Paprastųjų ir įvardžiuotinių būdvardžių opozicija lietuvių kalboje kaip apibrėžtumo sistema

Acta Linguistica Lithuanica LIX (2008), 109 – 139 The Opposition between Simple and Pronominal Adjectives in Lithuanian as a Definiteness System Birutė Spraunienė Vilnius University This article deals with the so-called pronominal forms of Lithuanian adjectivals in relation with definiteness. The pronominal morpheme added to the adjectival form is interpreted as a marker of definiteness, comparable to the definite article of other languages, and it is suggested that the adjectival forms themselves should be called definite. The aim of the article is to establish to what extent and in what contexts the use of these definite forms correlates with the semantic definiteness of the NP. An empirical investigation shows that in many cases the use of the definite adjective is at best optional, sometimes even ungrammatical, though the NP clearly functions as a definite description. This points to a major difference between the use of Lithuanian definite adjectives and the marking of definiteness in article languages. In Lithuanian the context is decisive for the definite or indefinite interpretation of the NP, and the selection of definite and indefinite adjectival forms depends, among other things, on information structure, i. e., on whether the adjective contains known (presupposed) or new information. 1. THE CATEGORY OF DEFINITENESS AND ITS SEMANTICS The literature recognizes that the primary function of the category of definiteness is to coordinate reference, which is necessary for communication to take place. Languages have various means that help participants in a communicative act solve problems of reference coordination. These include proper names, pronouns, and so on. In article languages, the principal means of reference coordination are formal markers of definiteness vs. indefiniteness—the articles. Christopher Lyons defines the concept of definiteness as follows: “de- 109 Birute Sprauniene finiteness is a morphosyntactic category, grammaticalizing a pragmatic category of identifiability” (Lyons 1999: 282). This means that definiteness is associated with the possibility of identifying the referent. The referent of a definite noun phrase is presented to the addressee as uniquely identifiable, whereas the use of an indefinite noun phrase means that its referent is being newly introduced into the discourse and therefore cannot be uniquely identified (see Chafe 1976; Hawkins 1978 and 1991; Haspelmath 1999; Lyons 1999; Hansen & Heltoft (forthcoming)). It is important to emphasize that in the case of definite descriptions1 the referent is presented to the addressee as identifiable. The referent may have been unknown to the addressee before the utterance, but the definite description itself is like an instruction to the addressee to look for the only possible referent until they find it. According to Hawkins, the search for the referent is conducted not across the entire discourse space, but within a particular pragmatic set of entities (Eng. pragmatic set of entities / P-set). Hawkins’s theory, developed on the basis of English-language facts, can be summarized as follows: A noun phrase is definite (i) if its referent is locatable in a pragmatic set of entities shared by the speaker and the hearer, that is if it is pragmatically anchored, and (ii) if the referent is unique (or, more precisely, inclusive) within this set. (Haspelmath 1999: 231) Thus, there must exist a particular pragmatic set of entities that is manifest (Eng. manifest) to both speaker and addressee, and in which the referent of the noun phrase exists and is unique. In this case, we have a definite noun phrase. By using the definite (i.e., marked) article, the speaker indicates that the addressee will be able to determine which referent or referents are intended by finding the pragmatic set in which it (they) exist(s). The logical property of uniqueness facilitates the search—there can be no other entities in this set that satisfy the content of the definite description (Hawkins 1991: 414). Such a set may consist of entities previously mentioned by the speaker and hearer (“the set of the previous discourse”) or of the immediate conversational setting in which the speaker and hearer are situated. Entities connected by whole–part or other associative relations also form pragmatic sets, e.g., a car and a steering wheel, a wedding and a bride. Hawkins argues that it is precisely in these 1 Dėl termino žr. Holvoet, Tamulionienė (2006: 13–14 ir išn. 3), taip pat Hawkins (1991: 406). 110 Paprastųjų ir įvardžiuotinių būdvardžių opozicija lietuvių kalboje pragmatic sets that the requirement of uniqueness is satisfied, helping the speaker and hearer cooperate and refer unambiguously to the individual that matches the content of the definite description (ibid., 408). Thus, definite descriptions are characterized by uniqueness, whereas an indefinite description, according to Hawkins, is neutral with respect to uniqueness (ibid., 420). For example: (1) A professor is drunk2. Sentence (1) will be true if there is at least one professor who is drunk. It is logically compatible with the claim that there may be more than one such professor, as well as with the claim that there is only one, so sentence (1) is “neutral with respect to uniqueness and does not contradict it”3 (ibid., 407). The use of the indefinite article generally means that the entity in question does not belong to any pragmatic set that is manifest to both speaker and hearer. However, Hawkins notes that the indefinite article can also mark membership in a pragmatic set; for example, the sentence Pass me a cup can be interpreted as: Pass me one of the cups in front of you on the table. Hawkins applies Grice’s theory of implicature to the analysis of English articles. He argues that phrases with the indefinite article can implicate membership in a pragmatic set, but this is a conversational implicature (Eng. conversational implicature), which can be cancelled, for example: (2) A senator resigns. Example (2), in a US newspaper headline, may carry the conversational implicature that it refers to a US senator. In example (3), we see that this implicature can be cancelled (Hawkins 1991: 418, example taken from the author): (3) A senator resigns, not a US senator, but a senator in Japan. With the use of the definite article, membership in a pragmatic set is a conventional implicature that cannot be cancelled (ibid., 414). Hawkins discusses somewhat paradoxical cases in which indefinite descriptions in English denote unique referents (examples taken from the author), cf.: 2 Pavyzdys paimtas iš Hawkinso (Hawkins 1991: 407). 3 „[…] is actually neutral to uniqueness and does not contradict it.“ 111 Birute Sprauniene (4) England has a prime minister, and America has a president. (5) This larger situation set contains a professor, a classroom, a textbook, etc. Predicates have, contain, comprise, etc., indicate that entities belong to a certain set. That uniqueness of the indefinites implied in examples (4–5) can be seen by replacing the unique entities in them with non-unique ones. The utterance becomes odd: (6)?England has a member of parliament, and America has a senator. Hawkins observes that definite articles can be used in examples (4–5), but the meaning of the sentences changes completely (Hawkins 1991: 421), cf.: (5’) England has the prime minister, and America has the president. Example (5’) implies that the prime minister of another country is visiting England, or that John Major is in charge of England’s affairs and George Bush of America’s. In other words, these examples are naturally interpreted as meaning that the referent of the definite noun phrase does not belong to the set defined in the examples with the corresponding indefinite descriptions, and that the listener must find another set in which the requirements of existence and uniqueness are met (ibid., 422). 2. TYPES OF DEFINITENESS Based on the grounds for identifying a referent, the literature distinguishes four main types of definiteness (cf. Chafe 1976; Hawkins 1978 and 1991; Haspelmath 1999; Lyons 1999; Hansen & Heltoft (forthcoming)): 1. anaphoric definiteness (English anaphoric uses, Danish anaforisk bestemthed), 2. deictic definiteness (English immediate situation uses, Danish deiktisk bestemthed), 3. situational definiteness (English larger situation uses, Danish baggrundsbestemthed), 4. inferential definiteness (English inferrability, associative-anaphoric uses, Danish inferentiel bestemthed). 