Postverbų pateikimo problema „Lietuvių kalbos žodyne“

ACTA LINGUISTICA LITHUANICA XLVIII (2003), 71 – 96 The Problem of Presenting Postverbs in the Dictionary of the Lithuanian Language ROLANDAS MIKULSKAS Institute of the Lithuanian Language The article deals with the treatment of the so-called ‘postverbs’ (i. e., particles occurring in phrasal verbs) in Lithuanian lexicography. In this connection, the North Lithuanian ‘perficientive’ constructions involving such postverbal markers are discussed in detail. The forms selected to act as postverbs are usually original illatives. Their functioning as aspect markers deprives them of their proper lexical meaning and of the possibility of functioning as adverbials. That is why, in dealing with the question of the grammatical interpretation of such forms (inspired by the needs of lexicographical description) one should first of all reject attempts to connect them with the process of adverbialisation of illatives in the West Lithuanian dialects. 0. When the twentieth volume of the Dictionary of the Lithuanian Language (hereafter LKŽ) appeared in print last June, I, like any other writer of LKŽ entries, was eager to see how my portion of the text had been edited. I was especially interested in how the editors had approached the ideas of an ethnolinguistic bent that I had tried to introduce into the organization of the text and the definitions of meanings, as well as innovations in the grammatical presentation of words. Unfortunately, more than once, as I read the edited text, I was disappointed: the impression was that my attempt to introduce new principles for organizing and presenting lexicographic material had gone unrecognized. Of course, the LKŽ editors followed an editorial tradition that had developed over many years, but does such an approach sometimes give rise to a certain inertia that blocks the way to innovation? Understandably, when preparing the final volume of the Dictionary, it may not be the right time to change the editorial principles radically, in order to preserve its methodological coherence. However, the innovations I proposed were not radical in nature. Two broader premises prompted me to introduce them. First, during the hundred-year period in which the LKŽ project was carried out, methodological and theoretical innovations were repeatedly applied to the writing of its entries in order to keep pace with developments in linguistics. Second, the word žemė, with its particularly complex semantics, is at issue—it, along with its closest derivatives, was precisely what I had to describe for the twentieth volume of LKŽ. Given their existential significance, contemporary semantic science—cognitive semantics—pays exceptional attention to words of this type (besides žemė, we can also mention dangus, namai, vieta, tėviškė, as well as the historically important Lithuanian words vakarai and rytai, etc.; cf. Gudavičius 2000: 39). This is because the world of meanings encompassed by such words encodes a people’s or linguistic community’s outlook, formed over many centuries, on matters important to its existence, as well as its mentality. When writing or creating such “foundational texts of Lithuanian culture” as LKŽ (Daujotytė 2002: 35), it is worthwhile to organize and present the meanings of these words, which are existentially important to us, so that their users’ mentality may be as clear as possible to readers and transmitted to future generations in as undistorted a form as possible, using the most appropriate lexicographic means. 72 | ROLANDAS MIKULSKAS In this respect, the experience accumulated in organizing meanings would also be useful in the future, as the LKŽ project is continued in one form or another. Here I would like to conclude these introductory remarks, whose rather abstract nature may begin to irritate the reader. I hope to return in the future to ethnolinguistic reflections and to applying the principles of this approach to dictionaries of the LKŽ type. Below I would like to dwell somewhat longer on one grammatical problem—the presentation of so-called postverbs in LKŽ. 1. While writing the entry for žemė, I noticed that in the northern Lithuanian dialects, the instrumental case of this word, or its analogue—the prepositional construction “į + accusative”—when used after verbs of a certain meaning, indicates that the action denoted by those verbs is carried through to completion, that is, it functions as an aspect marker. In the northwestern Samogitian dialect (Kretinga subdialect), this function is also performed by the adverb žemyn (downward) and its longer variant žemynais (downward). A review of analogous cases given in other volumes of LKŽ (e.g., lauka(n), šalin, viršuo, etc.) shows that the meanings of these words are defined inconsistently and that the editors’ view of the grammatical function of these forms has not been clearly formulated. In some entries these words are called emphatic particles, in others adverbs¹, while elsewhere examples of such usage are lost among the case forms of the corresponding nouns, which are used with various meanings. Seeing this inconsistent treatment of these words in LKŽ and wishing to present a more unified view of them, I tried, as far as possible, to present in one place all forms associated with the word žemė that perform the same grammatical function (this cannot be done completely, since some examples of such usage inevitably appear under the adverb žemyn). Thus, I conceived of establishing a separate sense in the entry for žemė. It could bring together the vast majority of instances of such usage, provided that the definition noted that the instrumental form of the headword, or its analogue, the prepositional construction “į + accusative,” is used chiefly after verbs denoting disappearance, destruction, or removal, and indicates that the action is nearing completion or is performed almost to completion. Place-name abbreviations would show the geographical distribution of this phenomenon. In light of recent studies of the expression of verbal aspect in northern Lithuanian dialects (Girdenis, Kačiuškienė 1986: 21–27), I consider it appropriate to call the forms in question, in their grammatical function, postverbs. Admittedly, adopting this formula for presenting the phenomenon in LKŽ meant turning a blind eye to the fact that in western Lithuanian dialects the instrumental is no longer used as a paradigmatic case; more precisely, only isolated instances of such usage are recorded. Many already regard the instrumental forms of some words used here as adverbs. However, as we shall see later, owing to their inherent semantics, the instrumental forms of different words have become detached from their declensional paradigms to differing degrees. For this reason, it is easier to treat žemė(n) “instrumentally,” as an intermediate aspect marker of verbs, than other words with this function, such as laūkan or šalin. 1 In the latter case, they too are treated inconsistently: they are presented as a separate sense of the corresponding adverb (šalin, viršuo) or given a separate entry as adverbs of a special function (žemė(n), žemyn). POSTVERBŲ PROBLEMA LKŽ 73 1.1. The editors of LKŽ chose a different solution. Only some examples of žemė(n) are treated “as illatives,” in accordance with the “grammatically correct” view that illatives are not possible west of a certain boundary of eastern dialects (see LKŽXX 313, sense 3). Only Samogitian dialect examples in which the function of a marker of aspect is performed by the prepositional construction “į+accusative” have been retained here. Sentences with žemėn (or northeastern Samogitian žemė) are listed by the editors under the adverb žemėn, which has been given a separate entry (ibid., 337). The function of the latter is not explained separately; instead, the reader is referred to žemyn (ibid., 344), where it is explained that this adverb (and its accentual variant žemýn) is used in the Samogitian dialect to emphasize a completed action². It should also be added that some instances of this usage are also listed under the adverb žemynais (žemynais), which has its own separate entry (ibid., 345). This editorial decision is logical. Its rationale is that it relies on a certain “knowledge of the position of things,” that is, on knowledge from historical grammar and dialectology concerning the development and fate of the illative case in present-day Lithuanian dialects. A certain inconsistency in this editorial principle will become clear later. For now, I would like to draw attention to the most striking shortcoming of this choice: forms that essentially have the same grammatical function are presented in no fewer than four places in the