7.10 THE CHIONITES.
According to Ammianus Marcellinus, after the third siege of Nisibis in 350 CE, the Sasanian king Šābuhr II was forced to quit the western borders of
Ērānšahr because a new danger had appeared in Central Asia:
For already the Massagetae had invaded Persia and were causing damage there.
It’s precisely at this point in history when the surviving text of Ammianus’
Res Gestae picks up, and for the following twenty-eight years (until the Roman defeat at Adrianople in 378 CE), we enjoy the rare luxury of having a high-quality contemporary source by a direct eyewitness covering the events in the Roman empire. In 350 CE, taking advantage of Šābuhr II’s troubles in Central Asia, Constantius II left the Middle East with his
praesentalis army to fight the usurpation of Magnentius in Gaul, while the western provinces who had been until then ruled by his brother Constans took sides; some aligned with Constantius II, others with Magnentius and Illyricum with the general
Vetranio, who proclaimed himself as augustus, allegedly at the insistence of Constans’ sister Constantina. Constantius II acknowledged Vetranio’s claim, probably in order to divide any possible support for Magnentius; as soon as he reached Illyricum, in December 25, 350 CE both emperors met in public before their assembled armies (either at Serdica, Naissus or Sirmium, the sources are contradictory about the location) and Vetranio abdicated of his titles, and accepted the comfortable retirement in Bithynia that Constantius II offered him. In September 351 CE, the armies of Magnentius and Constantius II clashed at
Mursa in Pannonia (modern Osijek, in Croatia) and in one of the bloodiest and most destructive battles in the history of Roman civil wars, Constantius II inflicted a decisive defeat on his foe. Magnentius though managed to feel to Gaul, where the final battle took place at
Mons Seleucus in July 353 CE (located in the modern commune of La Bâtie-Montsaléon, in the Hautes-Alpes department in southeastern France).
But this was not the end of Constantius II’s troubles in Europe. In order to defeat Magnentius as fast as possible, Constantius II had allied himself with the Franks, Alamanni and Sarmatians, who now refused to retreat from Roman territory, and so Constantius II had to embark in a series of costly and time-consuming campaigns along the Rhine and Danube to restore the European
limes. He also had to deal with the problem of succession. He was childless, so he decided to recall his cousin Constantius Gallus (who with his brother Julian were the only other male members of the House of Constantine still alive and was living in internal exile somewhere in Asia Minor) and had him proclaimed as
caesar in Sirmium on March 15, 351 CE. Gallus was sent to the East as Constantius II’s representative and presumptive heir, but according to Ammianus’ account, his “debauchery” and “misconduct” there was reported back to Constantius II, who recalled him to the West. While traveling to meet Constantius, Gallus was stopped, arrested, tried and executed at
Pola (modern Pula, in Croatia) in 354 CE, after which Constantius II, who was still childless, had no other option but raising his only remaining cousin Julian to the rank of
caesar. Julian was summoned to the West and he was proclaimed
caesar at
Mediolanum (modern Milan) on November 6th, 355 CE, and was married to Constantius II’s sister Helena.
Gold solidus of Constantius Gallus as caesar; mint of Thessalonica. In the obverse: D(ominus) N(oster) CONSTANTIVS NOB(ilissimus) CAES(ar). On the reverse: GLORIA REI PVBLICAE (lit. “glory to the republics”), a celebration of the twin capitals of Rome and Constantinople, symbolized by two female figures wearing mural crowns.
After the fiasco with Gallus, Constantius II did not send Julian to the East, but instead he sent him to Gaul, where the situation was still far from settled, with the Roman generals showing themselves incapable of containing the Frankish and Alamannic warbands that were ravaging the country; according to Ammianus, the situation was so chaotic among other things because Constantius II’s commander in Gaul, the
magister militum Silvanus, was “falsely accused” by Constantius II’s courtiers of plotting an usurpation and was subsequently “forced” to proclaim himself as
augustus in Cologne on August 11th, 355 CE, in a strange episode that some modern historians suspect could have been an invention by Ammianus to justify the role played in the execution/murder of Silvanus by his own patron, the
comes Ursicinus.
Šābuhr II did not restart hostilities against the Romans until 359 CE, and at that date Constantius II and the
caesar Julian were still bogged down in Gaul and Illyricum respectively. According to Ammianus, during these nine years Šābuhr II remained “encamped” in the eastern borders of his empire, without offering more details. According to Ammianus though, the Sasanian commanders along the Roman border launched frequent raids and plundering expeditions against the Roman eastern provinces and Armenia (
Res Gestae, XV, 13.4):
While these men were in league and enriching themselves (i.e. the Roman high officials in the East in Constantius II’s absence) by bringing mutual gain one to the other, the Persian generals stationed by the rivers, while their king was busied in the farthest bounds of his empire, kept raiding our territories with predatory bands, now fearlessly invading Armenia and sometimes Mesopotamia, while the Roman officers were occupied in gathering the spoils of those who paid them obedience.
Following Ammianus’ account, the Praetorian Prefect of the East
Strategius Musonianus opened on his own initiative peace talks with the Sasanian commander across the river
Tamsapor (
Tahm-Šābuhr) in 357 CE. Considering Constantius II’s personality, I find this a bit difficult to believe, and I think that it’s quite more credible that he was acting on Constantius II’s secret orders, in order to avoid any damage to the emperor’s prestige if the negotiations failed. This is Ammianus’ account (
Res Gestae, XVI, 9.1-4):
But the Persians in the East, rather by thieving and robbery than (as their former manner was) in set battles, kept driving off booty of men and animals; sometimes they were successful, being unexpected; again, they lost, overmatched by the great number of our soldiers; occasionally they were not allowed to see anything at all which could be carried off. None the less, Musonianus, the praetorian prefect, a man (as I have said before) gifted with many excellent accomplishments, but corrupt and easy to turn from the truth by a bribe, inquired into the designs of the Persians through emissaries of his who were adepts in deceit and incrimination; and he took into his counsels on this subject Cassianus, dux of Mesopotamia, who had been toughened by various campaigns and dangers. When the two had certain knowledge from the unanimous reports of their scouts that Sapor, on the remotest frontiers of his realm, was with difficulty and with great bloodshed of his troops driving back hostile tribesmen, they made trial of Tamsapor, the commander nearest to our territory, in secret interviews through obscure soldiers, their idea being that, if chance gave an opportunity, he should by letter advise the king finally to make peace with the Roman emperor, in order that by so doing he might be secure on his whole western frontier and could rush upon his persistent enemies. Tamsapor consented and relying on this information, reported to the king that Constantius, being involved in very serious wars, entreated and begged for peace. But while these communications were being sent to the Chionitae and Euseni, in whose territories Sapor was passing the winter, a long time elapsed.
Here for the first time we have the names of the peoples with which Šābuhr II was engaged:
Chionitae and
Euseni; these are not “antique” names like the term
Massagetae that Ammianus used earlier in his text (an archaism borrowed from older Greek authors), but names that are not found in earlier Graeco-Latin texts. The term Euseni has puzzled scholars for a long time, because it does not appear elsewhere in any other western accounts (Greek, Latin, Syriac. Armenian, etc.), and it’s been impossible to relate it to any eastern people known to us. In the early XX century, the German scholar Josef Marquart offered a possible solution by emendating it to
Cuseni, which was the name that Classical authors used to refer to the Kushans. But his proposal has not been accepted by everybody and is still disputed to this day, because it raises new problems. First of all, it’s linguistically impossible for
Cuseni to evolve into
Euseni in Latin, and so the only possible explanation for this evolution of a C into an E is a scribal mistake. That’s a possible explanation, but it’s impossible to prove because the work of Ammianus has arrived to us only through a single manuscript that was copied by a humanist in a German monastery in the XV century and was published in Florence a few years later. Also, the word
Euseni is repeated more than once across the text, and this raises serious doubts among the scholars: one scribal mistake is believable, but repeated mistakes in the same letter of the same word across the same text is quite hard to believe. And in addition to that, there’s of course the historical problem: in 357 CE, the Kushan empire had ceased to exist as such, and only a residual kingdom/principality in northern India remained, possibly beyond the Indus in the Punjab.
