William R. Schmalstieg
The Pennsylvania State University
CDU 801.52 : 808.1-25 : 803.0.023-25
THE SLAVIC AORIST ENDING IN -tb AND THE OLD PRUSSIAN PRETERIT IN -ts
The Slavic 2nd/3rd sg. aorist in -tb, e.g., pi-tb 'drank', po-vi-tb 'wrapped', u-mre-tb 'died', za-ce-tb 'began', ras-p§-tt> 'crucified', vb-z§-tb 'took', pro-stre-tb 'stretched', -zre-tb 'swallowed', vbs-pe-tb 'started to sing', kl§-tb 'swore', etc. have long been a subject of dispute. Some have thought that these forms were derived by the addition of a personal pronoun tb. Then since 2nd and 3rd sg. aor. forms vede 'led', bbra 'took', vide 'saw' were the same the tb was substituted back into the 2nd person also (Stang, 1942,219). Van Wijk, 1926,281, noted that the ending -tb was characteristic of verbs with an original circumflex accentuation and that the same verbs had past passive participles in -tb also. Roots with an original acute accentuation lacked the aorist ending -tb and had the past passive participle formation in -en-. Stang, 1942,65, wrote that he believed that the verbs with the -tb aorist ending were old root aorist forms (or imperfects from root presents). The only problem with this would seem to be the expectation that the Indo-European root aorist ended in *-t which should be lost in word-final position as a result of the action of 'the law of open syllables'.
Meillet, 1902, 139-140, considered the Slavic ending -tb of these verbs to be derived from an earlier 2nd sg. perfect ending *-to which is attested also in the Old Indie 2nd sg. perfect vet-tha, Gk. ois-thas 'you know'. Contaminated with the secondary 2nd sg. ending *-s an ending *-tos was created which led to the attested ending -tb. Stang, 1942, 221, writes that the number of instances of final *-to are so limited in Slavic that one cannot completely exclude the possibility that final *-to passed to Slavic -tb (It is assumed that *-os > -z> in the nom. sg. of *o-stem nouns, e.g., *orbhos > rabb 'slave'.).
Now the relationship between the 3rd sg. middle aorist ending *-to and the participles in *-t- was noticed by Hirt. According to Hirt, 1928, 102, "Die 3. sg. Med. des sogenannten Wurzelaoristes geht auf -to aus. Wir können die Formen des sog. starken Aorists im Griech., des Wurzelaoristes im Indischen dem Kasus indefinitus des Verbaladjektivs auf -to- gleichsetzen: ai. äkrta : krtds 'gemacht'; ai. avrta : vrtäs 'gewählt'; ai. ärta : rtäs 'geschickt'; ai. amrta: mrtäs 'gestorben'; ai. asprta: sprtäs 'gewinnen'; ai. ajukta : juktäs 'anschirren'; ai. amata 'denken': matäs; ai. aprkta 'mischen' : prktas] ai. abhakta 'zuteilen': bhaktäs\ ai. vikta 'zittern': viktäs-, ai. äspästa : spastas; ai. asrsta 'loslassen'; srstäs gr. etheto : ai. adhita: gr. thetos, ai. hitäs\ ai. ästhita : sthitäs von sthä 'stehen'; ai. astrta: astrtäs 'unüberwindlich'; ai. agurta: gurtäs 'begrüßt', 1. grätus; ai. agata : gatas, gr. batös 'gegangen'..."
197
Following Hirt I have assumed that this middle aorist ending is the same as the neut. sg. nom. participial ending *-to. One notes now that in Old Irish there is a preterit passive deriving from the participial forms, thus IE nom. sg. masc. *bhrtos > pre-Irish *brita(s) > 3rd sg. brith whereas the nom. pi. fem. *bhrtas > 3rd pi. bretha (Kerns and Schwartz, 1972,55).