112 Paprastųjų ir įvardžiuotinių būdvardžių opozicija lietuvių kalboje Cases are classified as anaphoric definiteness when the referent of a noun phrase is known from the linguistic context because it was mentioned earlier: (7) a miraculous light appeared <…> not one <…> caught sight of the miraculous light4 Anaphoric definiteness coincides with the demonstrative function of pronominal adjectives in Lithuanian distinguished by Adelė Valeckienė (Valeckienė 1986: 171). Anaphoric definiteness occurs not only when the repeated noun and its antecedent are identical nouns, but also when they are coreferential synonyms (e.g.: A certain young man had arrived …. That fellow) or when the antecedent is a hyponym and the repeated noun is its hyperonym (e.g.: A car crashed into a pole. The vehicle was uninsured) (Hansen & Heltoft, forthcoming). In the case of deictic definiteness, the referent is identifiable because it is visible, audible, or otherwise perceptible in the immediate conversational environment, e.g.: Hand me the hammer! Where is that sound coming from? Situational definiteness occurs when the basis for identifying the referent is the interlocutors’ shared knowledge of physical objects or social and cultural phenomena in their environment. For example, when saying krig-en (war-DEF), Danes will generally mean the Second World War, while the phrase the Prime Minister will mean ‘the British Prime Minister’ to the English. Inferential definiteness occurs when the referent of a definite noun phrase is identified through its association with another, non-identical referent that has already been introduced into the discourse. Nouns linked by inferential definiteness often stand in whole–part relations (where the part is unique to the entity), e.g.: house …. roof; car … engine. This type of definiteness is also called associative anaphora. It is important to note that in article languages the first two types of definiteness may be expressed by both definite articles and demonstrative pronouns, whereas situational and inferential definiteness are expressed only by articles (Hansen & Heltoft, forthcoming). 4 The example is taken from A Grammar of Contemporary Lithuanian (Valeckienė in: Ambrazas, ed., 1997: 175). 113 BIRUTĖ SPRAUNIENĖ 3. THE LIFE CYCLE OF THE DEFINITE ARTICLE In the literature, the emergence of definiteness marking is associated with the reanalysis of the adjectival demonstrative pronoun, through which the pronoun acquires an article function (Lyons 1999: 323). Once formed, the article gradually spreads into an increasing number of usage contexts. In the literature, this process is called the life cycle of the definite article and is divided into the following stages (cf. Greenberg apud Hawkins 2004: 84–86, Lyons 1999: 337–340): First stage. The demonstrative loses its deictic meaning, and the explicit or implicit contrast between objects near to or far from the speaker or addressee thus disappears. The newly formed article is restricted to anaphoric (a book... the book) and deictic contexts (Pass me the book, please!). Second stage. The article also begins to be used in situational (the king, the sun) and inferential contexts (a house... the roof). Third stage. The definite article extends into generic contexts, i.e., statements that say something not about individual entities but about their kinds5 (The lion is a mammal ‘A lion is a mammal’). According to Hawkins, modern Germanic languages are at this stage. Fourth stage. The definite article begins to be used in non-definite specific contexts, where it marks only the existence of the referent, not its uniqueness6. These four stages illustrate the gradual expansion of the definite article’s distribution in a language. Each stage retains the distributional possibilities of the preceding stage and adds new ones, thereby increasing the polysemy of phrases with articles. Article frequency rises because the article becomes obligatory in an increasing number of grammatical contexts and constructions (e.g., with predicatively used noun phrases: John is a teacher / the best teacher I’ve ever had). Ultimately, the definite article may lose any 5 Cf. the term “generic definiteness” used by Holvoet and Tamulionienė (Holvoet, Tamulionienė 2006: 19). 6 A similar phenomenon is observed in colloquial American English, where the stressed demonstrative pronoun this came to be used with referential objects, while the distribution of the indefinite article a became restricted to non-referential contexts (She wanted to buy this house – ‘She wanted to buy a house’ and She wanted to buy a house – ‘She wanted to buy (any) house’ (Givón 1995: 61, examples taken from Givón). 114 Paprastųjų ir įvardžiuotinių būdvardžių opozicija lietuvių kalboje become associated with definiteness and serve as an indicator of gender and/or number (cf. Greenberg’s notion of nominality marker). In many languages, this development has led to the emergence of a new definite article. The article life cycle partly explains the variation in the use of definiteness markers across different languages. This means that a noun phrase may be formally definite in one language, while its equivalent in another language is formally indefinite, even though they are referentially identical (cf. Lyons 1999: 160). The article life cycle also shows that the semantics of a definiteness marker gradually changes in a language. 4. DEFINITENESS MARKING IN LITHUANIAN Lithuanian is an articleless language, and the expression of definiteness in it is only weakly grammaticalized. As Mikulskas observes, Lithuanian grammars have not investigated in which cases and what kind of definiteness vs. indefiniteness marking is obligatory (Mikulskas 2006: 33). One means of coordinating reference in Lithuanian is the use of various pronouns. Demonstrative pronouns function as definiteness markers; some of them (e.g., tas and, in some cases, šis, šitas) have, according to Rosinas, become artroids (Rosinas 1972: 83ff.; 1996: 72). Artroids are linked to the definite articles of article languages by their loss of stress and faded deictic meaning, while they differ from genuine articles in lacking the feature of obligatoriness. Artroids tas, ta are very common in colloquial speech and are also making their way into written language. The function of indefiniteness markers in Lithuanian is performed by various pronouns with an existential meaning, e.g., vienas, kažkoks, kažin koks, and the demonstrative pronoun toks (Rosinas 1996: 69; cf. also Girdenienė 1971: 45; Holvoet, Tamulionienė 2006: 18; Mikulskas 2006: 33). Of course, in many cases definiteness in Lithuanian is expressed indirectly—by word order, case oppositions (nominative / accusative vs. genitive), or is simply inferred from context (see Holvoet, Tamulionienė 2006: 30–32; cf. also Rosinas 1996: 62–63). The only morpheme that could be associated with formal definiteness marking in Lithuanian is the pronominal morpheme of pronominal adjective forms. This article will focus primarily on the opposition between simple and pronominal forms in contemporary Lithuanian, with the aim of investigating the distributional properties of these forms, differences in their functions and meanings, and their relationship to the category of definiteness. If the pronominal morpheme of pronominal forms 115 BIRUTĖ SPRAUNIENĖ -(i)is, -ji is interpreted as a definiteness marker, we may ask at what stage of the article life cycle it is, in which contexts it is used, when its use is obligatory, when optional, and when it is ungrammatical. According to Mikulskas, when investigating issues related to definiteness in Lithuanian, it is useful to take as a point of comparison an article language in which definiteness marking is grammaticalized (Mikulskas 2006: 33). Therefore, this article compares pronominal adjective (and participle) forms with articles in article languages (Danish and English). 