Iš optatyvo istorijos

LIETUVIŲ KALBOTYROS KLAUSIMAI, IV LIETUVOS TSR MOKSLŲ AKADEMIJA THE HISTORY OF THE OPTATIVE J. KAZLAUSKAS § 1. In Lithuanian dialects, optative forms vary greatly. No grammatical category has as many variants as the optative. The following optative forms are now more widely used in Lithuanian dialects: singular 1st person būčiau, būčia, būčio, būč, būčiu, būtųu (būtau), būčiobe, būčiap (būtiap), būčiobi, būtum, būčium; singular 2nd person būtum, būtumei, būtumi, būtųi (būtai), būtumbe, būtumbei; plural 1st person būtum(e), būtumėm(e), būtumim, būtumiam, būtūm (būtūme), būtum, būtumbe, sukutumbėmės; plural 2nd person būtut(e), būtumėt(e), būtumit, būtųt (būtūte), būtūl, būtūtbe, sukutumbėtės; 3rd person būtų, būtų, butu, būt, būtum, būtun, būčube, būčubi. Adding the optative forms found in 16th–17th-century Lithuanian writings makes the picture even more varied: singular 1st person būčiau, būčia, būčio, būtumobiau; 2nd person būtumbei, būtumbi, būtumei, būtum, būtumb; plural 1st person būtumbim(e), būtum(e), būtumim(e), būtumb; 2nd person būtumbit(e), būtumit(e), būtumbo, būtumbet; 3rd person būtų (būtun). § 2. Such diversity of optative forms has long attracted the attention of researchers. Interest in the Baltic optative began, one might say, with the very emergence of the comparative-historical method. It was written about by A. Schleicheris1, F. Bopas2, A. Becenbergeris3, J. Šmitas4, O. Vydemanas5, V. Poržežinskis6 (whose analysis is based on rich dialectal and old 1 See A. Schleicher, Handbuch der litauischen Sprache, 228 ff., Prag, 1856; Compendium der vergleichenden Grammatik der indogermanischen Sprachen, 841, Weimar, 1866. 2 See F. Bopp, Vergleichende Grammatik..., 3³, § 685 ff., 1861. 3 See A. Bezenberger, Beiträge zur Geschichte der litauischen Sprache..., 214 ff., Göttingen, 1877; Lettische Dialekt-Studien, 72 ff., Göttingen, 1885; Litauische Forschungen, 16 ff., 39, Göttingen, 1882; Über die Sprache der preussischen Letten, 97 ff., Göttingen, 1888; BB 26, 175 ff. 4 See KZ, 24, 305, 1879. 5 See O. Wiedemann, Das litauische Präteritum, 184 ff., Straßburg, 1891; Handbuch der litauischen Sprache, 111, Straßburg, 1897. 6 See B. K. Поржезинский, Къ исторіи формъ спряженія въ балтійскихъ языкахъ, 56 ff., Москва, 1901. 73 (material from written monuments), E. Berneker⁷, J. Zubatis⁸, K. Brugmann⁹, F. Specht¹⁰, J. Endzelin¹¹, H. Pedersen¹², Ch. Stang¹³, V. Pisani¹⁴, J. Otrębski¹⁵. The formation of individual optative forms is often explained in different ways. There are almost as many differing opinions on this question as there are forms themselves. Time has rendered some of them obsolete; others are insufficiently supported by Lithuanian-language evidence. Almost all the authors had too little dialectological material on Lithuanian and knew little about the distribution of these forms in the dialects. § 3. All the scholars mentioned above, with the exception of F. Bopp and V. Pisani, agree that the Lithuanian optative is formed from the supine and the auxiliary verb -bime. E. Berneker alone, quite rightly unconvinced by V. Poržežinskis’s suggestion that a single supine could have been used with optative meaning for all three persons, conjectures that a form of the imperative with *-tōd (Sanskrit -tād, Greek -τω, Latin -tod, -to) was incorporated into the optative system in the Baltic languages; its remnant is Latvian 3rd person suktuo-s, Lithuanian suktu (with short -u)¹⁶. F. Bopp¹⁷ and V. Pisani¹⁸ believe that the first component of the Lithuanian-Latvian optative was the Indo-European imperative with -tu (cf. Indo-Iranian -tu, Hittite -du), which could be used with potential meaning in Old Indian. Different opinions have also been expressed about the second component of the optative (-bime, etc.). Most scholars (F. Bopp, A. Schleicher, J. Schmidt, V. Poržežinskis, E. Berneker, A. Bezenberger, O. Wiedemann, J. Zubaty) consider it to be the optative of the verb to be. K. Brugmann states unequivocally that this form is a preterite. Ch. Stang, too, tends to regard it as a preterite. J. Endzelin takes an alternative view of its origin, not ruling out the second opinion, although that opinion does not explain the presence of two stems (-bei and -bime) in a single paradigm¹⁹. The origin of the 1st person singular forms būčia, -čio, -čiau poses especially many difficulties for researchers. Despite considerable efforts, the origin of these forms has not yet been explained. No hypothesis has been proposed that can be considered certain. § 4. Imperative or supine? This question might not need to be raised were it not for V. Pisani’s recent discussion of it. He supports the view put forward by F. Bopp ⁷ See E. Berneker, AfslPh, 25, 485, 1903 (a review of the work by V. Poržežinskis mentioned in footnote 6). ⁸ See J. Zubaty, IFA, 16, 55 ff., 1904. (A review of the work by V. Poržežinskis mentioned in footnote 6). ⁹ See K. Brugmann, IF, 30, 348 ff., 1912; Grundriss der vergleichenden Grammatik der indogermanischen Sprachen, 2², 3, 513 ff. Strassburg, 1916 (hereafter: Grundriss). ¹⁰ See F. Specht, Litauische Mundarten, 2, Leipzig, 1922 (see below under Optativ). ¹¹ See J. Endzelin, Lettische Grammatik, § 695 ff., Riga, 1922; FBR, 15, 11, 1935; FBR, 17, 166, 1937; Latviešu valodas gramātika, § 695 ff., Rīga, 1951 (hereafter: Gramatika). ¹² See H. Pedersen, Études lituaniennes, 56, København, 1933. ¹³ See Ch. Stang, Das slavische und baltische Verbum, 249 ff., Oslo, 1942 (hereafter: Verbum); Die litauische Konjunktion jeib und der lit.-lett. Optativ (Norsk Tidsskrift for Sprogvidenskap, 18, 348 ff.). ¹⁴ See V. Pisani, L'ottativo (congiuntivo) baltico e il trattamento di ō in sillaba finale in gotico e baltico. — “Studi Baltici,” 9, 34 ff., 1952. ¹⁵ See J. Otrębski, Gramatyka języka litewskiego, 3, 230 ff., Warszawa, 1956. ¹⁶ See AfslPh, 25, 486. ¹⁷ See F. Bopp, cited work, § 685. ¹⁸ See V. Pisani, cited article, 40. ¹⁹ See J. Endzelins, Gramatika, § 695. 74 idea, and throughout his article he attempts to substantiate it with facts from Lithuanian and Latvian. The basis of this idea is an incorrect interpretation of the Lithuanian and Latvian singular third-person forms with -tu (short -u). Moreover, the formation of many Lithuanian and Latvian dialectal optative forms is unclear. Claims that fail to grasp its essence have been passed from book to book, from article to article, down to our own day. The short -tu arose here through the particular shortening of optative forms. Consequently, E. Berneker’s conjecture that Lithuanian -tu and Latvian -tuos derive from the imperative *-tōd is also contrary to fact. Both Lithuanian -tuos (turietō̜s ~ turētuos Vieksniai, suktu̜os Šakyna) and Latvian -tuos arose by analogy with other reflexive verb forms, since alongside the simple forms ending in -u, the reflexive forms always end in -uos (cf. Lithuanian suku: sukuos, Latvian velku: velkuos)20. As soon as the supine ending -ty shortened to -tu, a reflexive form ending in -tuos could readily arise here as well, by analogy with the reflexive forms of other persons. Both these theories are already a priori implausible simply because no imperatives with *-tu or *-tōd have survived in either the Baltic languages or the Slavic and Germanic languages related to them. Moreover, there is no Lithuanian dialect that would give us even the slightest reason to doubt that its optative forms were indeed formed on the basis of the supine. Since the time of V. Porzeziński, a widespread notion has obstructed the correct understanding of optative formation: that the supine could originally have been used as the optative form for all numbers and persons. E. Berneker