Dictionary, and are treated differently there. This unjustifiably fragments a unified phenomenon in verbal grammar. The newly proposed term postverb is retained only for the first case mentioned (i.e., under the third sense of the entry žemė). Incidentally, a closer look at how this sense has been edited also reveals certain inconsistencies and even errors. The semantic groups of verbs mentioned earlier, after which the postverbs žemė(n) and į žemę are usually used, have been “adjusted” in the definition so that verbs expressing derogation appear in place of those expressing disappearance. But that is not true! Nor has the principle of geographically delimiting the illative case been applied consistently: the sources indicate Lȳgumai, which lies far west of the area of paradigmatic use of the illative. Moreover, examples with postverbal use of žemėn include some that not only fall outside the distribution area of postverbs but do not even demonstrate such usage. Thus, the following examples are inappropriate: (1) A simple heart, so it won’t puff itself up: hands fall to the ground and collapse ML. (2) I beat and beat—it got beaten to pieces on the ground (ps.) DGLŠ. (3) You, dear brothers, my brothers-in-law, go to the new hill and forge a little golden sickle and cut my head off (d.) PRNG. (4) “May your head go to the ground”—he doesn’t say that to him, doesn’t recognize him ML. All these sentences were recorded in eastern dialects, where the illative is still actively used as a paradigmatic case. The illative plural in the second example also shows this. In the first example, the adverbial meaning of the word žemėn is clear: here it indicates the direction of the action, that is, it modifies the verb atsil eidžia with respect to direction in space. In the third ² Such a reference seems to assert the priority of the form žemyn over the form žemė(n), although geographical data show that the former is clearly a local variant. ³ In the usual place, that is, under the relevant senses of the adjective žemas, -a, the adverb žemyn is listed together with its longer variants (see LKŽXX 298 ff.). 74 ROLANDAS MIKULSKAS and in the fourth example, the illative meaning of the word žemėn is especially evident when these sentences are compared with geographically nearby illustrative sentences given under the second sense of the entry žemė: (5) The girl, angry, threw the rake to the ground and left Lb (308). (6) As soon as it falls to the ground somewhere [an apple] and bursts, it breaks apart ŠVnč (309). These remarks clearly show that the editors of LKŽ did not fully understand the point of my proposal concerning the forms in question used postverbally. It seems that enough arguments have been presented here to warrant a more serious discussion of how postverbs should be presented in LKŽ. So let us start at the beginning and proceed in order. 2.0. First, it would be useful to recall the main claims of the already mentioned article by Alekso Girdenis and Genovaitė Kačiuškienė. Drawing attention to parallel phenomena in the Latvian and northern Lithuanian verb systems, the authors state that a ternary aspectual opposition exists in northern and northwestern Lithuanian dialects, and that its development closely resembles the Latvian system. The essence of the ternary aspect system is that, in addition to the imperfective (durative aspect) and the perfective (completed-event aspect), there is also an intermediate category, which the authors of the article propose calling PER-FICIENTIVE. The perficientive is formed with certain adverbs of “bleached” semantics, similar in function and meaning to the prefixes of perfective verbs. The authors propose calling these POSTVERBS⁴. The general meaning of perficientive verbs and their postverbs is THE ACTION’S APPROACH TO COMPLETION OR, MORE GENERALLY, TO A CERTAIN BOUNDARY. Almost all postverbs used in the Lithuanian dialects mentioned above have Latvian equivalents. The most widespread postverb is žemė(n) and its local variants žemyn, žemynais, į žemę, corresponding to Latvian nuošts, zemē (e.g., northern Samogitian Kiršk koję žemėn TRK = Cērtiet kāju nuošt). With this postverb, the perficientive of so-called annihilative (destructive-meaning) verbs is especially common. Verbs with lauk (lauko(n), laukonais) ≈ Latvian ārā are also quite common among Samogitians. In northern Panevėžys dialects, forms with ōre are most often their equivalent. The perficientive with šalin (šalinais, į šalį) ≈ Latvian pruo(j)ām is only slightly less common. The authors give examples of perficientives with other, less frequent postverbs as well. The grammatical function of some of these, e.g., pry(šais), pušiau (pušiau), į rinkį, lygiai, is harder to discern—at least in the examples provided; compare: (7) My hands won’t rise against him! TRK. (8) Don’t break it in half—it’ll crumble ALS. (9) They’ll drive the horse into the ring like that Kl. (10) She went—and led me along with her KlK. Indeed, one can say that these words modify the predicates of the sentences given here. ⁴ Long before the article appeared, Girdenis had already used this term in identifying the grammatical function of the words in question, in slips submitted to the LKŽ card index with examples from northern Samogitian dialects. POSTVERBŲ PROBLEMA LKŽ 75 in terms of semantics, not aspect (e.g., in the last sentence lygiȧi could easily be replaced with the adverbial semantic equivalent kartų). 2.0.1. Having carried out a detailed analysis of deciphered tape recordings made in nearly all North Samogitian parishes, the article’s authors conclude that many perficientive constructions (more precisely, macroforms) are supposedly entirely usual and neutral only in the northern dialects. Further south (and southeast) from Viekšniai, Tirkšliai, Salantai, and Grūšlaukis, they gradually begin to lose their regularity and become increasingly dependent on context and the expressiveness of speech. This is presumably why the “annihilative” constructions miršta, bẽngi(a)si, sprógsta žẽmen. are by far the most widespread. 2.0.2. From all this, the article’s authors draw a single diachronic conclusion: the ternary aspect system of the northern Lithuanians (or its beginnings or rudiments) is supposedly a new grammatical phenomenon, arising under the influence of Latvian⁵. This influence may have been productive because, first, the new intermediate aspect increased the expressiveness and precision of the language and, second, its emergence fitted well with one of the most pronounced tendencies of the northern dialects—their syntax’s inclination toward the greatest possible concreteness. 2.1. Having briefly recounted the main ideas in Girdenis and Kači žkienė’s article, it is worth highlighting several points relevant to further discussion. First, words such as žẽme(n), laũko(n), õre, šalìn, used after certain verbs and functioning as a kind of marker of their intermediate aspect, are much more accurately called postverbs than adverbs, even if their semantics are “bleached.” Second, when explaining the structural features of perficientive constructions, it is important to remember that their general meaning is the approach of an action to its end or, more broadly, to a certain boundary. This idea accords with the fact that the most widespread postverb is žẽme(n) (į žẽmę). Compared with the others mentioned in the article, this postverb best represents the concept of a boundary: owing to its inherent semantic feature of locating the endpoint of an action, it can also signify the end or boundary toward which that action is approaching. The fact that the Latvian postverb nuõst, which, when used as an adverb, means ‘away’ and indicates only direction, is most often matched by Lithuanian žẽme(n), would itself seem to point to the tendency toward the greatest possible concreteness in the syntax of the northern Lithuanian dialects noted by the article’s authors. Third, for a uniform grammatical interpretation of postverbs, it is very important that the article’s authors regard perficientive constructions as certain grammatical MACROFORMS. Girdenis explained this concept in an earlier article (Girdenis 1984: 25–30). Macroforms are distinguished at a deeper morphosyntactic level; they directly “serve” syntax and encompass both synthetic forms and those periphrastic constructions that belong in a sentence to the same “horizontal” paradigms as synthetic forms and perform the same role. Macroforms are related to the level of microforms by realization (or manifestation) relations. A sentence is supposedly constructed directly from macroforms, but at the textual level it is treated as a combination of individual microforms (or simply words). This is precisely why various permutations of words belonging to one macroform are possible, as are their ellipsis and many other textual phenomena. The article’s author’s latter point is worth keeping in mind, since it explains some irregularities in the expression of perficientive constructions, such as occasional postverbs in pre-verbal position or their use after prefixed verbs. Agreeing with Girdenis, we could regard the prepositional construction as different manifestations of the same macroform, the periphrastic illative— 5 Regis, šia išvada dabar abejoja E. Lipartė (2000: 149–150). Jos manymu, trijų veiks lų sistemą šiaurinėse lietuvių tarmėse reikėtų kildinti iš baltų kalbų partik ulinių veiksmažodžių ir priešdėlinių veiksmažodžių koreliacijų sistemos. Ji plačiai paplitusi latvių kalboje. Jos liekanų ar užuomazgų esama ne tik šiaurinėse, bet ir rytinėse bei pietinėse lietuvių kalbos tarmėse, tik čia šios koreliacijos nesančios dar pakankamai ištirtos. Tačiau straipsnio autorės pateikiami negausūs veiksmažodžių su partik ulomis pavyzdžiai iš pastarųjų tarmių terodo lokalinę reikšmę kaip ir jų priešdėliniai atitikmenys, vadinamosios partikulos čia neatlieka veikslo rodiklio funkcijos. 