Zur Vorgeschichte des Inessivs im Urostbaltischen

ACTA LINGUISTICA LITHUANICA LI (2004), 59 – 67 On the Prehistory of the Inessive in Proto-Baltic1 ILJA SERŽANT Universität zu Köln The article deals with the prehistory of the inessive in East Baltic. The iness. sg. of ā-stems in *-āi̯ēn is shown to be morphologically a very old formation that must have arisen before the loss of laryngeals in Baltic, and must be the oldest among the inessive forms. The fact that the inessive postposition *-ēn also is unclear within Baltic entitles us to look for an explanation reaching back to IE. An immediate parallel can be found in Old Indic, where the locative sg. ending of the ā-stems -āyām provides a phonetically exact match for Baltic *-āi̯ēn. In Indic the postposition -ām is also restricted to the locative of ā- and ī-stems (the latter have not survived in Baltic). The distribution of the postposition and its phonetic shape point to an IE postposition *-ēm, an optional additional marker of the locative. 0. PRELIMINARY REMARKS The old locative ending of Lithuanian and Latvian, built into the inessive form singular of the ā-stems and to be reconstructed as *-āi (“´” stands for the acute accent), cf. Lithuanian galvoj-è and Latvian galvā, does not occur in this form in any other Indo-European language. In those Indo-European languages where the distinction between long and short diphthongs has remained intact, we find a short diphthong *-ai, cf., for example, Greek dialectal βοταμι̯ων, where -ε < *-ai < *-eh₂-i (Rix 1976: 133). The reconstruction of a Proto-Baltic *-āi (e.g. Stang 1966: 199) therefore requires an explanation. 1. THE SINGULAR OF THE *ā-, *-i̯ā- (è)-STEMS 1.1. The inessive ending is not archaic here simply because it contains an unshortened *-āi- (so Zinkevičius 1982: 21). The locative ending *-āi̯- contained in it can ONLY be explained as a pre-vocalic sandhi variant of the Indo-European locative morpheme: *-ah₂-i̯-V > *-ā́j-V. It is noteworthy that we find it precisely in pre-vocalic position as well, that is, before the inessive particle or postposition *-ēn (for this proposal, see below). Otherwise *-āi- would not have arisen at all, owing to a different syllabification. 1 This is a fuller version of two presentations: “Некоторые заметки о восточнобалтийском локативе”, XXXIII International Philological Conference, Saint Petersburg, 5–6 March 2004, section “Baltic Languages: Historical Grammar, History of the Literary Language”; and “Dažas piezīmes par latviešu lokatīvu”, 40th conference “Artura Ozola dienas”, Riga, 19–20 March 2004. I am sincerely grateful for the valuable critical comments and other assistance of Prof. J. L. García-Ramón (Cologne), Prof. D. Petit (Paris), and Prof. A. Holvoet (Vilnius), who nevertheless did not agree with me on every point of this paper. 60 | ILJA SERŽANT Although it is theoretically possible that the prevocalic sandhi variant also arose before other words beginning with a vowel, the fact that it has been preserved precisely before PIE *-ēn speaks for a very great antiquity of the close attachment of the postposition PIE *-ēn to the old locative form. This probably took place before the loss of the laryngeals, which is why the prevocalic sandhi variant was able to survive at all. 1.2. The preconsonantal sandhi variant PIE *-eḥ₂-i > *-ai > Proto-Baltic *-ai (?) expected in this context may still be preserved in the forms of the adessive; cf. mergaīp / mergāip < *-ai-p. 1.3. Furthermore, the following argues for the great antiquity of the singular inessive form of ā-stems: with a prevocalic sandhi variant *-āj, in accent class 2 a shift of the ictus to the ending would be expected before the postposition was attached, according to de Saussure, i.e. *raṅk-oj > **raṅkój. This form would then, after the form had been expanded by *-ę, have become **raṅkojē, or, by analogy with the mobile-stress forms, where the principle of marginality² applied, **raṅkojē. But no such accentuation is found, and one must therefore assume that the expansion by *-ę predates the operation of Saussure’s law, since in trisyllabic words Saussure’s law did not operate between the first and the middle syllable. 