kalinauskas was primarily responsible for transferring the three-part Russian model of classifying subordinate connections into Lithuanian linguistics and selecting its terminology. P. Gailiūnas adopted this model from him or directly from Russian and adapted it for secondary schools, while J. Balkevičius did so for higher education (the latter perhaps also via Latvian linguistics). This classification of subordinate connections became part of Lithuanian linguistics and remained established there throughout the occupation. The model continues to predominate in schools and scholarly works to this day... B. Kalinauskas’s works are also the first to feature distinctions among many types of syntactic relations (full and partial agreement, strong and weak government, full and partial juxtaposition), also most likely adopted from Russian grammar studies. The examination of these distinctions in later Lithuanian syntax works substantially defined the field of discussion in this area of grammar. During the period under consideration, subordinate syntactic relations were described in works representing two approaches to syntax: by characterizing sentence constituents, as one of their properties (in the works of J. Žiugžda and V. Sirtautas), and by examining word combinations, as the basis of their structure (in the works of B. Kalinauskas, J. Balkevičius, V. Labutis, and A. Valeckienė). It was only in V. Ambrazas’s 1985 Lithuanian grammar that a clearer move was made toward merging the two approaches. Throughout the period under consideration, syntactic relations were not distinguished from semantic relations; this is especially evident in the definition of government. Only at the very end of the period (in V. Ambrazas’s programmatic article) did scholars begin to recognize these shortcomings in the description of syntactic relations and attempt to distinguish syntactic and semantic units clearly. Nevertheless, the syntactic level of analysis continued to be closely linked to the morphological (expression) level (agreement in gender, person, and number; government of cases), with both regarded as “surface” levels of language pertaining to expression. ARTŪRAS JUDŽĖNTIS. Prijungiamasis ryšys sovietų okupacijos laikų lietuvių sintaksės darbuose 155 As early as the beginning of the 20th century, agreement was understood as the use of morphological forms, that is, as a phenomenon at the level of expression rather than semantics or syntax. Later, the term also came to denote a type of syntactic subordinate connection, without examining the semantic relations conveyed by this connection. Essentially, this multifaceted conception—initially conflating semantics and morphology, and later adding syntax as well—persisted throughout the 20th century, and is still partly adhered to today. Government was defined as a semantic relation between words both in independent Lithuania and in J. Žiugžda’s school grammars, which continued its syntactic scholarly traditions after the war. Moreover, the semantic nature of this relation was conflated with its morphological expression. This view was later supplemented by a syntactic interpretation of this type of relation. In the early period of occupation (up to and including the publication in 1976 of the third volume of the academic Lietuvių kalbos gramatikos), government and agreement were linked to an interest originating in Russia in word combinations as the basis of their structure. In Lithuanian linguistics, juxtaposition was distinguished (first by B. Kalinauskas) as a semantic, and probably also syntactic, relation between words not expressed by morphological means. Thus, it was based solely on expression—that is, on its absence—and lacked clear syntactic content. Syntactically, it could not be strictly distinguished either from government (when the dependent word is expressed by an indeclinable form functioning as an object) or from agreement (when an indeclinable form functions as the subject). LITERATŪRA Ambrazas Vytautas 1986: Lietuvių kalbos sakinio sintaksinės ir semantinės struktūros vienetai. – Lietuvių kalbotyros klausimai 25: Lietuvių kalbos sintaksės tyrinėjimai, 4–44. Balkevičius Jonas 1963: Dabartinės lietuvių kalbos sintaksė, Vilnius: Valstybinė politinės ir mokslinės literatūros leidykla. Dumašiūtė