Silver drahm of the Sasanian Šahān Šāh Šābuhr II. The king is wearing his usual crenellated crown with a large cloth korymbos atop it, studded with what look like pearls, a typical accoutrement of Sasanian royal regalia that symbolized the kings divine “farr”, their “royal fortune” that entitled them to reign. This crown is exactly the same one worn by his ancestor Šābuhr I, and Šābuhr II probably chose it deliberately, in order to make a statement about his intentions to bring Ērānšahr back to its former glory.
Some historians have proposed that perhaps Ammianus meant that Šābuhr II was campaigning “in the old lands of the Kushan empire”, as they would have been still known to the Romans as “the
Cuseni lands”, and again that’s perfectly possible (indeed, the name “Kushan” showed a remarkable resiliency in Central Asia, and survived into the V and VI centuries CE in Sogdiana and Ṭoḵārestān, as proved by epigraphy and numismatics), but what’s more puzzling is that according to Ammianus the Euseni allied themselves with Šābuhr II in 359 CE and invaded the Roman East with the Sasanian army. Yet another possibility (within the same proposal that
Euseni is an equivalent or mistake for
Cuseni) is that the name
Euseni referred to the Sasanian vassal kingdom of
Kušān Šāhr, and that the participation of the
Euseni in Šābuhr II’s 359 CE invasion simply reflects the participation of the
Kušān Šāh and his army alongside the main Sasanian army, as was to be expected of a vassal kingdom ruled by a cadet branch of the main Sasanian line. And finally, some scholars prefer to see in these mysterious
Euseni a revolt by the native elites of the eastern territories ruled by the Sasanians who identified themselves as “Kushans” (even if they were Bactrians, Sogdians or otherwise) in order to throw off the Sasanian rule over their lands.
Khodadad Rezakhani has also advanced the hypothesis that the
Euseni/
Cuseni could be the Kidarites (a clan or dynasty of the Huns), as in later coinage (still within the late IV century CE) that they minted in the territories that they’d conquered in Central Asia they called themselves
Kušān Šāhs and deliberately appropriated the iconography, regalia and titles of the old Kushan kings and their Sasanian successors. In support of this theory, Rezakhani points out that the Armenian author Faustus of Byzantium also refers to the Kidarites as “Kushans” in his chronicle. But I must repeat once more that none of these explanations have been wholly accepted to this day and so the identity of the
Euseni is still an open issue.
But with the
Chionitae, it’s a whole different issue, because this people has been identified and nowadays most historians accept that this is the first appearance of the Huns in the western and Central Asian historical accounts, although there are still some doubts about it. As I wrote in an earlier post, the defeat of the Xiongnu and the fall of their empire in the eastern steppes at the hands of the Han empire and the Xianbei tribes had been once considered the end of this people, but in the last twenty years a reappraisal of eastern sources (Chinese dynastic chronicles and Sogdian letters), combined with some archaeological evidence has led most scholars to reconsider their position.
In 220 CE, the Han dynasty of China fell after four centuries of rule, and the country disintegrated into several minor states; until the reunification of the territory by the Sui dynasty in 581 CE China stopped exerting a dominant political role over the Tarim Basin and was unable to counter the rise of successive nomadic empires in the eastern steppe, none of which though would be as prestigious, extensive and powerful as the old Xiongnu empire until the rise of the Göktürk Khaganate in the VI century CE. China fragmented into several minor states which were generally short-lived, and the ones in the northern part of the country came under increasing control of dynasties of northern nomadic origin, although their ruling elites soon became Sinicized and imitated the customs and practices of the old Han court, one of which was the redaction of dynastic chronicles.
Immediately after the end of the Han dynasty, China was divided into three kingdoms, this was the Three Kingdoms Period, which lasted from 220 to 280 CE. The northern part of China (the Gansu corridor and the Yellow River valley) became the state of Cao Wei, which lasted from 220 to 266 CE. Due to its geographical corridor and its control of the Gansu corridor which was the main land route between central China and Central Asia, the state of Cao Wei had an interest in the events in the northern steppe. The historian Yu Huan wrote between 239 and 245 CE the
Weilüe (meaning “A brief history of Wei”); this chronicle has been lost in its original form, but quite extensive passages have survived as quotes in later works. And in these quotations, the
Weilüe mentions specifically that at the time of its redaction (the mid-III century CE) the northern Xiongnu still existed as an organized political entity somewhere near the Altai mountains, to the west of their original homeland in the Mongolian steppe.
The Three Kingdoms (Wei, Shu and Wu) that succeeded the unified Eastern Han dynasty in China after its fall in 220 CE.
Another Chinese source that provides us with information about the fate of the northern Xiongnu is the
Weishu. The
Weishu (meaning the “Book of Wei”), is a Chinese historical text compiled by Wei Shou from 551 to 554 CE, and it describes the history of the Northern Wei and Eastern Wei states from 386 to 550 CE. The Northern Wei state, which accomplished the reunification of northern China after 439 CE before becoming split again into Western and Eastern Wei in 535 CE, was ruled by a “barbarian” dynasty, the Tuoba clan, which was of Xianbei origin. The original work had 114 volumes, and its has survived mostly, with minor lacunae which were later “patched up” with material taken from other contemporary works. Again, the geographical situation of the Northern and Eastern Wei states and the ethnicity of their ruling clan (although they became also increasingly Sinicized over time, to the extreme that all Tuoba names were officially forbidden under Emperor Xiaowen of the Northern Wei dynasty in 496 CE and changed to Han names). The
Weishu offers also more information about the whereabouts of the northern Xiongnu after they fled the eastern steppe: towards the beginning of the V century CE, to the northwest of the Rouran (then the ruling power in Mongolia) there were still in the vicinity of the Altai the remaining descendants of the Xiongnu.
Terracotta statuette of a cataphract found in a grave dated to the Northern Wei dynasty (386 – 584 CE).
The
Weilüe provides us with a clear sense of the geopolitical context in which these Xiongnu/Huns were situated in the III century CE. The
Weilüe notes that the
Zhetysu region (in modern eastern Kazakhstan, also called
Semirechye in Russian) directly to the southwest of the Altai (where the Xiongnu were located according to these Chinese sources) was still occupied by the
Wusun people, and the area to the west of this area and north of the
Kangju people (whose territory was centered around the middle valley of the Syr Daryā) was the territory of the
Dingling tribes. The
Wusun and the
Kangju are said in the
Weilüe to have neither expanded nor shrunk since Han times. The Dingling were a Turkic or proto-Turkic people which were quoted in Chinese sources from the II century BCE onwards; at that time, they were living in the are between Lake Baikal and northern Mongolia and were among the tribal groups conquered by Modu
Shanyu, the founder of the Xiongnu empire; from the I century BCE onwards they took part in several large revolts against the Xiongnu. Once more according to Chinese sources, at this time a fraction of the Dingling migrated to the west and settled “in Kangju”, and this is the group that we’re referring to in here.
