Know your Narts: cattle rearing and cattle raiding

« previous post | next post »

We here at Language Log know our Ossetians:  see "Know your Ossetians" (2/17/20), and be sure to read the informative comments to that post.  Today, let us go one step deeper into their language and lore.  We shall do so through getting to know some basic things about the Nart sagas (Abkhaz: Нарҭаа ражәабжьқәа; Nartaa raƶuabƶkua; Adyghe: Нартхымэ акъыбарыхэ; Nartxıme aqıbarıxe; Karachay-Balkar: Нарт таурухла; Nart tawruxla; Ossetian: Нарты кадджытæ; Narty kaddžytæ; Nartı kadjıtæ) are a series of tales originating from the North Caucasus. They form much of the basic mythology of the tribes in the area, including Abazin, Abkhaz, Circassian, Ossetian, KarachayBalkar, and to some extent ChechenIngush folklore.

The term nart comes from the Ossetian Nartæ, which is plurale tantum of nar. The origin of the root nar is of Iranian origin, from Proto-Iranian nar for 'hero, man', descended from Proto-Indo-European *h₂nḗr. In Chechen, the word nart means 'giant'.

Source:  Nart saga

On his blog, A Canadian in Ossetia, Life in the central Caucasus (4/28/20), Richard Foltz has an informative, insightful article on "The Narts: Ossetia’s National Epic".  Here is an especially revealing passage:

The Nart tales are known all across the Caucasus region and are held by many of its peoples—most of whom, such as the Circassians and Chechens, are not Indo-European— as their own. However, a comparison of the Nart cycle with other epics clearly demonstrates that it belongs within the Indo-European tradition. Centuries of cohabitation among the diverse peoples of the Caucasus have added many later elements borrowed from the Adyghe, Vainakh, Turks and others, but it is the version preserved by the Ossetes, whose northeast Iranian language and culture can be traced back to the ancient Scythians, that can be considered to contain the original core of the tales of the Narts. As noted by the Harvard-based comparative mythologist Calvert Watkins, the pastoral-nomadic world that gave birth to the Indo-European epics was one in which the bold exploits of the “hero” —whose principal achievement was to steal cattle from enemy tribes—were celebrated and memorialized by the “poet”, who, recompensed in his turn by gifts of that same livestock, was “the highest-paid professional in his society.” The Nart stories preserve this value system more directly and obviously than any other Indo-European epic tradition: the most heroic thing a man can do is to rustle cattle to bring back home to his community, and he does so in the hope that this will earn him “everlasting fame” (an outcome which depends upon the skill of the poet).

The theme of cattle-raiding is essential to the Indo-European epic tradition:

The act of cattle-raiding is quite ancient, first attested over seven thousand years ago, and is one the oldest-known aspects of Proto-Indo-European culture, being seen in inscriptions on artifacts such as the Norse Golden Horns of Gallehus and in works such as the Old Irish Táin Bó Cúailnge ("Cattle Raid of Cooley"), the paṇis of the Rigveda, the Mahabharata cattle raids and cattle rescues; and the Homeric Hymn to Hermes, who steals the cattle of Apollo.

Source:  Cattle raiding

See also:

Bruce Lincoln, "The Indo-European Cattle-Raiding Myth," History of Religions, 16.1 (Aug., 1976), 42-65.

Wikipedia, "Proto-Indo-European mythology"

In connection with our translation and study of the I ching / Yì jīng 易經 (Classic / Book of Changes), my brother Denis and I have been discovering evidence of cattle raiding in Asia.  It's not just in the I ching, but may also be found in other early sources.


Selected readings


  1. Victor Mair said,

    June 6, 2020 @ 12:43 pm

    From Clifford Coonan:

    Interesting — the Irish for ladybug (ladybirds in Ireland) is bóín Dé – little cow of God.

  2. Tom Dawkes said,

    June 6, 2020 @ 1:38 pm

    John Colarusso has written widely about the Narts and the Kabrdian language. His 'Grammar of the Kabardian language' [downloadable] has a text about a Nart hero, with grammatical analysis and a translation, and the book 'Tales of the Arts' is available in part as a Google book. Search Scholar Google under 'Colarusso Kabardian' and 'Colarusso Narts'

  3. Victor Mair said,

    June 6, 2020 @ 2:03 pm

    @Tom Dawkes

    Thanks so much for mentioning John Colarusso and several of his important works. He is a leading scholar of the Narts and some of the languages in which they are written.

