1. Gu, Ban , Han shu (Beijing: Zhonghua, 1983), 94A.3790–91Google Scholar.

2. 53 b.c.e. is the date Huhanye sent his son to the Han court as guarantor of the treaty (Han shu, 94B.3797). Di Cosmo uses a date of 51 b.c.e. (pp. 206, 226), which corresponds to the year Huhanye was received at the Han court for a state visit (Han shu, 94B.3798). The debate about the correct protocol for the occasion seems to have taken place ca. 52 b.c.e. (Han shu, 8.270), not 60 b.c.e. (Di Cosmo, p. 221).

3. Qian, Sima , Shiji (Beijing: Zhonghua, 1982), 110.2879–920Google Scholar.

4. kaogudui, Beijing shi wenwu yanjiusuo Shanrong wenhua , “Beijing Yanqing Jundushan Dong-Zhou Shanrong buluo mudi fajue jilue” , Wenwu 1989.8, 17–35, 43Google Scholar, Color Plate, PI. 3–5.

5. In reference to the use of the bird head décor in Maoqinggou (Liangcheng, Inner Mongolia), he notes that “Such findings can indicate differences in a society's development toward more widespread use of horses and in its range of contacts with other cultures” (p. 77). Apart from being unsubstantiated, this statement suggests that Di Cosmo views cultural exchange as taking place primarily through the movement of peoples, rather than through trade. Otherwise, were the Maoqinggou bird head indeed Saka, its appearance in Maoqinggou would not be indicative of nomadism. Di Cosmo's view is made explicit on page 37, where he writes, “Externally, nomadic-dominated societies tended to have more contacts with neighboring communities.” Again, Di Cosmo provides no substantiation, or even examples.

6. yanjiusuo, Shanxi sheng kaogu, wenhuaju, Lingshi xian , “Shanxi Lingshi Jingjiecun Shang mu” , Wenwu 1986.11, 7 Fig. 11:5Google Scholar.

7. Yun, Lin “A Reexamination of the Relationship between Bronzes of the Shang Culture and of the Northern Zone,” in Studies of Shang Archaeology, ed. Chang, K. C. (New Haven and London: Yale University Press, 1986), 242 Google Scholar.

8. Lin Yun “Reexamination of the Relationship,” 245.

9. Lin Yun “Reexamination of the Relationship,” 244.

10. Bokovenko, Nikolai A., “Scythian Culture in the Altai Mountains,” in Nomads of the Eurasian Steppes in the Early Iron Age, ed. Davis-Kimball, Jeannine (Berkeley: Zinat Press, 1995), 292 Google Scholar.

11. Chernykh, E. N., Ancient Metallurgy in the U.S.S.R.: The Early Metal Age, translated by Wright, Sarah (Cambridge: Cambridge University Press, 1992), 194–95Google Scholar.

12. Chernykh, Ancient Metallurgy, 224.

13. Chernykh, Ancient Metallurgy, 226.

14. For this explanation, see Chernykh, Ancient Metallurgy, 216.

15. Chernykh, Ancient Metallurgy, 228.

16. Chernykh, Ancient Metallurgy, 200, 210.

17. I follow the dating adopted by Okladnikov, A. P., “Inner Asia at the Dawn of History,” in The Cambridge History of Early Inner Asia, ed. Sinor, Denis (Cambridge: Cambridge University Press, 1990), 84 Google Scholar.

18. Chernykh, Ancient Metallurgy, 269.

19. Chernykh, Ancient Metallurgy, 269.

20. I am following the latest dating proposed by Vadeckaja, E. B. , Arxeologičeskie pamjatniki v stepjax srednego Eniseja (; Leningrad: Nauka, 1986), 101 Google Scholar. For examples of these bow-shaped objects, see Vadeckaja, Arxeologičeskie pamjatniki, 57, Fig. 5:32 (Karasuk); 88, Fig. 6:16, 89, Fig. 6:17 (both Tagar). Contrary to Di Cosmo's assertions that M. P. Grjaznov's chronologies of Altaian archaeology not only represent consensus for the Altai but implicitly apply to Tuva as well (pp. 35–36), there continues to be considerable debate about the chronology of Central Asian and steppe cultures. Vadeckaja provides a summary of the debate for the Yenissei River region. Jettmar, Karl, “The Altai before the Turks,” Bulletin of the Museum of Far Eastern Antiquities 23 (1951), 159–60Google Scholar, critiques Grjaznov and Sergei I. Rudenko's dating, which builds on Grjaznov's, noting that Rudenko ignored “objects which do not fit into Rudenko's picture.” Clearly, Grjaznov ignored parallels within the material from the few tombs then known in the High Altai that made impossible the wide spread of dates he advocated. This article by Jettmar is absent from Di Cosmo's bibliography.