The
Wusun on the other part are considered by Étienne de la Vaissière and other scholars to have been Iranian speakers (other scholars believe them to have been Tokharian or even Old Indic speakers) and were also one of the tribal groups subjugated by Modu
Shanyu. After their initial defeat by the Xiongnu, the Wusun became close allies of the Xiongnu, and expelled the Yuezhi from the Illi valley (in Zhetysu) where they settled; the Yuezhi fled southwards to Sogdiana and Bactria and from one of their ruling clans would rise the Kushan dynasty. But later, relations between the Wusun and their Xiongnu rulers soured and the Wusun took part in several great revolts against them; during the III century CE they were still settled in Zhetysu, between Lake Balkhash and the Tian Shan Range. Relations between them and the Kangju first and the Sogdian principalities don’t seem to have been very cordial.
Golden belt buckle found in a kurgan belonging to the Wusun culture, III – II centuries BCE, Kazakhstan.
In 1992 came to light the first of series of fragmentary inscriptions on ceramic plaques found at the Kazakh site of Kultobe, by the river Arys. The scholars Frantz Grenet and Nicholas Sims-Williams identified the script in them as Sogdian and translated then in 2006 – 2007. They were discovered in the remains of a citadel on the banks of the river, at the foot of a plateau that had been once inhabited by nomads, as shown by the big kurgans overlooking the valley. The inscriptions could not be dated with any certainty, but linguists believe that on paleographical and grammatical grounds they should be older than the Sogdian letters from the early IV century CE; and as De la Vaissière noted, the situation they seem to describe corresponds very closely to the one described by Chinese sources for the I century CE in that geographical area.
By then, the formerly unified Kangju state, centered on Čāč, had dissolved into a confederation of oasis-states. This situation seems to have lingered up to the late III century CE, as the Paikuli inscription of Narsē seems to describe the same confederacy under the name “
Kēš,
Soğd and
Čāč”. It seems clear that the inscriptions refer to a colonization effort originating from Kangju as a whole; they describe the creation of a border fortress by a general from Čāč with the agreements and participation, and maybe even a supply of colonists, by all the main southern oasis in Kangju (i.e. the Sogdian oasis, which are explicitly named in the inscriptions: Čāč, Samarkand, Naḵšab, Kēš and Bukhara). According to the inscriptions, this strategical move was conducted at the expense of a people called the
wδ’nn’p, which were identified by Étienne de la Vaissière as the equivalent of the Wusun of Chinese sources.
The central and eastern Eurasian Steppe according to the Weishu.
But the geopolitical picture provided by the later
Weishu is one that has drastically changed. According to it, a people called the
Yueban Xiongnu were now occupying the territory of the Wusun and it further points out that these Yueban were a horde of the
Shanyu of the Northern Xiongnu. It tells us that when the Northern Xiongnu were defeated by the Han imperial armies they fled westwards. The weak elements among them were left behind in the area north of the city of Qiuci (now in central Xinjiang). Afterwards, this weak group of Xiongnu is said to have subjected the land of the Wusun to form the new state of Yueban, while the stronger group of Xiongnu/Huns are reported to have headed further west. The
Weishu tells us that the remnants of the defeated Wusun were to be found in the V century CE in the Pamirs. Archaeology in addition to the written evidence shows that the main group of Huns/Xiongnu in the Altai region (i.e. the “strong” Xiongnu as opposed to the “weak” Yueban Xiongnu) had already started to absorb the Dingling Turkic tribes to their west, an area corresponding to modern northern/northeastern Kazakhstan, and the Irtysh and Middle Ob regions (western Siberia) in the III century CE. This corresponds exactly with the areas from which the Huns of Europe and the Huns of Central Asia would later start their trek to Europe and Sogdiana respectively. The
Weishu also confirms that the Central Asian White Huns originated from the Altai region and moved into Central Asia ca. 360 CE, at exactly the same time the European Huns were moving into Europe at the expense of the Alans and later the Goths. The
Weishu further specifically states that the V century CE rulers of Sogdiana, that is the White Huns, were of Xiongnu origin.
The identification of the European Huns with the Central Asian Huns and of both groups with the Xiongnu of the ancient Chinese chronicles has a long story in western historiography since it was first proposed by the French orientalist Joseph de Guignes in his
Histoire générale des Huns, des Mongoles, des Turcs et des autres Tartares occidentaux (1756–1758). This hypothesis has caused incessant debates since its first appearance and rivers of ink to flow, with some scholars being fiercely opposed to it (like the renowned Austrian-American Sinologist Otto J. Maenchen-Helfen) and others supporting it. The most recent trend seems to be in support of it, since the essays written on the subject by the French scholar Étienne de la Vaissière in the 2000s, which have been further expanded by other authorities. The identification of the Huns (European and Central Asian ones) with the ancient Xiongnu is based on several grounds.
First, there is the study of the Chinese sources, which I’ve already exposed above.
Secondly, there’s the linguistic evidence; I’m no linguist (and certainly not an expert in Iranian, Chinese or Turkic languages), but I will try to put together a brief summary in here. Today it seems clear to historians and linguists that the name
Hun was the name that the ancient Xiongnu/Huns used to refer to themselves. Arriving to this conclusion has taken a lot of time and effort to scholars, and I will try now to offer a small glimpse of the principal factors that have been a part in this process. In 1948, the German Iranologist Walter Henning published a new translation of a text that was already known: the second of the
Ancient Sogdian Letters. This letter was written by a Sogdian trader named
Nanaivande somewhere in the Gansu corridor, and was addressed to an unknown recipient in Samarkand; the letter has been dated very precisely to 313 CE. In this letter, Nanaivande informs his correspondent in Samarkand about the recent fall of the capital city of Luoyang in northern China at the hands of the
Xwn (pronounced “Khūn”). This event is also referred to in Chinese sources; according to them, the Southern Xiongnu (a fraction of the Xiongnu that had surrendered to the Eastern Han in the II century CE and had been resettled by them in the Ordos region) conquered in 311 CE the city of Luoyang, which was the capital of the Jin dynasty, burned the city to the ground and captured emperor Huai of Jin. The letter never arrived in Samarkand; together with other letters addressed to correspondents outside China, they were confiscated by the Chinese authorities in Dunhuang, stored and then forgotten until they were discovered in the early XX century.
One of the Ancient Sogdian Letters from Dunhuang.
This exact coincidence of events led scholars to consider that the “barbarians” that the Chinese called
Xiongnu were the same that the Sogdians knew as
Xwn; a name that is here first attested in Sogdian which which would appear with frequency in Sogdian texts in the two and a half following centuries. The Sogdian pronunciation of
Xwn is very close to the pronunciation of
Hun, but what’s even more interesting is that the reconstructed pronunciation of the name Xiongnu according to the pronunciation of Ancient Chinese is also quite similar to it. The Chinese script is a logographic script that has changed very little across the centuries (or at least before the introduction of Simplified Chinese in the PRC in the 1950s), but while the script has changed little, the spoken Chinese language has changed a lot since Han times. To put in in other words: today a cultivated Chinese reader and Sima Qian could understand each other with little problem in writing, but their forms of spoken Chinese would be mutually unintelligible. Thus, historically the spoken Chinese language has been divided into Old Chinese (1250 BCE – 600 CE), Middle Chinese (601 CE – 1200 CE) and Modern Chinese (1200 CE to the present). This means that the set of characters that is read in modern Mandarin Chinese as
Xiongnu was read in a different form in Old Chinese, when the Xiongnu existed and were given this name by the Chinese.
