We use cookies to distinguish you from other users and to provide you with a better experience on our websites. Close this message to accept cookies or find out how to manage your cookie settings.
Published online by Cambridge University Press: 24 December 2009
A central Problem of Bactrian Phonology, to which Georg Morgenstierne devoted the greater part of one of his last articles, is the origin of Postvocalic t, which is found in a small number of words of disputed etymology. Since OIr. postvocalic t generally gives Bactrian d, as in G[raeeo-Bactrian] οδο, M[anichean Bactrian] 'wd, 'wṭ[ud] ‘and’ <*uta or G κiρδo, M qyrd ‘made’ <*kṛta-,it seems that some oter orgin must be sought for Bactrian t—unless, indeed, G δ and τ may be regarded merely as alternative notations of [d], as has been Proposed. Such a Possibility was first suggested by G oτo ‘and’, which accoriding to W. B. Henning is ‘not differne from οδο ’, and which I. Gershevitch has explained either as a historical spelling or as a generalized sandhivariant of οδο. However, a crucial objection to the identifieation of οδο and οτο lies in the fact, the implications of which have not yet been fully appreciated, that they differ not only in form but also in function.
1 ‘Notes on Bactrian phonology’, BSOAS, XXXIII, 1, 1970, 125–31
2 SeeGershevitch, I., ‘The Bactrian Fragment in Manichean script’, Acta Antiqua, XXVIII, 1980 [1984] 273–80Google Scholar
3 BSOAS, XXIII, 1, 1960, 51
4 Asia Major, XII, 1, 1966, 104, n. 24
5 IF, LXXII, 1–2, 1967, 45–6, where Gershevitch retracts his proposal
6 Harmatta, J., Acta Antiqua, XII, 1964, 445–6, recognized that oδo and oτo differ in their usage and rightly cpnclude that they are distinct words, but his etymological solution (oτo < *rta. = Sogd. rty) is phonolgically unsatisfactory and essentially arbitrary. In fact Sogd. rty is certainly a compound with -ty <*uti (see BSOAS, XLVI, 1, 1983, 46, s.v. hade)Google Scholar
7 This and the following rules have been deduced from the great Surkh Kotal inseruotion. with which the usage of the remaining inscriptions of the Kushan period seems to agree. In the (much later) MS fragments the distintion between the two words for ‘and’ has been lost. Thus, M ‘wd is found with pronominal suffixes (1 sg.'wd-wm, 3 sg. 'wd-yyẖ), while oδo and oτo appear to the interchangeable in the so-called ‘Hephthalite ’ fragments
8 Similarly in the Kara-tepe inscriptions, which belong to the Kushano-Sasanian Perioed: ωρℴμℴζδℴ ℴδℴ μιρ⃛ ℴδℴβρμιρℴ ℴδℴ⃛ ‘Ohrmuzd and mihr⃛and Burzmihr and ⃛’ (but the reasings are not entirely certain), cfLivŠic, V. A., Buddijskie peŠČery Kara-tepe v Starom Termeze (Staviskij, B. Ja.), Moscow, 1969, 77–8Google Scholar
9 This seems to be true also of ℴτι in the Ajrtam inscription (for which see Turgunov, B. A.LivṦc, V. A., and Rtveladze, E. V., ‘Otkrytie baktrijskoj monumental’noj nadpisi v Ajrtame’, ObṦāestvennye Nauki v Uzbekistane, 1981, 338ℴ48).Google Scholar
10 cf. D. Weber, If. LXXVI, 1971, 81ℴ2
11 cf. n. 6 above
12 cf.Morgenstierne, ,BSOAS,XXXIII,1, 1970, 129,Google Scholar who tentatively suggested a derivation from‘ *uta + iu (?) with early syncope’
13 The context does not exclude Henning's interpretation of τa (kaλδo;) as ‘until’ (BSOAS, XXIII, I, 1960, 54 with n.7), which has been generally accepted (ef. Gershevitch, IF, lXXII, 1–2, 1967. 55–6, Sims-Williams, BSOAS XXXVIII, I, 1975, 138, etc.), but, ϊαδο implies the existence of ϊα ‘then’, Occam' razor should be applied
14 This etymology of τα was proposed already by A.Maricq, JA, CCVLVI, 1958, 362, but its relationship to ταδο has not previously been recognized
15 BSOAS, XXIII, I, 1960, 50, n. 4
16 GMS, pp. 13(§ 96) and 307.
17 cf. D.N., Mackenzie. the ‘Sūtra of the causes and effects of actions’ in Sogdian, London, 1970, 46–7Google Scholar
18 IF, LXXII, 1–2, 1967, 46
19 Henning, BSOAS, XXIII, I, 1960, 54; Harmatta,Acta Antiqua, XII, 1964, 437; Morgen. stierne, BSOAS, XXXIII, I, 1970, 129
20 Contrast σιιο, if this represents *čim+uti (=Sogd, 'cwty) is an equally likely etymology, cf. nn. 27*–8 below
21 See BSOAS, XXXVIII, I, 1975, 136–9
22 cf, in particular Bactrian ιο/ιοα ατο ‘so that’ and Sogd, m'yο'ty, m'δ ZY, Later m't ‘id.’