21. Chernykh, Ancient Metallurgy, 270–71.

22. Koshelenko, G. A., “L'Asie centrale au début de l'âge du fer: le problème des relations extérieures,” in L'Asie centrale et ses rapports avec les civilisations orientales des origines à l'âge du fer (Paris: de Boccard, 1988), 172 Google Scholar.

23. See Guangjin, Tian and Suxin, Guo , E'erduosishi qingtongqi (Beijing: Wenwu, 1986), 96 Google Scholar, Fig. 64:4, 6.

24. Early China 13 (1988), 201–40Google ScholarPubMed.

25. Bulletin of the Museum of Far Eastern Antiquities 57 (1985), 133–60Google Scholar, Fig. 1–25.

26. Jacobsen, “Beyond the Frontier,” 225.

27. Bunker, Emma C., Chatwin, C. Bruce, Farkas, Ann R., ‘Animal Style’ Art from East to West (New York: The Asia Society, Inc., 1970), 82n23 Google Scholar. Bunker appears to be following Jettmar, Karl, The Art of the Steppes (New York: Crown Publishers, 1967), 154 Google Scholar. Bunker provides this reference to Jettmar in ‘Animal Style’ Art, 109n40.

28. For Aržan, Di Cosmo refers to Grjaznov, M. P., Großkurgan von Aržan in Tuva, Südsibirien (Munich: Beck, 1984)Google Scholar and Nikolai A. Bokovenko, “The Tagar Culture of the Minusinsk Basin,” in Nomads of the Eurasian Steppes in the Early Iron Age, 302; for the remainder of this passage, Marina Moshkova, “Sarmatians, Concluding Remarks,” in Nomads of the Eurasian Steppes in the Early Iron Age, 185–88.

29. His subsequent description of Andronovo is credited to Hiebert, Fredrik T., Origins of the Bronze Age Oasis Civilization in Central Asia (Cambridge, Massachusetts: Peabody Museum of Archaeology and Ethnology, 1994), 135 Google Scholar.

30. Citing Anthony, David and Vinogradov, Nikolai B., “Birth of the Chariot,” Archaeology 48.2 (1995), 38 Google Scholar.

31. Unfortunately, Di Cosmo gives no reference for Andronovo's “unquestioned economic superiority.” He likewise provides no explanation of this phrase. Interestingly, he considers Andronovo sedentary “agro-pastoralists” (p. 28), despite his reference to the culture spreading “by horseback or by chariot” (p. 29).

32. Chernykh, Ancient Metallurgy, 233.

33. Following Chernykh Ancient Metallurgy, 125,132.

34. Following Chernykh Ancient Metallurgy, 194.

35. Chernykh, Ancient Metallurgy, 195, 232.

36. Chernykh, Ancient Metallurgy, 211.

37. Chernykh, Ancient Metallurgy, 210–11, 215.

38. Curiously, Di Cosmo writes that “The inefficiency of this [Sintashta] type of chariot is made even more evident by the probable availability of horseback riding, which clearly was a superior means of transportation, herd control, and warfare” (pp. 28–29). His initial assertion is speculative; his conclusion is invention.

39. The same is true in his political discussion of Shang foreign relations, of which he writes: “At this early stage, the northern frontier societies constantly interacted with the Shang and early Chou, and even though a frontier did exist, no sharp demarcation can be detected” (p. 6). Apparently, he does not see the marked disparity in the amount and type of data available, both written and archaeological, as skewing comparison with later periods.