The Chinese name
Xiongnu is formed by two characters: 匈奴. The meaning of these two logograms has not changed since Han times: “howling (or “fierce”) slave”. It was obviously a derogatory name, but modern scholars like De la Vaissière point out that the ancient Chinese, even when they were naming foreign enemy peoples with deliberately derogatory names, usually tried to make these Chinese names as phonetically close to the foreign original names as possible, but they were obviously constricted by the rigidities of the Chinese script (this is an important point; the constraint was caused by the availability of logograms in Chinese script, not by the phonetics of Old Chinese). And so, modern attempts at reconstructing the Old Chinese pronunciation of the first of these two logograms (匈) suggest that the original pronunciation might have been something close to
*ŋ̊oŋ (in linguistics, “*” before a word indicates a hypothetical, reconstructed word), which is again quite similar to
Hun or
Xwn.
But the linguistic clues don’t end here. Another important linguistic testimony in this subject has been transmitted by the Buddhist monk Zhu Fahu. He was born in Dunhuang (just on the Chinese border) in 233 CE, and at the age of eight he became a novice in a Buddhist monastery, where he later took the Sanskrit name
Dharmarakṣa. According to Buddhist tradition, he travelled extensively across the “western regions” and returned to China with a number of Buddhist texts which he then proceeded to translate into Chinese with the help of several associates, both Chinese and foreigners. Zhu Fahu first began his translation career in
Chang'an (present day Xi'an) in 266 CE, and later moved to Luoyang, the capital of the newly formed Jin Dynasty. He was also active in Dunhuang for some time as well and alternated between the three locations. It was in Chang'an that he made the first known translation of the Lotus Sutra and the Ten Stages Sutra, two texts that later became definitive for Chinese Buddhism, in 286 CE and 302 CE, respectively. Altogether, Zhu Fahu translated around one hundred and fifty-four
sūtras. Many of his works were greatly successful, they circulated widely across northern China in the III century CE and became the subject of exegetical studies and scrutiny by Chinese monks in the IV century CE.
The Gate of Jade (Yumenguan), west of Dunhuang, marked the Chinese border since the times of Emperor Wu of the Western Han. These are the remains of the fortress erected by the Han at this location, where the Great Wall ended.
Two of these texts that he translated were the
Tathāgataguhya-sūtra and the
Lalitavistara. The original Sanskrit versions of these two texts included the term
Hūṇa, and Zhu Fahu translated them into Chinese as
Xiongnu. De la Vaissière points out that this choice of words by Zhu Fahu is especially telling, because in the original Sanskrit versions of both texts, this term appeared in lists of peoples written from an Indo-centric point of view. When Zhu Fahu had to translate these lists into Chinese, he realized that it would be pointless to translate it verbatim, as many of these peoples would have been unknown and thus meaningless to a Chinese readership, so he only kept for direct translations those names that had sense for a Chinese audience:
- Sanskrit Pahlava (for the Arsacid Parthians) is translated as Anxi (a name already established in the Shiji).
- Sanskrit Tukhāra (the Kushans) is translated as Yuezhi (already a name established since Han times).
- Sanskrit Yavana (literally “the Greeks”, derived from “Ionian”) is translated as Daqin (the name used since Han times to refer to the Roman empire and for extension to the Hellenistic world).
So, Zhu Fahu’s decision to render
Hūṇa as
Xiongnu is quite significant in this context, and it should also be taken into consideration that
Hūṇa is not an Indian word, so it must’ve come into Sanskrit from a foreign language.
Then there’s also the Sasanian and Mediterranean perspective. As I already noted, the word
Chionitae employed by Ammianus is the first appearance of this word in any know Latin or Greek text. Its origin was a puzzle for historians and linguists until it was pointed out that the term bears a remarkable phonetical similarity to a term found in an apocalyptic Middle Persian text written during the Sasanian period, the
Bahman Yašt, a commentary (
Zand) of an Avestan text now lost. In its fifth and sixth chapters, this eschatological text describes the calamities that would befall the land of the Aryans at the end of the tenth millennium of the Zoroastrian calendar: it would be invaded by the Arabs, Romans, Turks,
Xyōn, Tibetans, Chinese and other foreign peoples who would cause the decay of religion, breakdown of social order, debasement of law and morality, and degeneration of nature.
What’s interesting in this text is the appearance of the term
Xyōn (or
Hyōn), which many historians and linguists think is the source for the Latin term
Chionitae of Ammianus’ text, and which bears obvious phonetic parallelisms with the Sogdian term
Xwn. As with the Chinese term
Xiongnu, the Middle Persian term
Xyōn also plays with a double meaning: on one side, it’s phonetically similar to
Hun, but on another side it’s a derivation of the Avestan term
X’iiaona (or
Hyaona), the name of a tribe which is described in the Avesta as a foe of king Vištāspa, the patron of Zoroaster. In this way, it was possible to establish a clear and deliberate parallelism between the Avestan king Vištāspa and his tribal enemies and the reigning Sasanian
Šahān Šāh and his Central Asian nomadic enemies. Further study of the Zoroastrian Middle Persian literature led to the discovery of further references to the Xyōn; in the Pahlavi tradition the Xyōn are counted among the enemies of the Sasanian
Šahān Šāh Pērōz (459 – 484 CE) in his struggle against the Hephthalites in the later V century CE.
The
Bahman Yašt also makes an interesting reference to further divisions among the Xyōn: the
Xyōn proper, the
Heftal (Hephthalites), the
Karmīr Xyōn (“Red Xyōn”) and the
Sped Xyōn (“White Xyōn”). Scholars have noticed clear analogies between these Middle Persian terms and Sanskrit terms of the Gupta era:
Śveta Hūṇa or
Sita Hūṇa (“White Huns”) and
Hala Hūṇa (“Dark Huns”, or “Red Huns”). The issue though has been further confused because the VI century CE Eastern Roman author Procopius of Caesarea used the expression
Leukoi Ounnoi (“White Huns”) to refer to the Hephthalites that were fighting the Sasanians in the east, while adding that they were referred to as “white” Huns because they were “European-like” in appearance and not “swarthy” like other Huns.
But today, Procopius' explanation has been thoroughly discredited. The American scholar Edwin Pulleyblank pointed out that these names are not artistical licenses or fantasies by Iranian or Indian writers, but that they probably reflect the symbolical and cultural value of colors in the social and political hierarchies of the Eurasian steppe nomadic societies. Pulleyblank pointed out that the color white was simply symbolic of “West” among steppe nomads. Black signified “North” and red the South, hence the existence also of Red Huns (referred to as
Kermichiones in later Greek sources or
Alkhon/
Alkhan -as we’ll see in later chapters- from the Turkic prefix “Al-“ for “scarlet” added to the term “Hun”, meaning “Red Huns”, in coins found in Ṭoḵārestān and Gandhāra), who were the southern wing of the Huns. The Ukrainian scholar Omeljan Pritsak also pointed out that in steppe societies the color black signifying “North” and the color blue signifying “East”, carried connotations of greatness and supremacy and almost always had precedence over white (West) and red (South). Thus, whichever group constituted the Black or Blue Huns (if they existed or are identifiable with known Hunnic groups such as Attila’s Huns in Europe or the Yueban Huns in Kazakhstan) probably possessed seniority over the White Huns, at least initially. The fact that the color black,
kara in Turkic, suggested elevated status among the European Huns also as it did among other Inner Asian Turkic peoples, seems to be confirmed by the report by Olympiodorus of Thebes, an Eastern Roman of the V century CE that the supreme king of the (European) Huns was called
Karaton.
