J. Harmatta
III. THE LANGUAGE OF THE SARMATIANS
4. Conclusions
If we sum up the results of our observations we get the following picture
|
|
|
|
||||
|
|
|
|
|
|
|
|
| *ar-y | ar-(y)- | al- | ir- | il- |
 īr- |
īr- |
| *au | au | ō | ō | u | ||
| *sp | sp | sf | fs | fs | fs | |
| *p- | p- | f- | f- | f- | ||
| *fri- | fli- | li- | li- | lə- | ||
| *ri- | r | l | l | l | ||
* n |
-n |
* n |
-on | -on | ||
| *m |
-m |
*m |
-om | -om | ||
| *gr | gr | rg | rγ, lγ | rγ, lγ | ||
| *vi- | vi- | i- | i- |
 |
||
| *ha- | ha- (χa-) | a- | χa-, a- | χa-, a- | ||
| *χš | χš- | š- | χs- | χs | ||
95
Thus the examination of the Pontic Greek inscriptions and the Iranian names preserved in classical sources on South Russia clearly shows that, as early as the first centuries A. D., the language of the Iranian tribes inhabiting the steppes of Eastern Europe was by no means homogeneous. The phonemic differences appearing in the names amply prove that these tribes spoke several dialects, obviously corresponding to the nature of their tribal division. This fact is important for several reason. First of all, the picture which we derive, after examining these names, about the linguistic condition of the Iranian tribes in South Russia is in entire harmony with the observations made by us above concerning the language of the Median and Persian tribes. Secondly, this result enables us to approach the examination of the realationship between the Finno-Ugrian and the Iranian languages from a new angle: the realization that there existed several Iranian languages or dialectes in South Russia will make it possible to interpret several phonemic features, hitherto unexplained, in the Iranian loan-words of the Finno-Ugrian languages.
Naturally, it would be extremely important, both for clearing up the
dialectology of the Iranian languages and the historical background of
Finno-Ugrian and Iranian linguistic connections if we could give a precise
ethnical delimitation to the various Iranian dialects. But this is a very
difficult task. At present most of the names cannot be assigned to any
definite tribes nor do we know how the various phonemic differences crystallize
into structural features which separate the dialects from one another.
But the problem is not insoluble. First of all, we can gather some indications
from the names themselves. Thus e. g. those dialects which
give us the names 
and 
,
in spite of the difference in the development of Old Iranian initial *p-
are nevertheless united by certain common features since they show a similar
development of the Old Iranian group of phonemes -ϑr-.
The same development of the Old Iranian group of phonemes -ϑr-
is seen, however, also in the name 
< 
,
so that we may assume its close connection with the former dialects. Moreover,
since in the people's name 
we see the same development of the Old Iranian initial group of phonemes
*ary- as in the name ,
this word, too, must be included in this group. Thus we are beginning to
see the outlines of a group of dialects which, on the strength of certain
phonemic criteria, is connected with Ossetian though it is clearly distinguishable
from the latter by other phonemic phenomena. At the same time, there are
some Sarmatian dialects which are fairly distant from either group (cp.
e.
g. Iazygian 
and Thatean 
).
Besides these phonemic connections arising from the names themselves, a
careful comparison of the geographical distribution of the names with the
historical sources, as well as the examination of the historical and ethnical
conditions in the various Greek settlements, a task recently attempted
by Knipovich in his book on Tanais (
, 
, 
1949) — all this will make possible the ethnical and historical evaluation
of the linguistic differences established in this essay. This much we may
safely say, in any case, that on the ground of phonemic criteria alone
one may distinguish at least four languages or dialects: through the various
concatenations of phonemic peculiarities this number will be doubled.
The fact that the Iranian tribes of South Russia spoke several languages or dialects, clearly distinguishable from one another, as early as the first centuries A. D., has important consequences in clearing up the linguistic relations between Sarmatians, Alans, and Ossetes. Although this question may be solved only by a close examination of Alanic linguistic remains and the history of Ossetian phonemes, we may confidently state that the simple, generalizing historical identification of the language of
96
the Scythians, Sarmatians, Alans, and present-day Ossetes is not a probable
proposition. Since the names hailing from the territory of the Alans as
well as of the other Sarmatian tribes point to the existence of several
dialects, it is obvious that the language of the Sarmatians or that of
the Alans as a whole cannot be simply regarded as being Old Ossetian. Moreover,
some of the Sarmatian dialects show certain phonemic peculiarities (e.
g. -ān
> -ōn)
which are quite recent developments in Ossetian. The same situation prevails
also in Alanic. Thus e. g. the name 
(Tanais 225 A. D., 
,
No. 327) shows already the change from Old Iranian, syāva-
to the form sau which is characteristic of Ossetian. At the same
time, however, we see in this word also the change -ān
> -ōn
which is a much later development in Ossetian. Thus this Alanic name from
Tanais (
:
sāuanōn
< Old Iranian *syāva-nāna-),
together with other names pointing in the same direction, is a clear proof
that Ossetian only represents the outcome of a single Alanic dialect group,
the historical development of which was different from that of the Sarmatian
dialects attested by the Iranian names occurring in the inscriptions of
the North Pontic Greek cities.
[Previous] [Next]
[Back to Index]