LIETUVIŲ KALBOTYROS KLAUSIMAI XXXIII (1995)

GRAMATIKA IR LEKSIKOLOGIJA

Gertrūda SAVIČIŪTĖ-NAKTINIENĖ

TRENDS IN THE DEVELOPMENT OF CAUSATIVES IN LITHUANIAN AND OTHER INDO-EUROPEAN LANGUAGES

1. Introductory remarks

The diachronic approach to research, so common in works on phonetics and morphology, is making its way into the fields of syntax and semantics. Modern typological linguistics would benefit particularly from going beyond the mere synchronic description of existing universals and taking into account both the genetic relationships among languages and general patterns of language development (cf. Климов 1975: 28).

The work of many authors and the material examined in our study show that causativity is a universal (panchronic) category of the Indo-European languages, expressed by means that have usually arisen in individual languages. Thus, the correspondences between the semantics and means of expression of causatives, as well as trends in their development, are best discussed from a typological perspective.

As in previous studies (1980¹, 1980²), the subject of this study is Lithuanian, Latin, and Russian, languages that are genetically related and typologically close, but belong to three different branches of the Indo-European family. Most of the material has been gathered from dictionaries, grammars, and other works on these languages; where necessary, data from other Indo-European languages have also been used.

From a grammatical perspective, causatives in the languages under study can be divided into morphological and syntactic, or periphrastic, types; from a typological perspective, into synthetic and analytic types; and from the perspective of syntactic semantics, into closed and open types. The latter are distinguished according to whether their surface structure corresponds to the deep, or underlying, structure, which is described by the paraphrase “do so that + a verb in the indicative mood.” This paraphrase is based on a subordinate clause expressing result. It was chosen because result clauses, rather than purpose or causal clauses, share the most features with sentences whose predicates are causative verbs or analytic causative constructions (for a fuller discussion of the causative paraphrase, see Savičiūtė 1986¹: 164—165; 1986²: 206—209).

Causatives whose surface structure does not correspond to the aforementioned paraphrase and which express causativity in a compact form are called closed; those that correspond to the paraphrase and are expanded at the surface level are called open. The former are characterized by morphological forms of expression, and the latter by syntactic, or periphrastic, forms.

110

2. Closed causatives

2.0. The tendency of the ancient Indo-European inflectional languages to express complex meanings synthetically is demonstrated by morphological causatives—causative and curative verbs that express two (1) or three (2) different actions in a single sentence. The root of a causative verb, which coincides with the root of the base verb, expresses the caused action, i.e. the action or state sought or brought about by the subject, the causer, e.g.: (1) to raise — to rise, to make sink — to sink, to make someone eat — to eat, (2) to have something woven — to weave, to have something sent — to send, to have something repaired — to repair. The suffix and/or vowel alternation marks the abstract meaning of causation, i.e. the causing action “to make it so that...”. The suffix itself does not indicate the manner of causation; this becomes clear from the broader context.

2.1. Closed causatives expressing simple causation

2.1.1. If each component of the deep structure of a causative sentence is represented by a symbol (S—subject, P—predicate, O—object, → causative relation), the deep structure of causative sentences can be represented by the formula (S + P) → (O + P). The aforementioned paraphrase “to make it so that...” helps to fill this formula with lexical and grammatical content. A causative sentence has four components expressing two situations: (1) the causing situation (the causer subject and its causing action) and (2) the caused situation (the object and the change of state or action caused in it by the subject). For example, the sentence The host feeds the horse can be analyzed as follows: “(1) The host + (2) does something so that + (3) the horse + (4) drinks”, or The sun faded the shirt: “(1) The sun + (2) made it so that + (3) the shirt + (4) faded”.

2.1.2. Closed causatives in Indo-European languages are verbs with a particular ablaut grade and/or a causative suffix. Not all causative suffixes can be formally derived from Proto-Indo-European.

A. Meillet distinguishes two causative suffixes in the ancient Indo-European languages: (1) *-eje/o- and (2) -ijo-/-ī- (Meillet 1951: 190). Reflexes of the first suffix are found in Indo-Iranian, Ancient Greek, and, in part, Latin. For example, Ancient Greek φεβομαι “I fear” — φοβέω “I frighten”. In Sanskrit, this suffix corresponds to -áya (Burrow 1959: 301—302, 330; Елизаренкова 1982: 367—370), e.g., vart-áya-ti “turns” — vārta-ti “turns around”; in Latin, it corresponds to *-eie/ō, e.g.: moneō “I remind” — moneō “I remember”, doceō “I teach” — disco “I learn” (Тронский 1960: 215; Safarewicz 1969: 212). Other, less numerous Latin causative verbs should also be noted, e.g.: augeo “I increase, enlarge”, suādeō “I advise, persuade, convince”, terreō “I frighten”, torreō “I dry” (Ernout 1927: 231—232). Rare Latin forms attest to the second Indo-European causative suffix, -ijo-/ī-, e.g., sōpiō “I put to sleep” (Ernout 1927: 235).

