Locality and Local Gazetteers in the Republic: A Case for the Continuity of Spatial Order

Abstract The way late imperial political elites in China positioned themselves in the tianxia—their life world—can be described as a balance between polity and locality, which was often accompanied by an enduring sense of local identity. This article argues that despite the fall of the tianxia concept in modern China, the age-old locality–polity relationship and the elite local identity did not disappear. Taking the flourishing local gazetteer production of the Republican era as a case, I suggest that instead of suppressing locality, the crisis of the polity and the coming of the nation-state in China brought it more to the foreground. The decline of locality in China's political culture occurred only after the communist takeover. The study makes use of the Local Gazetteer Research Tools (LoGaRT) developed by Max Planck Institute for the History of Sciences.

immediately before the downfall of the local, the local gazetteer was actually in full bloom during the Republican era.
As a genre that matured in the Song and flourished in the Ming and Qing, the gazetteer was deeply embedded in the social and political structure of the late empire: its production involved the imperial state's demand and guidance, local elites' active participation, and sometimes the local officials' initiative and organizing leadership; its key functions included information collection for the state, governing guidance for the local officials (who were invariably, due to the rule of avoidance, outsiders in the local community), and expression of local identity for the local elites. One is inclined to call it a quintessentially late imperial genre. Yet despite the fall of the imperial state, the ending of the rule of avoidance, and the decline of the traditional gentry elites, the production of local gazetteers surged in the Republic: with over 1,600 titles over a short period of thirty-eight years, its annual average was higher than it had been in the Qianlong  and Jiaqing (1796-1820) eras of the Qing, the widely acknowledged heyday of the genre. 3 Local elites' enthusiasm for compiling and the central government's efforts to administrate continued, and "local gazetteer studies" ( fangzhi xue) established itself as a clearly defined field of learning, with monographs in the publishing market and courses in the modern universities on the subject. Obviously, the late imperial genre renewed its life and vitality in the young republic.
The men who compiled the gazetteers and/or penned treatises on their nature in the Republican period clearly noticed that they lived in a new era, and they eagerly discussed-and adopted-new focuses of coverage and new categories of representation for the gazetteers that befit the modern times. Reading through the Republican gazetteers, one can hardly miss the narrators' nostalgia for the world they've just lost, often juxtaposed with the determination to adapt to the new one that has befallen them. The sensitivity to change and the urge to adapt, however, make it all the more compelling to ask: why did the Republican local elites and intellectuals bother with the compilation of the gazetteers and the exploration of the genre's nature at all? Weren't the localities-the raison d'être of the late imperial gazetteers-embedded in a world called "All-under-Heaven" (tianxia)? If that world had shattered, together with all its universalistic confidence and claims of centrality, what kind of locality was there in the Republic to continue providing the motivation and energy to engage the gazetteer?
The fate of locality in Republican China was no less odd. It has long since been declared in studies of nationalism in general that the dawn of nationalism ushered in p. 24), in which the then-Governor of Shandong province, Zhao Jianmin, was purged after having reportedly argued "I am a native of Shandong. I am for the people of Shandong and the cadres of the dusk of things local. 4 Studies of nationalism in China have rightly captured the prominence of "national unity" that featured both in national narratives where it "becomes complicit in the project of the nation-state" and "subordinates difference" 5 , and in nationalist politics where "local languages and customs, provincial warlords, and political federalism all appeared irremediably backward to centralists because they were all irrecoverably local." 6 However, against this chorus of its decline and oppression, the local in Republican China tagged along quite well with the national, its high profile "other" and alleged oppressor. Aside from the quiet but widespread celebrations of the local in the local gazetteers, there was the resilient thrust of local selfgovernance, endorsed by Sun Yat-sen even after his conversion to the party-state idea, and consistent calls for and reliance on localist institutions, such as tongxianghui (Native-Place Association) and huiguan (Native-Place Lodge) to serve the nationalist course both before and after the 1911 revolution. Even the cultural iconoclasts of the May Fourth/New Culture Movement largely spared the local in their denouncements of the old, targeting mostly filial piety and the patriarchal family system instead. The scope of the local in these debates and movements varied from the province to the county and even smaller, but the overall picture of locality's perseverance is clear enough to suggest that in Republican China, the local was not simply a force opposite to and conflicting with the nation-state.
This type of local-national relationship is understandable only if we view locality in the Republic as something inherited from the late empire: not as self-standing places to be annexed by the newly emerging territorial state, but as deeply rooted component parts of an enduring polity. The localist orientation of the literati tradition did contest the state's power and authority in defining local places, but the dynamics of localism always evolved within the parameter set by the imperial state. On balance, the late imperial locality was as much a part and parcel of the state system as it was a site where social elites contested the state power. 7 The dual Republican phenomena of locality and local gazetteer, both reminiscent of a by-gone age, prompt us to question how the late imperial locality experienced the modern metamorphosis that the country went through: did the same locality continue to exist in the new age? what changes did it undergo? Since "the local" always exists in relationship to some larger spatial entity such as "the national," an inquiry about the fate of locality in modern China is in fact an inquiry into the relationship between locality and polity. I use the term "spatial order" to describe the totality of relationships between various spatial entities such as the local, the polity, and the world. Locality's position in the polity, together with the polity's position in the tianxia world, were no doubt the two "pillars" of old regime China's spatial order. While the polity's imagined centrality in the world had completely evaporated by the Republican era, the inquiry here concerns the feasibility and durability of the other part of this traditional spatial order in modern times. Simply put: how did locality do after the fall of the tianxia? 8 The rest of this article is an effort to answer this question. Corresponding to the variation of scales in field administration, "locality" here refers to spatial entities originally demarcated by all levels of administratively by all levels of administrative units outside places outside the political center, i.e., the county, the prefecture, and the province. 9 Recent "localist turn" scholarship has made it clear that the late imperial locality had never been a matter of mere administrative entities solely defined by the state. Localist activities of the late imperial elites often inscribed different meanings to the local and turned it into social space of their own. 10 This balance between polity and locality imbued the latter with some irreducible dignity as the building blocks of the former. The enduring vitality of the local gazetteer in late imperial China, as a genre substantially (though not completely) "of the local, by the local, and for the local," was possible only against the background of this balance. In this sense the state of the gazetteer reflects the fate of locality, and vice versa. General studies of the gazetteer in both the Qing and the Republic abound, especially in Chinese-language scholarship, and need not to be reproduced here. However, a study of the correlated trajectories of locality and its gazetteers from the Qing to the Republic could shed new light on the modern changes of the spatial order in China. Following this approach and utilizing the "Local Gazetteer Research Tools" (LoGaRT) developed by the Max Planck Institute for the History of Science, as well as the more than 4,000 local gazetteers in its database (3,076 from the Qing, 776 from the Republic), I make a case for the continuation into the Republic of late imperial China's locality-polity relationship.