76 | ROLANDAS MIKULSKAS as well. A sentence is supposedly “constructed” directly from macroforms, but at the textual level it is treated as a combination of individual microforms (or simply words). This is precisely why various permutations of words belonging to one macroform are possible, as are their ellipsis and many other textual phenomena. The article author’s latter point is worth keeping in mind, since it explains some irregularities in the expression of perficientive constructions, such as occasional postverbs in a position before the verb or their use after prefixed verbs. Agreeing with Girdenis, we could regard the prepositional construction as different manifestations of the same macroform, the periphrastic illative. POSTVERBŲ PROBLEMA LKŽ 77 2.3. Given the view that the ternary aspect system in our northern dialects developed on the Latvian model, it is entirely understandable that Latvian perficientive constructions with postverbs of directive origin nuost, žemė, ārā, pruō(jā)m are matched in Lithuanian by analogous constructions with illative forms žemė(n), (į žemę), lauk(a)n7, šalìn. Since these postverbs in the constructions under discussion primarily serve as aspect markers, it would seem pointless to discuss their meaning as illatives of concrete nouns. However, when addressing the problem of presenting these forms lexicographically—a task that, among other things, necessarily entails a grammatical interpretation—one must do so. All the more so because, as noted above, casually assigning these words to the class of adverbs complicates their presentation in the Dictionary. 2.3.1. While agreeing with the authors of the article discussed that this is a new grammatical phenomenon, we nevertheless cannot say anything more definite about when perficientive constructions began to spread in the northern Lithuanian dialects. Since Daukantas already used them in his writings, we can only surmise that some of them were widespread in the northwestern Samogitian (Kretinga) dialect at the end of the eighteenth and beginning of the nineteenth century. Historical grammar data show that by then the illative, as a paradigmatic case, had already been replaced by the prepositional construction “į + accusative” (Zinkevičius 1980: 258). The replacement of synthetic illative forms by prepositional constructions probably intensified gradually westward. Even today, the older generation systematically uses the singular illative of some nouns considerably farther west than the line indicated above, as far as Ariogala, Tytuvėnai, Radviliškis, and Jūkiškiai; elsewhere in western Lithuania it occurs mostly in folklore (Zinkevičius 1966: 200–201). It is evident that the need for the illative as a macroform functioning at the morphosyntactic level of language has not disappeared. Only its general function—to indicate the place toward which an action is directed or to which it is being directed—is realized in the western dialects of Lithuanian by a periphrastic construction. This raises the question of how to classify the synthetic illative forms of certain words that, for one reason or another, continue to be used in these dialects. In attempting to answer this question, linguists are compelled to address another problem: the degree to which these forms have become adverbialized. 2.3.2. In grammar, an adverb is defined as an uninflected and unconjugated independent part of speech that modifies an action or state with respect to various circumstances (place, time, cause, etc.) (cf. Ulvydas, ed., 1971: 425; Ambrazas, ed., 1997: 410; LKE 504). However, when seeking to distinguish adverbs from other parts of speech, we often cannot use these defining features of the category as criteria. For example, in attempting to determine the degree to which a particular case form of a noun has become adverbialized, one is essentially deciding whether it belongs to an inflectional paradigm; therefore, the first criterion—that it be an uninflected part of speech—becomes tautological. Sometimes context, or rather the function of the form in the sentence, helps distinguish a non-paradigmatic noun form from a paradigmatic one This postverb is apparently presented in a literary-standard form in LKŽ (VII 176). The authentic Samogitian forms lauko(n), laukonais (ibid., 184), cited in the article by Girdenis and Kaukčienė, are not associated in the Dictionary with the grammatical function under discussion; there they are treated as adverbs indicating direction of motion in space, although in some sentences these words can also be interpreted as postverbs, cf.: Gana, lipk lauko [out of the tree]! ŽG.7 78 | ROLANDAS MIKULSKAS (Paulauskienė 1989: 55–56). However, in the specific case under consideration, both the illative and the adverb derived from it, with its directive meaning, modify the action in terms of direction in space; thus, both serve as locative adjuncts. Therefore, the second, functional, property of the adverb cannot in this case serve as a criterion for distinguishing the adverb from the noun’s illative. A more substantial basis for regarding the disputed word form as not wholly detached from the noun paradigm is the fact that diminutives or plural forms of it still occur. This latter criterion, however, applies only partially, since in the western dialects of Lithuanian the plural illative disappeared first, and we are dealing only with surviving singular forms. If we take the form žemė(n) specifically, then, given its connection—important for postverbal usage—with the primary⁸ meaning of the word žemė, ‘the surface of (our planet) on which we walk and stand’, it is generally difficult to expect a plural form of it. The situation is entirely different when this word is used with other meanings. For example, the fourth meaning of žemė, ‘the loose upper layer of our planet’s surface’ (LKŽXX 313), fully allows the plural illative, cf.: (11) We did not carry into the soil what we ate and drank. ŠvNČ (316). (12) Born in the forest, raised in the forest, it comes out into the field and burrows into the soil (a plough) SLK (316). 2.3.3. Their inability to take agreeing modifiers is cited as an effective means of distinguishing adverbs from the cases from which they derive (Laigonaitė 1957: 49; Ulvydas, ed., 1971: 426; Paulauskienė 1989: 56). This is a point we must accept, though it should be added that not all concrete nouns, not all meanings of a polysemous noun, and not all paradigm forms of a single meaning are equally inclined to take agreeing modifiers. Suppose that in speech some object is used only as a landmark for indicating the direction of movement; in that case, the illative used (or the prepositional construction “into + accusative”) normally has no modifiers, since the circumstances in which the utterance is formed do not require further modification of that object. 