Was ‘Scythian’ an ‘Iranian’ language?

Most people take it for granted that all ‘Scythians’ were Iranian-speakers but few are aware of the quality of the evidence on which this theory is based. There are no direct records or inscriptions in ‘Scythian’ and only a few words and names reached us in the accounts of early Greek historians (Herodotus, Strabo and others). The first extensive attempt to ‘reconstruct’ the ‘Scythian language’ was done in Vasmer (1923) but the largest vocabulary, numbering 247 ‘bases’ or radicals (approx. 200 considering duplicates), was given in Abaev (1949).

Both works are difficult to access since not everybody is proficient enough in German and Russian so I would like to bring up the Abaev’s dictionary for your attention. I have translated some of the entries so that you could decide whether you find his etymologies credible or not. Before proceeding with this I would like to present here a part of the introduction to the ‘Scythian’ dictionary by Abaev (1949:150-151) in the English translation (Rus. original is available here) with my comments (which are, of course, optional):

SCYTHIAN vocabulary elements are scattered in proper names, toponymic and tribal names, preserved for us in many historical, geographical and epigraphic sources. In the following we try to extract them and arrange in the alphabetical order. We shall thus obtain some rudiments of the SCYTHIAN vocabulary. (See the list of Greek words which Abaev used)

It should be noted that etymologies built on proper names, toponymic and tribal names and appellations are most unreliable and controversial. Almost every such etymology is contested. Etymologies of  the names ‘rus’, ‘slav’, ‘wend’ are good examples.

Although the preconditions for our endeavour have been done in works of our predecessors, yet we meet with difficulties. The transfer of the Scythian names with the Greek (or Latin) letters was, by force, very imperfect: too great was the difference in the phonetic systems. It is easy to imagine that there were frequent cases of distortion, misprints and errors. To restore, in these circumstances, the true face of the sounding of Scythian is not easy.

A very fair admission, indeed!

But if it is difficult to reproduce the real shape of Scythian words, the situation is even more difficult with their meaning. Our sources usually do not give us a hint about the meanings of Scythian words or names. These meanings must be established by the comparative way, based on the data of the Iranian and Indo-European linguistics.

Another frank admission, especially if we consider that at the time this text was written the “Yaphetic theory” of  Nicholas Marr, which rejected the traditional methods of “historical reconstruction”, was still very much alive. In fact, one will not find any reference to reconstructed IE *forms in Abaev’s vocabulary. Instead, there are numerous references to Marr in the book.

The main instances to which we can appeal, are, on the one hand, the ancient Iranian languages (Avestan and ancient Persian), on the other the Ossetic language, as the direct successor of SCYTHIAN.

Note that there is not a shred of doubt in Abaev’s mind that the Ossetian language is “the direct successor of SCYTHIAN”. Before Abaev even approached to the dictionary, he took it for granted that ‘Scythian’ was an Iranian language, therefore, the “SCYTHIAN vocabulary elements” which are “scattered in proper names, toponymic and tribal names” should only be viewed exclusively through Avestan, ancient Persian and, of course, Ossetic. This  attitude is not surprising because Abaev was Ossetian and  the Scythian origin of Ossetian was his idée fixe.

We had to contemplate more than once: what forms to show in our ‘dictionary’ as the cardinal ones? The sources from which we draw our material cover a vast period of about 1000 years. Of course, during this time the language was changing.

A very reasonable question. It is difficult to imagine that the Scythians, which was for the Greek a generic name applied to any ‘barbaric’ (in their opinion) people living north of their confines, had remained a single ethnos speaking a uniform ‘Scythian’ language over the vast territory for a thousand of years.

We often meet the same base-form at different stages of phonetic evolution. What form shall we put as the base-form in our dictionary?

This is another reasonable question. To answer it one has to establish first what were the ‘different stages of phonetic evolution’, what phonetic changes happened at different ‘stages’ and why.

We have found it most appropriate to propose as the original forms normalised ancient Iranian ones taking into consideration also the peculiarities of the SCYTHIAN group.

There are several questions I would like to ask at this point. What were the criteria to decide what was the ‘most appropriate’? What are the ‘normalised ancient Iranian forms’? How do we know about ‘peculiarities’ of the ‘SCYTHIAN group’ even before approaching to the analysis? Finally, what exactly is the ‘SCYTHIAN group’?

In other words, we are putting forms that we believe, would coincide with the ancient Scythian ones if the latter existed.

