Notes on an identification

Study of the ethnic processes associated with the history of nomadic societies in the Eurasian steppes always encounters a number of difficulties, especially when it comes to attempts to identify peoples living in different chronological periods in the same territory, and/or having common ethnic names (Németh, 1991; Akın, 1982).

L.N. Gumilev proposed the hypothesis of identifying the Chй-mй-kйn Ж^й tribe¹, which inhabited Semirechye in the 7th–8th centuries, and which is well-known from Chinese sources, with the Kimäk tribal group, known in the later period from Muslim sources

(Gumilev, 1993: 380–381, nt. 38). Orientalists were skeptical of this hypothesis, and found this identification lacking any proof (Kumekov, 1972: 32). We do not intend to study the history of the Kimäk , since this is a separate and large issue (Kumekov, 1972; Golden, 1992: 202–205; 2002), but consider it necessary to turn to reviewing some data concerning the early history of the Kimäk . Such data may confirm the presence of a certain sound insight in the hypothesis by Gumilev, which can be rejected owing to the lack of philological arguments alone.

Chй-mй-kйn and Yemak :

from personal names to ethnic names

All direct information about the Kimäk that scholars currently have has survived solely in Muslim sources.

Thus, the Persian author of the 11th century Gardizi cited the following legend on the origin of the people bearing this name: “kimakiyan)> their origin (asl) was this, that the leader (mehtar) of the Tatars (Tataran) died leaving [82C] two sons. The elder son seized the kingship (padsahi) and the younger son became envious of his brother. The name of that younger brother was *Šad. He tried to kill the older brother but was not able, [after which] he be-came afraid for himself.

[Now], this Šad had a girl (lit. concubine, or maid, maiden, kanizak ), [who] was his lover (or mistress, ‘asqqe ). He took away this girl and fled [257D] from before his brother. He went to a place where there was a great river (or lake ab-e bozorg ), many trees, and abundant game. There he pitched his tent ( xarqah ) [Cl] and settled down forud amad). Every day that man and girl, both of them, would go hunting and they would eat the flesh of the game [they killed] and they would make garments of [258A] skins of sable, grey squirrel, and ermine ( samur , senjab o qaqom ).

[And so it went] until seven persons from among the clients (* mawaliyan or the adopted, inferior [tribesmen], mowaledan in the sense of mowalladan ) of the Tatars [82D] came to them ( nazdik-e Isdn sodand ). The first was Imi; the second, Imak; the third, Tatar; the fourth *Bayandur ( BHandr ); the fifth, Qifcaq; the sixth, Laniqaz; the seventh, Ajlad. And these were a party ( qomi ") who had taken (lit. brought) their masters’ ( xodavandan ) horses ( soturan ) to graze, but where the horses were there was no pasturage left and so they had gone (lit. went) in search of grass to that region in which Šad was. When the maid saw them she came out and said ‘ertiš’ , which means ‘dismount yourselves’ for which reason this river has been named the Ertiš (Irtysh).

[Now] when this party recognized that girl, they all dismounted and put up [their] tents. [Then] when Šad returned feraz rasid ), he brought [258B] much game and [82E] entertained them, [so that] they stayed there until winter. When the snow came ( beyamad) they were unable to go back, [but] [83A] there was abundant grass in that place [and so] they were there all winter.

[At length] when the world became fair [again] and the snow went away, they sent a person to the abode (bongah) of the Tatars, that he might bring them news of that party. But when he arrived, he *saw [that] the entire place had become desolate and devoid of people, for the enemy had come and plundered and killed the whole nation (qom), [except for] that remnant which had been left (and came forward) towards him from the foot of the mountain. [These] he told of Sad (*hal-e Sad, ut Barthold, pro, xali šod) and his own comrades, and all that folk set out for the Ertiš. When they arrived there they greeted Šad at as their chief ( riyasat salam kardand) and held him in awe (u-ra bozorg dastand). Then other folk (qom) who heard this news [83B] began to come, [until at length] seven-hundred persons came together [258C] and stayed a long [C2] time in Šad’s service. Afterwards, when they became [more] numerous they spread out over those mountains and became seven tribes, named after those seven persons we have mentioned. <…> Now, all these Kimekis are bad tempered, ungenerous and inhospitable (garib-dosman). One day this Sad was standing on the edge of the Ertis with his attendants (qom-e xvis) [when] a cry came [saying] ‘O Sad, *give me [your] hand (Hab.: *ma-ra dast de; Bart.: *ma-ra didi, pro, morad sodi) in the water’. [But] he saw nothing except some hair that was floating (lit. going) on top of the water. He tethered his horse, went into the water and took hold of that hair. It was his wife, the Xatun. He asked her ‘How did you fall [in]?’ The woman said, [83C] ‘a water-dragon (nehang) seized me from the river’s edge’. [So now] the Kimek people revere that river, worship it [258D] and prostrate themselves to it and they say thus that the river is the god of the Kimek. To Sad they gave the name Tutuq which means that he heard the cry, entered the water and was not afraid.” (cited after (Martinez, 1982: 120–121 (English translation), 179–181 (Persian text), cf.: (Marquart, 1914: 89–91; Bartold, 1973: 27–28 (Persian text), 43–44 (Russian translation)).

