Diatezės alternacijos iš skirtingų teorinių modelių tyrimų perspektyvos

Acta Linguistica Lithuanica LXI (2009), 61 -111 Perspectives on different i§ theoretical models of diathesis alternances KRISTINA LENARTAITE Lietuviy kalbos institutas The article gives overview an the of approaches offered different in theoretical frameworks (Generative Grammar, Relational Grammar, the Lexical Rule Models, Construction Grammar and the Lexical-constructional Approach) to account for alternations argument in realisation (diatethetic alternations) such active passive as the alternation, vs the ‘spray/load’ type alternations, dative alternations etc. The final part the article presents of certain considerations favour the derivational view. speaking in of 0. PRELIMINARY REMARKS The article discusses You diatezés alternavimo! expressions: active subconstructiir ons, applicative syvinés alternations and other cases of argument alternation (e.g., locative alternances, etc.). ir So, how- | "Diathesis" savoka XX a. the nineteenth pradZioje decade was }vesta CholodoviGiaus, in order to universaliat apra’yti ries category. The term is apie used to describe a change in the nu semantic function of the subject, for example, an active action, where the subject is su an agent, a passive or o su — patientas ‘patyréjas’ (downgoing) (Matthews 2007: 104-105). However, diateze it can be understood more ir broadly as — a grammatical function (part of the sentence) distribution (semantiniy roly atrovyu) model, horn only o as one — grammatiy means, letZandiu change diateze (Holvoet, Seméniené 2004: 35-36). Yes (broader) diathesis is understood ir In this article. Syntax alterations in linguistic literature can be very different predicative vadinami The word valence is changed by processes realized by su morphological derivations of expressions Zymikliu or him. be All these processes can be further classified according to their functions: valence-increasing processes (e.g., causative constructions in j jy relation to non-causative ones), valence-decreasing processes (e.g., passive constructions are usually intransitive, while active structures are transitive); Alternatives can also be divided into interchangeable (e.g., active passive alternatives, applicative alternatives) and interchangeable (e.g., causative constructions) (Kroeger 2005: one case in question includes two more constructions, whereas the other — case involves two more constructions). j / ir 270-287). 61 KRISTINA LENARTAITE ar su ¢iu actionazodziu (with a significant morphological derivation marker) ar be his) alternatively realized two even several different ar diathesis, which in terms of action are to be considered as alternating. to Visus The theoretical models used to explain diathesis discussed in this article can be divided into two groups: ma j (i) design models (¢only discussed here) du i8 jy — Goldberg construction of grammar ir Iwata's lexical-constructive models) ir (ii) other models, such as generative grammar, relational grammar, and the ir so-called,,lexical rules” models (i8 holding The position of lexical-functional grammar on the question of alternance of diathesis, Levin ir Rappaport Hovav bei "Pinkerton". IS first A glimpse galéty appear that These model groups are distinguished by the recognition of the ar derivative relationship connecting the alternating constructions (when one construction (or its lexical-semantic representation) is considered to be basic, other o (ar other) — i§ its i8veds), no derivation of ar one construction of another is recognized i8 positive corresponding constructions ir arising independently of each other (this position is observed in the design models). Therefore, as I will explain later, only the ne derivative $i skirnon-derivate — is the / essential — criterion distinguishing the constructive models from the other We'll talk about it. Be to, Considering j the investigated i8 needs konstrukciniy modely position,,,alternacijy“ term, although This model representative traditionally ir used, seems misleading because — neither syntactic nor semantic level In these models there is no derivative connection between the structures*. Thus it is obvious, that by looking at the investigated j phenomena from the perspective of a constructive i model, I probably would not even have defined the investigated object in this way, thus, already the title of the article suggests my preference for non-constructive models. Thus, my theoretical preferences with regard to the subject under consideration and ir the choice of the relevant research direction I try to justify at the end of the article. > For example, in the Goldberg construction grammar, each (1995, 2006) such sentence is treated as a different generalization of the construction's specific implementations, or instances. The appearance of several concrete realizations determines the ig significance of a specific action. ne ,,suiletar* (to integrate) su with different generalized constructions, in ir each case of realization it is important only how the specific need fits with the needs of the corresponding ta generalized construction, the question of what relationship connects two constructions that have arisen on the basis of o the meaning of one action is not relevant to the Siam model. ar 62 Perspectives on different i8 theoretical models of diathesis alternances 1. TIRIAMUJY OF EVENTS DECLARATION The reference path of this study has become various cases of argumenty speech alternance, such as locality alternances, otherwise called ,,purk&ti / krauti* tipo alternacijomis spray load alternation),,k- ( nibzdéti* tipo / alternacijos (warm alternation) ir kt*. Attempting to 8i define t. y. rei8kinj, jo identify functions and find a jo place in the language system prompted the convergence ji su active passive applicative / alternations. ir The aforementioned different cases of argumenty ir alternance and active passive alternance exist in one form or another in many languages of the / world, applicative constructions (applicatives) ar are common to non-Indo-European languages 0 (it has many In African languages, some Indonesia ir etc. language). All the phenomena discussed, the su morphological derivations of the verb Symbols (in the case of ir passivation applicative construction) or their, be allowing to redistribute the grammatical ir functions of the subject's direct object without changing neither the predicate argumenty number, nor the semantic jy roles. Be therefore, it is often asserted that the alternative constructions in question have the same propositional meaning (Mel’tuk 1993: 11), because they convey the same ta conceptual event ar ta the same background information (Foley 2007: 362-370). This does not mean in any way that alternating sentences, for example, active and passive respectively, are absolutely synonymous, they first differ very markedly in their ir unsymmetry of actuality, for some of the discussed sentences there are other markedly — necessary differences, for example, in the case of localising alternation, the predicate of one sentence often has an additional holistic meaning, meaning that the argument o expressed by the direct object is the i§ whole of the action. ir i§ Furthermore, each investigated rei3kiniy is discussed separately. ®In this article ,Alternations of argumentative speech’ or simply palternations* I call only those things that meet aS the criteria discussed in chapter i8, although in general linguistic literature other things can be 11. called so ir (plg. 1 i8na8q). For example, one Winter English language arguments raiSkos alternacijy tyrinétojy Levin understands argumenty alternavima as all possible variants of one action: tia patenka argumenty poziy ir (subject, direct ir indirect object), ir circumstantial termy positiony variation cases of expression. Be to, in English in terms of form the same verb (t. y. be jokey 1required morphological derivations Zymikly) may realize such argumenty ir configurations, which in Lithuanian language can be played only with the appropriate verb derivatives, e.g. : su a. Janet broke the crystal ‘Dzanet sudauzé kristala’ ir b. Crystal breaks at the slightest touch The crystal sparkles at the slightest touch' (Levin 1993: 26). 63 KRISTINA LENARTAITE 1.1. Alternating arguments The definition of this term and the criteria for its classification have been formulated primarily on the basis of Lithuanian cases (see Lenartaite 2007), but it is also valid for similar terms in other languages. And it will be operated mainly on Lithuanian examples, besides you, also considering the so-called,,- datyvinés“ alternances of the English language and the locative alternances of the German language. Since argumentative speech alternations are best studied in European languages (first of all, of course, in English), it may appear that this phenomenon is unique to Indo-European languages, but this is not the case: §j speech, for example, also exists in Japanese (see Iwata 2005, 2008). Thus, in this argument, only those cases are considered to be alternations of expression that meet all of the following criteria: (1) the alternating sentences have the same lexical substance, i.e. the same terms with the same semantic roles are included in the predicate; (2) bit predicate structure parameters for realization of alternative constructions; (3) the changes in the expression of arguments must affect at least one of the main grammatical functions — subject and/or direct object (i.e., at least one of these positions must be realized by different arguments); (4) the alternating constructions of the relationship can be perceived as paraphrases, and in terms of the correctness of the logical statement often even as equivalents. Various Cases of argument and expression alternation that meet these criteria can best be classified based on the syntactic functions of the subject and direct object, to which different terms can be assigned: (i) argument and expression alternations of the subject and direct object: (1) a. b. Mother is overwhelmed with children. (ii) alterations of subject expression: (2) a. b. The courtyard smelled of honeysuckle. (iii) alternance of the expression of the direct object: * Dél Sios savokos ir jos taikymo plg. Chafe (2003: 107-108). 