  4. J. L. Wei said,

    June 6, 2020 @ 3:26 pm

    I don't know if this is germane to the discussion. Victor's post says, if I understand correctly, that cattle-raiding in the old Indo-European tradition was extolled and celebrated, and that there is evidence of cattle-raiding in the Yi Jing.

    Some years ago I wrote a paper linking the Yi Jing trigrams to Indo-Europeans or Proto-Indo-Europeans. The paper, "The Names of the Yi Jing Trigrams" (published in Sino-Platonic Papers, no. 161, now free online) maintained that the 8 Chinese names of the 8 Yi Jing trigrams were originally Indo-European or Proto-Indo-European words. In the paper I tried to explain that two of the names, namely GAN 乾 and KUN 坤 came from Indo-European or Proto-Indo-European words for "heaven" and "earth, land" respectively. I now revise that view and maintain that
    GAN 乾 "heaven, creator, father, male principle" came from the Proto-Indo-European root GEN- in *gen-e-ter "parent", whence Latin GENITOR "progenitor, creator, father", and that
    KUN 坤 came from the Proto-Indo-European word *G(w)EN "woman, wife" whence Anglo-Saxon CWEN "woman, wife".

    The Chinese GAN 乾 for "heaven, progenitor, creator, father, etc." and KUN 坤 for "earth, creatrix, woman, etc." would come from the Indo-European concepts of a Sky-Father and an Earth-Mother.

  5. martin schwartz said,

    June 6, 2020 @ 8:06 pm

    As to the etymology of Oss. nart(æ): While in contemporary Ossetic
    there is unquestionably a word nar 'hero', and in fact Prof. Foltz
    sent me a picture of the Nar Hotel in Vladikavkaz, where he was married, the word is a recent extraction (by back-formation) from
    nart(æ), which has the appearance of a plural, interpretable
    from context as a plurale tantum. In fact the word is not listed in
    Vsevolod F. Miller's Ossetic-Russian-German dictionary (1929-1934), which very much quotes from Nart tales and whose learned native
    assistants included the folklorist Abaev.
    This is not surprising, since Old Iranian nar- (nominative nâ, from the
    PIE nom.(!) which Victor Mair cites) is nowhere reflected in Middle
    and New Iranian languages, which however do reflect the adj, *narya-,
    'male', which as expected gives Oss. næl (with the Alan -l- from
    *-ry- as in the Alan ethnonym < *Arya- and many Ossetic examples).
    Recently in my article (in A. Korangy and C. Miller, eds., Trends in Iranian and Persian Linguistics) "On some Iranian secret vocabularies
    as evidenced by a fourteenth century Persian manuscript" (the whole article is online via googlebooks), in Section (20) at the end I have a long discussion with new evidence for the the verbal root of the
    PIE noun h2ner- , i.e. √h2ner 'to be strong, potent', with various interesting semantic developments. Inter alia, I derive the Oss. Nart
    word from Proto-Iranian *narØrâ (Ø = thete), a deverbal noun.
    As for heroes stealing bulls, in my (now largely outdated 1975)
    "Cautes and Cautopates, the Mithraic torchbearers" pp. 417-418,
    I note the epithet bouklópos theós 'the bull-stealing god' for the Roman god Mithras, and Jean Chardin's 1672 account from his visit to Georgia of how monks in Ilori shut up a bullin the church on the eve of St George's feastday, and how the next morning, opening the church. they announce that St George has stolen a bull, who is then immolated by a young man with a knife very much as per the Roman
    Mithraic iconography. Some Nart stuff on p. 416.
    We saw a ladybug on a blooming thistle yesterday — I quipped to my Russophone wife, punning on the Russ. word for God, that it's-a
    Bogbug, the bug in Russian again called "God's little cow".
    Martin Schwartz

  6. Victor Mair said,

    June 6, 2020 @ 8:48 pm

    From Gene Anderson:

    I've read Colarusso's book of Nart sagas, as well as a lot of Georgian and Armenian lit. And I can sing assorted Scottish ballads on the same themes. I see no reason to assume that cattle-raiding is somehow an Indo-European thing. Surely everybody did it and told tall stories about it, long before PIE emerged from proto-Eurasian or whatever it emerged from.