The last sentence, certainly, speaks about the “folk etymology” (Bartold, 1973: 44, nt. 14; Czeglédy, 1973: 259; Zuev, 2004: No. 2: 18); nevertheless this is a source reflecting such events as migration of a group of tribes of various origins to the Irtysh from somewhere else (this fact, albeit in a somewhat different aspect, was specially noted by S.M. Akhinzhanov (1995: 102, 103, 107, 115, 120)) and the formation of the Kimäk tribal union in that exact place. Without going into the discussion about the time and historical context of this migration (see (Golden, 2002)), we want to draw attention only to one point: whenever and wherever the representatives of the various tribal groups came to the Irtysh valley, the local population also participated in the formation of a new association. Since, in view of the specificity of social organization in the nomadic societies, all ethnic processes associated with their history appear to be much more complex than those in the sedentary societies (Németh, 1991: 38–44; Akın, 1982: 2–3), and any attempts to equate the peoples inhabiting the same territory, but in different historical periods, are ungrounded.

It has been established that the valley of the Emel River, in the area of the Chuguchak River, was the place where the Chd-md-kun tribe had settled (Chavannes, 1903: 34, nt. 3; p. 73, nt. 2; p. 270, nt. 1; Malyavkin, 1989: 38, 163, comm. 232). In the year of 656, the “Xin Tang shu” and “Ce fu yuan gui” mentioned the Chd-md-kun “town of Yan - Щ ” ( Yancheng Щ^ ), which apparently was the center of the tribal possessions (Chavannes, 1903: 267, 270, nt. 2; p. 294, 307); cf.:

(Zuev, 1962: 119)². However, if we make a connection between this center and the territory of the district ( zhou 州 ) Yànmiàn 咽麫 , created in 702, which apparently coincided with the territory of the Fúyán 匐延 province ( dudufu ##W ) , formed in 657 (Chavannes, 1903: 281, nt. 2; Zuev, 1962: 120, nt. 83; Malyavkin, 1981: 188–189, comm. 286; 1989: 38, 163, comm. 232)³, we may assume: yan-mian < EMC * ?£n^h-mjian^h , LMC * ?jian'-mjian' (Pulleyblank, 1991: 358, 214), MC *?ian-mjiän (Schuessler, 2009: 319 (32–9h = K. 370), 250 (23–31a = K. 223)), < * emän , which is comparable to the name of the Emel River ((Chavannes, 1903: 270, nt. 1; Malyavkin, 1989: 38, 163, comm. 232), cf.: (Zuev, 1962: 120–121)). This river now flows into Lake Alakol, which together with the adjacent lakes Uyaly and Sasykol, at least in the early second millennium AD, probably formed one large lake (Gagan jl^ in al-IdiisT)) in the central part of the Semirechye possessions of the Kimäk (Kumekov, 1972: 70–74, 75).