My argument is two-fold. First, the late imperial structure of the locality-polity relationship survived the political and social changes of twentieth-century China up until the communist takeover, with locality not only retaining its dignified status as the building blocks of the polity but also gaining more confidence as the site where the renewal 8 The polity, being dubbed the "central state" and constituting the majority of the known world to late imperial Chinese, was considered the core of the tianxia world and often interchangeable with it in common usage. This general perception remained into the late nineteenth century for the vast majority of people, despite earlier intellectual exchanges with the West since the late Ming. See Ge Zhaoguang 葛兆光, Zhongguo shixiangshi 中國思想史 (Shanghai: Fudan daxue chubanshe, 2009), 328-35. For a recent discussion of the tianxia concept, see Wang Ban, ed., Chinese Visions of World Order: Tianxia, Culture, and World Politics (Durham, NC: Duke University Press, 2017). 9 Subcounty level localism in both late imperial and Republican China, such as those crystalizing on the village and the township, certainly existed, as the studies of Mori Masao and others have made clear. But the articulation of sub-county-level local identity was not as conspicuous as those at higher levels. The vast majority of local gazetteers, for example, focus on the county level and above. Modern scholarship on imperial China's localist turn also focuses mostly on the county level and higher. It is for these reasons that subcounty level entities are not included in this discussion. See Mori Masao 森正夫, Konan Deruta Shichin kenkyū: Rekishigaku to Chirigaku kara no sekkin 江南デルタ市鎭研究：歴史学と地理学の接近 (Nagoya-shi: Nagoya Daigaku Shupaankai, 1992) for subcounty level local histories in the late imperial period. For a study on the modern times, See Satō Yoshifumi 佐藤仁史, Jindai Zhongguo de xiangtu yishi: Qing mo min chu Jiannan de defang jingying yu shehui 近代中國的鄉土意識：清末民初江南的 地方精英與社會 (Beijing: Beijing shifandaxue, 2017). of the polity should begin. If the millennium-old locality died in Mao's China, it was a sudden death, at the peak of its career, in 1949. Second, this continuation can be seen in the steady (though not linear) rise, between the heyday of the old regime in the late eighteenth century and the Republic in the early and mid-twentieth, of the Qing scholar Zhang Xuecheng's 章學誠 (1738-1801) thesis that local gazetteers should be fully accepted as "history" (shi 史) in the knowledge classification and production system, instead of "geography" as it had been conventionally held. The key sign that the gazetteer gained a generic status as "history," whose proper subject was the state in the Confucian tradition, is the dramatic increase in the number of the "chronicle" ( ji 紀) sections included in local gazetteers of the Republican era. Being an appellation intertwined with the monarch's sovereignty and historical narrative from the monarch's perspective in dynastic histories, the chronicle carried special ritual connotations the gazetteer had long been wary of; its open adoption in the Republic finalized the longterm change that made locality the subject of history.
The Local, its Gazetteer, and its Status The delicate balance between polity and locality came into being during the Tang-Song transition. In terms of political power, the center's domination over the local was unchallengeable. Institutional arrangements since the Song that meant to "strengthen the trunk and weaken the branches" turned county and prefecture officials into genuine agents of court authority throughout the realm. A Southern Song scholar-official described the central and local bureaucracies in a configurative way that "is similar to the way the body commands the arms and the arms command the fingers." 11 Outside the bureaucracy, the gentry elites, whose rise has mostly been attributed to the civil service examinations that penetrated deep into local societies, heavily depended on the symbolic rewards distributed by the court, and had neither the will nor the wherewithal to directly challenge the political authority of the center. 12 Thus the political center was strong in terms of its relations with both local governments and local societies. In terms of the realm's makeup, such a strong center meant indisputable political unity, whereby local places lost the power to stand as independent entities through political actions such as secession. They could exist only as component parts of the imperial whole, as a lower-level entity controlled from above, or, as it runs in Chinese, as "difang" (the local) vis-à-vis "zhongyang" (the center) or "chaoting" (the court). The local was defined as such by the center.
But the state was not the only agent that defined the local. Within the parameter set by the state and proceeding from it, actual life of people in local societies turned localities into communities through various activities and absorbed state-demarcations into 11 Songshi jishi benmo 宋史紀事本末, juan 2, "Shou bingquan" 收兵權. The power the court achieved through the early Song institutional re-arrangement often appeared overreaching in the eyes of contemporary civil officials, and often was critiqued for the effects of its over-centralization. For example of these critiques by well-known figures such as Zhu Xi 朱熹 and Su Zhe 蘇轍, see Zhu Xi, Zhu zi yu lei 朱子 語類 (electronic Sikuquashu edition), juan 128; and Song Shi 宋史 (electronic Siku quanshu edition), juan 339. their self-perceptions. Key to this process was the literati gentry, whose enduring localist orientation in family and career strategies-the orientation toward paying more attention to local affairs and building networks and influences in their home places-often generated such local activist projects as the construction of local lineages and local learning traditions. For individual places, recurring local activism gave full display to the gentry's agency and turned the local from the "marked identity" imposed by the state into "felt identity" of their own. For the realm as a whole, it rendered the local places increasingly distinct in time. 13 In this sense the local, though bound by the imperial framework and destined to form the building blocks of the empire, also had its own dignity, for it was never reducible to mere administrative tools.