2.3.4. It would be mistaken to think that the criteria listed here exhaust all the possibilities for distinguishing adverbs from the meanings of the specific cases that give rise to them. Semantic analysis of the entire utterance offers considerable possibilities here. Aldona Paulauskienė also acknowledges the influence of understanding and interpreting context and situation on the meaning of a grammatical form (ibid., 56). According to her, a case turns into an adverb when its objective meaning disappears altogether and the adverbial meaning becomes primary (ibid., 55). The “objective” interpretation of grammatical meaning also depends on the specific meaning of the noun. On the other hand, it is also required and supported by the semantics of certain predicates that are most often used in utterances with that noun. It is easy to understand that a noun such as 8 This meaning was placed second by editorial decision, with the meaning of Earth as a planet moved to first place. Such a decision would be understandable in an encyclopedic dictionary, an astronomical reference work, or, ultimately, a dictionary of contemporary language. However, this order of meanings is wholly inappropriate in a dictionary of the LKŽ type, where meanings are established primarily on the basis of folk-language utterances. Careful semantic and cognitive analysis of these utterances shows that for Lithuanians the fundamental concept and corresponding meaning is the earth as the surface on which we walk. It is from this concept that the strands of many other meanings of the word žemė emerge and to which they return. POSTVERBŲ PROBLEMA LKŽ | 79 žemė, whose fundamental meaning inherently denotes a boundary at which the action of all falling or thrown things naturally stops and ends, the illative used with predicates kristi, mėsti, lėistis, lipti and so on retains its object meaning more readily than, for example, the illatives of the nouns laukas or šalis. Because of the semantic structure of these nouns, their illative forms, when modifying an action, cannot locate its endpoint. Therefore, since they express the direction of an action in very abstract terms, they are more readily interpreted as adverbs. Nevertheless, only by examining each occurrence of such forms individually can we draw reliable conclusions about the relationship between their object and adverbial meanings. For example, even the illative forms in sentences like these from Samogitian, generally already considered adverbs (Laigonaitė 1957: 50), can be interpreted in different ways, cf.: (13) I’ll go outside to see what kind of breeze is blowing Vkš. (14) Go away! Vkš9. In the first example, it is easy to discern the object meaning of the form laukan because of its semantic relationship with the content of the subordinate clause. In the second example, the form šalin indeed has only an adverbial meaning. Next, cf. sentences from the LKŽxx: (15) When smoke pours from the chimney down to the ground like that, there’ll surely be rain PLs (309). (16) When you stoke the fire, the smoke falls to the ground—it’s a sign of rain coming Vdk (309). (17) Smoke is still falling to the ground; the weather won’t be fair yet Aln (309). (18) Smoke is going down; perhaps there’ll be more rain again Kp (309). The illative žemėn in the first sentence and the corresponding prepositional construction į žemę in the second both denote here the direction in which the smoke spreads, and thus have only a directive meaning. This is shown even more clearly by comparison with the examples of utterances given below: to describe an analogous situation, explicit directives are used here—the adverb žemyn (4) and the prepositional construction in žemę10. These examples were intended to show that even “genuine” illatives can have an “adverbial” interpretation. Such cases are indeed less common. Much more frequent are contexts that emphasize the object meaning of the illative of the word žemė, cf.: (19) As soon as it falls to the ground, something like an apple shatters and splits Švnč (309). (20) Audra muni raised, put into the ground Akm (309). The verbs in both sentences razsipleskoja, trūksta, dėjo ‘struck’, by predicating the property of hardness of the earth’s surface, also determine the “object” interpretation of the corresponding word’s illative forms. 9 Examples from A. Laigonaitė’s book are presented here unchanged, although they do not conform to the LKŽ rules for transposing material into the standard language. 10 In the dialects of eastern Lithuania, where the illative is used as a paradigmatic case, direction is expressed by the construction “į + accusative” (Laigonaitė 1957: 47; Zinkevičius 1980: 255). According to LKŽ data, such usage is also characteristic of the Southern Aukštaitian dialect, cf.: Jeigu in aukštų staugia, tai an gaisro, o jei in žemę, tai an mirties šuva staugia Mrc (LKŽxx 308). 80 | ROLANDAS MIKULSKAS 2.3.5. Acknowledging the influence of context on the meaning of a grammatical form, we cannot rule out the possibility that, in certain cases, even speakers of those dialects in which the illative is no longer used as a paradigmatic case are capable of understanding the original function of these synthetic forms. It would be just as pointless to deny that we understand the grammatical meaning of illative forms used in literature or colloquial standard language merely because this case is no longer alive as a paradigmatic case in the contemporary language (cf. Laigonaitė 1957: 48). 2.3.6. In general, the function of the illative can be defined as follows: in expressing an object, it modifies the action with respect to direction in space11. Its formal meaning can be represented as follows: movement + direction inward + interior of an object – location (Paulauskienė 1989: 121). Applying this schema to the illative of the word žemė, we must take its specific meaning into account. If we identify as its primary meaning ‘the surface of (our planet) on which we walk and stand’, the schema will be simpler. The form žemė(n) (į žemę) will then consist of these components: movement + direction toward an object – location. For example, using this schema, we could readily interpret the illative in the sentence: Nedainuok ne ant lietaus, lipk žemė (said to a rooster) GrNK (LKŽxx 337)12. However, this schema is even simpler. For example, in the sentence Vanduo krisdamas už daideliai, par užplūška krimta žemėn Kl (LKŽxx 337), the meaning of the illative form used can be represented as follows: movement + direction toward an object – direction. Here, all that remains of the illative meaning is the semantic component of the directive. It is unsurprising that in certain circumstances the form žemė(n) should express only direction. After all, owing to gravity, all objects moving toward the EARTH also move DOWNWARD. In the last example, it is now simpler to treat the form žemėn as an adverb13. 2.3.7. From what has been said, it is time to draw one important conclusion: the interpretation of the grammatical meaning of the forms under discussion also depends on how the material is presented lexicographically. If we postulate the aforementioned meaning of the word žemė as primary, few cases remain in which we could interpret the form žemė(n) in a purely “adverbial” way (that is, when it has only an abstract directional meaning). Therefore, one may view the form žemėn in the command Lipk žemėn! one way when reasoning abstractly—that the specific meaning of žemės has disappeared here (Laigonaitė 1957: 49)—but in a lexicographic description one must view it quite differently, since ‘floor’ and ‘ground’ are also identified as shades of the primary meaning (LKŽxx 312–313). In that case, the illative form in the sentence Lipk žemė nu lovos vaikai, kur užsikorei? KrŠ (LKŽxx 337) should also be interpreted concretely, “objectively”14. 2.4. Let us return to the main object of our discussion—perficientive constructions. It is easy to see that in these constructions the grammatical function performed by postverbs sets them apart from the lexical meaning these words would have if used freely in a sentence. The independence of lexical meaning is one of the characteristics of the adverb cate- 11 This article need not address the possibility that the illative also expresses temporal, manner, and purpose adjuncts (for more on this, see Laigonaitė 1957: 38–44; Paulauskienė 1989: 59). 12 In the article author’s opinion, this sentence does not demonstrate postverbal usage. 13 Perhaps because such an interpretation is so clearly suggested, this sentence is also out of place here among the other examples representing perficientive constructions. 