Chapter 1
Ethnonyms, Ethnic Origins and Ethnicity

The ethnonym Yeda 嚈噠 presents many different forms and transliterations in Chinese and Western historical writings.

In China, Weishu 魏書 gives the ethnonym as “Yeda 嚈噠”; Liangshu 梁書, as “Hua 滑”; Zhoushu 周書, as “Nieda 囐噠”; Suishu 隋書 and Xintangshu 新唐書, as “Yida 挹怛”; Luoyang qielanji 洛陽伽藍記, as “Heda 噠”; Xifanji 西蕃記 and others, as “Yitian 挹闐”; Youyang zazu 酉陽雜俎, as “Yanda 厭達”; Cefu yuangui 册府元龜, as “Yida 悒怛”or “Yida 挹達”; and Xu Gaosengzhuan 續高僧傳”, as “Yanda 厭怛”.

The historians of Byzantium, such as Procopius（I, iii） and others, rendered the ethnonym as “Ephthalitae（Hephthalitae）”; Theophanes Confessor（AM 5967）, as “Nephthalitai”, and Theophylactus Simocattas（VII, 7.8）, as Abdeli. Procopius（I, iii） and others also called the Hephthalites “Huns”; Procopius and Kosmas（XI） more specifically called them “White Huns”; and Priscus,（41）, “Kidarite Huns”.

This ethnonym was rendered “Kushan Huns”in Armenian historical writings, such as Eḷishē’s History of Vardan and the Armenian War （ch. 1）.

They were called “Hūṇa”in ancient Indian writings such as Mah ā bh ā rata （I, III, VI）, Vish ṇ u Pur āṇ a （177）, and Kālidāsa’s Raghuvaṃsa （IV, 67-68）, as they were in inscriptions of the Gupta period. Vārāhamīhira’s Brihatsaṁhita （XI, 61）, composed in the 6th century CE , designated them “Śveta Hūṇa”.

The Hephthalites were rendered as “Hunnāye”or “Kūnānāye（Kūšānāye） d-hinnōn Hunnāyē”in Syriac literature, The Chronicle of Joshua the Stylite （XI）, which was composed in 507 CE .

They were called “Heftal”in Middle Persian writings and “Hētal”in New Persian writings.

They were designated “Hayṭāl”（plural: Hayāṭela） in Arabic texts.

Moreover, the ethnonym was rendered as “Heβtalit”in Sogdian documents and as “Hebdel”in ancient Tibetan documents.

Seen overall, Yeda 嚈噠, Nieda 囐噠, Heda 噠, Yanda 厭達, Yanda 厭怛, Yida 悒怛, Yida 挹怛, Yida 挹達, Yitian 挹闐, Ephthalitai（Hephthalitai）, Nephthalitai, Heftal（Hētal）, Hayṭāl（Hayāṭela）, Heβtalit, Hebdel, Abdeli, and so on are obviously variant transcriptions of the same name. According to Theophanes Byzantios, the name of the Yeda 嚈噠 king who defeated Pērōz [459-484], the king of Sasanian Empire, was Ephthalanus, and he commented that “indeed the [name of that] race was derived from that name [Ephthalanus]”. Liu Fan’s 劉璠 Liangdian 梁典, quoting the item titled “Xirong 西戎 5”of the “Bianfang 邊防 9”section of Tongdian 通典（ch. 193）, also states:

The surname [of king of the] state of Hua 滑 is Yeda 嚈噠. His descendants named the state after his surname. As the name was mispronounced, the state also became known as “Yida 挹怛”.

Thus, we can see that the name of this state was derived from the surname of its king. It is stated that, in all writings, the earliest appearance of the ethnonym was in Weishu 魏書. According to the “Gaozong ji 高宗紀”of Weishu 魏書（ch. 5）, in the second year [456] of the Tai’an 太安 reign-period, the Yeda 嚈噠 first sent an envoy to pay respects at the Wei 魏 court.

The other terms, with the exception of “Hua 滑”as seen in Liangshu 梁書, all occur in connection with the “Huns”or “Hūṇa”. Procopius wrote:

They are the only ones among the Huns who have white bodies and countenances which are not ugly. It is also true that their manner of living is unlike that of their kinsmen, nor do they live a savage life as they do.（I, iii）

Therefore, he called the Yeda 嚈噠“Huns”, at the same time as he also referred to them as the “White Huns”. The contemporary Indians called the Yeda 嚈噠“Śveta Hūṇa”, probably for the same reason.

It is possible that the Yeda 嚈噠 had two ethnonyms. One was derived from the surname of their king, and the other was “Huns”or “Hūṇas”, which does not, however, indicate the ethnic group had the same origins as the Xiongnu 匈奴 of Chinese history or even the Huns of European history. Many nomadic tribes active across the Eurasian steppe from the 4th to the 8th century CE were known as “Xiongnu 匈奴”or “Huns”. Notable examples include the Chionitai, Khazār, Sabir, Onogur, Utigur, Kutrigur, Avar, and so on. Other examples include the Gaoche 高車, who were claimed as “nephews of the Xiongnu 匈奴”in the “Gaoche zhuan 高車傳”of Weishu 魏書（ch. 103）; the Rouran 柔然, who were described as “a detached stock of the Xiongnu 匈奴”in the “Ruirui zhuan 芮芮傳”of Songshu 宋書（ch. 95）; and the Tuoba-Xianbei 拓跋鮮卑, who were recorded as being “of the Xiongnu 匈奴 ethnicity”in the “Weilu zhuan 魏虜傳”of Nanqishu 南齊書（ch. 57）. These tribes’customs were similar to those of the Xiongnu 匈奴 or the Huns, and thus they often called themselves, and were called by others, “Xiongnu 匈奴”or “Huns”. Yeda 嚈噠 was another example of this naming practice.

However, historians at that time did not necessarily have this insight. Procopius, for example, distinguished the Hephthalites from the Huns, calling them “White Huns”or “Hephthalitai Huns”, but he regarded the Hephthalites as a branch of the Huns. The Indians at the time did not seem to have been acquainted with the ethnonym “Hephthalite”and always called them “Hūṇas”or “Śveta Hūṇa”. We cannot know how they viewed the relationship between the Hephthalites and the Hūṇas. As for the terms “Kidarite Huns”, “Kushan Huns”, or “Ḳūšānāyē d-hinnōn Hunnāyē”used by the historians of Byzantium, Armenia, and Syria, these ethnonyms obviously reflect the fact that after the Hephthalites entered Central Asia their rule was centered on Tokhāristan that had originally belonged to the Kushāns and Kidārite Kushāns.

Finally, in my opinion, the pronunciation of “Hua 滑”was the same as that of “Hu 胡”, because according to the item titled “Xia 黠”of the “Rusheng 入聲（entering tone） 14”section of Guangyun 廣韻（ch. 5）, the pronunciation of “Hua 滑”[hәt] is similar to that of “Hu 胡”[ha]. “Hua 滑”was, in fact, simply “Hu 胡”. On the meaning of “Hu 胡”, opinions vary. Chen Yinque 陳寅恪 pointed out: “‘Hu 胡’was originally the proper noun for Xiongnu 匈奴（Huna）. If ‘ na ’is omitted, only ‘ hu ’remains, and that was then simply transliterated as ‘Hu 胡’”. His theory was explicit and remains credible. Objectively, “Hua 滑”is simply a variant transcription of “Huna”.

On this point, historians have always held differing views and their theories can be summarized as falling into three categories, as follows:

1. The ethnonym “Hua 滑”was derived from the name of the king of the Hephthalites, Warz, who was killed by Sinǧibū, the Khan of the Türks, in the second half of the 6th century CE .

In my opinion, this theory is inadequate, because the Hephthalites first paid their respects to the Liang 梁 court in the fifteenth year [516] of the Tianjian 天監 reign-period when the Liang 梁 people already called them “Hua 滑”. Moreover, the “Xibei zhurong zhuan 西北諸戎傳”of Liangshu 梁書（ch. 54） clearly records that the name of their king was “Yandaiyilituo 厭帶夷栗陁”.

2. “Hua 滑”was derived from the name of the capital of the Hephthalites, i.e., an abbreviated transcription of Warwālīz, which was near present-day Kunduz.

In my opinion, this theory is also inadequate, because the capital of the Hephthalites was not Warwālīz, and the Hephthalites had not yet chosen a site for their capital when they first paid their respects to the Liang 梁 court.

3. “Hua 滑”was derived from a toponym, specifically being a transliteration of “Ghōr”, which was located along the upper reaches of the Hari River. “Hua 滑”, i.e., Ghōr, was originally a small state under the rule of the Hephthalites. The reference to “its king Yandaiyilituo 厭帶夷栗陁”in the “Xibei zhurong zhuan 西北諸戎傳”of Liangshu 梁書（ch. 54） and so on only indicates that the state was controlled by Yandaiyilituo 厭帶夷栗陁, the king of the Hephthalites. The state was later destroyed by the Hephthalites, and so thus “Hua 滑”, i.e., Ghōr, became an alternative name for the Hephthalites.

In my opinion, this theory is also inadequate. In the “Xibei zhurong zhuan 西北諸戎傳”of Liangshu 梁書（ch. 54） it is clearly recorded: “In the fifteenth year [516] of the Tianjian 天監 reign-period, its king Yandaiyilituo 厭帶夷栗陁 first sent envoys to present its local products”. The details such as Ghōr having been destroyed by the Hephthalites and so on are all conjecture by the proponent of this theory. The Hephthalites, generally speaking, did not destroy the original native regimes after they conquered Central Asia. The small states, such as Humidan 胡蜜丹, Hebatan 呵跋檀, Zhouguke 周古柯 and so on, which came under Hephthalite control, were all recorded in the “Xibei zhurong zhuan 西北諸戎傳”of Liangshu 梁書（ch. 54）. It would have been an exception to the rule for a nomadic tribe to assume the name of a small place it conquered as its own ethnonym or as the name of its state or polity.

There is evidence for the proposition that “Hua 滑”was a variant transcription of “Xiongnu 匈奴”（Huns）. The item titled “Xirong 西戎 5”of the “Bianfang 邊防 9”section of Tongdian 通典（ch. 193） records:

The state of Hua 滑: Its people were the detached stock of the Jushi 車師. ... [and] was known as Huatun 滑 during the time of the Later Wei 魏 [or Yuan-Wei 元魏] dynasty.

Thus, we know that “Hua 滑”was the abbreviation of “Huatun 滑 ”. The character “tun ”was simply “tun 囤”, whose old pronunciation had been [duәn]. The [d] and [n] were interchangeable, thus it could have been read [nuәn]. “Huatun 滑 ”might well have been read [hun-nuәn], the pronunciation being similar to that of “Xiongnu 匈奴”. ^[1]

The Chinese during the time of the Liang 梁 dynasty believed that “Yeda 嚈噠”was the surname of their king, whose name in complete translation was “Yandaiyilituo 厭帶夷栗陁”, but this was not their real ethnonym, which was rendered as “Hua 滑”or “Huatun 滑 ”. The reason why the ethnonym was not translated as “Hu 胡”or “Xiongnu 匈奴”is probably because people during the Liang 梁 dynasty attempted to distinguish between the Hephthalites and the real Xiongnu 匈奴. However, it is more possible that they only heard the pronunciation and did not know the true meaning. This would be a situation analogous to people in the Northern Wei 魏 dynasty referring to Sogdiana, when conquered by the Hephthalites, as “Wennasha 溫那沙”[uәn-na-shea].

Having clarified the ethnonym of the Hephthalites, we can further explore the ethnic origins and ethnicity of the Hephthalites.

The ethnic origins of the Hephthalites have always been the subject of debate. Among ancient Chinese historians, there were a number of theories—the Gaoche 高車 theory, Jushi 車師 theory, Da Yuezhi 大月氏 theory, and Kangju 康居 theory, and in Western ancient history there was the Hun theory. Eastern and Western scholars have extensively studied the topic since the middle of the eighteenth century, echoing or extending all the old theories apart from the Kangju 康居 theory, or devising their own hypotheses. Among the newer theories are the influential Rouran 柔然 theory, the Mongol theory, Türk theory, Iranian theory, Yueban 悅般 theory, and so on. Contemporary scholars have expended most of their efforts in determining ethnicity on the basis of an investigation of their ethnic origins.

I would first like to say something about the sound and less sound aspects of these various theories before advancing my own tentative views.

1. The Gaoche 高車 theory. This theory first appeared in the “Xiyu zhuan 西域傳”of Weishu 魏書（ch. 102）:

The state of Yeda 嚈噠: Its people were the detached stock of the Gaoche 高車. They originated north of the Great Wall and had moved south from the Jinshan 金山 Mountains; [they had been there for 80 or 90 years by the time of Emperor Gaozong 高宗（Wencheng 文成）]. Their clothing is similar to that worn by the Hu 胡 people, with the addition of tassels. They all clip their hair. Their speech is different from that of the Ruru 蠕蠕, the Gaoche 高車, and the various Hu 胡 people.