5. DEFINITE AND INDEFINITE ADJECTIVES The idea that the opposition between simple and pronominal adjective forms expresses the category of definiteness in Lithuanian is not new. The functions of pronominal forms in Lithuanian linguistics have been studied most comprehensively by Adelė Valeckienė (Valeckienė 1957 and 1986); her research also underlies the account of these forms both in the academic Lietuvių kalbos gramatikoje (Ul vydas, ed., 1965) and in Dabartinės lietuvių kalbos gramatikoje (Ambrazas, ed., 1997). The latter treats definiteness as an adjective category (cf. the statement that adjectives “have the categories of degree and definiteness” (Valeckienė in: Ambrazas, ed., 1997: 167)). This view is rightly criticized in the fourth volume of Lietuvių kalbos gramatikos darbų, Daiktavardinio junginio tyrimai, which contains two articles devoted to the issue of definiteness. The authors argue that definiteness in Lithuanian, as in typical article languages, should be regarded as a noun-phrase category. Mikulskas convincingly shows that the pronominal adjective morpheme acquires grammatical value not through the opposition of simple and pronominal adjective forms (e.g., mėlyna vs. mėlynoji), but through the corresponding noun phrases (e.g., mėlyna skarelė vs. mėlynoji skarelė) (Mikulskas 2006: 59, 61). Holvoet and Tamulionienė likewise argue that pronominal adjective forms convey definiteness not to the adjective but to the entire noun phrase. The authors conclude that “the locus of expression and the scope of the category of definiteness do not coincide in this case” (Holvoet, Tamulionienė 2006: 13). Incidentally, the wording used in Dabartinės lietuvių kalbos gramatikoje—“simple forms denote a property of an entity, while pronominal forms, in addition to the property of the entity, also indicate that the ENTITY (emphasis mine—B. S.) is known, definite” (Valeckienė: in Ambrazas, ed., 1997: 174)—also 116 Paprastųjų ir įvardžiuotinių būdvardžių opozicija lietuvių kalboje implies that the entity itself, rather than its property, becomes definite as a result of the pronominal form. It should also be mentioned that only qualitative adjectives can take pronominal forms (and, incidentally, not all of them: e.g., baltmargis, gerutis, saldinis have no pronominal forms); ordinal numerals, participles, and some pronouns can also do so (Paulauskienė 1994: 220). Relative adjectives cannot take pronominal forms, which distinguishes them from qualitative adjectives. One may ask whether an adjective’s ability or inability to take a pronominal morpheme depends on its derivational properties (qualitative adjectives that cannot take pronominal forms are linked to relative adjectives not by meaning but by derivation—both gerutis and saldinis are suffixal derivatives formed from a stem, while baltmargis is a compound formed from a stem) or on its meaning (cf. Paulauskienė’s claim that pronominal forms can be taken by “all words that denote a non-relative property inherent in the entity itself” (ibid.)? It should be noted that Lithuanian adjectives in the neuter gender have no pronominal forms (e.g., šalta, brangu), since they are not used with nouns. Pronominal forms are also not characteristic of half-participles (e.g., eidamas) or active past frequentative participles (e.g., eidavęs), since these likewise are not used to express properties of entities (Ambrazas, ed., 1997: 367). This could provide an additional argument in favor of the view that the domain of the category of definiteness is the noun phrase. Incidentally, the term “noun phrase” itself is defined differently in the Lithuanian linguistic tradition than noun phrase in Western linguistics. According to Labutis, “word combinations whose head is a noun are considered noun phrases” (Labutis 1998: 68). On this view, pronouns and their combinations (e.g., jis, šis, tas su geltonomis garbanomis) and adjectives (e.g., aktyvioji, pastarasis) in examples (8–12)7 should be considered pronoun phrases and adjective phrases, respectively. On the other hand, it is clear that, in terms of syntactic function, the words mentioned above represent a noun phrase in the sentence; cf.: (8) They found the priest Anupas almost lifeless. He was burning with a living fire. (9) After A. Valinskas gave him the floor, this one apologized on behalf of his colleague J. Jankevičius (10) For a time Bona had taken control of Plateliai and Kaunas (the latter was bought back from her in 1545) (11) one of them, the one with blond curls, is my cousin Stasys 7 Examples (8–11) are taken from KT. 117 BIRUTĖ SPRAUNIENĖ (12) It so happened that I was flying on a plane, and a pair of ladies was sitting in front of me. <...> One of them, leaning over the seat, tried to tell me something. I heard nothing and understood nothing. But my active fellow passenger would not leave me alone <...>. The active one nodded amiably and turned away. HK 200–201 It should be noted that in spoken Lithuanian, in certain cases even the simple adjective form can perform the function of coordinating reference and stand for a noun phrase, e.g.: (13) At the market I bought a red rose and a white one. The red one I will plant in my garden, while the white one I will take to my mother as a gift8. The adjectives red and white in example (13) should, from a functional (or syntactic) perspective, be regarded as elliptical, reduced noun phrases, since, like pronouns, they are used in place of nouns, that is, referentially. It is worth considering whether, in defining a noun phrase, one should take into account only its prototypical structure, in which the head is a noun, or whether it would be more appropriate to expand Labutis’s definition of a noun phrase so that it also covers non-prototypical (reduced, elliptical) noun phrases. In the remainder of this article, following Holvoet and Tamulionienė, the “simple” and “pronominal” forms will be referred to respectively as “indefinite” and “definite” forms, thus emphasizing the function of the inflectional morpheme under discussion. 6. ANALYSIS OF THE OPPOSITION BETWEEN DEFINITE AND INDEFINITE ADJECTIVE FORMS Valeckienė characterizes the contrast between the simple and pronominal forms as follows: “simple forms denote a property of an object, whereas pronominal forms, in addition to the property of the object, also indicate that the object is known, definite” (Valeckienė in: Ambrazas, ed., 1997: 174, Valeckienė 1986: 169). The author claims that simple and pronominal forms in Lithuanian constitute a binary, asymmetrical opposition, whose marked member is the pronominal form, bearing the feature [+Definite], while the simple form is the unmarked member, bearing the feature [−Definite] (it should be recalled that this author considers definiteness 8 Example (13) was devised by me – B.S. 118 Paprastųjų ir įvardžiuotinių būdvardžių opozicija lietuvių kalboje the adjective category, and therefore the definiteness/indefiniteness features are assigned to the adjective itself). This is illustrated by the following diagram: 1 schema. Paprastųjų ir įvardžiuotinių formų opozicija (plg. Valeckienė 1986: 169) Paulauskienė rightly criticizes this analysis, arguing that “the meaning of the marked member cannot be broader than that of the unmarked member” (Paulauskienė 1994: 224). I would suggest modifying Valeckienė’s diagram as follows: Diagram 2. Opposition between indefinite and definite forms This would mean that definite forms mark the definiteness of the noun phrase, whereas indefinite forms are indifferent or neutral with respect to this marking9, and thus have a broader functional meaning than the marked member of this opposition. A noun phrase with an indefinite adjective can be either semantically indefinite or definite; the context determines which interpretation applies, while the adjective form itself indicates nothing. I will also attempt to demonstrate that the pronominal form marks the definiteness of the noun phrase. One argument is that this form cannot be used with words indicating the indefiniteness of a noun phrase or in cases where the phrase is rhematic, i.e., where the noun modified by the adjective conveys new information (cf. Larsen’s term new rheme (Larsen 2005: 46)), e.g.: 9The essential difference between the marked and unmarked members of any asymmetrical opposition is similarly described as follows: “the difference <...> is not that between A and non-A, but between A and indifference between A and non-A” (Haspelmath 2006: 29). Hjelmslev also characterizes the contrast between the marked and unmarked members of an opposition in this way apud Jensen (2007: 304). 