Of precious swords and Old Sinitic reconstructions, part 7

« previous post | next post »

[This is a guest post by Chau Wu, with additions at the bottom by VHM and others]

On the akinakes* (Scythian dagger / short sword) and Xiongnu (Hunnish) horse sacrifice

Chinese historical records suggest that the akinakes, transliterated from Greek ἀκῑνάκης, may be endowed with spiritual significance in the eyes of ancient Chinese and Northern Barbarians, for it was used in solemn ceremonies.  Let me cite two recorded ceremonies and a special occasion where an akinakes is used to “finesse” an emperor.

In the Book of Han (漢書), Chapter 94 B, Records of Xiongnu (匈奴傳下), we see an akinakes is used in a ceremony sealing a treaty of friendship between the Han and Xiongnu.  The Han emissaries, the Chief Commandant of charioteers and cavalry [車騎都尉] Han Chang (韓昌) and an Imperial Court Grandee [光祿大夫] Zhang Meng (張猛) visited the Xiongnu chanyu** (單于) [VHM:  chief of the Xiongnu / Huns] in 43 BC.  Han and Zhang, together with the chanyu and high officials, climbed the eastern hill by the river Nuo (諾水)***, killed a white horse, and the chanyu using a jinglu knife (徑路刀) and a golden liuli**** (金留犁, said to be a spoon for rice) mixed the horse blood with wine.  Then they drank the blood-oath together from the skull of the King of Yuezhi, who had been defeated by the ancestor of the chanyu and whose skull had been made into a goblet.  Essentially, this jinglu knife was a holy mixer.

In Yi Zhou Shu (逸周書), Scroll 4, Chapter on Ke Yin (克殷 ‘Conquering the Yin-Shang’), a ceremony equivalent to coronation at the founding of the Zhou dynasty is described.  This event took place after the last king of the Shang committed suicide upon losing a battle, which by the way is the first battle in Chinese history described in great detail by contemporaries.  In the ceremony to enthrone King Wu of Zhou (周武王), Duke Zhou (周公) held a large broad-axe (dayue 大鉞), Duke Zhao (召公) a smaller one (xiaoyue 小鉞), to guard the king’s sides, and the three great officials who helped to found the Zhou Dynasty, San Yisheng (散宜生), Tai Dian (泰顛), and Hong Yao (閎夭), all took an akinakes (qinglü 輕呂) to present to the king, whereupon the king entered and took a seat on the ceremonial ground (王入, 即位於社).  Here the akinakes represents a symbol of power that commands respect and submission, not unlike the sword in Arthurian legend.

In Shiji (史記 Records of the Grand Historian), Chronicles of the Zhou (周本記) and in the same chapter of Yi Zhou Shu cited above, King Wu of Zhou is described as taking an action that is considered repugnant by modern standards of moral conduct.  He used an akinakes to mutilate / lacerate the face of the dead king of Shang.  One recalls an analogous custom among the ancient Northern Barbarians as well as Attila’s Huns that, when a king died, his followers would lacerate their own faces in deep grief.

In Han Shu, the dagger is 徑路 (jinglu), in Yi Zhou Shu it is 輕呂 (qinglü), and in Shiji 輕劍 (qingjian); they all refer to akinakes.  Jaroslav Průšek, in his Chinese Statelets and the Northern Barbarians in the Period 1400-300 B.C. (1971, page 133) states:

In this connection we must consider whether the name of the knife, or rather the sabre ching-lu, used on this occasion [referring to the friendship treaty ceremony], is not in fact a transcription of the Scythian name for a short sword – or rather of the Greek transcription of the Scythian name, akinakes. Although it is difficult to explain the a in the initial syllable, this etymology in not impossible.  As far I am aware, it was first suggested by Japanese authors in Inner Mongolia 1937, English resumé p. 7.

By the way, if one takes the Taiwanese pronunciation of Chinese graphs for derivation, like I do, it is rather straight-forward to see the correspondence between akinakes and the three Sinitic terms given above.

(1) The Taiwanese pronunciation (in POJ) of the three Sinitic names for *akīnaka-:

徑路 kèng-lō·

輕呂 kheng-lū

輕劍 kheng-kiàm (N.B. Here the 2nd element "-lū" of kheng-lū had been lost due to displacement, and replaced with 劍 'sword')

The following is my derivation from *akīnaka-:

akinaka > akinak- > (aphesis) -kinak- >(gemination) > *kin-nak >

1st element kin- > (pattern of sound correspondence -in > -eng; I have hundreds of examples) > keng 徑 / kheng 輕

2nd element -nak > (de-nasalization) > *lak > (loss of -k) > lō· 路 / lū 呂 (the 7th tone of both glyphs reflects the loss of -k).]

(2) The River Nuo:  諾水.  (The text reads: 匈奴諾水東山 'the eastern hill by the River Nuo in Xiongnu territory'.)