The scholars who analyzed the fragment quoted by Gardizi repeatedly paid attention to the report on the special status of water among the Kimäk (Ögel, 1995: 326; Zuev, 2002: 128–129; 2004, No. 2: 9–10). Mention of water in this context is curious, since it may give us an opportunity to reconstruct the original sound of the tribal name chU-mU-kun < EMC * tehia' - mawk-kwan, LMC *tghia' / t§hya'-mawk-kun (Pulleyblank, 1991: 60, 220, 282), MC *tshjwo-muk-kwan (Schuessler, 2009: 49 (1-18a = K. 85), 161 (11–24ae = K. 1212), 333 (34–1a = K. 417)), < *cumuqun. The following etymology is probable: *cumuqun ~ *comuqun *‘immersed in water’, *‘drowned (?)’ < com-uq- ‘to drown’ (middle voice) (see (Erdal, 1991: Vol. 2, p. 646)), < com - ‘to sink in (water, etc. Loc.)’ (Clauson, 1972: 422) + -(X)k- + -Xn. It is theoretically possible to imagine this form as primary, if we assume that the wide vowel is labialized under the influence of the adjacent nasal consonant /m/: * cam - > com - (Erdal, 1991: Vol. 1, p. 391). The hypothesis as to the presence here of the word çomuk (dialect. çumak) > comuk (Zeki Velidi Togan, 1946: 51, 428, dipnot 182, 183) leaves the presence of the third syllable without explanation. We should also compare the variants of reconstruction suggested by Zuev: < *tsiwo-muk-kuen < ? cumul qun (1962: 119), cumuq qun (1967: 18; 1981: 66). The attempts to link this ethnonym with the group of words (personal names, toponyms, ethnonyms, and social terms) containing a wide vowel in the first syllable, for example, A-^- (gmwk) gamuk in Arabic writing (see (Iskhakov, Kamoliddin, Babayarov, 2009: 8–10; Babayarov, Kubatin, 2010: 16; Otaxo‘jaev, 2010, 65-67)) raise some doubts. For example, al-Tabari mentioned “the people from the house of al-g.muk” uAj-A1 ^ J^l (’hl byt ’l-gmwkyyn) present at the funeral of the Türk Qaghan who was killed in 119 AH / 737 AD4. The change in the form of the ethnonym may be explained by its reinterpretation, since the proposed variant *camoq ~ *camuq may be interpreted as a derivative of the same verb * cam- by means of the corresponding affix -(O)k (Erdal, 1991: Vol. 1, p. 224–261), which, in turn, makes it possible to further make the form of *comuq. In this case, this abstract verbal name in its essence is synonymous with the form *cumuqun ~ *comuqun.

In 649, 651, 739, and 740, the leader of this tribe was called ChUmUkun [Qu] Ltt Chud ЖА^(®)Ш^ (Chavannes, 1903: 34, 60, 65, nt. 4: 84, 270; Ta^agil, 1999: 71, 96; Malyavkin, 1989: 39, 168, comm. 248), that is * kulug cor (see (Hamilton, 1955: 96, nt. 8)). Such a reconstruction of the reading of this title (instead of the written form ChUmUkun Ltt Chud ЖА^Ш^ ) makes it possible to reject the E. Chavannes’s suggestion (Chavannes, 1903: 285–286, nt. 3; Beckwith, 1987: 118, nt. 60) to correlate the leader of the ChU-mU-kun with the Turgis (with the nisba ’l-trqsy A-') commander named Kursul Jj^jj^ ( kwrswl ), who killed the Qaghan in a quarrel (119 AH / 737 AD), and who was mentioned by al-Tabari. It seems more sensible to make a comparison with the Turgis tribal leader Mdhe Dagan M^^A (< * baya tarqan ), well-known from the Chinese sources, who killed Sulu ^Ш Qaghan (738) (Marquart, 1898a: 38-39, Anm. 1; 1898b: 181-182) ( su-lu < EMC * sa-lawk , LMC sua-lawk (Pulleyblank, 1991: 294, 201), MC *suo-ljwok (Schuessler, 2009: 52 (1-31с = K. 67), 159 (11–15klm- = K. 1208)), < * suluq (cf.: (Hirth, 1899: 77, Klyashtorny, 1986: 166, 169); cf. with the vowels of the palatal type (Zuev, 1998: 66))). If we take into account the hereditary nature of the titles, which is suggested, for example, by the epitaph of some “lady from the Ashina ИАЖ clan” furen ashina shi ЖАИАЖЮ , daughter of the governor ( dudu ## ) of Shuanghe Ж A , named Shesheti Tun Chud ^^^^^ (* Ton cor from the tribe Shesheti ®A^ ⁵; cf. the form Shesheti

摄舍提 )⁶, who married one of the Tang high-ranking commanders⁷, and most likely this commander belonged to the tribe of Huluwu ЩШЖ , whose leader, mentioned in the year 651, was called Huluwu Que Chud ЩШ ЖИ^ (< * uluY oq kul cor ) (Marquart, 1898b: 182; Chavannes, 1903: 34; Malyavkin, 1989: 39, 166, comm. 245; Ta^agil, 1999: 96).