64 Diatezes alternacyyos i8 differenty teoriniy modely researchy perspectives (a) and the direct object, and the place of action> Alternacyys of the circumstantial thermo expression of Zyminéio: (3) a. b. The girls (ap-)spread their chests with gems. (b) alternances of direct object and indirect (recipient/- beneficial) object expressions: (4) a. Mother (over-, under-)maukslina hat to the child. b. Mama apmaukslina vaiikq hat. (c) alternances of the circumstantial term of the place of action/influence and the direct object: (5) a. Suo (per-)bégo per keliq. b. His perbégo kelig. (6) a. A child throws a ball. b. Child (pa-)spyré ball. Now, based on concrete examples of alternatives, we can discuss each of the above criteria in more detail. In the first case, it should be emphasized that the requirement of the same lexical substance of alternating constructions applies primarily to the identity of the terms (arguments) and circumstantial terms realized in alternating sentences, and the predicate expression of such constructions may differ, t. y. Sometimes one alternating i§ construction may be formed only with the prepositional verb, and the other with the corresponding non-prepositional verb (see 5ab) or the prepositions of the alternating construction may differ (see 4ab). However, even with different parts of action, the constructs realized by the zodiacs are in all other aspects (in terms of the criteria applied to all alternations) equivalent to the constructs made with the same predicates in terms of form, cf.: (7) a. b. The President awarded the students with books. (8) a. Mother (su-)puts apples around Zqsj. b. Mom decorates Zqsj with apples. ’This summarizes the content of the circumstantial term performing various semantic functions, such as the locative, directive or goal. 65 KRISTINA LENARTAITE The lexical meaning of the verb determines that the structure of some arguments can be realized without changing the form of the verb (see sentence pairs (1), (2), (3), (6)), while other changes in semantic structure must be indicated (signalized) by the corresponding symbol, in Lithuanian language — the preposition (see sentences (5b), (7b) and (8b)). However, the processes occurring in all the above-mentioned alternating cases can be described and represented equally. Moreover, in many cases in one alternating construction the preposition is only desirable, but not necessary (see sentences (3ab)) —this is another argument for the fact that all these cases belong to the same —alternative preposition can be realized only with verbs having a preposition be-*, cf.: (9) a. The rioters sprayed paint on the car. Vandals in the middle of a machine. The rioters sprayed the car with paint. ,Vandalai apipurské magina daZais‘ (Examples i§ Iwata 2008: 153) The localivinés alternations of the German language can also be used as an additional argument for the fact that the Lithuanian language alternating constructions, realized both ju — reiSkiniui ir turéty biti traktuojami vienodai. Cia dar pravartu paminéti vokieciy kalbos lokatyvines alternacijas, kur vienas i§ alternuojantiy with the same verbs in terms of form and with the corresponding prepositional verbs, should be treated equally. Furthermore, the so-called applicative variant (sentence (9b)) in the localative alternance of German (see also 6 and 8naSa), as well as, for example, the applicative constructions in African Bantu and other languages (about them in more detail in section 1.3.), are used by many researchers as an argument, based on the alternating relationships of the derivative relations of the constructions (derivation can be recognized as a constructive level, or as a verb-based lexical-semantic representation level): it is based on the fact that the iSoriSkai are partly independent of the verb-based lexical-semantic representation level. ®Because such a process of the German contextual term offering the position of the direct object, consistently accompanied by unprefixation, is very similar to non-Indo-European applicative constructions, where analogous processes of term offering the position of the primary object of the sentence are also necessarily accompanied by the process of adaptation of the applicative preposition to the verb-based semantic level, the corresponding German construction is also called applicative (e.g., Michaelis & zymima, 1999). However, just as it is common for the vast majority of languages, there are also exceptional cases of German language alternances. For example, in very rare cases, German allows alternative sentences to be expressed in the non-dependent verb form, even if the dependent form is also possible (e.g., in the case of the verb streichen, spread, smear) (see Iwata 2008: 153-156 for more details). This is very similar to the case of Lithuanian language alternancy. Furthermore, in German one can also encounter the case where the partial verb of the applicative construction has no non-partial equivalent, for example, the use of the verb beampeln,jrechtstellen Sviesoforus‘ (for the phrase: »apssfororinti- *) in the applicative construction, it incorporates in its meaning as an object and the substantive, so long as it is possible to use it only in such a construction, where the position of the direct object (in the passive construction respectively — the subject) is given to a circumstantial term having a semantic role of location, cf. The crossroads Ruppenhofer 2001). 66 Diatezés alternacijos i§ skirtingu teoriniy modeliy tyrimy perspektyvos bet néra atitinkamo sakinio su veiksmazZodZiu *“ampeln, ir tai galima paai8kinti: kadangi veiksmaZodis yra iSvestas i8 daiktavardZio Ampel ,sviesoforas‘, Arsenalstrafe wird beampelt,- Arsenalo gatvés sankryZa bus apriipinta Sviesoforais (pZd.,,apSviesoforinta‘)‘. (Example i§ Michaelis & Ruppenhofer 2001: 19.) Derivative Models This example poses a difficulty: there is a problematic inconsistent derivation, i.e., the crossing of one step of the derivation. : Traffic light —*ampeln —beampeln. In conclusion of the discussion of the criteria for the lexical nature of the uniform alternating construction, it should be added that such cases, when one term cannot be realized in one construction, are not considered as alternations, cf. Mergaité nuSluosté crumbs from the table and Mergaité nuSluosté stalq (?? from trupiniy). These sentences are most quickly realized with differently meaningful verbs: in the first case reiSkian@- iu,(something) by sweeping Salinti ka (from what)‘, and in the second —,(something) by sweeping valyti (8varinti, išdžiovinti) ka‘. The second criterion —the changes in the predicate structure that are typical for alternating constructions —can be considered as redundant: because the changes in the predicate structure can be determined only from the changes in the structure of its arguments, which are ti nuo aplinkybiniy termy raiskos ivairavimo atvejy (plg. Meskos per Ziemq defined by the third criterion. However, this criterion is used to distinguish Meška Ziemq 67 KRISTINA LENARTAITE from sleeping (the predicate structure is not changed by changes in the environment) and from other panaphic alterations (cf. They atplé vaikq from mother and They atplé moting from child —The structure of the predicate also remains unchanged, because the predicate in both cases opens identical positions, but they are occupied by different participants in the situation). The third criterion —that only the cases of expression driving the subject and/or direct object are to be regarded as syntactic alterations—numbers the boundary between two purely structural syntactic positions —the subject and the direct object—and the syntactic position of the i8 part, the semantic position of the i8 part of the indirect object. The grammatical function of the indirect object correlates with the compact group of semantic roles (recipient- /beneficiary/experiencer/addressee), and there is no case where several different terms included in one action word claim to occupy the syntactic position of the beneficiakiti tolimesniy objekty arba aplinkybiy pozicijoje realizuoti aplinkybiniai ry object. The term realized in the object of the beneficiary, like terms, can only be elevated by the position of the direct object (sentences (4), (7)), but the elevation of the circumstantial term realized in the Lower circumstantial position by the position of the beneficiary's object, or even more so the demotion of the argument realized in the position of the subject or object by the position of the beneficiary's object, is not conceived. And quite rightly, the grammatical function of the indirect object, the subject and the direct object, although prototypically perceived as correlating with the semantic roles of the agent and the patient, respectively, allows any other circumstantial terms with semantic roles to be promoted to these positions. When discussing the role of the objective noun in Lithuanian language alternances, it is useful to consider the,,dative” alternances of English language, also called,,dative shift”* (dative shift), cf.: (10) a. John handed the scarf to Mary. ,Dzonas padavé Salika Merei‘ b. John handed Mary the scarf. »Dzonas padavé Merei Salika‘ (Examples from Shaumyan 2006: 236). The sentence (10b) is also called double object, or ditransition. Here, a kind of,,indirect object* is given a narrower meaning than in traditional