1. Yan Xiuyu and Xie Yingju, 1:213-222; Mao Zedong 1982, 41-181. The text of Mao's "Xunwu Investigation" was first published in this 1982 collection, using an original copy from the Central Party Archives. Reports from Xunwu , an annotated translation by Roger Thompson, is scheduled for publication by Stanford University Press in 1990. [BACK]

2. Mao Zedong 1982, 44-55. For the concept of macroregion, see Skinner 1977a. Originally Skinner included Jiangxi within the Middle Yangzi macroregion as it appears on Map 1.1., but he has since come to consider it a macroregion in its own right. [BACK]

3. Ganzhou fuzhi 2:41a, 42a; Ruijin xianzhi gao , 54; Jiang Yuchang, Gongdu cungao 1:60a; Yang Chaolin, 1:28b-29a. [BACK]

4. This general description of the southern Jiangxi hill country is derived mainly from my earlier work. See Averill 1983, 1987, esp. 280-281. For a vivid series of reports on social unrest and antigovernmental activity in early twentieth-century southern Jiangxi, including Xunwu (then known as Changning), see Jiang Yuchang. [BACK]

5. Mao Zedong 1982, 56-61, 79-81, 87-88, 99. [BACK]

6. I accept Robert Hymes's definition of local elites as those "whose access to wealth, power, or prestige was, in the local scheme of things, especially privileged: whose control of material resources, hold over men's actions and decisions, or special place in the regard of their contemporaries, set them apart from ... society as a whole and made them people to be reckoned with." Hymes 1986a, 7. [BACK]

7. In the "Xunwu Investigation," Mao Zedong (1982) also distinguishes among local elites, dividing them into categories such as "merchants" and "landlords," then ranking them on the bash. of economic criteria such as amount of capital or quantity of grain produced on land owned. The present account draws heavily on the information Mao provides but seeks to set it in an analytical framework that more systematically includes noneconomic factors and other elite categories. [BACK]

8. The term shi (picul) as Mao uses it appears to be equivalent to a dan (also translated picul), which is usually considered to contain 100 jin (catties), each of which is approximately one-half kilogram. Chinese measures varied considerably by region, however, and Mao notes that in Xunwu the shi contained 180 jin (i.e., 90 kg.). I have therefore used this figure in the text when converting shi into metric figures. To put these figures into some context, Mao reported that two shi of grain per year was enough to support a person. See Mao Zedong 1982, 170. [BACK]

9. Mao Zedong 1982, 113-115.

10. Ibid., 113-114; Pan Shanfu, 25. [BACK]

9. Mao Zedong 1982, 113-115.

10. Ibid., 113-114; Pan Shanfu, 25. [BACK]

11. For a description of the central place hierarchy, see Skinner 1977a. [BACK]

12. Mao Zedong 1982, 67, 100, 102.

13. Ibid., 162-163; Zeng Biyi, 88. [BACK]

12. Mao Zedong 1982, 67, 100, 102.

13. Ibid., 162-163; Zeng Biyi, 88. [BACK]

14. Mao Zedong 1982, 118-119.

15. Ibid., 126-1300.

16. Ibid. See also Jiang Yuchang (c. 1908) for descriptions of local elite irregularities and antigovernment attitudes. [BACK]

14. Mao Zedong 1982, 118-119.

15. Ibid., 126-1300.

16. Ibid. See also Jiang Yuchang (c. 1908) for descriptions of local elite irregularities and antigovernment attitudes. [BACK]

17. Mao Zedong 1982, 119. [BACK]

18. Zeng Biyi, 88. [BACK]

19. Mao Zedong 1982, 105. Mao estimated that landlords constituted 3.45 percent of the total Xunwu population, a figure supported by extrapolations from information provided elsewhere in his text. If allowance is made for nonlandowning merchants, teachers, and other elites not included in Mao's estimate, a figure of 3.5 to 4.0 percent seems reasonable. [BACK]