quite rightly questioned this, since it is difficult even to imagine such a situation. Ch. Stang likewise believes that originally the pure supine served as the form for all optative persons and numbers21. J. Endzelynas also supports V. Porzeziński’s idea22. There is no need to rely on this idea any further. § 5. In almost all Lithuanian linguistic monuments of the sixteenth and seventeenth centuries, we find the optative with -b-. Forms with -b- have also survived in old Latvian written monuments and dialects. It may therefore be assumed that during the common Lithuanian-Latvian period the optative was formed from the supine and an auxiliary verb whose stem began with the consonant b-. Over time, this auxiliary verb attached itself to the preceding supine, giving rise to synthetic optative forms. Eventually, the auxiliary verb, as a separate inflected verb form, disappeared and, having fused with the supine, became a postposition, which lost its morphological optative meaning and was destined to disappear23. The third-person form (suktu), which had lacked a postposition since ancient times, also hastened the leveling of the postposition’s morphological meaning (see § 16). The optative inflection system and its considerably shorter third-person form also encouraged the postposition’s disappearance. In the Baltic languages, the third-person form is the center of the verb. On its basis, both the Prussians and the Lithuanians and Latvians have created numerous innovations. The optative postposition disappeared more quickly in long forms of polysyllabic verbs (pasivaiškėtuombime and others). In the eastern Dzūkian dialects, where the optative with b- is used, its second component has survived particularly rarely in polysyllabic verbs. 20 For Latvian suktuos, cf. J. E n d z e l i n s, Gramatika, § 705. 21 Ch. S t a n g, Verbum, 251. 22 See J. E n d z e l i n s, Gramatika, § 696. 23 For the disappearance of elements that have lost their morphological meaning, see K. B ū g a, Rinktiniai raštai, 1, 578 ff., edited by Z. Zinkevičius, Vilnius, 1958. K. Būga would probably have reached a similar conclusion had he completed the article cited here. 75 As the optative postposition was disappearing, the following optative conjugation system must have developed in separate Lithuanian and Latvian dialects: First type Second type sg. 1st person suktumbiau (where suktum occurred, it had formed), 2nd person suktumbei; suktum, 3rd person suktų, pl. 1st person suktumbime, suktum, 2nd person suktumboite. suktum. Alongside the new short forms, the old long forms could undoubtedly still have been used for some time. Traces of the second type have survived in Lithuanian; cf.: sg. 1st person suktum (Biržai), būčium (Viekšniai, a contamination of būčiu and būtum), 2nd person suktum (widespread), 3rd person suktų, pl. 1st person suktum (widespread), 2nd person suktum24, nebutum-te ChB 161; žynotum-te ChB 2620. In Latvian dialects the second optative type is fairly widespread25, but here the form with -tum was introduced even for the 3rd person because -tum became indifferent to person. This is entirely understandable. This type must also have been widespread in Samogitian and East Prussian Aukštaitian dialects; cf. the 3rd-person forms neštum < neštum and neštõm < neštumi26, used across almost the entire Samogitian-speaking area, and East Prussian Aukštaitian paskubin-tum, darytum27. In Latvian dialects, alongside suktum, suktu is also used for all three persons. The same form suktu for all optative persons arose through the use of suktum with the meaning of all three persons. For alongside the new sg. 3rd-person form suktum, the old suku could also be used in the sg. 3rd person; cf. East Prussian Aukštaitian darytum || darytu || daryt28, and Samogitian nešto < neštu < neštų ir neštūm < neš-tum. By analogy with suktum, the sg. 3rd-person form suktu too began to be used for all persons; this was further facilitated by the fact that suktu was very well suited to the sg. 1st person, corresponding to the sg. 1st person of all verbs29. In Latvian dialects, alongside the sg. 1st person in -tu, a sg. 2nd person in -ti is also used, newly formed on the model of the sg. 2nd person of o- and a-stem verbs30. § 6. The second optative type (see § 5) could have been used in all Lithuanian and Latvian dialects: in some earlier, in others later. Only Latvian was 24 See A. Schleicher, Handbuch, 229. 25 See J. Endzelīns, Gramatika, § 697. 26 neštumi arose by adding the i-stem verb ending -i to neštum (by analogy with the other persons, whose forms coincided with those of i-stem verbs); cf. also the reflexive form neštumis>neštomes, which shifted to the ia-stem neštumias. 27 A. Bezzenberger, Litauische Forschungen, 39 and footnote 1. 28 Ibid. 29 J. Endzelīnas is mistaken in explaining suktu differently, relying on the previously mentioned view of V. Por-žezinskis; see J. Endzelīns, Gramatika, § 698. 30 J. Endzelīnas is mistaken in attempting to connect this -ti with the late Lithuanian -tai. The second assumption of J. Endzelīnas is valuable; see Gramatika, 896. Latvian sg. 2nd-person -tib may be a contamination of -tub and -ti, as J. Endzelīnas (896) thinks, but more likely it was formed from -ti with -b- added later; cf. the 2nd-person plural būtūtėbe of the Eastern Dzūkian dialects and p. 82 of this article. 76 and is more tolerant of identical forms for different persons. Lithuanian dialects, by contrast, could not tolerate forms indifferent to person (except the sg. 2nd person with -tum). Therefore, the second type was remodeled: sg. 2nd person suktum, 3rd person suktų, pl. 1st person suktum, 2nd person suktut. The old postpositional optative forms, naturally, did not disappear immediately once new short forms arose. This situation, in which the old forms were still used alongside the new optative forms, is vividly reflected in the 1653 Kėdainiai publications; cf. “Knigos No-bažnistes” sg. 2nd person -tumbei (Nepażieystum bey 80) and -tum (noretum 68), “Sumos” pl. 1st person -tum(e)31 (apturetume 13) and -tumbim(e) (garbintumbim 8). A similar situation has also persisted in the Eastern Dzūkian dialects; cf.: sg. 2nd person sùktum || suktum̃ and sùktumbėis, pl. 1st person sùktum || suktum̃ and sùktumbėmės, 2nd person sùktūt || suktū̃ (newly formed) and sùktumbėtės32. As the postposition disappeared, new forms lacking person marking arose, and since Lithuanian could not tolerate this situation, yet another new optative type developed through blending the new and old long forms used side by side. In the 1653 “Maldose Krikščioniskose” and “Sumoje” we already find the sg. 2nd person -tumei (būtumei 2), formed by blending -tum and -tumbei; “Sumoje” also uses the new pl. 1st-person form tumim(e) (galetumime 9), which occurs no less frequently than -tum(e) or -tumbim(e). In K. Širvydas’s writings, the sg. 2nd person occurs only with -tumei, the pl. 1st person only with -tum(e) and -tumim(e), and the pl. 2nd person with -tumit(e); but these forms clearly show that if the optative with -bei, -bime, -bite. was not used during Širvydas’s lifetime (Širvydas may have avoided forms with -bime, -bite for reasons of standardization), then it was used somewhat earlier in Širvydas’s dialect. In the aforementioned Eastern Dzūkian dialects, too, forms arising through blending are used; cf.: sg. 2nd person sùktumės alongside sùktum̃ and sùktumbėis, pl. 1st person sùktumėnės alongside sùktum̃ and sùktumbėnės, 2nd person sùktumėnės alongside sùktumbėtės. Through blending, the 1st-person plural form with -tumim, found in Samogitian dounininkai dialects, also arose, as did the fairly widespread 2nd-person plural form with -tumi in those same dialects. In Samogitian dounininkai dialects, the 1st-person plural form with -tumiām is much more widespread than the form with -tumim and has an ia-stem, like i-stem verbs that have shifted there to ia-stems (cf. tyliam). In the Samogitian dialect, by analogy with i-stem verbs, the 1st-person singular optative form may also have developed; cf. būčiau (Viekšniai and elsewhere). 