Lithuanian grammar, and only objects of the beneficiary are in mind, having the semantic roles of recipient / beneficiary, etc. 68 Diatezes alternacuos i§ skirtingu teoriniy modeliy tyrimy perspektyvos These constructions play a decisive role in distinguishing recipient/beneficial objects from other circumstantial elements of the sentence (oblique) and in assigning to them the grammatical function of an indirect object. In other words, if not the transitive constructions, where the term having the semantic role of recipient/beneficiary is realized as the primary object of the sentence and thus loses its prepositional meaning (cf. Mary in the sentence (10b) and that Mary in the sentence (10a)), there is no longer any serious basis for the prepositional construction realized in the term in the sentence (10a) iSredi iS kity, also with the prepositions iSreiSkiamy circumstantial termy. Thus, in English it is easier to deduce the object — and circumstances than in Lithuanian: in English circumstances are considered to be everything that is realized by prepositional constructions. As already mentioned, Gia forms kalboje skirt) tarp pagrindiniy gramatiniy funkcijy — subjekto ir tiesiogionly prepositional constructions correlating with recipient/beneficial objects. Further speaking about the “dative” alternances of the English language, it is worth mentioning that they involve verbs, which by meaning are generally assigned to the semantic group of “giving” verbs®. If we try to compare, in terms of verb meaning, the English,,dative’’ alternances with the Lithuanian alternances, in which terms with recipient (see sentence (7)) or beneficiary (see sentence (4)) roles are involved, we will first see that,,Datyviniy* alternacijy’’ transitive constructions in English and other languages are most often realized with physical transfer-requiring verbs, such as give, grant, sell, give away, etc. Such verbs encode an event where an agent causes an object to pass into the possession of another living object (i.e. recipient, recipient, or possessor of the object). Benefactive situations (i.e., performed for someone’s benefit) are not easy to distinguish from tangible action words: from what is only discussed physical transfer of a good. The verb action word, for example, the English transitive construction She brought me a coffee She brought me a coffee to me can be paraphrased as a,dative’ construction She brought a coffee to me, and the benefactive construction She brought a coffee in many other languages benefactive situations can be realized in transitive constructions, and this often makes it difficult for recipient and benefactor to understand each other. for me ,Ji atne3é kavos man (del manes)'. Kaip ka tik matéme angly kalboje, tai pat ir (beneficiary) semantic role differentiation. However, unlike in the case of verbs that involve physical transfer, the semantic role of the benefactor can be realized even with intransitive verbs, e.g., She sang for me (dél manes)’ (Malchukov, Haspelmath & Comrie 2007). 69 KRISTINA LENARTAITE the phenomenon is defined as the promotion of the circumstantial term j the position of the direct object (or main object). This process in the structure of the verb is always marked by an applicative suffix. Let's look at a few Giéeva (chichewa) language (bantu language groupé) example": (19) a. I-na-tumiz-a letter to the king. 1SG. SUBJ-PRAET-SYYST-ASP laigkas virgininkui I sent a letter to the chief. I-na-tumiz-ir-a mfumu kalata. 1SG. SUBJ-PRAET-systi-- APPL-ASP __virsininkas _laiškas,I sent an officer a letter (20) a. Msangalatsi a-ku-yend-a ndi ndodo. entertainer(1) SUBJ(1)-PRAES-child-playing-AsP with a stick,Entertainer playing with a stick‘ b. Msangalatsi a-ku-yend-er-a ndodo. entertainer(1) SUBJ(1)-PRAES-children-APPL-ASP stick,e.g. Entertainer walking with a stick (Examples i§ Kroeger 2004: 64.) The circumstantial terms (kwa mfumu, ndi ndodo) realized by prepositional constructions in the sentences (19a) and (20a) acquire a different syntactic status in the sentences (19b) and (20b) —Gia they are already direct objects. Such a process of promotion of the circumstantial term to the position of the direct object is always an applicative suffix. Signification in the structure of the predicate — in the Igeva language this is a suffix (-ir-~ -er-). Here the question may arise whether the terms mfumu and ndodo are objects in the sentences (19b) and (20b). We will soon see that the objectiveness of the aforementioned term is evidenced by the order of the words, the coordination of the verb with the object and the reflexes of passivation. In the sentence (19a) without the preposition iSreikStas (i.e. in the aspect of expression not Zymétas) the term (kalata) takes the position if immediately after the verb, and in the sentence (20b) Sia pozicija already takes the term (mfumu), also realized by the non-prepositional ‘construction. The sentence (20b) is ditransitive (or, in other words, a construction of a double object), because Gia is two unsymmetric object complements, but in applicative constructions (i.e., in which the verb has an applicative suffix) the target object, having the semantic function of the recipient/beneficiary, is realized in the first 4 Data on Gieva language application constructions are provided by i8 Kroeger (2004: 63-71). 76 Diatezés alternacijos i§ skirtingy teoriniy modeliy tyrimy perspektyvos in the same position to another, also in terms of expression unsyméta objective complement, possessed by the semantic function of the theme. In most languages with double-object constructions, the main object is also realized in the first effect-small position (cf. English transitive construction (10b)). In addition, in the ci¢eva language, the verb is compulsorily conjugated in the genus with the subject, 0 with the object This conjugation is optional. If actionWord does not have an object matching indicator, the object must be realized immediately after actionWord. However, when the action word has a necessary coordinarealizuotas pirmoje poveiksmazodinéje pozicijoje. Ciéeva kalbos aplikatyviniy konstrukcijy paaukstintasis (arba aplikatyvis (applied object)) objektas, tion with the object Marker, the object does not have to be if the action word does not have a necessary coordination marker, realized immediately after the action word, and if the coordination with the object Marker is insufficient, the object can be realized elsewhere or not at all, cf. : (21) Amayi [a-ku-mu-tumb-ir-a mtsuko],,,, Female(1) SUBJ(1)-PRAES-OBJ- (1)-lieti-APPL-ASP cup(3) (Child). child(1) »Woman pours a cup for him (child)* (Example i§ Kroeger 2004: 67.) Here (21) the verb indicates a connection with the main object (mwana), which is realized beyond the bounds of the action verb phrase, or can even be ignored altogether. The passivation data discussed below also confirm that the terms used in the applicative construction are direct (or main) objects, cf. : (22) a. Ndodo i-ku-yend-er-edw-a. SUBJ(9)-PRAES-vaikStioti-APPL-PASS-ASP sLazda is vaikStoma’ b. *Ndodo i-ku-yend-edw-a (ndi). SUBJ(9)-PRAES- -paikStioti-PASS-ASP with (,p2d. The stick is a childStoma with‘) 2 The Gigeva language has a complex system of genus categories, consisting of 18 genera Classy, when glossing the genus classé is indicated in brackets with Arabic numerals. 77 KRISTINA LENARTAITE In the passive sentence (22a) (passive verbWord Signifikation suffix ~ -edw-) subjektas yra instrumento semantinj vaidmenjturintis termas ndo- -idwdo, tèaieu This sentence's verbWord has also an applicative suffix (-er-), which shows that ndodo is an elevated object, cf. sentences (20a) and (20b), in which the term in question is the circumstance and the object respectively. The verb of the sentence (22b) has no applicative suffix, which indicates that ndodo is the object of the action, so it cannot be used as a subject in a passive sentence. An applicative ditransitive sentence predicate, such as (19b), allows only the recipient, but not the subject, to realize the semantic role of the object (i.e. the main object) in the passive construction of the subject. Thus, let us summarize: the purpose of the applicative construction is to offer the circumstantial term in the syntactic position of the object. This process is always accompanied by a change in the structure of the predicate, which in the ¢i¢eva language Zymi always Zr. 6 i8nagq). In this paper, applicative constructions are used not only to demonstrate the wide variety of alternance phenomena in diatesis, iSrei8kiama aplikatyvine priesaga. | aplikatyviniy konstrukcijy rei8kinj labai panaiios skyriuje 1.1. aptartos vokie¢iy kalbos lokatyvinés alternacijos (dar but also because they are often used by various theoretical models to support their hypotheses about other arguments for alternance phenomena. 