The suffix -ja- and vowel ablaut in the root are characteristic of ancient Germanic causatives, e.g.: Gothic bi-leiban, Old English belifan, Old Frisian b(i)liva, Old High German bi-līban “remain” — Gothic bi-laibja-n, Old Icelandic leifa, Old English læfan, Old Frisian lēva, lāvi-a “leave” (СГГЯ IV 142, 169).

In modern German, causatives expressed by root vowel ablaut have survived—

111

remnants of causativity, e.g.: liegen “to lie” — legen “to lay, put,” sitzen “to sit” — setzen “to seat,” trinken “to drink” — tränken “to water,” etc. Remnants of morphological expression of causativity are also found in present-day English, which has an analytic structure. The verb wake (woke, waked; waked, woken, woke) “to wake up; to awaken” has both a causative and a non-causative meaning. Of the past-tense forms woke and waked, the latter is more often chosen to express causativity. The verb rise (rose, risen) “to rise, get up” differs from raise “to raise, erect, awaken” in its vowel grade, although R. Lakoff (1972: 187) argues that most English speakers do not feel that raise is derived from rise.

In Russian, causativity can be expressed irregularly by suffixes, prefixes and suffixes, and vowel alternation, e.g.: висеть “to hang” — вешать “to hang (something),” гаснуть “to go out” — гасить “to extinguish,” гибнуть “to perish” — губить “to kill,” мерзнуть “to freeze” — морозить “to chill,” спать “to sleep” — усыплять “to put to sleep,” сохнуть “to dry” — сушить “to dry (something).”

The class of morphological causatives in Lithuanian is particularly large. These are causative verbs whose causative meaning is formed by vowel alternation and/or the Baltic-origin suffix - (d)inti, -(d)ути (Skardžius 1941: 529—548; Jablonskis 1957 I: 288—290; Jakaitienė 1968: 227; Якайтеne 1968: 11; LKG 1971: 25, 264; LKM 1976: 110, etc.). Correspondents of the suffix -yti are sometimes found in Slavic languages as well (Skardžius 1941: 535).

2.1.3. The origin of causative verbs is a complex problem.

A. Desnickaja, relying on N. Marr’s claim that nouns arose earlier than verbs, argued that causative verbs are nominal in origin (1941: 138— 165). However, linguistic data and the work of many researchers show that causative verbs can derive from both nouns and verbs. Causatives may be associated: (1) with verbs, (2) with both nouns and primary verbs, (3) with nouns, e.g.: (1) Skt. sadáya-ti “seats” — sīdati “sits,” (2) Skt. rādháya-ta “satisfies” — rādhoti “accomplishes” — rādhah “gift,” (3) Old High German wermen “to warm” — warm “warm” (cf. Э. Менчер 1966: 73—79).

The indubitable derivational connection between causative and simple verbs, clearly perceptible in present-day Lithuanian, is reflected in the entries of the “Dictionary of Contemporary Lithuanian” (1972) (hereafter DŽ) and the “Dictionary of the Lithuanian Language” (hereafter LKŽ), e.g.: laužti caus. lūžti LKŽ, pūdyti caus. pūti DŽ. The semantic link between morphological causatives and nouns is also confirmed by the fact that many Lithuanian suffixal verbs can be paraphrased as “make something [adjective],” e.g.: kreivinti “make crooked” — kreivas “crooked,” platinti “make wide” — platus “wide” (Якайтеne 1968: 11). This is precisely how LKŽ and DŽ explain verbs of this type, which are not marked with the abbreviations caus. or prž., although most of them have a corresponding verb of the same root meaning “become [adjective],” e.g.: silpninti “make weak, weaker” — silpnas — silpti, silpnėti; slapinti “make wet” — šlapias — šlapti. The fact that dictionaries choose this method of presentation indicates an understanding of their nominal origin. However, it is sometimes genuinely unclear what causative verbs of this type should be related to (cf. Skardžius 1941: 541—544), and dictionary authors choose a compromise; for example, beside sendinti they add the abbreviation caus. and refer to the verb—

112

senti, while explaining the meanings with the paraphrase “make [something] [adjective]” and relating it to the adjective senas (LKŽ XII 380).

There is a known hypothesis that causative verbs derive from iterative verbs. B. Delbrück (1897: 119) argued that the causative verb vardháyati “to raise” arose through a change in meaning from the verb vardháyatē “to grow vigorously,” which had developed from várdhatē “to grow.”

B. Sebrennikov also considered it possible for causative suffixes to derive from frequentative suffixes, on the basis of the frequent co-occurrence of these suffixes and the proximity of their meanings (Серебренников 1974: 175–176). The closeness of causative and iterative meanings is also confirmed by the Lithuanian suffix -(d)inti, -(d)yti, which can have both meanings, e.g.: spaudýti, nardýti, šaudýti and gydýti, guldýti, žudýti, etc. Derivatives with this suffix are sometimes even called iteratives-causatives (Кауkене 1991: 21).