Weishu 魏書 was composed from the second year to the fifth year [551-554] of the Tianbao 天保 reign-period of Emperor Wenxuan 文宣 of the Northern Qi 齊 dynasty. At that time, the state of Yeda 嚈噠 had not yet been vanquished and destroyed. According to the “basic annals”of Weishu 魏書, the Yeda 嚈噠 sent envoys to the Northern Wei 魏 on at least thirteen occasions. The Northern Wei 魏 also twice sent envoys to the Yeda 嚈噠. Their exchanges were thus not infrequent. The records in Weishu 魏書 should thus have some basis and substance, and they therefore warrant our attention.

（1） The Yeda 嚈噠“originated north of the Great Wall”: The Hephthalites came from beyond the Great Wall, and thus were heavily influenced by ethnic groups speaking Altaic languages. They not only used titles commonly seen among Altaic tribes, such as Khan, Khatun, Yehu 葉護, and Teqin 特勤, but also followed customs commonly seen among these Altaic tribes, such as live interment with the dead, facial scarification, and ear lopping.

（2） The “detached stock of the Gaoche 高車”: Since the meaning of “detached stock”was often the same as that of “separate tribe”, the two cannot be regarded as being the same ethnic group. The language of the Hephthalites was different from that of the Gaoche 高車, which serves as corroborative evidence.

（3）“[The Hephthalites] moved south from the Jinsha 金山 Mountains; [they had been there for 80 or 90 years by the time of Emperor Gaozong 高宗（Wencheng 文成）]”: The Hephthalites first paid tribute to the Northern Wei 魏 in the second year [456] of the Tai’an 太安 reign-period of Emperor Gaozong 高宗. Therefore, we can infer that the Hephthalites migrated south from north of the Great Wall and crossed the Jinshan 金山 Mountains eighty to ninety years before 456 CE , approximately between the years 366 and 376 CE . In addition, the Hephthalites had already taken control of Tokhāristan when they first paid tribute to the Northern Wei 魏 dynasty. If they indeed moved south from the Jinshan 金山 Mountains, they certainly would have previously occupied Sogdiana, as evidenced by a record in the “Xiyu zhuan 西域傳”of Weishu 魏書（ch. 102）.

（4）“Their clothing is similar to that worn by the Hu 胡 people, with the addition of tassels. They all clip their hair”: Since the addition of tassels and the clipping of hair were Iranian customs at that time, the “Hu 胡 people”in this statement must be a reference to known Iranian tribes. The Hephthalites first occupied Sogdiana then founded their state in Tokhāristan after they had moved south from the Jinshan 金山 Mountains, and so their customs were influenced by the Iranians. The “Xibei zhurong zhuan 西北諸戎傳”of Liangshu 梁書（ch. 54） records the Iranian states near the state of Hua 滑 and states that their clothes were “the same as those of Hua 滑”, which can be regarded as circumstantial evidence.

（5）“Their speech is different from that of the Ruru 蠕蠕, the Gaoche 高車, and the various Hu 胡 people”: It is generally acknowledged that the Gaoche 高車 were a Turkic linguistic group, and the Rouran 柔然, a Mongolic linguistic group, but one cannot, on this basis, assume that the Hephthalite language was neither Turkic nor Mongolic, or still less that the Hephthalite language did not belong to the Altaic language family. This is because the Altaic family was not simply confined to the Turkic and Mongolic groups; nor did a language branch necessarily simply have one language. The ancients obviously did not follow the present scientific classificatory systems, and their descriptions of languages were inevitably general. This monograph similarly also states that the Hephthalite language was not the same as that of “the various Hu 胡 people”, namely the known Iranian tribes at that time, but one cannot rule out the possibility that the Hephthalite language belonged to an Iranian or Indo-European linguistic group.

In sum, the value of the above-quoted record in the “Xiyu zhuan 西域傳”of Weishu 魏書（ch. 102） is that it points out that the Hephthalites originated from beyond the Great Wall, and crossed the Altai Mountains and moved south in the 360s or 370s CE . However, based on this record, one cannot determine their ethnicity and language.

It has also been suggested that the Hephthalites were the Gaoche 高車（Chile 敕勒）. This theory maintains that the most of the Xiongnu 匈奴 moved to Europe after they were expelled from north of the Great Wall. The remnant Xiongnu 匈奴 moved to Aksu and Kashgar, and then moved west to Caspian Sea and the eastern boundary of Persia. The remnants were the Tiele 鐵勒 or Chile 敕勒, who lived between the Amu Darya and the Syr Darya, where there were many marshes, and thus were called “Ab-te-lé”, the meaning of which is rendered in French as “Telites d’eaux”（i.e., the Tiele 鐵勒 of the Waters）. The ethnonyms Abtelites, Euthalites, and Nepthalites were all derived from this.

In my opinion, this theory seems barely tenable. As previously mentioned, the “Xiyu zhuan 西域傳”of Weishu 魏書（ch. 102）, in fact, did not identify the Gaoche 高車 as the Hephthalites. Moreover, the “Xiyu zhuan 西域傳”of Weishu 魏書（ch. 102） clearly recorded the differences between the Gaoche 高車 and the Hephthalites. Language aside, it also recorded that the state of Yeda 嚈噠 was without chariot pulled by horses but had cart pushed or pulled by persons, as well as an abundance of camels and horses. The “Gaoche zhuan 高車傳”of Weishu 魏書（ch. 103） recorded of Gaoche 高車: “Their carriage wheels are tall and large, with a lot of spokes”. It is obvious that the two would not have been confused.

2. The Jushi 車師 theory. This theory first appeared in the “Xibei zhurong zhuan 西北諸戎傳”of Liangshu 梁書（ch. 54）:

Its people were the detached stock of the Jushi 車師. In the first year of the Yongjian 永建 reign-period of the Han 漢 dynasty, Bahua 八滑 followed Ban Yong 班勇 to attack the northern savages and rendered great service. Thus [Ban] Yong [班]勇 submitted a written report to request [the Emperor] make Bahua 八滑 as Marquis Qinhan 親漢（Enjoying Friendly Relations with Han 漢） in the tribe of Further Jushi 車師. From Wei 魏 to Jin 晉 times it did not communicate with the Middle Kingdom. In the fifteenth year of the Tianjian 天監 reign-period, its king Yandaiyilituo 厭帶夷栗陁 first sent envoys to present its local products.

The deeds of “Bahua 八滑”are recorded in the “Xiyu zhuan 西域傳”of Houhanshu 後漢書（ch. 88）:

In the first year [126] of the Yongjian 永建 reign-period of Emperor Shun 順, [Ban] Yong [班]勇, leading Jiatenu 加特奴 and Bahua 八滑, the sons of the king of Further Jushi 車師, Nongqi 農奇, and others, called out crack troops to attack King Huyan 呼衍 of the northern savages and defeated him. [Ban] Yong [班]勇 sent a written message to the Emperor, asking permission to establish Jiatenu 加特奴 as king of the tribe of Further Jushi 車師 and Bahua 八滑 as Marquis Enjoying Friendly Relations with Han 漢 in the tribe of Further Jushi 車師.

“The tribe of Further”was an abbreviated reference to the tribe of Further Jushi 車師. Based on this, the “Hua 滑”people were the progeny of Bahua 八滑, Marquis Enjoying Friendly Relations with Han 漢 of the so-called “Tribe of Further”in Eastern Han 漢 times, and thus the Hua 滑, i.e., the Hephthalites, were derived from the Jushi 車師.

The main basis of the Jushi 車師 theory is the above-quoted record in the “Xibei zhurong zhuan 西北諸戎傳”of Liangshu 梁書（ch. 54）. In addition to this, there is some evidence regarding physique, facial features, and language. For example, as a proponent of the theory has pointed out, the complexion of the Hephthalites was whiter and so they were known as the “White Huns”, and therefore they should have been a white-skinned group. Moreover, their language was different from that of the Rouran 柔然, Gaoche 高車, and the Iranian tribes, and so they could have spoken Tocharian, because it is generally acknowledged that the Jushi 車師 people were West-Eurasianoid who spoke “the so-called Tocharian language”. In other words, the record on the ethnic origins of the Hua 滑, i.e., the Hephthalites, in the “Xibei zhurong zhuan 西北諸戎傳”of Liangshu 梁書（ch. 54） is credible.

In my opinion, although there is no lack of favorable footholds for this theory, there are also many flaws in the theory. The related records in the “Xibei zhurong zhuan 西北諸戎傳”of Liangshu 梁書（ch. 54） do not derive from the envoys of the Hephthalites, and there is no other evidence, so we can say that the theory was simply determined arbitrarily by Pei Ziye 裴子野. In the “Pei Ziye zhuan 裴子野傳”of Liangshu 梁書（ch. 30） it is recorded:

At that time, envoys came via the Minshan 岷山 Mountain Route from the states of Baiti 白題 and Hua 滑, both beyond the northwestern frontier, to present tribute. These two states had never had any contact with [the Middle Kingdom] and no one knew anything about their origin. [Pei] Ziye [裴]子野 said, “Marquis Yingyin 穎陰 of Han 漢 killed a Hu 胡 general from Baiti 白題. Fu Qian’s 服虔 commentary reads, “Baiti 白題 is a Hu 胡 name”. On another occasion, Marquis Dingyuan 定遠 attacked the savages, and Bahua 八滑 followed him. Could this [Hua 滑] have been a descendant of [Bahua 八滑]”? The people at that time were impressed by his expansive knowledge. The Emperor therefore ordered him to compile Fangguoshitu 方國使圖, which included twenty states in total, from the remote areas to overseas, in order to provide a sweeping description of the splendor that attracted many states to pay their respects [to Liang 梁].

From this, it can be seen that the Chinese in Liang 梁 times were utterly ignorant of the real origins of the Hua 滑（“no one knew anything about their origin”）. The reason why Pei Ziye 裴子野 decided that the “Hua 滑”were descendants of “Bahua 八滑”was simply because of the coincidental similarity of the Chinese characters used in the names.

The “Xibei zhurong zhuan 西北諸戎傳”of Liangshu 梁書（ch. 54） also records: “The state of Baiti 白題: Its king is surnamed Zhi 支 and named Shijiyi 史稽毅. Its ancestors were the detached stock of the Xiongnu 匈奴. When he fought with the Xiongnu 匈奴 in Han 漢 times, Guan Ying 灌嬰 killed a cavalryman from the Baiti 白題 [state]. At present, the state is to the east of the state of Hua 滑 at a distance of a six-day’journey. To the west one proceeds and reaches Bosi 波斯”. Needless to say, this passage also adopted Pei Ziye’s 裴子野 theory. “Baiti 白題”seems to have been the state of “Bati 跋提”as recorded in the “Travels of Song Yun 宋雲”quoted in Luoyang qielanji 洛陽伽藍記,（ch. 5）. From the “Zhi 支”, the surname of their king, it is possible that the state was a regime of a Yuezhi 月氏 leader, but it surely would have had nothing to do with the “general from Baiti 白題”mentioned in the “Guan Ying zhuan 灌嬰傳”of Hanshu 漢書（ch. 41）. The reason why the mistake could have arisen is similar to the reason why the proponent of the theory determined that the “Hua 滑”were the descendants of “Bahua 八滑”.

Such erroneous analogies can also be found in the “Xibei zhurong zhuan 西北諸戎傳”of Liangshu 梁書（ch. 54）, for example, the identification of the “state of Mo 末”as the state of Qiemo 且末 in Han 漢 times.

Since the Hua 滑, i.e., the Hephthalites, had nothing to do with the Jushi 車師, there is no need to examine seriously the ethnicity and language of the Jushi 車師. The record on the language of the Hephthalites in the “Xiyu zhuan 西域傳”of Weishu 魏書（ch. 102） is insufficient to indicate whether the language of the Hephthalites belonged to either the Altaic family or the Indo-European family. In the same way, simply on the basis of the Byzantine record that the complexion of the Hephthalites was whiter（paler）, one cannot determine that they were a white ethnicity, because complexion is not an indicator of ethnicity, nor is it of any importance.

In sum, the above-mentioned Jushi 車師 theory is difficult to establish.

3. The Da Yuezhi 大月氏 theory. This theory first appeared in the “Yiyu zhuan 異域傳 B”of Zhoushu 周書（ch. 50）:

The state of Nieda 囐噠: Its people are descended from the stock of the Da Yuedi 大月氐. It is located to the west of Yutian 于闐, and to the east it is 10,100 li 里 from Chang’an 長安. The seat of the king’s government is at the town of Badiyan 拔底延, [which means something like “the town of Rājagṛha”]. This town is over ten li 里 square. Its penal laws and customs are roughly similar to those of the Tujue 突厥（Türks）.

The “Xiyu zhuan 西域傳”of Suishu 隋書（ch. 83） also records:

The state of Yida 挹怛: Its capital is over 200 li 里 to the south of the Wuhu 烏滸 River. Its people are descended from the stock of the Da Yuezhi 大月氏.

