The goal of motion in the Indo-European languages: Dative and accusative

Acta Linguistica Lithuanica LVI (2007), 65 — 71 The goal of motion in the Indo- European languages: Dative and accusative’ WiLLiAM R. SCHMALSTIEG State College, Pa. The author's contention is that the attested IE accusative and dative sg. forms of o-stems go back to original sandhi variants of one single case ending. These phonetic variants were subsequently reanalyzed as differ- ent case forms. The Indo-European *o-stem accusative case results from the reinterpretation of some of the original semantic dative case uses as syntactic accusative case uses (in the following N = m or n). Etymological automatic phonological variants were reanalyzed, each variant gaining a separate syntactic significance. For the most part the accusative case is represented by *-oN (< *-oN in prevocalic sandhi position with retention of the final nasal) whereas the dative case is represented for the most part (but not exclusively) by *-6 (< *-oN in etymological preconsonantal sandhi position with loss of the nasal and lengthening of the preceding vowel), cf. Lat. acc. sg. serv-um (< “-oN < **-oN + V...) ‘slave’ vs. dat. sg. servo (<*-0 < **-oN + C...), see Schmalstieg (2000, 2004). A partial phonological and syntactic parallel is furnished by the English I should like to thank herewith Prof. Vytautas Ambrazas for reading an earlier version of this paper and helping with the Lithuanian examples. This does not imply, of course, his endorsement of the theory here proposed or his responsibility for any mistakes. ‘I accept the Indo-Hittite hypothesis and assume that this particular phonological development might not have taken place in Indo-Hittite. Nevertheless with regard to the nasals the Hittite development seems to be very similar to that which I have proposed for Indo-European. 65 WiLLiaM R. SCHMALSTIEG possessives my and mine, which now have different functions, although originally the distribution was guite automatic, my resulting from the preconsonantal sandhi variant and mine resulting from the prevocalic sandhi variant (Strang 1970: 198). The original distribution of morpheme alternants depending upon sandhi phenomena is still observed in the use of the English indefinite article, i.e., a before consonants, an before vowels. The Slavic -o stem dative singular ending -u derives from *-o (< *-uN < *-oN which, although from the Indo-European point of view existed only in prevocalic sandhi position, became possible again in Slavic preconsonantal sandhi position) with loss of the nasalization in root-stressed paradigms. This denasalized -u was then spread to all the -o stem paradigms regardless of stress pattern. (This would have a partial parallel in the situation described by Zinkevičius 1966: 75-78 for certain western Lithuanian dialects (Klaipėda, etc.) in which the retention or loss of a final nasal seems to depend on whether the final syllable was stressed or not.) The old original dative ending (*-oN-) is retained in the etymological adjectival and pronominal definite forms where it is protected by the following vowel, e.g., Slavic dat. sg. masc.-neut. t-om-u ‘to that’ (cf. the Lithuanian equivalent dat. sg. masc. t-am-ui ‘id.’).? The Slavic acc. sg. masc. definite form t-»-jb ‘that one’ results from a secondary restructuring. The Slavic -o stem acc. sg. ending -b also derives from *-uN (< *-oN) in specific Slavic prevocalic sandhi position. Thus *-uN (in prevocalic position) > -» vs. *-uN (in preconsonantal position) > *-¢. Evidence for the ‘According to Thumb-Hauschild (1959: 138) the Old Indic dat. sg. masc.