  7. martin schwartz said,

    June 6, 2020 @ 9:42 pm

    I should add that Proto-Iranian *rØ (*r-theta) regularly gave Ossetoc
    rt, so that *nar-Ørâ 'strength, viritility' would give *narrt(æ) > nart(æ)
    *'the mileu of herosim', i.e. the Nart (tales). The suffix is cognate with
    the Skt., Gr., etc. deverbal derivative -trV-.
    As for cattle-rustling, it was probabaly heroized in some less settled sectors of Indo-European society. It was pejorized in Vedic, and posed
    a problem in early Iran, as I believe is reflected in the Gathas, which seem to have vehemently opposed it (but that requires a very long discussion).
    Martin Schwartz

  8. martin schwartz said,

    June 6, 2020 @ 9:44 pm

    To whatever robot is monitoring ths: No, I did not already say this.
    I should add that Proto-Iranian *rØ regularly gave Ossetoc
    rt, so that *nar-Ørâ 'strength, viritility' would give *narrt(æ) > nart(æ)
    *'the mileu of herosim',. The suffix is cognate with
    the Skt., Gr., etc. deverbal derivative -trV-.
    As for cattle-rustling, it was probabaly heroized in some less settled sectors of Indo-European society. It was pejorized in Vedic, and posed
    a problem in early Iran, as I believe is reflected in the Gathas, which seem to have vehemently opposed it (but that requires a very long discussion).
    Martin Schwartz

  9. Germanist said,

    June 7, 2020 @ 12:37 am

    I'm a bit surprised this hasn't come up yet: there is a webcomic artist doing a series based on Nart sagas. It's been going weekly for over three and I've found it quite enjoyable. Spoiler: there are some "adult" moments.

  10. Tamerlan Salbiev said,

    June 7, 2020 @ 2:02 am

    L. A. Lelekov seems to be the first to pay special attention to the motive of cattale raiding in Nart Sagas as a key one (Лелеков Л.А. Ранние формы иранского эпоса // "Народы Азии и Африки"б №3, 1979). In Ossetian epics there even was a special word for such a raid – balts. Saying that, it is, of course, important to bear in mind the difference that balts had, compared to the modern semantics of the word raid that is used for its translation. Two very important differences should be mentioned. First of all, balts was conducted in a special way in order not to be taken for a simple theft. Before the the allien cattle could be driven away, the intruder was supposed to make the host realise what was happening. According to the ethnographic material he was supposed to make noise or even shout out to give the owner of the cattle to protect his property and respond. At the same time the offenden side was given the rifgt of 'first shot" cause otherwise the intruder would be disgraced. The other thing is that cattle driven away from the others was not supposed to make its new owner rich, but should be used for common ritual feasts with religious underground. Tese are features that made Ossetian balts a noble expedition respected by everyone, sort of a contest for testing one's bravery and readiness to die in a fight.

  11. Tamerlan Salbiev said,

    June 7, 2020 @ 2:10 am

    This was the only way to aquire Old Greek κλέος ἄφθιτος; κλέος ἄφθιτον | Old Indian. áksiti śŗávah ; śŗávah…áksitam, i.e. "eternal fame" that Adalbert Kuhn found in Homer and in Rigveda.It was believad to be a formula of "Indogermanische Dichtersprache".

  12. Tamerlan Salbiev said,

    June 7, 2020 @ 2:27 am

    It is also worth while mentioning that Abaev found an avestan term, used in Gathas, corresponding to Ossetian balts – aešma-, which he derived from the verb aeš- 'to move quickly' (Abaev V.I. Scythian life and the reform of Zoroaster // Selected Works. Vladikavkaz, 1990, p.16). It was used to differentiate Iran and Turan.

  13. Tamerlan Salbiev said,

    June 7, 2020 @ 2:33 am

    As for Nar it is a place name, and in Ossetia there two villages that bear this name and it seems to have no direct links with the epic tradition. One of those villages is situated in central Ossetia, in Tual gorge, and is a birthplace of Khetaggaty Khosta. Another one is situated in Digore gorge and is believed to have a counterpart settlement – Fas-nal, i.e."(situated) after Nar/l", which i supported by its geographical position.