40. Di Cosmo groups bells (and bell-shaped ornaments), “head ornaments and masks,” and ornaments depicting horses together with horse gear (pp. 64–65). I do not know to which objects he is referring as “head ornaments and masks,” nor do I understand why bells are necessarily horse (or chariot) gear. No specific references are provided. For Nanshan'gen bits, see gong-zuozhan, Liaoning sheng Zhaowudameng wenwu, gongzuodui, Zhongguo kexueyuan kaogu yanjiusuo dongbei , “Ningcheng xian Nanshan'gen de shiguo mu” , Kaogu xuebao 1973.2, PI. 11:13 (M101:80)Google Scholar; gongzuodui, Zhongguo shehui kexueyuan kaogu yanjiusuo dongbei , “Neimenggu Ningcheng xian Nanshan'gen 102-hao shiguo mu” 102 , Kaogu 1981.4, 306, Fig. 4:4–6Google Scholar (M102:52, presumed to be part of a bit; M102:54, 58). For nine Nanshan'gen tombs without bits or other horse gear, see yanjiusuo, Zhongguo shehui kexueyuan kaogu gongzuodui, Neimenggu , “Ningcheng Nanshan'gen yizhi fajue baogao” , Kaogu xuebao 1975.1, 117–140, PI. 1–7 Google Scholar. Note that one of the ornaments depicting horses (in bronze) to which Di Cosmo refers is M3:6 in this report (137, Fig. 19:4); the incised bone plaque showing horses is M102:18, in the Kaogu 1981.4 report cited above (307, Fig. 6). Di Cosmo does not provide illustration-specific references.

41. Use of the word “chariot” assumes that the type of vehicle to which these axle caps were attached is known; it is not. Further, since such vehicle components were used by the Chinese as well, axles and other vehicle remains cannot as a form be markers of nomadism, regardless of the ultimate origin of the vehicle. Although the excavation reports associate these pieces with vehicles, it is unclear why, particularly when a tomb yields only one cap, rather than a pair. For examples, see Tian and Guo, E'erduosishi qingtongqi, 156 (citing Hulusitai [United Middle-Rear Banner of Wulate, Inner Mongolia] M2:17, Yulongtai [Zhungar Banner, Inner Mongolia], and Sujigou [Zhungar Banner, Inner Mongolia]); PI. 15:1,102:1.

42. In contrast, Di Cosmo implies that the ge is an indicator of nomadism (p. 76).

43. For an index to shortswords from southern Inner Mongolian sites, see Tian and Guo, E'erduosishi qingtongqi, 11.

44. yanjiusuo, Neimenggu wenwu kaogu , “Liangcheng Guoxianyaozi mudi” , Kaogu xuebao 1989.1, 79–80 Google Scholar; Tian and Guo, E'erduosishi qingtongqi, 306–13.

45. Tian and Guo, E'erduosishi qingtongqi, 227.

46. Although Di Cosmo notes late nineteenth century critiques of the earlier emphasis on evolution in the social sciences (p. 21), he himself clearly views history as evolutionary. Throughout his book, he refers to nomadism as the result of evolution (particularly pp. 23, 36, 37, 59).

47. Curiously, Di Cosmo refers to Taohongbala as a culture (p. 78). He incorrectly states that “In the [excavation report] reprint of 1986 [i.e., in Tian and Guo, E'erduosishi qingtongqi] the site was attributed no longer to the Hsiung-nu but to the Pai Ti” (p. 76, n. 94). The reprinted excavation report mentions various branches of Di in a summary of historical references, but concludes that the site may be considered Early Xiongnu or Proto-Xiongnu (Tian and Guo, 219).

48. It is commendable that Di Cosmo (pp. 151-5–2) places the sites of Taohongbala (Hangjin Banner, Inner Mongolia), Hulusitai (United Middle-Rear Banner of Wulate, Inner Mongolia), Yulongtai (Zhungar Banner, Inner Mongolia), Xigouban (Zhungar Banner, Inner Mongolia), and Sujigou (Zhungar Banner, Inner Mongolia) in geographic context in relation to the walls built by the Chinese states of the Warring States (ca. 5th–3rd cent, b.c.e.); on p. 151, he mistakenly locates Xichagou (Xifeng, Liaoning) in the Ordos. Throughout this discussion, Di Cosmo refers to these sites as Xiongnu, but concludes that although these sites constitute “a relatively homogeneous culture,” the latter would be “more correctly defined simply as ‘early nomadic”’ rather than Xiongnu (p. 152).

49. yanjiusuo, Gansu sheng wenwu kaogu , “Yongchang Sanjiaocheng yu Hamadun Shajing wenhua yicun” , Kaogu xuebao 1990.2, 237 Google Scholar.

50. Di Cosmo seems to include Qingyang (Gansu) within his “Guyuan culture” (pp. 80–81; see also p. 151). However, neither Guyuan nor Qingyang forms an archaeologically homogeneous region, nor do the two regions together constitute a cultural whole.