The intensity of the causing action expressed by causative verbs and result clauses has been noted (Savčiūtė 1986: 159). This too points to a connection between iteratives (at least those that have an intensive nuance) and causatives. However, none of the facts mentioned here proves that causatives originated from iteratives. Their similarity is apparently based on the activity of the causing and iterative actions themselves (in cases where iteratives have an intensive nuance).

Thus, in the absence of a clear and concrete historical prototype from which the various causatives of Indo-European languages could be derived, the origin of causativity came to be associated with the development of semantic oppositions. J. Kurylowicz (1964: 86–88) regarded the semantic opposition between state and caused action as the starting point for the development of causativity in Indo-European languages (state > intransitive action > causativity).

Authors who investigate transitivity–intransitivity on the basis of oppositions discuss causativity as one case of transitivity–intransitivity. S. Krinickaitė showed that an opposition of external and internal diathesis exists between transitive verbs with vowel alternation and intransitive verbs, e.g., versti – virsti. This opposition, which constitutes the oldest layer of the transitivity–intransitivity category, initially expressed the semantic opposition between the state of an object or other inanimate subject and the action of a human or animate subject. A later opposition, non-causative verb / causative verb, became more pronounced (Криницкайте 1979: 46–49).

A. Paulauskienė (1979: 21, 26) pagrindine laiko tranzityvumo–intranzityvumo, o ne kauzatyvumo kategoriją. Ji skiria tris tranzityvumo tipus: (1) keisti – kisti, (2) augti – auginti, (3) keisti – keistis ir kalba apie tranzityvumo ir kauzatyvumo rodiklių bei funkcijų sutapimą.

J. Stepanov linked the emergence of Indo-European morphological causatives with the reconstruction of Proto-Indo-European simple sentence types. According to him, during the period of active alignment the following sentence types existed: (1) active subject + active verb, (2) inactive subject + inactive verb, (3) active subject + active verb + inactive object. However, at that time there was no grammatically formalized type (4) active subject + active verb + active object, e.g.: A warrior kills an enemy. In J. Stepanov’s view, causatives such as Lithuanian žudyti (: žūti), marinti (: mirti), Russian

113

“to make die” ‘to kill, destroy’ (: -merėti, -mirati (umirati) ‘to die’, Sanskrit maráuati ‘makes [someone] die, lets [someone] die, kills’ (: márati, márate, mṛyáte ‘dies’) are related precisely to the formation of the aforementioned sentence type (4). As the nature of sentence structure itself changed, sentences of this type became commonplace. The fact that the object of many Lithuanian causatives can be a person, e.g., klaidinti — klysti, is likewise regarded as evidence for this hypothesis. The emergence of morphological causative derivation is considered one of the means that helped avoid certain difficulties as the structure of Indo-European languages changed from active to accusative.

2.2. Closed causatives expressing more complex causation (curatives)

2.2.1. Of the inflectional Indo-European languages discussed here, only Lithuanian has developed a class of curatives, or verbs of procuring (verba curativa seu causativa passiva). It is distinguished from causative, or causal, verbs (verba causativa activa).

Causative and curative verbs have the same suffix -dinti (curatives are rarely formed with the causative suffix -dyti and especially -yti), and generally have the same sentence structure, so they are often confused (for a fuller discussion of their meaning and the criteria for distinguishing them, see Jakulienė 1975: 78 — 81; Savičiūtė 1984: 73 — 78, 1985: 236 — 251).

Since the two classes have much in common, and the causative class is much larger and more varied in form and meaning, this suggests that curative verbs arose from causative verbs.

The causative suffix formant -d-, once merely a means of eliminating hiatus, later became a characteristic marker of curative meaning (Stang 1966: 374), while the causative suffix began to be added to an already formed morphological causative, e.g., lakinti — lakindinti, or to a factitive verb, which may be called an inflectional causative, e.g., siūti — siūdinti. Similar secondary causative derivation has also been observed in other languages (cf. Недялков 1969: 27).

The differences between causative and curative sentences are especially evident in the case of ternary oppositions, e.g.: juokindinti — juokinti — juoktis, kalbindinti — kalbinti — kalbėti, lauždinti — laužti — lūžti, lodindinti — lodinti — loti. Compare sentences with verbs in ternary opposition: I will have shingles torn from the maple branches (cur.), and I will have planks sawn from the trunk LTR (Lkv) — We tear willows (caus.) and we weave bast shoes Švnč — The bast is already splitting (intr.) J (LKŽ X 166, 179, 248).