1.4. General observations After reviewing all the investigated phenomena, it is possible to summarize and compare how these aspects are reflected in them: (i) changes in alternating sentence meanings and functions, (ii) which syntactic position expression is touched and (iii) how the change of argument status is reflected in the verb word structure. Considering all the investigated cases, the first thing to be noticed is the alternating paraphrasical relationship of the sentences, and in general, it can be said that all these cases function as means of expression of the subordinate actual division. Active/passive alternances typically involve changes in the expression of both the main syntactic positions — both the subject and the direct object. The repeated alterations realize only the variation of the direct object. In the section 1.1. discuss various arguments and cases of alternating expressions allowing to vary the expression of the subject and/or direct object. Moreover, if in activation/passivation and applicative alternances arguments of syntactic status changes are always marked accordingly in the structure of the verb, then other arguments of expression alternances behave very differently in this respect: for example, in German language the localising action-word in English language alternances never has any external sign, in Lithuanian language it behaves inconsistently — often the same verb allows alternance both without any sign, and with the corresponding preposition, but there are also many cases where 78 Diatezés alternacijos i8 skirtingy teoriniy modeliy tyrimy perspektyvos one of the alternating constructions can be formed only with the corresponding preposition verb. Different theoretical models treat these inequalities in the structure of the action-word of the verb in a alternacijy didZioji dauguma atveju Zymima veiksmazZodzio prieSdéliu be-, different way. Obviously, all the investigated phenomena have a lot in common, but is this really a sufficient basis to compare them and seek a common principle regulating these processes? And there can't be one answer to j $j questions. However, on the basis of the definition of all the studied verbs, we can consider these verbs as different syntactical positions of the verbs, which are involved in semantic variables (terms). The following sequence is as subcategories of one supercategory — the category of verbs — if the verb is defined as a variable (inflectional) category, specifying the basic diathesis modifications of the verb lexicon (Mel’¢uk 1993: 11). Here it is necessary to draw attention to the fact that one diathesis is treated in this case as basic, and the other as its modifications, and this is most often explained by the action of the corresponding signs in the given lexem. ‘8 Savoka, kaitybiné kategorija’ should be understood: somewhat differently than usual (i.e. not as a full morphological predicate paradigm), because, for example, in Latvian the predicate of the active and passive races is valid only for participants, but not for finite action-word forms, and the differences in the meaning of the state are determined by the action-word combination, and this is one of the reasons why the oppositions of the active and passive races in Latvian should be viewed rather as a semantic category, but not as a predicate. Moreover, in Lithuanian it is not possible to establish a direct paradigmatic correspondenklus tokios pozicijos laikymasis netgi leidZia visus Siuos rei8kinius traktuoti ce between the active and the corresponding passive form (cf. Holvoet & Seméniené 2004: 36-41). RiSies,,caityviskumas* ¢ia nereiktu susieti su pagrindinés (basic) veiksmažodžio diatezes (i.e. argumenty semantiniy roles and syntax functiony configurations) modifikacijos, kurios dažniausiai yra susijusios su atitinkamais žymėjimais veiksmažodžio struktūroje (cfr. Mel’Cuk 1993: 5-13). The fact that in some cases of alternance in some languages the verb has no additional symbol is not considered a reason to exclude such cases from the content of the corresponding Symbols (cf. Mel’éuk 1993: 25-29). 79 KRISTINA LENARTAITE Considering i8 from such a perspective, the investigated reiSkiniy approximation is certainly justified. 2. THEORETICAL MODEL APPLICATION Before moving on to different theoretical model positions on the issue of diathesis alternances, it should be emphasized that the main emphasis will be placed on the cases of argument expression alternances (i.e. the phenomena discussed in section 1.1), while active/passive and applicative alternances will be used mostly for argumentation purposes. As already mentioned in the previous section, all the theoretical models discussed can be divided into two groups —constructive models and all others, for which it is bitty difficult to find a common pavadinima. 2.1. Generative, relational grammar and lexical rules models All these theoretical models have in common the fact that they aim at one or another level, for example, at the level of the sentence or the semantic structure of the predicate, i.e., at the level of lexical-semantic representation, to define alternating constructions in connecting relationships. Let us first discuss models that recognize multiple syntactic representation units of organization, such as generative and relational grammar. Generative grammar distinguishes between two types of syntactic structure: deep structure and surface structure, the former providing the basis for the semantic interpretation of the sentence, the latter providing the basis for the phonological representation. Transformation rules operate between Siu dviejy lygmeny. For example, it is asserted here that the respective active and passive constructions contain a common deep syntactic structure, only in the realization of the passive construction, before the transformation of the structure organization, special transformational, noun’’ transfer (NP movement- ), rules are applied (see Radford 2002: 420-435). | $i The theoretical model is a grammar of relations, only differently than in generative grammar, where the predicate and 1s Gia Sis term is interpreted in a purely syntactic sense and includes only complements of the verb (Zodiius or jy compounds). 80 Diatezés alternacijos i§ skirtingy teoriniy modeliy tyrimy perspektyvos Its termy relations are defined through their structural configuration (for example, subjective relations have the noun dominated directly by the node of the sentence, and objective — the noun dominated directly by the verb of the sentence phrase), in relational grammar the termy relationship with regard to the predicate — subject, object and indirect object — functions in the syntactic definition directly, i.e. as primitive, and not derived from anything and not defined on the basis of any other element, adjectives. In relational grammar, each term has a grammatical function, and the corresponding rules defining the changes in taksinés reprezentacijos lygmenyje, arba sluoksnyje (stratum), identifikuojathe grammatical function link the existing representation to the previous, underlying representation and re-evaluate the grammatical functions of the term if the existing representation is not of the original layer, i.e. if it encompasses more than one layer (Blake 1990). For example, Gia positive passive constructions are composed of at least iS dveejy slooksniy, plg. : (23) a. John sent the letter to Mary. ,DJohn sent Mary the letter‘ b. John sent Mary the letter. ,Dzonas nusiunté Merei laiska’ c. The letter was sent to Mary by John. ,LaiSkas buvo Merei atsiystas Dzono‘ d. Mary was sent the letter by John. ‘pzd. Meré was John’s letter’ (e.g. Perlmutter & Postal 1986: 92-93). Only the sentence (23a) can be considered as a single-layered one, where the term John performs the functions of subject, letter —direct object, 0 (to) Mary —indirect object. The sentence (23b) is a double sentence. In the final layer, which is represented by the sentence in question, its terms have the following grammatical functions: John —subject, Mary —direct object, and letter —chomeur (French, ‘unemployed’), which is the basic function of the sentence. In the following, sentence (23c) is the passive equivalent of sentence (23a), and sentence (23d) is the passive variant of sentence (23b). The sentence (23c) is a two-layered one, because it has as its base a single-layered representation of the sentence (23a), in which the term functions are distributed as follows: letter —subject, (to) Mary —indirect object, and (by) John —chdmeur. O Stai passive sentence (23d) includes three 81 KRISTINA LENARTAITE layers, since already his basic representation (23b) was two-layered, the termy functions (23d) are divided as follows: Mary —subject, letter —chdmeur, (by) John —chomeur (cf. ibid., 92-95). The sentence pair (23a) and (23b) is a case of »datyviniy* alternancy (cf. sentences (10) in section 1.1. ), so in grammar both,,dative,, and active/passive alternances of relations are realized by sentences that are evaluated equally — one jy is fundamental in relation to the other (i8vestinis). However, one important aspect of relational grammar seems to be undefined: i.e., how the predicate of the relevant predicate changes with respect to different layers of representation: whether it remains the same (although the morphological changes in the predicate structure required in the respective alternating constructions (e.g., the passive form Zymikliai, etc.) are not taken into account when terms are changed)—probably abstracted from it, as from the linear structure of the word in a sentence (cf. Blake 1990: 2). Regarding how generative grammar treats the alternances of the diathesis, we should note that only in the case of matinéms funkcijoms, tai vargu ar imanoma), ar predikato pakitimus turéty savaime atskleisti termy gramatiniy funkcijy pakitimai. Be to, Gia, regis, active/passive alternances a derivative relationship is undoubtedly established, 0 In the case of the so-called locative and,,datyviniy* alternances, even based on the same theoretical model, different researchers give different interpretations. For example, in the Government & Binding version of the grammar, the Projection Principle regulates that all the lexical properties of the verb are encoded at the beginning of the derivation, i.e. in the deepest structure. Subcategorization characteristics of each verb (i.e. complements of the verb, jy tipas (realized by prielinksnine construction (PP) or without preposition (NP)) and the order of complementary linearity in the sentence (Radford 2002: 339-361)) can be deduced from the thematic characteristics of its argument, so thematic roles (for example, agent, subject, etc.) participate in organizing the syntactic projection of the verb. And in the case of Stai’s argument about the