20. Mao Zedong 1982, 115, 359n. [BACK]

21. Hsien-chin Hu 1948; Hui-chen Wang Liu 1959b; Mao Zedong 1982, 106-108. [BACK]

22. Hsien-chih Hu, 20-30; Mao Zedong 1949a, 22; Jiang Yuchang, 64b; Xiao Hua, 227. [BACK]

23. Mao Zedong 1982, 129-131, 145-149; Mao Zedong 1949a, 22-24. [BACK]

24. Mao Zedong 1949a, 22; Polachek, 812-813. [BACK]

25. See essays by William Rowe and Timothy Brook in this volume, and Beattie 1979b. [BACK]

26. Xunwu, and doubtless many other peripheral Jiangxi counties, had relatively few gentry during the Qing period. By the late nineteenth century the county elite could claim only 4 civil and 1 military jinshi (metropolitan graduate), 10 civil, 4 "imperial favor," and 52 military juren (provincial graduate), and 186 gongsheng (senior licentiate) degree winners. The Pans (present in the county from Song times onward and whose patriarch, Pan Mingzheng, was the dominant elite figure in the county in the early twentieth century) had won no major degrees whatsoever during the Qing and had only a handful of the honorific gongsheng degrees presented to elderly men who had repeatedly taken but failed the regular examinations. Even acknowledging the likely presence in the county at any one time of several hundred shengyuan (whose names are not listed in the local gazetteers) there must have been many local elites who were not degree holders and whose elite status was based on other attributes. Changning xianzhi, juan 2. [BACK]

27. Deng Zihui, 60. [BACK]

28. Qiu Zhou, 64-65. [BACK]

29. Interview, Xiao Zhengqing et al., Xingguo county, May 26, 1984. [BACK]

30. These comments on county magistrates are derived from such Republican period Jiangxi gazetteers and other sources as Ninggang xianzhi, houzhi :3-14; Fenyi xianzhi , 10; Jiangxi minzheng gongbao , 8 (Apr. 16, 1928). For Qing conditions, see Watt 1972. [BACK]

31. Zhonggong Ruijin xianwei, 6a.

32. For other examples, see ibid., 5a-b; Jiang Yuchang, 1:60a. [BACK]

31. Zhonggong Ruijin xianwei, 6a.

32. For other examples, see ibid., 5a-b; Jiang Yuchang, 1:60a. [BACK]

33. E.g., Tse-tsung Chow 1960. This and most other relevant works concentrate upon the political impact of urban institutions of higher education. As this essay hopes to show, the politics of the May Fourth era and after were also shaped by events in lower-level, more rural schools. [BACK]

34. This section on education follows Averill (1987, 282-285) and the sources cited therein. [BACK]

35. For examples of provincial government inspection of schools, see Jiangxi jiaoyu gongbao , 2.14 (Jan. 1, 1929):25-28; 3.2 (May 11, 1929):37-45. [BACK]

36. For examples of such networks, see Yang Yuanming, 35, and Jiang Bozhang. [BACK]

37. It is unclear just when this involvement in factional politics began, but certainly the late Qing reformist and revolutionary movements set in motion various changes—the formation of study associations ( xuehui ), the New Policies (including

the establishment of the new schools themselves), and the founding of organized political parties—that encouraged its growth. [BACK]

38. Comments on county-level schools are based on republican period gazetteers; comments on old-style village schools are based on similar biographical materials published in the Taiwan journal, Jiangxi wenxian , and a variety of PRC sources. For an example from Huichang county, on Xunwu's northern border, see Liu Lucheng, 116-119. Information on Xunwu is from Mao Zedong 1982, 160-163. Altogether in 1930 the county had 30 university graduates (6 of whom had also studied abroad), 500 middle-school graduates, and 800 upper elementary-school graduates. It also had 400 living Qing shengyuan and one juren . [BACK]

39. Mao Zedong 1982, 125-126, 161-162.

40. Ibid., 126-131, 161-162. Mao's impressionistic comments on educational attitudes should be read with caution. But his conclusions are plausible, consistent with other available evidence, and well informed. He drew his information from local CCP cadres, such as Gu Bo, who had past personal involvement in Xunwu's educational circles and considerable recent opportunity to reflect upon the county's class structure. [BACK]

39. Mao Zedong 1982, 125-126, 161-162.

40. Ibid., 126-131, 161-162. Mao's impressionistic comments on educational attitudes should be read with caution. But his conclusions are plausible, consistent with other available evidence, and well informed. He drew his information from local CCP cadres, such as Gu Bo, who had past personal involvement in Xunwu's educational circles and considerable recent opportunity to reflect upon the county's class structure. [BACK]