14 For this reason, the inclusion of this sentence among the illustrations of perficientive constructions is also questionable. POSTVERBŲ PROBLEMA LKŽ 81 goriją apibrėžiančių požymių (Ul vydas, red., 1971: 425; Ambrazas, red., 1997: 410). Būtent dėl turimos leksinės reikšmės prieveiksmis ir sugeba reikšti įvairias veiksmo aplinkybes, t. y. modifikuoti sakinio tarinį erdvės, laiko, būdo ir kitais atžvilgiais. Akivaizdu, kad perficientyvinėse konstrukcijose postverbais einantys žodžiai žemyn (žemyn), žemynais (žemynais), laũk, nors formaliai ir nesiskiria nuo atitinkamų prieveiksmių, tačiau yra daugiau ar mažiau apriboti reikšmės atžvilgiu ir tokiems prieveiksmiams būdingos vietos aplinkybės funkcijos par excellence neatlieka. Kita vertus, ir minėtos iliatyvo formos žemė(n), lauka(n), (veikiau – laũko(n)), šali(n) dėl tų pačių priežasčių beveik15 nemodifikuoja sakinio veiksmo krypties erdvėje atžvilgiu. Ypač perficientyvinėse konstrukcijose yra desemantizuota forma žemė(n). Neaišku, kuo ją laikyti, prie kurios kalbos dalies priskirti. Geografinio kriterijaus taikymas nepadeda atsakyti į šį klausimą. Juk vargu ar galėtume analogiškuose pasakymuose iš rytinių tarmių, kur iliatyvas dar gyvas kaip linksnis, formą žemėn interpretuoti kitaip: čia vartojamose perficientyvinėse konstrukcijose ji desemantizuota lygiai taip pat kaip ir vakarų pusės pavyzdžiuose. Šios ir kitų panašių formų interpretacijai labai reikšmingas tas faktas, kad žemė(n) šiaurės rytų žemaičiai perficientyvuose vartoja pramaišiui su prielinks nine konstrukcija į žemę. Kadangi konstrukcijos su šiais postverbais yra labiausiai paplitę ir, kaip matysime toliau, aiškiausiai reiškia tarpinį veikslą, tai pagrįstai galime manyti, kad šiaurinėse lietuvių šnektose formuojantis ternarinei veikslo sistemai būtent iliatyvas suvaidino svarbų vaidmenį. Žinoma, kalbėti reikėtų tik apie semantinę minėtų iliatyvo formų potenciją, kuri, kaip minėta anksčiau, net vakarinėse tarmėse tam tikruose kontekstuose tolydžio gali būti atgaivinama. Sprendžiant apie formų žemė(n) ir į žemę semantinių potencijų panaudojimą perficientyvui reikšti, jas derėtų suprasti labai plačiai: gebėdamos reikšti nukreipto ar kreipiamo žemyn veiksmo ribą kartu jos gali reikšti ir kai kurių kitų veiksmų pabaigą, o veikiau – artėjimą prie pabaigos. Inherentiškai turėdamos semantinį ribos požymį kalbamosios formos geriausiai tinka tarpiniam veiksmažodžio veikslui reikšti. Taip įtikinamai galima būtų paaiškinti, kodėl būtent su postverbais žemė(n) ir į žemę perficientyvinės konstrukcijos yra labiausiai paplitusios. Nėra abejonės, kad ribos požymį šios iliatyvo formos paveldėjo iš tų junginių, kuriuose jos modifikuoja vertikalaus judėjimo žemyn reikšmę turinčius veiksmažodžius. Kaip žinome, perficientyvines konstrukcijas su kalbamomis formomis sudaro tik tam tikros reikšmės veiksmažodžiai. Tačiau tik dalyje jų galima įžvelgti vertikalaus judėjimo žemyn reikšmę. Suprantama, kad kuo sunkiau ši reikšmė įžvelgiama, tuo aiškesnė yra perficientyvo raiška. 2.4.1. Based on the data in the LKŽ card file, we can divide the predicates that occur with the postverbs žemė(n) and į žemę into three semantic groups. The first group consists of verbs meaning ‘perish’, the second of verbs meaning ‘destroy’, and the third of verbs meaning ‘remove (separate)’. The most frequently recorded verbs in the first group are: mirti, sprógti, dvėsti, šalti; in the second group – šauti, mušti, bengti; and in the third group – imti, vilkti(s), kinkyti, mėsti, griauti, kirsti, pjauti (see Tables 1, 2, 3, pp. 90–95). It is only in the third group that we may expect to find clearer verbs denoting downward vertical motion, cf.: 15 As we shall see later, the degree of desemanticization of all these forms vs the possibility of semantic interpretation also depends on the meanings of the particular verbs after which they are used. 82 | ROLANDAS MIKULSKAS (21) Mėšk žėmen, don’t eat it, you’ll get sick PLN. (22) Rītink toki seni žėmė, why would anyone need that much RDN. (23) Mažus numus griauna žėmen, building five- and nine-storey ones ŠTS. (24) Neškiava žėmen [the pot with potatoes, so it doesn’t burn KL. However, a verb or two with this meaning may also occur in the first two semantic groups, cf.: (25) Ejo anudu į suejo į tokį liūgną – smunka, skėsta į žemę that child VKŠ. Since the verbs in these sentences have retained the meaning of downward vertical motion, it remains possible to interpret semantically the combinations they form with the forms žėmė(n) and į žėmę, and therefore their perficientive meaning is not very prominent. Nevertheless, the vast majority of meaningful combinations formed by these so-called annihilative verbs with the illative forms in question do not occur; in other words, the periphrastic constructions they form do not meet the conditions for semantic compatibility. This is best illustrated by examples in which these forms occur after verbs of the opposite direction, that is, verbs denoting upward motion, cf.: (26) My eyes have already gone white, well, then he started to hang me down, and that’s enough LC. (27) Ráusu the grass down and I’ll throw it out KRŠ. (28) You have to pull flax out into the ground KLVRŽ. Thus, the grammatical function of the constructions formed by the verbs under discussion with the illative forms žėmė(n) and į žėmę is very clear. 2.4.2. As we already know, in the northwestern Samogitian (Kretinga) dialect, the postverb žėmen alternates with žėmyn (žėmyn) and its longer variant žėmynais (žėmynais). Since this form, when used freely—that is, adverbially—has a somewhat different semantic structure from the illative forms discussed above, we might reasonably ask whether the meaning of the predicate in an utterance sometimes determines its selection for postverbal use. However, examination from this perspective of the utterances in the LKŽ card file shows that the predicates forming perficientive constructions with the postverbs žėmyn (žėmyn) and žėmynais (žėmynais) do not differ in meaning from those discussed above. They too can be divided into the three semantic groups mentioned. Some of them, for example: mirti, sprógti, šálti; šáuti, pílti, kárti, múšti, ganýti; ráuti, máuti, gráuti, kìršti, are recorded with both illative and adverbial-origin postverbs, often in the same or nearby locations (cf. data in Tables 1, 2, 3). This shows that the choice of postverb form in utterance formation is not (at least at present) determined by the meaning of the predicate. From the perspective of meaning, however, the forms žėmyn (žėmyn) and žėmynais (žėmynais) are clearly derived from the illative forms and should therefore be regarded as secondary to them (when used freely, they have only an abstract directive meaning). They should also be regarded as secondary because they are local in distribution, compared with the illative-origin forms, which are widespread throughout the northern dialects. Other geographical facts also support this line of argument. Based on the data of the Atlas of the Lithuanian Language, the illative form žėmė(n) in the southern distribution POSTVERBŲ PROBLEMA LKŽ 83 The boundary (west of the area where the illative is used as a paradigmatic case) could be represented by the line Pagėgiai – Katyčiai – Švėkšna – Judrėnai – Plungė – Telšiai – Luokė – Pavandenė – Padubysys – Radviliškis – Šeduva – Pociūnėliai – Surviliškis – Kėdainiai (LKA11 55). We can see that this boundary essentially coincides with the boundary given earlier for the distribution of perficientive constructions in the northern Lithuanian dialects. Consequently, we can conclude that the spread of the form žemė(n) west of the area where the illative is used is associated solely with the use of perficientive constructions. The occasional instance in which this form can here be interpreted purely “illatively” (or even adverbially) merely demonstrates its semantic potential, that is, it should be regarded as a case of resemanticization. The adverb žemyn is quite different. It is widely used throughout Lithuania and is one of the most frequent Lithuanian adverbs (Ulvy das, ed., 1971: 535). It may have entered perficientive constructions later as a synonymous variant of the postverb žemėn, since it partly corresponds to its semantic structure. This idea is supported by the fact that this postverb (and its longer variant) currently has two accentuation patterns. As an adverb in the Samogitian dialect, as indeed throughout Lithuania, this form is stressed only on the ending, thus žemyn (žemynais). However, when used in perficientive constructions—that is, in a certain degree of isolation—it is often stressed on the root as well, thus žemyn (žemynais). Most likely, žemyn (northern Samogitian žėmin) in postverbal position also came to be stressed like žemyn (northern Samogitian žėmin), following the fairly frequent form in this position žemėn (northern Samogitian žėmėn). 