This utterly confusing “salad of names” (
Chionites,
Huns,
Xiongnu,
Xyōn,
Xwn,
Karmīr Xyōn,
Sped Xyōn,
Śveta Hūṇa,
Kermichiones,
Alkhons/
Alkhans, etc.) gets even messier by the fact that it’s quite possible that these Hunnic groups were called also at given times by the name of their ruling clan, like it happened centuries earlier with the Kushans, who were initially just one of the five ruling clans of the Yuezhi people, and whose name was imposed over that of the whole people. This seems to be the case with the Kidarites, Hephthalites and Nēzak Huns, who were named thus after their ruling clan or dynasty. We should keep this in mind when dealing with the Central Asian Huns in the following chapters, or else things will become really confusing.
Archaeology has also played a part in linking the Huns with the Xiongnu, that’s quite remarkable because in the case of nomadic peoples the archaeological record is usually quite poor. Fortunately, the European Huns have left behind a very recognizable type of religious object that is without doubt linked to them, and which has also been found in the central Eurasian Steppe, in Central Asia and in Mongolia, the original home of the Xiongnu. These objects are the Hunnish cauldrons, which have been found always in similar locations: they’ve been deliberately buried near a river or a spring, suggesting some sort of sacral usage. Their form is very characteristic, especially the mushroom-shaped adornments along their rim. Very similar cauldrons found buried in Mongolia have been dated by archaeologists to the times of Xiongnu rule, and they disappear from the archaeological scenario after the demise of the Xiongnu empire. But they’ve been found in the Altai, Dzungaria, the Kazakh Steppe, the southern Urals, the Volga valley and finally in Hungary. Some examples have also been found in Sogdiana. According to De la Vaissière, the distribution and the typologies of these cauldrons suggest two migration or expansion routes for the Huns: firstly, a “Nordic” one that ran across southern Siberia along the limit of the taiga, all the way to Yekaterinburg and the Upper Kama valley in the Urals. The cauldrons associated with this route are not found anywhere else further to the west. The southern route runs all the way from the Altai to Dzungaria, to the Kazakh Steppe, to the lower Volga valley and then to Hungary.
Sacrificial bronze cauldron of Hunnish origin found at Törtel, Hungary.
Another point that has caused lots of controversy among scholars for the past two centuries has been the issue of the Huns’ ethnical identity. Today, this issue has faded into the background, as today we have a more nuanced and detailed understanding about how steppe empires functioned, and it’s quite clear that Eurasian nomads gave little importance to the issue of ethnical or linguistic identity, contrary to what happened in academic circles during the XIX and XX centuries, in the heyday of nationalism. Today, the question of the ethnical identity of the Huns is still an unresolved one. The amount of Hunnic words that has reached our day is quite scarce, to say the least: exactly three words, plus some personal names and tribal ethnonyms. That’s all that remains.
With this meager evidence, some scholars are reluctant to reach any conclusion and just label the Hunnish language as “unclassifiable”, but others are willing to make some speculations. It’s clear that the Huns that reached Europe had their own “Hunnic” language. Most of what we known of the customs of the European Huns comes from the account of the historian and rhetor Priscus of Panium, who was sent as part of an embassy of the Eastern Roman empire to the court of Attila in 448/449 CE. Priscus wrote Attila’s “Scythian” (i.e. Hunnic) subjects spoke:
(…) besides their own barbarian tongues, either Hunnish, or Gothic, or, as many have dealings with the Western Romans, Latin; but not one of them easily speaks Greek, except captives from the Thracian or Illyrian frontier regions.
The three words we have from the Hunnic language are preserved in the writings of Priscus and the VI century CE author Jordanes. In Priscus’ account, we learn that:
In the villages we were supplied with food (millet instead of wheat) and “medos” as the natives call it. The attendants who followed us received millet and a drink of barley, which the barbarians call “kamos”.
And according to Jordanes’ account of Attila’s funeral:
When the Huns had mourned him (Attila) with such lamentations, a “strava”, as they call it, was celebrated over his tomb with great reveling.
According to linguists, these three words (
medos,
kamos and
strava) are of Indo-European origin; either Germanic, Iranian or Slavic. But in the case of personal names and ethnonyms, Hyun Jin Kim points out that the overwhelming majority of them are Turkic ones and concludes that most probably the “European” Hunnic elite was Turkic-speaking, following what seems to be the predominant trend nowadays among historians and linguists.
Hyun Jin Kim though warns against any attempt to seek any sort of genetic/racial continuity between the Huns and Xiongnu. He stresses that “Hun” (or “Wusun”, “Yuezhi”, “Kangju”, etc.) was a political category, not a racial or hereditary one; at its heyday the Xiongnu empire must’ve included speakers of many different languages (Turkic, Mongolic, Yeniseian, Tungusic, Iranian, Tokharian, etc.) and many of them must’ve considered themselves as “Xiongnu” without further issue.
One of the extant Chinese sources seems to indicate that some of the Xiongnu, in particular the Jie tribe of the wider Xiongnu confederation, spoke a Yeniseian language. The
Jin Shu (meaning “The Book of Jin”, an official Chinese historical text covering the history of the Jin dynasty from 265 to 420 CE), compiled in 648 CE in the court of the Tang dynasty, gives us a rare transliteration of a Xiongnu Jie song composed in a language that modern linguists believe to be most likely related to Yeniseian languages. This fact has led scholars such as Edwin Pulleyblank and Alexander Vovin to argue that the Xiongnu had a Yeniseian core tribal elite, which ruled over various Tocharian, Iranian and Altaic (Turco-Mongol) groups. However, whether the Jie tribe and the language they spoke is representative of the core ruling elite of the Xiongnu Empire remains uncertain and other scholars strongly argue in favor of a Turkic, Mongolic or even Iranian ruling elite.
Hyun Jin Kim thinks that it seems rather likely that the core language of the Xiongnu of old was either Turkic or Yeniseian (or maybe even both), but that no definitive conclusions can as yet be made about which linguistic group constituted the upper elite of their empire. He stresses that the attempt itself may in fact be irrelevant since the Xiongnu were quite clearly a multi-lingual and multi-ethnic hybrid entity.
And furthermore, Hyun Jin Kim states that it’s quite possible that the European Huns were equally as heterogeneous as the old Xiongnu of Mongolia. According to him, it’s quite likely that their core language was Oghuric Turkic (a branch of Turkic whose only extant member is Chuvash) given the names of their kings and princes, which are for the most part Oghuric Turkic in origin as the following list shows:
- Mundzuk (Attila’s father, probably from Oghuric Turkic Munčuq, meaning “pearl/jewel”).
- Oktar/Uptar (Attila’s uncle, from Oghuric Turkic Öktär, meaning “brave/powerful”).
- Oebarsius (another of Attila’s paternal uncles, from Oghuric Turkic Aïbârs, meaning “leopard of the moon”).
- Karaton (already quoted above; Hunnic supreme king before Ruga, from Oghuric Turkic Qarâton, meaning “black cloak”).
- Basik (Hunnic noble of royal blood from the early V century CE, probably from Oghuric Turkic Bârsiğ, meaning “governor”).
- Kursik (Hunnic noble of royal blood, from either Oghuric Turkic Kürsiğ, meaning “brave” or “noble”, or Quršiq meaning “belt-bearer”).
Hyun Jin Kim also points out that all three of Attila’s known sons have probable Turkic names: Ellac, Dengizich, Ernakh/Hernak, and Attila’s principal wife, the mother of the first son Ellac, has the Turkic name Herekan, as does another wife named Eskam. The heavy concentration of Turkic peoples in the areas where the Huns dwelled before their major expansion into Europe and Central Asia is likely to have led to the consolidation of a Turkic language as the dominant language among the European Huns.