The dates when Zhoushu 周書 and Suishu 隋書 were composed are later than the date of the composition of Weishu 魏書, and the reliability of the records on the ethnic origins of the Hephthalites in the former two books should fall short of the latter account. The Chinese in Northern Zhou 周 and Sui 隋 times were utterly ignorant regarding the origins of the Hephthalites. The Hephthalites consistently ruled with Tokhāristan as their center, from shortly after their move into Central Asia. Tokhāristan was, moreover, successively under the Kushāns and Kidārite Kushāns from the Eastern Han 漢 onwards and the Chinese, “basing themselves upon the old appellation”, described the Kushāns and Kidārite Kushāns as the “Da Yuezhi 大月氏”. Therefore, the Hephthalites were also mistakenly regarded as being of “Da Yuezhi 大月氏”stock. The related records in the “Yiyu zhuan 異域傳 B”of Zhoushu 周書（ch. 50） and the “Xiyu zhuan 西域傳”of Suishu 隋書（ch. 83） all describe the situation after the Hephthalites chose the site for their capital south of the Amu Darya, and it is possible that in the minds of their contemporaries they had always lived there.

Apart from citing the records in Zhoushu 周書 and Suishu 隋書, the main evidence put forward by those who uphold the Da Yuezhi 大月氏 theory is simply that “Yuezhi 月氏”and “Yeda 嚈噠”were variant transcriptions of the same name. For example, it is suggested that the pronunciation of “Yeda 嚈噠”[I-ta, Y-tien or Yeta] was the same as that of “Yuezhi 月氏”[Yé-tha or Yue-tchi], and that groups both lived on the banks of the Amu Darya. They were successively called “Yuezhi 月氏”and “Yeda 嚈噠”in Chinese histories, but they were, in fact, one and the same ethnic group. Another variant of the theory proposes that the original pronunciation of “Yuezhi 月氏”or “Yuedi 月氐”may have been Ghuthal, which was mistakenly changed into Yuttal or Yettal, as well as Haythal or Ephthal, and further changed into Khuttal, Khotthal, Khottalan, and so on. Nieda 囐噠, Heda 噠, and so on were thus transliterated from Ghuthal（Khuttal）, and Yeda 嚈噠, Yida 挹怛 and so on, from Yuttal（Yet-tal）, ^[2] etc.

In my opinion, the records on the ethnic origin of the Hephthalites in the “Yiyu zhuan 異域傳 B”of Zhoushu 周書（ch. 50） and the “Xiyu zhuan 西域傳”of Suishu 隋書（ch. 83） were founded on a misunderstanding, as discussed previously. There is also the record that the state of Da Yuezhi 大月氏 under King Kidara were driven from Tokhāristan by the Hephthalites in the “Xiyu zhuan 西域傳”of Weishu 魏書（ch. 102）. It is obvious that, simply on the basis of the coincidentally similar pronunciations of ethnonyms, one cannot readily determine that both stemmed from the same source or were the same ethnic group.

In addition, on the basis of the Da Yuezhi 大月氏 theory, it has been further inferred that the Hephthalites were Tibetans. This theory has many errors and the only evidence, quite forced, would seem to be that, according to the “Yiyu zhuan 異域傳 B”of Zhoushu 周書（ch. 50） and the “Xiyu zhuan 西域傳”of Suishu 隋書（ch. 83）, the Hephthalites practiced polyandry, and the Tibetans also had this custom.

In my opinion, polyandry is an exceptional form that appears in a particular developmental stage of the family and it has nothing to do with ethnicity. In fact, this custom was not an original custom of the Hephthalites, and was not unique to Tibetans, so we cannot infer that the Hephthalites were Tibetans.

4. The Kangju 康居 theory. The item titled “Xirong 西戎 5”of the “Bianfang 邊防 9”section of Tongdian 通典（ch. 193） states: “Wei Jie’s 韋節 Xifanji 西蕃記 records that he ‘personally asked the natives, [who told him] they all called their state Yitian 挹闐’. In Hanshu 漢書, the viceroy of Kangju 康居, named Yitian 挹闐, plundered the provisions of Han 漢 arms when Chen Tang 陳湯 marched against Zhizhi 郅支. This may mean that they were descendants of the Kangju 康居. However, the information has come from remote countries and foreign languages are subject to corruption and misunderstanding and, moreover, it concerns events that were very remote in time. So we do not know what is certain. [In this way] it is impossible to determine [the origins of the Hephthalites]”.

In my opinion, the advocate himself has severe doubts about this theory, and its mistakes are obvious. Chen Tang’s 陳湯 march against Chanyu 單于 Zhizhi 郅支 was recorded in the “Fu, Chang, Zheng, Gan, Chen, Duan zhuan 傅常鄭甘陳段傳”of Hanshu 漢書（ch. 70）, and “Yitian 挹闐”is read as “Baotian 抱闐”in the current edition of the work.

5. The Hun theory. The Hephthalites were often known as “Huns”or “Hūṇa”in Western historical writings and modern historians have elaborated on this identification to propound the Hun theory. Apart from citing Byzantine records and Indian inscriptions, some historians also proposed the following as evidence proving the theory:

（1） The Huns who invaded Europe were the Türk-Tartars.

（2） Jauvla, the title of Toramāṇa, the king of the Hūṇas who invaded India, was Türk.

（3） According to the Táríkhu - l Hind of Bírúní, Kabul was ruled by the Türks when it was invaded by Islamic forces from 664 CE onwards. Wukong 悟空 also recorded that Gandhāra and Kabul were under Türkic control in 753 CE .

In my opinion, this theory is unsustainable because on the basis of names alone it is impossible to determine whether two ethnic groups both derive from the same ethnic group. For this reason, there is no need to enter into a deep discussion of the ethnicity of the Xiongnu 匈奴 or the Huns. Since the Hephthalites originated from the area north of the Great Wall, they would have been inevitably influenced by the tribes speaking languages of the Altaic family, including the Turkic branch, but the similarity of one or two words is not enough to verify that the Hephthalites did in fact belong to a Turkic speaking ethnic group. As for the records of Bírúní and Wukong 悟空, these discuss the situation of the Western Türks, had nothing to do with the Hephthalites.

6. The Rouran 柔然 theory. It is said that Shelun 社崙 created the first dynasty of the White Huns and styled himself Khan Qiudoufa 丘豆伐 in 402 CE . After a few generations the throne was passed on to Wuti 吳提, Khan Chilian 敕連. Wuti 吳提 invaded the Sasanian Empire during the reign of Bahrām V. After Wuti 吳提 died, Khan Chu 處 ascended the throne. After this Khan died, Yucheng 予成, Khan Shouluobuzhen 受羅部真, ascended the throne. The latter Khan was simply Khūshnāwaz who helped Perōz assume the throne, as recorded in Persian historical writings. Emperor Xianzu 顯祖 of the Northern Wei 魏 destroyed Yucheng 予成 and killed over 50,000 troops of the White Huns in 470 CE . When Yucheng 予成 died in 485 CE , his son, Doulun 豆崙, ascended the throne and styled himself Khan Fugudun 伏古敦. His general, Afuzhiluo 阿伏至羅, rebelled in 494 CE and the first dynasty of the White Huns thus fell. Afuzhiluo 阿伏至羅 was simply Ephthalanus who defeated Perōz, as recorded by Theophanes Byzantios. The theory maintains that the name of the Hephthalites was derived from Ephthalanus, and so on. The error can be seen at a glance, but from this, we can see the progressive trajectory of the study of the Hephthalite history.

7. The Mongolic theory. Its proponent suggests that, apart from calling the Hephthalites “Śveta Hūṇa”, Indian records also referred to the Hephthalites as “Hāra Hūṇa”. “Hāra”would have been qara （black） in Turkic or Mongolic languages, and Hāra Hūṇa was simply Qara Qūn. From this, it can be seen that the term “Śveta Hūṇa”was originally Čaγan Qūn, Čaγan meaning “white”in Mongol. Čaγāniyān, a town in Tokhāristan, gained its name, because it was occupied by the Hephthalites. For these reasons, the Hephthalites would have belonged to an ethnic group speaking a Mongolic language of the Altaic family.

In my opinion, opinions vary regarding the annotation of the name “Hāra Hūṇa”, and there has been no general agreement to date. The conclusion of the proponent of this theory is not necessarily credible. It is difficult to imagine that the Indians called the Hephthalites “Sveta Hūṇa”at the same time as they called them “Hāra Hūṇa”. Much less, even if the above-mentioned theory is correct, can this be taken as evidence determining the linguistic attribution and ethnicity of the Hephthalites.

8. The Türk theory. The proponents put forward the following evidence:

（1） The Hephthalites who invaded the Sasanian Empire during the reign of Bahrām V [421-439] were called Türks by Ṭabarī.

（2） The ruler of the Hephthalites in Gandhāra took the title “Teqin 特勤”（Tigin）, which was a common title for an official position of the Türks.

（3） According to Tāranātha’s History of Buddhism in India , in the times of Asaṅga and Vasubandhu the ruler in Kashmir was Mahāsammata, a king of the Türks. The king was an adherent of Buddhism; he occupied Tokhāristan, Ghazni, and so on, and established temples throughout Kashmir where the disciples Vasubandhu, Saṁghadāsa, and others were highly esteemed. However, the father of Mahāsammata persecuted Buddhists and damaged halls, stūpas, and monasteries when he ruled over Kashmir. Since Vasubandhu lived around Gupta times, the above-mentioned king of the Türks should have been a king of the Hephthalites.

（4） Chinese historical writings record that the customs of the Hephthalites were the same as those of the Türks.

In addition, some scholars argue that because the Hephthalites originated in the Altai Mountains, this is enough to prove that the Hephthalites belonged to an ethnic group speaking a Turkic language of the Altaic family because the Altai Mountains were the birthplace of the Turkic branch. As for the several customs that were similar to those of the Iranian ethnic groups, these were the result of influences to which they became subject after moving west.

In my opinion, the above-mentioned evidence is not sufficient to verify that the Hephthalites belonged to an ethnic group speaking a Turkic language:

（1） The “Türks”who invaded the Sasanian Empire during the reign of Bahrām V, as recorded by Ṭabarī, were indeed the Hephthalites. However, this cannot be taken as evidence to verify that the Hephthalites and the Türks were the same ethnic group. “Türks”for Ṭabarī was the general term used by the Iranians for non-Iranian nomadic tribes.

（2） Similar customs, titles, and so on may have been the result of cultural influences, and cannot provide evidence for determining ethnicity. Of the customs of the Hephthalites, some were similar to those of the Iranians, as some titles were the same as those of the Iranians, but those upholding this theory have not determined that the Hephthalites were an Iranian ethnic group.

（3） Tāranātha lived in the eighteenth century CE , and what he wrote must be taken with a grain of salt. The Hephthalites neither destroyed Buddhism nor believed in Buddhism. This alone indicates that Mahāsammata and his son had nothing to do with the Hephthalites.

（4） That the birthplace of the Turkic tribes was in the Altai Mountains has not yet been confirmed. Weishu 魏書 simply recorded that the Hephthalites originated from north of the Great Wall and moved south by the way of the Altai Mountains, and nothing more.

In addition, it is suggested that “-l”, the suffix of “haftal”or “haptal”（i.e., Ephthalitae） is the suffix of all or part of a collective or plural noun form in Altaic languages. In my opinion, the Hephthalites did not hand down their literature, and so we cannot know what term they used to describe themselves. Even if it can be verified that the ethnonym in Western literatures is a combination of “hafta”（hapta） and “-l”or “Eftal”and “-itae”, it is impossible on this basis to identify the Hephthalites as an ethnic group speaking an Altaic or Turkic language.

9. The Iranian theory. This theory has a great influence among scholars. Here I would like to present my own views briefly on its representative points.

The most important evidence presented by those proposing this theory is that the birthplace of the Hephthalites was located in the eastern part of Tokhāristan, which would deny the record in the “Xiyu zhuan 西域傳”of Weishu 魏書（ch. 102） that the Hephthalites originated from north of the Great Wall and moved south from the Altai Mountains.

Based on the research regarding the composition of the original manuscript of the “Xiyu zhuan 西域傳”of Weishu 魏書（ch. 102）, the proponents point out that in the report of Dong Wan 董琬 and Gao Ming 高明 there is no mention of the Hephthalites, which shows that the Hephthalites had not yet appeared in Sogdiana and Tokhāristan（although in fact this was not the case） as late as 437 CE . Therefore, the statement that they “moved south from the Jinshan 金山 Mountains”in the “Xiyu zhuan 西域傳”of Weishu 魏書（ch. 102） is untenable. Moreover, they argue that the Hephthalites did not come into conflict with the Gaoche 高車 in Dzungaria and annex the various states on the Southern and Northern Route until the end of the fifth century or the beginning of the sixth century, which indicates that the Hephthalites originated in Tokhāristan, then expanded eastwards from the west.

