Curatives in the Old-Lithuanian Bible translations

The subject of the paper is curatives, a special type of causative verbs in Lithuanian, in the Bible translations by Johannes Bretke (1579–1590), Samuel Boguslaw Chylinski (1660), Samuel Bythner (1701), Philipp Ruhig (1727) and Juozapas Arnulfas Giedraitis (1816). Curatives are a special type of causation implementing the causee-suppressing valency (CSP) pattern. In the analyzed texts, one can notice several morphological, semantic or syntactic features coinciding with contemporary curative constructions. However, in addition to the morphology and semantics of curatives, the article discusses also some rare and interesting cases that can be interpreted as expressing the causee and includes remarks on the lexicalization processes.


Introduction
The article aims to characterize a special group of causative verbs called curatives in Lithuanian translations of the Bible produced from the 16th to the 19th century. The analysis includes examples from the texts by Johannes Bretke (1579-1590), Samuel Boguslaw Chylinski (1660), Samuel Bythner (1701), Philipp Ruhig 1 (1727) and Juozapas Arnulfas Giedraitis (1816). In addition, the selected fragments were compared with the contemporary translation by Rubšys and Kavaliauskas (1998). The article will discuss the morphology and syntax of the construction with particular emphasis on how the curatives affect the valence structure of the sentence and the possible ways of expressing the causee. Furthermore, the issue of the semantics of curative constructions and traces of the lexicalization process will be discussed. It should be noted that curatives are peculiar constructions and it is often difficult to distinguish them from the regular causative. Thus, their identification based solely on one criterion: morphological, semantic or syntactic, might prove problematic.
The base of the prototypical causative construction is an intransitive verb whose main argument (S) 2 is devoid of agentivity, volition and control over the event in question. During causation, a volitional agent (A) is introduced into the sentence structure and claims the semantic role of the causer. The primary subject of the base sentence becomes the patient (P) of the new causative construction: it claims the role of the causee and occupies the syntactic role of the direct object. In this way, the prototypical intransitive sentence becomes transformed into a prototypical transitive sentence. Such a transformation can be considered prototypical because it is both syntactically and semantically complete. Additionally, derivation from intransitive verbs is possible in any language that has morphological tools to express causation; in case of other verb classes, however, it is subject to additional restrictions (Kittilä 2009, 9). This situation is prototypical also for Old-Lithuanian, where the morphological causatives are derived with the suffixes -(d)in-and -(d)y-. Below I present the prototypical causativization of an intransitive verb, where a. is an intransitive sentence, b. a causative sentence, and c. a valency pattern: ( As can be seen, an additional argument A, the causer, is introduced into the structure of the causative sentence (1b) and it claims the position of the new subject. The original argument S, the causee, is moved to the position of the direct object.
2 The morphology and syntax of Curatives Lithuanian seems to confirm the typological data from other languages and regularly creates morphological causatives based on intransitive verbs, while the derivation of causatives from a transitive base is limited by a number of restrictions. However, a small set of causatives can be derived from a transitive base (mainly ingestive verbs, but not only these; Nedjalkov & Sil'nickij 1973, 7-8;Dixon & Aikhenvald 2000, 13;Holvoet & Nau 2015, 26). In this context, the now-disappearing group of verbs called curatives is a particularly interesting case. Curatives describe the situation of indirect causation, in which the causer is not a direct cause of action but they only order someone to perform the action for them (Savičiūtė 1985, 236;Toops 1989, 249;Arkadiev & Pakerys 2015, 55-56;81-82).
As for the morphological criteria, curatives are usually formed on the basis of transitive verbs, with the morphology of causative verbs, mainly with the suffix -(d)in-. However, researchers often find that distinguishing curatives from causatives on the grounds of semantics and syntax might prove rather problematic. Below I present examples of curative constructions where a. is a transitive sentence, b. a curative sentence and c. a valence scheme: (2) a.  (Savičiūtė 1985, 242;Toops 1989, 260-275;Arkadiev & Pakerys 2015, 56). Interestingly, unlike causatives, they do not change the valence structure of the base sentence but reorganize it ). An external causer, fulfilling the syntactic role of the subject, is introduced into the causative sentence structure (2b). While the original argument P retains its function of the direct object, the original argument A is omitted and can be interpreted as a proper causee of the curative construction because the causer directly influences its actions.