——————–conclusion of Chau Wu's guest post———————

*Since no Old Persian or other Old Iranian form is attested, we have no choice but to give the Greek form (in Roman transliteration). If one prefers the Iranian form *akīnaka-, don’t forget an asterisk and a hyphen to show it’s a stem form. (HK)

**People (Chinese and others) pronounce 單于 both as shanyu and as chanyu, but most critical Western scholars pronounce it as chányú.

Old Sinitic

(BaxterSagart): /*dar  ɦʷa/

(Zhengzhang): /*djan  ɢʷa/

Cognate with Mongolian ᠳᠠᠷᠤᠭ᠎ᠠ (daruɣ-a) / дарга (darga, “chief; head; governor”), Persian داروغه‎ (dâruğe, “governor”), Old Turkic ‎ (tarqan, “commander”).

Related to 答剌罕 (“tarkhan”), 達魯噶达鲁噶 (“darugha”), 達魯花赤达鲁花赤 (“darughachi”)


***nuò 諾

Old Sinitic

(BaxterSagart): /*nˤak/

(Zhengzhang): /*naːɡ/


See below for more on the probable identification of this river as the Orkhon in Central Mongolia.

****liúlí 留犁 (spoon or other implement in which to mix the blood of the horse with the wine; this word has the look of a borrowing from a foreign language)

Old Sinitic

(BaxterSagart):  /*C.ru/      /*[r]ˤ[i]j/, /*[r][i]j/

(Zhengzhang): /*m·ru/      /*riːl/, /*ril/

(source:  here and here)

They are uncertain about the initial consonant or onset of the first character.

Now, to focus on the akinakes:

The acinaces, also spelled akinakes (Greek ἀκῑνάκης) or akinaka (unattested Old Persian *akīnakah, Sogdian kynʼk) is a type of dagger or short sword used mainly in the first millennium BC in the eastern Mediterranean region, especially by the Medes, Scythians and Persians, then by the Greeks


In the Hàn shū 漢書 (Book of Han; History of the [Western / Former] Han [Dynasty]) (finished in 111 AD), the equivalent term for akinakes is jìnglù 徑路.  Superficially, that would mean "route; path", but it makes no sense in the context in which it is found, where it must refer to a sword or knife.  Indeed, in one occurrence in the quoted passage, it is explicitly referred to as a jìnglù dāo 徑路刀, a "jinglu knife".  Moreover, as we have seen above, there are multiple orthographic variants of the term, so there can be no doubt that it is a transcription of a non-Sinitic word.  I believe that word is the Greco-Scythian akinakes,

jìnglù 徑路 (akinakes < *akīnaka-)

Old Sinitic

(BaxterSagart):  /*[k]ˤeŋ-s/      /*Cə.rˤak-s/

(Zhengzhang): /*keːŋs/      /*ɡ·raːɡs/

(source here and here)

Note that the Sogdian form of akinakes lacks the initial "a" and the final "s".

As I have stressed repeatedly in the long series of posts on Old Sinitic reconstructions here at Language Log, historical linguists must take seriously into account the phonological data available from words that have been borrowed into Sinitic from other languages.  Such data will provide solid anchorage points for further analysis.  Chinese historical linguists cannot continue to operate as though Sinitic were a completely closed, hermetically sealed system, basing their reconstructions primarily on rime tables, which amount to a set of Procrustean beds.

Now that we have witnessed the adoption of an Iranian implement (weapon) and the word for it by the Xiongnu / Huns, let us see whether the same thing happened with Iranian culture and customs.

A good starting point for our inquiry is Wilhelm Koppers (1886-1961), "Pferdeopfer unde Pferdekult der Indogermanen", which is in his (as editor) Die Indogermanen und Germanenfrage (1936), pp. 279-411. This is Kulturkreislehre (culture sphere) stuff at its most exhaustive and comparative and, as can be seen from the length of the article, it is pretty thorough, although it concentrates especially on evidence from Central Asia.  While Koppers himself later abandoned the Kulturkreislehre methodology in favor of a more historical and ethnological approach, the sheer accumulation of massive amounts of data and comparison among them provides a solid foundation upon which to build new models with more recently discovered evidence, as adduced below. (JPM)

In the survey of Iranian-Hunnish cultural interactions conducted below, here are some of the themes we will be tracking:

  1. chieftains / rulers drinking together (whether from a skull or other vessel) to cement a friendship between different polities
  2. mixing of blood — whether of animals or humans — with other substances before consumption
  3. sacrifice of horses, especially white ones
  4. solar symbolism
  5. lavish display of gold
  6. the language of the Xiongnu / Huns

The Xiongnu were a confederation. It is likely that they had speakers of different languages and ethnicities under their command and control. The late Elling Eide worked for decades on the question of the language, culture, and people of the Xiongnu, and he had assembled a considerable amount of evidence that there were Iranian elements within the confederation. So far as I know, he never had a chance to publish any of this material, but some of his notes on the subject — which he discussed with me during numerous phone calls and visits — may be in the Elling Eide Center in Sarasota.