119 BIRUTĖ SPRAUNIENĖ (14) Someone called him one *the famous / famous journalist. (15) A sound was heard: some *the strange / strange sound. (16) On the windowsill sat *the white / white cat. A definite form is possible in such contexts only if the adjective denotes an inherent species-specific property of the object, that is, in terminological compounds: (17) Ant mūsų stogo buvo nutūpęs baltasis gandras. The compound white stork means “some (any) member of the white stork subspecies.” Thus, in this case the referent of the compound is indefinite (the addressee cannot identify it), but the white stork subspecies itself is known to the addressee (or is presented as uniquely identifiable), which is why the definite form of the adjective is used. As Holvoetas and Tamulionienė state, “taxonomic definiteness” in such compounds is “in a sense fossilized” (Holvoet, Tamulionienė 2006: 28). Similar phenomena can be found in other languages as well; for example, in Hebrew, a species name can be used with the definite article even when any member of the species is meant (ibid., n. 10). Many Lithuanian terms are formed with definite adjective forms. If a term is formed with a definite form, that form is obligatory in all contexts; however, in colloquial speech there is a tendency to drop the pronominal element, perhaps because of a tendency to shorten words. One often hears laptop, mobile phone or simply mobile. Of course, such usage is nonstandard, but it shows that the indefinite form, being unmarked, can displace the definite form even in terminological compounds. Some terminological compounds occur with both the indefinite and definite forms, for example, red / red wine. One might suspect that some pairs of more conventional modifying compounds increasingly express qualitative rather than taxonomic distinctions, a fact reflected in their names through the use not of “terminologized” definite adjective forms, but of indefinite forms, which are unmarked in this respect; cf. white vs. red wine, white vs. black bread, sweet vs. sour milk. Do definite forms always express definiteness? In the Grammar of Contemporary Lithuanian and in Valeckienė’s work, it is argued that in certain cases these forms have an emphatic meaning. Examples of this usage abound in folklore and literature: dear good mother, old father, swift steed, bright sun, terrible dragon, black abyss. It is claimed that here 120 Paprastųjų ir įvardžiuotinių būdvardžių opozicija lietuvių kalboje They are virtuous with fixed epithets and merely emphasize a characteristic property of the thing (Valeckienė 1998: 266ff.). I would suggest that these compounds also denote particular types; the adjectives in them are usually uninformative and often express a property that follows from the meaning of the noun itself (the sun is always bright, a pine is always terrible, and so on)10. Mikulskas does not rule out the possibility that these fixed compounds, too, can be interpreted as definite, except that their definiteness is supposedly understood in a particular sociocultural, say, ethnographic context, and resembles the role definiteness discussed in the literature (Mikulskas 2006: 54–55, n. 16). 7. EMPIRICAL STUDY An empirical study was conducted to compare the expression of definiteness in Lithuanian and article languages (Danish and English). The research material was collected from Icchokas Meras’s novel A Stalemate Lasts for a Moment. This novel was selected because it has been translated into both Danish and English. All noun phrases with adjectival modifiers in the Lithuanian text that had been translated into the article languages—Danish and English—as definite noun phrases with definite articles and other definite determiners (demonstrative pronouns and possessives), i.e. interpreted by the translators as semantically definite, were extracted. Only phrases with qualitative adjectives and participles that can have a definite form in Lithuanian were selected. The aim of the study was to establish how often and in which cases semantically definite phrases are used with the definite form of the adjective in Lithuanian. It also sought to determine how often the definite form of the adjective is used to express different types of definiteness (see Section 2). Generic noun phrases were not examined. The research material consists of 216 Lithuanian noun phrases and an equal number of Danish (and English) noun phrases. 10 Karin Larsen, who studied the distribution of short (indefinite) and long (definite) adjective forms in Old Russian, observes that in certain cases the adjective in phrases with the long form is redundant; for example, in the phrase многоценное богатство (English a very valuable object of value), the meaning of the adjective многоценное follows from that of the noun богатство, while in the phrase волк хищный the adjective хищный denotes a property generally associated with a wolf (Larsen 2005: 238; see also Larsen 2007: 37). 121 Birutė Spraunienė 8. RESEARCH RESULTS The results of the study are presented in Table 1: Table 1. Lithuanian noun phrases with definite and indefinite adjective forms Phrases with ADEF¹¹ | Phrase structure | Number of examples |% | Phrases with AINDEF | Phrase structure | Number of examples |% Adef | 6 | 2.8 | Aindef | — | — ADEF + N | 33 | 15.2 | AINDEF + N | 110 | 50.9 DEM + ADEF + N | 2 | 0.9 | DEM + AINDEF + N | 14 | 6.5 Pos + Adef | 2 | 0.9 | Pos + AINDEF | — | — Pos + ADEF + N | 9 | 4.2 | Pos + AINDEF + N | 39 | 18 DEM + POS + ADEF + N | — | — | DEM + Pos + AINDEF + N | 1 | 0.5 Total: | 52 | 24 | Total: | 164 | 76 As Table 1 shows, phrases with the definite form of the adjective constitute only a quarter of all phrases (24%). The remaining 76% of phrases use the indefinite form of the adjective. Thus, if we assume that all the phrases examined denote unequivocally identifiable referents, since they are rendered as definite noun phrases in Danish and/or English, this would mean that in most cases context alone is sufficient to identify the referent in Lithuanian, and the definite form of the adjective is unnecessary. Since demonstrative pronouns and possessives themselves make a phrase definite, the definite form of the adjective may be perceived as redundant with these pronouns (cf. Holvoet, Tamulionienė 2006: 26); this is probably why the indefinite form is used in most such phrases (incidentally, Latvian has chosen a different strategy for marking definiteness in such cases—in attributive phrases determined by demonstrative pronouns or possessives, the definite form of the adjective is obligatory, ibid. 23–24). However, if we set aside phrases with pronouns, we see that noun phrases with the definite form (with or without a noun) constitute 18 per 11 Explanations of the abbreviations used in the table: ADEF – definite adjective, AINDEF – indefinite adjective, N – noun, POS – possessive pronoun or genitive noun, DEM – the demonstrative pronoun tas, ta. 122 Paprastųjų ir įvardžiuotinių būdvardžių opozicija lietuvių kalboje … percent of all phrases, while phrases with the indefinite form make up more than half (50.9 percent) of all phrases. Thus, according to the frequency criterion, the definite form is marked. The analysis also sought to determine how often definite adjectives are used to express different types of definiteness. It proved not so easy to do this, and in some cases it was even impossible to assign a noun phrase unequivocally to any one type. Kari Fraurud (1990) also reports encountering theoretical and methodological difficulties when she tried to apply Hawkins’ classification to noun phrases in authentic nonfiction texts. I will illustrate these difficulties with a few examples. Let us suppose that a typical case of anaphoric definiteness occurs when an indefinite article is used at the first mention of an object and a definite article at a subsequent mention. In article languages, however, there are many cases in which a referent is introduced into the discourse using a definite noun phrase (Eng. first mention definites, e.g.: the king, the sun). The first mention is classified as situational definiteness, but if the same phrase is mentioned several times later in the text, should the subsequent mentions be classified as situational or anaphoric definiteness? Fraurud attempts to find a solution, arguing that an anaphoric noun phrase must be linked to an antecedent in order to be interpreted as such. This link exists when the antecedent is an indefinite phrase. But how should we interpret cases in which the antecedent is a proper noun or a phrase with a possessive pronoun, that is, inherently definite, as in examples (18) and (19): (18) – Er Ina Lipman din søster? – Min søster... – Den berømte sangerinde? DST 10 – Ina Lipman your sister? Buzia asks. – Sister... – That famous singer? LTA 11 (19) To brødre havde gravet dette beskyttelsesrum. Der gemte de deres lamme mor <...