2.4.3. From what has been said, it is clear that forms of both illative and adverbial origin now used in perficientive constructions should be viewed solely as markers of intermediate action, without overlooking the semantic interpretations of these forms that are still occasionally possible. The formal label postverb is entirely sufficient for naming these forms, since it makes it possible to avoid artificially “forcing” them into any part of speech on semantic grounds. Since the place of the forms in question in the Dictionary nevertheless depends on their relationship to traditional parts of speech, and their names are often indispensable in defining their meaning (function), in such cases one can use the non-committal terms illative form (not case!) or adverbial form. Knowledge of historical grammar should be used with great caution in the categorical treatment of these forms. While generally expressing developmental tendencies, it more or less schematizes the description of language. Above all, the lexicographer’s duty is to describe as fully as possible the semantics of every word (or form) included. As we have seen, the forms under discussion have undergone different semantic changes. It would be excessively rigoristic for the Dictionary to explain the semantic content and grammatical function of these forms according to the pre-established schemas proposed by historical grammar. This only obscures the true nature of these forms. 3. This article would also like to touch, at least briefly, on other more common postverbs. In the northern Lithuanian dialects, the function of an intermediate-action marker is quite often performed by the illative-origin forms lauka(n), lauk(o)n, laukonais and the purely adverbial-origin form laũk-16. According to LKŽ data (including the reviewed supplementary card file), perfi- 16 LKŽ (VII 176) associates only the form laukan (and its shorter variant lauka) with a postverbal function. It is called an intensifying particle here, but is also interpreted semantically as the adverb šalin, an inconsistent approach. 84 | ROLANDAS MIKULSKAS The perficientive constructions with these postverbs are most often formed by the same verbs with an annihilative meaning. They too can be divided into the three semantic groups mentioned above; cf.: (i) predicates denoting disappearance: (29) Džiūvo eglės laukan (laukon? – R. M.), išraudonavo ŠtS (LKŽᵥᵢᵢ 177). (30) From those old ones they are running away YL. (ii) predicates denoting destruction: (31) Plough up those meadows (laukon? – R. M.) there is little grass here; the soil will be better ŠtS (LKŽᵥᵢᵢ 176). (32) In the forests of Courland they shot the wolf cubs away (laukon? – R. M.), but not the elk ŠtS (LKŽᵥᵢᵢ 176). (33) And I no longer stash [money]; I have them letting out JDR. (iii) predicates denoting removal (separation): (34) If you find moldy wax and dead bees or brood, then cut them out (laukon? – R. M.) S. DAUK. (LKŽᵥᵢᵢ 176–177). (35) We’ll have to pull up the orchard out PKR. (36) That one was taken to collect mines away after the war AKM. (37) Well, since he had already drunk himself into poverty, when the bailiff was moving him out of the farmstead LŽ. (38) Climb into the barn, into the bear’s flax outdoors SKD. Among the perficientive constructions formed with verbs of the latter group are quite a few ambiguous cases in which postverbs may also be considered to perform the function of modifying the direction of an action in space; cf.: (39) Darys operaciją, reik imti tą blužnį laũk TRK. (40) I only hear them – they’re banging on the door outside (outside? – R. M.) Dr (LKŽᵥᵢᵢ 176). Such a double interpretation is possible particularly in cases where the sentence also contains a construction that locates the beginning of the action, “iš + genitive”; cf.: (41) The sons are banging the door outside, out of the yard; not a single person showed up TRK. The phrase mūša laũk iš kiėmo can be interpreted as: [mūša laũk] iš kiėmo = ‘knocks (someone) out of the yard’ and mūša [laũk iš kiėmo] = ‘(someone) knocks outside from the yard’. In the latter case, the form laũk modifies not so much the verb (mūša) as the noun (kiėmas) with respect to space. 3.1. We showed earlier that the postverbal form žemė(n) is genetically related to predicates denoting downward vertical movement. Similarly, the forms laũka(n), laũko(n), laũkonais, laũk could be associated with verbs of motion, that is, verbs denoting horizontal movement. This time, however, the genetic relationship seems stronger. Indeed, the vast majority of the forms in question used with verbs denoting horizontal movement must already be regarded as spatial adverbs; cf.: POSTVERBŲ PROBLEMA LKŽ 85 (42) Children, let’s run outside! RDN (LKŽVII 184). (43) He headed outside MŽK (LKŽVII 184). (44) We ran out of the hut, afraid that the buildings might catch fire TRKN. (45) Drive the chickens out of the hut SLNT (LKŽVII 184). In one or another such utterance with a motion predicate, a perficientive interpretation might also be discerned, cf.: (46) And you, go to the devil from my head, go, you wretch, outside! PLN. 3.2. With verbs of motion, the forms under discussion are used throughout the area west of the locative’s area of use. Here they generally mean only movement outward and are to be regarded as adverbs. The rapid adverbialization of the locative form laukon was probably driven by the development of the semantic structure of the noun laukas ‘field’ towards abstraction. The most abstract meanings of this noun, ‘an open-air place’ and ‘outside’ (LKŽVII 178), usually designate only one side of the spatial opposition inside vs. outside. This noun no longer has an object-related meaning. Its locative form laukon abstractly indicates direction in space (from the inside to the outside) and has in effect already become an adverb. Such adverbs cannot locate the endpoint (result) of the modified action17. Perhaps because of this inherited semantic feature, the forms lauka(n), laukon, laukonais (and the derivative lauk) are chosen less often than žemė(n) (and the derivative žemyn) for the postverb role in perficientive constructions, since they are less suitable than the latter for expressing the action’s approach to its endpoint. In addressing the treatment and presentation of these postverbial forms in lexicographic description, it should be noted that, because the object-related meaning is lost in one case but retained in the other, the semantic difference between the locative forms lauka(n) (et cet.) and the adverb lauk is much smaller than that between the corresponding forms žemė(n) and žemyn (et cet.). 4. The locative form šalin could also be associated with the expression of an intermediate verb aspect. In LKŽ (XIV 446–447), this form, when used postverbially, is called an adverb that emphasizes the meaning of the verb. Reviewing the examples given here and those in the supplementary card index makes it clear that perficientive constructions with the form šalin consist mainly of verbs meaning ‘remove’ (separate), cf.: (47) He pushed that door away with his shoulder TRK (LKŽXIV 447). (48) Then you’ll pour that water away again VVR (LKŽXIV 447). (49) You walk around like a scarecrow—cut your hair off JDR (LKŽXIV 446). (50) You’ll wash the clothes and won’t take them off until the sun goes down DR (LKŽXIV 447). Only one or another utterance with predicates meaning ‘disappear’ and ‘destroy’ has been recorded, cf.: 17 In utterances with these adverbs, the beginning of the (motion) action is usually located, with this further indicated by internal-location prepositional constructions such as from the hut, through the door, and so on. 86 | ROLANDAS MIKULSKAS (51) The bulbs are drying awayxiv, 447). (52) Those looms have worn out; one must make others Šts (LKŽxiv, 447). In many utterances with predicates meaning removal, the form šalin can be regarded as also performing the function of a spatial adverb. This is not surprising: as an adverb, šalin is an ideal semantic fit with verbs of this meaning. If verbs of this meaning are modified by the form šalin in dialects distant from the area where perficientive constructions are used, then it should be considered a purely adverbial form, cf.: (53) I say, will we have to cut off the whole arm? KIVRŽ (LKŽxiv 446). (54) You’ll have to cut off the branches of the lilac apple tree. SNT (LKŽxiv 446). Given the possible double interpretations of most such examples, it must be acknowledged that perficientive constructions formed with šalin (except, of course, for those rare cases with verbs meaning disappearance or destruction) are not as clear-cut as, for example, those formed with žemė(n). This also makes them difficult to present in a lexicographic description18. The weak expression of the perficientive in such constructions is also evidenced, among other things, by the fact that šalin quite often follows prefixed verbs which, on their own, can already denote a completed action, cf.: (55) They ordered the old barns to be torn down KRT. (56) After that they knocked down that tower LK (LKŽxiv 447). (57) We peel off the bark, then cut [the apples] into large pieces TV. 4.1. As in the previous case, the reasons why the form šalin is unsuitable for expressing the perficientive should be sought in its semantics. As