1.4. Given the fact that the inessive ending -oje appears to be old, a remark on Latvian is also in order here. One must assume that a similar locative allomorphy (*-ai before # or K, *-āj-V) also existed in Proto-Latvian. The second variant *-ai must then have yielded *-i through the general shortening of word-final vowels in Proto-Latvian, which is not attested but can certainly be reconstructed. This *-i, however, also coincided phonetically with the locative ending of o-stems, *-o-i > *-ie > *-i; and since this ending was ambiguous³, the ā-stem ending was transferred to the o-stems according to the following pattern: ā-stems o-stems Primary ending *-i *-i Extended ending *-ājī (< *-ājén) x x = *-ājī. Only the merger of the primary locative endings of the o- and ā-stems made it possible to transfer the extended ending of the ā-stems to the o-stems in Latvian⁴. Latvian thus also indirectly provides evidence, on the one hand, that in Proto-Baltic- 2 According to this principle, in trisyllabic word forms the mobile accent could stress only the first or the last syllable, never the middle one; cf. Stang (1966: 290 and 490 n. 1); Zubatý (1896: 276). 3 It was identical to the dative singular and nominative dual of ā-stems, the locative singular of o- and ā-stems, and the nominative plural of o-stems. 4 Vanags (1994: 125) erklärt die Gleichheit der Lokativendungen der o- und ā-Stämme durch den phonetischen Zusammenfall der alten Illativendungen -an > -ā und *-ana > *-ān > -a. Dies kann aber nicht für diejenigen Mundarten gelten, die die alte Lokativendung bewahrt haben, z. B. die Endung des Lok. Sg. der o- und ā-Stämme -ę < -ai < -āji im Niederlettischen. ZUR VORGESCHICHTE DES INESSIVS | 61 that there was a double reflex of the Indo-European singular locative ending in the ā-stems, and, secondly, that the extended Lithuanian singular locative of the o-stems miškė must be quite recent (cf. 2.1.3). 2. THE OTHER INESSIVE ENDINGS The following (2.1–2.3) shows that the ancient postposition *-ę́ originally occurred ONLY in the singular locative of the -ā-stems (and *-i̯ā- (-ē)-stems), and spread relatively late by analogy to the plural and to the singular and plural of the other stems. 2.1. Singular of the *o-stems 2.1.1. The old Indo-European singular locative ending is preserved in fossilized forms such as namiẽ, ankstĩe (attested in Zinkevičius 1966: 209 § 278), as well as in the singular adessive namiẽp(i). On the basis of the primary forms just mentioned, the intonation should be reconstructed as circumflex. This continues precisely the Indo-European locative ending *-o-i > *ōi > iẽ. Following Stang (1966: 182–3), one might envisage the development as follows: *-oi (cf. Sanskrit aśve < *eḱu̯oi) > *-ai > *-ę̄ > -iẽ. The common inessive ending is -e, e.g. miškè (‘in the forest’). This ending contains the original Indo-European ending, which was additionally extended, like all other locatives, by the inessive postposition *-ę́⁵ (cf. Stang loc. cit.). This inessive form (miškè) is to be explained as follows: *miškē (< *-ai < *-o-i) + ę́ > *miškē⁶ > (Leskien’s Law) miškè (Stang loc. cit.) or > miškì in those dialects where every *en > i. It follows that the extension of the o-stems by *-ę́ took place relatively late. The development of a form mišk-è at an earlier stage, when *-ai (< Indo-European *-oi) was still a diphthong, would be inexplicable, since *-ai + ę́ would have become **-aję. This is also supported by the fact that a large number of fossilized singular locatives of the *o-stems have been preserved in Lithuanian, unlike the -ā-stems. 