>. Den gamle kvinde var lam i benene, men hendes hænder kunne udrette hvad der skulle være. DST 119 Two brothers dug that shelter. They hid their paralyzed mother here, their paralyzed mother <...>. The old woman’s legs were paralyzed, but her hands could do any kind of work. LTA 111 I would argue that examples (18–19) can also be classified as cases of anaphoric definiteness, but this requires revising Fraurud’s definition of an anaphoric noun phrase 123 BIRUTĖ SPRAUNIENĖ … phrase definiteness. Examples (18–19) are problematic also because the adjective in the anaphoric noun phrase was not mentioned together with the antecedent. The choice of definite adjective forms presupposes that the addressee knows that Ina Lipman is a famous singer and that the brothers’ mother is an old woman. In example (18), the demonstrative pronoun ta allows the definite suffix on the adjective to be omitted without disrupting the text’s coreferential relations; in example (19), however, replacing the definite form with an indefinite one breaks the coreferential relation between their paralyzed mother and the old woman, or at least creates the ambiguity that a new referent is being discussed; cf.: (19’) Two brothers dug that shelter. They hid their paralyzed mother here <...>.?The old woman’s legs were paralyzed, but her hands could do any kind of work. Examples (20–20’) also illustrate that the coreferential relation between the antecedent and the anaphoric phrase may break down if the definite form is replaced by an indefinite one without adding a demonstrative pronoun; cf.: (20) On August 12, 1955, the German writer and Nobel Prize laureate Thomas Mann died. Fate had brought the famous writer to our country as well. http://www.ziniur.lt/apie/pranesimai?item=506 (20’) On August 12, 1955, the German writer and Nobel Prize laureate Thomas Mann died.?Famous writer / This famous writer fate had brought to our country as well. Thus, in examples (19–20), the definite form of the adjective is obligatory. Because of the possibility of double interpretations (it is particularly difficult to draw a boundary between situational and deictic definiteness), it is not useful to provide statistics on how the phrases examined are distributed across definiteness types. Nevertheless, the research material makes it possible to state that the definite form of the adjective is used to express anaphoric and situational definiteness; however, as already mentioned, it competes with the indefinite form, and in cases of anaphoric definiteness, with demonstrative pronouns as well. 124 Paprastųjų ir įvardžiuotinių būdvardžių opozicija lietuvių kalboje 8.1. Anaphoric definiteness (21) He left us the windmill, a few horses, a young widow and a baby. <…> The young widow married again, moved somewhere far away, and left us the baby. DST 7 He left us a mill, a couple of little horses, a young widow and a baby. <…> the young widow married and went away somewhere far off, and gave the baby to us. LTA 8 In example (21), we have a classic case of anaphoric definiteness. Here, it seems, we could not replace the definite form with an indefinite one (cf.:?a young widow married and went away somewhere far off). Perhaps the topical status of the noun phrase prevents this. On the other hand, the data included quite a few cases in which an anaphoric phrase in Lithuanian is used with an indefinite adjective form, e.g.: (22) Next he selected a white pawn and a black one <…> The white pawn slipped from his palm. DST 5 Only then did he take one white pawn and one black one <…> From his handful slipped a white pawn. LTA 6 In article languages, the definite form in example (22) is obligatory. In the Lithuanian text, we could replace the indefinite adjective with a definite one (cf., From his handful slipped the white pawn), but the indefinite adjective form is sufficient to ensure textual coreference. Perhaps the indefinite form was chosen by the author because the adjective phrase is rematic in this case. The great distance between the anaphor and its antecedent (18 sentences) may also have determined the choice of the indefinite form. There are differing views in the literature on the maximum distance that can separate noun phrases linked by an anaphoric relation. Some authors argue that it cannot exceed 1–3 sentences, but Chafe gives an extreme case in which as many as 105 pages separate the anaphoric phrase from its antecedent. Chafe emphasizes the importance of the nature of the context for identifying the referent: In fact, it would appear that context or scene is all-important, and that definiteness can be preserved indefinitely if the eventual context in which the referent is reintroduced is narrow enough to make the referent identifiable (Chafe 1976: 41; išskirta mano — B. S.). 125 BIRUTĖ SPRAUNIENĖ The material examined also included examples in which the distance between phrases linked by anaphoric relations was several pages. In example (23), the distance between the antecedent and the anaphoric phrase is 3 pages. In these pages, one of the characters in the work, Janek, tells his friend Izaokas about what happened on the day when all the Jews were herded into the ghetto. Before beginning his story, Janek kicks a round, worn-down pebble. The pebble’s mention 3 pages later seems to frame the account of past events (English flash back) and mark the transition from one time frame to another. In the Danish (and English) translation, the definite form is necessary to ensure the anaphoric relation; cf. Danish den lille runde sten (ART.DEF small.DEF round.DEF stone, “the small round stone”), whereas Lithuanian uses the indefinite adjective form (of the round pebble is no longer there). (23) And just as he stops, he kicks a stone. It was a round, worn-down stone. We could really pretend it was a ball and start playing football with it. <…> We walk on, Janek and I. The small round stone is no longer in front of us. And that is good. It was no ball, after all, and we are not football players. DST 28–31 And, stopping, he kicks a pebble.That pebble is round, worn smooth. We could really imagine it was a ball and play football. <…> We walk on. Janek and I. The round pebble is gone. And that is good. It is no ball at all, and we are not football players. LTA 27–30 In example (24), we see an anaphoric chain in which the demonstrative pronoun that and the indefinite adjective form are used on the second, third, and fourth mentions of the referent. On the fifth and sixth mentions of the same referent, the pronoun that is omitted; the noun alone with the indefinite adjective form is sufficient to ensure textual coreference. Danish consistently uses the definite form. The seventh mention of the referent in the Lithuanian text (I need to bury a sleeping man) is ambiguous, since the phrase a sleeping man can also have a generic interpretation. In the Danish translation, this interpretation is ruled out: the phrase den sovende mand (ART.DEF sleeping man) refers only to that particular man who has already been mentioned. (24) A bearded man leans close to my face and winks. <…> A bearded man leans close to my face and winks. <…> That 126 Paprastųjų ir įvardžiuotinių būdvardžių opozicija lietuvių kalboje The bearded man has overpowered me. <…> I get up and see the bearded man who winked at me. <…> He is dead, the bearded man. <…> I spread out my jacket and pull the bearded man onto it. <…> We crawl toward the trench. I lower the bearded man into it. <…> Now he looks as if he is sleeping. I am afraid. I have to bury the sleeping man. DST 112–114 The bearded man bewitched me. <…> I get up and see beside me that bearded man, who winked <…> He is dead, that bearded man. <…> I spread out my jacket and pull over it the bearded man. <…> We crawl over to that trench. I lower into it the bearded man. <…> Now it seems to me that he is asleep. I am afraid. I need to bury the sleeping man. LTA 105–106 Valeckienė claims that “in contemporary standard Lithuanian, the use of demonstrative adjectives in the pronominal form has become the norm” (Valeckienė 1959: 191). The material examined shows that anaphoric definiteness in Lithuanian is not consistently expressed by definite adjectives. 