In my opinion, the fundamental weakness of this theory is that it overlooks the fact that the Hephthalites were a nomadic tribe. As its proponents argue, the date when the Hephthalites extended their influence to the Tarim Basin and Dzungaria was indeed after they had taken control of Tokhāristan, i.e., at the end of the fifth century or the beginning of the sixth century. However, this fact does not contradict the records in the “Xiyu zhuan 西域傳”of Weishu 魏書（ch. 102）, because the Hephthalites could certainly have expanded eastwards after having moved south from the Altai Mountains to Tokhāristan and gaining a firm foothold there.

The proponents of the theory, quoting copiously from many sources, identify four centers of Hephthalite activity in Tokhāristan in order to strengthen their argument—Ghōr, Balkh, Warwālīz（the town of Huo 活 or Ahuan 阿緩）, and Ximodaluo 呬摩呾羅（Himatala）, and point out that only Ximodaluo 呬摩呾羅 could have been the birthplace of the Hephthalites. However, as the “Xiyu zhuan 西域傳”of Weishu 魏書（ch. 102） stated, the Hephthalites “migrate to cool lands in summer and to warm places in winter”, as long as there is abundant water and fertile grasslands where they can graze and settle. For this reason, even if, as the proponents argue, there were the four centers of activity in Tokhāristan, this does not help establish the theory that the Hephthalites originated in Tokhāristan.

As for the argument that Ximodaluo 呬摩呾羅 was the birthplace of the Hephthalites, the main basis for this is the following record in Datang xiyuji 大唐西域記（ch. 12）:

The state of Ximodaluo 呬摩呾羅（Himatala）, which is a former territory of the state of Duhuoluo 覩貨邏（Tukhāra）, is more than three thousand li 里 in circuit. The land has a chain of mountains and valleys, and the soil is fertile, good for growing crops and producing abundant winter wheat. Vegetation of diverse kinds thrives, and fruits of every variety are profuse. The climate is bitterly cold, and the people are violent and impetuous in disposition, not knowing what is iniquitous and what meritorious. They are short and ugly in their features, and their ways and manners, as well as their garments made of felt, fur, and hempen cloth, are quite the same as those of the Tujue 突厥（Türks）. Their married women wear a wooden horn about three feet high as a headdress, with two branches in front to represent their husbands’parents, the upper branch indicating the father and the lower one standing for the mother. When one of the parents is dead, the branch indicating the deceased person is removed, and when both are dead, the horn headdress is discarded altogether. Formerly this was a powerful country and the royal lineage was of Shi 釋（Śākya） stock. Most of the states west of the Congling 葱嶺 Mountains were subject to this state. As the land is linked with the territory of the Tujue 突厥（Türks）, the people of this state are influenced by Turkish customs. As they have to guard their own land against invasion and pillage, they wander about leading a vagrant life in different regions. There are several tens of strong cities, each having its own ruler. The people live in domed felt tents and move from place to place.

The proponents of the theory make the following points on this passage:

（1）“Most of the states west of the Congling 葱嶺 Mountains were subject to this state”: This refers to the history of the Hephthalite conquest of Central Asia.

In my opinion, this is not true. There is a detailed description of the history of the king of Ximodaluo 呬摩呾羅 in Datang xiyuji 大唐西域記,（ch. 3）, and according to this chapter, the king “fully occupied his territory and ascended the throne in the six hundredth year after the Tathāgata’s Nirvana. He planted his mind in the earth of the Buddha and poured his sentiments into the sea of the Dharma”. Hearing that the Krītas had overthrown the law of Buddha, he marched and destroyed them. “When order was restored in the country, the monks were invited back, and monasteries were built as peacefully as before”. Unfortunately, the Hephthalites did not believe in Buddism, and could not have played a role in protecting the Dharma.

（2） The horned headdress of the women of Ximodaluo 呬摩呾羅 must have been the Hephthalite women’s horned hat. It is known that Ximodaluo 呬摩呾羅 was the largest settlement of the Hephthalites by Xuanzang’s 玄奘 time.

In my opinion, this theory is inadequate. This is because the horned hat was not only worn by the Hephthalites. Moreover, the shape and structure of the horned hat worn by the Hephthalite women was unique, according to the description in the “Yiyu zhuan 異域傳 B”of Zhoushu 周書（ch. 50）. The Ximodaluo 呬摩呾羅 hat had two horns signifying the father and mother of her husband, whereas the Hephthalite hat had many horns to signify the number of husbands of the wearer. Moreover, the Hephthalites ruled over Tokhāristan for many years, so it also no wonder that they remained there.

（3） The site where Song Yun 宋雲 met the king of the Hephthalites should have been close to Ximodaluo 呬摩呾羅, from which we know that the site was the winter camp of the Hephthalite king.

In my opinion, even if this hypothesis were true, it also has no bearing on the origins of the Hephthalites.

（4） The fact that King Kidara was driven to the west demonstrates that the Hephthalites came from east of Balkh. Of the four centers of activity mentioned above, only Ximodaluo 呬摩呾羅 fulfills this criterion.

In my opinion, such an interpretation is too mechanistic. If the Hephthalites came from north of the Amu Darya, why would King Kidara necessarily not have been able to move west?

（5） The Hephthalites practiced polyandry, which signifies that they should have come from some isolated region in the Hindu Kush to the south of Ximodaluo 呬摩呾羅.

In my opinion, this theory is also inadequate. For details, see below.

In addition, there is another related problem. In the item titled “Anbei duhufu 安北都護府”of Tanghuiyao 唐會要（ch. 73） it is recorded:

On the third day, in the tenth month of the twenty-third year [649] [of the Zhenguan 貞觀 reign-period], the tribes of the Türks submitted to the rule of the Tang 唐 dynasty. [The settlement of] the tribe of Shelituli 舍利吐利 was established as Sheli 舍利 Sub-Province; the tribe of Ashina 阿史那, Ashina 阿史那 Sub-Province; the tribe of Chao 綽, Chao 綽 Sub-Province; the tribe of Helu 賀魯, Helu 賀魯 Sub-Province; and, the two tribes of Geluolu 葛邏祿 and Yida 悒怛 were established as the Geluo 葛邏 Sub-Province. All were subordinate to the Area Command of Yunzhong 雲中.

Moreover, rgyal - rabs for Several Monarchs in the North , an ancient Tibetan document（No. 1283）, held in the Bibliothèque Nationale de France records:

In the past, the king of the Hor issued the edict: How many monarchs are there to the north? Then he commanded five Hor people to go to reconnoitre. This is their report, which is copied from the storehouse for old documents.

A place, called ǰi-’ur in Chinese, Ba-ker Balïq in Dru-gu language. To the north of this site, there are twelve tribes of ’Bug-čhor of the Dru-gu:

The royal tribe of Zha-ma mo-ṅan, the tribe of Ha-li, the tribe of A-sha-ste, the tribe of Shar-du-li, the tribe of Lo-lad, the tribe of Par-sil, the tribe of Rṅi-ke, the tribe of So-ni the tribe of ǰol-to, the tribe of Yan-ti, the tribe of He-bdal, the tribe of Gar-rga-pur.

The small tribes have no monarchs and have 6,000 persons able to bear arms in all.

According to research, this document was composed after the Huihu 回鶻 replaced the Türk tribe seeking hegemony in the area north of the Gobi between the eighth and the ninth century CE . Some consider that the record can be taken as circumstantial evidence that the Hephthalites originated to the north. The advocates of the Iranian theory believe that the remnants mentioned in this document were those left behind when the Hephthalites expanded from west to east and entered Dzungaria. In other words, the record cannot be used to prove that the Hephthalites originated to the north.

In fact, regarding the origins of the Hephthalites there is the clear record in the “Xiyu zhuan 西域傳”of Weishu 魏書（ch. 102）, and there is no need to find evidence from the remnant groups. These remnants could have been left behind when the Hephthalites moved west, as well as after they later advanced east; both scenarios could have occurred, and there is no need to further examine these possibilities here. The problem is that the advocates of the Iranian theory have attempted to prove that the Hephthalites originated in Tokhāristan because of the possibly remnant Hephthalites there, although they know that the presence of Hephthalite remnants in the Jinshan 金山 Mountains and elsewhere are insufficient to prove that the Hephthalites originated to the north, such are the contradictions in their argument!

In sum, the record on the origins and migration of the Hephthalites in the “Xiyu zhuan 西域傳”of Weishu 魏書（ch. 102） is the most reliable, and it cannot be easily denied.

Another important basis put forward by the advocates of the Iranian theory is that there are Iranian elements in the physique, facial features, and culture of the Hephthalites. In my opinion, this assertion ignores the possibility of interaction and blending between different ethnic groups. The following is an analysis of some of the specific questions put forward by the theory’s proponents.

（1） The physical characteristics.

The proponents cite two items in the main. The first is the statement by the Byzantine historian Procopius that the Hephthalites “have white bodies and countenances which are not ugly”. However, those who “have white bodies”did not necessarily belong to an Iranian ethnic group. The “Fujian zaiji 苻堅載記 B”of Jinshu 晉書（ch. 114） records that “the people of the [Fu]Jian [苻]堅 dynasty called the Xianbei 鮮卑 White Savages”. From this, it can be seen that there were people with white bodies among the Xianbei 鮮卑, an ethnic group that spoke an Altaic language.

The second is the above-quoted statement in Datang xiyuji 大唐西域記 that the king of Ximodaluo 呬摩呾羅 was a Shizhong 釋種（Śākya）, i.e., Sakās. The proponents of the Iranian theory assume that Xuanzang 玄奘 was commenting on ethnical differences. In fact, since the proponents accept Procopius’record and acknowledge that the Hephthalites’countenances were not ugly, why ignore Xuanzang’s 玄奘 statement that the appearance of the people of Ximodaluo 呬摩呾羅 was “short and ugly in their features”? Moreover, the concept of the “Sai 塞 ethnic group”was also fairly fuzzy at that time. Menander records that the Persians once described the Türks as Sacae（Fragment 10.2: Exc. de Leg. Rom. 7）. Therefore, it cannot be inferred that the Hephthalites were Iranians simply because Xuanzang 玄奘 called the people in Ximodaluo 呬摩呾羅“Sakā”（Śākya）. In other words, even if it can be proved that the people of Ximodaluo 呬摩呾羅 were Hephthalites, how does this help in establishing the Iranian theory?

The record of the “Xibei zhurong zhuan 西北諸戎傳”of Liangshu 梁書（ch. 54） stated: “For all states that are near Hua 滑, their clothes and appearance are the same as those of Hua 滑”. While the similarity in clothing was the result of mutual influences, the “same”appearance was because the Hephthalites had white skin.

At present, the only data enabling us to know the physical characteristics of the Hephthalites is that provided by the portrait of the envoy from the state of Hua 滑（Hephthalites） in the painting titled Liangzhigongtu 梁職貢圖. The image is of a man with tousled hair that has been cut; he has no facial hair and his nose does not have a high bridge. Even the main proponent of this theory has to admit that the man resembles a Mongol rather than an Iranian and so suggests that the envoy might not have actually been a Hephthalite. He does not realize that the Hephthalites sent envoys from Sogdiana or Tokhāristan at that time and they were surrounded by people of Iranian ethnicity, but howcould they select someone of an ethnic group speaking a Altaic language to serve as an envoy? Moreover, the Hephthalites sent their envoys to present tribute at the Liang 梁 court more than once, when the Liang 梁 people wanted to depict the appearance of envoys, they would have undoubtedly chosen typical examples of the envoys.

（2） Language.

The “Xiyu zhuan 西域傳”of Weishu 魏書（ch. 102） records that the language of the Hephthalites was different from that of the Rouran 柔然 and Gaoche 高車. The proponent rules out that possibility that the language of the Hephthalites belonged to the Altaic family because the Rouran 柔然 spoke a language of the Mongolic branch of that family and the Gaoche 高車 spoke a language of its Turkic branch. This hypothesis, as previously mentioned, cannot be established. Moreover, the “Xiyu zhuan 西域傳”of Weishu 魏書（ch. 102） records at the same time that the language of the Hephthalites was different from that of “the various Hu 胡 people”; combining the above, the “Hu 胡”should refer to Iranian peoples. According to the logic of the proponent, shouldn’t the Iranian peoples known at that time also be ruled out?

Because very few words or their remnants can be confirmed to be Iranian and these are insufficient to identify the original language of the Hephthalites, the proponent has to draw indirect support from what was denoted as “Hephthalite”written in quasi-cursive Greek script and he states that the documents and inscriptions written in Hephthalite cursive are very similar to an Iranian language, with not one identifiable Altaic word to be found among them, and that this indicates that the Hephthalites were an Iranian speaking ethnic group. In my opinion, this argument is also not thought through.