At first glance, sentence (2b) might look like an ordinary causative sentence, where 'Pilate' is the causer and 'Jesus' is the causee. The situation, however, is not so obvious. 'Jesus' performs the semantic role of a Patient, but the actual causee of the sentence is suppressed. Based on pragmatic knowledge, we can assume that 'Pilate' did not flog 'Jesus' personally; he had to entrust the execution of the sentence to his subordinates. Therefore, we may consider 'Pilate' to be a causer, but the actual causee (the presumed executor of the judgment, the original argument A) is excluded from the syntactic structure of the sentence. Therefore, the curatives can be described in the following way: "the causer orders the causee to perform the action x" or, in other words, the curatives can be conventionally called "double causatives" (Savičiūtė 1985, 250, translation mine).
Apart from the more conventional curatives, the discussed texts include a number of sentences worth analysing. Let us consider the following example: (3) Old Lithuanian (BrB Song of Songs 3.9) Karalius Salamonas pa-dari-din-oia ſaw weßimm{ạ} king.nom.sg pn.nom.sg pfx-make-caus-pst.3 rpo.dat chariot.acc.sg iſch medʒio Libano of wood.gen.sg pn.gen.sg 'King Solomon made himself a chariot of the wood of Lebanon.' The sentence can be understood as follows: "Solomon commissioned an undefined causee to make a chariot for himself." Therefore, we can observe the coincidence of the Initiator of the action (the causer) and the Beneficiary. This sentence is an example of an autobenefactive construction (Geniušienė 1987;Kemmer 1993;Kulikov 2013) with a heavy marker sau. This marker is heavy both in the phonetic and syntactic sense and it is called "heavy" because it occupies a separate syntactic place in the sentence structure, contrary to the light middle voice marker -si-. By comparing this passage with a modern translation, one can see a shift from the domain of reflexivity to the middle voice: (4) Lithuanian (RKB Song of Songs 3.9) Karalius Saliamonas pa-si-dirb-din-o nešamą king.nom.sg pn.nom.sg pfx-rfl-make-caus-pst.3 mobile.acc.sg sostą iš Libano medžių. throne of pn.gen.sg wood.gen.pl 'King Solomon made himself a chariot of the wood of Lebanon.' As shown in example (4), the heavy marker sau has been replaced by the light -si-marker, thus moving from the domain of the natural reflexivity to the middle voice. Due to the removal of the one argument from the argument structure of the sentence, the roles of the causer and the beneficiary equate with each other. non-valency increasing causatives, called covert causativization (Nedjalkov & Sil'nickij 1969, 50;Kittilä 2009, 75-79;Arkadiev & Pakerys 2015, 83).
According to Kittilä (2009, 78), covert causativization is usually based on ditransitive verbs, which accommodate three participants in the event, and the fourth argument is beyond the limitations of many languages. However, the situation is different in Lithuanian. Although in most cases the causee is not expressed at all, there are some cases in modern Lithuanian where the causee might be interpreted as expressed indirectly by several tactics, including locative phrases (5), prepositional constructions with pas (6) or per (7), or the instrumental (8)  However, according to Arkadiev & Pakerys (2015, 85), the only instance of the genuine expression of the causee is sentence (6) with the preposition pas.
In the analyzed texts, the number of constructions that can be interpreted as expressing the causee is limited. As in the modern language, they do not explicitly express the causee, but background it. The first such construction is similar to example (5)  We can conclude that if Nebuchadnezzar ordered to bring people to Babylon, the Babylonian soldiers are likely to be the causee. Unfortunately, as in the case of sentence (5), interpreting a locative phrase as the actual causee, is very questionable. It should rather be said that the sentence refers to the place where the activity was performed or, more precisely, in the case of the illative, to the destination where the activity will be concluded.
In the context of the expression of the causee discussed above, the following sentences seem peculiar: Such constructions can only be found in the translations by Bretke and Ruhig, and are always combined with the verbs siųsti 'send' and sakydinti 'order to tell'. The example mentioned above may raise doubts as to whether sakydinti can be interpreted as a curative verb at all. However, let us recreate the deep structure of the curative sentence (10) as proposed by Savičiūtė (1985, 237): Causer ('Joab') + P1 ('sent') → Causee ('messengers') + P2 ('so they say') → Object ('to David') + P3 ('that Joab have taken the city') As evident above, although the performer is not explicitly expressed in the sentence structure but only by an ellipsis to the previous sentence, we can see that the causees are 'messengers' (10) and then 'disciples' (11). Therefore, sakydinti in the above sentences can be understood as 'to tell one someone else's words', 'to tell one something that another person asked us to'. A curative construction of this type assumes the causation of a complex predicate. In this case, we deal with the mixed curative: apart from the morphological coding in the form of the curative verb sakydinti, it includes also the analytical coding using the verb siųsti 'to send'. In the modern translation of the Bible, examples (10) and (11)  The sentence can be understood as follows:

Causer ('servants') + P1 ('tell') → Causee ('through men') + P2 ('so they tell') → Object ('to David') + P3 ('what happened')
In other words: 'The servants made the men pass information to David.' Assuming that the base sentence would be (15a), the sentence pattern should look as follows (15b) (7) where the person expressed by the prepositional phrase did not send the letter himself, but only delivered it (Arkadiev & Pakerys 2015, 85), 'the men' in sentence (14) may be interpreted as direct 'messengers' who passed on information to 'David' on behalf of the agent, 'servants'.

Semantic shifts and lexicalization
As has been already mentioned, Lithuanian curatives are a unique construction because they suppress the causee in the sentence structure. However, as reported in the literature (Naktinienė 2011, 158;Arkadiev & Pakerys 2015, 87-88), there are some deviations from what one could call the prototypical curative.
In addition to the prototypical meaning presented in example (2), the analysed texts include interesting and rare constructions meaning: "the causee performs the action x on behalf of the causer": for the LORD hath delivered them into the hand of Israel. And Jonathan climbed up upon his hands and upon his feet, and his armourbearer after him: and they fell before Jonathan;] and his armourbearer slew after him.' c. V TR <A i :Sbj; P j :DO> → V CAUS <Causer: Ø; Causee i :Sbj; P j :DO> The difference between sentences (2) and (16) lies in the degree of the agent's control over the described situation. In sentence (2), 'Pilate', who ordered his soldiers (not expressed in the syntactic structure of the sentence causee) to flog 'Jesus', was both the causer and the subject of the sentence. Sentence (16) presents the opposite situation. It might be interpreted that the causee, 'armourbearer', kills enemies on the orders of the causer, 'Jonathan'. However, in sentence (16), the causee takes the role of the argument A and the causer is suppressed. Therefore, here we can talk about a rare type of valency pattern -let us call it the causer-suppressing pattern. In the case of sentence (16), the syntactic structure is not reorganized at all, and the causer, 'Jonathan', is backgrounded only by the ellipsis to the previous sentence.
In example (16), the action is performed by the agent, the causee. However, since the curative verb is used, it is implied that the driving force of the action must have been some external causer. However, due to the already very flexible usage in the 17th century, the verb užmušdinti 'to kill by order' could be used both in the usual prototypical curative context based on the CSP model (17) and as an ordinary transitive verb with a controlling agent (18) one. nom.pl.m another.acc.pl.m pfx-kill-caus-irr.3 '…and power was given to him that sat thereon to take peace from the earth, and that they should kill one another.' Of course it is worth to keep in mind already mentioned flexibility of the curatives in the 17 th century. One could argue that užmušdinti in (16) is used in the same manner as (18). This fact only proves the abovementioned difficulties in the interpretation of the curatives.
As Arkadiev & Pakerys (2015, 90) noted, the semantic change may be caused by the fact that the base verb may be relatively polysemic. Therefore, the addition of a causative suffix may result in the new, specialized meaning. This assumption can be confirmed by the example of the verbs nukirsti ('to cut') (19)  As can be seen above, Bretke's translation includes two variants, while the translations by Bythner and Giedraitis include only a transitive verb variant, which would indicate that Herod carried out the execution himself. Interestingly, Chylinski and Ruhig also used the curative variant, and eventually, the curative variant prevailed as the modern translation also uses the curative nukirsdinti, indicating that Herod was the one ordering the beheading of John the Baptist. Thus, this version has lexicalized and assumed a specialized meaning.
At this point, it is worth adding that translations by Bythner and Giedraitis do not include curatives at all, allegedly, due to the sources of both translations. Although there has been no detailed research on the sources of Bythner's translation of the New Testament, it is assumed that he translated from Greek, Polish and German. The history of the sources used by Giedraitis also has not been fully explored, however, according to the historical data, the New Testament was published by the Protestant Bible Society in Vilnius and had to be compared with Bythner's text 3 . Let us compare some longer passages of the Gospel of Luke with the text of the Gdańsk Bible (1632) his servants. However, based on the derivational semantics of their base words, and by comparing the texts with the Gdańsk Bible, it should be assumed that these constructions are devoid of the semantic element of curativity and are lexicalized to some extent. Curative constructions are also found in Latvian, where many of the curatives have developed a new meaning by abandoning the causative element (šūdināt ← šūt 'sew'; darināt ← darīt 'make/do') (Holvoet and Nau 2015, 27). A similar tendency can be observed in Lithuanian, therefore verbs suvadinti and užsodinti have assumed a specialized use and, at least in the idiodialects of the authors, they have moved from the word formation to the word-stock.

Conclusions
Curative constructions are a rare type of causativization derived from transitive verbs and are formed by adding the suffix -din-to the base verb. From the point of view of syntax, the characteristic feature of curatives is the fact that, unlike proper causatives, they do not increase the valency of a sentence, but only reorganize it. A new argument, the causer, is introduced into the sentence structure, but the original argument P, the causee, is removed. Interestingly, in the analyzed texts, there are a number of examples where the causee can be considered as expressed. Some of them, such as the locative expressions in (9), can be considered questionable because they indicate the place of action rather than the actual causee. However, the situation becomes clearer in examples such as (10) and (11), where the causee can be identified to some extent by the pronoun anaphora. This may prove that although the causee does not occupy a syntactic place in the sentence, it is still backgrounded in the semantics. An interesting example in this context is sentence (14), where the causee takes a separate syntactic place and is introduced into the sentence with the preposition per. Although such a construction can be seen only in Bretke's translation, it may denote the process of the normalization of curative constructions created according to the CSP pattern.
The processes of normalization and lexicalization can also be seen in other translations. Constructions cited in Part 4 show that curatives, in addition to meaning "the causer orders the causee to perform the action x", can take on specialized functions such as "the causee performs the action x on behalf of the causer". Over time, verbs that have the curative morphology may take on