8.2. Situational definiteness (25) Two guards are almost always stationed at the large gate. DST 14 Two guards are always stationed at the large gate. LTA 15 The Great Gate is mentioned for the first time in the novel, yet the definite form is used both in the article-language version and in Lithuanian. Since the action of the novel takes place in the Vilnius Ghetto, it is assumed that the addressee knows that the reference is to the great gate of the ghetto, of which there is only one. It seems that the definite adjective form in example (25) could not be replaced by an indefinite one, since a phrase with the indefinite form may also refer to a non-unique referent or have a generic interpretation, cf.: (25’) ?At large gates two guards are always stationed. In example (26), situational definiteness is likewise expressed in Lithuanian by a definite adjective form: 127 BIRUTĖ SPRAUNIENĖ (26) <…> whether she had managed to tell Gita about the new German machine gun they had buried right by the riverbank. They had wrapped a cloth around the firing mechanism before hiding it in a hollow in the old linden tree. DST 105 <…> whether she had managed to tell Gita about that new German machine gun which they had buried on the riverbank and whose bolt they had wrapped in a rag and hidden in the hollow of the old linden tree. LTA 98 From a certain pragmatic set (the linden trees known to the characters in the novel), one unique tree is singled out, bearing the property indicated by the adjective. On the other hand, the novel contains many examples in which the indefinite form of the adjective in Lithuanian, being unmarked, can very well be used to express situational definiteness: (27) I wanted there to be fewer people around me with those yellow stars on their clothes. DST 6 (I wanted there to be fewer people around me marked by those yellow stars. S 3) I wanted there not to be so many people around me with yellow stars. LTA 8 (28) – Janek, did people wear yellow stars back then too? DST 29 (ang. “Janek, did people wear the yellow patches then?” S 26) – Janek, were there already yellow patches? LTA 29 (29) – I can still see the narrow street before me, says Janek. DST 29 – I can still see a narrow street, says Janek. LTA 28 (30) I have been walking for a long time and grew tired long ago of the yellow moon, which twists and turns like a pancake <…>. DST 85 I have been walking for a long time, and I have long since grown tired of the yellow moon, which turns like a pancake <…>. LTA 79 Example (27) mentions the yellow stars with which the Nazis marked the clothing of Jews living in the ghetto. Lithuanian uses the indefinite form of the adjective (with yellow stars), whereas translations into article languages use a definite nominal phrase with demonstrative pronouns12 12 Since definiteness is marked here by demonstrative pronouns, this example 128 Paprastųjų ir įvardžiuotinių būdvardžių opozicija lietuvių kalboje (Danish disse gule stjerner, English those yellow stars). In example (28), the same yellow stars are mentioned (yellow patches), but in this case definiteness is retained in the English translation (English the yellow patches), whereas an indefinite noun phrase is chosen in Danish (Danish gule stjerner). Of course, the possibility cannot be ruled out that the translators made mistakes in some cases by interpreting certain Lithuanian phrases with non-definite adjectives as definite. On the other hand, there are too many such examples for all of them to be considered translation errors. Such seemingly unmotivated selection of definite descriptions has a certain stylistic effect, since it creates the impression that the addressee is, as it were, a participant in the story who sees all the events together with the narrator and can therefore identify the referents associated with them; the addressee is “local,” familiar with all the little streets, squares, streams, meadows, and so on. All this increases the vividness of the narrative, as the reader identifies with the characters and the narrator. 8.3. Objectual definiteness This type of definiteness occurs mainly in dialogues, when the participants in a conversation refer to objects in their immediate surroundings, e.g.: hand me the long knife! Such cases are harder to find in literary texts; I have assigned to this type examples containing the deictic adverbs there / here / look, but this classification is provisional, and examples (31–33) could also be interpreted as cases of situational definiteness: (31) Here is the narrow street. DST 20 Here is the narrow little street. LTA 21 (32) Does it matter now, after we have lived together for a whole month, thirty days, here by the bank of the shallow stream DST 106 Does it matter now, when we have both spent a whole month, thirty days, here, of the rushing stream bank LTA 100 (33) Meanwhile, three people had been hanged in the town, where there was no ghetto fence, in the large square DST 121 At that time, in the town, where there was no ghetto fence, in the large square, three people were hanging LTA 113 could be assigned to objectual definiteness. On the other hand, no clear boundary can be drawn between these types of definiteness; see below. 129 Birutė Spraunienė In examples (31–33), the indefinite adjective form is used, but it could be replaced by the definite form. 8.4. Inferential definiteness Only one example was found in the material examined in which inferential definiteness is expressed by the definite form: (34) Han [Chaim] blev taget levende og kørt til fængselshospitalet. <…> De bedste læger søgte at sikre Chaims helbredelse DST 70–71They took him [Chaim] alive and took him to the prison hospital. <…> The best doctors tried to ensure that Chaim recovered LTA 66 In the other examples assigned to this type, the indefinite form is used; the definite form would sound strange in them. Compare: (35) Han kunne tage ud til markerne, til den store, vidstrakte eng. Han kunne sidde i det bløde græs på engen DST 31 could go out to the fields, to the large, wide meadow. He would sit there in the meadow on soft /?the soft grass LTA 31 (36) Han er ikke ung, bare en soldat. Hans hænder er store <…>. Tyskeren stirrer længe på Esther, og jeg stirrer på den tunge hånd og det falmede, grønne uniformsærme som ligger på min skulder DST, 92–93 He is no longer young, that soldier. His hands are large <…> The German stares at Esther for a long time, while I stare at the heavy /?the heavy hand and the green /?the green, faded /?faded military sleeve, lying on my shoulder LTA 87 (37) Han tog om pistolen ved løbet, løftede den højt og knaldede det stumpe skæfte mod min tommelfinger DST 82 He took the Browning by the barrel, raised it, and struck the blunt /?the blunt handle against my thumb LTA 77 It is worth noting that in many Lithuanian examples the indefinite form cannot be replaced by the definite form, even though the noun phrase clearly provides a definite description. This shows that the marking of definiteness on the adjective form in Lithuanian follows somewhat different principles from the marking of definiteness by articles in article languages. The noun phrases in examples (35–37) are definite, but the adjective provides new 130 Paprastųjų ir įvardžiuotinių būdvardžių opozicija lietuvių kalboje informaciją, jam būdingas predikatyvumas (plg. Larsen 2007: 40). (35) pavyzdį galėtume pakeisti taip: Jis sėdėtų toje pievoje ant žolės, kuri yra minkšta. Daiktavardis žolė yra apibrėžtas dėl sąsajos su pieva, tačiau minkštumas nėra identifikacinis požymis, padedantis nustatyti, apie kokią žolę kalbama. Todėl lietuvių kalboje neteiktina būdvardžio apibrėžiamoji forma. Panagrinėkime dar vieną pavyzdį, kuriame būdvardžio predikatyvumas matyti dar aiškiau, nes būdvardis pateiktas kaip apozicija: (38) Og han tænkte at dette var den første gang han havde set Basias ben så tæt på, mens de fjedrende bevægede sig for hvert skridt, set hendes lange, smukke hals, hendes katteøjne som glim­te­de så forunderligt og hendes dybt røde læber så nær ved. DST 64 it seemed to him that for the first time he was seeing Basia’s legs from so close up, springing with each step, her neck, long and warm, her catlike eyes, shining strangely, and her red lips, all red. LTA 61 The phrase Basios kaklą ‘Basia’s neck’ can be classified as