“Tribe,” “subtribe,” and “clan” are troublesome terms, inadequately refracted in our sources. This is widely known, accepted, and noted in my review. David Sneath continues to ignore Kâšgharî, an “insider” who is explaining his tribal society to the larger Muslim world. One has to look at the original texts (Turkic, Arabic and Persian), which Sneath has not done, to see the context. One cannot arbitrarily devise new meanings for words to suit a theory. The language of kinship (real, politically “negotiated,” or “constructed”) was used within steppe society to express political bonds. Our sources, Chinese, Greek, Arabic, and Persian, simply reflect what their local informants tell them.
According to Kâšgharî, boy (rendered by him as “kinsmen, clan, people,” “kin, subtribe, clan”) consists of “subtribes” (Arabic qabîla, pl. qabâ'il; baṭn, pl. buṭûn), named after founding ancestors, and even smaller subdivisions (1983–85, 1:267, 2:219; facsimile ed., 170, 505). The subtribes bear the names of eponymous ancestors (ajdâd) “who gave birth to them in olden times. They trace their ancestry back to them” (Kâšgharî, facsimile ed. 40–41; see also Kâšgharî 1983–85, 1:102, facsimile ed., 624; Kâšgharî 1983–85, 2:363). Kâšgharî is describing ever-smaller units of descent groups, using appropriate Arabic terms. Yes, one can find a number of meanings for qabîla, baṭn, and so on, but the kinship context of these terms in our sources is clear and well known to the Arabo-Persian historico-geographic literature. Arabic baṭn (pl. buṭûn) can mean “belly, abdomen, womb, interior part” and “a tribe below that which is termed qabîla” (Wehr Reference Wehr and Cowan1994, 78–79; Lane Reference Lane1986–93, 1:219–21). Obviously, Kâšgharî is not discussing “bellies” or “wombs.”
Pre-Soviet Qazaq nomads, when meeting, still inquired qay ruwdan keldingiz (what clan do you come from? cf. Schatz Reference Schatz2004, 28). Post-Soviet Qazaqs, when introduced, perform a similar kind of “interrogation,” reciting “at least some of their ancestral lineage” (Privratsky Reference Privratsky2001, 115), including jüz (horde), clan (ruw < urugh “seed, pip, kernel … metaphor. progeny, descendants … clan, i.e. descendants of a common ancestor,” Clauson Reference Clauson1972, 214). Ruw located one in Qazaq society, determined social status, even one's place at feasts, order of giving toasts, and so on (Masanov et al. Reference Masanov2008, 78–82). Every Qazaq, noble and commoner, is supposed to be able to recite his or her jeti ata, “seven forefathers.” Given the disruptions in Qazaq society over the past three centuries, many now cannot and resort to “inexpensive booklets” that provide genealogies of the major clans. Nonetheless, clan and subclan names are still used in daily conversation to designate “a third party” when the family name is unknown (Privratsky Reference Privratsky2001, 116–17). Tribe, clan, and knowledge of ancestors are equally important in Türkmen society among others (Edgar Reference Edgar2004). Clan-based political alignments remain a powerful force in Qazaq domestic politics (see Schatz Reference Schatz2004).
Rashîd al-Dîn's descending order of qowm/qawm > qabîla > shu‘ba (< Arabic “a branch of a tree, branch,” cf. cognate sha‘b “a tribe,” Lane Reference Lane1986–93, 1:2555–57) should be seen in the same light as Kâšgarî's subdivisions. Sneath's claim that Clauson (Reference Clauson1972, 297–98) offers “a wide range of possible terms for bod/bodun, (clan/tribe), many with no tribal or kin connotation,” is misleading. The Turkish bod (> boy ) had the original meanings of “body, stature; self; kin; tribe, color” (Starostin et al. Reference Starostin2003, 1:365), but these homonyms represent at least two distinct semantic groups. Of bod/boy, Clauson states unambiguously, “from the earliest period it also clearly means ‘a clan,’ the Sing. of bodun, q.v.; the meaning underlying the latter is perhaps ‘lineage’ that is the size of a clan as a number of natural families.” Bodun (1972, 306) denotes “‘clans,’ in practice a semi-technical term for ‘an organized tribal community, a people,’ in the sense of a community ruled by a particular ruler …” p. 322: “the tribes, clans, common people.” It is often prefaced by an ethnonym (e.g., Toquz Oghuz bodun “the Toquz Oghuz people”). In discussing sociopolitical organization, it clearly means “kin, tribe.”
No one denies the existence of “nobles” and “commoners” in steppe society, but Sneath has not explained how his “aristocratic orders” arose and acquired power (if not from clan or tribal bases), or how they functioned in “headless,” stateless confederations, the default form of organization in much of the central and western steppes. Chinggis Khan and his sons broke up the earlier tribal/clan-based order seeking to focus the bonds of loyalty exclusively on the altan urugh, the Chinggisid house. This was a revolution and even changed the naming system of these groups that subsequently retribalized. These dispersed groups can be seen in the “Qïpchaq” subdivisions among the Uzbeks, Qara Qalpaqs, Qazaqs, Qïrghïz, Altay Turks, Noghais, and Crimean Tatars, the “Qitan” subdivisions among the Uzbeks, Qara Qalpaqs, Qïrghïz, and Crimean Tatars, and so on.
A state requires centralized institutions, a juridical system, a military, and a monopoly over the use of force, a system to extract taxes and other agencies of governance, all propped up by an ideology. Modern states, imposing their authority on the nomads, named administrative units after the names of the peoples, tribes, and clans inhabiting the regions they conquered rather than the reverse.
The Headless State has more than its fair share of factual errors (some of which are noted in my review) or misleading oversimplifications. I might add here: Ozmïsh was not a “usurper” of power from the Türk Qaghan (p. 20). As an Ashina, he had the right to assert his claim to the throne. The Timurids did not rule “much of the Middle East until the sixteenth century before finally collapsing in the face of the waxing power of the native Persian Safavid dynasty in the east and the Ottomans in the west” (p. 33). Their center of power was Transoxiana and Khurasan (today eastern Iran/western Afghanistan) with capitals in Samarqand and heart, respectively. Timurid Transoxiana fell to the Uzbeks in 1500. The Safavids took Herat from the Uzbeks in 1510, but it remained contested for some time after that. The Ottomans conquered the Aq Qoyunlu, victors over the Qara Qoyunlu, who had gained control of these areas that were leaving the Timurid orbit already during the reign of Tamerlane's son and successor Shahrukh (1405–47). Safavid origins are complex. The dynasty's ancestor may have been Kurdish or Persian. They were Turkish speakers (and heavily intermarried with Turks) before 1500. Their movement was based on Turkish-speaking tribes from Anatolia, Syria and Azarbaijan. “Native Persian” does not really explain them or the forces they brought under their control.
Sneath's schema may help us understand postimperial Chinggisid-centered Mongolian groupings working within the ruins of a state, but such a system cannot be anachronistically imposed on earlier periods and posited as an analytical tool for understanding the internal governance of the steppe as a whole, in particular the pre-Chinggisid steppe.