So much for the listing of the themes of Iranian-Hunnish cultural interactions.  By Iranian I intend primarily Sogdian and other Middle Iranian languages as well as their Old Iranian and Indo-Iranian predecessors, and by Hunnish I intend mainly the Xiongnu and the Huns, but also to a lesser degree their successors.  I will pay due attention to customs and beliefs pertaining to the sacrality of horses throughout the Indo-European world as supporting evidence for phenomena within Iranian.

Now to lay out an abundance of evidence to document the interaction between these two large cultural and linguistic agglomerations.

There is a scene on the Anqie 安伽 mortuary panels where the Sogdian Anqie and a Turkish chieftain drink together under a tree to mark their friendship. (YY)

The entire kapala ("skull") / khaṭvāṅga ("skull-topped club, a skull-mounted trident, or a trident-staff on which three skulls are impaled") ritual of Tantric Buddhism (India, Nepal, Tibet, China, and Mongolia) is worthy of consideration. In it the yogin prepares the elixir in the cup (kapala) mixing the eight meats (including horse & human) together with the khaṭvāṅga which dissolves into the elixir and then consumes the elixir. I made a graphic video of the process that we showed with the Circle of Bliss exhibition. (JH)

In the Vedic horse sacrifice (aśvamedha), the blood of the sacrificial horse is not drunk but poured upon the head of an ugly and bald old man sticking out of a river. While this is done, the old man is addressed as Jumbaka, supposed to be another name of the god Varuṇa. (AP)

Herodotus gives a fair amount of detail about Scythian rituals but as far as I recall his only description of horse sacrifice is when they are stuffed with hay and mounted on spears in a circle around the buried chieftain's kurgan. (RF)

No animal sacrifice forms a part of Zoroastrian rituals (like the Aśvamedha or equus october). However, apparently pre-Zoroastrian legends incorporated into the Zoroastrian scriptures speak of sacrificing “a hundred male horses, a thousand bulls, and ten thousand sheep” to a deity; see Shapour Shahbazi’s article ASB “horse” in EIr. (HK)

The places of sacrifice vary, but rivers and lakes are certainly mentioned. In a new translation of the Avesta by Pierre Lecoq, it is “sur le piémont de la Harā” (Yt 5.21), “sur les hauteurs du mont Hukairya” (Yt 5.25), “dans le pays de Bawri” (Yt 5.29), “dans le pays de Varəna aux quatre coins” (Yt 5.33), “devant le lac Pišinah” (Yt 5.37), “dans sa caverne souterrraine” (Yt 5.41), “sur le mont Ǝrəzifya” (Yt 5.45), “devant le lac Čaēčista” (Yt 5.49), “au défilé du Xšaθrō.suka” (Yt 5.57), “près du grand ahura, le splendide souverain Apąm Napāt” (Yt 5.72), “en face de l’île de la Raŋhā” (Yt 5. 81), “devant la rivière Frazdānu” (Yt 5.108), “devant la rivière Dāityā” (Yt 5.112), “près du lac Vourukaša” (Yt 5.116). (HK)


Elizabeth Barber:

I seldom get as far east as Indo-Iranian, so I have no direct answers to any of your questions; but I have run into sacred white horses all over the place among other Indo-European groups.  Here's a brief data-dump, for what it's worth:

1) Celtic kings had to have intercourse with a white mare at some point.

2) Europeans prior to the arrival of the Indo-Europeans had many of the same animal rituals, it's just that they used bulls / cattle where IEs used horses: e.g., head-and-hoof sacrifice — sorry, color unknown), and royalty mating with said animal, as fertility / abundance rite (cf. Celtic king with white mare, and Cretan queen with bull — and note that Europa was hijacked to Crete by a white bull).

3) Various IE divinities rode specifically white horses: Celtic Rhiannon and Epona (both female), Slavic Svarovich and Svantovir (both male), Germanic Odin (actually, Sleipnir was grey, but I'm told that truly white horses, like the Lipizzaners, are very rare, and that light grey was taken as white).

4) Horses had a special pipeline to the Other World, for IEs, in at least 2 ways (and of course, this would make them specially potent for oaths — like swearing on a Bible): they could ensure prosperity directly, or could divine for mortals what the gods had in store.  E.g., Achilles's divine horses, sired by the West Wind, foretell Achilles's death; Tacitus and Saxo Grammaticus describe northern priests (Celtic? Germanic?) saddling a white stallion, and leading it riderless(!) past various obstacles to divine the future by which foot the horse used first (young Russian peasants used to do the same, to learn about their future mates!); Albanians used white horses to detect vampire graves; and the oracular horse on the Baltic island of Ruegen was white (I don't happen to know if this site was N. Germanic or Baltic or both). The terrifying Welsh Mari Lwyd, which comes out just before New Year — a cosmic crack in the calendar when the spirits can escape the spirit world — is a Grey Mare (literally), and it is made from a horse skull and pelt (or blanket) and chases girls in particular–fertility magic again.

