Studies in the History and Language of the Sarmatians
J. Harmatta


3. The Sarmatian Dialects of the North Pontic Region


We see also a twofold development in the case of the Old Iranian groups of phonemes *-n- and *-m-. In Modern Ossetian the corresponding groups of phonemes are usually -on, -om (see Miller, Die Sprache der Osseten, 20); this, however, is the result of a fairly recent development, because in the Ossetian place-names which survived in the Balkar, etc. territories formerly inhabited by Ossetes we usually find a instead of o; the same position is revealed also by the other Ossetian loan-words in Balkar and Karachay (see recently Abaev, IAN SSSR OON 1935, 890). The names in the inscriptions, too, still preserve largely the group of phonemes -n-, -m- where


there is the same group in Old Iranian; it is very important to note, however, that we already come across names here which show the correspondence -on-, -om-.

Old Iranian *-n-, *-m- > -n-, -m-.

see above: < Old Iranian *aryana-; - Ossetian iron.

Tanais,  No. 14 (220 A. D.): ~ Ossetian andon 'Stahl' (Vasmer, op. cit., 32).

Panticapaeum, Latyshev, IOSPE II, 29: < Old Iranian *bānu-: ~ Ossetian bon 'Tag'.

Tanais,  No. 223 (228 A. D.): ~ Ossetian zärond 'ajt' (Vasmer, op. cit., 39).

Olbia, Latyshev, IOSPE I, 55: ~ Ossetian nomgin 'namhaft' (Vasmer, op. cit., 45).

see above: ~ Ossetian fidon 'väterlich'.

Olbia, Latyshev, IOSPE I, 75: ~ Saka sāna-, Sogdian s’n, Ossetian son 'Feind' (Harmatta, Ant. Hung. II, 35).

Sangibanus 'Prince of the Alans' Jordanes, Getica, 37: ~ Avestan saha- 'Spruch, Gebot, etc.' + bānu- 'Lichtstrahl, Strahl' = 'durch Gebote glänzend', (Vasmer, op. cit., 50): ~ Ossetian bon 'Tag'.

see above: ~ Ossetian fidon.

see above: ~ Ossetian iron.

Old Iranian *-n-, -m- > -ōn-, -ōm-.

Olbia, Latyshev I, 52: according to Justi (op. cit. 39) ~ Ossetian art + mon 'Feuergeist'. This interpretation is rendered unacceptable by the fact that Ossetian does not possess the word mon 'Geist'. The form quoted by Justi is only an etymological abstraction from the words dälimon, däluimon, etc. given by Miller (op. cit., 35); but these words must probably be divided as däl-uimon, etc. (On this point see Schmidt, FUFAn:. XVIII 95—6, XIX, 19; Abaev, IAN SSSR OON 1935, 885), Vasmer compares the name with the Avestan word arəϑamant- 'Kläger'; this should have developed however, into the form . The difficulties disappear, however, if we regard the name  as the development of the Old Iranian compound *arta-mana-, the elements of which correspond to the Avestan words arəta- 'Gesetz, Recht, heiliges Recht' and mana- 'Art und Weise' ('dessen Art und Weise das heilige Recht ist').

'a town beside the river Tyras' Ptolemy III 5,15: ~ Avestan maēϑana- 'Aufenthaltsort, Wohnung, Haus' (Vasmer, op. cit., 63).

Tanais,  No. 333 (beginning of 3rd cent. A. D.). Jacobsohn (KZ LI V. 273) has tried to explain this name from an Old Iranian form *syāvamaka-: this interpretation seems, however, unconvincing since we cannot demonstrate the phonetic change -āva- > -ō- from the names in the inscriptions. On the other hand, the word seems to be an exact equivalent of the Avestan name syāmaka- 'Name eines Bergs oder Gebirgs' (Bartholomae, AirWb., 1931). The corresponding form is known also from Old Indian: yāmaka- "dunkelfarbig" while the form yāma- (without the sutfix -ka-) is used also as a proper name. Thus the name  may be regarded as the development of a supposed Old Iranian word syāmaka- 'dunkelfarbig' and may be read as syōmaχ.

Additional Notes

To p. 92. The interpretation of Sangi- in the name Sangibanus by the help of Avestan saha- can hardly be correct. Nor is the attempt of Abaev to identify Sangi- with Ossetian cong 'hand' far better.

To p. 92. On the name  cf. L. Zgusta: Die Personennamen griechischer Städte der nördlichen Schwarzmeerküste. 70. I do not regard the separation  as convincing. Dividing, however, the text into , we can only restore the form .

Zgusta, op. cit. 269 tries to eliminate the examples for -an- > -on-. In his opinion  is grecized on the pattern of  etc. This assumption is, however, hardly correct. An Iranian form *mēton(i) could, of course, be transcribed as  in Greek, but if the Iranian prototype would have been *mētan(i), surely this would have been grecized on the pattern of  etc. or on that of  etc. Accordingly, the evidence of the name  for -an- > -on- remains valid further on.

In the case of  the argumentation of Zgusta is hardly comprehensible. He adopts my interpretation  < *Syāmaka-, but he refuses to use this name as an evidence for the phonemic development -am- > -om- for that reason that a in the name  also developed into ō among other circumstances. Where is a logic here?

To p. 92. The form syōmaχ cannot go back to Old Iranian *syāmaka-, but only to *syāma-aχva- 'evil-minded' (lit. 'dark-minded').

[Previous] [Next]
[Back to Index]