This balance between the grandeur of the polity and the power of its imperial center on the one hand and the irreducibility of locality and the vitality of gentry's local identity on the other is vividly described by the philosopher Wang Yangming 王陽明 as "the All-under-Heaven [tianxia] is a collection of prefectures and counties." With minor variations in formulation, this phrase was commonly used in the Ming and Qing dynasties, often followed by a statement to the effect that as long as all these locallevel components are put in good order, the whole shall be in good shape. 14 What this phrase conveys is a view of the realm that, while paying homage to the grandeur of the All-under-Heaven, put the local prior to it, and honored the local as its foundation. Corresponding to this view is the weight carried by the local as social space. Scholar-officials, including those with national reputations and unusual success, usually started their career from the home locale (as a local candidate of the civil service examination system) and returned to the same place for retirement and burial after death. Ritual, social, and intellectual activities that evolved in the local space, such as learning and lineage building, were often assigned universalistic meanings. 15 The mature local gazetteer, as a genre about this specific kind of locality in late imperial China, was born out of this balance and served to facilitate its renewal. The timing of its emergence, during the Southern Song, should not be a surprise if we consider that the same period also witnessed the rise of the new gentry elites and the latter's localist orientation. 16 The genre's growth in late imperial China was steady: it gained relatively 13 Bol, "The Rise of Local History." The phrases "marked identity" and "felt identity" were used by David Howell, Geographies of Identity in Nineteenth-Century Japan (Berkeley: University of California Press, 2005), 10-13. 14 Wang Yangming 王陽明, "Gaoping xian zhi xu" 高平縣志序, in Wang wencheng quanshu 王文成全 書 (Electronic Siku quanshu edition), juan 29. Wang's phrase in the original text reads "天下之大州县之 积." Other formulations include "tianxia zida, zouxian zi he" 天下之大州縣之和 or "tianxia zida, yixian zhiji" 天下之大一縣之積.

15
This can be confirmed by the biographies of people ranging from Wang Yangming in the Ming to Zeng Guofan in the Qing. The Neo-Confucian movement actually involved a view that took the local as the beginning place for their moral and social reforms of the "All-under-Heaven." See Robert Hymes and Conrad Schirokauer, eds., Ordering the World: Approaches to State and Society in Sung Dynasty China (Berkeley: University of California Press, 1993). See also Peter Bol, New Confucianism in History (Cambridge, MA: Harvard Asia Center, 2008). 16 The local gazetteer as a genre could be traced back to the early imperial era, but its intended audience, authorship, and content all went through substantial transformations during the Southern Song. For a discussion on this transformation, see James M. Hargett, "Song Dynasty Local Gazetteers and Their Place in the History of Difangzhi Writing," Harvard Journal of Asiatic Studies 56.2 (1996), 405-42. The correlation between this newly transformed genre and the initial emergence of the literati localist orientation during the Song dynasty is vividly illustrated by the preface to the Xin'an zhi 新安志 (pub. 1175), the earliest extant gazetteer of the Huizhou 徽州 prefecture, written by the Huizhou native Luo Yuan, who also single-stable scope and format-as a mix of history, geography, literature, and government records-in the Song; saw the beginning of the central government's request for local gazetteers and guidance on the "rules of compilation" in the Yuan; and enjoyed a substantial increase in the number of gazetteer projects in the Ming (with over a thousand extant titles in contrast to the dozens during the Song and Yuan). In the Qing, a sophisticated gazetteer production system was fully established. It included the repeated promotion and detailed guidance from the imperial government, much keener attention by local officials of various levels, the convention of periodical (i.e., every 60 years) renewal of the gazetteer, and the participation in the compilation projects by some first rate scholars. By the end of the eighteenth century, the local gazetteer had covered almost all local administrative entities. Among the 8,500 pre-1949 gazetteer titles catalogued in the China Local Gazetteer Union Catalogue, about 5,700 are from the Qing. 17 Bibliographers in late imperial China often catalogued the gazetteer under the category of "geography," which itself was usually a subcategory under "history." This treatment probably can be explained by a stage in the local gazetteer's pre-Song development, when it was called "tujing" (annotated map) and functioned as a geographical handbook on specific places. Occasionally, Ming and early Qing private bibliographers did catalogue the gazetteer as a separate subcategory under "history" but parallel to "geography," or even as an independent category parallel to "history," but the status as a subsection beneath the "geography" subcategory (shibu dililei 史部地理類) remained the standard view, and was adopted as such in the siku quanshu project in the 1770s. 18 Since "geography" was usually regarded as subordinate to "history" in the traditional classification of knowledge, this rendered the gazetteer lower in status, particularly in comparison with the official dynastic history (zhengshi 正史). 19 Every now and then, there were passing remarks comparing the gazetteer to the official dynastic history. These could come from famous scholars such as Wang Shizhen in the late Ming, or powerful figures such as the Yongzhen emperor in the Qing. 20 These handedly compiled the gazetteer. Luo was aware that his work differed from previous writings about the place, which he viewed as no more than administrative data compilations made by clerks serving as manuals for government officials. On the contrary, his work was a genre with profound meanings that could only be understood by the learned, and worthy of the commitment of a real gentleman. See Xin'an zhi, juan 1. Notable examples in the unorthodox cataloguing practices include Zhu Mujie 朱睦㮮, Wangjuantang yiwenzi 萬卷堂藝文志, which set a "gazetteer" category under "history", and Zhang Xuan 張萱, Neige cangshu mulu 內閣藏書目錄, which made the gazetteer a separate category, called zhicheng, in parallel to the usual four categories of "jing," "shi," "zhi," and "ji."