According to the “Xibei zhurong zhuan 西北諸戎傳”of Liangshu 梁書（ch. 54）, the Hua 滑, namely the Hephthalites, “have no script, but use wooden slips as tallies. In communications with neighboring states, they make the barbarians from neighboring states prepare documents in Hu 胡 languages, using parchment instead of paper”. Song Yun 宋雲 also wrote of the Hephthalites: “The natives were illiterate, completely lacking in etiquette or culture”.（ Luoyang qielanji 洛陽伽藍記, ch. 5） From this, we know that the Hephthalites did not yet have their own script as late as the beginning of the sixth century CE . Moreover, as the proponent of the theory argued, the date when the above-mentioned quasi-cursive appeared is far earlier than the date when the Hephthalites appeared in Tokhāristan. This indicates that the cursive script was not created by the Hephthalites. Therefore, even if the so-called Hephthalite documents were left behind by the Hephthalites, there would be nothing to be surprised at that the original language of the Hephthalites cannot be found from them. It is indeed difficult to imagine that the Hu 胡 people of neighboring states would use their own script to write the original language of the Hephthalites. Much less could a nomadic tribe as culturally backward as the Hephthalites totally adopt the local language after entering Tokhāristan. The proponent believes that, from the point of view of linguistics, no evidence to disprove the Iranian theory can be found, but surely no evidence has also ever been found to disprove the Altaic theory?

（3） Religion and customs.

According to the proponent, the description of Hephthalite religious beliefs by Song Yun 宋雲（in Luoyang qielanji 洛陽伽藍記, ch. 5） and the record in the “Xibei zhurong zhuan 西北諸戎傳”of Liangshu 梁書（ch. 54） that the Hephthalites “worship the God of Heaven and the God of Fire”provide a basis for asserting that the Hephthalites were of Iranian ethnicity, because the worship of fire was an important characteristic of the Persians and other Iranians.

In my opinion, this was in fact not the case. There is no necessary connection between ethnicity and religion, because religious beliefs can spread beyond ethnic confines. No trace has been found of the original beliefs of the Hephthalites, but their funeral customs was entirely different from those of orthodox Fire-Worshippers. The latter must expose the body of the deceased on the top of a mountain to consecrate the body as a sky burial. The Hephthalites buried their dead in the ground. This is surely very clear evidence that the Hephthalites were originally not Fire-Worshippers.

Moreover, the “Xibei zhurong zhuan 西北諸戎傳”of Liangshu 梁書（ch. 54） records that the Hephthalites “are without walled towns, living in felt tents with a door open to the east”, and according to proponents of the Iranian theory, this was also the custom of the Iranians.

In my opinion, most of the ethnic groups inhabiting the northern steppe of China had this custom. For example, the “Houfei zhuan 后妃傳 B”of Beishi 北史（ch. 14B） records that “the custom of the Ruru 蠕蠕 is to value the east. ... the curtains of their camps and household seating all face east”. The Wuhuan 烏桓 and others had a similar custom. Since this was not a unique custom of the Iranians, it cannot be used to determine ethnicity.

On the subject of Hephthalite customs, it should also be noted that Procopius mentioned that the Hephthalites practiced live interment with the dead, and the “Xibei zhurong zhuan 西北諸戎傳”of Liangshu 梁書（ch. 54） mentioned the Hephthalite custom of lopping off ears. There are the following lines praising Raghu’s military accomplishments in the Raghuvaṃśa of Kālidāsa:

The redness on the cheeks of the Hūṇa queens testified to Raghu’s achievements in which his prowess was displayed to their husband.

The description obviously refers to the “Hūṇa”, namely Hephthalite's custom of wounding the face（scarification）. The three practices, namely live interment with the dead, ear lopping, and facial scarification were all specific customs of the nomadic tribes of the northern steppe of China.

As for the comment that the Hephthalites “all clip their hair”, this could no doubt refer to an Iranian custom, but it is not impossible that they were influenced by this fashion after they entered Central Asia, and imitated “the Hu 胡 people”. This custom sporadically appeared among the nomadic tribes to the north of China. For example, the “Xiyu zhuan 西域傳”of Weishu 魏書（ch. 102） records that the Yueban 悅般 observed the “custom of clipping their hair and even their eyebrows”.

Finally, I wish to discuss the issue of the polyandry of the Hephtalites.

From the passages cited by the proponents of the Iranian theory, we know that this marital custom was not unique to the Hephthalites. It was widespread in various parts of Central Asia and among different ethnic groups and, as the proponents of the Iranian theory have pointed out, the existence of the practice can hardly serve to trace the origins of a tribe. However, the proponents nevertheless attempt to explore the environment in which this custom was spread to assist them in inferring the spread of particular ethnic groups. Since the custom of polyandry was a phenomenon that survived in communities that were geographically and therefore culturally isolated, they conclude that the Hephthalites were originally a tribe found in such an isolated environment, and specifically identify this as a mountainous location in the Hindu Kush.

In my opinion, this view is inadequate. Firstly, regarding the reason why the Hephthalites practiced polyandry, the “Xibei zhurong zhuan 西北諸戎傳”of Liangshu 梁書（ch. 54） pointed out that this was because there were few women. This might simply have been rational conjecture. The reason for the practice might have been primarily economic, and the geographical environment that reinforced the practice might have been a secondary factor. From the perspective of geography, the fact that this custom was so prevalent would surely also suggest that the environments sustaining this custom were ubiquitous. Why then only single out a mountainous region of the Hindu Kush as the birthplace of the Hephthalites? Secondly, the proponents of the Iranian theory invariably seem to overlook the fact that the Hephthalites were a horse-riding nomadic tribe, so how could they have been in an isolated environment over such a long period?

There are other problems that the proponents cannot explain. The polyandry of the Hephthalites is not mentioned in the “Xiyu zhuan 西域傳”of Weishu 魏書（ch. 102）, and so we have reason to suspect that this was not a custom of the Hephthalites, or was at least not an original custom of the Hephthalites.

10. The Yueban 悅般 theory. This theory can be summarized as follows:

（1） After the Xiongnu 匈奴 had collapsed, the several tribes who were originally subject to the Xiongnu 匈奴 crossed the Altai Mountains and moved west to what is today Kazakhstan and Uzbekistan along the northern side of the Tianshan 天山 Mountains. The Hephthalites who “moved south from the Jinshan 金山 Mountains”and the Yueban 悅般, whose “forebears were a tribe of the Northern Chanyu 單于 of the Xiongnu 匈奴”as recorded in the “Xiyu zhuan 西域傳”of Weishu 魏書（ch. 102）, were actually one and the same ethnic group, although each had its own origin legends and the descriptions of them were different in Chinese histories.

（2） The date when the Hephthalites appeared in Chinese histories for the first time was the second year [456] of the Tai’an 太安 reign-period of Emperor Gaozong 高宗 of the Northern Wei 魏 dynasty, and the date when the Yueban 悅般 appeared in Chinese historical writings for the last time was the second year [448] of the Taiping Zhenjun 太平真君 reign-period of Emperor Taiwu 太武 of the Northern Wei 魏 dynasty. After that the Yueban’s 悅般 whereabouts became a mystery. Therefore, we can suppose that the appearance of the Hephthalites was the result of the Yueban 悅般 moving westward over an eight-year period. The northern boundary of the Hephthalites abutted with the territory of the Chile 敕勒, i.e., the Gaoche 高車, according to Song Yun’s 宋雲 account, which suggests that the Hephthalites had already occupied the northern side of the Tianshan 天山 Mountains. Only when we identify the Hephthalites as the Yueban 悅般 does such a large-scale exploitation of territory become explicable.

（3） The fact that “Yeda 嚈噠”was also read as “Nieda 囐噠”, “Heda 噠”, “Yida 挹怛”, “Yida 揖怛”and so on indicates that the old pronunciation of “yan 厭”was similar to that of “he ”and “nie 囐”, or “yi 挹”and “yi 揖”. The old pronunciation of “da 達”could well have been “dan 旦”. The original pronunciation of “Yueban 悅般”would have been [yet-pat]. Even if the old pronunciation of “ban 般”were not [pat], but [pan], in line with present Japanese readings, the ethnonym could also have been identified as “Yida 挹怛”, an alternative or erroneous transliteration of “Yeda 嚈噠”. For this reason, it would not be far-fetched to regard “Yueban 悅般”as another transliteration or variant of “Yeda 嚈噠”or “Yida 挹怛”.

（4） The sixteen Area Commands in the Western Regions are listed in the “Dili zhi 地理志 3”of Jiutangshu 舊唐書（ch. 40）. One of them was the Area Command of Yueban 悅般, which was established in the town of Yan 豔, the seat of the king’s government of the state of Shihanna 石汗那. There is a similar record in the “Dili zhi 地理志 7B”of Xintangshu 新唐書（ch. 43B）. Shihanna 石汗那, i.e., Čaγāniyān, was always in the main part of Hephthalite territory. The Chinese people in Tang 唐 times linked the site with the name Yueban 悅般, which must have had some basis reason. It is very possible that the Yueban 悅般 were simply the Hephthalites in the Tang’s 唐 mind. The link at least indicates that the pronunciations of the two names were very similar.

（5） The “Gao Hu zhuan 高湖傳”of Weishu 魏書（ch. 32） records that Gao Hu 高湖 was sent on a diplomatic mission to the Hephthalites, and acquired the fine horses of Poluona 破洛那 and Wusun 烏孫 during the Yanchang 延昌 reign-period [512-515]. This event occurred only a few years before Song Yun 宋雲 went to the Western Regions, at the time when the Hephthalites were close to completing their territorial expansion. “Wusun 烏孫”and “Poluona 破洛那”were named only because they were areas that produced fine horses in ancient times, and this did not necessarily mean that the two states continued to maintain their independence. On the other hand, it is very possible that Dong Wan 董琬 went to Wusun 烏孫 and Poluona 破洛那 through the agency of the Yueban 悅般, after Dong Wan 董琬 was sent west and reached Yueban 悅般 during the Taiyan 太延 reign-period. The same areas producing fine horses（Wusun 烏孫 and Poluona 破洛那） were contacted by the Northern Wei 魏 through the Yueban 悅般 during the Taiyan 太延 reign-period and appeared in the records of the Northern Wei 魏 through the Hephthalites during the Yanchang 延昌 reign-period, which would indicate that the Yueban 悅般 had something to do with the Hephthalites, even if the areas occupied by them were different when they were known by the Northern Wei 魏 dynasty.

In my opinion, the theory is unconvincing:

（1） The “Xiyu zhuan 西域傳”of Weishu 魏書（ch. 102） records of the Yueban 悅般: “Their forebears were a tribe of the Northern Chanyu 單于 of the Xiongnu 匈奴”. So the proponents of this theory argue that the Yueban 悅般 were probably subject to the Xiongnu 匈奴. However, it is difficult to see the Hephthalites as also being a tribe of the Xiongnu 匈奴. The Hephthalites had called themselves “Xiongnu 匈奴”, but there were many tribes that called themselves “Xiongnu 匈奴”and they did not necessarily have political or consanguineous relations with the Xiongnu 匈奴. Even if the Hephthalites, as the Yueban 悅般, had once also been subject to the Xiongnu 匈奴, one cannot also identify them as one and the same tribe, because the tribes which were subject to the Xiongnu 匈奴 were many when the Xiongnu 匈奴 were powerful and prosperous; many tribes also moved west after the Xiongnu 匈奴 collapsed.

（2） The “Xiyu zhuan 西域傳”of Weishu 魏書 records（ch. 102） clearly that the Yueban 悅般“pursued by the General of Chariots and Cavalry of the Han 漢 dynasty, Dou Xian 竇憲, the Northern Chanyu 單于 crossed the Jinwei 金微 Mountains and went west into Kangju 康居. Those who were too thin and weak remained north of Qiuci 龜茲”. From this, we know that the former lands of the Yueban 悅般 were to the north of Qiuci 龜茲. Later the Yueban 悅般 moved north and occupied the valleys of the Naryn River and the Ili River because the Wusun 烏孫 moved west. Their eastern border was in the valley of the Yuldus River and adjoined the western border of the Rouran 柔然. However, the “Xiyu zhuan 西域傳”of Weishu 魏書（ch. 102） records clearly that the Hephthalites originated from the north of the Great Wall, and then “moved south from the Jinshan 金山 Mountains”. The difference between the two is very obvious.