2.1.2. The addition of the postposition *-ę́ must have taken place before or during the operation of Saussure’s law, since only through the fusion of the ending *-ę̄ and the postposition *ę́ in rate (accent class 2) could the acute in the second and final syllable arise, as required for the shift. Before the contraction, the ictus would not have shifted, since under de Saussure’s rule of shifting it cannot skip intervening syllables. 2.1.3. The evidence in Latvian (cf. 1.4) also indirectly supports the relatively recent date and exclusively Lithuanian development of the inessive form of the o-stems. 5 Here and below, <´> denotes intonation (acute), not ictus. 6 Der Einwand von Mažiulis (1970: 133.1), nach dem eine Kontraktion unbedingt zirkumflektierte Intonation hätte ergeben müssen, ist nicht ganz berechtigt. Es geht hier erstens um eine wesentlich spätere Kontraktion, als z. B. diejenige der indogermanischen Vokale (wie z. B. im Dat. Sg. der o-Stämme *-o-ei), und zweitens erfolgte hier die Anfügung des *-ę̄n in Analogie zu dem Lok. Sg. der ā-Stämme, folglich kann man hier auch die analoge Übertragung der akutierten Intonation auf das Kontraktionsergebnis (*-ę́-én > *-ēn analog zum Lok. Sg. der ā-St. *-āj-én > -én) nicht ausschließen. 62 | ILJA SERŽANT Here there was presumably no analogical form at all; rather, Latvian simply used the inessive form of the -ā-stems. 2.2. The Inessive Endings of i- and u-stems 2.2.1. The Indo-European ending of the u-stems was *-ōu; cf. Sanskrit sūnāu, Slavic сынѫ. Thus Lithuanian sūnujè/sūnųjì, High Latvian tirguô (< *tirgôji), etc., are young, since the -j- of the postposition can only be explained from the locative singular of the -ā- or -i-stems. The locative *sūnù, which is also contained, for example, in sūnujè/sūnųj, can likewise only be explained from *-ōu-# by Leskien’s law7. The extension -je must have arisen subsequently. 2.2.2. The situation is different with the i-stems. Although the modern ending -yje (avyjè) has been remodeled after other stems and is therefore recent, in the Samogitian dialect one finds -ie < -ē in its place (Zinkevičius 1966: 231), -ie, -ē in some Latvian dialects (Endzelin 1951: 412), and nakteie, schirdie, naktie, ma-nieie in Mažvydas’s catechism (Stang 1966: 209, 211). These forms go back to *-ējē or, in part, *-iejē, and thus contain the old i-stem locative singular ending *-ēi; cf. Vedic agná, Ancient Greek πόλη, 8. It is difficult to determine with certainty whether these forms belong properly to the -i-stems or were borrowed via the *-(i)jas-stems from the -ē-stems9. The extension with *-ē could be explained on the basis of the variation between the proto-forms *-ējē and *-iejē (Stang 1966: 210f): the old locative form *-ēi was extended by -je, an extension abstracted from the locative singular of the ā-stems, and initially developed into *-ējē10. The individual dialects then treated this form differently: some simplified it—*-ējē > *-ēje > Lith. -ēje, > Latv. -ē—while others retained the diphthong and later changed it by regular sound change to -ie-: *-ējē > *-iejē > -ie. The Samogitian ending -ie (< ē < *ēje) may also have been transferred from the *(i)jo-stems to the i-stems; in the *ijo-stems, in turn, it may derive from the ē-stems, which partly coincide in Samogitian and Latvian (Endzelin 1922: 31). One must posit such or a