19
On a recent discussion on the relative status of the genres "history" and "geography," see Ba Zhaoxiang 巴兆祥, Fangzhi xue xinlun 方志學新論 (Shanghai: Xuelin, 2004), 28. The most authoritative statement on this is probably the siku quanshu project's "general statement on [the category of]'history'" which states that "the dynastic history" is the "main framework" (dagang) of history, while "geography," together with other subcategories such as "astrology" and "governance manuals," are "derived from and comparable to the treatises [inside the dynastic history]." See Siku quanshu zhongmu tiyao 四庫全書總目提要, vol. 45, "shibu zongxu" 史部總序. exceptions and variations notwithstanding, the issue of the gazetteer's generic nature was largely absent from serious scholarly discussions despite the growth in its production. It looks as though with regard to the gazetteer, as long as the practical needs of each party-the imperial government, the local elites, and the local officials-were fulfilled, people were willing to leave aside such scholastic issues as the gazetteer's nature and status.
The first serious intervention into this quiet and uncontroversial scene occurred in the late eighteenth century, when Zhang Xuecheng, the posthumously recognized master of historical theory, launched a set of bold proposals and powerful arguments over the nature of the local gazetteer, the proper arrangement of its content, and the institutional reforms to facilitate its production. The essays that best articulated these arguments were written by Zhang in the late 1780s and early 1790s, but the basic ideas had been with him for decades, spanning his career as a talented theorist of history and a devoted compiler of local gazetteers.
The core of Zhang's intervention is the motto that essentially, as a genre, "the local gazetteer is history" (zhi nai shi ti 志乃史體) 21 . Along this line he made two specific arguments: first, the gazetteer originated from the "total history of the ancient feudal states" ( yiguo zhi quanshi), and should function just like that for the contemporary prefectures and counties even though the hereditary feudal lords had gone, and local places had lost the political status as "states" (guo), because in terms of submitting their local documents to the court of the Son-of-Heaven there is no difference between the prefectures/counties and the feudal states; second, the production of gazetteers should be part of the production of dynastic history, in the senses that the compiling of the gazetteer prepares materials for the compiling of the dynastic history, and that the existence of detailed information in the local gazetteers makes it possible for the dynastic history to be succinct and focused. 22 Obviously, the target of such an argument was the prevalent view of the time that took the gazetteer as merely a handbook of local information and catalogued it under "geography"; and the purpose of it was to mark the gazetteer as a normal form and integral part of (dynastic) history.
In ancient China, history writing was both official and moralistic. All historical documents (that is, all written materials) were handled by state functionaries (scribes), and the writing of history was a state function that provided a narrative of the state's own. The official historian was supposed to help set the political order of the state through condemning (the evil) and praising (the good). Thus there was an inherent solemnity in historical narrative that was proper only for the state. After the founding of the empire, since there was only one dynastic state, the genre "history" became part of the trappings of the latter: only the narrative of the state deserves the appellation "history," to the effect that the two words "state" and "history" were often tied together as "state history"(guoshi). The appellation "gazetteer" (zhi), as the writing of and about one local place, therefore denotes a difference from "history" (shi), the writing of and about the state, just like the writing of and about a clan had its own distinctive appellation, "genealogy ( pu)." In this sense, the common saying in late imperial China, "the state has its history, the place has its gazetteer, and the clan has its genealogy" (guo you shi, fang you zhi, jia you pu), marks the difference between these three types of writings. 21 Zhang Xuecheng 章學誠, "Yu zhen xiucai lun xiuzhi di yi shu" 與甄秀才論修志第一書, in Wenshi tongyi 文史通義 (Shanghai: Shanghai Shudian 1988), juan 8: 1-4. 22 Zhang Xuecheng, "Fangzhi li sanshu yi" 方志立三書議, in Wenshi tongyi, juan 6: 1-7, in Zhang Xuecheng, Wenshi tongyi, (Shanghai guji: 1988).
Against this background, Zhang's argument was an emphasis on the similarity and connectedness between the history and the gazetteer genres, and hence granted the latter not only a higher status but also a stronger sense of stately solemnity.

Zhang Xuecheng's Intervention in Historical Perspective
As a theorist of history, Zhang fully identified with the statist approach to "history," and actually made a name for himself in that way. He viewed history in ancient times as the outgrowth of the state's records of its own activities, and highly appreciated the situation when there was no private writing at all, and the state monopolized not only the genre of "history" but also the enterprise of history-writing. The complaint that drove his intervention into the gazetteer issue was not that the relation between "history" and the state was too tight, but rather that between the two of them and the gazetteer was too loose. To understand the significance of this intervention, one has to view it against the social and intellectual context of the late eighteenth century when Zhang lived and wrote: the swelling of the ranks of the literati and the widespread sense of frustration in career prospects befalling most of them, the intellectual predominance of the evidential scholarship that featured highly technical yet somehow fragmented textual analyses, and the normalization of local gazetteer production throughout the empire that provided employment opportunities for less prosperous scholars and fueled the specialization of the field.
Zhang was convinced of his talent as a historian from an early age, and he was critical of the practice of history-writing prevalent in his day, which, according to him, had become no more than mechanical piling of materials in the bureaucratized dynastic-history compiling offices. For him, a true historian should be an individual genius able to grasp the significance of materials and bring it out through proper selection and arrangement. Meanwhile, he insisted on a historical approach to the Confucian classics, which held that the classics are not mystical and abstruse texts, but historical records of the historically formed institutions and policies of the ancient sages, who were themselves coping with actual problems in historically specific situations. Both approaches made him a critic of the dominant evidential scholarship, and hence a strident voice of his time.