What a bold admission! In my view, this phrase should be placed as a sort of a disclaimer: ‘Please be warned that this dictionary is wholly based on our subjective opinion’.

In cases where there was not enough data for reconstructing ancient Iranian forms, we put the most archaic of the known attested forms.

It becomes clear by now that by ‘ancient Iranian forms’ Abaev, probably, meant a precursor of Gathic Avestan, otherwise, why reconstruct them? Therefore, he de-facto presupposed that ‘Scythian’ derived from an ‘Iranian’ language older than Avestan which, would be the ‘pra-Indo-Iranian’. This admission needs to be analysed in the context of the current theory.

There are conflicting views on this. I, for one, have my own opinion, but let us stick to what is called ‘the mainstream theory’ according to which the eastward migration of pra-Indo-Aryans most probably originated in the Pit-grave culture on the northern shore of the Black Sea, supposedly, at the end of the 3rd and the beginning of the 2nd millennia BC (Mallory 1989, Kuz’mina 2007).

It is generally accepted that the Pit-Grave culture was an organic Eastern continuation of the earlier Tripolye (Cucuteni) culture (4th millennium BC) and that it did not cease after the hypothetical departure of the proto-Indo-Iranians first to southern Siberia (Andronovo culture?) and then to today’s Iran and Hindustan. Therefore, we can reasonably presuppose that in the Circum-Pontic area there always remained an ethnos directly continuing the ‘proto-Sanskrit’ (non-Iranian) dialects (Трубачев 2003, 51). (See my post “Iranian loans in Slavonic“)

In his analysis Abaev completely ignored any possibility that on the vast territory of what the Greeks referred to as ‘Scythia’ could also live numerous IE tribes who never migrated to Iran and, therefore, never returned from there bringing with them the already changed ‘Iranian’ dialects. According to Trubačev (idem), they were represented by the so-called ‘old Scythians’ while the newer returning wave of Iranians constituted the ‘young Scythians’. The biased ‘pro-Iranian’ approach is clearly seen in the next Abaev’s assertion:

We include in our lexicon also the Scythian names which were witnessed outside the Scythia – Thrace, Iberia, etc., as long as they have luminous (lit. “яркие” [bright]) signs of their Scythian origin.

We do not know anything about the Scythian language so how  can we see its “luminous (lit. “яркие” [bright])” signs? Following this logic any appellation or name found in ancient texts relating to a vast territory spanning from the Altai to Iberia should be included here as long as it appeared ‘Scythian’ enough to Abaev.

In our interpretations of Scythian names and appellations, we are partly contiguous with our predecessors, partly give new ones. In the first case, we refer to the comprehensive paper by M. Vasmer Die Iranier in Südrussland (1923), in which whoever is interested can find further bibliographical details. We avoided strained and dubious affiliations and interpretations, and ventured to do the analysis only when we felt solid ground under the feet.

In plain words: ‘we selected only what corresponded to our views’.

Yet in some cases, marked with a question mark, we assume the possibility of other interpretation than ours.

Effectively means: ‘we allow for other interpretations only in a few cases, otherwise our interpretations are rock-solid’. Let us analyse one of such ‘doubtless’ etymologies. On page 154 Abaev proposes a ‘Scythian’ word ar– :

ar – ‘to find’, ‘to procure’, ‘to give birth’; Ossetic. aryn | jerun (from bar-→war-?)

– Γώαρ ‘Alanian leader in the 5th century BC.’ (Olympiodorus and others) = go-ar ‘procuring cattle’; see. gau ‘cattle’;

– Ξησσάγαρος (0) = Osset. xsæz-sag-ar ‘one who procures six deer’ i.e. ‘lucky hunter’; see xšas ‘six’ sāka ‘deer’;

– Sagaris (Ovidius), Saggarius (Plinius) ‘river Berezan’ lit. ‘where they find  (аg) deer (sag)’.

Perhaps there are people who would accept this but for me this sort of reasoning amounts to sheer speculation.

We  usually take Ossetic parallels from the Digor dialects as the Digor mountain forms are generally more archaic, and, therefore, closer to Scythian.

Ossetic is a generally accepted as an Iranian language although it can hardly be considered a true genetic continuator of ‘Iranian’ having become an agglutinative language with a phonetic system closely resembling Caucasian languages and only a limited inherited IE lexicon.