[VHM:  There are many different competing etymologies for Mari Lwyd, but I believe that the most convincing by far is simply that the two words mean "mare" and "gray", nothing more arcane or obscure than that.

When we were in elementary school, my siblings and I used to get teased by classmates singing "The Old Gray Mare she ain't what she used to be, ain't what she used to be, ain't what she used to be…".  No one, including the ones who taunted us with those lines, knew what they meant, but we all knew for sure that it wasn't nice to be compared to an Old Gray Mare".

The Old Gray Mare is referenced in Aristophanes "Lysistrata", where it refers to wives who are starting to rebel.


From Middle English mare, mere, from Old English mīere (“female horse, mare”), from Proto-Germanic *marhijō (“female horse”) (compare Scots mere, meir, mear (“mare”), North Frisian mar (“mare, horse”), West Frisian merje (“mare”), Dutch merrie (“mare”), Danish mær (“mare”), Swedish märr (“mare”), Icelandic meri (“mare”), German Mähre (“decrepit old horse”)), from *marhaz (“horse” [Proto-Germanic]) (compare Old English mearh).


"female of the horse or any other equine animal," Old English meare, also mere (Mercian), myre (West Saxon), fem. of mearh "horse," from Proto-Germanic *marhijo- "female horse" (source also of Old Saxon meriha, Old Norse merr, Old Frisian merrie, Dutch merrie, Old High German meriha, German Mähre "mare"), said to be of Gaulish origin (compare Irish and Gaelic marc, Welsh march, Breton marh "horse").

[VHM:  Irish marcshlua and marcra mean "cavalry".  Proto-Celtic would be *markos. This is generally reconstructed as being from Proto-Indo-European *márkos (“horse”)]

The fem. form is not recorded in Gothic, and there are no known cognates beyond Germanic and Celtic, so perhaps it is a word from a substrate language. The masc. forms have disappeared in English and German except as disguised in marshal (n.).

(Online Etymology Dictionary)

Middle English mere, mare, from Old English mȳre, mīre (influenced by forms of mearh, mēar-, horse); see marko- in Indo-European roots.



    1. marshal from Old French mareschal, from Frankish *marha-skalkaz, horse-servant (*skalkaz, servant).
    2. mare1 from Old English mere, miere, mare, from Germanic feminine *marhjōn-.

[Pokorny marko- 700.]

(AHD 5th ed.)

Note that in some European languages, cognates of "mare" just mean "horse", not "female horse".

N.B.:  I pay particular attention to this word ("mare", etc.), not because my schoolmates used to kid me with it, but because I still believe that it, or an IE word related to it, is the source of Sinitic mǎ 馬 ("horse"):

Old Sinitic

(BaxterSagart): /*mˤraʔ/

(Zhengzhang): /*mraːʔ/



None of this is Indo-Iranian, but it gives you a good sense of what the IEs thought and did about white horses!  They are special messengers from and tie-lines to the spirit world, where oaths are sacred.

Oh, I almost forgot: I now recall that the Scythians were said to have to give any white foal that was born to their king.  Must be in Herodotus somewhere, and of course Scythians were Iranians.  So that should be directly relevant!


Pita Kelekna:

White-stallion sacrifices:  across the entire breadth of Eurasia wherever IE peoples have penetrated, there is an immense abundance of ritual surrounding the White stallion as a cosmic symbol of the sun, from endless Whitehorse (but never black, brown, or greyhorse) taverns in England to herds of pure white horses at Xanandu. Please note above below an earlier exchange I had with Richard Bulliet on whitehorse rituals, in which I summarized pretty much everything I knew on the subject–which possibly might be of help.  Remarkably, I do not have any specific information on Iranian whitehorse sacrifices; Indian asvamedha yes, but Iranian no.   However note the images of Russia and Turkmenistan, immediately north of Iran.  The golden horse perched atop the white mountain in modern Turkmenistan is certainly tentatively reminiscent of the Persian sacrifice posited in your text.

Zarathustra’s clear opposition to cruel forms of sacrifice has led some scholars to believe that he forbade animal sacrifice altogether; but, like haoma, the weight of tradition is testimony that it was not proscribed.  Evidently though, such cruel acts were severely punished.  This may account for less flamboyant horse sacrifice in Iran compared to India.

[VHM:  In the course of her investigations on the Ashvamedha (Sanskrit aśvamedha अश्वमेध) sacrifice and analogous rituals among other Indo-European groups, Pita also sent me these remarkable notes on pan-Eurasian whitehorse solar symbol, which she had written in late May, 2016 in response to a query regarding "a whole bunch of white horses being harnessed to a single (parade) vehicle":]

It just so happens a day or two ago on the London news, I spotted an aerial video of her Britannic Majesty, Queen Liz II leaving Buckingham Palace for the ceremonial Opening of Parliament in a golden stage coach drawn by four pairs of white horses serially harnessed in tandem, not abreast— admittedly a rather paltry showing compared to the Ming spectacle.  It would appear then that today’s Turkoman Communist dictator, Gurbanguly Berdymukhamedov, is not the only pompous potentate into the white / golden cult.  Aside from constitutional monarch Liz II, both conquerors Charlemagne and Napoleon are depicted victoriously posed astride magnificent white chargers.  In fact, white horses abound across the entire IE realm:

In Zoroastrianism, the Avestan benevolent divinity Tištrya appears as a white stallion associated with life-bringing rainfall and fertility.  According to myth, Tištrya is locked in cosmic battle against the drought-bringing demon Apaosha, who assumes the form of a terrifying black stallion. With the timely intervention of Creator Ahura Mazda, Tištrya finally prevails.  In Buddhism, the white stallion Kanthaka was the loyal servant and favorite horse of Prince Siddhartha, later Gautama Buddha.  As shown in images below, in Greece Bellepheron, bequeathed Athena’s golden bridle, mounted the snowy winged horse Pegasus to combat and kill the fiery Chimera. In the Celtic world, the white mare Epona is revered as goddess of horses, (Note in epona2, the IE triad of three white horses within the solar circle). And let us not forget pious St George, whose dragon-slaying feats early earned him fame as protector of Crusader and Eastern Christian knights, later launching him as patron saint, not just of England, but also of Portugal, Germany, Aragon, Catalonia, Genoa, and Venice.

[VHM:  In a separate note of transmittal to me on June 9, 2016, Pita added the following remarks, accompanied by ten striking illustrations, ranging from Queen Elizabeth II's gilded royal carriage drawn by eight white horses, through Napoleon astride his famous rearing white war horse Marengo (c. 1793-1831), a soaring winged white horse, a triskelion formed of three white horses, to two gigantic, extended, hill figures of horses, one prehistoric made of crushed white chalk and one completed in June 2003 made of limestone slabs.]

Thus, it would seem that the extraordinary IE solar symbolism of White Horse plus accompanying golden wheels of the war chariot / parade vehicle not only span the Eurasian land mass but share a tradition of untold millennia:  the Aryan-Iranian split occurring 4,000 years ago, contributing respectively to the Vedic asvamedha and Yuan White Feast; the centum (Celtic)-satem split even earlier, more like 5,000 years.  Even more amazing, this tradition still prevails into modern times.  One final point, white chalk horses as territorial markers are quite numerous in Britain. Note after the 360-foot Uffington, the final Folkestone image—a flying gallop, no less.


Francesco Brighenti

Regarding sacrifices of (yes, white!) horses among some ancient Iranian peoples, see S. Fuchs, The Vedic Horse Sacrifice in Its Culture-Historical Relations (New Delhi: Inter-India Publications, 1996), pp. 59-61, 63 (notes). [VHM:  I have scans of these pages.]

As to ancient Greek ἀκĩνάκης ‘short sword of Persians and Scythians’ (Herodotus), if it is not Pre-Greek but an Iranian loan (Beekes leaves open both possibilities):

Bailey (Transactions of the Philological Society 1955, p. 69) derives Old Persian *akīnaka- ‘sword, knife’ (thus reconstructed on the basis of Greek ἀκĩνάκης) and Sogdian kyn’k ‘sword’ (< *kīnaka-) from a putative Proto-Iranian form *kīna- ‘wedge’ corresponding to Vedic Sanskrit kīla & khī́la ‘spike, nail, stake, peg, wedge’ on the assumption of alternating suffixes -n- and -l- in the Proto-Indo-Iranian etymon (but note that variation of k- with kh- in Vedic Sanskrit rather supports non-Indo-Aryan origin). Bailey speculates: “From both the words we thus get *kī-, beside which one would expect to find an accented *kāy. This may be found in the Latin cae- of caedō ‘cut’. […] Here then we have an Old Iranian akīna- [with initial a- unexplained, as Bailey notes] and Middle Iranian kīna- with a suffix -aka-, meaning a ‘cutting tool’. The k- before -ī- indicates an Indo-European *kəi̯-, to which latin cae- corresponds regularly, hence here Latin cae-d- provides a satisfactory connexion.”

Stephen Fuchs, in the same book quoted just above, writes that the (Turkic) Yakuts of northeastern Siberia have traditions about a white stallion and white mares being sacrificed yearly in earlier times (p. 47); similarly, the (Mongol) Buryats of southeastern Siberia prefer a white horse as their sacrificial victim (p. 48).

As for the ancient Persians, Parthians and Scythians, here are some important links:

Pierre Briant:

Discussion of sacrifices of horses (especially white horses) as an exceptional Persian ritual practice.

Albert de Jong:

“Herodotus mentioned that [Persian animal] sacrifices were performed on a mountaintop.”

Albert de Jong:

“Xenophon explicitly refers to bulls being sacrificed to Zeus (and horses to the Sun [i.e. Mithra?]), which seems wholly plausible.”

Albert de Jong:

“In the Yašts, many heroes from Iranian mythology perform elaborate sacrifices (of horses, cows and small cattle) to obtain certain favours from the deities they invoke.”