For Zhang the marginalized scholar and dedicated statist historian, the local gazetteer, as a lesser form of historical writing, was the only chance he (and scholars like him) had to try his hand. 23 While a student in the guozijian, the young Zhang, then in his twenties, once stated that "even if a great man does not become an official historian, he should still serve the people of power with his pen, make a name for himself, and contribute to the contemporary time; compiling local gazetteers is one of these acceptable engagements." 24 Indeed, throughout his career, although Zhang made insightful remarks on history writing at the general/state level, he personally practiced it only at the local level. No evidence suggests that Zhang's theoretical thinking was determined by the limitations to his own space of career development, but to argue that the gazetteer be treated like history proper certainly made his works more meaningful. Even if he was to a certain degree compelled to promote the status of the 23 This was an observation shared by both the Chinese in the late Qing and Republican periods who admired him, and foreign scholars who were able to afford a distance. See David Nivison, The Life and Thought of Chang Hsueh-ch'eng (Stanford: Stanford University Press, 1966), 218. 24 Zhang Xuecheng, "Da zhen xiucai lun xiu zhi di yi shu" 答甄秀才論修志第一書, in Zhao Gengqi, Xiuzhi wenxian xuanji, 25. gazetteer, the seriousness and effectiveness with which Zhang argued for his case was unprecedented, as illustrated by his well-known debate with Dai Zhen 戴震, the most famous scholar of the late Qianlong era. 25 As a result, Zhang and his "gazetteer-is-history" thesis became the single most recognizable in the emerging field of gazetteer compilation.
In addition to this core argument on the gazetteer's status, Zhang made bold proposals about the gazetteer's production. One was the three-way division of the gazetteer, with one part providing historical narrative as succinct and formal as the dynastic history, and the other two devoted to documentation and details. Another called for the establishment of a special "office of the gazetteer" in prefectural and county local governments, to collect and preserve documents from the local government and local society, and to participate in the preparation for the dynastic history of the whole realm. With these proposals, not only would the gazetteer's format more closely resemble the official history, but also its production process would further integrate into that of the latter and constitute a bridge between the imperial state and local societies.
In general, therefore, the event of Zhang's intervention can be viewed as the birth cry of the gazetteer as a self-conscious field, announcing its arrival and claiming its seat in the existing regime of history-writing. 26 His argument revolved strictly around the issues of the genre instead of the locality itself. Yet in building up this argument, Zhang frequently referred to the locality-polity relationship, both as the foundation of and as an analog to the gazetteer-history relationship. For example, justification of the resemblance between the gazetteer and dynastic history came from the parallel between structures of local and central governments: Although the localities are small, they have all the six branches of government similar to the central government-those of "personnel," "taxation," "rites," "military," "justice" and "construction"-to carry out tasks they receive from above. This is the so-called "same structure in smaller size 具体而微" [emphasis mine]. The dynastic history shall draw its materials from these local offices, just like the Spring and Autumn drew from the documents of the feudal states. 27 The reason for establishing special offices for the gazetteer in local government was the parallel between governance and history writing: Since the major issues of governance in "All-under-Heaven" start from the prefectures and counties, the assigning of responsibilities for history writing should also start from the gazetteers of the prefectures and counties. 28 With the phrase "same structure in smaller size," Zhang drove home the notion that locality is the microcosm of the polity. His review of locality as the site where governance of the polity starts reveals exactly the locality-polity balance mentioned above: the dignified locality, though building blocks of the polity, is also its foundation and 25 Zhang Xuecheng, "Ji yu Dai Dongyuan lun xiuzhi," in Wenshi tongyi, juan 8: 22-26. 26 Indeed for the Republican scholars such as Liang Qichao, gazetteer studies came into being with Zhang Xuecheng. See Liang, "Longyou xianzhi" 龍游縣誌序, in Xiuzhi wenxian xuanji, 80-84; and Zhongguo jin sanbainian xueshu shi 中國近三百年學術史 (Beijing: Dongfang, 2004), 324-39. 27 Zhang Xuecheng, "Fangzhi li sanshu yi", juan 6: 3. 28 Zhang Xuecheng 章学诚, "Zhouxian qing li zhike yi," 州县请立志科议 juan 6: 9. prior to it. His argument on the gazetteer follows the logic that it deserves a "history" level status because its subject, the locality, functions just like history's subject, the state. In the proposals, the status of locality was a non-issue; it was used as evidence rather than pitched as the point of argument. Yet this very lack of focal attention, and the unquestioned deployment of it as evidence, makes it even clearer that for Zhang (and his audience the Qing literati) this mode of locality-polity relationship was deeply ingrained in their imagination of the spatial order. Zhang did not highlight the locality's status in his reasoning because he did not expect this to be an issue of controversy at all. And he was right on this. Few people took issue with Zhang's evidence. In the heyday of the Qing, the balance between locality and the polity was taken for granted.