When the aforementioned causative and curative sentences are expressed using the general causative periphrasis, or their deep structure is represented by a formula, their principal differences become clear:

Causation: (1) “I make it so that...” + (2) “The bast splits” (S + P)→(O + P)

Curation: (1) “I will make it so that...” + (2) “[Someone] will make it so that” + (3) “The shingles will split” (S₁ + P₁)→(S₂ + P₂)→(O + P₃)

In causative sentences, four syntactic-semantic components were identified,

114

describing 2 situations. Curative sentences have 6 components describing 3 situations:

(1) the initiating situation (subject, curator—initiator S₁ and its curative action P₁),

(2) the intermediate curated situation (subject, intermediary or executor S₂ and its directly performed action P₂, which was curated by the initiating subject),

(3) the final curated situation (object O and its change of state or action, caused by the intermediary subject and curated by the initiating subject).

Thus, we can refine the deep-level formula for curative sentences:

(S₁ + P₁) cur. → {(S₂ + P₂) caus. → (O + P₃)}

The object of a curative sentence is affected twice—directly by the intermediary subject and indirectly by the initiating subject. Therefore, we have provisionally called curation “double causation”.

From the semantic perspective, it seems possible to assert that the curative meaning arose on the basis of verbal (or, in D.A. Cruse’s terminology, requestive) causation, in which an indirect, non-physical influence of a living subject on another living subject was expressed.

Such indirect (in V. Nedialkov’s terminology, non-contact) causation, close to the meaning of Lithuanian curatives, is also found in other languages, e.g.: German lassen + infinitive (Миссюра 1952: 10; Недялков 1971: 19 tt.), French faire + infinitive (Абросимова 1954: 7—10).

The surface level of Lithuanian curative sentences developed on the basis of sentences expressing indirect, non-physical (verbal) causation, in which S₁ is a person, and P₁ means “order, command, ask...”; e.g.: Make (caus.) the children crawl into the space behind the stove, Šts (LKŽ VII 119). As curation developed, the semantics of the components narrowed: S₁ is only (1) a person with the authority to give orders, e.g., king, lord, etc., or (2) a person unable or unskilled to do a particular job, a client; P₁ means (1) “order, command” or (2) “commission, ask”.

The intermediate situation in curation “surfaces” only in rare and atypical cases. E.g.: He had three barrels quickly carried into the reception room by servants K. Donel (LKŽ VIII 702); He had the thief torn apart by dogs for his third theft S. Dauk (LKŽ II 644). Most often it is inferred from context or from the grammatical meaning of curative verbs. This intermediate situation in curation, usually present only at the deep level, is close to direct physical causation, in which S₁ in the causative sentence is a person, object, or force of nature, and P₁ denotes various physical, and rarely non-physical, actions: The child (wind, apples) breaks the branch.

In curatives, S₂ is only (1) a person carrying out an order, e.g., a servant or executioner, or (2) a person able and skilled to do a particular job, e.g., a tailor or barber; P₂ denotes (1) various verbs expressing services and punishments, e.g., have carried in, have beheaded, or (2) verbs denoting various professional actions requiring a particular skill or expertise, e.g., weave, build (a building). E.g.:

115

(1) Herod had beheaded Saint John

BP II 366 (LKŽ II 854).

(2) H a v e t h e m b u i l d a field hut at the far end J (LKŽ XIII 674).

In this way, sentences of indirect, nonphysical (verbal) causation were adapted to express a more complex situation in reality. Curation and curative (care-taking) verbs developed through (1) the narrowing and specialization of the meaning of their components and (2) an increasing complexity of deep structure—with the emergence at the deep level, and in atypical cases also at the surface level, of one more subject and its action.

3. Open causatives

3.0. Causatives that express more deep-level components at the surface level and whose form is closer to a causative periphrasis are called open. Open causatives include analytic causative constructions and subordinate result clauses.

3.1. Open causatives expressing the causing and caused actions with separate words

3.1. The general tendency in the development of inflectional Indo-European languages to shift from a synthetic to an analytic type is also evidenced by the formation of periphrastic causatives (cf. Lakoff 1972: 172 — 198; Reklaitis 1974: 197—214; Ambrazas 1977: 149 et al.). Causative constructions with auxiliary verbs, which replace causative suffixes and/or vowel gradation, are a late Indo-European development.

These constructions clearly distinguish the causing and caused actions by expressing them with separate words, cf.: Jmdn singen machen — make someone sing K II 40, Jmdn weinen machen — make someone cry K II 40, binden lassen — have something tied K.

Periphrastic causatives arise in the absence of a morphological causative class. Of the three languages discussed here, only one had causatives of this type. In Latin, whose structure was predominantly synthetic, the analytic causative construction facere + inf. was used; it was atypical in the archaic and classical periods, though already used by Plautus and widely prevalent in the post-classical period. Another analytic causative construction, laxare + inf., also appeared in Late Latin (Абросимова 1954: 4—7).

These auxiliary verbs often have a lexical meaning denoting an active action, as in Latin facere (cf. Гончарова 1974: 136—155). In causative constructions, their original meaning was apparently ‘force to do’. It was precisely with this meaning that, during a certain period, the verbs leten, don, goren, mohen, later get, have were used in English to convey causation (Ярцева 1945: 60).