Zelma 1963: Dėl sakinio dalių ir jų tarpusavio santykių. – Lietuvos TSR MA darbai. A serija 2(15), 237–248. Gailiūnas Pranas, Žiugžda Juozas 1969: Lietuvių kalbos gramatika. II dalis: Sintaksė, 2-asis patais. leidimas, Kaunas: Šviesa. GLJ 1985: Grammatika litovskogo jazyka, vyr. red. V. Ambrazas, 1985: Vilnius: Mokslas. Judžentis Artūras 2020: Sintaksinis veiksnio ir tarinio ryšys: sampratos kaita. – Bendrinė kalba 93, 1–25. Prieiga internete: http://journals.lki.lt/bendrinekalba/index. Judžentis Artūras 2021: Prijungiamasis ryšys lietuvių sintaksės darbuose (XX a pirmoji pusė). – Bendrinė kalba 94, 1–15. Prieiga internete: http://journals.lki.lt/bendrinekalba/index. 156 BENDRINĖ KALBA | 95 Kalinauskas Bronius 1954: Žodis ir žodžių grupės sakinyje. – Tarybinė mokykla 2, 17–25. Kalinauskas Bronius 1961: Žodžių junginys kaip sintaksės objektas. – Tarybinė mokykla 5, 31–39. Kalinauskas Bronius 1963: Žodžiai ir žodžių junginiai sakinyje. – Kai kurie gramatikos ir stilistikos klausimai, Kaunas: Valstybinė pedagoginės literatūros leidykla, 3–23. Kalinauskas Bronius 1972: Lietuvių kalbos žodžių junginių sintaksė: mokymo priemonė, Vilnius. Labutis Vitas 1976: Žodžių junginių problemos, Vilnius: Vilniaus valstybinis universitetas. LKG 1976: Lietuvių kalbos gramatika 3: Sintaksė, vyr. red. K. Ulvydas, Vilnius: Mokslas. Sirtautas Vytautas 1969 [2019]: Veiksmažodinis tarinys dabartinėje lietuvių kalboje, Šiauliai: Litera [Bibliotheca actorum humanitaricorum universitatis Saulensis 18]. Sirtautas Vytautas 1972: Dėl derinimo sąvokos. – Kalbotyra 24(1), 108–110. Sirtautas Vytautas 1976a: Dėl veiksnio ir tarinio santykių. – Kalbotyra 27(1), 78–80. Sirtautas Vytautas 1976b: Sakinio dalys. Problematika, pagrindinės sąvokos, Šiauliai: Šiaulių K. Preikšo pedagoginis institutas. Sirtautas Vytautas 1978: Pagrindinių sakinio dalių derinimas (Vientisinio sakinio gramatinės struktūros analizė), Vilnius: Lietuvos TSR aukštojo ir specialiojo vidurinio mokslo ministerijos Leid ybinė redakcinė taryba. Sirtautas Vytautas 1981: Sakinio dalys. Problematika, pagrindinės sąvokos, Kaunas: Šviesa. Sirtautas Vytautas, Grenda Česys 1988: Lietuvių kalbos sintaksė, Vilnius: Mokslas. Vaskelaitė Ramunė 2009: Sintaksinių ryšių modeliai lietuvių kalbos sintaksėje: humanitarinių mokslų daktaro disertacija, Vilnius: Lietuvių kalbos institutas, Vytauto Didžiojo universitetas. Žiugžda Juozas 1945: Lietuvių kalbos gramatika. II dalis: Sintaksė. Progimnazijoms ir gimnazijoms, Kaunas: Valstybinė pedagoginės literatūros leidykla. Žiugžda Juozas 1954: Lietuvių kalbos gramatika. II dalis: Sintaksė. Vidurinėms mokykloms, 8-asis leidimas, Kaunas: Valstybinė pedagoginės literatūros leidykla. Žiugžda Juozas 1960: Lietuvių kalbos gramatika. II dalis: Sintaksė, 14-as leidimas, Kaunas: Valstybinė pedagoginės literatūros leidykla. Gauta 2022 03 20 Priimta 2022 10 17 ARTŪRAS JUDŽENTIS. Prijungiamasis ryšys sovietų okupacijos laikų lietuvių sintaksės darbuose 157 ARTŪRAS JUDŽENTIS SUBORDINATION IN THE WORKS ON LITHUANIAN SYNTAX IN THE YEARS OF SOVIET OCCUPATION S u m m a r y The article analyses, how the types of subordination are described in the works on Lithuanian syntax, during the Soviet occupation, during the years between the middle of the 20th century ant the end of eighties. The investigation opens with the Grammar of the Lithuanian language (1945) by Juozas Žiugžda and ends with the manual for higher schools Syntax of the Lithuanian language (1988) by Vytautas Sirtautas and Česys Grenda. Initially, under conditions of occupation school grammar continued the traditions of syntax of independent Lithuania and distinguished only two variants of subordination – agreement and government. This tradition was continued in the works of the ideologue of the occupation regime Juozas Žiugžda. It was only when Pranas Gailiūnas came to Žiugžda’s help that the three-variant model of subordination, which had already been adopted in scientific works from Russian linguistics, was applied to the school. Bronius Kalinauskas, a lecturer at the Pedagogical Institute, transferred the three-variant Russian syntax model to Lithuanian in the 1960s and early 1970s. Pranas Gailiūnas took over this model from him (or directly from the Russian linguistics) and applied it to secondary, and Jonas Balkevičius – to high school. In Kalinauskas’ works we also find many differences in the types of subordination (full and partial agreement, strong and weak government, full and partial adjunction), that are most likely taken over from Russian linguistics. Their analysis in later works of Lithuanian syntax has largely defined the field of consideration in this area of grammar. Subordinate syntactic relations were described in works on syntax of two directions during the considered period: as one of the characteristics of the parts of sentence (Žiugžda, Sirtautas), and as the structural basis of word groups (Kalinauskas, Balkevičius, Labutis and Valeckienė). Only in the Grammar of the Lithuanian language, edited by Vytautas Ambrazas (1985), was more clearly turned in the direction of merging both directions. Throughout the considered period, syntactic relations were not distinguished from semantic relations (this is especially obvious when defining the relation of government). And only at the very end of the period (in the article of Vytautas Ambrazas) did these shortcomings of the description of syntactic relations begin to be realized, and an attempt was made to clearly distinguish between syntactic and semantic units. However, the syntactic level of analysis continues to be confused with the morphological (expression) level (agreeing in gender, 158 BENDRINĖ KALBA | 95 person, number; governed inflections), both of which are considered as surface levels of language. Already at the beginning of the twentieth century agreement was perceived as the alignment of morphological forms, rather than semantics or syntax. Later, it became known as a variant of a syntactic subordination, without delving into the semantic relationships that this connection conveys. In essence, this multifaceted concept, first merging semantics and morphology, later on – and syntax, has survived throughout the twentieth century and in part continues to this day. Government in independent Lithuania was defined as a semantic relationship, and in the first decades after the war it was generally perceived in a similar way. Nevertheless the semantic (later on and syntactic) nature of the relationship was fused to its morphological expression. At the early period of occupation, agreement and government were associated with the study of word groups, as they made the basis for their structure. Adjunction (šliejimas) in Lithuanian linguistics was singled out as a semantic (later on and as a syntactic) relationship of words that is not expressed by means of morphology. So it was based on surface expression (or not-expression, exactly) only and didn’t have any distinct syntactic content. The article analyses, how the types of subordination are described in the works on Lithuanian syntax during the years of Soviet occupation – approximately between the middle of the 20th century ant the end of eighties. The investigation opens with the Grammar of the Lithuanian language (1945) by Juozas Žiugžda and ends with teaching aid for higher schools Syntax of the Lithuanian language (1988) by Vytautas Sirtautas and Česys Grenda. Initially, under conditions of occupation school grammar continued the traditions of syntax of independent Lithuania and distinguished only two variants of subordination – agreement and government. This tradition was continued in the works of the ideologue of the occupation regime Juozas Žiugžda. It was only when Pranas Gailiūnas came to help for Žiugžda that the three-variant model of subordination, which had already been adopted in scientific works from Russian linguistics, was applied. KEYWORDS: subordination, agreement, government, adjunction (šliejimas), Vytautas Ambrazas, Jonas Balkevičius, Pranas Gailiūnas, Zelma Dumašiūtė, Česys Grenda, Bronius Kalinauskas, Vitas Labutis, Vytautas Sirtautas, Juozas Žiugžda. ARTŪRAS JUDŽENTIS Lietuvos mokslų akademijos Vrublevskių biblioteka Žygimantų g. 1, LT-01102 Vilnius [email protected] ARTŪRAS JUDŽENTIS. Prijungiamasis ryšys sovietų okupacijos laikų lietuvių sintaksės darbuose 159 ARTŪRAS JUDŽENTIS PRIJUNGIAMASIS RYŠYS SOVIETŲ OKUPACIJOS LAIKŲ LIETUVIŲ SINTAKSĖS DARBUOSE S a n t r a u k a Straipsnyje apžvelgiama, kaip sovietų okupacijos laikais, maždaug nuo XX a. vidurio iki 9-ojo dešimtmečio galo, lietuvių sintaksės darbuose