  14. martin schwartz said,

    June 7, 2020 @ 6:09 pm

    I was aware (via Miller's dictionary) of Nar as an Ossetic place name ,
    and I'm glad to see Prof. Salbiev's expert opinion that it seems to have no direct link to the epic tradition, which was my inexpert assessment,
    and thge reason I did not mention it. For my next remarks, forgive the
    casual transcriptions: Pace Abaev, as seen already from the Gathas,
    aês﹩3)ma- (﹩ = s-hacek, 3 = schwa) means 'fury, wrath, disorderly
    or violent behavior'; the Parthian, Sogdian, and Middle Persian cognates mean 'wrath, anger', cf. Persian xe﹩m, xa﹩m, the ordinary word for 'anger'. Thepara-Biblical demon Ashmedai (Asmodeus)
    (Asmodues) has his name from Avestan aê﹩ma- daêuua- 'the demon Wrath'. Zarathushtra's Gathas do not use this word , which occurs a good number of times, for 'raid'. As I have shown in several articles,, the Gathic Yasna 32.10 seq. and Yasna 48.10 seq. are parodic/pejorizing citations of the pre-Zarathushtrian Hymn to Haoma, of which Yasnas 9 and 10 are later versions; Yasna 10.8, whose protoform was demonstrably known to Zarathushtra, who contests it,
    claims that all intoxicants/intoxications other than Haoma are accompanied by aê﹩ma-. The opposition of Iran to Turan is a quite la
    quite late topos in Iranian literature. As to Abaev's general idea,
    however, it does seem relevant fthat in Vedic iSm-ín- (S = underdotted
    s) is a positive heroic epithet of martial gods. Indo-Iranian *i﹩ma-, alongside *ai﹩ma-, may be the source of the Middle Iranian
    words for 'wrath' alluded to above, and show a Proto- Iranian
    (rather than Zarathushtrian) deprecation of an aggressive force
    which had been a heroic feature in some archaic stratum of Vedic
    history. Thus Abaev's thesis may be accordingly reframed.
    Martin Schwartz

  15. martin schwartz said,

    June 7, 2020 @ 6:29 pm

    It seems that the demon of disorder has wrathfully messed with what I wrote. I used a dollar sign for s-hacek, eliminated by the diabolic program of th Comment format. So, damage control:
    the Avestan word is, grosso modo, aêshma-. It does not mean 'raid'
    in the Gathas, but 'fury, wrath, disorderly or violent behavior'.
    Cf., inter alia, Persian xeshm, xashm 'anger'. The name of the demon Ashmedai (Asmodeus)in Jewish tradition is from Avestan aêshma-daêuua- 'tha demon of wrath'. I have shown that
    aêshma- occurs in a passage of the Hymn to Haoma, known
    in its protoform to Zarathusthra, who contested it in his Gathas,
    which states that all intoxications other than that of haoma
    are accompanied by aêshma-. Enough, I'm outa here.
    Martin Schwartz

  16. David Marjanović said,

    June 8, 2020 @ 5:13 am

    I used a dollar sign for s-hacek, eliminated by the diabolic program of th Comment format.

    It's still there, it's just oddly small.

    But may I suggest you open your character map? Copy and paste from there whatever you need. I don't know off the top of my head where to find it on a Mac; in Windows its Start > All Programs > Accessories > System Programs > Character Map.

  17. Christopher Culver said,

    June 8, 2020 @ 2:16 pm

    The reconstruction *nar-θrâ 'strength, viritility' > *narrt(æ) > nart(æ) is problematic with regard to the Mari data: it has been suggested that Meadow Mari onar and Hill Mari narə̑ ‘giant; mythical hero’ represent borrowings of the Alanic word. This etymology assumes that the name of the Narts really goes back to a bare root nar with a plural in -tä.

  18. martin schwartz said,

    June 9, 2020 @ 6:46 pm

    @Christopher Culver: In looking at the (online) work of Sampsa Holopainen and others, I am impressed by the many chronological
    strata of loanwords, from Prot0-Indo-European, Indo-Iranian, and
    Old and Middle Iranian languages (which are many), that may be reflected in Mari. Since the (common) noun nar is not attested in
    earlier 20th cent. Ossetic, what makes you choose Alanic (whether you mean that term generically or in reference to the poorly attested Middle Alanic) as the source of your Mari words? WHO suggested such a source? You speak of "a bare root". Indo-Europeanists use the term "root" to mean a very abstract, sort of verbal, conception behind
    attested noun/adjs. and verbs, which are found in stems that are inflected. Old Indic and Old Iranian had a noun stem *nar- 'male man, hero', nom. nâm acc. nâram, which are NOT continued in ANY Middle Iranian language, which would include Middle Alanic.
    All we find in Middle Iranian are reflexes of Old Iranian *narya- 'male'
    (adj.), represented in Osstic as næl, with -l- as in Al- (of the "Alan"
    ethnomym). On the other hand, my article in Trends in Iranian and Persian lingusitics adduuced many Ossetic and other New Iranian
    verbs from an underlying root **H)nar 'to be strong, potent, massive',
    which seemed to be lacking (Cf. Johnny Cheung's Etym. Dict. of the Ir. Vb., which denied that verbal forms are attested. Unless one really is on top of the "tricky" (Holopainen's word) (indo-)Iranian details (I assume you know Uralic details–I don't) and can adduce specifics foms,-preferably not reconstructions–
    it is inadvisable to make Iransitic inferences from Mari, Meadow or Hill.
    Martin Schwartz