31 -e in the pl. 1st person arose by analogy with einame, nešėme and so on. 32 J. Kardelytė provided the author with a considerable number of facts from the eastern Dzūkian dialects. Some were taken from diploma theses by students of Vilnius University; the author collected others himself in the environs of Palūšė. 77 As a result of the crossbreeding of saktum and saktumbei, the 2nd person singular form suktumei arose; alongside saktum, it is widely used by the Žemaitian ūnininkai and almost all Aukštaitians, except the Dzūks. Through the crossbreeding of saktum and saktumbei, the 2nd person singular form suktumi33 (reflexive suktumės) arose as well; it is widespread in the Žemaitian donininkai dialects. § 7. If the 1st person plural form with -umime arose through crossbreeding, it is quite reasonable to ask whether the form -tumėm(e), too, might have arisen through the crossbreeding of -tumbėme and -vum. The 1st person plural form with -tumėm(e) is widespread throughout almost the entire territory of the Lithuanian SSR, except for a large part of the Žemaitian donininkai, the Dzūks, and the areas adjacent to Dzūkija. In some places it is the only form used; elsewhere, alongside it, a form with -tum is used (the zanavykai kapsai34, Dotnuva, Palomenė, Baisogala, Alunta, Kupiškis, Lyduokiai, Raguva, Skapiškis, Svėdasai, Troškūnai, Gerviniai, etc.). -tumėm(e) could also be a new formation, created on the model of the 1st person plural past frequentative form of ē-stem verbs, under the influence of the 2nd person singular -tumei, whose ending -ei coincided with the 2nd person past-tense form35 (cf. § 10). But there are Lithuanian dialects in which the 2nd person singular has -tum, while the 1st person plural has -tumėm(e) (Alunta, Dusetos, Pašvitinys, Pandėlys, Palomenė, Biržai, etc.). Of course, the form with -tumei may have disappeared there, just as the 1st person plural form with -tum has disappeared in some dialects; but it may also never have been formed. Here, most likely, the form -tumėme arose through the crossbreeding of the forms -tum and -tumbėme. That -bē- was also widespread in the plural is shown by the Latvian 1st person plural forms celtubem, celtubemes and the 2nd person plural form butubet36. Similar optative forms are used in the eastern Dzūkian dialects: 1st person singular sùktóbė, sùktóbės, 2nd person singular sùktumībės, 3rd person singular sùktī, 1st person plural sùktumībėmės, 2nd person plural sùktumībėtes. J. Endzelīns correctly surmised that Lithuanian būtumēm may derive from būtumbēm, and Latvian byutumem from *būtumbēm with būtubēm37, but he explained their emergence by an implausible dissimilation. In fact, our forms in -tumėm, as well as those in -tumim, may have arisen through blending from coexisting -tum and -tumbėme. Of course, it would be mistaken to suppose that wherever forms in -tumėm are used, they developed in this way. 33 suktumi has a close -i in those dialects, deriving from -ī; evidently, it was influenced by the 2nd person singular forms of other verbs (a- and i-stems), with -i < -ī. 34 See J. Senkus, Lietuvos TSR MA Darbai, Serija A, 2 (5), 133, 1958. 35 Ch. Stang thought that in many dialects -tum had been understood as the optative stem, to which present- or past-tense endings began to be added. This is how the forms -tumei, -tumi, -tumėme, -tumime supposedly arose; see Ch. Stang, Verbum, 250. But, as the preceding analysis of the emergence of -tumime showed, such a development could hardly have occurred; see § 6. J. Otrębski explains the emergence of -tumėme differently; see J. Otrębski, Gramatyka języka litewskiego, 3, 233. 36 Later these forms, together with ē-stem verbs, may have become a-stem forms; cf. sanggathubam, see J. Endzelīns, Gramatika, § 696. The short -u-(celtubem) may have arisen under the influence of the 3rd person celtu; see J. Endzelīns, ibid. By a similar route, that is, on the basis of the 3rd person, the very rare forms found in Daukša’s Postil may have arisen: dariŕubit 499₂₇, estŕubit 499₃₈, giartubit 499₃₈, etc. In some Latvian dialects, the loss of the postposition from forms of the celtubem type, owing to the loss of its morphological meaning, may also have given rise to the optative celtu (see also § 5). 37 See J. Endzelīns, FBR, 15, 11. 78 having arisen. Nor should we forget that the form with -tumėm may have arisen in many places in place of the form -tumim(e), which originated through contamination of forms with -tum and -tumbime. Through leveling of optative inflection, -tumim may have been remodeled on the model of the past-tense forms of the ē-stem, because the singular 2nd-person form -tumei (which arose through contamination of -tum and -tumbei) coincided with the singular 2nd-person -ei of the ē-stem. Moreover, -tumėm may in some places have been borrowed from neighboring dialects; cf. the spread of Dzūkian forms with -tau, -tai in the southern Dzūkian Kapsai dialect, where they are used alongside forms with -čiau, -tum, -tumėi³⁸. § 8. Mažvydas attests to yet another type of optative, with -bi- not only in the plural but also in the singular: sing. 2nd person butumbi, 3rd person butu, plur. 1st person butumbim, 2nd person butumbit. The singular 2nd-person form suktumi of the Samogitian douinininkai arose through cross-contamination of suktumbi and suktum (see § 6). That Mažvydas’s singular 2nd person had -bei is suggested by the singular 1st-person form issakiumbiau Mž 521¹⁸. § 9. In the old monuments of the Lithuanian language (the writings of B. Vilentas, J. Bretkūnas, M. Daukša, S. Vaišnoras, and M. Petkevičius, the 1605 catechism, and the 1653 Kėdainiai publications), we find a third type of optative: sing. 2nd person būtumbei, 3rd person būtų, plur. 1st person būtumbime, 2nd person būtumbite. § 10. Of the last three types, the last is the oldest, since it preserves the ablaut of the old Indo-European athematic optative: the *-(i)jē grade in the singular and the -ī- grade in the plural; cf. the Old Latin siēm, siēs, siēt, sīmus, sītis, sint. *-bīme, *-bīte was shortened to -bime, -bite either because of its enclitic position or under the influence of i-stem verbs³⁹. Ch. Stang’s idea that būtumbei, būtumbiau were formed on the model of ē-stem verbs⁴⁰ is wholly unreliable, if only because the form būtumbei is widespread in almost all the old monuments, while optative inflection could have had nothing to do with ē-stem verbs (the present tense of i-stem verbs would more likely have influenced it), if we accept Ch. Stang’s view that -bime is an old preterite. Moreover, on this account it remains unclear why only the singular 2nd- and 1st-person forms adopted the ē-stem endings, while the plural remained unaffected by their influence. The strongest doubts about whether -bime is of optative origin arise from the fact that Lithuanian, having eliminated ablaut from the inflection of athematic verbs, could hardly have preserved the old Indo-European ablaut in the inflection of the verb’s -bi- optative as well. But the beginnings of the formation of our new optative go back to ancient times, to the period when Lithuanian and Latvian were still one language. We should agree with J. Endzelīns that the optative was forming at a time when word-final *-m had not yet turned into 38 See J. Senkus, cited work, 133. 