an adverb, this form is widely used throughout the area west of the range where the illative is used. Like lauka(n), it most often modifies verbs of motion or, more generally, horizontal movement. The rapid adverbialization of the illative form šalin was probably brought about by the abstraction of the meaning of the noun šalis. The most abstract meaning of this noun, ‘a place or space situated away from something, a secluded place’ (LKŽxiv, 456), denotes one side of the spatial opposition between center (where the subject is located) and periphery. This noun has already lost its object meaning. Its illative form šalin abstractly indicates a direction in space (from the place where the subject is toward the edge) and is, in effect, already an adverb. Like lauka(n), this form is inherently unable to locate the endpoint of the action it modifies, and perhaps for this reason it is less suitable than žemė(n) for expressing the perficientive. On the other hand, in seeking an answer to the question of how words with such semantics can serve as markers of aspect at all, it is worth noting that the perficientive in verbs of motion modified by the directives laukan and šalin is associated not so much with locating the end of the action as with locating its beginning, which is always presupposed in these constructions. 18At the same location, LKŽ provides examples from both northern and southern dialects. One or another clearer perficientive seems to be mixed in among the examples of adverbial use, cf.: They grind me away, and that’s enough ŽEML (LKŽxiv 446); I picked up a stick to beat that dog away. END (LKŽxiv 446). POSTVERBŲ PROBLEMA LKŽ 87 5. Not all postverbs used in northern Lithuanian dialects are associated with the illative case. The form viršuo used by the Northern Samogitians is most likely of inessive origin19. In the Northern Panevėžys dialect, it corresponds to the forms viršū (viršum), viš. The Dictionary of the Lithuanian Language (LKŽ) interprets all these forms semantically and presents them as adverbs. Since they often recur after certain verbs, they are frequently treated in the Dictionary as constituents of phraseological expressions formed with those verbs (especially if the verbs are used figuratively). This, however, is done inconsistently; cf.: (58) He lunges over, barks: a furrow turned over Všk (LKŽXIX 605). (59) I can’t say anything; in a moment he lunges over Všk (LKŽXIX 621). (60) When he sees it, he lunges over like a boar TL (LKŽXIX 626). The latter example is presented in the Dictionary not as a phraseological expression (as, incidentally, are the other Samogitian sentences nearby). On the other hand, phraseological expressions are quite unjustifiably made here out of utterances that can be understood literally, such as The dog climbs over, if you say nothing Klk (LKŽXIX 626), and so on. 5.1. It is clear that the forms under discussion are not associated with the category of aspect in the LKŽ. A review of the Samogitian utterances presented here with the form viršuo shows that it is most often used with verbs denoting horizontal movement (to lunge, jump, go, drive) and modifies them not only in terms of direction (indicating movement straight toward something, an attack directly at something), but also in terms of intensity (indicating a rapid approach, a sudden attack). Precisely as action intensifiers indicating the sudden approach of an object or phenomenon toward something, they could also be interpreted as grammatical markers of aspect, since they also express an action of movement rapidly approaching its endpoint (a collision or contact). 5.2. Analysis of the utterances shows that if a verb denotes direct (physical) contact with an object, the form viršuo is generally easier to interpret as an adverb of direction; cf.: (61) [The machine] roared, I think, it’s coming straight toward me; I shuddered Krš (LKŽXIX 626). (62) There were roosters, — he’s lunging straight toward me! Rnv (LKŽXIX 626). (63) We’ve run from those bulls, — they’ll charge straight at us Žd (LKŽXIX 626). It may be assumed that the form viršuo inherited the function of a modifier of action direction from prepositional (postpositional) use in utterances of essentially the same kind; cf.: (64) The [machines] race along, and one is driving straight toward the man RDN (LKŽXIX 626). (65) That cow was very wild, it charged straight at people AKM (LKŽXIX 626). 5.3. When these verbs are used figuratively, that is, when they do not denote physical contact, the form viršuo modifying them is more readily treated as a postverb; cf.: 19 A form viršūon is also attested, where n is probably from the illative form viršun: Now the children lunge onto the older one Krš (LKŽXIX 627). 88 | ROLANDAS MIKULSKAS (66) I would not want to give away those children, on top of that mound SD (LKŽxix 626). (67) He jumps straight at you, tearing at your eyes [the little boy] Krš (LKŽxix 626). 5.4. The expression of the perientive is clearest when the form viršuo follows verbs of motion and denotes the (rapid) approach of someone or something directly to the place where the speaker is, cf.: (68) The Germans are coming straight this way, but the Russians are in a muddle Šts (LKŽxix 626). (69) The wind is blowing that stench straight this way Trk (LKŽxix 626). (70) The rain is coming straight this way, hurry up and rake Akm (LKŽxix 626). This is understandable, since the event denoted by verbs modified by the form viršuo necessarily involves the arrival of the agent (or the bringing of the object) at a particular point in space (along a route). 6. Finally, a practical question should be discussed, arising in the application of the term postverb to the forms under consideration when they occur before the verb, cf.: (71) The hen may have something to crow about, [but] its head is cut off Krš. (72) The village began to burn, enslaving some of the inhabitants and driving others out (outward? – R.M.) S. Dauk (LKŽvii 177). Analysis of perientive constructions reveals the following tendency: the purer the expression of the intermediate action, the more rigidly the postverb’s position after the verb is fixed. Conversely, the more the grammatical function performed by the forms in question is mixed with the expression of direction20, the freer their position in the sentence. However, even in the case of clear perientives, the occasional use of the postverb before the verb is entirely understandable if we interpret it as a textual phenomenon, that is, as dependent on the circumstances in which the utterance is produced. Such “inversion” is possible when a strict rule exists that fixes the order of the components of the periphrastic perientive only at a deeper morphosyntactic level. Thus, the use of the term postverb to describe the forms under discussion is entirely justified. CONCLUSIONS 1. There is no doubt that perficientive constructions exist in the northern Lithuanian dialects. They must be represented appropriately in the text of the LKŽ. 2. In the LKŽ, perficientive constructions should be presented alongside the forms of their verbs that indicate intermediate aspect. These forms should be treated as uniformly as possible grammatically. This achieves the greatest possible concentration of the constructions in question in one place, making it possible to determine their geographical distribution. Consistent presentation of these perficientive markers would also make electronic searches in the newly prepared edition of the LKŽ more transparent. 20 Perficientive constructions with postverbs of a different semantic origin, such as kiaurai, ilgai, kiekai and so on, have not been examined. POSTVERBŲ PROBLEMA LKŽ 89 3. The expression of aspect in perficientive constructions used in the northern Lithuanian dialects varies in degree of distinctness. The purity of the perficientive depends on the degree of desemanticization of the forms that indicate its grammatical function. The more desemanticized these forms are, the purer the expression of intermediate aspect. The possibilities of resemanticizing these forms—that is, interpreting them as spatial adverbs—depend in part on the semantics of the predicates in the utterances. 4. There is no theoretical basis for considering the forms in question to be adverbs. In the dictionary, it is sufficient to use the formal term postverb (abbreviated as postv.); where necessary, the definition may use noncommittal terms such as illative form, adverbial form, etc. (depending on the semantic origin of those forms). 