The clarity of the lexical meanings of auxiliary verbs often makes it possible to use them to express one particular type of causation and to specialize them accordingly.

The construction laxare + inf., which arose in Late Latin, came to express permissive causation, in which the causing action is ‘allow, not hinder’, and so on. This meaning of the construction is apparently based on the auxiliary verb’s

116

with the meanings of the verb lasare “to release, allow (to rest), ease, weaken, reduce, leave” and so on.

There is a close connection between the German causative construction lassen + inf. and the semantics of its auxiliary verb. Some causative constructions can be associated with the meaning of the verb lassen “(to) tell, order, urge,” for example: schmieden oder hämmern lassen — have (something) forged K, sich einen Anzug machen lassen — order, have a suit made K II 40, tragen lassen — have (something) carried K, etc. When compared with their Lithuanian translation equivalents and with the periphrasis of causation, this German construction could often be called curative rather than causative. Cases of passive causation may perhaps be related to the meaning “to allow” of the verb lassen, for example: tüngedinti caus. trägen sein lassen K, timpsodinti caus. lümmelhaft daliegen lassen K.

By contrast, constructions with machen express more active causation, for example: schweigen machen — silence (someone) K, Jmdn springen machen — make someone jump K II 40.

Periphrastic causatives correspond to the semantic types of synthetic causatives, but have shifted toward syntactic openness. However, German constructions corresponding to Lithuanian curatives (sich die Haare schneiden lassen K — have one’s hair cut) also imply a periphrasis of causation, since at the surface level they do not express the subject acting as intermediary, just like Lithuanian curatives.

3.2. Open causatives that explicitly express a causative situation

3.2.1. Causation is also expressed by complex subordinate result clauses, which at the surface level contain at most the components of the deep level. Like other types of subordinate clauses, they arose and developed within the separate Indo-European languages, but share many structural and semantic features.

The period when Russian complex sentences of consequence were formed has been established fairly precisely—the constructions with the consecutive conjunction так что arose only in the 12th–17th centuries. The conjunction of consequence is not found in earlier written sources (ИГРЯ 322, 332). The connection between the main and subordinate clauses was weak, since the main clause had no correlative words. Coordinating conjunctions were used, or the clauses of sentences of consequence were joined without conjunctions (ИГРЯ 323; Пометун 1971: 24—25).

The grammatical structure of sentences expressing consequence is likewise not clearly defined in early Lithuanian texts. They had the conjunctions teip jog, teip kad, jog, kad, teip idant, jeib, jeneg (Drotvinas 1967: 65—67). Most of these are no longer used to express consequence in contemporary Lithuanian, or are used only very rarely.

The late development of this type of sentence is also indicated by sentences in old Lithuanian and Russian texts in which the meanings of consequence and cause, and of consequence and purpose, are not clearly distinguished.

The academic grammar of Lithuanian does not even distinguish a type of complex sentence of consequence; these are considered descriptive sentences: “Descriptive sentences—

117

…often mark cause–result relations: the main clause presents a particular phenomenon, while the subordinate clause explains it by describing the consequences that follow from it” (LKG III 836).

3.2.2. (1) The development of subordinate clauses with a resultative meaning was encouraged by the need to express subject and object relations explicitly. In a result clause, the semantic subject of the caused situation is the syntactic subject of the subordinate clause, whereas in constructions with closed causatives and analytic causatives it occupies the object position, e.g.: It rained so much that I was soaked in no time – The rain soaked me through (DŽ 552).

(2) In the result clauses of all three languages, the intensity of the causing action (or “causing property”) can be expressed explicitly. Most clauses of this type have correlative words accompanying the conjunction, which emphasize intensity, strength, abundance, and the like: Lithuanian taip, toks, Latin ita, sic, ejusmodi, adeo, tantopere, talis, tantus, tam, hic, is (Соболевский 1948: 236), Russian так, столь, до того, до такой степени (ГРЯ II(2) 338). For example: Such fear seized me that I turned around and ran BV III 24 (LKŽ manuscript); Tantus in curia clamor factus est, ut populus concurreret Cic. Verr. 2,47 (Соболевский 1948: 236) “Such a great outcry arose in the Curia that people came running”; Общее волнение до такой степени сообщалось Буланину, что он даже позабыл о несчастном фонаре и о связанных с ним грядущих неприятностях. Куприн (ГРЯ II(2) 339) “The general excitement affected Bulanin so much that he even forgot about the wretched lantern and the future troubles associated with it.”

Intensity can also be indicated by the correlative word tai, used with net in the subordinate clause or with kad in the main clause: kad... tai net, tai kad... net, kad... tai... net, e.g.: I scolded that rascal so thoroughly that he even started to cry (Drotvinas 1967: 62).

(3) In result clauses, as in analytic causative constructions, the causing and caused action or state are expressed by separate words.