Genetically, Ossetians are also similar to their Caucasian neighbours with only traces of the male M17 (R1a1) haplo-group which is typical for the Scythians: “Y-haplogroup data indicate that North Ossetians are more similar to other North Caucasian groups, and South Ossetians are more similar to other South Caucasian groups, than to each other” (Nasidze et. al. 2004). The only link connecting Ossetians with Iranians can be found in their maternal mtDNA: “with respect to mtDNA, Ossetians are significantly more similar to Iranian groups than to Caucasian groups” (idem). Therefore, by their genetic profile Ossetians are a typical mixed people and their Iranian affiliation can only be established from the maternal side. This is confirmed by the clearly mixed character of the Ossetic language.

The extreme ‘Osseticocentrism’ lead to some truly preposterous assertions. Such was Abaev’s theory that the Hungarian aladar and the Mongolian (!) aldar were Alanian (read Ossetic) loans:

ardar ‘master, ruler, knjaz’; Oss. ældar, ærdar id. From ærm–dar(←armadāra) ‘hand-keeper’, ‘Handhaber’; from Alanian adopted into Hungarian (aladar) and Mongolian (aldar): […].

Let us now move to the dictionary which Abaev entitled as ‘Dictionary of Scythian bases’. The full text of the dictionary can be found here. Since the dictionary is in Russian I translated some parts taken randomly. In my opinion they may give an understanding of the kind of argumentation used by Abaev.

(Capital letters in parentheses mean location: G – Gorgippia, O – Olvia, P – Panticapeai, T – Tanaida).

bala ‘military force’, ‘team’, Old. Ind. bala-, Oss. bal:

– Πάλοι ‘Scythian tribe’ (Diodorus. Sid.);

– Ούαστόβαλος (Т) varzta-bala ‘loved by the team [meaning military team or troop]’;

– Ουαρζβάλακος (О) = Old. Osset. warz-bal ‘one who loves the team’ or ‘loved by the team’; see vārz ‘to love’ (ср. Miller “Эпиграфические следы…”, ЖМНП, 1886, окт., стр. 254).

bānu ‘light’, ‘day’; Avest. baxta, Osset. bon ‘day’, ‘force’, ‘power’, ‘wealth’, Alanian ban (παν) in the Alan. salutation ταπαγχας (tä ban xwarz) ‘good day’ in Byzantine writer Tzetzes (see. below, page. 256):

– Βάνας (P);

– Sangibamis ‘Alanian tzar of the V century in Gallia (Jordan) = čаn-gi-ban, Osset. congi-bon “possessing power” (bon) in hand (соngi)”; see. čang ‘hand’.

axšaina ‘blue, dark grey’, O.Persian, axšaina-, Ossetic. œxsin id., œxsinœ‘pigeon’:

– ΄Αξεινος in the name of the Black Sea Πόντος ΄Αξεινος (Pindar,. Ευριπιδ, Strabon) ‘blue Pont’, has later been rethought as Πόντος Ευξεινος ‘hospitable Pont’, this witty explanation belongs to Vasmer.

– Χαράξηνος (O) = xar-axšen ‘dark-grey donkey’ (Old. Iranian хаra, Oss.хœrœg ‘donkey’); this name might not contain anything humiliating; cp. Pers., gūr‘wild donkey’ as an epithet adorning a king (Bahrām-Gūr); such names were also given for magic purposes so as to render a child unattractive for bad spirits; for example the Ossetic name Kuʒœg from kuʒ ‘dog’ is of such origin.

az ‘to chase’, ‘to rule’, Avestan az-:

– Νάβαζος (Т) = Old. Iran. nav-āza– ‘steersman’, ‘one who steers a ship’; see.nav ‘ship’ (Vasmer, 45).

bad– ‘to sit’, Probably from Old Persian upa-had– (Miller), Oss. badun ‘to sit’:

– Βάδαγος (О) = Oss. ‘one who sits’ (used as a proper name to this day); here also belongs Βαδάκης (О) (Vasmer, 35).

baga ‘god’, Old. Pers, baga-, Avest. baγa-; not preserved in Ossetic:

– Βάγης (G), reduced form of some name, containing baga, for example Bagadata ‘god-given’ etc., (Vasmer, 35); here also probably belongs Βάγιος (G).