Here I would like to point out that the Indo-European participle *-ont- is exclusively active in all the Indo-European languages except Hittite where -ant- is active for intransitive verbs, but passive for transitive verbs. Thus kunant- 'killed' (from kuen-'to kill'), appant- 'seized' (from ep- 'to seize'), dant- 'taken' (from da 'to take'), but pant- 'gone' (frompai- 'to go'), akkant- 'died' (from ak- 'to die'). Exceptionally the participle adant- (from ed- 'to eat') can mean 'having been eaten' and 'having eaten' the participle aknuant- (from eku- 'to drink') can mean 'having drunk' and 'having been drunk' (Friedrich, 1960,144—145). The active meaning of the participles derives from the time prior to the period when the stems ed- and eku- became transitive. The passive meaning is later, since it assumes possible transitivity on the part of the stem.
Similary the *-to participles were originally active intransitive and then could later be interpreted as passives of transitive verbs. Thus Latin po-tus can mean 'who has drunk' or 'what has been drunk, drained' (Gonda, 1960,66). Likewise Latinpran-sus means 'who has breakfasted'. Old Indie bhuk-ta can denote 'the thing eaten, or enjoyed, food' or 'one who has eaten a meal' (Monier-Williams, 1899,759). Old Indie pl-ta can denote 'having drunk' (Monier-Williams, 1899,629). Thus one could imagine that a Slavic 3rd sg. aor. pitt might derive from a middle aorist form which was identical with the masc. nom. sg. participle *-tos, rather than the neuter *-to. Since the ending would derive from a participle the identity of the 2nd and 3rd person forms would present no problem. The participle, or middle aorist form, losing whatever special function it may have had, united with the aorist. Thus Slavic pith (like Latin potus) could mean either 'who has drunk' or 'what has been drunk, drained'.
The following verbs do not show the formal identity with the past passive participle: u-mre-tb 'diedpro-stre-tb 'stretched', -zre-tb 'swallowed'. All of these verbs have a zero-grade root in the present tense, thus zirgtb 'they swallow', mhrQtt 'they die', -sthrQtb 'they spread' (Lunt, 1955, 117). The zero-grade forms which were ety-mologically marked for aspect pushed the full-grade forms into the aorist function. This is the origin of the full-grade root aorists. In the verbs u-mre-tb, pro-stre-tb, -zre-tb the final -tb was used to hypercharacterize the aorist. This -tb was taken from forms such as pi-tb 'drank', po-vi-tb 'wrapped', etc. where the -tb had its origin in forms similar to the past passive participle. (Forms such as u-mre 'died', pro-stbre 'spread out' [where the b is probably an orthographic addition] attested in the Codex Suprasliensis [Diels, 1932, 249] could reflect the full-grade root aorist *mer-t and *ster-t respectively, but they may be the result of later development.)
In Old Prussian an ending -ts appears as an ending to certain verbs. There are six ways of explaining this form: (1) it has its origin in an incorrect use of the past passive participle; (2) it is German -(e)s which has penetrated from German into Old Prussian
198
in the first most important Christian writings; (3) it is a pure Old Prussian pronoun *tas or *dis literally translated from German er, es. (4) it is a relic of the hypothetical Baltic thematic 3rd sg. ending *-f; (5) it is an innovating form of the conjunctive; (6) it is an innovating form of the preterit (Palmaitis, 1989,128). Palmaitis tries to show that all of these suggestions are incorrect and that the form denotes the modus relativus. He notes that in Old Prussian the ta-participles of intrasitive verbs have active meaning, so that a form such as dmkauts can pass in meaning from 'thanked' to *'thankful' to *'the one thanking, the one who thanks, thanker'. Once the -ts came to be felt as a marker of the modus relativus it could be used in the present, thus (esti-ts- (=111 astits) vs. the preterit *bu-ts. As far as its etymological origin is concerned the Old Prussian ending -ts could then be cognate with the Slavic aorists in-ft, viz. both could go back to an old 3rd sg. middle aorist ending *-tos (rather than the neuter *-to encountered in Greek edo-to, Old Indie ddi-ta).