39 See J. Endzelīns, Gramatika, § 683; Ch. Stang, Verbum, 198. 40 See Ch. Stang, Verbum, 250. 79 into -n, since, on the contrary, -un- would have become ū- (cf. top <*tanpa), and we would not have the Latvian dialect forms sukum (see § 5). Both Old Prussian and Lithuanian records preserve only the language of the sixteenth century—a fairly late period—and yet even in them we find traces of the old ablaut, cf. esti and the participle santyji, eiti ‘goes’ and the participle plural nominative entis, the gerund enti < *jent (cf. Old Indian yant41). See also § 14. Of course, it is scarcely plausible to suppose that in some Lithuanian and Latvian dialects the vowel ablaut in the optative was leveled at a time when b- had not yet coalesced with the supine (cf. Old Indian plural syāma, dadhyāma, Greek εἴημεν instead of εἴμεν, Latin sim instead of siēm). A comparison of the evidence from sixteenth-century Lithuanian dialects with that from present-day dialects also argues against this (in the old records we have forms with -tumbim(e), -tumbit(e); -tumim(e), -tumit(e), whereas today forms with -tumėm, -tumėt are most widespread in Lithuanian dialects). Thus, the leveling of the optative personal-ending system, its adjustment to the verbal conjugation system, took place rather late. In some dialects this happened before the postposition disappeared, cf. Mažvydas’s singular second-person form būtumbi, and the eastern Dzūkian plural first-person form suktumbėmės (and Latvian celtubem); in others, it occurred only after the postposition had disappeared and contaminative forms had arisen. The leveling of optative conjugation in the dialects proceeded in two directions: in some places it was based on plural forms that coincided with the corresponding present-tense forms of i-stem verbs (būtumb-ime=tyl-ime); elsewhere, it was based on the singular second-person form, which coincided with the corresponding past-tense form of an ē-stem (būtumb-ei or būtum-ei=kel-ei). In some dialects the singular second-person form būtumbi also arose on the model of the present-tense conjugation of i-stem verbs; in others, the plural first-person forms suktumbėmės and second-person suktumbėtės, and plural first-person būtumėm and second-person būtumėt arose on the model of the past-tense conjugation of ē-stem verbs. Thus the optative postposition must have developed into the forms of the old stem bhū- optative: *bh(ų)īē->*bjē->*bē and *bh(ų)ī->*bi-. Those who regard this as a preterite rather than an optative rely on bi(i) ‘was’ in old Lithuanian records and dialects and Latvian biju ‘I was’, since the change in meaning optative → preterite is supposedly harder to explain than preterite → optative. But the development of the Lithuanian-Latvian optative bi- into a preterite is readily comprehensible. For the singular conjugation of the optative *bhū- must have coincided with that of the preterite bē-, cf. nābe (Lit. Mundarten I 154), Old Prussian be, bei (with extended -i), Old Church Slavonic bē (an aorist used with imperfect meaning), in conjugation. As a result, with the old optative disappearing, the stem bi- could easily acquire a preterite meaning. Later, the singular third-person form of this verb could have acquired the present-tense third-person ending -ti42 from athematic verbs, since a monosyllabic form had to be avoided and the distinction between “primary” and “secondary” personal endings had disappeared in the Baltic languages. In Latvian, bi- was extended by the secondary -jā. § 11. Most likely, the preterite *bē- has survived in the conjunction bet and the pronoun bet kas, eastern Lithuanian bēt kas (K. Būga Aist. stud. 56). The singular third-person form *bē, like *bi, may have been extended by the present-tense ending -ti. That we have here a remnant of *bēti is shown by the following sentences: There was someone to bow here! ‘he need not have bowed; why should just anyone bow here!’; 41 See E. Fraenkel, Sprachliche, besonders syntaktische Untersuchung des calvinistic Lithuanian catechism of Malcher Pietkiewicz von 1598, 28, Göttingen, 1947. 42 See Ch. Stang, Verbum, 198; E. Fraenkel, ZfSlPh, 20, § 61, 1950. 80 There was someone to ask here! ‘he need not have asked him; why ask just anyone here?’ (Birštonas, Paliūšė). The combinations buvo kam, buvo ko in the sentences cited above seem to be on their way to acquiring the meaning of the pronoun bet kas. In some dialects, the conjunction nes is used in the sense of bet⁴³, cf.: being disheveled, he did not begrudge another the leaves, because he himself had not been eager to gobble them up nu gōslibės DB 237. In general, the nuances of meaning of nes and bet are close. J. Endzelynas⁴⁴ explained the etymology of the conjunction nes: nes derives from *neest(i) (cf. Slavic je beside jestb), used as a question expecting an affirmative answer, cf. French n’est-ce pas? If nes < comes from *neest(i), then bet may derive from *bēti. In proclitic position, *bēti was shortened to bet, while in the eastern Aukštaitian dialects the long ē in the pronoun bēt kas was preserved because bet retained its stress⁴⁵. Thus we too had an aorist *bē- alongside the aorist bū⁴⁶, cf. Old Indian abhūt, Greek ἔφυ, Old Church Slavonic by; just as the Slavs, alongside the aorist bē, have the aorist by. If bet kas arose only in the way just described, one may seriously doubt whether the first component by of the pronoun by kas was borrowed from Polish⁴⁷. It could have developed in a way similar to bet kas, from the optative of the root *bhū- (perhaps already used with preterite meaning), back when *bī had not yet become bi. The following sentences suggest that the optative too could serve as the first component of the pronoun mentioned: There was someone to help here! ‘there is no need to help; why help just anyone!’; There was something to ask for here! ‘there is no need to ask; why ask for just anything?’ (Birštonas). Only būt is ambiguous here: in this dialect it may derive either from būtų or from būta, cf. There was something to ask for here! (Paliūšė). This question requires special investigation. § 12. The 3rd-person forms of the optative of the stem *bhū- that survived the longest are be and bi. be < *bē (cf. the 2nd-person singular forms praustumbe and praustumbės in the eastern Dzūkian dialect) is the 3rd-person singular form of the optative of the stem *bhū-, while bi may be its 3rd-person plural form: bi < *bī < bint (cf. the forms -be and -bi used for the same purpose alongside it in Apse; see below). As modal particles, they may have continued to be used for quite a long time. With the modal particle be, which gives an action a nuance of unreality, we have also formed what is called the compound initial present tense: I was about to leave home, but guests arrived, so I stayed at home; and I was about to leave home, but it began to rain heavily, so I had to stay at home; he was about to take that knife, but look—a worm appeared and went away⁴⁸. Later the particle be lost this modal nuance and became an ordinary prefix be-, cf. We were sowing, but it began to rain—and we stopped sowing⁴⁹. With such a modal particle be or bi, the eastern Dzūkian dialects have also formed an optative, used alongside the optative of the old formation, cf.: ⁴³ See E. Fraenkel, KZ, 60, 237 ff., 1933. ⁴⁴ See J. Endzelīns, IMM, 2, 122, 1927. F. Specht agreed with him; see KZ, 57, 281, and E. Frenkel, see KZ, 60, 237, 1933. ⁴⁵ E. Frenkel mistakenly believes that bēt has a long ē because of the influence of by; see ZfslPh, 20, § 60. ⁴⁶ For bū, see F. Specht, KZ, 59, 58; 62, 142. ⁴⁷ E. Frenkel holds this view; see E. Fraenkel, Litauisches etymologisches Wörterbuch, 42, Göttingen, 1955 ff. ⁴⁸ See J. Jablonskis, Rinktiniai raštai, 1, 311, 316, edited by J. Palionis, Vilnius, 1957. ⁴⁹ See ibid., 316. 