Albert de Jong:

“[T]he species offered [in Persian animal sacrifices] (horses, cows and sheep) were certainly all offered once, even though there does not seem to be direct evidence for horse sacrifice.”

Magi in Thrace:

Herodotus (7.113.2) writes that the magi embedded in Xerxes’ army performed a sacrificial consultation in Thrace using white horses as victims.

Parthian king in Babylon wishes to sacrifice a Nisaean horse:

Philostratus (Life of Apollonius of Tyana 1.31) mentions a lavish horse sacrifice offered to the Sun (i.e. Mithra?) by the Parthian king Vardanes in Babylon. The king, says Philostratus, “was on the point of sacrificing to the Sun as a victim a horse of the true Nisaean breed, which he had adorned with trappings as if for a triumphal procession.”

[VHM:  This is the same breed as the Chinese emperor Wudi of the Han sent an army to Central Asia to obtain.  The Chinese called the Nisaean breed "Tiānmǎ 天馬" ("Heavenly Horse").

The War of the Heavenly Horses (simplified Chinese: 天马之战; traditional Chinese: 天馬之戰; pinyin: Tiānmǎ zhī Zhàn) or the Han–Dayuan War (simplified Chinese: 汉宛战争; traditional Chinese: 漢宛戰爭; pinyin: Hàn Yuān Zhànzhēng) was a military conflict fought in 104 BC and 102 BC between the Chinese Han dynasty and the kingdom known to the Chinese as Dayuan, in the Ferghana Valley at the easternmost end of the former Persian empire, now the easternmost part of Uzbekistan. The result was a Han victory, and a temporary expansion of its hegemony deep into central Asia.

Emperor Wu of Han had received reports from diplomat Zhang Qian that Dayuan owned tall and powerful horses ("heavenly horses") that could help fight against the Xiongnu. He sent envoys to survey the region and purchase horses. Dayuan refused the deal, resulting in the death of one of the Han ambassadors, and confiscated the gold sent as payment. So the Han sent an army to subdue Dayuan. Their first incursion failed, but a second, larger force defeated Dayuan, installed a regime favorable to the Han, and won enough horses to later build a cavalry strong enough to defeat the Xiongnu in the Han–Xiongnu War.

[VHM:  Han only got a measly 13 Nisaean horses out of the deal, but emperor Wudi, upon beholding them, thought it was worth all of the expense, effort, and bloodshed.]


It appears that the name "Yuan" was simply a transliteration of Sanskrit Yavana or Pali Yona, used throughout antiquity in Asia to designate Greeks ("Ionians"), so that Dayuan would mean "Great Ionians" or "Great Greeks".


The "Heavenly Horse" is also called the Ferghana Horse, after the name of the valley where Dayuan, where they flourished, was located.  In Chinese, they were also known as "Blood Sweating Horses" (Hànxuè mǎ 汗血馬), which has given rise to much speculation about how they could sweat blood (probably triggered by a parasitic nematode, Parafilaria multipapillosa).]

Herodutus on Scythian horse sacrifice to the sun:

Herodutus (1.216) writes of the Scythians, “Of gods they reverence the Sun alone, and to him they sacrifice horses.”

Juliet Clutton-Brock:

“Much of what Herodotus wrote about the Scythians has been corroborated from the excavations of [Pazyryk] tombs.”


The Nuo river, on whose banks the 49 BCE treaty between the Xiongnu chanyu Huhanye and the Han emissaries Han Chang and Zhang Meng was concluded, is identified by some scholars with the river Orkhon in central Mongolia – cf. R. Michaud, S. Michaud & M. Jan, The Great Wall of China (New York: Abbeville Press, 2001), p. 75.That area of Mongolia is located some 800 km to the east of the homeland of the Altai Scythians.

The sacrificial practices of the Scythians of western Central Asia and Ukraine, who strangled the horse by means of a rope and stick as per Herodotus’ description, differed from those of the Altai Scythians (Sakas), who slew the horse by an axe-blow to the forehead which may have been followed by throat slitting, though no identifiable traces of this subsist today. It therefore seems that two different sacrificial practices coexisted among the ancient Northern Iranians: one which draws blood and the other, which does not – see p. 318 in S. Lepetz’s paper here.

If the sacrificed horse’s blood really used to be ritually mixed with wine and drunk from a skull-cup by Xiongnu royal sacrificers as stated in the Book of Han, then the Xiongnu may have followed the sacrificial tradition involving bloodshed (for which there is evidence in the princely tomb of Berel’ in Kazakhstan as discussed in Lepetz’s paper).



Wherever domesticated horses spread, they took with them aspects of the Indo-European peoples who learned how to hitch them to chariots and carriages and plows, ride them, and milk them, who developed elaborate mythologies surrounding them, who sacrificed them to the sun and other potent forces with special weapons, and — above all for us linguists — gave names to all of these equine related things, activities, and ideas.