Map of the Eurasian Steppe.
However, this does not mean that the ethnic composition of the Huns in the central Eurasian Steppe before their entry into Central Asia and Europe in the mid IV century CE was exclusively Turkic. There was also an important Iranian element within their ethnic mix, and this is borne out by the fact that the Central Asian Huns and the Iranian-speaking Alans (the first recorded opponents of the Huns during the Hunnic expansion west into Europe in the mid fourth century AD) shared a very similar material culture. Both groups also practiced the custom of cranial deformation (the origin of which is obscure). Archaeologically it is often very difficult to make a clear distinction between a Hun, an Alan and later even a Germanic Goth due to the intensity of cultural mixing and acculturation between all the major ethnic groups that comprised the population of the European Hunnish empire: Oghuric Turkic, Iranian and Germanic. Just as the Xiongnu accommodated Chinese defectors into their empire, the later Huns also provided refuge for Graeco-Roman defectors and also forcibly settled Roman prisoners of war in their territory. Priscus of Panium leaves us with a vivid image of the heterogeneity of Hunnic society. In the fragment I quoted above, he tells us that at the Hunnic court Hunnic (presumably Oghuric Turkic), Gothic and Latin were all spoken and all three languages were understood by most of the elite to some degree, so much so that Zercon the Moor, the court jester, could provoke laughter by jumbling all three languages together at a Hunnic banquet in the presence of Attila. Remarkably the Hunnic Kidarite empire in Central Asia, which was contemporary with Attila’s Hunnic empire in Europe, also used multiple languages; we know for instance that Sogdian, Bactrian, Middle Persian and Sanskrit on different occasions were all used for administrative purposes.
The fulgurant rise of the Sasanian dynasty in Iran under its two first kings Ardaxšir I and Šābuhr I happened not only because these two rulers were exceptionally good sovereigns (which they were), but also because the geopolitical situation in Eurasia had taken a turn that led to the weakening of the three great powers that until then had dominated the political scene from the Atlantic to the Pacific, and compared to which the Arsacid empire was a relatively minor participant. In 220 CE, the Eastern Han dynasty, which had been experiencing troubles for decades, finally collapsed and China split into three kingdoms and entered three centuries of inner stability that would not end under the reunification of the country by the Sui dynasty in 581 CE. The collapse of the unified Han empire brought about the loss of control over the Chinese military protectorate over the oasis of the Tarim Basin, which had a serious impact over the “Silk Road” trade. When the Chinese armies of emperor Wudi of the Western Han first occupied these oasis, he set up a complex array of Chinese military garrisons designed to control these oasis and defend them from Xiongnu attacks; the soldiers stationed in these remote outposts were mostly paid in silk, which was collected as tribute in kind by the bureaucracy of the Han in inner China and then transported to the borders in form of raw silk rolls that were handed over to the troops as salary. In turn, the troops sold these silk rolls to obtain the commodities needed for subsistence (them and their families), and this quickly created a lucrative trade in which rolls of first-grade raw Chinese silk became an abundant commodity in the Tarim basin and began to be exported to west of the Pamirs, where they were in high demand. The abrupt stop of this system after 220 CE must have hurt a lot the trading communities along the “Silk Road” and must’ve had serious repercussions further west.
Approximately at the same time, the Roman empire entered the “troubles” of the III century CE, originating in a complex array of causes, among them the constant debasement of the silver denarius with which the Roman state paid its armies, which in turn was probably caused by two centuries of deficit trade with the East, in which the Roman empire exported silver in exchange for luxury Indian and Chinese items. In turn, the increased internal instability within the Roman empire, the onset of continued civil wars and foreign invasions brought about probably a sharp stop in the Roman eastern trade.
The state that probable was hit the hardest by these two concurrent crises on both ends of Eurasia was the Kushan empire, which was located in the central Asian space and controlled all the major trade routs and mercantile emporia in Central Asia and northern India. As we’ve seen, there are signs of debasement in Kushan coinage in this century, and one can only imagine the disastrous effect that the slowing down of the Eurasian trade network must’ve had in the trading cities of Bactria and northern India. To the north, the Kangju confederacy (basically Čāč and the Sogdian oasis to the south must’ve been also badly hit, as they stood right on the main gate towards China, at the western approaches to the Pamir Plateau.
This means that the two first Sasanian kings were confronted on all fronts by enemies that were undergoing deep internal crises, and they lost no time in taking advantage of it. The Sasanians managed to inflict humiliating defeats on the Romans, but expansion into the Roman east does not seem to have been a priority for them, as they were mostly content with annexing Mesopotamia and Armenia as bulwarks protecting the approaches to Media in the north and Āsūrestān in the south, and asserting their supremacy over the Caucasus, to seal a possible invasion route for northern nomads into the Iranian Plateau and controlling the branch of the “Silk Road” that arrived there across the Caspian Sea from Khwarazm.
But in the East, the Sasanians took a completely different approach. Ardaxšir I and Šābuhr I basically demolished the tottering Kushan empire and annexed all of its territories north of the Hindu Kush (at the very least, if Šābuhr I did not manage to make inroads into India, as the Manichaean tradition states and the Rag-e Bibi relief in northern Afghanistan seems to suggest). Given that the Arsacid empire in the East at the fall of the dynasty did not extend further than Sistān, Arachosia and Merv, is we are to believe the ŠKZ, these two kings managed to subjugate Makurān, Pāradān, Turān, Hindustān, Gandhāra, the Afghan highlands, Ṭoḵārestān, Sogdiana “all the way to Čāčestān” and Xvārazm, which basically means all the Iranian sedentary peoples and communities in the eastern Iranian Plateau, Baluchistan and the Turanian Basin of Central Asia, and controlling also a part (or even the totality) of the Indus valley. This is a phenomenal expansion that pushed the borders of
Ērānšahr to the same limits of the old Achaemenid empire, and which substantiated the claim of the two first Sasanian kings to be “kings of the Iranians and the non-Iranians”; at the death of Šābuhr I, the only Iranians who were not Sasanian subjects would have been northern nomadic tribes like the Wusun, the Alans of the Sarmatians.
But of course, scholars have been skeptical about the triumphalist claims of Šābuhr I in the ŠKZ. There’s no hard evidence of Sasanian rule north of the Āmu Daryā at this time. Unlike in Bactria, there are no rock reliefs, remains of buildings, or seals or public documents attesting to the presence of Sasanian officials. There are only coins, but there’s a problem with this. Before the reign of Šābuhr III (383 – 388 CE), the practice of stamping mint marks on Sasanian coins was quiet a haphazard one, and most coins issued prior to his reign don’t bear mint marks. And most importantly, there are no Sasanian mints located north of the Āmu Daryā at this time, Numismatics play a very important part in the reconstruction of events in Central Asia in Late Antiquity due to the lack of written sources, and so I will get back in more detail to the issue of Sasanian coinage in later chapters.
Even if we take the ŠKZ at face value and accept that Šābuhr I managed to impose direct or indirect Sasanian rule (through cadet branches of the House of Sāsān like the
Kušān Šāhs), it seems to me rather improbable that his successors would’ve managed to keep such immense tracts of land under their control, considering that all of them had either very short or troubled reigns until Šābuhr II reached his majority of age. In his
Res Gestae, Ammianus Marcellinus describes the Sasanian empire and his list of territories subjected to Šābuhr II includes all these lands, but it’s also suspiciously similar to the description of the old Achaemenid empire that was given by old Greek geographers and is full of anachronisms and mistakes.