6, Kalbotyros klausimai, IV. 81 1st sg. sùkčōbe, 2nd sg. sùktùmbe || sùktùmbeis, 3rd sg. sùktūbe, 1st pl. sùktùmbe || sùktùmbėmės, 2nd pl. sùktūbe || sùktūbėtės. It seems that in the eastern Dzūkian dialects, as the postposition -bėme… disappeared, the new optative forms (cf. 2nd sg. sùktum) were insufficiently marked as optative, and the modal particle be was used to emphasize their form. In some dialects, the particle bi performs this function, cf. praustùžbi, eitùbi (Lazūnai). In some other dialects, both particles—be and bi—were used to mark the optative, cf. 3rd sg. būtùbe and papraustùbi (Apsas)50. Shortening of the postpositional particle yielded 1st sg. nemiegóčiap, užgírščiap (Apsas)51, gyventup (Lazūnai), and so on. Shortening of such a modal particle also gave rise to optative forms in old written monuments: 2nd sg. peržegnotumb K1M 147; 1st pl. atneštumb K1M 142, užmigtumb K1M 186, atsigautumb K1M 18552. Finally, traces of similar forms are also found in Latvian, cf. 1st sg. kolpotub, 2nd sg. bywutùb, sacytùb; duib, sistìb53 (see § 5, footnote 28). Parallel phenomena are found in the Slavic languages. In Old Church Slavonic, bimb was still conjugated; later it was influenced by the aorist forms bychb and was eventually replaced by them. Later still, bychb ceased to be conjugated, and the 1st- and 2nd-person singular form by was generalized to all three persons of the optative, becoming a modal particle. § 13. If the Lithuanian optative is formed from the optative of *bhū-, rather than from the past tense, then one should assume that the Slavic optative is formed in the same way. E. Frenkel considered Slavic bi to derive from the optative form *būi and connected it with Lithuanian bui ‘be’, plural buite, which Volter recorded in Linkmenys54. But in Linkmenys the form is būj, 2nd pl. būjite, and these forms are innovations. This is shown by the system of imperatives in that dialect55, cf.: 1. bū, būn, būj; griū, griūn, griūj; 2. griau, griaun; mau, maun; 3. šluo, šluoj; atsista, atsistoj; 4. šìžau, šìžaun; riešutai, riešutaun; 5. knáisioj, láipoj; 6. didėį, mažėį; 7. blogin, gerin. How did such a variety of forms arise? In all Indo-European languages, the 2nd-person singular form of the root imperative was the bare stem. Most likely, such an imperative 50 See E. Ginaveckienė, Lietuvos TSR MA Darbai, series A, 2 (9), 180, 1960. 51 See E. Ginaveckienė, Lietuvos TSR MA Darbai, series A, 2 (9), 180, 1960. 52 See also Pirmoji lietuvių kalbos gramatika, 87; A. Bezzenberger, BGLS, 214, 7. 53 See J. Endzelīns, Gramatika, 894 ff. 54 See E. Fraenkel, Sprachliche, besonders syntaktische Untersuchungen..., 33. 55 See J. Kardelytė, Lietuvių kalbotyros klausimai, 1, 184, Vilnius, 1957. 82 The remnants of the optative in the Baltic languages are duo, išěi, nèj, Latvian ej56. Some of its forms in Lithuanian coincided with the stem of the verb infinitive (cf. duo and duo-ti), which also began to be used with the meaning of the imperative mood. This gave rise to bū, griū, šluo, šūkau, etc. Their origin is not recent, as shown by the retraction of the acute accent to the circumflex or by shortening; cf. griau < *griau, riešutau < *riešutau, atsista < *atsistā. In the Linkmenys dialect, as in other Eastern Dzūkian dialects, the old optative forms are also used with the meaning of the imperative mood; they often coincide with the 2nd-person forms of the present tense, cf. žālgai, ródai, žadĭes, barĭes, etc. As a result, a close connection developed between present-tense forms and the imperative57. The 2nd-person present-tense forms influenced the imperative forms: alongside bū, riešutau, etc., būnĭ, riešutāuni, etc., arose on the model of the 2nd person; and būn, riešutaūn, šluo, atsistoj, knáisjoj, láipioj, etc., developed as well. Alongside šluoj, stoj, etc., būj, grī̄j could readily have arisen from bū, griū. As is clear, the Lithuanian optative *bui could not have existed. § 14. The 1st-person singular forms būčiau, būčia, etc., are a genuine puzzle because of their unusual shape (-t-j- and the absence of -b-). Justifiably rejecting V. Poržezinski’s explanation, E. Berneker would suppose that sukčiau arose through the blending of the forms suktumčiau and *suktu (cf. Latvian suktu); E. Berneker did not venture to explain sukčia and sukčio (“vor sukcza und sukczo muss ich die Waffen strecken”)58. Ch. Stang, considering the origin of this form one of the more difficult questions in Lithuanian historical grammar, regards the form in -čia as the singular nominative of nouns derived from verbs with ējā stems, and acknowledges that it remains unclear why such a nominative came to function here as the 1st-person optative59. Of all the hypotheses proposed so far on this question, the aforementioned hypothesis of Ch. Stang is the most plausible60, although even it is unlikely to correspond to reality. Thus, the form remains unexplained. The forms of the 1st-person singular in -čiau, -čia are now the most widespread in Lithuanian dialects, occurring alongside one another in many Western Aukštaitian and Eastern Aukštaitian dialects. Forms in -č are also often used alongside them (in the Veliuona dialect and some Eastern Aukštaitian dialects). In some dialects both occur: those in -čiau and those in -č (the Lithuanian fishermen’s dialect, Biržai, Molėtai, Utena, Pagramantis, Gerviniai, Nemakščiai, etc.). In some dialects -č occurs alongside -čia (Alunta, Kuktiškės, Dusetos, Pasvalys, etc.). Finally, in some dialects only -čiau is used (Dotnuva, etc.), in others only -čia (Žasliai, etc.), and in still others only -č (Debeikiai, etc.). In Žemaitian Dounininkai dialects the form -čio is used. There is no doubt that -č arose through shortening. Judging from its distribution across the dialects, it is difficult to say which form it was shortened from (-čiau or -čia). One thing is clear: this shortening is not a regular phonetic shortening. It is difficult to say which form is older: -čiau, -čia or -čio. It is true that in Eastern Aukštaitian dialects we find a reflexive form in -čias or -čios (Svedasai, Alunta, Pa- 56 See F. Specht, KZ, 55, 182; E. Fraenkel, Balticoslavica, 2, 41; IF, 55, 85. 57 In some dialects (Dieveniškės, Apse), the 2nd-person singular present form began to be used with the meaning of the imperative mood. 58 See AfslPh, 25, 488. 59 See Ch. Stang, Verbum, 252 ff. 60 J. Otrembski explains this form in a wholly implausible way; see J. Otrębski, Gramatyka języka litewskiego, 3, 233 ff.; V. Pizani tends to regard *-tjau as a “deuterotetic” borrowing; see Studia Baltici, 9, 41, 1952. 