similar development; otherwise one would face the difficulty of having to date the extension with *-ē—abstracted as *-jē—relatively chronologically to two far-removed points in time in different areas: either to the period when *ēi existed (Samogitian and Latvian dialects), or to the period when ié existed (Latvian dialects). It should be borne in mind, however, that a complicated development intervenes: *ēi > *ei > *ē > ié. Whichever proposal one adopts, 7 Zinkevičius (1982: 27) erklärt diese Endung als in Analogie zu den i-Stämmen (-i) entstanden, wo sie aber auch (Zinkevičius 1982: 20–21) aus den konsonantischen Stämmen stammen soll. Die beiden Formen lassen sich doch aber unter Heranziehung des Leskienschen Gesetzes lautgesetzlich erklären: -u (< *-ūo < *-ōu), -i (< *-ie < *-ēi). Zumal die Zwischenstufen -uo und -ie verbaut in den lettischen und žemaitischen Mundarten überliefert sind. 8 The old ending is certainly preserved in infinitives in *ti; cf. northwest Samogitian Lith. -tę, reflexive -teis. 9 Cf. debesesu in Mažvydas’s catechism, 291₃ (apud Kazlauskas 1958: 82), a locative plural with an ending in ē- (< *ijā-) stems on a synchronically i-stem (originally an s-stem). 10 Nur so rekonstruiert sie Kazlauskas (1968: 154). Anders Stang (1966: 210). ZUR VORGESCHICHTE DES INESSIVS 63 One must assume that the extension with *-ę́/-ję occurred substantially later than in the locative singular of the -ā-stems, namely by analogy with the latter. 2.3. The Inessive Plural The old plural ending is preserved in only a few dialects; cf. šird-išū, rañkosu (Zinkevičius 1966: 237). Here the form of the o-, i- and u-stems was first phonologically assimilated to the illative plural, by analogy with the ā-stems as follows: Ill. Pl. -osna / Loc. Pl. -osu: Ill. Pl. -uosna (-ysna, -üsna) / Loc. Pl. x, x = -uosu (-ysu, -ūsu → -uose / -uosū after the o-stems), cf. Stang (1966: 186, 213, 218)¹¹. This also occurred in Latvian; cf. dārbūos. Among the -ā-stems, the old Indo-European locative plural ending has remained unchanged in only a few dialects; cf. Tverečius (Otręb ski 1934: 228) šakāšū, -ui, run̄kāšui. The -i here is by analogy with the shortened locative singular forms run̄kāi < rañkoje. In most dialects and in the standard language, final -u in the endings of all stems has been replaced by the -e of the inessive singular; cf. rañkose; in other dialects it has instead been replaced by the -a of the illative plural (Stang 1966: 186). Probably the plural form of the -ā-stems was renewed first: rañkosu > rañkose (replacement, not addition: **rañkosve), followed by the other stems. The formations with the extension *-ę́ are therefore recent in the inessive plural forms. 3. THE URBALTIC POSTPOSITION *-ēn 3.1. The following considerations lead to the reconstruction of the Uro-Baltic inessive postposition as *-ēn (< *ḗ): a) the Iness. Sg. form vaik-i in those dialects where *-en > i and *-an > u, as well as the extended forms of the adjective, cf. dangųje jis, point to a nasal in final position (cf. Stang 1966: 182); b) the intonation in the Iness. Sg. of the *o-stems must be reconstructed as acute on account of the effect of Saussure's law. This is also confirmed by the fact that Leskien's law (final accented *-ę́ > -e) must have operated in all inessive forms as well. Compare here the forms of the accusative and instrumental singular of the -ā-stems: Acc. Sg. rañką (< *-ań < *-am < *-eh₂m) versus Instr. Sg. rankà (< *-án < *-ān). If, then, one assumes that the acute intonation arose only in the singular of the -o-stems as a result of contraction, one faces the difficulty that the other, non-contracted, locative forms end not in **-ę (< **ēn), but in -e (< *ę < *ēn). Since the intonation can be reconstructed as acute, one may assume *-ēn for Urbaltic (Stang 1966: 128). 