-neut. tasmai, Umbrian esmei ‘huic’, pusme ‘cui’, Goth. pamma, Old Prussian stesmu, kas- mu have an extension in sm(a)-, whereas Lith. t-am-ui, Latv. tam, Old High German demu, Slavic t-om-u show the expected adjectival *-o stem declension with no trace of an original -s-. See also Haudry (1982: 30). Another possibility is also imaginable. Haudry (1982: 41) has written: “On nomme hypostase le fait de traiter une forme fléchie ou adverbialisée come une base de flexion ou de dérivation”. An example of this would be the use of the indefinite nominative of the adjective to which another case is added. Zinkevičius (1966: 283) gives some examples of the indefinite adjecti- val nominative case to which other case endings are added, e.g., šį pirm-ds-j jésim in karuze... ‘Let us ride this first one into war...’. Note the addition of the definite acc. sg. ending -į to the nom. sg. form pirm-as- (with stress as in def. nom. sg. pirm-as-is). One might assume for Old Indic dat. sg. kdsmai ‘to whom’ may derive from nom. sg. kdh, just as Old Prussian kasmu might derive from the nom. sg. kas. Other forms such as dat. sg. Old Indic tdsmai, Old Prussian stesmu might also be analogical. 66 The goal of motion in the languages etymologically automatic nature of the distribution of *-» and -o is furnished by the reflexes of the preposition and prefix *svn- which developed either into sb- or s9 — depending on the environment. In prevocalic position the final -n was transferred to the beginning of the following word, thus s»n- imb ‘with him’ is now analyzed as s nim ‘id.’ in modern Russian, whereas in preconsonantal position a nasal vowel resulted, thus so-prog-» ‘spouse’. The etymological prevocalic form was finally spread to all positions in the function of the preposition su, thus, e.g., sv gradoms ‘with the city’. The etymological preconsonantal form is retained in modern Russian compounds in etymological preconsonantal position, e.g., in the compound suprug ‘spouse’ (< *sg-prog-v) and other words with the prefix su-, e.g., su-stav ‘joint, articulation’, su-merki ‘twilight, dusk’, etc. After the split into two cases in the *o-stem class other noun stem classes (e.g., the consonant stems) assigned the old dative in *-N to the new accusative function and a form of the old locative in *-(e)i took over dative function, cf. Lat. acc. sg. homin-em <*-m ‘man’, dat. sg. homin-i < *-ei, Gk. acc. sg. mowpéva < *-m ‘shepherd’, dat. sg. mowévi < *-i. This etymological locative marker *-i was also added independently to the etymological *-6 to give standard Lithuanian -ui and in Greek to give -. As a relic of an earlier epoch before the *-o stem dative and accusative were differentiated the goal of motion may be expressed by either case in the oldest attested Indo-European languages, cf. Lat. domum (acc. sg.) ire ‘to go home’ (replaced in late Latin by in (ad) domum, where the preposition, added later, reinforces the original meaning), tollitur in caelum (acc. sg.) clamor ‘a shout is raised to heaven’ (Virgil, Aeneid 11, 745), but also it clamor caelo (dat. sg.) ‘the shout goes heavenward’ (Virgil, Aeneid 5, 451). Further Latin examples include qué ‘whither’, alio ‘to another place’, intro ‘to the inside’, ultro ‘to the farther side’, retro ‘backwards’, etc., all etymological -0 stem dative singular forms. (The etymological Latin qué when combined with the old locative marker *-i gives the dative sg. cui ‘to whom, for what’.) Greek examples include oikov (8opov) (acc. sg.) iévar ‘to go home’ (Hofmann & Szantyr 1972: 49), but also AA@ed (dat. sg.) péoow katapaic exadeooe Iloceidav' edpupiav ‘going down into the middle of the Alpheos he called upon widely ruling Poseidon’ (Pindar, Olympian Ode 6, 61-62). Delbriick (1893: 177) gives the Old Indic examples, gramam (acc. sg.) gacchati vs. gramaya (dat. sg.) gacchati ‘he goes to the village’. Danylenko (2003: 67 WiLLiAM R. SCHMALSTIEG 295) gives the following examples from Old East Slavic: pride Volodimers Haličeskoi (acc.) i pride Kyevu (dat.) “Volodimer came to Galicia and arrived at Kiev'. In later Slavic the accusative of direction was strengthened with a preposition (as in Latin), e.g., ve dom ‘into the house’. A relic of the old dative is preserved in the contemporary Russian adverb domoj ‘home’ < domovi ‘to the house’. In their discussion of this topic