  19. martin schwartz said,

    June 9, 2020 @ 11:40 pm

    @ C. Culver
    OOPS! The nominative of OIr. nar- is nâ.
    Martin Schwartz

  20. Christopher Culver said,

    June 10, 2020 @ 12:38 am

    Martin, it is traditionally assumed that Iranian loans in Mari not shared with the wider Uralic family are borrowings from Alanic. (Also, some late Iranian loans into Ob-Ugric not shared by other branches have also been termed thus.) This tradition stems from Abaev, for example, who was rather proud of his ancestry and stated outright on various occasions that no other Iranian language could have been the source, since only Alanic was spoken in the South Russian steppes. This is probably not entirely true – I would argue for least some trade-related borrowings in Mari from Khwarezmian, and of course some Iranian words could have been borrowed into Proto-Uralic but lost everywhere but in Mari. However, some Iranian borrowings in Mari show the Ossetic metathesis of fricatives, and therefore it is valid to term them Alanic loans.

    Yes, "bare root" was my misphrasing, stem would have been a better word. The point is that the Mari etymology assumes that the -ta in the Ossetic word is the plural marker. I won’t insist on the Mari etymology, which is much less secure than other Iranian loanwords in Mari (Meadow Mari and Hill Mari show irregular correspondences, suggesting a late borrowing, but the end of Proto-Mari unity is traditionally dated to the Mongol invasion, and thus probably too late for any Iranian loans, unless maybe those Burtas who fled the Mongol invasion towards the Komi lands were Alanic speakers, but their linguistic affiliation is mere speculation.) So, please take my comment above as a suggestion that things don’t add up here somewhere, and that may be on the Mari side and not the Iranian side.

  21. martin schwartz said,

    June 10, 2020 @ 1:43 am

    @Christopher Culver. Thanks, Christropher, for your educative comments. I would be grateful for a list of the exclusively Mari Dept of Near Eatern Studies).
    Btw when you say "Mari etymology assumes that the -ta in the Ossetic word is a plural marker", you mean the scholarly etymology assumes
    that, and not the Mari people metanalyzing what they percieved, as plural marker based on other words they knew of Alanic, yes?
    Martin (Schwartz)

  22. martin schwartz said,

    June 10, 2020 @ 1:51 am

    @C. Culver: ho, my computer keeps on pranking around.
    I meant to write that you may send me a sepearte e-mai via my UC Berkeley account–or you may do it in the Language Log chain–
    listing exclusivly Mari Uralic borrowings from Iranian, and also
    a list of Mari words you think are from Khwarezmian, an old flame of mine. Dunno if the Khwarezmian ethnicity of the Khazar garrison
    is of relevance.
    Martin (Schwartz)

  23. Christopher Culver said,

    June 10, 2020 @ 10:43 am

    Work on Middle Iranian loans in Mari falls squarely into Western and Soviet schools. The Western tradition was summarized in Aulis Joki’s 1973 work Uralier und Indogermanen, and little has been done since. Cheung lists some Mari data in the index to his 2002 book on Ossetic vocalism, but mainly this consists of just the work Joki summarized.

    The Soviet tradition of research on Mari–Iranian contacts was sparked when Abaev published the first volume of his Ossetic etymological dictionary. The key names here are F. I. Gordeev and D. E. Kazantsev who published a number of works on the subject. (I. G. Dobrodomov was working in a similar vein on possible Alanic loans in the neighboring Chuvash language.)

    The problem with both the Western and Soviet traditions is that they are based on a pre-modern understanding of both East Iranian and Mari historical phonology, and they were published before Khwarezmian, Sogdian and Bactrian were as well described as they are now. In my own work I’m trying to sift out the comparisons which are fallacious, preserve those which seem valid, and add a few etymologies of my own. If you’d like, you can follow me on, I would love to have more input from Iranists.

    With regard to Khwarezmian loans, the obvious source for these would be the sort of trade between Central Asia and the Middle Volga which was so vividly described by Ibn Fadlan. I have one example of an Iranian word with a distinctly Khwarezmian sound shift that I need to finally sit down and write up. Note that V. V. Ponjadarov has been working along similar lines with regard to Udmurt and Komi, see his recent-ish article "Khorezmijskie zaimstvovanija v permskix jazykax", though I don’t necessarily endorse it.

RSS feed for comments on this post