Zhang's argument on the gazetteer's nature and status, "gazetteer is history," rarely encountered open challenge either. Conscious adoption of the motto increased remarkably in the Qianlong and Jianqing period discussions over the nature of the local gazetteer (see Figure 1 29 ). As a private scholar, Zhang's name was referred to sixty times in  Erudition's Zhongguo Fangzhi ku, searched with LoGaRT. The dotted line refers to the frequency of chronical sections, the solid line refers to the frequency of reference to the notion that "gazetteer is history" in its various phrasings. The frequency for each period refers to the number of counts divided by the number of gazetteers in the database. Conventional reign periods are used as the base, but reign periods with less than 100 sets of gazetteers were combined with neighboring reign periods following historiographical conventions, e.g., Xianfeng and Tongzhi as one period, Guanxu and Xuantong as one period, etc. On searching methods: Zhang's original phrase for the statement "Gazetteer is history" is "志為史體." Keyword search for this statement is extended to a set of phrases including "志為史," "志屬史," "志史體," "志史也," "志者 史." The search was limited to sections with titles containing the following words: "序," "跋," "目例," "大 事," "編年," "事紀," "郡紀." Sections of藝文志 are not included because they could include many references to these phrases but had nothing to do with the discussion of the gazetteer's nature as a genre. The search for chronicle sections used the keywords set that includes: "大事," "編年," "郡紀," "縣紀," "事紀," the database's near 1,400 sets of gazetteers compiled between the Jiaqing reign when he died, and the end of the Qing dynasty. So, in general, his intervention into the practice of the local gazetteer was effective. But as an event, its impact was limited to the relatively small and eventless gazetteer circle. In the beginning of the twentieth century, Zhang's historical approach to the classics was rediscovered and earned him the reputation of one of the greatest thinkers of the Qing dynasty, who held a rather "modern" outlook. In the light of this delayed vindication, his gazetteer intervention appeared an insignificant episode in a field of lower stakes, and the fact that his reputation had been limited therein was regarded as an unfair negligence befalling a genius. 30 Yet this very lack of attention to Zhang's intervention into the gazetteer, just like Zhang's own lack of anxiety over using the status of locality as his evidence, more effectively illuminates a salient feature of the spatial order in late imperial China: to grant gazetteers the status of "history" is, like citing the polity-locality balance, to recognize something obvious.
Zhang's comparison of contemporary prefectures and counties to the pre-Qin fengjian states calls to mind Gu Yanwu's proposal to combine the fengjian spirit with junxian reality in the early Qing. 31 While Gu sought to solve the problems of local administration inherent in the centralized empire, Zhang addressed a problem of the local's representation in the same centralized empire. Both problems were peculiar to a mature junxian system, and both, while fully accepting unity of the whole as given, tried to secure what the parts duly deserve. Gu's certainly had a broader resonance, but Zhang's apparently moved ahead more smoothly. If the dignified status of locality was the foundation for the vitality of the gazetteer, here the genre's growth and birth as a field was requesting a "rectification of name" for its subject. In practice, however, implementing the motto "gazetteer is history" is much more complicated. Zhang's major proposals on the gazetteer's production-the three-way division of the gazetteer and the establishment of "gazetteer office"-involved too drastic a change of convention, and too-complicated institutional reform. None got even close to realization. It was rather one of his other proposals, requiring neither structural change of the gazetteer nor institutional remodeling of local governments, yet addressing more directly the gazetteer's "history" status, that carried the torch forward. This is his promotion of a special section-the section of the chronicle-to be included in the gazetteer.

Rituals and Politics of the Chronicles
A chronicle section in the gazetteer is almost a logical corollary if one agrees that "gazetteer is history," for it allows the gazetteer to give a historical narrative of the local place that evolves in time, and it helps to organize the rest of the gazetteer contents more clearly. In Zhang's own words, "the main task of historiography is to record events, "事記," "故事志," "紀事志," "前事志," "史事志," "舊事志." Basically the chronicle had been a highly unregulated field in both the late empire and the Republic. For those literati concerned with ritual problems, the concern reached not only the word "ji" but also the very idea of having a local chronicle. Similarly in the Republic, when people decided to install chronicles in their local gazetteers, they used various names despite the ordinance of the central government. In this circumstance, the section search of chronicles follows a maximalist principle and includes all known variants of the phrase "dashiji." the main way of recording events is through chronicles (lit., to braid events along years)." 32 Although there are other obviously more "historical" elements in the gazetteer (e.g., biographies), there is none as powerful and direct in making it "history" as the chronicle. In the light of this, the usual absence of chronicles in the gazetteer is an obvious problem. Ming and Qing gazetteer writers long before Zhang had sourly felt the troubles it caused. Lin Dachun, the late Ming scholar-official and gazetteer writer from Guangdong, had lamented that in ancient times all feudal states had their histories to help order their affairs … nowadays the gazetteer is not regarded as history, yet still the counties descend from ancient feudal states. Histories record things clearly because they are equipped with chronicles, but gazetteers of the counties do not have their chronicles, they are left messy and never put in order. For this reason there is no good history-writing enterprises in our times, and the state rarely has worthy histories. 33 Zhang himself complained about the fragmented narrative in gazetteers where, without a chronicle, major events scattered in different sections such as those on natural disasters and military affairs. Therefore, in his last major gazetteer project, the gazetteer of Hubei Province, he weaved together all major events of the province into an integrated chronicle (biannian ji), and put it upfront in the book's opening chapter so that readers can "know things of the past and the present as they know their hands and fingers." 