alternance of expressions, the fundamental question arises whether there are several verb Zodiacs, each of which specifies thematic roles related to different syntactic structures in a different way, or whether there is only one verb Zodiac, and it should be associated with a single syntactic structure, and the other structure should be considered as derived from the first. Since the alternate skins encompass many different situations, it is not surprising that some of them apply one of the strategies mentioned above and others another. In the case of localative alternations (e.g., a. I loaded hay onto the truck,a3 I loaded Siena ji vezima’ and b. I loaded the truck with hay,a I loaded veZima Sienu /Sieno’) it is more likely to assume that the basis of such a realization is a different action-verb lexicon (cf. Radford 2002: 372-384). O Stai dél,,datyviniu“ alternacijy (cf. (10), (23ab)) unanimous consensus néra’’, so we will look more broadly at the §j case. When the alterations 82 Diatezés alternacijos i8 skirtingy teoriniy modeliy tyrimy perspektyvos of,,dativne* are recognized as related by derivative relation, then the construction of a bitent double object (10b, 23b) is considered to be i8ved from i§,dativne* (10a, 23a). When the cases under consideration tend to be treated as lexical alternations (i.e. as syntactic projections of different verb lexics, not connected by a derivative relationship), the following aspects are often cited as arguments for such a choice: (i) regularity and productivity are usually necessary for the derivative operation, while neither is necessary for,,dativne“* alternations (These alternations include only certain semantic groups of verbs, but it is also difficult to define the limitations that allow / do not allow these alternations in relation to that specific group); (ii) derivatively related alternating constructions are incompatible with the operation of the Projection depth (see above). The Salininkai claim that without recognizing the derivative relationship it is impossible to adequately explain such regular and productive alternances as the applicative alternances (2r. section 1.3.), by the way, the latter are used in the analysis of,,datyviniy“ alternances. In comparing the German language’s localative alternances, where one construction is regularly realized with the action word, which has a limitation in the double object constructions, it was decided to interpret these constructions in English and other German languages as the realizations of the zero applicative suffix. In other words, the predicate in the construction of the double object is derived from the corresponding predicate of the,,dative* construction, but it is realized with a null derivation symbol, i.e., the derivation symbol is not readable. Firstly, this allowed us to identify analogies between many different linguistic phenomena and to explain them on the same basis, and furthermore, such an interpretation helps to solve some of the problems arising in the generative context: when a prepositional phrase is realized in a »dative« construction (cf. 10a, 23a), 6 the different evaluation of the localative and,dative, alternatives will, apparently, in large part determine the uneven development of the constructs with regard to the aspect of paraphrasism, about which more in detail in section 1.1. nacijy realizavimo procesus. Jau matéme, kad su aplikatyvinémis alternaci- (r. skyriy 1.1. ir plg. 6 i8naSa), tad siekiant pagristi transformacinj datyvo 83 KRISTINA LENARTAITE and prepositions, as asserted, are lexicalized in a deep structure and they do not allow their possessive noun-word transformation to move them to the position of the direct object, which seems to contradict the transformative origin of the double object construction —because in the position of the direct object Gia is precisely realized that noun-word (cf. 10b, 23b), which in the corresponding prepositional, “dative” construction (10a, 23a) is expressed as an adjective of the preposition (its possessive). However, this problem is solved if the derivation incorporates a null predicate affix in the construction of the double object, replacing the directional preposition of the,,datyvinés* construction with the semantic postposition of the PATH, and the corresponding transformation in the construction of the double object elevates the object of the depth structure (the prepositional suffix of the,,datyvinés“ construction) to the position of the direct object. Furthermore, the applicative effect of the null affix can be ir compared with the,,conversion’ of derivative morphology, when, for example, in English, also by applying the null affix, lexical biwords are,,converted’’ into verb words, e.g. brown jtudas* and netiesioginj brown,rudinti’. In this way, the problem of non-productiveness of,,datyviniy* alternation is transferred to the level of the lexicon (more Zr. Emonds & Whitney 2006: 73-144). It is not jsivaizduoti, difficult to see how such an interpretation can be applied to other types of alterations, for example, to locative alterations. At this point it is also useful to discuss a slightly different, but also related to the change of prepositional status in alternating constructions — atypical passive constructions, or prepositional passive constructions (prepositional /non-canonical /peculiar passive /pseudopassive), cf. : (24) a. This spoon has been eaten with. This SaucSt was eaten' b. This pub hasn’t been smoked hash in before. ,Sioje smukléje prie’ tai nebuvo rikyta /rilkoma haéi8o‘ (Examples are Kageyama 2006: 88-90.) c. This pot has been cooked in (by generations of Smiths). Sis puodas buvo verdamas ([i8tisy] Smity karty)’ d. This bed hasn’t been slept in (by anyone). This bed was not (no one) sleeping in.’ (Examples in Anderson 1997: 202) ¥ As for the exact location of the zero affix in the configuration structure, there is no unanimous opinion, but it is agreed that it should be directly in the node nés, dominuojamas V (Emonds & Whitney d. 2006: 73-144). 84 Diathesis alternation 15 differenty teoriniy modelsy research perspectives (25) a. Caves are Zvériy inhabited. (Example i§ Ambrazas, ed., 1997: 322.) b. This room was inhabited by him. c. Sitas Saukstas your eating. This road (Zmoniy) is no longer used. e. Sunelis Siandie tebéra nothing given. These examples in English and Lithuanian show that recipient/beneficial (25e), instrumental (24a, 25c) or locative (24b-d, 25abd) circumstantial terms can be realized in atypical cases of passive construction subjects. I8 At first glance it may appear that the activation of subjective nouns (e.g. this spoon, this pot) in English passive constructions (24) corresponds to the activation of prepositional constructions (e.g. with this spoon, in this pot), and the preposition must not allow its suffix to be transferred by the transformation to the position of the subject (as in the case of the above discussed,,datyviniy“ alternance), besides, the transformation of the transfer of NP (nameZodinés phrases) to the position of the subject of the passive construction is possible only in the first effect of the position of the passive NP (usually the direct object is realized in it). But this is not the case: the prepositional suffix is the subject in the corresponding passive construction, and the preposition usually does not disappear in such a passive construction. These cases of passivation are explained by the rule of reanalysis: in the deep syntactic structure after the lexicalisation process (i.e. also after the lexicalisation of both the preposition and its complement- ), the rule of reanalysis allows to reinterpret the relationship between the simple verb and its indirect complement (circumstantial term), represented by the prepositional construction, for example, the constituent structure [V eat] [PP with [NP spoon] that existed before the reanalysis becomes [V eat with] [NP spoon] after the reanalysis (Radford 2002: 427-432). Thus, after such reanalysis, the simple verb word is transformed into a subject, and the noun word in the first position of the action is not prevented from being transferred to the position of the subject during the transformation. So how should we interpret the English sentence (24b), where the subject position of the noun pub is moved not from the first affective position (in this position even in the passive construction the direct object hash is realized), but from the circumstantial position, and how should we interpret the analogous constructions of the Lithuanian language? In Lithuanian, all passive sentence subjects in the respective active 85 KRISTINA LENARTAITE time interval, the capacity will not be filled). However, such interpretation is more difficult to apply to analogous cases, where the simultaneous occurrence of events is more obvious, and therefore the events connecting the initial connection than the development (cf. Iwata 2005: 375-376). In Pinker’s model, the corresponding lexical rule also regulates the fact that the action ‘a word whose semantic structure core is x causes y to move i/on z’ is transformed into a new action ‘a word whose semantic structure core is already x causes z to change its behaviour in such a way that y moves j/on it’ (Pinker 1989: 79). In comparison with the previously discussed explanation, the new Gia is that in the case of a locative alternation the derivation can take place in both directions: both in the i8 locative variant (31a) j with variant (31b), and in the i8 with variant j locative (this is the case of,,inverted subordination“). The direction of the derivation is determined by which argument cannot be omitted, i.e. which argument can go with the only objective complement: if the theme has a semantic role — then the derivation takes place from the locative variant, 0 if the goal — then from the i8 with variant. However, it turns out that this test method is unreliable, because there is a case where in alternating sentences the complements occupying the same syntactic position can both be omitted (or neither can). According to Pinker, such cases as alternances with the verb stuff,(pri-)kimSti‘ show the derivation i§ with of the (32b) variant from the (32a) locative, cf.: (32) a. He stuffed the breadcrumbs into the turkey. jJis prikimSo duonos trupinius j turkey‘ b. He stuffed the turkey with breadcrumbs. is prikimSo kalakuta duonos trupiniuy‘ c. *He stuffed the breadcrumbs. d. He stuffed the turkey. The sentences (32c) and (32d) show that the verb stuff's only objective complement can be murkey, because the sentence (32c) is a non-grammatical one, so the grammatical form is (32b). In this action, the semantic representation of the word has a more complex event structure ('x causes z to change behaviour in such a way that y moves j /on it (z)'), and when the corresponding rule removes the primary dimensions ('xcausesz-changes behaviour in such a way that y moves j /on it (z)'), the subordinate demu may be promoted to the primary dimensions (Iwata 2005: * Referred to 18 Iwata (2005: 358). 