41. Hobsbawm, chaps. 2, 6; Liu Xiaonong 1980; Zou Fuguang 1984.

42. Ibid.; Jiang Yuchang, 1:58a-60a. Many themes mentioned in this discussion of hill-country banditry have been discussed at greater length in Billingsley 1988, which appeared after this article was completed. For an account of banditry elsewhere that also emphasizes connections between bandits and local elites, see Lewin 1979. [BACK]

41. Hobsbawm, chaps. 2, 6; Liu Xiaonong 1980; Zou Fuguang 1984.

42. Ibid.; Jiang Yuchang, 1:58a-60a. Many themes mentioned in this discussion of hill-country banditry have been discussed at greater length in Billingsley 1988, which appeared after this article was completed. For an account of banditry elsewhere that also emphasizes connections between bandits and local elites, see Lewin 1979. [BACK]

43. Hobsbawm, 35. [BACK]

44. Liu Xiaonong, 106-107; Zou Fuguang, 94. [BACK]

45. The following account of the revolutionary movement in Xunwu is based primarily on Zeng Biyi, 91-98; Bousfield, 109-124; Zhonggong Jiangxi shengwei dangshi yanjiushi, 183-185; Mao Zedong 1982, 67, 100, 127-129. [BACK]

46. Averill 1987, 284-290.

47. Ibid., 283-284; Zhonggong Ruijin xianwei, 6b; Li Jishan, 26:123; Yifeng xian difang zhi bianzuan wei, 8-9. [BACK]

46. Averill 1987, 284-290.

47. Ibid., 283-284; Zhonggong Ruijin xianwei, 6b; Li Jishan, 26:123; Yifeng xian difang zhi bianzuan wei, 8-9. [BACK]

48. For sources discussing these various issues, see Chen Qihan, 408-410; Huang Muxian 1981; Qiu Zhou, 77-79; Mao Zedong 1949a, 90; Guo Qi and Dong Xia 1982. [BACK]

49. Deng Zihui, 57; Xiao Hua, 239; Chen Qihan, 409; Qiu Zhou, 75. [BACK]

50. Huang Muxian 1981; He Changgong 1959; Chen Qihan, 408-410; Liu Xiaonong, 116-119. [BACK]

51. The general point that Communist cadres used their family status and lineage ties to further the movement is well established. E.g., Chen Qihan, 410; Fang Zhichun, 145-146. Although much more indirect, evidence of the use of small lineages against large ones is considerable. E.g., Mao Zedong 1949a, 90; Yan Yaxian 1959; Hu Yueyi, 90; Shao Shiping, 62. [BACK]

52. Mao Zedong 1982, 127-130. [BACK]

53. In Mao's listing of over 125 landlords from around the county, he explicitly identified seventeen people belonging to or "colluding with" the New Xunwu clique: two from upper-elite, fourteen from middle-elite, and one from lower-elite families. As of 1930, eight of the seventeen were also active members of the GMD. Mao also mentions names of eleven people who were members of or "linked with" ( jiehe ) the

Cooperative Society and four who were linked with the Zhongshan clique or the Zhongshan School: twelve of these fifteen men were from lower-elite and two from middle-elite family backgrounds. Mao also specified that nine of the eleven people explicitly connected with the Cooperative Society were from declining households, a fact that he found characteristic of all people in the society. The vast majority of these fifteen individuals ultimately became CCP members and/or active participants in the revolution, though a few later abandoned the movement. Mao Zedong 1982, 67, 115-124, 128-129. [BACK]

54. Data gathered on thirty-one pre-1927 Jiangxi CCP members whose biographies contain usable family-background information show that two-thirds of this group (all themselves well-educated individuals) were from families of local school teachers, small merchants, peasants, or artisans. The biographies indicate that some of these families (particularly the school teachers) represented declining branches of formerly more prosperous elites. Other Communists, especially those from peasant or artisan background, were only able to obtain the schooling that made them elites because of money borrowed from elite patrons, lineages, or distant relations. These figures are derived from research still in progress: the main sources are collections of biographies such as Zhonggong Jiangxi shengwei; Jiangxi sheng minzheng ting 1960, Jiangxi sheng minzheng ting 1980. [BACK]

55. See Averill 1987, 294-295. [BACK]

56. Somewhat exceptional in this regard are Yung-fa Chen 1986 and Galbiatti 1985. Neither, however, makes local elites a central part of his discussion. [BACK]

57. See, among other studies, Moore 1966, Skocpol 1979, and Goldstone, ed. 1986. [BACK]