Selected readings


The several spectacular congregational sacrifices or yajñas known from early first millennium B.C. South Asia, viz. the rājasūya, vājapeya, gavāmayana or atirātra, included numerous rites, rituals and customs with magical connotations, and also had significant political implications like reasserting a monarch’s positions or promoting him through several stages of kingship. One such sacrifice, the aśvamedha, often referred to in the context of empires and metropolises c. 8–7th century B.C., was rather spectacular, involving the letting loose of a horse, its military peregrination for one year, and its subsequent immolation in a rite involving mock necrophilia and bestiality. The many studies of the rites, procedure, and symbolism of aśvamedha are marked by two apparent gaps. The first of these is that most studies, apart from some notable works like of Puhvel or, more recently, Witzel, are preoccupied with the mature form of the aśvamedha and do not adequately consider the evolution of the rites, procedures and symbolisms. The second gap is that most studies, including the ones referred to above, concentrate on the various ritual symbolisms of the sacrifice, usually overlooking its ‘military’ aspect, i.e. the ritualised military context of the yearlong circuit, that had actually made the sacrifice so spectacular. The aim of this paper is to go beyond the ritual and ceremonial aspects of the aśvamedha as known from its mature form, and seek the roots of its politico–military procedure far back in time in the nomad world of endemic warfare.


[Thanks to Elizabeth Barber (EB), Hiroshi Kumamoto (HK), Asko Parpola (AP), Yutaka Yoshida (YY), J. P. Mallory (JPM), Richard Foltz (RF), and John Huntington (JH)]




  1. Victor Mair said,

    January 11, 2021 @ 1:38 pm

    From Daniel Waugh:

    "One recalls an analogous custom among the ancient Northern Barbarians as well as Attila’s Huns that, when a king died, his followers would lacerate their own faces in deep grief."

    I am reminded of the striking image in Mogao Cave 158 of the mourners at the bier of the Buddha, where clearly the artist is depicting those slashing their faces as foreigners from somewhere to the West.

    On drinking from skulls, there is a well-known passage under the year 6430 (972) in the Russian Primary Chronicle (Povest' vremennykh let), here in modernized orthography from the Laurentian manuscript:

    Поиде Святослав в пороги, и нападе на нь Куря Князь печенежский, и убиша Святослава, взяша главу его, и во лбе его, сделаша чашю, оковаше лоб его и пьяху по нем…

    In the late Horace Lunt's draft, unpublished translation (which is being edited for publication now):

    "Sviatoslav went to the rapids. And Kuria, the Pecheneg prince, attacked him. And they killed Sviatoslav. And took his head, and they made a cup of his skull, putting metal around the forehead, and they drank from it."

  2. Benjamin E. Orsatti said,

    January 12, 2021 @ 8:54 am

    So many questions:

    (1) Is the suggestion, then, that the Huns / Οὖννοι / Xiongnu were IE's (probably not, since they were a "confederation" of mixed ethnicities), but can historical linguistic investigation tell us anything about the poorly attested Hunnic language, i.e. was it IE, Altaic, etc.?
    (2) …and that the Minoan Cretans (the bull-jumpers) were non-IE's? Who were they, then, or are we just as unlikely to find out where they came from as we are to find out where the Basques came from?
    (3) How does one treat a human skull to prevent one's beverage from leaking out the intracranial sutures (asking for a friend)?

  3. Chau said,

    January 12, 2021 @ 10:59 am

    @ Benjamin E. Orsatti. Q (3). Around two years of age, a child's skull bones begin to join together because the sutures become bone. When this occurs, the suture is said to “close.” The cups (more correctly, chalices) mentioned in history were mostly fashioned from enemy chiefs' skulls who were adults of course. Therefore, a skull cup is leakproof. Wikipedia gives a list of better known skull cups in history (https://en.wikipedia.org/wiki/Skull_cup).

  4. Benjamin E. Orsatti said,

    January 12, 2021 @ 12:39 pm

    Oh, good! We're having company over later, and we wanted to bring out the "good cups", and my wife had just bought a new rug that really ties the room together.

    It's always comforting to read ancient history; however savage people are to each other now, good grief was there some shocking cruelty back in the "good old days" of the Bronze Age!

  5. cameron said,

    January 12, 2021 @ 12:59 pm

    Early in the OP we read: "Essentially, this jinglu knife was a holy mixer."

    I think "mixer" isn't quite the right word. "Stirrer", perhaps; or, even better "swizzle stick".

    I'm imaging a sort of Scythian tiki bar, where one sips horse-blood cocktails out of skull mugs.

  6. Chau said,

    January 12, 2021 @ 1:10 pm

    Here's your Bloody Mary. Cheers!

  7. Philip Taylor said,

    January 12, 2021 @ 1:31 pm

    Regarding Benjamin's "however savage people are to each other now, good grief was there some shocking cruelty back in the "good old days" of the Bronze Age!", I don't experience any real emotion when I read of human sacrifices, but the sacrifice of horses fills me with absolute horror. Given that the sacrifice was almost certainly effected by cutting the neck, it is beyond my comprehension how anyone could bring themselves to do such a thing to a horse.