92 Diatezes alternacijos i§ skirtingy teoriniy modeliy tyrimy perspektyvos 358-375). In the Pinker model, forms and meaning correspondence are considered to be thematic cores, i.e. schematic representations of the type of event or relationship that constitutes the corresponding action-word class. Cores such as,,x affects y“ or,,x conditiones that y turétu 2‘ are associated with the corresponding syntactic rhymes [NP V NP], [NP V NP NP]. Thus, the verb Word appears in any syntactic rém if its theme Heart is connected with that syntactic rém (ibid., 363). Furthermore, Pinker classified the English locutive verbs that allow and do not allow alternances: some are,,oriented to the content” and others are,,oriented to the corresponding,,added” thing. (content-oriented) biitinai (locatum) argument, while other,,,container-oriented,'' bits specify the location argument. Pasak According to Pinker, action types in which arguments with a goal semantic role can be more easily perceived as undergoing change are more likely to realize their goal arguments as patients, and therefore can participate in alternations (cf. Arad 2006: 474-477). In summary, it should be noted that in both Levin, Rappaport Hovav, and Pinker’s models of “lexical rules” two different, but indirectly connected, semantic representations of the same verb are associated with different implementations of the argument. In the case of non-functional Taigi nors tiesioginio derivacinio rySio tarp dviejy realizacijy, kaip ir leksigrammar, it cannot be determined, but the indirect linking relationshgiai — per leksines taisykles — atitinkama leksines-semantines reprezentaciip can be traced. The fundamental difference between the discussed models of generative and relational grammar and the models of,,lexical rules* is that the former model detects the differences in alternating constructions on the syntactic level, while the latter, on the contrary, detects that the differences in alternating constructions are determined by the processes taking place on the semantic level, which are then correspondingly combined with different syntactic representations, so while the models of,,lexical rules* are single-layered in terms of syntactic representation, the models of generative and relational grammar are multi-layered. 2.2. Design models The constructive models are usage-oriented, i.e., usage-based, and are largely inductive (i.e., from individual cases to generalizations), and as the main determining mechanisms of various linguistic processes, they lead to 93 KRISTINA LENARTAITE vistiniai modeliai, pripazista laipsnisko kalbos i’mokimo procesus, kurie digeneralizations. Furthermore, these models do not recognize such an important process for other models as the structural derivation of one language from another, because, unlike the models discussed before, the constructive models recognize as basic linguistic units the structures established in the usage of the language, which cannot be further broken down into any atomic primitives. Constructions, i.e., learned specific linguistic units with their own form and meaning or discourse function, are recognized as such structures (constructions are also considered to be morphemes, words, idioms) whose aspects of form or meaning cannot be fully deduced from their separate components or from other constructions existing in the language. Furthermore, even a fully-exposed specific case of form-requirement interaction can be recognized as a construction, as long as it is used frequently enough (Goldberg 2006: 5). In this chapter we will discuss the role of diathesis in alternation in the context of Goldberg’s constructional grammar and Iwata’s lexical-constructive model. In the grammar of Goldberg constructions as the main means of sentence expression are distinguished, for example, the following English language argument structure constructions: (33) a. Ditranzityviné konstrukcija _,x kauzuoja y gauti 2‘ Subj V Obj Obj2 Pat faxed Bill the letter. ,Pat faksu nusiunté (nufaksavo) Bilui laiska‘ b. Caused motion construction,x causes y motion j z* Subj V Obj Obl Pat sneezed the napkin off the table. ,p2d. He even stripped the napkin off the table' (Goldberg 1995: 3-4). Such argumentative structure constructions are more likely to be associated with a closely related group of meanings than with a single abstract meaning, although such a hierarchy of meanings in inherited constructions is likely to be distinct (Goldberg 1995: 108-110). In constructive grammar, as in other variants of constructive models, constructions largely determine the differences in sentence requirements, 0 94 Diathesis alternances 18 different theoretical models of study perspectives actionazZodis™ only semantically integrated in its construction. Constructions exist independently of the verb, have their own meaning sintaksine struktira. Pavyzdziui, kauzuoto judéjimo (caused motion) konsand specific traction, x causes y move j z’ argumenty roles —cause, goal, theme —are directly related to the construction, 0 and the DETI participatory roles of the verb —the acting, the being acting, the place of acting —are related to the verb. By “merging*” or integrating the argument roles with the action roles, we obtain a syntactically-semantic structure, e.g.: i CAUSED <cause, goal, theme> MOVEMENT | | | | | DETI <depositing, place, where to be placed, being placed > J 4 \ 4 | Angry. V SUBJ OBL OBJ : (cf. Goldberg 1995: 52) The process of merging the argument role of the construction and the participle role of the verb is regulated by the following: (i) Semantinés dermés principle (Semantic Coherence principle): only semantically compatible roles can be merged*; (ii) The Correspondence Principle: every participant role that is lexical profiled and expressed must be merged with the construct profiled argumenty roles. If the verb has three participatory roles, then one of them may be merged* with a non-profilable role in the structure of the construction argument (Goldberg 1995: 50). Here it is worth noting that the lexical roles of the arguments in the sentence, and the construcprofiliuojami dalyviy vaidmenys yra tie, kurie biitinai reikalaujami i8reiksti tive roles (e.g., part (agent) purpose, subject) are associated with the basic syntactic positions — subject and direct object. Cases of alternation in constructional grammar are interpreted as the meaning of one and the same verbword (i.e., the action perceived on the basis of the relevant semantic-interpretative frame and the participatory roles involved in it). * The meaning of the verbword in constructive models is bitwise linked to the semantics of the relevant semantic-interpretative frame, on the basis of which the limitations of the participatory role of the verbword are determined (see Goldberg 1995; 25-31; 43-44). 95 KRISTINA LENARTAITE the ability to integrate with two or more different arguments in the construction of structures. For example, in the case of the English locutive alternance, or otherwise spray /load, the same action load can be integrated with the causative movement construction (31a) or with the causative + prepositional with construction” (31b), in other words, the sentences in question are to be considered as instances of different constructions: sentence (31a) — the causative movement construction, and sentence (31b) — the causative + prepositional with construction. Differences in the meaning of alternating sentences in this case are determined by differences in the meaning of the construction (cf. Goldberg 1995: 175-179; Goldberg 2006: 34-38). The active/passive alternances Gia are also interpreted as instances of different constructions. The purpose of the passive construction is to shade the role of the highest-ranking participant associated with the action. Thus, the specific passive sentence is to be regarded as an instance of a construction that obscures Gija’s corresponding role as a participant, i.e. a passive construction (Goldberg 1995: 56-57). And the fact that such alterations as passivation or applicative constructions are usually accompanied by morphological derivations is explained in constructive grammar as follows: “The classes of grammatical morpheme related to the above-mentioned phenomena, such as passivation of verbs or applicative suffixes, are comparable to the mazZodzio integracija yra analogiska konstrukcijos ir veiksmaZodzio integraarguments of structural constructions”’, to the semantics of the corresponding morpheme and function (Goldberg 1995: 22-23). Interesting how in this model biityu treated in the section 1.3. discussed