83 Iušė, Tverečius, Kuktiškiai, etc.). It would seem to indicate that -čia < *-tjá, but this may be misleading. Facts from the old monuments of the Lithuanian language help somewhat in sorting out this difficult situation. Mažvydas uses forms with -čia (negalėčia 10₁₈). B. Vilentas has -cžia and -cʒo, J. Bretkūnas -cžia, -cʒau, and -cʒo, M. Daukša — -cʒe and -cʒio, K. Širvydas -čia and -čio, publications from 1653 — -čia and -čio, and D. Kleinas -čiau. As can be seen, the form with -čio, which today is apparently found only in the dounininkai dialects of Samogitia, was formerly much more widespread and used alongside -čia. This form must, of course, have arisen anew by analogy with reflexive forms. Alongside the simple form with -čia, the reflexive form with -čiás(i) was formed, since where a simple verb form has -a, the reflexive form has -ās, cf. the 1st person plural prausiama and prausiamos, etc. It was formed in exactly the same way as the reflexive turėtuos alongside the 3rd person turėtu (see § 4), or the reflexive turėčiuos alongside the 1st person singular neologism turėčiu (Viekšniai). In some dialects (among the dounininkai of Samogitia and in some old monuments), the simple form -čio⁶¹ also arose on the model of the reflexive forms; cf. neologisms of the same kind, patrepšėmo, pašokėmo (TŽ III 433), Kupiškis, etc. Moreover, the Samogitian 1st- and 2nd-person singular forms sakāu, sakāi; darāu, darēi likewise arose from the reflexive forms sakāusi, sakāisi; dariāusi, darēisi in place of the older sakāu, sakāi, etc. The Samogitian acute word-final accent cannot be explained otherwise. It is possible to suppose that the reflexive -čiás(i) was newly created alongside the simple -čia, while -čio came from the reflexive form. When the modal particle be was added to this form, the 1st-person singular sūkčiōbe also arose in some eastern Dzūkian dialects. It remains to clarify the relationship between -čiau and -čia. -čia may have developed from *-tjáu. *-tjáu could have shortened to *-tjá along with the acute long vowels and diphthongs ie, uo at the end of a word. Although acute -áu at the absolute end of a word would be expected to become -āu (cf. sakāu), here, in the optative form, it could have shortened irregularly. This is because irregular shortening is characteristic of optative forms. The optative 3rd-person ending -tų has shortened to -t in the Dzūkian and Central Lithuanian dialects and in the eastern Aukštaitian dialects. In many dialects that do not shorten endings, the shortened 1st-person singular form būč is used. Also compare the 1st-person singular šākčia, bēgčia || šōkčia, bēgčia; 2nd person šāktum || šōktum; 3rd person šākt || šōrt (Linkmenys), etc. Such irregular shortening is also characteristic of imperative and future-tense forms (cf. 1st-person singular būs, eīs in dialects that do not shorten endings). That áu can shorten to -a is shown by palā < palāuk. Later, being isolated within the verb conjugation system, the form with -čia could again have become -čiau in the dialects (following sakiau). The accentuation dėčia, būčia, found in dialects and in Daukša’s “Postilė,” may have arisen later under the influence of the word-final stress of 3rd-person forms ending in -tų. Thus, the basis of all dialectal variants may be the form with -čia < -*tjáu. The 1st-person singular form has had an optative meaning from ancient times. There are no examples in the old monuments of the Lithuanian language of its having the postposition b-. The 1st-person singular formed from the supine and -biau is very rare in the old monuments (only nusidetumbi-au (1 x, 1605 catechism) and išsakitumbi-au Mž 52₁₈). 61 Cf. Ch. Stang, Verbum, 251. 84 All this indicates the archaic character of the optative 1st-person singular form and its ancient optative meaning. If we assume that its basis is a verbal noun, we encounter several facts that clearly contradict its nominal origin. If it is a noun, why does it have no formative -b-? Why was a noun introduced for the 1st-person singular rather than a supine? These questions have not been answered, and it is unlikely that they can be answered at all. Our old optative became a permissive and an imperative mood. The change in the meaning of the old optative, its transition into the imperative, is also connected with the supine’s transfer into the optative system (see § 16). The dialects of Lithuanian preserve remnants of forms for all persons of the old optative: 2nd-person singular sākai, 3rd person tesākai, 1st-person plural sāk aime, 2nd-person plural sākaite. Remnants of the 1st-person singular form, however, have not yet been found. Because of its imperative meaning, this form could not pass into the imperative system. If it could not acquire an imperative meaning, it could have survived and entered the new optative system as its 1st person. The 1st-person singular of the athematic verbal optative must have been: *likjęm ‘I would remain’, *migjęm ‘I would sleep’ (cf. Sanskrit iyām, Old Latin siēm, etc.). As the “secondary” endings disappeared, being replaced by the endings of thematic verbs, *lik-jēd or *likjēuo arose, which must have developed into *likjau (cf. the development of sakiau). It could also have remained in the new optative system, cf. 1st-person singular likčiau, 2nd person likt umbei, 3rd person liktum, and so on. When vowel gradation in verbal forms was leveled in the Baltic languages, the 1st-person singular of the optative of almost all athematic verbs was identical, cf. *duojau, *dėjau, etc. The t could have entered the 1st person from other persons by analogy, thus giving rise to *liktjau, *duotjau, etc. From this, later, when the sound sequence un in the 3rd-person singular form was understood as an ending, the 1st-person ending -jau began to be added to the supine stems daryt-, miegot-, and so on. *-tjau could also have been added directly to the infinitive stem, cf. the formation of the future tense. Latvian -tjau must have changed into -šu and merged with the 1st-person singular form of the future tense. It is not surprising that the form of athematic verbs spread here. Similar phenomena occur in other languages as well. In some Slavic languages (Serbian, Slovenian), the ending -mi, originally found only in a few athematic verbs, spread to all verbs already in historical times. One may ask why the form of athematic verbs spread rather than that of thematic verbs. This happened for entirely understandable reasons: the 1st-person singular of thematic verbs (*sākam, *vežėm, *tylīm) was well suited when the 3rd person had the form of the old optative. But because of this variation in forms, it was entirely unsuitable for the new optative system, in which verbs of all stem classes have the same forms. The 1st-person singular form of athematic verbs best matched the new system here, and it spread as the 1st-person singular of the optative. This origin of the 1st-person singular form further confirms that the optative postpositions -bei, -bime, -bite are of optative origin (see § 10). § 15. The Southern Dzūkians have a distinctive optative, whose formation has likewise remained not entirely clear to this day: 85 sing. 1st person būtáu, 2nd person būtái, 3rd person būtų. pl. 1st person būtūm (būtūmė), 2nd person būtūt (būtūtė). The 1st- and 2nd-person plural forms are undoubtedly new. They were formed by adapting the older būtų, still found in this dialect, to the 3rd-person form, and the older būtum, būtut: the 3rd-person ending -ų was introduced into the forms of all persons⁶². The 1st- and 2nd-person singular forms were likewise newly formed on the model of the conjugation of a-stem verbs. Just as alongside the 3rd-person form bėgo there are the 1st-person singular bėgau and 2nd-person singular bėgai forms, so alongside būtų arose the 1st-person singular būtau and 2nd-person singular būtai. The acute accent on the word-final syllable is explained by the fact that, under the influence of the stress in būtų, the stress shifted to the end here at a time when the post-stress acute accent was still clearly distinguished. § 16. K. Brugman brilliantly surmised that the Slavic languages, which have an optative formed from a participle and an auxiliary verb, have preserved a more archaic state than the Baltic languages, which replaced the participle with a supine⁶³. Even earlier, A. Leskien had identified from Bretkūnas’s “Postil” a type of unreal conditional sentence in which the conjunction jeib occurs only with a participle, though he regarded this as accidental⁶⁴. In an excellent article, Ch. Stang⁶⁵ presents this type of sentence from the writings of M. Mažvydas and M. Daukša; compare: jeib Pons Diews mums nepagelbas... Ir nepretelu neramdas Jau butumbim praszuwe Mž 337; If our little child had not been born || The whole world would have perished Mž 192. On the basis of these sentences, Ch. Stang drew the bold and entirely well-founded conclusion that the consonant b in the conjunction jeib is of verbal origin (as is the case with the optative -bime, -bite), while the constructions jeib mums bernelis negimęs, jeib Pons... mums nepagelbas show that Lithuanians, too, had the type ašė bi bylo. My native dialect (Birštonas), in which the conjunction jép < jeib is still used, also supports this excellent conclusion of Ch. Stang; for example: 1. if he does not come, it is unknown what you would do; if it had not frozen, you would have finished the work; if you had loved, you would have lived well; 2. if he comes, you would have left; if I know that he will trip me up, I would not be alive. In sentences of the second type, present-tense verb forms are used instead of the present participle, since the nominative of the present participle has disappeared in this dialect. This is a replacement of the participle—the participle found in Mažvydas’s writings. We must agree with Ch. Stang that we once had a periphrastic present and preterite optative. It seems to me that the evidential mood in Lithuanian and Latvian is also a remnant of such an optative. Its meaning was aptly described by J. Jablonskis: “A person who relates or recounts something in reported speech presents himself as the person who experienced that thing ⁶² Cf. J. Otrębski, Gramatyka, 232. ⁶³ See K. Brugmann, IF, 30, 349; Grundriss..., 2. 2³, 513. ⁶⁴ A. Leskien, IF, 14, 103, 1903. ⁶⁵ See Ch. Stang, Die litauische Konjunktion jeib und der lit.-lett. Optativ.— “Norsk Tidsskrift for Sprogvidenskap”, 18, 348—356. 86 Not having investigated, truly not knowing, he seems to have said only what he himself had heard or found somewhere. This “indirect,” reported mode of speech is used most often by “straightforward,” frank, cautious people—those who, when they themselves do not know something for certain, say so when recounting it: that they do not know it for certain⁶⁶. This meaning could have been expressed by the optative. Oratio obliqua is also expressed by the optative in other Indo-European languages (Germanic, Greek, Italic). Later, in this kind of periphrastic optative, the burden of optative meaning was transferred to the participle, and the particle bi or be, left without meaning, could disappear. Thus, Lithuanian preserves remnants of an older periphrastic optative, taking us back to a time when the supine was not yet used to express the optative. The question arises: why and how was the participle replaced here by the supine? The supine’s entry into the optative system is connected with a change in the meaning of the old optative. The 3rd person of the old optative acquired a permissive meaning, while the 2nd person acquired an imperative meaning. That the optative had long since changed its earlier meaning is shown not only by the East Aukštaitian dialects, where the optative constitutes the entire imperative system, but also by most of our dialects, where we still find surviving optative forms with an imperative meaning; compare kláusai (listen) Linkuva, tylý (be quiet) Svedasai, imý, bėgý, nešý (take, run, carry) Kretinga⁶⁷. Only later, in many dialects, were the optative forms apparently replaced by a newly formed imperative with the suffix -ki. Latvian, too, preserves traces of the optative being used with an imperative meaning. We do not know what meaning the present and preterite formations of the optative had, just as we do not know what role the optative played in subordinate clauses in general. That it existed there is shown by sentences such as nėr kas valgai (there is no one to eat), atsiras kas darai, nėr kas dirbie (dirbų)⁶⁸. The sentences cited earlier⁶⁹ show that the present optative, too, is used with a preterite meaning, exactly the same meaning as the preterite optative; compare jép ateina (he is coming), būtum išvažinėję. That an optative formed from the present stem could indeed have a preterite meaning is also shown by the following sentence: Kad tasai ne būtų ėssąs piktadėja, ne išdavę būtumbime io taw DP 163. It should be assumed that both the present and the preterite formations of the optative had a preterite meaning. Apparently, in Lithuanian, as in other Indo-European languages (Germanic, Avestan, Latin), a new optative preterite was created. In Lithuanian, it was formed from the optative of the verb * bhū- and a present or past participle. As the old optative forms turned into the imperative, the periphrastic optative forms, because of their preterite meaning, could not replace the simple optative forms. The participle here had to be replaced by a verb form indifferent to tense. Both the infinitive and the supine are tense-indifferent forms. Why did the supine, rather than the infinitive, replace the participle? ⁶⁶ See J. Jablonskis, Rinktiniai raštai, 1, 524, edited by J. Palionis, Vilnius, 1957. ⁶⁷ See E. Grinaveckienė, Lietuvos TSR MA Mokslo darbai, series A, 2 (9), 181, 1906; J. Aleksandravičius, Vilniaus Valstybinio pedagoginio instituto Mokslo darbai (Lietuvių kalba ir literatūra), 8, 117, 1959. ⁶⁸ See J. Jablonskis, Rinktiniai raštai, 1, 460. ⁶⁹ See p. 86. 87 The meaning of the supine must have been more closely associated with the optative than with the infinitive. In some subordinate clauses expressing purpose, where the main clause contained a verbum movendi, the supine may have been used from ancient times; cf.: Nuvejo turgun, kad žirnių nusipirkt (nusipirktų). Išėjo, kad arklių parvest (parvestų). Nuvejo pas kaimyną, kad pieno parsneštų (Palūšė). In other sentences of a similar type, where the main clause did not contain a verbum movendi, the optative could be used, e.g., *pirksim duonos, kad jie valgai. As the old optative meaning weakened, the supine could easily replace the optative here by analogy with primary clauses. The supine could thus acquire a shade of optative meaning in sentences of a similar type. Perhaps that is why the 3rd-person form of the new optative is, from ancient times, a pure supine. Ch. Stang does think, however, that the 3rd person of the supine also had bi, but that it was used after a conjunction and, over time, because of its short form, became an enclitic and fused with the conjunction; that is, for a while people said: jeib žinotų: jei žinotumbime, until finally jeib became a conjunction⁷⁰. This is possible. But first of all, in the dialect mentioned earlier, where the conjunction jėp = jeib is used, no one would say *jep žinotų, ateitų, but only jei žinotų, ateitų. This construction is archaic. Second, it remains unclear why in the other persons -bime, -bite came to occupy an enclitic position. After all, the optative of the verb *bhū- must have had a free position in the sentence, just as it did in the Slavic languages. It seems that in ancient times the 3rd person was a pure supine. Later, in some places, following the example of the other persons, the 3rd-person supine also began to be used with the optative bi or be. But since it was not needed here, it could soon fuse with the conjunction. Thus, most likely, we originally had: * if we knew, and * if you knew we did, * if you all knew, and * if you knew you all did. * if you knew. Later, in some places, the following may have arisen: * if we knew, and * if you knew we did, * if you all knew, and *if you knew you all did, * if you knew. * if you knew. Still later, the optative meaning of *bime weakened, and the supine had to take on an ever greater share of its functional load. The ending was not needed in the third person, since other 3rd-person verb forms also had no ending. To meet the language’s requirement that persons be distinguished by their endings, the enclitic position of the optative bhū- was generalized in the other persons, and the supine fused with it into a single form. The third-person form be, bi fused with the conjunction. Abbreviations Archive for Slavic Philology, Berlin, 1876 ff. BB — Contributions to the Knowledge of the Indo-European Languages, Göttingen, 1877—1906. ⁷⁰ See Ch. Stang, cited work, 353. 