Gamkrelidze and Ivanov note as a typological parallel the fact that the direct and indirect object and the object of motion can all be expressed by the same case (-s) in Georgian (1984: 285- 286): k’aci azlevs c’ign-s bavsv-s ‘the man gives the book (c’ign-s) to the child (bav$v-s)’ and midis kalak-s ‘goes to the city’. According to Friedrich (1960: 120) the Hittite accusative is very rare and perhaps archaic in its usage as the object of motion. He gives as examples nu-smas HUR.SAG-an (acc. sg.) parhanzi ‘und sie werden euch ins Gebirge jagen, and they will chase you into the mountains’; GŪ-zu “*APIN-an (acc. sg.) Ser tizzi ‘sein Nacken kommt auf einen Pflug, one places his neck on a plow’ (Hoffner 1997: 133). The older language has a directive in -a for the object of motion as opposed to an -i to denote location at rest, thus arun-a ‘to the ocean’, nepis-a ‘to heaven’ as opposed to arun-i ‘in the ocean’, nepis-i ‘in heaven’. Ordinarily the directive is the case of the object of motion and the dative denotes location (Friedrich 1960: 121). In my view this Hittite directive in -a (not the dat.-loc. in -i) and the accusative in -an have a common origin. One would expect perhaps Hittite *-un < Indo-European *-on, but it is possible that the old *-un was replaced by -an on analogy with the nom. sg. -a (Kronasser 1956: 99). The alternation between final -a and -an is observed in the Hittite particles anda/andan ‘in, within, into’, appa ‘afterwards, again’ /appan ‘behind, after’, katta/kattan ‘afterwards’ (Kronasser 1956: 157-160). Tischler (1983: 539) writes that it is usually assumed that katta is an old directive, kattan an old accusative of direction and katti an old locative. Tischler also mentions the suggestion that katta may be derived from kattan with loss of final -n as a result of sandhi phenomena. See also Sturtevant (1927: 250). According to Friedrich (1960: 35) final -n is ordinarily assimilated to the initial consonant of a following enclitic (with or without the doubling of this initial consonant, cf. *iStamanan-San ‘his ear’ (acc. sg.) > ištamanaššan; man ‘if + the particle ua of quoted speech > madua. Friedrich notes also 68 The goal of motion in the languages (1960: 34) that n in the interior of a word is freguently omitted, e.g., kar- pa-zi beside karpanzi ‘they raise’. In this regard Sturtevant (1951: 25) writes, ‘It is probable that in some dialect of colloquial Hittite, n before certain consonants had been lost’. One can imagine then that originally word final -n was retained before words beginning with a vowel, but that word final -n was lost before words beginning with certain consonants. This original sandhi variation was morphologized such that -an came to serve as the accusative singular, whereas -a became the directive case. Whether the Hittite directive reflected a long or short final -a cannot be determined (see Laroche 1970: 46 and 48, fn. 36). In principle it could be long and the correspond exactly to Indo-European *-0. In her study of case and prepositional usage in the Gospel of Luke in Bretkūnas' Bible translation Gelumbeckaite (2002: 150) writes that the Latin use of the accusative to denote motion into the inside of something helped to retain the similar Lithuanian construction (Luke 7, 36), ieies [> ieijes] namus (acc. pl.) Pharifeia [corrected to > Pharifeufcho] ‘having entered the Pharisee’s house’ = Lat. ingressus domum (acc. sg.) pharisaei. But note, however, Lith. Einu einu ir prieinū mišką (acc. sg.) ‘I go and go and arrive at the forest’, (LKZ 1100); privagiavome dvarelį (acc. sg.) ‘we came to the small manor’ (KZ... 449); AS pats keliausiu tolimą Salele ‘1 myself will travel to a distant land’ (JD 2704); Lėkčiau žalią giruzéle “I would fly to a green forest’ (JD, 341); Ryto josiu jomarkélj ‘Tomorrow I will ride to the market’ (LKZ, 358, JD 521); Balnok, tėveli, bérq Zirgelj, josiu svečią Salele ‘Saddle up, o father, the bay horse, I will ride to a foreign country’ (LKZ 358, JV 985). The Lithuanian preposition į ‘to, in’ strengthens the simple accusative case in the function of object of motion just as ad, in do in Latin, see above.