  8. Scott P. said,

    January 12, 2021 @ 2:08 pm

    The Minoans are known to be pre-IE due to the evidence of place names. Linear A, of course, has yet to be deciphered.

    Another example of a skull cup is reported by Paul the Deacon as being in the possession of one of the Lombard kings, but I can't find the reference at the moment.

  9. Abbas said,

    January 12, 2021 @ 3:07 pm

    @ Benjamin E. Orsatti. Q (3). As Chau duly noted before, the most common and convenient vessel is made from the top of the head (parietal-frontal bones). There is no need to plug the foramen magnum or caulking minor holes, although the resulting cup is less impressive.

    If you want to share with your buddies, there are other proven means to enjoy skull drinking in style. Each year, in a small village near my home in Navarre (Spain) liberal amounts of wine are funneled through a silver head containing parts of the head (and assorted bones) purportedly from St. Wilhelm. Wine is shared and drunk (with sandwiches).


    Nor far from this place, water is passed through another silver head (this time St. Gregory's). Here the water is not usually drunk, but aspersed on fields.


    You can be confident that nobody takes seriously the crop protection properties of water or the soul cleansing of wine. And yes, you don’t need time travel to the Bronze Age to see some things…

  10. Victor Mair said,

    January 12, 2021 @ 5:05 pm

    From Zhang, He:

    t took me several hours to read the entire thread. Very interesting indeed! Thank you for sharing.

    I am not linguist, so cannot contribute anything specific to "akinakes". But it is always important for me to know issues like these.

    White horse sacrifice is interesting, and Barber's Celtic horse ritual is particularly interesting.

    I saw several reconstructed Scythian horse sacrifice sites at the national museum of Kazakstan in Astana and Almaty. The decorations on the horses are expensive and elaborate, full of gold and silver. But I do not remember their emphasis on the white horse, at least none of the reconstructed horse models is in white color.

    Anyang Yinxu 安阳殷墟 has discovery of extensive horse sacrifices, I do not know if anybody did any connection with the north other than a general comparison.

    I like your mention of “mare” and 馬, and I think they are related.

    I do have a puzzle about the horses from Dayuan. When I was visiting Fergana valley, I expected to see horses or at least a horse farm for tourist. But I did not see any. When I asked people around, they all show a big puzzled face. They said, "we do not raise horses here. Turkmens do. You must go to Turkmenistan to see good horses." In addition, I also noticed that the climate there was not suitable for raising horses. The valley is fertile for farming, but too hot and humid for horses. There is not grassland there either. I just do not understand why the famous horse of Dayuan could be from that place. I even doubted if the place is identified correctly as Dayuan. I wish that somebody notice the same problem and do some research on it. Where did the Dayuan horse go?

  11. Chris Button said,

    January 12, 2021 @ 11:08 pm

    The nasal coda in 徑 would probably have been closer to palatal *-ɲ (from *-ŋʲ via palatalization of the velar -ŋ), which would at least better approximate the coronal /n/ in "akinakes". The earlier *-s at the end of 徑 *káɲs might well have been immaterial if already surfacing as /h/. The *r- (shifting to l-) in 路 is still curious although it would have perhaps served to pull the palatal *-ɲ away from its palatal location to a coronal one as *-n. So, I suppose the phonetics aren't impossible. But using -ɲr- to approximate -n- does seem a little roundabout to say the least!

  12. Benjamin E. Orsatti said,

    January 13, 2021 @ 9:35 am

    Abbas: Fascinating! How curious that what amounts to essentially the same practice is used by some to honor (drinking from the skull of the conquered) and by others to dishonor (drinking from a Saint's relic) the former bearer of a skull. I wonder if both practices come from the same tradition.

  13. Chris Button said,

    January 24, 2021 @ 8:03 am

    In Han Shu, the dagger is 徑路 (jinglu), in Yi Zhou Shu it is 輕呂 (qinglü), and in Shiji 輕劍 (qingjian); they all refer to akinakes. Jaroslav Průšek, in his Chinese Statelets and the Northern Barbarians in the Period 1400-300 B.C. (1971, page 133) states:

    In this connection we must consider whether the name of the knife, or rather the sabre ching-lu, used on this occasion [referring to the friendship treaty ceremony], is not in fact a transcription of the Scythian name for a short sword – or rather of the Greek transcription of the Scythian name, akinakes. Although it is difficult to explain the a in the initial syllable, this etymology in not impossible. As far I am aware, it was first suggested by Japanese authors in Inner Mongolia 1937, English resumé p. 7.

    Pulleyblank (1962:222 "Old Chinese Consonantal System") notes Egami Namio (1948) as trying to make such comparison but calls "the phonetic resemblance very vague".

    I suppose the Turkish "qingiraq" connection in Hirth and Bailey might still be the most promising despite Pulleyblank's cautionary observations there too.

    Pulleyblank does include an interesting note on

RSS feed for comments on this post