One notes that Kazlauskas, 1968,303, wrote that Old Prussian -ts could formally be equivalent to Slavic -tb. Sometimes the Old Prussian -ts is added to the same stem as the infinitive, thus (prei)dlnkaut 'to thank', 3rd pret. dmkauts-, infinitive dat, datwei 'to give', 3rd pret. daits (= /dats/), beside dai (- /da/). Probably the same is true for poquoitets 'begeret, wishes', although this latter form may be a nom. sg. masc. past psv. participle. But one also encounters examples in which the final -ts seems to have been added to a preterite form which already has a vocalic suffix, thus 3rd pret. llmauts, limatz, lymuczt 'broke' probably to be phonemicized as /limats/ vs. the infinitive lim-twev, and ymmeits, ymmeyts, ymmits, jmmitz 'took' (also imma [cf. Lith. 3rd pret. erne]) vs. the infinitive lm-t (cf. Lith im-ti). These Old Prussian preterits could represent the addition of -ts (extracted from forms similar to the participle such as dmkauts) to the 3rd person preterit ending in a suffix vowel. Cf. the similar principle of contamination in which the -tj> from forms similar to the participle such as pi-tK 'drank' was added to the suffixless preterit thus deriving, e.g. from Slavic u-mre the form u-mre-th 'died'. There is then evidence of an etymological ending *-tos derived originally from either a middle ending or a participial ending which came to function as a preterit marker in both Slavic (aorists in -tb ) and Baltic (Old Prussian -ts).
Bibliography
Diels, Paul. 1932. Altkirchenslavische Grammatik. Heidelberg, Carl Winter. Gonda, J. 1960. Reflections on the Indo-European medium I. Lingua 9.30-67. Friedrich, Johannes. 1960. Hethitisches Elementarbuch. Heidelberg, Carl Winter. Hirt, H. 1928. Indogermanische Grammatik. Doppelung Zusammensetzung Yerbum.
Heidelberg, Carl Winter. Kazlauskas, Jonas. 1968. Lietuviy kalbos istorine gramatika. Vilnius, Mintis. Kerns, J. A. and Schwartz, Benjamin. 1972. A sketch of the Indo-European finite verb.
Leiden, E.J. Brill.
199
Lunt, Horace G. 1955. Old Church Slavonic Grammar. 'S-Gravenhage, Mouton. Meillet, Antoine. 1902. Études sur l'étymologie et le vocabulaire du vieux slave. Paris, Librairie Émile Bouillon, Éditeur. Monier-Williams, Monier. 1899. A Sanskrit-English dictionary. Oxford, Clarendon Press.
Palmaitis, M.L. 1989. Borussica: 2. astits, billäts - das prassische Modus Relativus?
Baltistica 25. 128-132. Stang, Chr.S. 1942. Das Slavische und Baltische Verbum. Oslo, Jacob Dybwad. Trautmann, R. 1910. Die altpreussischen Sprachdenkmäler. Göttingen, Vandenhoeck and Ruprecht.
Van Wijk, N. 1926. Die slav. Partizipia auf -to- u. die Aoristformen auf -te. Indogermanische Forschungen 43. 281-289.
Povzetek
SLOVANSKO OSEBILO AORISTA -tb IN STAROPRUSKI PRETEKLIK NA -ts
Slovansko osebilo aorista -tb in staropruski preteklik na -ts izhajata mogoče iz prvotnega sklonila za imenovalnik ednine moškega spola *-tos, ki je bil glede na rabo podoben osebilu za tretjo osebo ednine aorista srednjika *-to (npr. v grškem édoto, staroindijsko äditd). Slovensko pi-tb so morda v nadaljevanju tolmačili prehodno ali neprehodno, podobno kakor latinsko potus (ta pomeni bodisi 'ki je pil' ali 'ki je bil popit, izpit').
200