inferential definiteness; cf. Larsen’s claim Modification by possessive genitives and possessive adjectives can also be seen as an expression of semantic inference (Larsen 2006: 363). It is clear that the adjectives ilgą ir šiltą ‘long and warm’ provide a new characterization of the referent and do not help identify it. In article languages, attributive adjectives in definite noun phrases necessarily fall within the scope of the definite determiner, are used in the definite form (with the ending -e in Danish), and are understood as presupposed (cf. hendes lang-e, smukk-e hals ‘her long-DEF, beautiful-DEF neck’, ‘her long, beautiful neck’ in example (38)). This is also the case in Latvian, incidentally; see Holvoet, Tamulionienė 2006: 23–24. On the other hand, we cannot conclude that definite adjectives are not used at all in Lithuanian in cases of inferential definiteness, since there are examples of this, for instance: (39) Wash the avocado seed. Immerse it, blunt end down, in a glass filled with water. http://gyvenimas.delfi.lt/archive/article.php?id=6446869 In example (39), bluntness is an identifying feature that helps distinguish the referent (the end of the seed relevant to the speaker) from the corresponding pragmatic set, thu- 131 BIRUTĖ SPRAUNIENĖ s the definite form of the adjective is used. The use of a definite adjective presupposes that every avocado seed has a blunt end, and that it is the only such end. 8.5. Proper names Proper names are inherently definite and in article languages are most often used without a definiteness marker (cf. English *the Mary). In Lithuanian, attributive adjectives used with proper names may be either definite or indefinite, as examples (40–41) show: (40) Today Rudi is bothering me. His name is something else, but everyone calls him Rudi. He is red all over. His hair is like fire, his face looks like a red flag <...>. My angry gaze pierces red Rudi. DST 68–76 Today Rudi is bothering me. His name is something else, but everyone calls him Rudi. He really is red all over. His hair is like fire, his face like an ember <...>. I stab the red Rudi with angry eyes. LTA 72 (41) Jeg går uden at se mig tilbage, for at slippe for at se bjælken, hvor Esther nu sidder alene tilbage. DST 146 ‘I walk away without looking back so I won’t see the log where Esther is left sitting alone.’ I walk away without looking back, so I won’t see the log and lonely Esther. LTA 135 In example (40), redness is an inseparable feature of the referent (He is red all over). Therefore, both Danish and Lithuanian use the definite form (røde Rudi / raudonąjį Rudį), and there is no alternative. In example (41), being lonely is not a permanent, inseparable feature of the referent. Rather, it is a contingent feature—at the moment when Isaac leaves, Esther is left alone. Therefore, the definite form of the adjective is inappropriate in example (41) (cf. *vienišosios Esteros), and the Danish translator made the entirely correct choice by moving the adjective into predicative position. Erik Hansen also discusses interesting cases of proper names used with adjectival modifiers in Danish, where the adjective may be either definite or indefinite (Hansen 2001: 199 ff.; examples taken from Hansen): 132 Paprastųjų ir įvardžiuotinių būdvardžių opozicija lietuvių kalboje (42a) en förkölet Lassen (ART.INDEF ‘Lassen, who has caught a cold’) (42b) den förkölede Lassen (ART.DEF ‘Lassen, who has caught a cold’) (43a) *en rødhåret Simonsen (ART.INDEF ‘red-haired Simonsen’) (43b) den rødhårede Simonsen (ART.DEF ‘red-haired Simonsen’) According to Hansen, in examples (42–43), proper names retain their definiteness, i.e. in this case we cannot speak of an appellativized proper-name lemma. Hansen explains their use with articles and the corresponding forms of adjectival modifiers in terms of the information-structural properties of noun phrases: the definite article (and the definite form of the adjective, indicated by the ending -e) signals that the attribute is known information, i.e. that the addressee already knows that Lassen has caught a cold (example 42b). When a noun’s attribute provides new information, describing the individual in a new way, the indefinite article (followed by the indefinite form of the adjective) is used (example 42a). Examples (42–43) illustrate a certain (pragmatic) restriction: an adjective with the indefinite article can be used only when referring to a temporary (accidental) property (42a), whereas the definite form of the adjective can denote either a temporary or a permanent (characteristic) property of the individual (42b and 43b). The adjective rødhåret (Lith. ‘red-haired’) could be used with the indefinite article as an attribute of a proper name if red hair were a new, accidentally acquired, transient property of the individual. If the property in question is characteristic and permanent, the definite form is preferred. According to Hansen, the use of proper names with indefinite articles and indefinite-form adjectives in Danish is an innovation. It is a stylistically marked, new construction, found mainly in the media (Hansen 2001: 204). Hansen argues that similar phenomena are also observed in attributive constructions with common nouns, in which the definiteness of the noun is clear from the context13 (Hansen 2001: 201, examples taken from the author): (44a) hund-en stak snud-en frem dog-DEF stuck muzzle-DEF forward ‘the dog stuck its muzzle out’ (the noun ‘muzzle’ has inferential definiteness) (44b) hund-en stak den våd-e snude frem 13 All the examples with common nouns discussed by Hansen can be classified as inferential definiteness, although the author does not use this term. 133 Birutė Spraunienė dog-DEF put ART.DEF wet-DEF muzzle forward “the dog stuck out its wet muzzle” (N is inferentially definite, A presents old (known) information) (44c) hund-en stak en våd snude frem dog-DEF put ART.INDEF wet muzzle forward “the dog stuck out its wet muzzle” (N is inferentially definite, A presents new information) Nevertheless, constructions in which the form of attributive adjectives depends on whether they convey new or old information are rare, and attributive adjectives in Danish are generally simply matched to a definite or indefinite noun phrase. 9. THE STATUS OF MEMBERS OF THE CATEGORY OF DEFINITENESS IN TERMS OF MARKEDNESS As mentioned above, the unmarked member of the opposition between indefinite and definite adjectives in Lithuanian is the indefinite form. Indefinite forms are used more often than definite forms: in the material examined, phrases with an indefinite form make up 76 percent, while phrases with a definite form make up only 24 percent of all phrases. Moreover, only semantically definite noun phrases were selected for the study; if indefinite phrases were included, the percentage of indefinite adjectives would be even higher. The distribution of definite and indefinite forms could be represented by the following diagram: Diagram 2. Distribution of definite and indefinite adjective forms Definiteness Indefiniteness Definite adjectives Indefinite adjectives The distribution of indefinite adjectives is considerably wider than that of definite adjectives, and, being unmarked, indefinite forms can replace definite forms in many contexts, but not vice versa. Interestingly, in article languages the relationship between members of the category of definiteness in terms of markedness is the opposite—the definite form is unmarked, 134 Paprastųjų ir įvardžiuotinių būdvardžių opozicija lietuvių kalboje o neapibrėžiamoji — žymėta (Van Langendonck 1979: 33 ir 2007: 172). Pagrindinis Van Langendoncko argumentas yra tas, kad neapibrėžtumo semantikai būdingas egzistencinis kvantifikatorius, kuris apibrėžtumui nebūdingas: „definiteness can be rendered by a free variable and indefiniteness by an existential quantifier binding variable“ (Van Langendonck 1979: 41). Apibrėžto daiktavardinio junginio vartojimas reiškia, kad referento egzistavimas diskurso erdvėje yra presuponuotas. Neapibrėžtas daiktavardinis junginys įveda naują referentą į diskurso erdvę. Referento egzistavimas šiuo atveju yra ne presuponuojamas, o teigiamas („not presupposed, but asserted“) (ibid., 38). Anot Van Langendoncko, „apibrėžtumas yra natūraliausia referentinio junginio būsena“14, t. y., yra didelė tikimybė, kad referentinis junginys bus apibrėžtas. Remdamasis Greenbergo suformuluotais žymėtumo kriterijais (raiškos sudėtingumas (ang. surface complexity), dominavimas (ang. dominance), įvairovė (ang. diversification) ir dažnumas (ang. frequency), autorius įrodinėja, kad neapibrėžtumas yra semantiniu ir sintaksiniu požiūriu sudėtingesnis nei apibrėžtumas. Vienas iš įrodymų nekelia abejonių — apibrėžti daiktavardiniai junginiai yra vartojami žymiai dažniau nei neapibrėžti (Van Langendonck 1979: 49). The Danish linguist Eva Skafte Jensen (Jensen 2007) distinguishes two oppositions in the definiteness paradigm of modern Danish: 1. the article-bearing noun form — unmarked, the articleless (bare) form — marked; 2. the form with a definite article — unmarked, the form with an indefinite article — marked. Jensen also shows that the markedness status of noun phrases changes depending on the development of the category of definiteness and the degree to which definiteness markers have grammaticalized. When the definite article emerged in Old Danish, the articleless (bare) form was unmarked in relation to the article-bearing form. The newly emerged definite article underwent a long process of grammaticalization, during which it spread into new contexts of use, and ultimately, in modern Danish, compared with Old Danish, an inversion of the markedness of indefinite and definite forms took place (English markedness reversal) — the articleless form became marked, while the form with an article became unmarked. 14 „Definiteness is the most natural state of a referring expression“ (Van Langendonck 1979: 41). 135 BIRUTĖ SPRAUNIENĖ 10. CONCLUSIONS The article aimed to reveal the semantic contrast between indefinite and definite adjective forms in Lithuanian, to investigate the distribution of definite forms, and to examine the connection between these forms and the category of definiteness. The article argues that definite forms provide a noun phrase with definiteness, although in some cases it is expressed at another — taxonomic — level (in noun phrases with a generic meaning). Indefinite forms are indifferent or neutral with respect to the marking of definiteness. A phrase with an indefinite adjective may be either indefinite or definite; one interpretation or another depends on the context. Thus, the definite adjective form is the marked member of the opposition, while the indefinite form is the unmarked member. This is also confirmed by the distribution of these forms: indefinite forms are used much more frequently than definite forms with semantically definite noun phrases. There are few contexts in which the definite form is obligatory: with vocatives when there is no nominal form of address, when the definite adjective is used substantivally (although exceptions can again be found), and in certain cases of anaphoric definiteness in which the antecedent and the anaphoric noun are synonyms and the adjective was not mentioned together with the antecedent (see Section 8); these contexts have not been fully investigated. The analysis showed that in some cases the definite adjective form is almost invariably not used in Lithuanian, even though the noun phrase modified by the adjective provides a definite description. Exactly what limits the possibilities of using the definite form in such cases is not entirely clear, but it may be related to properties of the information structure of the noun phrase. It seems that an adjective can be used with a definite postfix if it denotes a property of the referent that is known from the context or presupposed and enables the referent to be identified unambiguously (distinguished from a particular pragmatic set). If the adjective has a predicative character, i.e., if it provides new information and newly characterizes the referent, the indefinite adjective form is generally used, despite the fact that the noun phrase as a whole is semantically definite. 136 Paprastųjų ir įvardžiuotinių būdvardžių opozicija lietuvių kalboje ŠALTINIAI DST — Icchokas Meras, Det sidste træk. Oversat fra engelsk efter Stalemate af Kirsten Vagn Jensen15. København: Vindrose, 1981. HK — Herkus Kunčius, Pilnaties linksm ybės: eseistinė proza. Vilnius: Rašytojų sąjungos leidykla, 1999. KT — Dabartinės lietuvių kalbos tekstynas, http://donelaitis.vdu.lt/ LTA — Icchokas Meras, Trys romanai. Lygiosios trunka akimirką, romanas; Ant ko laikosi pasaulis: romanas-baladė; Ties gatvės žibintu, vaikiškas romanas. 3-asis papild. leidimas. Vilnius: Lietuvos rašytojų sąjungos leidykla, 1998. S — Icchokas Meras, Stalemate. Translated by Jonas Zdanys. Revised edition. New York: Other Press LLC, 2005. LITERATŪRA Ambrazas, V., ed., 1997: Dabartinės lietuvių kalbos gramatika. Vilnius: Mokslo ir enciklopedijų leidybos institutas. Chafe, W. L. 1976: Givenness, contrastiveness, definiteness, subjects, topics, and point of view. In: C. N. Li, ed., Subject and Topic. New York: Academic Press, 25–55. Fraurud, K. 1990: Definiteness and the processing of noun phrases in natural discourse. Journal of Semantics 7, 395–433. Girdenienė, D. 1971: Aktualioji skaida ir jos vaidmuo bei reiškimo būdai lietuvių kalboje. Kalbotyra 22.1, 35–47. Givón, T. 1995: Functionalism and grammar. Amsterdam–Philadelphia: John Benjamins. Hansen, E. & L. Heltoft, spausdinama: Grammatik over det Danske Sprog. Hansen, E. 2001: Overbestemt og ubestemt proprium. In: H. G. Jacobsen & H. Jørgen- 15 Į danų k. iš anglų k. pagal Stalemate išvertė Kirsten Vagn Jensen. 137 BIRUTĖ SPRAUNIENĖ SEN, eds., Glæden ved Grammatik. Udvalgte artikler og afhandlinger. København: Hans Reitzels Forlag, 199–205. HASPELMATH, M. 1999: Explaining article-possessor complementarity: economic motivation in noun phrase syntax. Language 75.2, 227–243. HASPELMATH, M. 2006: Against markedness (and what to replace it with). Journal of Linguistics 42, 25–70. HAWKINS, J. A. 1978: Definiteness and Indefiniteness: A Study in Reference and Grammaticality prediction. London: Croom Helm. HAWKINS, J. A. 1991: On (in)definite articles: implicatures and (un)grammaticality prediction. Journal of Linguistics 27, 405–442. HAWKINS, J. A. 2004: Efficiency and Complexity in Grammars. Oxford: Oxford University Press. HOLVOET, A. & A. TAMULIONIENĖ, 2006: Apibrėžtumo kategorija. In: HOLVOET & MIKULSKAS, red., 2006, 11–32. HOLVOET, A. & R. MIKULSKAS, red., 2006: Daiktavardinio junginio tyrimai. Vilnius: Lietuvių kalbos institutas. JENSEN, E. S. 2007: Der var en gang en mand – om markeret og umarkeret artikelbrug i moderne dansk og gammel dansk. In: H. JØRGENSEN & P. WIDELL, eds., Det bedre argument. Festskrift til Ole Togeby 7. marts 2007. Århus: Wessel og Huitfeldt, 299–320. LABUTIS, V. 1998: Lietuvių kalbos sintaksė. Vilnius: Vilniaus universiteto leidykla. LARSEN, K. 2005: The Evolution of the System of Long and Short Adjectives in Old Russian. München: Verlag Otto Sagner. LARSEN, K. 2006: The distribution of long and short forms of adjectives in Old Russian (as represented by four chronicle texts). Russian Linguistics 30.3, 359–400. LARSEN, K. 2007: The loss of the attributive short form in Old Russian as a process of grammaticalization. Scando-Slavica 53, 33–57. 138 Paprastųjų ir įvardžiuotinių būdvardžių opozicija lietuvių kalboje Lyons, C. 1999: Definiteness. Cambridge: Cambridge University Press. Mik ulskas, R. 2006: Apibrėžiamųjų būdvardžių aprašo perspektyva. In: Holvoet & Mikulskas, red., 2006, 33–66. Paulauskienė, A. 1994: Lietuvių kalbos morfologija: paskaitos lituanistams. Vilnius: Mokslo ir enciklopedijų leidykla. Rosinas, A. 1972: Žymimasis artikelis lietuvių kalbos tarmėse ir senuosiuose raštuose. Kalbotyra 24.1, 83–89. Rosinas, A. 1996: Lietuvių bendrinės kalbos įvardžiai. Vilnius: Mokslo ir enciklopedijų leidykla. Ul vydas, K. red., 1965: Lietuvių kalbos gramatika 1. Fonetika ir morfologija. Vilnius: Mintis. Valeckienė, A. 1957: Dabartinės lietuvių kalbos įvardžiuotinių būdvardžių vartojimas. Literatūra ir kalba 2, 159–308. Valeckienė, A. 1986: Apibrėžtumo / neapibrėžtumo kategorija ir pirminė įvardžiuotinių būdvardžių reikšmė. Lietuvių kalbotyros klausimai 25, 168–189. Valeckienė, A. 1998: Funkcinė lietuvių kalbos gramatika. Vilnius: Mokslo ir enciklopedijų leidybos institutas. Van Langendonck, W. 1979: Definiteness as an unmarked category. Linguistische Berichte 63, 33–55. Van Langendonck, W. 2007: Theory and Typology of Proper Names, Berlin–New York: Mouton de Gruyter. Birutė Spraunienė Vilniaus universitetas, Lietuvių kalbos katedra Universiteto 5, LT-01513 Vilnius, Lietuva [email protected] 139