11 Zinkevičius (1966: 238) dagegen meinte, daß der Lok. Pl. auf -uose auf dem Akkusativ Plural und der Postposition *-ēn beruhe. Diese Meinung hat Stang (loc. cit.) mit schlagenden Argumenten kritisiert. Zu Stangs Erklärung könnte man noch einen formalen Grund hinzufügen: man findet keine Formen vom Typ **ratūose/ **ratuoše (2), wie sie nach Zinkevičius’ Erklärung ähnlich wie im Illativ Plural (s.d.) hätten entstehen müssen. Später schließt sich Zinkevičius (1982: 28) der Meinung Stangs an. 64 | I L J A S E R Ž A N T 3.2. Furthermore, one must take into account that postpositions in Lithuanian, and also in Ur-Latvian, underwent a particular weakening in comparison with the corresponding (prepositions or) prefixes, cf. Lith. dievóp(i) (not from -piē), Latv. mājup ‘homewards’ (< *pi, not *-pie, which would have become > **-pi) as against priē / (pie), Latv. pie, Slav. прu, Lith. raṅkon(a) as against Lith. nuõ, Latv. nùo, Slav. на. Leskien's law does not explain this weakening, since all the words just mentioned have circumflex intonation. This weakening must be very old, since it reflects the Indo-European ablaut. Lith.-Latv. nuo would have been shortened to nu in both languages, not to -na. The form *na can be explained as a Baltic full grade corresponding to *nō, interpreted as a lengthened grade (< (?) *noH, *nō). The Indo-European form *som ‘together’ also belongs to this group; in prefixed forms with nouns in Lithuanian, Latvian, and Slavic it shows the full grade, cf. Lith. sañdas, Latv. sùods (< *sā́nds), and elsewhere Latvian prefix sa- (< *sam-), Sl. сѫдъ, as against the prepositional form Lith. su (Latv. ar—not related) and Slav. съ. By analogy, one would expect a weakened postposition corresponding to the well-attested Latvian prefix je- (< *en-), namely *-in (< *-n), also represented in Lithuanian by the preposition į¹². But we have *é < *-én. This fact speaks, first, against the etymological connection between *-é and Indo-European *en, Latvian ie-, and, second, in favor of keeping the formation of the inessive singular of the ā-stems outside this ablaut system, because, as explained in 1.1 and 1.3, it had already been inherited in this form from the proto-language. 4. THE NON-BALTIC COMPARISON 4.1. The intralinguistic examination of the forms of the locative and inessive in Baltic shows that the inessive singular form of the -ā-stems (as well as the -ē-st. < *-ijā-st.) is by far the oldest among the inessive forms. Its age reaches back to the period before the loss of the laryngeals. From this form, the element *-é (or, abstracted: -je) spread to all the other locative endings, and thus a “marked locative,” i.e. inessive, arose in Uro-Baltic. The question of the origin and etymology of this postposition remains unresolved. Stang (1966: 182) connects it with the preposition *h₁(e)i(ī) (Latv. ie-kšā, Gk. ἐν, εἰς, Lat. in, etc.). The difficulty, however, lies in the fact that, as Stang (loc. cit.) himself also notes, length (*ēn) must be reconstructed for our postposition on the basis of the acute accent (3.2). The preposition *en, however, does not occur in this form in Indo-European. 4.2. On the basis of the original distribution (only Loc. Sg. of the -ā-stems¹³) and the phonological form¹⁴, I would like to place the Baltic postposition *ēn in direct 12 Girdenis (1980) stellt fest, daß in der Mundart bei Kretinga (im Ziwatas, XVIII Jh.) für die aukšt. Präposition į die Präposition y gebraucht wird, die auf *en zurückgeht. Daneben, jedoch seltener, kommt auch i = aukšt. į vor. Diese Allomorphie bezeugt m. E. nur, daß in dieser Mundart die ursprüngliche Verteilung der Voll- (*en) und Nullstufe (*n̥) nicht mehr aufrechterhalten ist. 13 In Old Indic, the postposition -ām originally also occurs with the ī-stems (devyām) and with the pronoun (tāsyām). The -ī-stems have not been preserved in Baltic, and pronominal inflection in the Baltic locative singular was replaced by nominal inflection (Stang 1966: 244-245). 