88 BGLS — A. Bezzenberger, Contributions to the History of the Lithuanian Language... Göttingen, 1885. ChB — Chyliński’s Lithuanian Bible. New Testament. Volume II, edited by Czesław Kudzinowski—Jan Otrębski, Poznań, 1958. DP — Postilla Catholica... Per Kuniga Mikolai D a v k s z a. Wilniui 1599. Naudotasi 1926 m. fotografuotiniu leidimu. FBR — Proceedings of the Philologists’ Society, Riga, 1921 ff. IF — Indo-European Research, Strasbourg, 1892 ff. IFA — Anzeiger für indogermanische Sprach- und Altertumskunde, supplement to the Indogermanischen Forschungen, Strasbourg, from 1892. IMM — Monthly magazine of the Ministry of Education. KIM — D. Klein, Naujos Labbay priwalingos.... Maldu knygeles.... Karalauczuje, 1704. KZ — Zeitschrift für vergleichende Sprachforschung auf dem Gebiete der indogermanischen Sprachen, from 1852. Mž — M. Mažvydas, The Oldest Monuments of the Lithuanian Language up to the Year 1570, Kaunas, 1922. Journal for Slavic Philology — Journal of Slavic Philology. FROM THE HISTORY OF THE OPTATIVE И. И. КАЗЛАУСКАС Abstract The formation of the Lithuanian-Latvian optative has not yet been clarified. The question of what the second component of optative forms represents—an optative or the preterite of the stem *bhū-—is disputed. It is not entirely clear what the first component represents—a supine or an imperative (following the appearance of V. Pisani’s article). Nor is it clear how and why the supine could have been drawn into the optative system. Scholars also disagree about the origin of individual dialectal forms of the optative. The origin of the 1st person singular form (in -čiau, -čia) is especially unclear. The present article is devoted to these questions. The article demonstrates that there is no reason to doubt that the first component of the optative is the supine. The second component of optative forms is the optative, not the preterite of the stem *bhū-, because: 1) this is indicated by the meaning of the optative’s postposition and the vowel alternation (-bei: -bime), which cannot otherwise be explained without straining the evidence, nor is there any need to do so, since in the optative’s distinctive conjugation this alternation could have persisted until a late period; 2) the optative stem *bi- (<*bī-) could easily have acquired the meaning of a preterite, since the singular forms of this stem (*bē<*bjē) coincided with the corresponding forms of the preterite bē-, cf. Lithuanian dialectal niabē ‘was not’, Prussian be, bei, Old Church Slavonic bē. The modal particle -be< -*bē, by means of which the optative can be formed in East Dzūkian dialects, should be considered a remnant of the 3rd person singular form of the optative of the stem 89 *bhū-, while the particle -bi, used for the same purpose in some dialects alongside -be, is probably a remnant of the 3rd person plural form (-bi<*-bi<*-bint). The optative of the stem *bhū- must have fused with the supine into a single word no later than the Lithuanian-Latvian period. All the dialectal forms of the 2nd person in both numbers and the 1st person plural of the Lithuanian optative known to the author go back to this compound form. Under the influence of the 3rd person form (būtų), which never had a postposition, the postposition into which the optative of the stem *bhū- had developed lost its morphological meaning and began to disappear. Thus, alongside the long postpositional forms, short forms also appeared (cf. dialectal 2nd person singular sùktū̃ || sùktū̃beis, 1st person plural sùktū̃ || sùktū̃bėmės). This new optative type in -tum is used in Latvian dialects (where the 3rd person, too, has a form in -tum, since it has become indifferent to person). Clear remnants and traces of it have survived in Lithuanian dialects (cf., for example, 1st person singular suktum, 2nd person singular suktum, 1st person plural suktum, 2nd person plural suktum or suktut (a new formation replacing suktum, under the influence of other 2nd person forms of the verb), 3rd person suktų, suktum). In the Latvian literary language, the form suktu, used for all persons, may by origin be a 3rd person form that began to be used for all persons by analogy with suktum (or it arose from suktubem... in connection with the disappearance, for morphological reasons, of the postposition -bem...). Under the influence of the system of verb conjugational endings, another type of optative developed in many Lithuanian dialects through the blending of shortened and long postpositional forms, e.g. 2nd person singular -tumei (from the concurrently used -tumbei and -tum), -tumi (from -tumbi and -tum), 1st person plural -tumim (from -tumbim and -tum), -tumėm (from -tumbėm and -tum), and so on. The 1653 Kedainiai publications clearly point to such a process of blending; cf. the forms of the 1st person plural used concurrently there: -tum(e), -tumbim(e) and -tumim(e). The leveling of the optative ending system in Lithuanian dialects proceeded in two directions: in some, it was leveled on the basis of the 2nd person singular, following the pattern of conjugation of verb stems in -ė; in others, on the basis of plural forms, following the pattern of conjugation of stems in -i. The South Dzūkian optative būtai, būtái... is a recent formation. The acute-accented ending is explained by the fact that, under the influence of būtų, the stress shifted to the end of the word when the acute intonation of post-stress syllables was clearly pronounced. Variants of the 1st person singular form in -č, -čio, -čiau are explained as new formations going back to -čia. -čia may have developed from *-tjàu, which may have shortened to *-tja along with other long acute-accented sounds, given that optative forms characteristically undergo irregular shortening (cf. būt<būči, būč<būčia, etc.). The form in *-tjàu 90 is explained as a transformed form of the athematic optative (cf. OInd. iyām, OLat. siēm, etc.), i.e. *likjēm>*likjēō (*likjēuo)>*likjau ‘would have remained’. As it corresponded well to the conjugation system of the new optative, it spread to all verbs. Under the influence of the forms of other persons (liktumbei, etc.), it acquired -t-, i.e. *likjau>*liktjau. The possibility that the form in -č- is of optative origin is also shown by the fact that the optative of the stem *bhū- was not added to it. The appearance of the supine in the optative system is closely connected with the change in the meaning of the ancient optative. Judging by the evidence of ancient texts and some Lithuanian dialects, in Lithuanian-Latvian, as in some other Indo-European languages, there may have existed a compound preterite optative, formed from a participle and the optative of the stem *bhū-. As the meaning of the ancient optative changed, these preterite forms could not replace it because of their preterite meaning. The participle in these forms could therefore be replaced by the supine, a verb form indifferent to tense. The supine may have acquired optative meaning in certain subordinate clauses.