“ The use of the accusative to denote the object of motion is an ancient Indo-European syntactic feature. Gelumbeckaite (2002: 243) writes further that the Latin use of appropinquare ‘to approach’ + the dative stimulated a similar usage of Lith. artintis with the dative (Luke 24, 28), prifiartinaia miefteliui (dat. sg.) [corrected to > miefteliap kurap eija] ‘they approached the village, where they were going’ = Lat. appropinquaverunt castello, quo ibant. “My colleague, V. Ambrazas, suggests that the meaning of the accusative of direction gave rise to the illative with its strengthening element -na. 69 WiLLiAM R. SCHMALSTIEG As far as I know the dative as an object of motion is not retained in any modern Lithuanian dialect, but the dative in this function has an ancient Indo-European origin so it cannot be proven that it was impossible at an earlier epoch in Lithuanian. ABBREVIATIONS JD = Liėtūviškos ddjnos užrašytos par Antaną Juškevičę. 3 volumes. Kazan’, 1880- 1882. The numerals denote the number of the song. JV = Liėtūviškos svotbinės ddjnos užrašytos par Antaną Juškevičę ir iSspdudintos par Joną Juškevičę. St. Petersburg ... 1883. The numerals denote the number of the song. LKŽ = Lietuvių kalbos žodynas. 1-20 Ed. by J. Barčikowis, J. Kruoras, K. ULvypas, V. VirkaAuskas. Vilnius, 1941-2002. VD = A. VikErniūNas, Kupiskénos dainos (see the journal Tauta ir Žodis, III and IV). The numerals denote the number of the song. REFERENCES DANYLENKO, A. I. 2003: Predykaty, vidminky i diatezy v ukrajinks'kij movi. Istoričnyj i typolohičnyj aspekty. Xarkiv: Oko. DELBRūck, B. 1893. Vergleichende Syntax der indogermanischen Sprachen I. Strassburg: Karl J. Triibner. FrieoricH, J. 1960: Hethitisches Elementarbuch. Heidelberg: Carl Winter. GAMKRELIDZE, T. & V. Ivanov. 1984. Indoevropejskij jazyk i indoevropejcy. Tbilisi: Izdatel’stvo tbilisskogo universiteta. GELUMBECKAITE, J. 2002. Linksniy ir prielinksnininiy konstrukcijų sintaksė Jono Bretkūno Biblijos Evangelijoje pagal Luką. Vilnius: Lietuvių kalbos institutas. Haupry, J. 1982. Prėhistoire de la flexion nominale indo-europčenne. Lyon: Institut d'études indo-européennes de I'Université Jean Moulin. 70 The goal of motion in the languages HorMANN, J. B., A. Szantyr et al. 1972 : Lateinische Syntax und Stilistik. Munich: C. H. Beck Horener, H. 1997: The laws of the Hittites. Leiden-New York-Cologne: Brill. Kronasser, H. 1956: Vergleichende Laut- und Formenlehre des Hethitischen. Heidelberg: Carl Winter. LarocHg, E. 1970 : Etudes de linguistique anatolienne III: 9. Le directif. Revue hittite et asianique 28, 22-49, SCHMALSTIEG, W. R. 1997: The origin of the neuter nominative-accusative singular in *-om. Journal of Indo-European Studies 24, 3/4, 401-408. SCHMALSTIEG, W. R. 2000: Dative or Accusative. A Latvian parallel to Proto-Indo- European. In: J.-D. Range, Hrsg., Aspekte baltistischer Forschung (Schriften des Instituts fiir Baltistik, Ernst-Moritz-Arndt-Universitat Greifswald, Bd. 1), Essen: Die Blaue Eule, 291-300. ScHMALSTIEG, W. R. 2004. The common origin of the *-o stem dative, accusative and instrumental cases. Baltistica 39, 5-11. Strang, B. 1970: A history of English. London: Methuen and Co. SturTeEVANT, E. H. 1927: Hittite katta(n) and related words. American Journal of Philology 48, 247-257. STURTEVANT, E. H. 1951: A comparative grammar of the Hittite language. New Haven and London: Yale University Press. Tiscurer, J. 1983: Hethitisches etymologisches Glossar. Part I. A-K. Innsbruck: Institut fiir Sprachwissenschaft der Universitat. (Innsbrucker Beitrige zur Sprachwissenschaft, vol. 20, part 1.) Tuums, A. R R. HauscHiLb 1959: Handbuch des Sanskrit. Bd. II. Formenlehre. Hei- delberg: Carl Winter. ZINKEVICIUS, Z. 1966: Lietuvių dialektologija. Vilnius: Mintis. William R. Schmalstieg 814 Cornwall Road State College, Pa. 16803, USA 71