（3） The Hephthalites had already moved west to Sogdiana in the 360s or 370s CE , gone south across the Amu Darya and invaded Sasanian Empire in the mid-420s, and then occupied Tukhārestan at the end of the 430s. They then defeated Yazdgird II in the mid-450s, and, following up this victory, crossed the Hindu Kush to invade the Gupta Empire. However, the “Xiyu zhuan 西域傳”of Weishu 魏書（ch. 102） records that the Yueban 悅般 still sent an envoy to pay his respects to the Wei 魏 court, and “they begged leave to attack the Ruru 蠕蠕, cooperating with the government troops from east and west”until the ninth year [448] of the Taiping Zhenjun 太平真君 reign-period. From this, we know that the Yueban 悅般 were still active between the valleys of the Naryn, Ili, and Yuldus Rivers as late as the end of the 440s CE . Therefore, if one wants to identify the Yueban 悅般 as the Hephthalites, the only way of doing so would be to prove that the Yueban 悅般 did not actually move west but retained the regions to the west of Dzungaria at the same time as they advanced west. They only moved their center of activity into Tokhāristan after 448 CE . In fact, one advocate of the Yueban 悅般 theory indeed describes the history of the development of the Hephthalites in this way, but unsuccessfully:

Firstly, the crossing of the Jinwei 金微 Mountains by the Northern Chanyu 單于 is recorded in the “Dou Xian zhuan 竇憲傳”of Houhanshu 後漢書（ch. 23）, and it happened in 91 CE . From this, we know that the Yueban 悅般 were already active north of Qiuci 龜茲 as early as the end of the first century. They did not move south from the Jinshan 金山 Mountains down to the 360s or 370s CE . The proponents of this theory cannot explain how “the Northern Chanyu 單于 crossed the Jinwei Mountains and went west into Kangju 康居”. The event is not recorded in earlier histories and is indeed suspicious. However, the Northern Chanyu’s 單于 crossing of the Jinwei 金微 Mountains is an event recorded in the “Dou Xian zhuan 竇憲傳”of Houhanshu 後漢書（ch. 23）, and we cannot deny that he crossed the Jinwei 金微 Mountains because it sounds suspicious that the Northern Chanyu 單于 moved west to Kangju 康居. Since it is a fact that he crossed the Jinwei 金微 Mountains, it also seems credible that the Northern Chanyu 單于 left behind those who were too thin and weak north of Qiuci 龜茲 at the same time as he moved west or even earlier. Even if it is pure legend that the Yueban 悅般 made contact with the Northern Chanyu 單于, there is every reason to think that this legend was based on the fact the Yueban 悅般 appeared north of Qiuci 龜茲 at a very early date and that this date roughly conformed to when the Northern Chanyu 單于 crossed the Jinwei 金微 Mountains. If this record is entirely doubted, it is even harder to prove that the Yueban 悅般 were a tribe subject to the Xiongnu 匈奴. Therefore, even if, as the advocates of the Yueban 悅般 theory maintain, the first stopping place of the Hephthalites when they moved south from the Altai Mountains was a region of the Tianshan 天山 Mountains, but not in Sogdiana, this does not allow us to identify the Hephthalites as the Yueban 悅般.

Secondly, as I will demonstrate in the following chapter, the date when the Hephthalites moved south is consistent with the date when they appeared in Sogdiana. This shows that the Hephthalites who moved south from the Altai Mountains first reached Sogdiana and did not tarry for any length of time on the way there.

Thirdly, the Wusun 烏孫 were active in the valleys of the Naryn and Ili Rivers to the north of Tianshan 天山 Mountains in the 360s or 370s CE （even if we do not take the Yueban 悅般 into account）. Therefore, there is little possibility that the Hephthalites remained north of the Tianshan 天山 Mountains after they moved south from the Altai Mountains. They would mostly have proceeded west to Sogdiana along Lake Balkhash.

Fourthly, the date when the Yueban 悅般 north of Qiuci 龜茲 occupied the valleys of the Naryn and Ili Rivers should be later than the date when the Wusun 烏孫 were driven out by the Rouran 柔然 and moved west, and the date when the influence of the Rouran 柔然 extended to west of Dzungaria and brought them into contact with the Wusun 烏孫 was during reign of Shelun 社崙, the Khan of the Ruru 蠕蠕, i.e., the beginning of the fifth century at the earliest. Obviously, prior to then, the Yueban 悅般 could not move west, but they could also not advance west. The proponents of the theory try to question the veracity of the statement in the “Xiyu zhuan 西域傳”of Weishu 魏書（ch. 102） that the Wusun 烏孫 were “invaded several times by the Ruru 蠕蠕, so they migrated west into the Congling 葱嶺 Mountains”, thereby moving the date when the Wusun 烏孫 moved westward to the time when the Xianbei 鮮卑 arose in the wake of the Xiongnu 匈奴 in the area north of the Gobi. This is also not thought through.

Fifthly, the news that the Yueban 悅般 had advanced north to the valleys of the Naryn and Ili Rivers from north of Qiuci 龜茲 and that Sogdiana had been conquered by the “Xiongnu 匈奴”had all been learned by Dong Wan 董琬 when he went on his western mission during the Taiyan 太延 reign-period. Dong Wan 董琬 did not arrive in Sogdiana, but it is possible that he arrived among the Yueban 悅般. If the Yueban 悅般 were the “Xiongnu 匈奴”who conquered the state of Sute 粟特（Sogdiana）, then it would have been impossible for Dong Wan 董琬 to be utterly ignorant about this. If the Yueban 悅般 were not the “Xiongnu 匈奴”who conquered the state of Sute 粟特, then we cannot argue that the Hephthalites（i.e., Yueban 悅般） invaded the Sasanian Empire in the 420s CE （which advocates of the Yueban 悅般 theory see as the link with the history of the development of the Hephthalites）, because the state of Sute 粟特 was under the rule of the “Xiongnu 匈奴”at that time.

As for the statement that the northern border of the Hephthalites adjoined the lands of the Gaoche 高車 in Song Yun’s 宋雲 report, this was the situation at the beginning of the 6th century CE , which had nothing at all to do with the Yueban 悅般. The proponents of the theory mentions this only in order to hint that the Yueban 悅般 did not move west, but only advanced west at the same time as they retained the area north of the Tianshan 天山 Mountains and confronted the Rouran 柔然. Obviously, the theory does not hold water.

（4） The Northern Wei 魏 was very familiar with the Yueban 悅般, and their contacts with the Hephthalites were also frequent. It is difficult to imagine that they could have confounded two ethnic groups after only an interval of eight years. Much less, according to the inevitable logic of the advocates of the Yueban 悅般 theory, the Yueban 悅般, namely the Hephthalites, never abandoned the area north of the Tianshan 天山 Mountains. it would be even harder to imagine how the Northern Wei 魏 were so careless.

（5） The “Xiyu zhuan 西域傳”of Weishu 魏書（ch. 102） records that the language of the Hephthalites, “speech is different from that of the Ruru 蠕蠕, the Gaoche 高車, and the various Hu 胡 people”, and also that the Yueban’s 悅般“customs and speech resemble those of Gaoche 高車”. It is very clear that both were not one and the same tribe. However, the advocates of the Yueban 悅般 theory believe that this difference was the result of the incremental Iranization of the language of the Yueban 悅般, namely the Hephthalites, after they extended into Sogdiana and Tokhāristan. However, this also fails to conform to reality because there was only an interval of eight years in the Chinese histories from the time of the disappearance of the Yueban 悅般 to the appearance of the Hephthalites. It is difficult to imagine that the language of a tribe could change so significantly in such a short time! Let alone, according to the logic of the Yueban 悅般 theory, could the Yueban 悅般 have continued to retain the regions north of the Tianshan 天山 Mountains during this time.

（6） The “Gao Hu zhuan 高湖傳”of Weishu 魏書（ch. 32） records:

[Gao] Hui [高]徽, and styles himself Rongxian 榮顯 ... during the Yanchang 延昌 reign-period [512-515 CE ], as Acting Supernumerary Cavalier Attendant-in-ordinary, was sent on a mission to Yeda 嚈噠. All the various states in the Western Regions respected and feared him, and Poluohou 破洛那 and Wusun 烏孫 presented famous horses through him.

Moreover, the “Xiyu zhuan 西域傳”of Weishu 魏書（ch. 102） records:

[Dong] Wan [董]琬, traveled through the nine states, and then went northwards and reached the state of Wusun 烏孫, whose king received a reward from the imperial court, did obeisance signaling his acceptance, and was duly pleased by this. Then he said to [Dong] Wan [董]琬, “I hear that both Poluona 破洛那 and Zheshe 者舌 long for the virtue of Wei 魏, and want to swear fealty and present tribute, but are worried there is no way to reach Wei 魏. Now that you have arrived here, it is proper to go to these two states to encourage them in their sincere admiration”. Thereupon, [Dong] Wan [董]琬 went in person to Poluona 破洛那 and sent [Gao] Ming [高]明 to Zheshe 者舌. The king of Wusun 烏孫 provided them with guides and interpreters for each of the two states. [Dong] Wan [董]琬 and the others read out the imperial edict intended to pacify and reward them. Later, when [Dong] Wan [董]琬 and [Gao] Ming [高]明 returned eastwards, Wusun 烏孫, Poluona 破洛那 and the like sent envoys with [Dong] Wan [董]琬, to come to pay tribute, sixteen states in all.

The two paragraphs are quite clear but contain no reference to relations between the Yueban 悅般 and the Hephthalites. Even if, as the proponents of the theory maintain, Dong Wan 董琬 went to Wusun 烏孫 and Poluona 破洛那 actually on the recommendation of Yueban 悅般, it would indicate that the neighboring states of Wusun 烏孫 and Yueban 悅般 during the Taiyan 太延 reign-period became vassal states of the Hephthalites during the Yanchang 延昌 reign-period, at the latest. In fact, the reason why the proponents contrast the above-mentioned events is to suggest that the Yueban 悅般 did not move west but developed towards the west and eventually occupied the area producing fine horses that had always been occupied by the Wusun 烏孫 and Poluona 破洛那 during the Yanchang 延昌 reign-period. As I explain earlier, this conflicts with the historical facts and is unacceptable, because this contradicts the known historical facts.

（7） In Chinese transliteration, “Yeda 嚈噠”and “Yueban 悅般”do not match phonologically. Specifically, “yue 悅”can acceptably be equated with “ye 嚈”, “nie 囐”, “he ”and “yi 挹”, but the identification of “ban 般”with “da 噠”or “da 怛”cannot be accepted, regardless of whether the pronunciation of “ban 般”was [pat] or [pan]. Of course, in terms of phonetic identification, “Yueban 悅般”[jiuat-peәn] can also not be regarded as the transcription of “Ephthal”, “Abdel”, or “Hebdel”. It is erroneous to regard “Yeda 嚈噠”and “Yueban 悅般”as different Chinese transliterations of the identical name only on the basis of phonetic identifications. Much less can the same name equate with the same item.

（8） The names of Area Commands in the Western Regions that were established in the first year [661] of the Longshuo 龍朔 reign-period of Emperor Gaozong 高宗 of the Tang 唐 dynasty were for the most part freely and unhesitatingly adopted from toponyms and ethnonyms from before the Southern and Northern dynasties period and were not squarely based on local realities of that time. An obvious example is that, according to “Dili zhi 地理志 B”of Xintangshu 新唐書（ch. 43B）, “the Area Command of Gumo 姑墨 Sub-Province”was established in the state of Damo 怛沒. From this, knowing that “the Area Command of Yueban 悅般”was established in the former territory of the Hephthalites is irrelevant.

In brief, the existing historical data show that the Yueban 悅般 cannot be identified as the Hephthalites. The information on the Yueban 悅般 was interrupted after 448 CE and they did indeed move west. The Yueban 悅般 who moved west were simply the Avars in the Western historical record. On this, I have written a monograph, and do not propose to discuss this issue here.

As mentioned above, the various theories on ethnic origin and ethnicity of the Hephthalites proferred to date are all untenable. The main reason for this is that those proposing the theories completely ignore or do not pay enough attention to the following factors: Each nomadic tribe admittedly had a primitive blood clan as its core, but when it rose up it would often annex or enslave many tribes around it, forming a political and military complex. When the tribe embarked on the path of migration, expansion, and conquest, this feature became particularly prominent. Therefore when, as historians today, we discuss the ethnic origins and ethnicity of a particular nomadic tribe in history, in addition to conduct a comprehensive rather than a one-sided examination of such aspects as nomenclature, language, customs, institutions, religion and so on of the tribe, we must also consider the course of formation and migration of the tribe, paying special attention to the interplay between it and other tribes, as well as the possibility of consanguineous and cultural integration or assimilation. Otherwise, the more precise our conclusions, the further we might stray from the truth.

In my opinion, it is possible that the Hephthalites and the Yifu-Xianbei 乙弗-鮮卑 came from one and the same source. Here I will first examine the Yifu 乙弗, then examine whether both sprang from one and the same source.

1. The Yifu 乙弗 was a tribe of the Tuoba-Xianbei 拓拔鮮卑.

In the “Guanshi zhi 官氏志”of Weishu 魏書（ch. 113）, it is recorded: “The Yifu 乙弗 clan later changed its clan name to Yi 乙”. The Yi 乙 clan was one of the “various surnames that submitted to be included into the lineage's inner”. From this, we know that the Yifu 乙弗 would have been a tribe of the Tuoba-Xianbei 拓拔鮮卑.

In the “Taizu ji 太祖紀”of Weishu 魏書（ch. 2） it is recorded: “[In the fifth month of the first year [386] of the Dengguo 登國 reign-period], Houchen 侯辰, the chief of the tribe of Hufuhou 護佛侯, and Daiti 代題, the chief of the tribe of Yifu 乙弗, revolted and left”. We thus know that the tribe of Yifu 乙弗 would have submitted to the Tuoba 拓跋 tribe before the Dengguo 登國 reign-period.