It is curious that in the year 649, the Chinese sources mention BasaifU Chdmdkun MOheduo Q^jm ЙШ^ЖЖ ^МЙЖ^^ (in “ Xin Tang Shu”, Chumukun MOheduo Qijin ЖЖГьММ^ ) (Bichurin, 1950: Vol. 1, p. 263; Liu Mau-tsai, 1958: B. I, S. 155, 208; B. II, S. 585, Anm. 804, S. 646, Anm. 1139; Ta?agil, 1999: 40, 90) among the surrendered tribal chiefs ( qiuzhang ЙЖ)— the companions of the Cheb ЖМ Qaghan (< * cavis ) (see

(Ecsedy, 1980: 27; Kasai Yukio, 2012: 89)), who formerly resided on the northern slopes of the Mongolian Altai (see (Zuev, 2004: No. 2, p. 11-12)), where basai ЙШ is undoubtedly the transcription of the word bars (see, e.g., (Harmatta, 1972: 270, Malyavkin, 1989: 39, 169, comm. 251)); fú 匐 is the transcription of the Turkic word bäg ((Hirth, 1899: 107, Hamilton, 1955: 148–149), see also: (Harmatta, 1972: 270; Malyavkin, 1989: 41, 169, comm. 251)) (cf. the personal name bars bäg (Drevnetyurkskiy slovar, 1969: 84)); mOheduo М^Ж is the transcription of the word ba^atur (Chavannes, 1903: 83-84, 90, 346); and qijin ^^ is the transcription of the title of irkin (Hirth, 1899: 103, 109, 111–112; Pelliot, 1929: 227–228; Hamilton, 1955: 98, nt. 1; Kasai Yukio, 2012: 90)⁸. This makes it possible to consider the word Chumukun exclusively an element of a personal name. Thus there is every reason to believe that this name, being once the personal name of an individual leader, formed the basis for the name of the group under his leadership. This is a fairly well-known phenomenon among the nomads of the Eurasian steppes (Németh, 1991: 58–65).

One more point is remarkable in this respect in connection with the history of the Kimäk. While enumerating the peoples inhabiting the territory to the north of the Altai, “Tong Dian” mentions the combination of YanmO Nian Duolu Que Qijin ^^^Ж^И^ 斤 (Zuev, 1962: 105–106; cf.: Kyuner, 1961: 54)). In this combination, the last three hieroglyphs (que qijin) certainly denote the title of *kül irkin (see, e.g., (Zuev, 1962: 118)); the fourth and fifth, that is, DuolU, like all other forms of this combination used in the name of one of the tribal confederations of Western Türks, taken together, make it possible to reconstruct here the sounding of *tölük (see (Golden, 2012: 167)) or *türük (cf.: (Klyashtorny, 1986: 169)); the third character niàn < EMC *nemh, LMC *niam' (Pulleyblank, 1991: 225), MC *niem (Schuessler, 2009: 365 (38–24a = K. 670)), which, as Zuev pointed out (2004, No. 2: 3), is tempting to link with the Sogdian nam (n’m) ‘name’ (Gharib, 1995: 232); while the first and second characters, that is, yán-mò < EMC, LMC *jiam-mak (Pulleyblank, 1991: 357, 218), MC *jiäm-mâk (Schuessler, 2009: 347 (36– 5n = K. 609), 74 (2–40ad = K. 802)), < *yemäk (Zuev, 1962: 118). With a significant degree of certainty, it may be assumed that the reconstructed *yemäk nam tölük (/türük) kül erkin, which originally had clearly designated a personal name, in the Chinese text marked some subordinate group, which was the subject of a certain leader. The word *yemäk here may act as an element of the personal name of that leader, and denote the name of the tribal group from which he originated. The first option is preferable. If the interpretation of the second element in the reconstructed combination of the Sogdian lexeme is correct, then *yemäk nam may literally be interpreted as ‘the one bearing the name of yemäk’ (Zuev, 2004: No. 2, p. 3). The former assumption seems more logical, if we take into account that, further into the source, the combination of yán-mò (< *yemäk) occurs independently.