Causative verbs denote a single action, e.g., to wake, aimed at a single result, to awaken. Here the causing action and its result are very close in time or coincide: the result is either to be achieved immediately (I am putting out the candle), or has already been achieved (I watered the horse). Expressing the causing and caused actions with verbs from different roots makes it possible to describe actions or states that are more distant from each other. Often the causal subject neither anticipates the result (i.e. the consequence or caused situation) nor seeks to bring it about, e.g.: When it is time to spin, they chatter so many tales that, before long, they forget even to draw the wool with their hands K. Donel (LKŽ XV 671); But the rose smelled so sweet that, as you sniffed it, you had to forget all the troubles and cares of this world J. Balč (Balkevičius 1963: 357). The distance between the causing and caused situations is especially clear in the latter example, since the inanimate subject mentioned here cannot seek or anticipate any result. It is precisely to express such situations that result clauses are needed in the first place.

Describing the two causative actions with separate verbs makes it possible

118

to express explicitly the manner of causation, which closed causatives cannot express; compare: I sang such a beautiful song that everyone is laughing (LKG III 839) – I made everyone laugh.

Separating the actions makes it possible to specify clearly the temporal sequence of the causing and caused situations. The result occurs either after the action that caused it or at the same time. This is how tense is expressed in Lithuanian and Russian. In Latin, tenses in result clauses are governed by consecutio temporum, but the perfect subjunctive, pluperfect subjunctive, and periphrastic future occur rarely. The usual tenses in result clauses are present and imperfect subjunctive (Соболевский 1948: 236). The pluperfect subjunctive is very rare because its meaning conflicts with the notion of a result: a result can never precede the situation that causes it (Лосев 1965: 155).

The fairly frequent cases of non-observance of consecutio temporum are explained by the fact that the result is described in relation not to the tense of the sentence but to the time of the speech act, i.e. it depends on the writer’s or speaker’s intention (Соболевский 1948: 237).

Expressing actions with separate verbs makes it possible to state explicitly another important feature of causation—the reality of the causing and caused situations. Result clauses in Lithuanian are most characteristically expressed with indicative forms. The perception of the reality of the result is also discussed in Russian grammar (ГСРЛЯ 723).

In Latin subordinate result clauses, the subjunctive may be used, giving them a resemblance to purpose and causal clauses (with the conjunction cum) (Draeger 1878 II: 629). This feature is explained by the fact that the result expressed in such clauses is hypothetical (Боровский 1975: 183). However, although the result is most often expressed with a verb in the subjunctive, it is nevertheless perceived as real. For example: Epaminondas adeo fuit veritatis diligens, ut ne joco quidem mentiretur Nep. 15, 3, 1 (Соболевский 1948: 236) – “Epaminondas loved truth so much that he would not lie even in jest.” Here we see a certain mismatch between meaning and grammatical form: a real result is conveyed in the sentence by a mood that usually expresses unreality. This is so not only because the result is perceived as hypothetical, but apparently also because the subjunctive is used to express the relation between the clauses of a complex sentence.

4. Conclusions

4.1. In the development of Lithuanian and other Indo-European languages, a shift can be observed from closed to open means of expressing causativity. Over time, situations that can be described implicitly by morphological causatives are increasingly tending to be expressed explicitly.

4.2. The perception of causation at the beginning of the development of causatives and later was apparently, and has remained, connected with time. When the causing action and the caused state (or action) were expressed by a single verb form, that action and state were described as simultaneous or close in time. Causation was also understood and expressed in this way by open causatives. Only in the latter

119

case was it possible to express more varied results, rather than just one, as with morphological causatives (girdyti→gerti).

4.3. Closed causatives describe a situation that is just as complex as that described by open causatives. As open causatives develop, the deep level does not become more complex; it merely acquires explicit expression. Describing the actions of the causer and the causee with two separate verbs is characteristic of analytic causative constructions and complement result clauses, whereas shifting the object of causation into subject position is characteristic only of the latter.

By contrast, in the development of curatives, the implicit mode of expression was retained, but the deep level became more complex. Curatives arose from morphological causatives expressing indirect, nonphysical (i.e. verbal) causation, adapted to express a more complex situation. Since an intermediary subject and its action appear here at the deep level, curation can be interpreted as a specialized “double causation”.

4.4. The formation of open causatives was encouraged by the small size of the class of closed causatives. Analytic causative constructions arose in languages in which there were no (or almost no) causative verbs. Apparently, for this reason, Lithuanian has relatively few open causatives—result clauses expressing consequence—and they are not strictly formalized. Because the class of morphological causatives is large, they have been and remain somewhat less necessary in Lithuanian than in Latin or Russian.

The emergence of open causatives also somewhat broadened the very concept of causation. Historically, the primary meaning of closed causatives is “an active action by a living subject, directed precisely and strictly at an object.” Later, causatives began to mean (especially in result clauses): “a state (or action, more precisely—an effect) of any subject, directed imprecisely and weakly at an object.”