14 Urbaltisches auslautendes *-m wird lautgesetzlich zu *-n (Stang 1966: 114). ZUR VORGESCHICHTE DES INESSIVS 65 This may be connected with the Old Indic postposition -ām (cf. in Old Indic śénā (f.) ‘army’, locative singular śénāy-ām), with the common proto-form to be reconstructed as Indo-European *ēm. The fact that this postposition occurs only in these two language families could be explained by its having been used, before the individual languages separated, only optionally alongside the locative form without a postposition, as must also be assumed for Proto-Baltic and even Proto-Lithuanian and Proto-Latvian—presumably it served to specify local relations. 4.3. The following facts support the antiquity of the extension -ām in Indic: a) the extended form is obligatory in the locative singular of the -ā stems not only in Classical Sanskrit but also in Vedic. As a rule, the later extended forms of other cases occur in only one of the two languages; b) die bisherige Erklärung (Wackernagel–Debrunner 1930: 121; Thumb-Hauschild 1959: 46–7) des indischen Elements -ām (in senāyām), die von einer etymologischen Zugehörigkeit zur avestischen Postposition (im Indischen Prä- und Postposition) -ā / ā ausging, stößt auf erhebliche Schwierigkeiten, und zwar muß diese Hypothese von einer doppelten Erweiterung – zunächst um -ā (das auch im Avestischen als Post- und Präposition bzw. Präfix fungiert), dann noch um ein unklares -am –, ausgehen. Bei einer solchen, erst indischen Kontraktion sollte prinzipiell auch die zweisilbige Lesung in den vedischen metrischen Texten auftauchen, da ja hier chronologisch viel frühere, indogermanische Kontraktionsfälle (z.B. Genetiv Plural *-o-om > *-ōm) zum Teil noch zweisilbig sind. Dies ist jedoch nicht der Fall. Oldenberg (1888: 185) erwähnt in seiner Liste aller Fälle, bei denen die zweisilbige Lesung im Vedischen vorkommt, den Lok. Sg. nicht. Whereas in Indic only the locative singular form of the -ā- and -ī stems contains the postposition *-ēm15, while the other locative forms are only occasionally specified by -ā, e.g. Vedic locative plural ápsv-ā, in Iranian and, above all, in Old Persian an “inessive” ā developed in a similar way to that in Baltic, with the help of the latter postposition, as can be seen from the fact that the postposition -ā had become obligatory with most locative endings in Old Persian. It therefore seems that in Avestan and Old Persian, in the locative singular of the -ā stems, the postposition -ā replaced the older postposition *-ēm, inherited in this form from Indo-European > **-ām. 5. CONCLUSION The investigation of the Proto-Baltic locative singular of the -ā stems leads to the following conclusions: 1) The preservation of the pre-vocalic sandhi variant of the locative morpheme *-eh₂i-V precisely in a pre-vocalic position indicates that the form of the inessive singular of the -ā stems in Baltic is extremely ancient. 2) The unextended ending must be reconstructed for Proto-Baltic on the basis of the adessive forms 15 Möglicherweise enthält nach Wackernagel–Debrunner (1930: 121) die Form des D. Abl. Instr. Du. auf -bhyām ebenfalls diese Postposition. 66 | ILJA SERŽANT may be posited as *-ai (1.2). Unfortunately, the type of intonation cannot be determined. However, one might suppose that this was a circumflexed *-ai. In this case, this ending would continue the Indo-European locative ending *-eh2-i before a pause or a consonant. 