23 See above, nn. 7–9.
24 BSOAS,XXIII, I,1960,49Google ScholarDifferently Harmatta,Acta Antiqua, XII, 1964, 444Google Scholar, and G.D., Davary, Bakrisch: ein Wōrterbuch, Heidelberg, 1982, 164Google Scholarneither of whom recognized the necessity of deriving –ðο from *–uti.Google Scholar
25 See the bibliographical indications apud Davary, op. cit., 210, 279–80.
26 The assumption that OIr.*–ah gave Bactr. –i (later–φ) is justified by the 3sg. preterites (in origion nom. sg. m. past participles) θιστι, πιδοριλδι, ζο[ο͝αστι in version ‘B’ of the Surkh Kotal inscription
27 cf. Av.čĪm, Khot. cu and ju. There is no need to reconstruct an unattested *āam (thus E.Benveniste,Essai de grammaire sogdienne, II, Paris, 1929, 132). Cf. also Studia Iranica, X, 2, 1981, 174–5, on traces of *čim in Middle Persian
28 I see no advantage in deriving κι– and δι– from genitive forms ([G.D. Davary and ] H., Humbach, Die baktrische Inschrift IDN 1 von Dasht-e Nāwūr (Afghanistan). Wiesbaden. 1976. 12)Google Scholar.
29 Acta Antiqua, XII, 1964, 439
30 cf. Davary, Baktrisch, 280–1. But ταδ– can harldy be OIr. *tat/tad as suggested by Harmatta, loc. cit., and Gershevitch,Asia Major, XII, I, 1966, 106 (with a significant qualification in n. 25a), since the final dental would hardly have been preserved as d
31 JA. CCVLVI, 1958, 353. No reliance can be placed on the reading or interpretation of the sequence –ιδο– in the inscriptions of Dasht-e Navur and Tochi (Davary, op., 181–2)
32 Greshevitch, Asia Major, XII, I, 1966, 96. Differently Humbach, op. cit., II, who again (cf. n. 28 above) starts from a genitive (*ahya)
33 At any rate, ειµο cannot simply represent *iyam or *ayam as proposed by Livṭic, , Drevnjaja Baktrija, II, Moscow, 1979, 108Google Scholar.Baktrisch, 183, since the preservation of final*–m would be unparalleled.
34 Asia Major, XII, I, 1966, 104, n. 24)Google Scholar). This etymology was in fact first proposed by Humbach, cf. his Baktrische Sprachdenkmaāler, I, Wiesbaden, 1966, 95.
35 Acta Antiqua, XII, 1964, 442–3Google Scholar
36 BSOAS, XXXIII, I, 1970, 128–9Google Scholar
37 BSOAS, XXIII, I, 1960, 53Google Scholar
38 Two asyndetic preterites, cf, οτηι…αχρτριλο κιρδο αλβαρλο ωστδο ‘and he made a water-wheel (and) installed a water-tank’
39 G θοβιχτο, M ηοyxt. ‘written’ <*ni-pixšta–
40 In view of the pres, stem φροχοαδρ– ‘to*withdraw’, whose š may go back to *rš or *rs, the 3 pl. pret. φροχορτιθδο ‘they *withdrew’ probably derives from a pp. in*–šta– (cf. Morgenstierne, BSOAS, XXXIII, I, 1970, 129, following suggestions by Harmatta and Gershevitch). Since none of the etymologies so far proosed is enitirely satisfactory, I am glad to be able to quote a suggestion of Dr. Gershevitch's that the Bactrian verg may be connected with Oss. āūrsyn, Dig, änxwärsun ‘to budge, make way, stand aside’, past stem änxwarst (cf. V.I. Abaev, Istoriko-ϋtimologičeskij slovar' osetinskogo jazyka, I, Moscow and Leningrad, 1958, 125, with references to other possible cognates within Ossetic). Dr. Gershevitch further suggests that G αμο– contain –mṛρτ– (cf. IF, LXXII, 1–2, 1967, 46, n.27)and M w(w)mw(w)rt may contain *–mṛšta–, to the root marz–. [On φροχοαρ–φρ–/φροχορτ– see now ch. de Lamberterie, Studia Iranica, XIII, 2, 1984, 203–5].
41 SeeGershevitch, , IF, LXXII, 1–2, 1967, 34–5.Google Scholar
42 SeeHenning, Z.V.,Togan'a Armaĝan, Istanbul, 1956, 259Google Scholar.
43 BSOAS, XXIII, I, 1960, 53.Google Scholar
44 Morgenstierne, ,BSOAS, XXXIII, I, 1970, 127; Davary, Baktrisch, 259.Google Scholar
45 See BSOAS, XLII, 1960, 133–4.Google Scholar
46 See BSOAS, XLII, I, 1960, 53, n.5.
47 One could of course resort to the assumption of an early dissimilatory loss of the second r.
48 Asia Major, XII, I, 1966, 105–6, with translation on p. 107.Google Scholar
49 On Chor prwz- see Henning, Asia Major, v, I, 1955, 45. Although it is beyond the scope of this note to discuss all possible instances of postvocalic t in late Bactraian, I cannot forbear mentioning the possibility that the sequence ωταδο in a ‘Hephthalite’ fragment (VII, 5)may be the equivalent of Sogd, ‘wšt't ‘stood’.