In the “Yifu Lang zhuan 乙弗朗傳”of Beishi 北史（ch. 49） it is recorded: “[Yifu Lang’s 乙弗朗] ancestors were from the Eastern Xianbei 鮮卑. He was an aristocrat of the tribe, who moved to Dai 代 together with the Wei 魏 [i.e., the Tuoba-Xianbei 拓拔鮮卑]. He later made his home in Shangle 上樂”. From this, we know that the Yifu 乙弗 originated in Liaodong 遼東 and later moved to north of the Great Wall.

2. The Yifu 乙弗 and the Yifudi 乙弗敵 had shared origins.

In the “Tufa Wugu zaiji 禿髮烏孤載記”of Jinshu 晉書（ch. 126）, it is recorded: “Wugu 烏孤 attacked the two tribes, Yifu 乙弗 and Zhejue 折掘, and was victorious. He sent his general Shi Yigan 石亦干 to build the castle of Lianchuan 廉川 and made it his headquarters”. According to the “Jin ji 晉紀 30”of Zizhi tongjian 資治通鑒（ch. 108）, this event happened in the 7th month of the twentieth year [395] of the Taiyuan 太元 reign-period of Emperor Xiaowu 孝武 of the Jin 晉 dynasty. Hu Sanxing’s 胡三省 commentaries state: “The two tribes Yifu 乙弗 and Zhejue 折掘 were located west of the tribe of Tufa 禿髮. Lianchuan 廉川 was located in Huangzhong 湟中”. In addition, in the “Tufa Nutan zaiji 禿髮傉檀載記”of Jinshu 晉書（ch. 126）, it is recorded: “Nutan 傉檀 advocated launching a military campaign against the Yifu 乙弗 in the west”. From this, we know that there was the tribe of Yifu 乙弗.

In the “Tuyuhun zhuan 吐谷渾傳”of Weishu 魏書（ch. 101） it is recorded: “To the north of the Tuyuhun 吐谷渾 there is the state of Yifu（wu）di 乙弗（勿）敵. Its customs are similar to those of the Tuyuhun 吐谷渾. They have no knowledge of the five grains and only eat fish and perilla plants”. In the item titled “Xirong 西戎 2”of the “Bianfang 邊防 6”section of Tongdian 通典（ch. 190） it is also recorded: “The [state of] Yifudi 乙弗敵 was known in Later Wei 魏 times. It was to the north of the Tuyuhun 吐谷渾. The state had a lake, Quhai 屈海, with a perimeter of more than a thousand li 里. It had approximately ten thousand households. Its customs were similar to the Tuyuhun 吐谷渾. They had no knowledge of the five grains and only ate fish and perilla plants”. Note: Quhai 屈海 refers to Lake Qinghai 青海. North of Tuyuhun 吐谷渾 was to the west of the Southern Liang 涼 dynasty. “Yifu 乙弗”is the abbreviation of “Yifudi 乙弗敵”.

In the “Xiwei Wendi Wenhuanghou Yifushi zhuan 西魏文帝文皇后乙弗氏傳”of Beishi 北史（ch. 13） it is also recorded: “The ancestor [of Empress Wen 文] was the commander of the Tuyuhun 吐谷渾. He resided at Lake Qinghai 青海 and was known as King Qinghai 青海”. Note: Both Houweishu 後魏書, quoted by Taiping yulan 太平御覽（ch. 104）, and the “Xiwei Wendi ji 西魏文帝紀”of Beishi 北史（ch. 5） describe Empress Yifu 乙弗 as belonging to “Empress Yi 乙”. From this, we can know that the Yifu 乙弗 north of the Great Wall and the Yifudi 乙弗敵 at Lake Qinghai 青海 originated from the same source.

3. Yifu 乙弗 and Qifu 乞伏 were “detached”from one and the same tribe.

In the “Qifu Guoren zaiji 乞伏國仁載記”of Jinshu 晉書（ch. 125） it is recorded: “Qifu Guoren 乞伏國仁 [? -388 CE ] was of Xianbei 鮮卑 descent from Longxi 隴西. In the past, the three tribes of Rufusi 如弗斯, Chulian 出連 and Chilu 叱盧 had migrated south to the Great Yinshan 陰山 Mountain from north of the Gobi. They encountered an enormous creature during their journey. ... The creature suddenly disappeared and they saw a small child there. At that time there was also the tribe of Qifu 乞伏, in which a childless old man requested to be allowed to adopt this baby. Everyone approved of this act. ... By the time he was only ten years old, he had grown to be a fierce warrior skilled in both horsemanship and archery. His bow had a pull of five hundred jin 斤. All four tribes admired his strength and nominated him as their leader. He was known as Khan Qifu 乞伏, Tuoduo Mohe 託鐸莫何. ... His descendants included Youlin 祐鄰, who was the fifth generation ancestor of Guoren 國仁”. In Xiqinlu 西秦錄 quoted in the item titled “Zhi 之”of “Shangping sheng 上平聲 7”section of Yuanhe xingzuan 元和姓纂（ch. 2） it is recorded that “the Qifu 乞伏 clan and the Qiyin 期引 clan both migrated to Yinshan 陰山 Mountain from north of the Gobi”. There is a parallel passage from Xiqinlu 西秦錄, quoted in the item titled “Zhi 支”of “Shangping sheng 上平聲 5”section of Gujin xingshishu bianzheng 古今姓氏書辯證（ch. 3）, except that in the latter work “Qiyin 期引”is written as “Siyin 斯引”. However, the reference in Xiqinlu 西秦錄, quoted in Tongzhi 通志（ch. 29） reads that “the tribe of Rufu 如弗, from which the ancestor of Qifu Guoren 乞伏國仁 came, together with the three tribes of Chulian 出連, Siyin 斯引, and Chiling 叱靈, migrated south to Yinshan 陰山 Mountain from north of the Gobi”. This informs us that the text in the “Qifu Guoren zaiji 乞伏國仁載記”of Jinshu 晉書（ch. 125） is missing the character “yin 引”following “si 斯”. But the character reads “qi 期”or “si 斯”in Xiqinlu 西秦錄 quoted in various books, and we do know which one is correct. Since “Qiyin 期引”or “Siyin 斯引”do not appear in the “Guanshi zhi 官氏志”of Weishu 魏書（ch. 113）, it is possible that the quotations are all incorrect. “Siyin 斯引”or “Qiyin 期引”might actually have been a textual error for “Huyin 胡引”. There is a clan of Hugukouyin 胡古口引”in the “Guanshi zhi 官氏志”of Weishu 魏書（ch. 113）. It is suggested that “Hugukouyin 胡古口引”should be “Huyin 胡引”, which would seem to be correct. This is because “hu 胡”was first misread as “qi 期”, then as “si 斯”. Therefore, the original text of Xiqinlu 西秦錄 should have been recorded as follows:

The tribe of Rufu 如弗, from which the ancestor of Qifu Guoren 乞伏國仁 came, together with the three tribes of Chulian 出連, Huyin 胡引, and Chiling 叱靈, migrated south to Yinshan 陰山 Mountain from north of the Gobi.

In addition, the original text of the “Qifu Guoren zaiji 乞伏國仁載記”of Jinshu 晉書（ch. 125） should be recorded as follows:

In the past, the tribes of Rufu 如弗, [together with] the three tribes of Huyin 胡引, Chulian 出連, and Chilu 叱盧, had migrated south to the Great Yinshan 陰山 Mountain from north of the Gobi.

From this, it can be seen that the Qifu 乞伏 were formed from the four tribes of Rufu 如弗, Huyin 胡引, Chulian 出連, and Chilu 叱盧（or Chiling 叱靈）.

The above-quoted Xiqinlu 西秦録 says that the ancestor of Qifu Guoren 乞伏國仁 came from the Rufu 如弗. The “Qifu Guoren zhuan 乞伏國仁傳”of Weishu 魏書（ch. 99） also records: “Qifu Guoren 乞伏國仁, who had descended from the Xianbei 鮮卑, came from Longxi 隴西. His ancestor came from the tribe of Rufu 如弗, and travelled south from north of the Gobi”. From this, we can know the Qifu 乞伏 stemmed originally from the Rufu 如弗. The reference “there was also the tribe of Qifu 乞伏”in the “Qifu Guoren zaiji 乞伏國仁載記”of Jinshu 晉書（ch. 125） is a supplementary reference. On the basis of the same chapter, those who later held high positions included Chilu Nahu 叱盧那胡, Chulian Gaohu 出連高胡,（Si）[Hu]yin Wuni（斯）[胡]引烏埿, and among the tribe of Qifu 乞伏, but no person of the Rufu 如弗, which also indicates that the Qifu 乞伏 and the Rufu 如弗 were in fact two faces of the one coin. Qifu 乞伏, which was also read Qifo 乞佛, and Rufu 如弗 were different transcriptions of one and the same name. However, “Qifu 乞伏”actually became the name of the tribal confederacy composed by the four tribes of Rufu 如弗 and so on, and was thus different from “Rufu 如弗”.

In my opinion, the Yifu 乙弗 should be simply the Qifu 乞伏. The “Guanshi zhi 官氏志”of Weishu 魏書（ch. 113） lists Qifu 乞伏 and Yifu 乙弗 respectively, because they were “detached”from the same tribe.

（1） According to the “Qifu Guoren zaiji 乞伏國仁載記”of Jinshu 晉書（ch. 125）, Youlin 祐鄰, the chief of the Qifu 乞伏 tribe, “[Youlin 祐鄰] led five thousand families in a migration to Xiayuan 夏緣 in the early years of Taishi 泰始 [c. 265], and the population of the tribe gradually grew. The Xianbei 鮮卑 tribe, the Lujie 鹿結, with a population of more than 70,000 households, settled in the valley of the Gaoping 高平 River and frequently fought with Youlin 祐鄰. Lujie 鹿結 was defeated and moved south to Lüeyang 略陽. Youlin 祐鄰 annexed all his remaining people and moved his tribe to the valley of the Gaoping 高平 River. After the death of Youlin 祐鄰, his son Jiequan 結權 assumed his position and moved to Qiantun 牽屯 [Mountain]. After the death of Jiequan 結權, his son Lina 利那 assumed his position. He launched an attack on Tulai 吐賴, a Xianbei 鮮卑 tribe, at Wushu 烏樹 Mountain, as well as the Yuchi Kequan 尉遲渴權 in the valley of the Dafei 大非 River. He took in more than 30,000 households of the defeated tribe members. After the death of Lina 利那, his brother Qini 祁埿 took over his position. After the death of Qini 祁埿, Lina’s 利那 son, Shuyan 述延 assumed his position. He launched an attack on Mohou 莫侯, a Xianbei 鮮卑 tribe in Yuanchuan 苑川, and utterly routed him, forced his population of more than 20,000 households to surrender, and thereupon lived in Yuanchuan 苑川. ... With the death of Shuyan 述延, his son Nudahan 傉大寒 assumed his position. When Shile 石勒 destroyed Liu Yao 劉曜, Nudahan 傉大寒 moved to Jigu 旡孤 Mountain in Maitian 麥田 out of fear. With the death of Nudahan 傉大寒, his son Sifan 司繁 assumed his position and then moved to Dujian 度堅 Mountain. Not long afterwards, he was attacked by Fu Jian’s 苻堅 general, Wang Tong 王統. His people betrayed him and surrendered to Wang Tong 王統. ... went to Wang Tong 王統 in order to surrender to Fu Jian 苻堅”. From this, we know that the tribal alliance of Qifu 乞伏 conquered many other tribes in the course of its migrations. Its influence and power slowly increased, as it also became increasingly multi-ethnic. Such a huge tribal alliance did not have a strong foundation. If it were to confront a powerful enemy, it could simply crumble without a fight. As Sifan 司繁 stated earlier, “the foundations have been destroyed without even a battle”. Thus, part of the tribe most probably surrendered to Fu Jian 苻堅, while the rest dispersed; they possibly joined the Tuoba 拓拔 or sought asylum among the Gaoche 高車. They dispersed and went their own way. The date when Sifan 司繁 surrendered to Qin 秦, according to the “Jin ji 晉紀 25”of Zizhi tongjian 資治通鑒（ch. 103）, was the first year [371] of the Xian’an 咸安 reign-period of Emperor Jianwen 簡文 of the Jin 晉 dynasty（the seventh year of the Jianyuan 建元 reign-period of Fu Jian 苻堅 of the Former Qin 秦 dynasty）. It is very possible that the Yifu 乙弗 separated from the Qifu 乞伏 at this time. Thereafter the Yifu 乙弗 appeared in historical texts.

（2） Above we quoted a passage from the “Taizu ji 太祖紀”of Weishu 魏書（ch. 2） referring to the revolt of theYifu 乙弗 and the chief of the tribe of Hufuhou 護伏侯, Houchen 侯辰. “Houchen 侯辰”was an alternative rendering of “Huyin 胡引”, and the “Huyin 胡引 clan”used the name of the tribal chief as the name of a clan. From the relationship between Yifu 乙弗, Huyin 胡引 and the others, it can be affirmed that Yifu 乙弗 came from Qifu 乞伏.