Even if we accept these dubious etymologies, ‘Scythian’ emerges here not as much as an ‘Iranian’ but, primarily, as an Indo-European dialect. It is significant that out of the 200 radicals given by Abaev at least 70 (more than a third!) have clear Sanskrit and Slavonic cognates. Here are only the most obvious ones:

Abaev Meaning Skr. Skr. Meaning Rus Rus. meaning abra sky abhra cloud, sky obla(ko) cloud ad to eat ada food eda food agar many, abundant agra foremost, prominent ogro(men) great, big aluθ beer ali spirituous liquor ol spirituous liquor arma hand īrma arm ramo shoulder arta deified truth artha aim, purpose; art sense ašta eight aṣṭā eight osm’ eight baga god bhaga dispenser, Lord bog god bala war-force bala power, strength bole more bora(a) yellow, yellow-brown bhūr earth bur brown brin to shave bhrī to injure bri- to shave da to give dā to give da- to give dar to hold der- to hold darana to demolish, break daraṇa cleaving, breaking dran’ smth. torn darga long dīrgha long dolog long dasa ten daśa ten desjat’ ten dvara door dvār, dvāra door dver’ door fri to love pri to please, gladden pri- to treat well, like garu heavy guru heavy, weighty gru(z) weight gasti guest goṣṭhī assembly, meeting gost’ guest gau cow gava cow, cattle gov- bovine hvarz good hṛṣu glad, happy xoroš good kuti dog kutja little dog madu honey madhu honey mjod honey mān to think, to want man to think, imagine mni- to think, imagine mār to kill mṛ to die, kill mor- to kill māta mother mātā mother mat’ mother mizda reword mīḍha prize, reword mzda reword, bribe nava new nava new nov new pada foot-print pāda foot, bottom pod under, sole panti way patha path, way put’ path, way paradata to establish parādā to give over, deliver pereda- to give over pari- through pari beyond, out of pere- over, beyond pasu sheep paśu cattle , kine pas- to graze pay- to drink pāya water poi- to give to drink porata river (big) paratā highest degree porato much, many sata hundred śata hundred sto hundred šati happiness śita satisfied syt full (not hungry) stāna staying sthāna abode, dwelling stan place of staying syāva black śyāva dark-coloured, dark siv dark-grey taka thread, cloth tka- to weave tap to worm up tap to make warm tap- to worm up ugra- strong, powerful ugra mighty, violent, ugro(za) thread, danger vāč word vaca speech veče council varka wolf vṛka wolf, jackal, dog volk woolf zevak lazy jabh to one’s jaws wide zevaka gaper zura force sāra force, energy sila force

It is well known that “Historical linguistics often resorts to generalizations based on limited evidence, making statements that are far from obvious and often subject to discussion and various interpretations.” (Wencel 2011). The study by Abaev, in my opinion, is a typical case of such ‘generalization based on limited evidence’.

The very Term “Scythian” has been correctly defined by Adrienne Mayor et al. as: ““Scythian,” a fluid term even in antiquity, does not describe a single ethnic group but is a conventional collective term for the extensive network of loosely connected, culturally similar peoples of the vast territory of “Scythia,” which stretched from the Black Sea and Caucasus region to Central Asia.

There is no doubt that some Iranian tribes and elements were present in the Circum-Pontic area in the last centuries B.C. and first centuries A.D. but it would be very short-sighted to call the whole multitude of peoples populating the vast area from the Altai mountains to the Baltic Sea,  whom the Greeks routinely named ‘Scythians’, as ‘Iranians’ and even more so to view them exclusively through the Ossetic language.

References

Абаев, В. И. (1949), Осетинский язык и фольклор, Ленинград: из-во Акад. Наук СССР.

Трубачев, О. Н. (2003), Этногенез и культура древнейших славян: Лингвистические исследования, Москва: Наука.

Kuz’mina, E. E. (2007), The origin of the Indo-Iranians, Leiden, The Netherlands; Boston: Brill.

Mallory, J. P. (1989), In Search of the Indo-Europeans, London: Thames and Hudson Ltd.

A. Mayor. J. Colarusso, and D. Saunders (2014), “Making Sense of Nonsense Inscriptions Associated with Amazons and Scythians on Athenian Vases”, Hesperia 83.3.

Nasidze, I., Quinque, D. Dupanloup, I. Rychkov, S. Naumova, O. Zhukova, & O. Stoneking, M. (2004), “Genetic evidence concerning the origins of south and north Ossetians”, Annals of Human Genetics 68.588–599.

Vasmer, M. (1923), Untersuchungen über die ältesten Wohnsitze der Slaven, Leipzig, chapter “Die Iraner in Südrussland”.

Wencel, Maciej (2011) “Making archaeology speak – archaeology and linguistics“.ANTHROJOURNAL – The Collegiate Journal of Anthropology, 1, October 2011.