34 But there was a ritual problem with the chronicle as a genre in history writing. The earliest chronicle in China, Confucius' Spring and Autumn Annals, which takes the reign periods of the Lu state as the temporal axis to construct historical narrative, had already exemplified history's function of upholding the social/moral order through condemning and praising. Shortly after Confucius, some texts started to apply to the chronicle the appellation "ji" 紀, a word that, according to the Tang dynasty historical theorist Liu Zhiji 劉知幾, denotes "ordering normal hierarchies, controlling various things, and conveying unsurpassable grandness to the title of a written piece." After the coming of the empire, it was Sima Qian who first tied the "ji" chronicle with the universal sovereign and called it "root chronicle" (benji 本紀). The subject of this type of chronicle is the sovereign, and the content of it the major events of the universal state from the sovereign's perspective. From here comes the well-known term "sovereign's root chronicle [diwang benji]" in historiographies throughout the imperial era. By the time of Liu Zhiji, the term "ji," the genres chronicle and history, and the notion of state sovereignty had become inseparable; as Liu wrote, "the essence of ji, like that of the Classic of Spring and Autumn, is to create order of time by weaving together days and months, and mark the orthodoxy of the state by recording events related to the sovereign." 35 Against this background, a chronicle in the gazetteer immediately makes it more stately, but it also carries the air of ritual transgression. Chronological recording of some issues in the gazetteer, such as natural disasters (zaiyi 災異), was inevitable. Concerns of ritual appropriateness made many Qing gazetteers put zaiyi recording under such categories as "miscellanies" or "geo-astrological fields" instead of "chronicle." 36 Another such issue, a locality's administrative evolvement ( yange 沿革)was often understood as more about the state than about the local place. Many gazetteers followed the example of the Song Jingding era Jiankang zhi (1261) to put recordings of natural disaster, administrative change, and local events all in one chronological table. But in Ming and Qing gazetteers one often sees complaints that, except for the yange, too little is known about a locality's past. 37 To those who felt strongly the need for a chronological recording of local events, anxiety was common. For example, the late Ming Zhaoqing prefectural gazetteer stated that "compared with the state, the prefecture and county is like a small pond compared with a big lake; for a county or a prefecture to weave its events into a ji chronicle is a risk of transgression." 38 The Kangxi era Guanyang County gazetteer, after asserting the ji chronicle as the trademark of history, admitted that "the history and the classics are all books for the Son-of-Heaven, not what a county gazetteer can aspire to." 39 People could find excuses to drive through the anxiety, or could take some special maneuver to avoid it. Lin Dachun in the late Ming cited his profound concern that, without a chronicle, the local past of Chaoyang would be lost and the situation would further deteriorate. The Zhaoqing gazetteer adopted a justification that is close to sophistry: was not the Spring and Autumn, which ordered the whole realm, written by a man from the Lu state (i.e., from the perspective of Lu, instead of the "Son of Heaven")? The tone of the Guanyang gazetteer was more humble but no less sophistic: "yet the county is the Son-of-Heaven's county, the people are the Son-of-heaven's people, all within the four seas are now one family and go through good and bad times together, who can say the county gazetteer is not also the Son-of-Heaven's book?" Zhang Xuecheng's own solution was to include "ji" but drop the word "ben." His analysis of this problem vividly reflects the position of a statist historian promoting the status of the local gazetteer: The inclusion of ji in history … has been established like a strict law. Nowadays the violation of this law and the lack of ji in local gazetteers is mostly due to the confusion caused by the classification of it under "geography." This is like shaking a cloth without holding its collar, or providing details yet without clarifying the main outline … It is appropriate for the local gazetteer to include a ji and hence give the whole book a warp, though it would be confusing if the gazetteer decorates its ji with the word ben, for that will deviate from the principle of honoring the state history. … All in all, while the ji in the gazetteer is the warp vis-à-vis other contents of the same gazetteer, it becomes the weft vis-à-vis the "benji" of the state history. Frequent discussions on the problem make it clear that that the concern was real. On the other hand, this "law" of history-writing that Zhang referred to was a soft "law," visible to the gazetteer compilers but not enforced by any state agency. Some simply adopted a chronicle for the home locale without saying anything. 41 For those who articulated their concerns, the very moment they did so was often also the moment they were already determined to maneuver through it. To the local gazetteer, picking up the chronicle section was like performing its "history" status in action, bold yet without risk; to the state, this type of provocation was like kids crying for attention, carrying no political threat at all. (With the junxian's embeddedness in the tianxia and the center's political domination both beyond question, who would be scared by a pretentious appellation inside the gazetteer of an innocuous local place?) The absence of any serious arguments against this type of transgression indicates that the literati in general were ready to allow localities to partake in the grandiosity of the empire and to have their own narratives, just as they were ready to grant local gazetteers the status of "history." Zhang Xuecheng's intervention occurred at just the right time: in the heyday of China's ancient regime, when everyone felt comfortable with the age-old locality-polity balance in a centralized empire, and the gazetteer production was flourishing everywhere. All he needed to do was to launch his arguments; the rest would ensue like ripened fruits falling off the tree. 42 As a testament to Zhang's success, nineteenth-century local gazetteers largely stopped apologizing for their chronicles. On some occasions, the chronicles were dubbed "biannian" (lit. braiding of years) instead of the catchier "ji." But in either case gazetteer writers simply took it for granted that what they were doing was "history." On the other hand, however, the percentage of gazetteers with a chronicle section did not increase after the Jiajing reign (see Figure 1). This stagnancy attests to the other side of Zhang's easy success: the issue of local gazetteers' status was, like that of the status of the locality itself, an issue of low stakes in a largely uncontroversial field. People could agree with Zhang and make some change to their gazetteers; but they could also neglect him and keep on doing gazetteers the old way. When the stakes were low, so was motivation, and simple inertia could easily carry the day. It was only during the Republican period that this scenario of quietness was replaced by one of excitement.