（3） The pronunciation of “Yifu 乙弗”, together with those of “Qifu 乞伏”or “Qifo 乞佛”, were different transliterations of one and the same name. The “Qifu Bao zhuan 乞伏保傳”of Weishu 魏書（ch. 86） records that “Qifu Bao 乞伏保 came from the Gaoche 高車 tribe”. Moreover, the “Gaoche zhuan 高車傳”of Weishu 魏書（ch. 103） records that “the tribe of Gaoche 高車 had twelve surnames. The first was Qifuli 泣伏利”. “Qifuli 泣伏利”[kiәp-biuәk-liet] should have been an alternate transliteration of “Qifu 乞伏”. In my opinion, “Yifu 乙弗”was an abbreviated rendering of “Yifudi 乙弗敵”. The pronunciation of “Yifudi 乙弗敵”was the same as that of “Qifuli 泣伏利”, which is also evidence that “Yifu 乙弗”was an alternate transliteration of “Qifu 乞伏”.

4. The Yifu 乙弗 were a tribe of miscellaneous barbarians.

As previously mentioned, the Qifu 乞伏 were composed of the four tribes of Rufu 如弗, Huyin 胡引, Chulian 出連, and Chilu 叱盧. Of them, the Rufu 如弗 was the Xianbei 鮮卑, so their descendants were known as the Qifu-Xianbei 乞伏鮮卑. Huyin 胡引, it is suggested, “belonged to the Jie 羯 ethnic group”. The Jie-Hu 羯胡 were mostly barbarians from the Western Regions. The Chulian 出連, it is suggested, were “originally the Tuge 屠各 tribe”, which indicates that they were originally the Xiongnu 匈奴. The Chilu 叱盧 are believed by some scholars to be of Gaoche 高車 descent. From this, we know that the tribe of Qifu 乞伏 included the Xianbei 鮮卑, Jie 羯, Xiongnu 匈奴, and Gaoche 高車, and thus was a tribe throughout of “miscellaneous barbarians”. Each tribe maintained its original tribal name, but the blend of ethnic groups is self-evident. The Yifu 乙弗, who stemmed from the Qifu 乞伏, were no exception.

5. The migration process of the Yifu 乙弗.

Of the Yifu 乙弗 who separated from the Qifu 乞伏, a part remained behind north of the Great Wall and depended on the Tuoba 拓拔. The majority moved west, with one branch reaching the shores of Lake Qinghai 青海 and forming the state of Yifudi 乙弗敵.

It should be pointed out that the Yifu 乙弗 clan was one of the surnames that submitted to the inner Tuoba-Wei 拓跋魏. According to the “Guanshi zhi 官氏志”of Weishu 魏書（ch. 113）, the so-called “various surnames that submitted to be included into the lineage's inner”refers to the “various surnames from other tribes that submitted to the inner during the reign of Emperor Shenyuan 神元 [Liwei 力微, r. 220-277]”. However, “various surnames that submitted to be included into the lineage's inner”might not have all been inducted during the reign of Emperor Shenyuan 神元. An obvious example would be the Qutu 屈突 clan. The Wei 魏 dynasty’s first record of the Qutu 屈突 clan was in the tenth year [395] of the Dengguo 登國 reign-period. In my opinion, the date when the Yifu 乙弗 submitted to the Tuoba 拓跋 should be between 371 and 386 CE .

If the origins and ethnicity of the Yifu 乙弗 are indeed as outlined earlier, it would not be difficult to determine whether the Yifu 乙弗 and the Yeda 嚈噠 had shared origins. The reasons why both could have shared origins are as follows:

1. Ethnonyms.

“Yeda 嚈噠”, that appears in Western historical texts as “Ephthal”and “Abdel”, which could at a glance also be transliterated as “Yifudi 乙弗敵”or be abbreviated as “Qifu 乞伏”, “Qifo 乞佛”, or “Yifu 乙弗”. Interestingly, there is a variant transcription of “Rufu 如弗”. The old pronunciation of “ru 如”was [njia]. Coincidentally, “Ephthal”was also transliterated as “Nephthal”in the Western data.

“Yeda 嚈噠”was derived from the surname or name of their king, while “Yifu 乙弗”was the tribal name. The tribal names of nomadic tribes were often derived from the surname or name of their chief. However, the Yifu 乙弗, who migrated west to Central Asia, first called themselves the “Xiongnu 匈奴”, and neighboring tribes also regarded them as Xiongnu 匈奴. After conquering Tokhāristan, their influence and power increased greatly and their original name became better known. The Byzantines thus mistakenly thought that this name derived from the king of the Yeda 嚈噠 in the latter half of the fifth century, the very king who triumphed in the war with Pērōz. There must been numerous difficulties and dangers in the course of the Yifu’s 乙弗 migration to Central Asia from north of the Great Wall, but the tribe always maintained the same ethnonym, meaning that the ethnonym would indeed not be lost even in troubled times.

2. Former lands.

The “Xiyu zhuan 西域傳”of Weishu 魏書（ch. 102） states that the Yeda 嚈噠 originated beyond the Great Wall. This is the only existing reliable record of the origins of the Yeda 嚈噠. Although the origins of the Yifu 乙弗 can be traced back to Liaodong 遼東, it was only when they had migrated to the north of the Great Wall and became a tribe of “miscellaneous barbarians”that they became the Yifu 乙弗 we know of today. Thus, we can conclude that the Yifu 乙弗 also originated from north of the Great Wall.

3. Migration.

The date when the Yeda 嚈噠 tribe crossed the Altai Mountains and migrated west to Sogdiana should be between 366 CE and 376 CE . This dating, as speculated above, conforms to the dating deduced above [371] for the Yifu’s 乙弗 break from the Qifu 乞伏 and their migration west. At that time, the Yifu 乙弗 people who migrated west can be divided into two branches. One branch（perhaps the weaker branch） reached the shores of Lake Qinghai 青海 and became the later state of Yifudi 乙弗敵. The other branch crossed the Altai Mountains and reached the northern banks of the Gui 嬀 River to later become the Yeda 嚈噠.

4. Ethnic origins and ethnicity.

The “Yeda 嚈噠”were known as “Hūṇa”or “Hun”in Western historical texts. The “Xiyu zhuan 西域傳”of Weishu 魏書（ch. 102） states that the Yeda 嚈噠 tribe was the “detached stock of the Gaoche 高車”. These statements seem difficult to reconcile at first glance. However, if seen from the perspective of the shared origins of the Yeda 嚈噠 and the Yifu 乙弗, the issues become clear. The Qifu 乞伏, the Yifu’s 乙弗 predecessor, who probably shared a mixed ancestry of the Xiongnu 匈奴, were also subject to the Xiongnu 匈奴. Therefore, they made use of the name of the Huns and were called “Huns”by neighboring tribes when they moved west, but they also had something to do with the Gaoche 高車 and, thus, in the eyes of the Northern Wei 魏 people, even though the Yeda 嚈噠 people spoke a different language from the Gaoche 高車, they were still regarded as “detached stock”of the Gaoche 高車.

As for the reason why the Yeda 嚈噠 were known as the “White Huns”, there are two possibilities. The first is that they were a tribe of “miscellaneous barbarians”, but their core, the Rufu 如弗, was the Xianbei 鮮卑, and some Xianbei 鮮卑 people were originally known as “White Savages”; the second is that this would have been the result of intermarriage between the Rufu 如弗 and the Jie-Hu 羯胡. It should be pointed out that regardless to what extent the skin color of the Yifu 乙弗, also known as the Yeda 嚈噠 was being white as a result of intermarriage, their physique and facial features would have been typically “East-Eurasianoid”, at least for a certain period of time. The descriptive portraits of the emissaries from the state of Hua 滑 in Liangzhigongtu 梁職貢圖 are clear evidence of this.

5. Language.

The “Xiyu zhuan 西域傳”of Weishu 魏書（ch. 102） records that the language of the Yeda 嚈噠 was different from that of the Gaoche 高車, Rouran 柔然, and other Hu 胡 tribes. Based on this record, as previously mentioned, one cannot determine the attribution of the Hephthalite language. Thus, we can see that the language spoken by the Yeda 嚈噠 people was very special, which is possibly the result of ethnic groups blending over a long period of time. However, the “Xibei zhurong zhuan 西北諸戎傳”of Liangshu 梁書（ch. 54） states: “Their language is intelligible only through oral interpretation conducted by the people of He’nan 河南”. He’nan 河南 here refers to the Tuyuhun 吐谷渾. The reason why the Tuyuhun 吐谷渾 could translate the language of the Yeda 嚈噠 people was not necessarily because the Yeda 嚈噠 language was the same as that of the Tuyuhun 吐谷渾 language, but clearly because the Tuyuhun 吐谷渾 were in frequent contact with their northern neighbors, the Yifudi 乙弗敵. In my opinion, linguistically, this strengthens the possibility that the Yeda 嚈噠 were of the same origin as the Yifu 乙弗.

6. System, customs and religion.

As previously mentioned, the Yeda 嚈噠 tribe had titles of high rank, such as Khan, Khatūn, Yehu 葉護（Yabgu）, and Teqin 特勤（Tigin）. Their unusual customs included cutting the face to display sorrow, piercing the ears, and sacrificial suicide, being all common customs among the Altaic tribes and indirectly confirming that the Yeda 嚈噠 and the Yifu 乙弗 had shared origins. This is because the Yifu 乙弗 originated from north of the Great Wall, and were formed mainly by the Xianbei 鮮卑, Xiongnu 匈奴, and Gaoche 高車, who all belonged to the Altaic family. Even the God of Heaven believed in by the Yeda 嚈噠 might also be traced to the Yifu 乙弗. The Yifu’s 乙弗 predecessors, the Qifu 乞伏, had some Jie-Hu’s 羯胡 blood, and they too subscribed to this belief. ^[3] It is thus possible that the Yifu 乙弗 were influenced by this particular tribe when they lived in the area north of the Great Wall. Some of the Iranian elements displayed in the customs of the Yeda 嚈噠 tribe were probably received influences from the Iranian tribes after the Yeda 嚈噠 migrated into Central Asia. These elements should not be seen as evidence disproving that the Yeda 嚈噠 and the Yifu 乙弗 had shared origins.

The above six points serve to support the hypothesis that the Yeda 嚈噠 and the Yifu-Xianbei 乙弗鮮卑 had shared origins. The course of the formation and migration of the Yifu 乙弗 discussed in this chapter can be read as the prehistory of the Hephthalites.

NOTES

[1] Enoki first believed that “tun ”of “Huatun 滑 ”was a suffix in Iranian languages that meant “place”. Later he believed that the original text of Tongdian 通典 should have been “... was known as Yeda 嚈噠 in Later Wei 魏 times”because the name “Huatun 滑 ”does not appear in Weishu 魏書. In my opinion, his former theory is incorrect. Moreover, since the original appearance of the “Xiyu zhuan 西域傳”of Weishu 魏書（ch. 102） has already been lost, the latter theory also has no basis. In addition, following the entry on “Yeda 嚈噠”there is “Yida-tong 挹怛同”in the item titled “Xirong 西戎 5”of the“Bianfang 邊防 9”section of Tongdian 通典（ch. 193）. Enoki identifies “tong 同”with -don or -dōna, the suffix in Ormuri, Parachi, Yagnobi, and Šughni, meaning “place”, to prove that the Hephthalites belonged to the Iranian language family. In my opinion, his theory is also incorrect. The character “tong 同”was an error resulting from the character “guo 國”being damaged. The statement “during the time of the Sui 隋 dynasty”and so on should follow the entry titled “Yeda 嚈噠”, The text in Liangdian 梁典 quoted at the end can be taken as evidence. In other words, it was the state of Yeda 嚈噠 in Later Wei 魏 times, but not the “Yida-tong 挹怛同”, that was known as “the state of Yida 挹怛”. Therefore, this item in Tongdian 通典 should read as follows: “The state of Yida 挹怛: [The state of Yeda 嚈噠] was also known as the state of Yida 挹怛 during the time of the Sui 隋 dynasty. ... The surname [of king of the] state of Hua 滑 is Yeda 嚈噠. His descendants named the state after his surname. As the name was mispronounced, the state also became known as ‘Yida 挹怛’”. For Enoki’s theory, see Enoki1952 and Enoki1959（e）.

[2] Huichao wangwutianzhuguozhuan （ canjuan ） jianshi , pp. 36-38; Fujita1943（d）, esp. pp. 38-39 and Fujita1943（g）, esp. pp. 94-96.

[3] In the “Shi Jilong zaiji 石季龍載記 B”of Jinshu 晉書（ch. 107） it is recorded: “General Imposing Manner Sun Fudu 孫伏都 and Liu Zhu 劉銖, and the others, mustered 3,000 of the Jie 羯-barbarians, and ambushed in the temple of the God of Barbarian Heaven to want to kill [Ran] Min [冉]閔 and the others”. There is an identical reference in Houzhaolu 後趙錄 cited in Taiping yulan 太平御覽（ch. 120）. QO1EMFhlAMO1+nYrD/S5jfw+jbIT5sDqAOomSIaB8soWiP9St8FfRSCr5Eh8BGPh