So, all things considered, it’s quite difficult to know with any certainty what was the extension of Sasanian rule north of the Āmu Daryā when Šābuhr II was forced to leave his never-ending war against Constantius II and go to the east in a hurry to defend his Central Asian borders against these nomadic invaders who had apparently appeared all of the sudden out of nowhere. But, as Korean historian Hyun Jim Kin writes in his book
The Huns, they did not appear out of the blue, and there’s archaeological proof that from their base in the Altai, the Huns/Xiongnu had been steadily expanding westwards and southwards for some decades before 350 CE.
The Hungarian scholar Miklós Érdy studied the findings of Hunnic cauldrons and concluded that the Huns had reached the Urals in their westward expansion already in 270 CE, which is much earlier than thought before. The Hunnic expansion to the west stalled between the Urals and the Volga for the following century, and the Huns did not cross the Volga in force until 370 CE, to invade the lands of the Gothic Greuthungi. Ands as Hyun Jin Kim points out, this suggests that all the states and tribes between the Altai and the Urals had succumbed to Hunnic conquest by the early IV century CE. But despite this expansion across the steppe, the Huns did not really start moving
en masse to the west and south until after 350 CE. This sudden mass movement is attested by written accounts (Ammianus Marcellinus and Chinese sources) and archaeology. Let’s see first the written evidence, then the archaeological one and finally we’ll try to get a glimpse of what might have prompted this sudden migration.
Distribution of archaeological findings attributable to the Xiongnu/Huns across Eurasia, according to the Hungarian scholar Miklós Érdy.
The first written source is of course Ammianus Marcellinus’
Rerum gestarum Libri XXXI. In it, Ammianus offers a very precise date for the sudden departure of Šābuhr II from the Roman border: 350 CE, the same year of Magnentius’ usurpation, dated in the classical Roman way (consular years). Ammianus also gives us the name of the peoples that caused so many problems to the Sasanian king, to the point that, as we will see, Šābuhr II would need nine full years to bring the situation on his eastern border under control, and then without a clear, smashing victory, but having concluded a peace and an alliance treaty with his enemies that later events would prove to be extremely fragile.
Now the Chinese sources. The
Weishu, describing the situation in Central Asia west of China in the early V century CE, offers this interesting bit of news:
Land of the Yeda, of the race of the Great Yuezhi, it’s also said that they’re another branch of the Gaoju. Originally, they came from a region north of the Sai. They departed the Altai towards the south, they settled to the west of Khotan, their capital is located more than two hundred li south of the Oxus, at 1,100 li from Chang’an.
Chinese authors were prone to the same sort of archaisms as their Graeco-Roman colleagues, which leads sometimes to confused etymologies and garbled ethnological relationships (like the supposed Tokharian Yuezhi ancestry of the
Yeda).
Yeda was the name that the Chinese gave to the Hephthalites, perhaps the more famous of the Hunnish Central Asian dynasties. To this information of the
Weishu, another Chinese source, the
Tongdian (a Chinese institutional history and encyclopedia, covering a wide array of topics from high antiquity through the year 756 CE; the book was written between 766 and 801 CE), offers a key precision: that the departure of the Yeda from the Altai happened between eighty and ninety years before the reign of emperor Wencheng of the Later Wei (r. 452 – 466 CE), which puts us squarely in the 360 – 370 CE time frame, very near to the dates given by Ammianus.
Moreover, the
Weishu, in an entry dated precisely to 457 CE, states:
Formerly, the Xiongnu killed the king (of Sogdiana) and took the country. King Huni is the third ruler of the line.
Soviet and post-Soviet archaeological digs in Central Asia have also confirmed the news of the literary sources about northern invaders moving south. The excavated sites of the sedentary culture of Džetyasar on the Syr Daryā delta shows signs of large-scale destruction and abandonment that have been dated to the late III or IV centuries CE, with the abandonment of irrigation networks. The displaced populations fled in two directions, carrying with them a characteristic style of pottery that has allowed archaeologists to trace their steps; one part settled in the Caucasus and the other sought refuge to the southeast as far away as Ferghana and Sogdiana. In the middle Syr Daryā, the sites of the Kaunči culture, dated back to the establishment of the Kangju domination, show the same signs of destruction, population loss and abandonment of villages and irrigation works; in this case the displaced populations seem to have sought refuge entirely in Sogdiana, where Kaunči pottery becomes commonplace during the mid-IV century CE. At the middle course of the Syr Daryā the city of Kanka, until then one of the most important Sogdian/Kangju cities, diminished to a third of its initial surface area. This timeframe (late III century CE and first half of the IV century CE) also coincides with important changes in the archaeological register of Khwarazm: the palaces of Toprak Kala were abandoned and according to Bīrūnī the Afrighid dynasty rose to power; Soviet and post-Soviet archaeological excavations detect the abandonment of irrigation works, the destruction and abandonment of sites and a general tendency towards the building of fortresses in central points scattered across the agricultural landscape (a process similar to what’s known in Italian medieval history as the
incastellamento).
Further south in Sogdiana the excavations have provided further confirmations for these movements of people. The Kaunči type ceramics become abundant in the central Sogdian oasis, especially in Samarkand. And the mass migration of people from the Džetyasar culture is even more marked; their influence is marked in the ceramics and aspects of domestic architecture in the Bukhara, Naḵšab and Guzar oasis, where archaeologists ascribe to them the appearance of small fortified settlements occupying all the available space within the irrigation network, and increasing to the limit the population densities of these oasis.
Archaeologists have observed that during the IV century the city of Samarkand shrank in size to about a third of its previous surface within the Hellenistic-period walls, and the same can be observed in other Sogdian sites like Erkurgan. The second half of the IV century CE especially witnessed the interruption of regular monetary emissions that dated back to the era of the Graeco-Bactrian kingdom in Sogdiana and Bactria. In the region of Samarkand Soviet archaeologists dated to this age the apparition of a wave of kurgan burials. Everywhere in Sogdiana, but it’s been especially studied in the upper Zarafšān valley, archaeologists have also dated to this period of time the generalization of a new typology of rural habitat formed by small fortified manor houses built in natural defensive locations. Archaeological layers of destruction and fires have been found in many sites in eastern Bactria, while in northern Bactria urban sites saw substantial parts of their walled precincts become depopulated and turned into burial grounds.
Most scholars, like Étienne de la Vaissière and Frantz Grenet, agree that these disruptions were caused by northern invaders, which must’ve been the
Chionitae of Ammianus, the
Xiongnu descendants the of Chinese dynastic chronicles, and the
Xwn of the Sogdian letters. Frantz Grenet and Nicholas Sims-Williams even suggest a complete interruption of the overland trade routes with China as a result of the sack of Luoyang by the Southern Xiongnu in 311 CE, a situation that would not have been remedied until 437 CE. If something even remotely similar to this really happened, the impact on the trading cities of Sogdiana and Bactria would’ve been disastrous and would have left them deeply weakened in front of the northern onslaught. According to the Japanese scholar Kazuo Enoki, in an unknown date during the second half of the IV century CE, the shock of the invasion reached Samarkand, where its king was killed by the invaders according to the
Weishu.
Étienne de la Vaissière argued in the 2000s, that the Huns migrated south in a single wave in the mid-IV century CE from the Altai, in a view that has since then become accepted in academic circles, against the prevalent view until then that saw in the Chionites, Kidarites and Hephthalites successive waves of invaders from the north. What caused this sudden movement?
View of the Altai mountains, which today stand at the geographic junction between Russia, Mongolia, China and Kyrgyzstan.