5. The most widespread are perficientive constructions with the postverb of illative origin žemė(n) (into the ground) and its local variants of adverbial origin (in the northwestern Samogitian dialect) žemyn (downward) and žemynais (downward). These postverbs usually follow verbs meaning to disappear, destroy, and remove (separate). The perficientive meaning of these periphrastic constructions is the most distinct. Only a few cases are ambiguous, when the aforementioned verbs also express vertical downward motion (in which case the forms in question may also be understood as spatial adverbs). 6. Perficientive constructions with the postverb žemė(n) are used throughout Lithuania’s border region with Latvia. The southern boundary of the distribution of these constructions may be taken as the general southern limit of perficientive use in the northern Lithuanian dialects, and used as a guide in cases where the periphrastic constructions in question can be interpreted in two ways. 7. Postverbs of illative origin lauka(n), (lauko(n), laūkonais) and šalin should be regarded as forms of adverbial origin, as should the form laū́k. Perficientive constructions with these postverbs are mixed in type, since they can often also be interpreted as modifiers of the direction of an action in space. The perficientive expression is somewhat clearer in constructions formed with these postverbs and the same verbs of annihilative meaning. Forms following verbs of horizontal (sliding) motion are most readily interpreted adverbially, so their postverbal function is the most “obscured.” 8. A limited number of verbs denoting attack or sudden motion form perficientive constructions with the Samogitian form of inessive origin viršūo. In many cases, this form modifies the event expressed by the sentence both in terms of quantity (indicating the intensity of movement) and direction in space (straight toward the object), and therefore is not a purely grammatical (aspectual) marker. The postverbal function of the form viršūo is most clearly revealed when it follows verbs of motion. By expressing the sudden approach of something directly toward the subject of the utterance, it also expresses the end of the motion. 9. In the dictionary, these forms, which function as markers of intermediate aspect, should be listed separately from adverbs under the label postv.. Ideally, of course, they would be given separate entries. To present as many perficientive constructions as possible in one place, local variants should also be listed alongside the main forms wherever possible. For example, the LKŽ methodology does not allow the forms žemė(n) and žemyn (downward) to be included in one lexicographic entry, but it is appropriate (and, as we have shown, there is also a theoretical basis for this) to list alongside perficientives with the synthetic illative form žemė(n) 90 ROLANDAS MIKULSKAS examples with the periphrastic illative form į žemę (or vice versa, as was done here in the entry žemė). Similarly, alongside constructions with the postverb žemyn (žemyn), one could also include examples with its longer variant žemynais (žemynais) in a single entry, as is done with adverbs in the entry žemas, -a TABLE 1. UTTERANCES WITH PERIPHRASTIC CONSTRUCTIONS PREDICATES (VERBS OF DISAPPEARANCE) Verb | žemėn | žemė | į žemę | žemyn | žemyn | žemynais | žemynais die | PLNG, LG, SKD, TRKN, ŠTS, YL, TRK, AKM, VGR, MŽK x 2 | KRŠ, TRŠ | TL | DR, GRŠL, PLT | SD, VGR | LNK | SK burst | KLVRŽ, TRK x 3 | RDN, KRŠ | — | DR, SKD, BRS | SD, TRK | — | — fade | KRT, PLT, x 2, KLVRŽ, AKM | KRŠ | — | SD | — | — | — benġtis | TRK | — | — | — | — | — | — šlóṗti | AKM | — | — | — | — | — | — káršinties | ŠTS | — | — | — | YL | — | — kandėti | ŠTS (of fur) | — | — | — | — | — | — šálti | ŽEML, MŽK, TRK x 3 | — | PP | — | KLVRŽ x 2, TRK | — | — drown | — | — | VKŠ, LŽ | ŠTS | — | — | — go ‘slide’ | VGR (of a moustache), RDN (of hair) | — | — | — | — | — | — POSTVERBŲ PROBLEMA LKŽ 91 CONTINUATION OF TABLE 1 Verb žemėn žemė į žemę žemyn žemyn žemynais žemynais šėrties Šts birti Ms (about blades) to cut YL to fall SKD (about heads in war) to tumble SKD to deteriorate SKD TABLE 2. PREDICATES OF UTTERANCES WITH PERFICIENTIVE CONSTRUCTIONS (VERBS OF DESTRUCTION) Verb to the ground earth into the earth down down down down shoot neighbor KL, [person] PLT, [person] KRTN, [person] ALS, [person] YL, [person] ŽEML, that snake TRK, [person] SK, woman VŠK, parents PS, communist PPL, man KPR [person] KAL, [person] SD, [person] GD such a snake SD, person ŽD x 2, [person] YL, [person] SD beat child TRK, father TRK [person] TRKN [person] YL 92 ROLANDAS MIKULSKAS CONTINUATION OF TABLE 2 Verb down to the ground ground to the ground downward downward downwards downwards [object] TRK, [person] YL, flies END (to) beat that person PLNG bengti [person] ŽEML, woman TRK oneself SD take a life [a person] SD hang [a person] LC [a person] YL strangle [a person]TRK [a person] TRK, [a person] Ms scream a boar KRŠ shoot ‘to shoot’ a man-killer KRŠ murderers PVN kill [a person] YL torment [crow] PLT (to) mow Vištytis ŠTS to beat old woman VKŠ (to) beat the old men ŽEML to graze clover ŠTS crops PLNG POSTVERBŲ PROBLEMA LKŽ 93 TABLE 2 (CONTINUED) Verb earthward earth to earth downward downward downwards downwards press (of a crowding crush in a church) SLNT (of pain) ŠTS thresh flax ŠTS freeze (of severe frost) ŠTS strike [a person] ŠTS cut (of illness) Sd TABLE 3. PREDICATES OF UTTERANCES WITH PERIPHERAL CONSTRUCTIONS PREDICATES (VERBS DENOTING REMOVAL (SEPARATION)) Verb downward earth to the ground downward downward downwards downwards take a skin YL, a cross TrK, a scarf KlK, a hat KlK, an arm KlK, valuables Vgr, a shell JDP, beets AKN, dust Pp pull (oneself) a jacket JNŠ, [clothes] VŠK a jacket LK harness a horse KlK, a horse ŽG, mares KlvrŽ [a horse] TrŠ 94 | ROLANDAS MIKULSKAS TABLE 3 (CONTINUED) Verb earthward earth to earth downward downward downwards downwards throw [food] PLN, a patch LNKV, beer KRČ smoking KRŠ to cut trees VŽ, head KL, leg TRK, head VB head KrŠ head LK leg TrK to demolish [house] KRTN, houses ŠTS, school KLK, little house SLČ [houses] YL to cut beard SD, ports SD, [material] YL, material ŽG to saw [crops] BDR, plaster BDR, pear ŽG, legs VKŠ hazel AKM leg KRG to pull up flax KLVRŽ weeds KRŠ potatoes KRTN, flax JDR, flax KLP to bite potatoes SD potatoes TrK to break arm TrK to pick bees MŠK (to) tear off onion skin Č to scratch pimple VKŠ to carry pot KL POSTVERBŲ PROBLEMA LKŽ 95 TABLE 3 CONTINUED Verb down to the ground ground to the ground downwards downwards downwards downwards to raise pot KL frying pan Lk [pot] SD máuti(s) a ring TrK trousers ŠAUK socks GrŠL rėdyti(s) [a person] TrK daryties LKŽ Ms ‘to get dressed’ to gape burnosus PlT lūpti birch trees AkM tėpti honey ŠtS (nu)mėsti village head JDR ritinti old man RDN versti government Skp ABBREVIATIONS LKAIII – Atlas of the Lithuanian Language 3, Morphology, ed. K. Morkūnas, Vilnius: Mokslas, 1991. LKE – Encyclopedia of the Lithuanian Language, ed. V. Ambrazas, Vilnius: Institute of Science and Encyclopaedia Publishing, 1999. LKŽVII – Dictionary of the Lithuanian Language 7, ed. J. Kruopas, Vilnius: Mintis, 1966. LKŽXIV – Dictionary of the Lithuanian Language 14, ed. K. Ulvydas, Vilnius: Mokslas, 1986. LKŽXIX – Dictionary of the Lithuanian Language 19, ed. V. Vitkauskas, Vilnius: Institute of Science and Encyclopaedia Publishing, 1999. LKŽXX – Dictionary of the Lithuanian Language 20, ed. V. Vitkauskas, Vilnius: Institute of the Lithuanian Language, 2002. LITERATŪRA Ambrazas, V., red., 1997: Dabartinės lietuvių kalbos gramatika. Trečiasis pataisytas leidimas, Vilnius: Mokslo ir enciklopedijų leidybos institutas. Daujotytė, V. 2002: Pamatinis kultūros tekstas. Dvidešimt lietuvių kalbos žodyno tomų, sudarė J. Zabarskaitė, Z. Šimėnaitė, Vilnius: Lietuvių kalbos institutas, 34–35. Girdenis, A. 1984: Stratifikacinis lietuvių kalbos veiksmažodžio morfologijos modelis. Kalbotyra 35(1), 25–30. Girdenis, A., Kačiuškienė, G. 1986: Paraleliniai reiškiniai latvių ir šiaurinių lietuvių veiksmažodžio sistemo­se. Kalbotyra 37 (1), 21–27. Gudavičius, A. 2000: Etnolingvistika, Šiauliai: Šiaulių universitetas. 96 | ROLANDAS MIKULSKAS LAIGONAITĖ, A. 1957: Vietininkų reikšmė ir vartosena dabartinėje lietuvių kalboje, Vilnius: Valstybinė politinės ir mokslinės literatūros leidykla. LIPARTĖ, E. 2000: Inatyvo resp. prieveiksmio laukan reikšmės ir galimi sinonimai lietuvių kalbos tarmėse. Acta Linguistica Lithuanica 43, 144–151. PAULAUSKIENĖ, A. 1989: Gramatinės lietuvių kalbos vardažodžių kategorijos, Vilnius: Mokslas. ULVYDAS, K., red., 1971: Lietuvių kalbos gramatika 2, Vilnius: Mintis. ZINKEVIČIUS, Z. 1966: Lietuvių dialektologija, Vilnius: Mintis. ZINKEVIČIUS, Z. 1980: Lietuvių kalbos istorinė gramatika 1, Vilnius: Mokslas. Gauta 2003 04 22 Rolandas Mikulskas Lietuvių kalbos institutas P. Vileišio g. 5, 2055 Vilnius, Lietuva