The panchronic semantic type of causation is “an active action of a subject, directed at a particular situation as an object (an object's action or change of state).”

The diachronic prototype of causation can be considered the opposition “active action / state” and linked to the hypothesis of an “active” Proto-Indo-European sentence structure.

Abbreviations¹

Highland dialects — Aukštaitian; Phrygian — Phrygian; Gothic — Gothic; Indo-European — Indo-European; Iranian — Iranian; Icelandic — Icelandic; Lithuanian — Lithuanian; French — French; ancient — ancient; Sanskrit — Sanskrit.

¹ Abbreviations not listed here are used as in LKŽ.

120

LITERATŪRA

1. Ambrazas V. Lietuvių kalbos istorinės sintaksės metodo problema. — Lietuvos TSR MA darbai. A serija. T. 4(61), 1977, p. 139—154.

2. Balkevičius J. Dabartinės lietuvių kalbos sintaksė. V., 1963.

3. Burrow T. The sanskrit language. London, 1959.

4. Cruse D.A. A note on English causatives. — Linguistic Inquiry. Vol. 3, N 4, 1972, p. 520—528.

5. Delbrück B. Vergleichende Syntax der indogermanischen Sprachen. Bd. 1, Strassburg, 1987.

6. Draeger A. Historische Syntax der lateinischen Sprache. Bd. 1,2, Leipzig, 1878.

7. Drotvinas L. Padarinio konstrukcijos lietuvių literatūrinėje kalboje. — Lietuvos TSR aukštųjų mokyklų mokslo darbai. Kalbotyra. T. 17, 1967, p. 61—80.

8. Ernout A. Morphologie historique du latin. Paris, 1927.

9. Jablonskis J. Rinktiniai raštai. T.1, V., 1957.

10. Jakaitienė E. Deverbatyviniai dabartinės lietuvių kalbos veiksmažodžiai su priesaga -inti. — Baltistica. T. 4(2), 1968, p. 221—229.

11. Jakaitienė E. Veiksmažodžių daryba (Priesagų vediniai). V., 1973.

12. Jakulienė A. Priežastinių ir parūpinamųjų veiksmažodžių skyrimo kriterijai. — Lietuvos TSR aukštųjų mokyklų mokslo darbai. Kalbotyra. T. 26(1), 1975, p. 78—81.

13. K-Kurschat Fr. Littauisch-deutsches Wörterbuch. Hale, 1883.

14. Kl, K II-Kurschat Fr. Deutsch-littauisches Wörterbuch. Hale, Bd.1, 1870; Bd.2, 1874.

15. Kurylowicz J. The inflectional categories of Indo-European. Heidelberg, 1964.

16. Lakoff R. Another Look at Drift. — In: Linguistic Change and Generative Theory. Bloomington, 1972, p. 172—198.

17. LKG — Lietuvių kalbos gramatika. V., T. 1, 1965; T. 2, 1971; T. 3, 1976.

18. LKM — Jakaitienė E., Laigonaitė A., Paulauskienė A. Lietuvių kalbos morfologija. V., 1976.

19. Paulauskienė A., Gramatinės lietuvių kalbos veiksmažodžio kategorijos. V., 1979.

20. Reklaitis J. K. Evidence for directed drift from Lithuanian. — Scandoslavica. T. 20, 1974, p. 197—214.

21. Safarewicz J. Historische lateinische Grammatik. Halle (Saale), 1969.

22. Savičiūtė G. Parūpinamieji veiksmažodžiai ir sakinio struktūra. — Kn.: Jaunųjų filologų darbai. T. 1, V., 1984, p. 73—78.

23. Savičiūtė G. Parūpinamųjų veiksmažodžių semantika. — Kn.: Lietuvių kalbotyros klausimai. T. 24, 1985, p. 236—251.

24. Savičiūtė G. Dėl lietuvių kalbos kauzatyvinių konstrukcijų semantikos. — Kn.: Lietuvių kalbotyros klausimai. T. 25, 1986, p. 157—167.

25. Savičiūtė G. Dėl lietuvių kalbos priežastinių veiksmažodžių reikšmės aprašymo. — Kn.: Jaunųjų filologų darbai. T. 2, V., 1986, p. 206—209.

26. Skardžius Pr. Lietuvių kalbos žodžių daryba. V., 1941.

27. Stang Chr. S. Vergleichende Grammatik der Baltischen Sprachen. Oslo-Bergen-Tromsø, 1966.

28. Абросимова Т. А. Способ выражения каузативности во французском языке. Автореф. дисс.... канд. филол. наук. Ленинград, 1954.

29. Боровский Я. М., Болдырев А. В. Учебник латинского языка. Москва, 1957.

30. Гончарова Н. А. Функционально-семантическая характеристика латинского глагола facere (на материале классической латыни). — Дисс.... канд. филол. наук. Киев, 1974.

31. ГРЯ — Грамматика русского языка. Т.2, ч. 1, 2, Москва, 1960.