the applicative constructions of the Giéeva language: after all, when integrating Gia action words with the same grammatical morphem, more precisely with the group of applicative prefixes (-ir-~ -er-), completely different constructions are obtained, for example, ditransitive (19b) and causative (20b)? In other models, in one way or trukcijy gramatikoje grindziamas taip: (i) geriau A generuoti tiesiogiai, o ne another (by syntactic or semantic means), the abandonment of the derivative relation connecting the phenomena in question constitutes a *Such construction of the argument structure is to be regarded as complex, i.e. consisting of two independent constructions: the causative construction and the prepositional with controlled construction (Goldberg 2006: 36). * In constructive models, as in cognitive grammar, there is no sharp distinction between lexical and grammatical units. Diathesis alternation 18 different theoretical models of study perspectives i8vesti i8 C, if there exists a model B, having the same (target) syntactic properties as A, but obviously not i8vesti i§ C; (ii) it is better not to enter A i8 C if there exists a model B with the same syntactic properties as C, but without an input from which A can be entered (Goldberg 2006: 22-25). In this way, the problem facing other models that recognize derivation at one or another level is solved: the construction of argument structures. In this model, surface generalizations act as a transformation in relation to their specific instances. As a flaw in Goldberg's construction of grammar, it is pointed out that the higher-ranking abstract constructions are chosen as the basic linguistic units, although the facts of child language learning and other facts show that the lower-ranking constructions related to specific verb classes function as the basic units in this respect. Abstract constructions of a higher rank cannot exist without those of a lower rank, so that the former cannot be considered more adequate in relation to psychological realities than the latter (Iwata 2008: 123-129). Another flaw referred to concerns the operation of the Conformity Principle discussed above: firstly, as the application of the omission test shows, there is a problem in determining what is lexical when profiling the participant role of the action, i.e. the bit that must be a participant role, and, furthermore, according to Iwata, this principle duplicates the principle of semantic dermis and can generally be done without it (i.e. the Conformity Principle) (Iwata 2005: 382-387). It may also appear that the information conveyed in the constructional grammar on different levels is repetitive, redundant, for example, the participle roles related to verbs seem to convey enough information about the syntactic projections of arguments, but all this is also repeated on the constructional level. Moreover, excess information forms the essence of the instance mechanism, which is extremely important in cognitive and constructive grammar. The purpose of abstract constructions (functional motivation) is to sanction more concrete constructions of the Lower level, for example, sentences, as well as newly formed ones. In this way, the binary constructions perform the same function as the rules in generative models. * For constructive models, such as usage-based models, it is very important that the recognized grammatical units and statements are adequate in relation to other psychological, cognitive processes, i.e. that they are psychologically real to the speaker/listener. 97 KRISTINA LENARTAITE In Iwata's lexical-constructional approach, although Goldberg's constructional grammar is followed, all its main aspects are slightly modified: (i) in the constructional grammar, the upper-ranking abstract constructions are chosen as the main unit of grammatical expression, while in the lexical-constructional approach, the lower-ranking constructions related to a specific class of verbs are chosen as the main units; (ii) in constructional grammar, the roles of the actors in the action can be reduced to the purpose of the verb, i.e. the purpose of the verb can be represented by the role of the actor in the action, 0 in mathematics, the relationship between the verb and the construction is determined by how the purpose of a particular verb is compatible with the purpose of the construction, leksinéje-konstrukcinéje gramatikoje veiksmaZodiio reik8mé neapribojama vien dalyviy vaidmenimis, ¢ia ji suvokiama placiau; (iii) konstrukcijy gra0 in the lexical-constructive model it is important how the whole sentence, for example, He sprayed the wall with paint He sprayed the wall with paint, is compatible with the construction of a particular class of verbs (see Iwata 2008: 123). The advantage of using lower-ranking constructions has already been mentioned above, and the fact that the parameters traditionally associated with the action word lexicon (e.g., argument structure) cannot adequately reflect compatibility with the construction is shown by such cases when this compatibility is obviously determined not only by the action word, but also by other elements of the sentence, i.e. with some terms action word can be realized in a concrete construction, and with others — not, cf. : (34) a. *He spread the sleeping child with a blanket. He covered the sleeping child with a blanket. He spread the bread with butter. He (ap-)greased bread with butter‘ (Examples i$ Iwata 2008: 124.) Both of Iwata’s works discussed in this article (Iwata 2005; 2008) deal with the problem of localiative alternation, and the main questions raised in this regard are: why do some verbs allow localiative alternation and others do not; i8 From the latter question other questions arise: which aspects of the semantics of verbs participating in localization alterations determine this possibility and why. It is therefore obvious that Iwata’s study, after considering the statement by Pinker and others that the so-called with variant (see (31b)), in which the member having the role of location, or yra labiau orientuotas j lokatyviniy alternacijy semantikos analize. Detaliau 98 Diatezés alternacijos i§ skirtingy teoriniy modeliy tyrimy perspektyvos rather the role of the locative, goes to the object, can be characterized as a Significant,,- (locative) change of state“ (change of state) of another alternating sentence, the sign,,(theme) change of location* (onto /into variant, see (31a)), Iwata came to the conclusion that,,(locative) change of state* characteristics and dél its too general form, and dél the significant localitive alternances in the semantic differences between the prototypical,,change of state* verbs, such as break, break, semantics cannot adequately describe the semantic aspect of the case under study (Iwata 2008: 22-25). Furthermore, in his proposed lexical-constructive model, Iwata maintains that only the Lower-ranking, i.e. with concrete alternative with variant (31b), where the locative goes to the object, is to be described as realizing the semantics of the cover verb class (0 in some cases — the fill verb class), 0 into /onto variant (31a), where the location-changing object (locatum) goes to the object, realizes the semantics of the put verb class. A construction related to a specific class of verbs is to be considered as a sanctioning scheme’’ in relation to the concrete realization Gia veiksmazodzZiu klase susijusios, konstrukcijos gali adekvatiai atspindéti lokatyviniy alternacijy semantinius ir sintaksinius procesus. Lokatyviniu that instances it, cf. : (35) a. The vandal sprayed paint onto the sculpture. "Vandal sprayed daZus on sculpture' b. The vandal sprayed the sculpture with paint. 'Vandal (api-)pure sculpture dazais' (Example i8 Iwata 2008: 48.) In use-oriented theoretical models, constructions are nothing more than abstract schemes of use events (sayings) and combinations of form and meaning. The constructions, perceived as schemes encompassing their specific implementation communities established in the usage, are in those implementations in origin and sanction new implementations that correspond to the specifications (Iwata 2008: 38). Both the concrete use of the verb involved in the alternances and the use of the construction in relation to a specific event of use are to be regarded as certain abstractions. Furthermore, the meaning of an action word is difficult to distinguish from the més, t. y. jos persikloja, Tad konkretaus verksmazodzio reik’mé jsivaizduotina kaip meaning of the construction of the minimum of information associated with that action word (ibid., 87-99). 