The word yemäk , which has been mentioned in the Chinese sources at least since the mid-7th century (Kyuner, 1961: 55), makes us turn again to the hypothesis (which has become commonly accepted by the scholars) of considering this word a secondary form of the word kimäk. The former word has reliably and independently appeared in sources written in Arabic since the second half of the 11th century AD in the form ^-j' ( ’ymak ) as the name of one of the main tribes of the union called ^Ц£ ( kymak ) (sometimes ^/^ ( kymyak )) in Gardizi, and in the form ^Ц> (ymak ) as the name of this entire union in Mahmud al-Qasyari, who did not know any Kimäk (see (Zuev, 1962: 121–122; Kumekov, 1972: 39– 41; Golden, 1992: 202; 2002)). Differences in writing can be explained by the rules of Turkic phonetics: *imak < * yemäk ~ * yimäk , which is adequately linked with the data of the 7th century.

According to K. Czegledy, the narration of Gardizi about the Kimäk , like his stories about other Turkic tribal groups, may refer to events that happened between 745 and 766 (1973: 263–267). Notably, Czeglédy dated the “Turkic episodes” only on the basis of information about the tribes of Qarluq and Yayma . Nevertheless, as Lurje has shown, the date proposed by Czeglédy is generally confirmed by the indirect evidence (2007: 189–190).

Abu Sa‘id Gardizi, who had no knowledge of the Turkic languages, as Czeglédy had shown, borrowed information about the Turks from the author of the “Kitab Rub’ al-dunya” (“The Book of the Inhabited Quarter of the World”) by the name of Abu Muhammad ‘Abd Allah Ibn al-Muqaffa‘ (720 to ca 757) (Czegledy, 1973: 259, 260–261, 263). More convincing is the suggestion of Lurje (2007: 189-190), according to which Gardizi might have taken this information from another source that he mentioned, “Kitab al-Masalik w’al-Mamalik” (“Book of Roads and Kingdoms”, which has not survived) by Abu ‘Abd Allah Jayhani (first half of the 10th century), who served as a wazir at the court of the Samanids.

The name of Kimäk has been reliably recorded in the most common form of ^Ц£ ( kymak ) since the 9th century, although it might already have been known in the second half of the 8th century (Kumekov, 1972: 11–13, 36, 56). The latter date appears in the list of the Turkic tribes, which was given in the book “Kitab al-Masalik w’al- Mamalik” (“Book of Roads and Kingdoms”) by Ibn Khordadbeh (the 880s), which was one of the sources of Gardizi.

Conclusions

The above analysis makes it possible to conclude that the hypothesis of Gumilev as to identifying the Semirechye tribe of Chumukun with the Kimak , which was based solely on the data regarding the coincidence of the territories inhabited by them, may find additional, albeit indirect, confirmation in the reconstruction of the Chinese sound of the name of the Semirechye tribe of Chumukun as * cumuqun ~ * comuqun with the meaning *‘immersed in water’, *‘drowned (?)’, which echoes the story about why the Kimäk worshipped water, which was cited by the Persian author of the 11th century Gardizi. The formation of the Kimak tribal union, according to Gardizi, occurred exactly in the Irtysh valley, where the representatives of various tribal groups arrived. The most important of these tribal groups was the group of Yemäk . Its name occurred in the Chinese sources in the form of Yánmò as early as the mid-7th century. Initially, this name was mentioned as the personal name of a certain leader.

Without addressing the issues of migration-processes associated with the formation of a new tribal community and of their dating, we should emphasize that the formation was complex and involved both local and migrant populations.

We should also pay attention to the doubts voiced by Zuev, contrary to the opinion of most scholars, about the impossibility of identifying the names of Yemäk and Kimäk as forms of the same word. The identification of these two forms as *yimäk <  *kimäk has been accepted by the scholars on the basis of reduction of the initial * k- > 0 , observed by the philologers in some Middle Kipchak dialects, which has not been found in the Old Turkic period. Together with the indirect data on the existence of both forms (for the 7th and 8th–9th centuries, respectively), this circumstance suggests another explanation for the consonance in the names used in relation to the same tribal group.

Acknowledgement

This study was supported by the Russian Foundation for Basic Research (Project No. 16-31-01029а2).