32. ГСРЛЯ — Грамматика современного русского литературного языка. Москва, 1970.

121

33. Десницкая А. В. Каузативные глаголы. — Уч. зап. Ленинградского университета. Сер. филологических наук. Вып. 5, 1941, с. 138—165.

34. Елизаренкова Т. Я. Грамматика ведийского языка. Москва, 1982.

35. ИГРЯ — Историческая грамматика русского языка. Синтаксис. Сложное предложение. Москва, 1979.

36. Каукене А. А. История корневых глаголов с основой на -а в литовском языке (типы СеС, СёС, СаС, СоС, СуоС). Автореф. дисс.... д-ра филол. наук. Вильнюс, 1991.

37. Климов Г. А. О понятии языкового типа. — Вопросы языкознания. № 6, 1975, с. 21—28.

38. Кримицкайте С. А. Классификация глагольных оппозиций по переходности / непереходности в литовском языке. — В кн.: Проблемы синтаксиса и структуры слова. Москва, 1979, с. 42—53.

39. Лосев А. Ф. О законах сложного предложения в латинском языке. — В кн.: Статьи и исследования по языкознанию и классической филологии. Москва, 1965, с. 131—195.

40. Мейе А. Общеславянский язык. Москва, 1951.

41. Менчер Э. М. К вопросу о происхождении каузативных глаголов. — Ученые записки Томского университета. № 59, 1966, с. 73—79.

42. Миссера Н. А. Каузативные глаголы и каузативные конструкции в немецком языке. Автореф. дисс.... канд. филол. наук. Москва, 1952.

43. Недялков В. П. Каузативные конструкции в немецком языке. Аналитический каузатив. Ленинград, 1971.

44. Недялков В. П., Сильницкий Г. Г. Типология каузативных конструкций. — В кн.: Типология каузативных конструкций. Морфологический каузатив. Ленинград, 1969, с. 5—19.

45. Пометун Т. Е. К вопросу о формировании в русском литературном языке сложноподчиненных предложений с придаточным следствия. — Уч. зап. Таджикского университета. Сер. филологическая. Вып. 4, 1971, с. 23—27.

46. Савичюте Г. С. Предикаты цели и предикаты каузации (на материале флективных индоевропейских языков). Автореф. дисс.... канд. филол. наук. Москва, 1980.

47. Савичюте Г. С. Предикаты цели и предикаты каузации. — Изв. АН СССР. Сер. лит-ры и яз. Т. 39, № 6, 1980, с. 539—548.

48. Серебренников В. А. Вероятностные обоснования в компаративистике. Москва, 1974.

49. СГГЯ — Сравнительная грамматика германских языков. Т. 4. Москва, 1866.

50. Соболевский С. И. Грамматика латинского языка. Т. 1, Москва, 1948.

51. Степанов Ю. С. Балто-славянское и индоевропейское предложение (проблемы реконструкции). — Baltistica. 24(2), 1988, с. 116—130.

52. Тронский И. М. Историческая грамматика латинского языка. Москва, 1960.

53. Якайтене Э. Суффиксальные глаголы современного литовского языка. Автореф. дисс.... канд. филол. наук. Вильнюс, 1968.

54. Ярцева В. Н. Категория каузативности в английском языке. — В кн.: Тезисы докладов по секции филологических наук. Ленинград, 1945, с. 60—62.

122

LIETUVIŲ IR KITŲ INDOEUROPIEČIŲ KALBŲ KAUZATYVŲ RAIDOS POLINKIAI

Reziumė

Ide. kauzatyvai rutuliojosi nuo uždarųjų prie atvirųjų ekspl icitinės raiškos link: (1) išskaidant kauzacijos veiksmus, išreiškiant juos atskirais veiksmažodžiais (analitiniuose kauzatyvuose ir padarinio sakiniuose), (2) kauzacijos objektą perkeliant į subjekto poziciją (padarinio sakiniuose).

Kauzacijos panchroninis semantinis tipas — ”Subjekto aktyvus veiksmas, nukreiptas į tam tikrą situaciją kaip objektą (objekto veiksmą ar būsenos kitimą)”.

Kauzacijos diachroninis prototipas — opozicija ”aktyvus veiksmas / būsena” kaip proide. aktyvo / pasyvo opozicijos atspindys.

ON THE TRENDS OF DEVELOPMENT OF LITHUANIAN AND INDO-EUROPEAN CAUSATIVE CONSTRUCTIONS

Summary

Indo-European close causatives are developing into open causatives, towards the explicit expression: (1) by decomposing the actions of causation, expressing them by separate verbs (in the analytic causative constructions and sentences of result), (2) by moving the object of causation to the subject position (in the sentences of result).

Semantic panchronistic type of causation is defined as ”Active purposive action of subject, strictly directed to object (its action or change of state).

Diachronic prototype of causation — ”active action / state” can be regarded as a reflection of active / passive opposition in Proto-Indo-European.

123