99 KRISTINA LENARTAITE Sentence (35a) is considered to be a word class of rank constructions, ‘put* ,déti® semantic component, instance, o sentence (35b) verb class — rank constructions containing cover ,cover’ semantic nj component, concrete realization (ibid., 27-51). Alternatives are available when the action is paties (its specifications and boundaries are defined by the corresponding semantic-interpretative framework) bit compatible gali (integrated) in two different su designs. Other significant localiative alterations permesianéiy verbazodziy semantics poZnaky are Sie: action holistic polymorphism ir veiksmo (t. y. assumed end of the action). In kaip sentences, e.g. (31b) ir (35b), where the locative goes object, These sentences sanctioning ,,cover“ (35b) or ,,son* (31b) semantic verb class condition the appearance of additional semantic componentholist effect, to — — t. y. the perception that the action-experiencing object is completely to encompassed by the action. Tatiau, for example, sentence (35b) leidzia double — ir holistic, ir non-holistic — interpretation: even the slightest spray of paint on the sculpture, as a piece of art, Zelis can be considered as a sculptural biiklés, t. y. estetinés i$szerep, változás, ir dél to possible realization (35b). Taéiau, pasak Iwatos, only realizations with a holistic effect are sanctioned ,,cover“ / ,,fill* semantic verbazodiy classés construction, then the same sentence (35b), kai leis holistine interpretacija, bus vertinamas kaip minétosios grupés konstrukciju instanciacija, kai neleis reik8, kad sankcionuoja kurios nors o kitos veiksmazodiiy — tai (pavyzdziui, ,,bjaurojimo“ ji veiksmazZodziy) klasés konstrukcija. Be that, if a particular sentence choosing a locative object does not allow a holistic interpretation, in such a case, in reverse evaluation, it is not quite a true case of locative alternance, because it is sanctioned ne ,,cover“ / ,,fill" semantic action-based class constructions. Similarly, the semantic ir component of the fullness reflex is explained: if the locative object of the chosen sentence is full veiksmas (non-testi- / ng final), that they are sanctioned ,,cover“ / , fill semantic action class constructions, if any other action class 0 construction is present — (ibid., In examining such cases 55-64). of locative alternances, where a sentence is composed of more morphologically complex verbs, such as the German locative alternances su iy (plg. sentence or (9)) Japanese language localising alternations, Iwata behaves inconsistently: German prepositions su (without-) verbazodiiais locative in the object position realizes sentences 100 Diatezés alternacijos i§ skirtingy teoriniy modeliy tyrimy perspektyvos (9b), in general, strictly speaking, it is not considered in the case of locative alternances, because they are not sanctioned by the construction of the,,cover“ /,,fill* semantic verb class, because they are not subject to a holistic interpretation (ibid., 2008: 152-156); o Here, for example, in Japanese sentences of the locative object formed with the prepositional (-tsukusu) verb Word, since the realizations with these verb Words can be considered as instances of the semantic construction of the verb class cover / fill and the preposition -tsukusu is not a grammatical morpheme, these cases of alternance are considered in the same way as the cases of the locative alternance in English (ibid., 2008: 188-192). The lexical-constructive model can be based on the following aspects: (i) all alternating sentences, where in one of their object positions is realized a sentence member having a semantic role of a locative (in the general sense), can be called,,locative alternances“, and Iwata tends to consider only those cases that are relevant to his semantic restrictions as localative alternations; (ii) the role of a certain aspect of the semantic structure of a word — holisticism and holisticism — seems to be overestimated here: holisticism is always relative — it is always assessed by the speaker, and the possibility of several (holistic/- non-holistic) interpretations of such a sentence, such as (35b), just shows that holisticism does not necessarily play the decisive role that Iwata would like to attribute to it in the case of a locative alternation (see also Michaelis & Ruppenhofer 2001: 23-27). The holistic approach is, of course, suitable for the vast majority of Latvian alternatives, but not for all of them. Thus, although constructive models can undoubtedly explain argumentatively the phenomena of alternation, rather than,,alternations“ in the literal sense of this term, in these models the phenomena in question do not exist, because there is no distinct relationship between them. In constructive models, in general, all possible implementations with the same action word can be called bytes alternances. 2.3. General observations It is now necessary to comment briefly on why all the theoretical models discussed in Gia (sections 2.1. and 2.2.) are refused to be classified on the basis of derivation, i.e. i derivative and non-derivative. Dél generatyvinés ir relaciju gramatiky bei »leksiniy rules* modely has already been mentioned that there are derivation processes 101 KRISTINA LENARTAITE ComRIE, B. & M. POLINSKY, eds. 1993: Causatives and Transitivity. AmsterdamPhiladelphia: John Benjamins. CHAFE [CEJF], W., L. 2003: Meaning and Structure of Language, Moscow: USSR. DESCLES, J.-P., Z. GUENTCHEVA & S. SHAUMYAN 1985: Passivization in Applicative Grammar. Amsterdam-Philadelphia: John Benjamins Publishing Company. EMONDS, J. & R. WHITNEY 2006: Double Object Constructions. In: EVERAERT & VAN RIEMSDIJK, eds., vol. 2, 2006, 73-145. EVERAERT, M. & H. VAN RIEMSDIJK, eds., 2006: The Blackwell Companion to Syntax, vols. 1-5. Malden-Oxford-Victoria: Blackwell Publishing. FOLEY, W. 2007: A typology of information packaging in the clause. In: SHOPEN, ed., vol 1, 2007, 362-446. GIvON, T. 1979: On understanding grammar. New York: Academic Press. GOLDBERG, A., E. 1995: Constructions: A Construction Grammar Approach to Argument Structure. Chicago: University of Chicago Press. GOLDBERG, A., E. 2006: Constructions at work: The nature of generalization in language. Oxford-New York: Oxford University Press. HOLVOET, A. 2003: Current problems of partitioning. In: HOLVOET & JUDZENTIS, red., 2003, 79-98. HOLVOET, A. 2009: Argument hierarchy and grammatical functions. In: HOLVOET & MIKULSKAS, red., 2009, 1-36. HOLVOET, A. 2009: Diffuse entities and objects. In: HOLVOET & MIKULSKAS, red., 2009, 37-68. HOLVOET, A. & A. JUDZENTIS, red. 2003: Syntax relations studies. Vilnius: Lithuanian Language Institute. HOLVOET, A. & R. MIKULSKAS, red. 2009: Grammatical functions and their nature and expression. Vilnius: Vilnius University & Association,Academia Salensis. (Records Salensia, 1.) HOLVOET, A. & L. SEMENIENE, 2004: RiSies category: medium and passive. In: HOLVOET & SEMENIENE, red., 2004, 35-60. 108 Diatezes alternacijos i8 skirtingy teoriniy modeliy tyrimy perspektyvos HOLVOET, A. & L. SEMENIENE, eds. 2004: Grammatical category research, Vilnius: Lithuanian Language Institute. IwATA, S. 2005: Locative alternation and two levels of verb meaning. Cognitive Linguistics 16-2, 355-407. IwATA, S. 2008: Locative Alternation: A Lexical-Constructional Approach. Amsterdam—Philadelphia: John Benjamins Publishing Company. JABLONSKIS, J. 1997: Grammar of the Lithuanian language, Vilnius: Mokslo ir enciklopedijy leidykla. JACKENDOFF, R. 1990: Semantic structures. Cambridge MA: KAGEYAMA, T. 2006: Property description as a voice phenomenon. In: TSUNODA & KAGEYAMA, eds., 2006, 85-114. MIT Press. KAPLAN, R., M. & J. BRESNAN 1982: Lexical-Functional Grammar: A formal system for grammatical representation. In: BRESNAN, ed., 1982, 173-281. KEENAN, E. L. & M. S. DRYER 2007: Passive in the world’s languages. In: SHOPEN, ed., vol. 1, 2007, 325-361. KLAIMAN, M.H. 1991: Grammatical voice. Cambridge: Cambridge University Press. KROEGER, P. R. 2004: Analyzing Syntax: A Lexical-Functional Approach. Cambridge: Cambridge University Press. * KROEGER, P. R. 2005: Analyzing Grammar: An Introduction. Cambridge: Cambridge University Press. LENARTAITE, K. 2007: The argumentative alternance of syllables in Lithuanian language and the applicative grammar tools for it aiSkinti. Acta Linguistica Lithuanica 57, 17-44. LENARTAITE, K. 2009: Direct objects and the question of transitive construction in Lithuanian language. In: HOLVOET & MIKULSKAS, red., 2009, 69-98. LEVIN, B. 1993: English Verb Classes and Alternations: A Preliminary Investigation. Chicago-- London: University of Chicago Press. LEVIN, B. & M. HOVAV REPORT 2005: Argument Realization. Cambridge: Cambridge University Press. 109 KRISTINA LENARTAITE MALCHUKOY, A., M. HASPELMATH & B. COMRIE 2007: Ditransitive constructions: a typological overview. http://www.haspelmt.de/~haspelmt/Ditransitive.html MATTHEWS, P., H. 2007: The Concise Oxford Dictionary of Linguistics. Oxford-New York: Oxford University Press. MEL’CUK, I. 1993: Voice: Towards a a rigorous definition. In: COMRIE & POLINSKY, eds., 1993, 1-46. MICHAELIS, L. & J. RUPPENHOFER 2001: Beyond Alternations: A Constructional Model of the German Applicative Pattern. Stanford: CSLI Publications. PERLMUTTER, D., M. & P. M. POSTAL 1986: Toward universal a characterization passivization. of In: PERLMUTTER, ed., 1986, 3-29. PERLMUTTER, D. M. & P. M. POSTAL 1986: Some proposed laws basic of clause structure. In: Perlmutter, ed., 1986, 81-128. PERLMUTTER, D., M., ed. 1986: Studies in Relational Grammar 1. Chicago-London: The University of Chicago Press. PINKER, S. 1989: Learnability and Cognition: The Acquisition of Argument Structure. Cambridge MA: MIT Press. REPORT, M. & B, LEVIN 1988: What to do with theta-roles. In: W. WILKINS, ed., Syntax and Semantics 21: Thematic Relationships. New York: Academic Press, 7-36. Press release, M. & B. LEVIN 2008: The English dative alternative: The case for verb sensitivity. Journal of Linguistics 44, 129-167. RADFORD, A. 2002: Transformational Grammar. Cambridge: Cambridge University Press. SHAUMYAN, S. 1987: A Semiotic Theory of Language. Bloomington to Indianapolis: Indiana University Press. SHAUMYAN, S. 2006: Signs, Mind and Reality: A Theory of Language as the Folk Model of the World. Amsterdam to Philadelphia: John Benjamins. SHOPEN, T., ed., Language 2007: Typology and Syntactic Description. Vol. 1: Clause Structure. Cambridge: Cambridge University Press. 110 Diatezés alternacijos is skirtingy teoriniy modeliy tyrimy perspectives SILVERSTEIN, M. 1976: Hierarchy of features and ergativity. In: R. M. W. DIXON, ed., Grammatical Categories in Australian Languages. Canberra: Australian Institute of Aboriginal Studies, 112-171. TSUNODA, T. & T. KAGEYAMA, eds., Voice 2006: and Grammatical Relationships. Amsterdam to Philadelphia: John Benjamins. VAN VALIN, R. D. & R. J. LAPOLLA 1997: Syntax: Structure, Meaning and Function. Cambridge: Cambridge University Press. Kristina Lenartaité Lietuviy kalbos institutas P. VileiSio 5, LT-10308 Vilnius [email protected] 11