51. On page 81, however, he refers to “Guyuan” as “pastoral.” Given that he uses the same term to define Xiyuan as “not quite ‘early nomadic,”’ it is impossible to know what meaning he assigns to the word. Normally, “pastoral” used alone implies sedentarism.

52. Trever, Camilla Excavations in Northern Mongolia (1924–1925), Memoirs of the Academy of History of Material Culture 3 (Leningrad: J. Fedorov, 1932), 14–15 Google Scholar; PI. 29:2.

53. Harvard Journal of Asiatic Studies 21 (1966), 68–173 Google Scholar.

54. Conquerors and Rulers (Leiden: Brill, 1952)Google Scholar.

55. Di Cosmo does not mention that Owen Lattimore identifies this narrative as an “epic or heroic account” in Inner Asian Frontiers of China (New York: American Geographical Society, 1940), 463 Google Scholar.

56. For both the Shi ji and the Han shu, Di Cosmo provides only title, sometimes chapter (juan) number, and page number; author and edition are not given, although page numbers are edition-specific. The first footnoted reference I can find in his work is page 95, note 5 for the Shi ji and page 143, note 40 for the Han shu. Neither work appears in the bibliography. The bibliography itself is poorly organized, with some works cited by author, others by title, and still others, particularly Chinese archaeological reports, by journal name and date. Important reference books may appear only as individual chapters, cited by chapter author and title, making some familiar volumes difficult to recognize.

57. Di Cosmo spends considerable time on the emergence of the Xiongnu as an international power under Modun (pp. 178–90); he confuses this emergence of empire with the emergence of statehood. We know essentially nothing of the Xiongnu prior to Modun. When or how they became a state is entirely a matter of conjecture.

58. Di Cosmo seems to view “Scythian” as a generic term equivalent to the Chinese hu (nomad). His use of “Scythian” (with or without quotation marks) to mean “nomadic” is both confusing and inappropriate, since the term properly refers to a specific culture. He further confuses the date of Scythian culture as known from archaeological finds with the date at which the term appears in Greek texts, as on page 32, where he writes, “… Greek sources of the ninth and eighth centuries b.c. … assign these [nomadic] groups names such as Cimmerians, Scythians, and Sakas.” This further comment suggests that the names are interchangeable, rather than referring to discrete cultures and peoples, not all existing in the same place or at the same time. Notably, on page 32, footnote 53, seemingly unaware of the discrepancy, he claims that “the first historically documented [i.e., textually documented western] steppe nomads are the Scythians, treated extensively in Greek historiography (sic) from the fifth century [b.c.] onward.” On pages 37–38, he speaks of “a Scythian culture” in the Northern Caucasus in the early 8th cent, b.c.e.; since he does not mention any sites by name, it is not immediately clear that he does in fact mean Scythian in the proper sense. On occasion, he also employs the term “Scytho-Siberian” as a synonym for “early nomad”: “‘early nomads’ or Scytho-Siberian peoples, as they are sometimes called by archaeologists” (p. 32). This, too, is a misuse of terminology. “Scytho-Siberian” is an adjective used primarily by art historians to encompass the entire steppe from west to east; it refers often, but not necessarily, to nomadic cultures and is not era-specific (i.e., “early” or “late” nomadic).

59. This comment follows a presentation of non-Chinese archaeological sites in southern Inner Mongolia. No data is cited in support of the remark.

60. Does Di Cosmo mean the emperor (i.e., of China) or is this in error for chanyu (i.e., of the Xiongnu)?

61. In his discussion of the heqin on page 217, Di Cosmo does offer a political, rather than economic, interpretation (“The ho-ch'in policy … was a long-term strategy aimed at absorbing the next generation of the Hsiung-nu leadership within the Han court's political sphere and neutralizing the Hsiung-nu as an independent and inimical state”). However, it, too, is marked by the same weaknesses noted above.

62. In this context, Di Cosmo also refers to “the major challenge in 156 b.c.” (p. 230), which we understand from context had to do with the emperor facing rebellion among his subordinates; no other information is provided and no reference of any kind. He may be referring to the 154 b.c.e. Rebellion of the Seven Kingdoms.

63. Canberra: Australian National University, 1984, 194–218.

64. See Cosmo, Di, “The Economic Basis of the Ancient Inner Asian Nomads and Its Relationship to China,” journal of Asian Studies 53.4 (1994), 1092–1126 CrossRefGoogle Scholar.