3) The linguistic-historical comparison of all the present-day inessive forms has shown that the form of the inessive singular of the ā-stems is the oldest, and that it formed the point of departure for the spread of the abstracted postposition *-ḗ or -ie. 4) Points 1 and 3 make it possible to connect the form of the inessive singular, Proto-Baltic *-ājēn, with the Old Indic locative form in -āyām, on the basis of phonological and distributional correspondence. This requires the assumption that the Old Indic distribution of the postpositions -ā and -ām, in contrast to the Iranian languages, represents the original state (lectio difficilior), that the generalization of the postposition -ā in Iranian for the marked locative is based on later analogy, and that the unextended locative form was also still alive in Indo-Iranian. BIBLIOGRAPHIE ARUMAA, P. 1930: Litauische mundartliche Texte aus der Wilnaer Gegend. In: Eesti Vabariigi Tartu Ülikooli Toimetused. Acta et Commentationes Universitatis Tartuensis, B Humaniora XXIII. Dorpat (Tartu). ENDZELIN, J. 1922: Zur baltischen Deklination der „ablautenden“ (i)jo-Stämme. Kuhns Zeitschrift 50, 13–35. ENDZELIN, J. 1951: Latviešu valodas gramatika, Rīga: Latvijas valsts izdevniecība. GIRDENIS, A. 1980: Dėl vieno prielinksnio formos XVIII a. šiaurės žemaičių kretingiškių tarmėje. Baltistica 16 (2), 111–115. KAZLAUSKAS, J. 1968: Lietuvių kalbos istorinė gramatika, Vilnius: Mintis. MAŽIULIS, V. 1970: Baltų ir kitų indoeuropiečių kalbų santykiai (Deklincija), Vilnius: Mintis. ME = Mülenbachs-Endzelīns: K. MÜLENBACHS, Latviešu valodas vārdnīca. Rediģējis, papildinājis, turpinājis J. ENDZELĪNS. 2. negröts izdevums. 6 Bde. Čikāga: Baltu filologu kopa, 1954. OLDENBERG, H. 1888: Die Hymnen des Rig-Veda. Met­rische und textgeschichtliche Prolegomena, Berlin: Wilhelm Hertz. OTRĘBSKI, J. 1934: Wschodniolitewskie narzecze Twereckie. Część I: Gramatyka, Kraków: PAU. RIX, H. 1976: Historische Grammatik des Griechischen. Laut- und Formenlehre, Darmstadt: Wissenschaftliche Buchgesellschaft. ROSINAS, A. 1999: Ines yvo ir adesyvo formų kilmės ir raidos klausimu. Baltistica 34 (2), 173–183. SERŽANT, I. 2003: Intonationen in den suffixalen und Endsilben im Lettischen. Synchronie und Diachronie. Baltu filoloģija 12 (1), 83–123. SERŽANT, I. 2004 К вопросу об образовании адессива. Acta Linguistica Lithuanica 51, 49–57. STANG, Chr. S. 1966: Vergleichende Grammatik der baltischen Sprachen, Oslo etc.: Universitetsvorlaget. THUMB–HAUSCHILD 1959: Handbuch des Sanskrit. Eine Einführung in das sprachwissenschaftliche Studium des Altindischen von Albert Thumb. II. Teil: Formenlehre. Dritte stark umgearbeitete Auflage von Richard Hauschild, Heidelberg: Winter. TOPOROV, V. V. 1968: Локатив в славянских языках, Москва: Наука. VIDUGIRIS, A. 1959: Kai kurios Zietelos tarmės ypatybės. Lietuvių kalbotyros klausimai 2, 195–230. WACKERNAGEL–DEBRUNNER 1930: Altindische Grammatik von Jacob Wackernagel. III. Band: Nominalflexion–Zahlwort–Pronomen von Albert Debrunner und Jacob Wackernagel, Göttingen: Vandenhoeck u. Ruprecht. ZUR VORGESCHICHTE DES INESSIVS 67 Vanags, P. 1994: Die Entwicklungstendenzen der Kasusendungen in den ältesten lettischen Sprachdenkmälern. Linguistica Baltica 3, 121–130. Zinkevičius, Z. 1966: Lietuvių dialektologija, Vilnius: Mintis. Zinkevičius, Z. 1982: Lietuvių kalbos postpoziciniai vietininkai. Baltistica 18 (1), 21–38. Zubátý, J. 1896: Baltische Miszellen. Indogermanische Forschungen 6, 269–307. Ilja A. Seržant Gauta 2004 11 04 Universität zu Köln Institut für Linguistik Historisch-Vergleichende Sprachwissenschaft D-50923 Köln [email protected]