De la Vaissière mostly attributed this migration to climate events. Published findings regarding accumulations of pollen in the lakes of the Altai range point towards a sharp drop in temperatures combined with a rise in humidity that lasted from the middle of the IV century CE through the VI century CE, causing significant change in the vegetation. Likewise, from 340 CE onwards, glaciers advanced in the valleys. The accumulated snow must’ve destroyed most herds in the high plateaus; although the Mongolian horse is able to dig through the snow to feed, its capacity to do so is strictly limited by the depth of the snow cover, and contemporary ethnography has shown the enormous impact that prolonged winters and their blizzards can have on herds of horses: eight million horses, twenty percent of the stock, died for this reason in Mongolia in the winter of 2010. Chinese sources report Hun invasions from the Altai happening exactly in the middle of the IV century CE, without giving any reason for their incursions. De la Vaissière adds that quite plausibly additional factors contributed to the destabilization of Hun societies in the Altai region, but little is known of them. The north slope of the Altai was well beyond the reach of knowledge for the Chinese observers, the only exception being the
Weishu text mentioned above. Actually, this is one important point that should also be stressed out, as has been done by critics of De la Vaissière and this theory: actually, the
Weishu is a completely isolated testimony, and there are absolutely no other sources that can be used to verify it: similar references that can be found in later Chinese chronicles have actually been borrowed straight from the
Weishu and so are worthless in this respect.
Another possibility, that is not mutually exclusive with the climate hypothesis, is pressure by other nomadic groups. It’s known from Chinese sources that the Rouran/Avar khaganate became active in the IV century CE in Mongolia, even if its power only truly began to develop at the end of that century. De la Vaissière quotes as evidence for this possibility a passage from the V century CE East Roman author Priscus of Panium:
At this time the Saraguri, Urogi and the Onoguri sent envoys to the eastern Romans. These tribes had left their native lands when the Sabiri attacked them. The latter had been driven out by the Avars who had in turn been displaced by the tribes who lived by the shore of the Ocean.
De la Vaissière proposes that the
Sabiri could be a Xianbei people (again, the Chinese characters for the name
Xianbei would’ve been presumably pronounced
*Sarbi in Old Chinese), chased out of Mongolia by the developing power of the Rouran/Avars, and they would’ve chased in turn the tribes further west. In this case, the Hunnic groups cited by Priscus (
Saraguri,
Urogi, and
Onoguri) paused in the Kazakh steppe before moving further westward in the middle of the V century CE, later than the main group of the European Huns who had crossed the Volga in 370 CE.
As I wrote before, it’s unclear where did the real borders of
Ērānšahr lay in this area. Frantz Grenet believed that direct Sasanian rule only included the southern bank of the Āmu Daryā (as far downstream as Khwarazm, where according to the Middle Persian text
Šahrestānīhā ī Ērānšahr the Sasanian king Narsē built a royal city, and perhaps the most westernmost part of Sogdiana (the lower part of the Zarafšān valley, i.e. the Bukhara oasis). He also believed that the Sasanian vassal kingdom (or viceroyalty, as some authors like to call it) of
Kušān Šāhr did not extend north of the Āmu Daryā and must’ve stretched from Herat in the west across Bactriana, crossing the Hindu Kush by way of the Kabul Basin and including Gandhāra.
According to Khodadad Rezakhani, the two last
Kušān Šāhs were:
- Pērōz 2 (303 – 330 CE).
- Bahrām (330 – 365 CE).
After Bahrām
Kušān Šāh, the monetary emissions of these kings in the former Kushan territory cease to exist and become replaced by the coinage of the Sasanian
Šahān Šāh Šābuhr II. According to Rezakhani’s chronology, this happened after 363 CE, and so after Šābuhr II’s victory over the Roman emperor Julian, when he was forced to return once more to the East. Numismatists first suggested and historians seem to have accepted their views that Šābuhr II was somehow forced to terminate the “viceroyalty” of his eastern cousins because
Kušān Šāhr was crumbling under the Chionite/Hunnish attacks and he decided to take personally control over the whole East and recentralize the government, administration and defense of these territories. Numismatic evidence seems to suggest that Bahrām
Kušān Šāh only reigned north of the Hindu Kush, because coins dated to his reign have found in Gandhāra issued by the local satraps
Mēzē and
Kawād in the name of Šābuhr II.
Coin of Pērōz 2 Kušān Šāh.
Gold dinar of Bahrām Kušān Šāh.
The problem, of course, is that all these are educated guesses with little evidence to support them. In the absence of written sources, numismatics and archaeology can only take us so far. Rezakhani summarizes the probable course of events thus: the first wave of invasion targeted the area north of the Āmu Daryā and the invaders initially overwhelmed eastern Sogdiana and the Badaḵšān Valley, east of the Iron Gates, before crossing the Āmu Daryā and moving southwest towards Termeḏ and Balḵ, into Ṭoḵārestān (and thus into the Sasanian “viceroyalty” of
Kušān Šāhr). Rezakhani thinks that is probably the moment when Šābuhr II made a peace treaty with at least some of the clans of these invaders, as attested in Ammianus Marcellinus’
Rerum Gestarum Libri XXI (XVII, 5, 1-2):
In the consulship of Datianus and Cerealis (i.e. 358 CE), while all provisions in Gaul were being made with very careful endeavor, and dismay due to past losses halted the raids of the savages, the king of Persia was still encamped in the confines of the frontier tribes; and having now made a treaty of alliance with the Chionitae and Gelani, the fiercest warriors of them all, he was on the point of returning to his own territories, when he received Tamsapor’s letter, stating that the Roman emperor begged and entreated for peace.
Here a new name appears, that of the
Gelani, a tribe or people that has been impossible to identify. Some historians think that they could have been the ones to give their name to the region of
Gilān in northeastern Iran, but Rezakhani and other historians refute this on the basis that the name
Gilān is not attested until the IX century CE, five hundred years later.
As we will see in the following chapter, after failed peace talks with Constantius II, Šābuhr II marched again against Roman Mesopotamia at the head of an army formed by forces of the eastern part of his possessions, and that included his new Gelani and Chionite allies. The Chionites were commanded by their own king
Grumbates, who rode besides Šābuhr II as an equal, a rare honor in Sasanian state ritual. The etymology of the name
Grumbates has puzzled historians since the XIX century. Ammianus’ Latin form
Grumbates is a direct loan from the Greek form
Grumbatés, but before the 1990s this name was unattested anywhere else. But with the discovery and publication of the Bactrian Letters, it appeared in one of the documents in its Bactrian form,
Gorambado, as the name of a Bactrian aristocrat of the V century CE. The problem is that as this letter is chronologically later than Ammianus’ text by a century, it’s not possible to ascertain if the name already existed in Bactria or if it was adopted in this country as a result of the Hunnish conquest. What’s certain is that until now the name has not been attested in any Middle Persian, Sogdian, Khwarazmian, Chinese or Sanskrit texts. As a result, scholars have proposed historically two different etymologies for this name. Initially, an Iranian etymology was proposed, but more recently a Turkic etymology for this name has gained acceptance, from Old Turkic
*Qurum-pat, meaning "ruling prince". Old Turkic
qurum, meaning "rule, leadership, administration" in the form of
Krum is attested also to in the names of the Turkic Bulgars that appeared in the Volga and the Balkans in the VIII century CE. What seems clear is that according to Ammianus’ text Grumbates and his Chionites were not Zoroastrians (and hence the were probably not Iranians) because they cremated their dead, which was the ultimate abomination for a practicing Zoroastrian.
.