aprašomos prijungimo rūšys. Pradedama nuo Juozo Žiugždos Lietuvių kalbos gramatikos (1945), o baigiama Vytauto Sirtauto ir Česio Grendos parengta mokomąja priemone aukštosioms mokykloms Lietuvių kalbos sintaksė (1988). Iš pradžių mokyklinė gramatika ir okupacijos sąlygomis tęsė Nepriklausomos Lietuvos sintaksės tradiciją ir skyrė tik dvi prijungiamojo ryšio atmainas – derinimą ir valdymą. Ši tradicija buvo tęsiama okupacinio režimo ideologo Juozo Žiugždos darbais. Tik kai J. Žiugždai į pagalbą atėjo Švietimo ministerijoje dirbęs Pranas Gailiūnas, mokyklai 1964 m. buvo pritaikytas mokslinės gramatikos iš rusų kalbotyros jau perimtas trinaris prijungiamojo ryšio skaidymo modelis. Trinarį rusų kalbos sintaksės modelį 6-ojo dešimtmečio ir 7-ojo dešimtmečio pradžioje į lietuvių kalbą perkėlė ir terminiją parinko daugiausia Pedagoginio universiteto dėstytojas Bronius Kalinauskas. Iš jo (arba tiesiogiai iš rusų kalbos) šį modelį perėmė ir vidurinėms mokykloms pritaikė P. Gailiūnas, o aukštosioms – Jonas Balkevičius. B. Kalinausko darbuose randame ir daugelį sintaksinių ryšių atmainų skirtybių (pilnasis ir nepilnasis derinimas, stiprusis ir silpnasis valdymas, pilnasis ir dalinis prišliejimas), greičiausiai perimtų iš rusų kalbotyros, kurių nagrinėjimas vėlesniuose lietuvių sintaksės darbuose nemažai apibrėžė šios gramatikos srities svarstymų lauką. Prijungiamieji sintaksiniai ryšiai nagrinėjamuoju laikotarpiu aprašyti dviejų krypčių sintaksės darbuose: apibūdinant sakinio dalis, kaip vieną iš jų ypatybių (J. Žiugždos, V. Sirtauto darbuose), ir nagrinėjant žodžių junginius, kaip jų sandaros pamatas (B. Kalinausko, J. Balkevičius, V. Labučio, A. Valeckienės darbuose). Tik 1985 m. išėjusioje lietuvių kalbos gramatikoje Vytauto Ambrazo buvo aiškiau pasukta abiejų krypčių suliejimo kryptimi. Visą nagrinėjamąjį laikotarpį sintaksiniai ryšiai neskirti nuo semantinių santykių, ypač tai akivaizdu. apibrėžiant valdymo prijungiamąjį ryšį. Ir tik pačioje laikotarpio pabaigoje (programiniame V. Ambrazo straipsnyje) imta suvokti šiuos sintaksinių ryšių aprašo trūkumus, mėginta aiškiai skirti sintaksinius ir semantinius vienetus. Vis dėlto sintaksinis nagrinėjimo lygmuo ir toliau painiotas su morfologiniu (raiškos) lygmeniu (derinama gimine, asmeniu, skaičiumi; valdomi linksniai), abu jie laikyti „paviršiniais“, raiškai priskirtinais kalbos lygmenimis. Derinimas jau XX a. pradžioje buvo suvokiamas kaip morfologinių formų sutaikymas, t. y. kaip raiškos, o ne semantikos ar sintaksės lygmens reiškinys. 160 BENDRINĖ KALBA | 95 Vėliau juo imta vadinti ir sintaksinio prijungimo ryšio atmainą, nesigilinant į semantinius santykius, kuriuos šis ryšys perteikia. Iš esmės tokia daugialypė – iš pradžių į vieną suplakant semantiką ir morfologiją, o vėliau dar ir sintaksę – jo samprata išliko visą XX amžių ir iš dalies tebesitęsia ir šiandien. Valdymas dar Nepriklausomoje Lietuvoje ir jos sintaksės mokslo tradicijas po karo tęsusiose mokyklinėse Į. Žiugždos gramatikose apibrėžtas kaip žodžių semantinis santykis. Be to, semantinis šio ryšio pobūdis buvo suplaktas su jo morfologine raiška. Toks požiūris vėliau buvo papildytas sintaksine šios ryšio atmainos interpretacija. Ankstyvuoju okupacijos laikotarpiu (iki pat akademinės Lietuvių kalbos gramatikos trečiojo tomo pasirodymo 1976 m.) valdymas ir derinimas sieti su iš Rusijos atėjusiu susidomėjimu žodžių junginiais kaip jų sandaros pamatu. Šliejimas lietuvių kalbotyroje buvo išskirtas (pirmą kartą B. Kalinausko) kaip morfologijos priemonėmis nereiškiamas semantinis, vėliau – ir sintaksinis, žodžių ryšys. Tad jis rėmėsi tik raiška (jos nebuvimu), o aiškaus sintaksinio turinio neturėjo. Sintaksiškai jo nebuvo galima griežtai atskirti nei nuo valdymo (kai priklausomasis žodis reiškiamas papildiniu einančia nekaitomąja lytimi), nei nuo derinimo (kai nekaitomoji lytis eina veiksniu) atvejų. ESMINIAI ŽODŽIAI: prijungimas, derinimas, valdymas, šliejimas, Vytautas Ambrazas, Jonas Balkevičius, Pranas Gailiūnas, Zelma Dumašiūtė, Česys Grenda, Bronius Kalinauskas, Vitas Labutis, Vytautas Sirt autas, Juozas Žiugžda. ARTŪRAS JUDŽENTIS Lietuvos mokslų akademijos Vrublevskių biblioteka Žygimantų g. 1, LT-01102 Vilnius [email protected] ARTŪRAS JUDŽENTIS. Prijungiamasis ryšys sovietų okupacijos laikų lietuvių sintaksės darbuose 161