The Dual Thriving of the Local Gazetteer and its Chronicle in Republican China
The conventional view that takes the Qianlong and Jiaqing eras as the heyday of the local gazetteer must have been caused, at least in part, by the fact that the most influential modern studies of the genre's history, by scholars ranging from generalists such as Liang Qichao to specialists such as Fu Zhenlun and Li Taifen, were conducted during the 1920s and early 1930s. By then the real flourishing period of the local gazetteer had not yet finished, and its historians probably did not take into consideration their own time period. In hindsight, though, it appears that the heyday for the studies of the 41 See Fengrun xianzhi 豐潤縣志 (1570), Hangzhou fuzhi 杭州府志 (1686), and Dengfeng xianzhi 登封 縣志 (1787). 42 The famous debate over the gazetteer between him and Dai Zhen was focused on technical issues such as which section of the gazetteer should carry more weight. The other prominent figure in the gazetteer field during the Qianlong-Jianqing era, Hong Liangji 洪亮吉, who was often regarded as the champion of the evidential scholarship in the gazetteer field, also featured a chronicle in his signature project, the Dengfeng county gazetteer of 1787. gazetteer was also that of its production. Statistics made at later dates indicate that about 40 percent of the Republican gazetteers were produced during the short period of political stability in 1927-1937, with the year 1936 alone there were 80 titles. If we confine the survey to the twenty-five years before the outbreak of total war in 1937, the annual average of over forty titles is far higher than any period in the Qing, even if counting in the estimated number of non-extant titles from the earlier age (see Table 1). 43 Meanwhile, the special field of local gazetteer studies that was just emerging at the turn of the nineteenth century became fully established in the early twentieth. Studies of the gazetteer were published in geographical journals such as the Yu Gong (The Tribute of Yu), as well as some more specialized journals such as the Fangzhi Yuekan (Gazetteer Monthly). An incomplete list counted close to 400 journal articles on this subject published between the 1920s and mid-1940s. 44 In addition, there were about a dozen monographs on this subject published before 1949.
The vast majority of the Republican gazetteers' coverage were at the county level. This matches the change in field administration, since the new republic abolished the prefecture as a territorial entity. Considering that about 10 percent of all gazetteers between the Yuan and the Qing recorded in the Union Catalogue are at the prefecture level, county level gazetteer production must have benefited from the removal of the prefectural entity per se. The province entered the Republic largely intact, so did the production of provincial gazetteers. The two decades from 1927-1949 alone witnessed more than fifty provincial gazetteers (many unpublished). Since there were less than thirty province-level entities in the Republic, this should be considered a quite high number. 45 The social and institutional background of Republican gazetteer production was similar to that of the late imperial past. The central government, despite its own weakness, made its first request for gazetteers of localities in 1917. After the restoration of political unity through the Nationalist Revolution, the Nanjing government renewed the request in December 1928. Following that, in December 1929, the Executive Yuan issued an ordinance on gazetteer compilation submitted to it by the Ministry of Interior, the "Essentials of Organization and Style for Gazetteer Compilation" (xiuzhi shili gaiyao), to be applied to all provinces, counties, and municipalities. An ordinance with similar content was issued in 1944, under the title "Methods for the Compilation of the Local Gazetteer." 46 At the local level, the enterprise to compile and publish the gazetteer still depended on the joint efforts of the local officials and local elites, as they had in the past. Overall political instability of the time could make the turnover of local officials Table 1. Local Gazetteer Production in the Republican Era. Source: Fu Dengzhou, "Minguo shiqi fangzhi xiuzhuan shulue". Note that Fu's estimate of the total is a modest 1,400 Time period 1912-1926 1927-1937 1937-1945 1945-1949 Annual average 32 56 27 26 43 The bibliographer Zhu Shijia 朱士嘉 estimated that the ratio of extant and lost gazetteer titles in the Qing is 1:1. On the number of gazetteer titles in the different phases of the Republic, see Fu Dengzhou 傅登 舟, "Minguo shiqi fangzhi xiuzhuan shulue" 民國時期方志修撰述略, in Wenxian, 1989.4: 114-54. 44 "Zhongguo difangzhi lunwen suoyin (1911-1949)" 中國地方誌論文索引 (1911-1949), in Zhongguo difangzhi lunji, 191-229. 45 Ba Zaoxiang, Fangzhi xue xinlun, 168-94. 46 Zhao Gengqi, Xiuzhi wenxian xuanji, 125-26, 161-62. Yongtao Du high and the compiling/publishing process unpredictable, but the commitment of local elites often made up the difference. 47 Almost all the widely acclaimed Republican gazetteers had stories of extraordinary devotion of the local elite behind them. Some of the projects could last an unusually long time. The Chuansha County gazetteer (pub. 1936), with Huang Yanpei 黃炎培 as chief editor and Lu Binglin 陸炳麟 as the head of the gazetteer bureau, took the relatively stable team twenty-one years of persistent work. The Anyang county gazetteer project started in 1921, went through four local gentlemen (shen) as the heads of the gazetteer bureau, the destruction of the first draft in a conflict between warlords, a renewed initiative by the local elites (shishen 士绅) and approval by the county magistrate's executive council to restart in 1931, and was eventually finished in 1933. 48 The passage of time did have its effects. The makeup of the local elite itself, for example, became more complicated, and motivations for participating in the gazetteer projects varied broadly. Dedicated Qing loyalists supported gazetteer projects because they interpreted local agency as resistance to the new regime, and writing local history was one of the few public engagements compatible with their mood as the "leftovers" from the previous dynasty. Chen Botao 陳伯陶, who had served as literary advisor to the Guanxu emperor and lived the life of a recluse after the fall of the Qing, agreed to join the gazetteer project in his home county of Dongguan only after he was assured that it was supported by funds of the local community and "there is no official appointment involved hence no suspicion of serving the new regime." 49 In the loyalist cases, as in Zhou Zuoren's more famous promotion of local literature in resistance to the mainstream May 4th discourse, the local carried a link to the past and an alternative to the unacceptable present evolving on the national stage. The strategy of resorting to the local in times of stress and frustration resembled the localist strategies of the late imperial literati, though the forces from which they sought shelter were no longer those of a new dynasty. 50 For those more open to republicanism, engagement in local gazetteer projects became a natural channel of political participation and was pursued with remarkable enthusiasm. Members of the literary organization South Society (nanshe), for example, initiated numerous local gazetteer projects across the country. 51 In most cases, however, the gazetteer projects saw collaboration of people with diverse educational backgrounds and cultural outlooks. In the Chuansha case mentioned above, for example, Lu Binglin held an examination degree from the Qing and was a member of the local assembly in the Republic, while Huang Yanpei combined old and new style education and had overseas revolutionary experience as a member of the Tongmenghui. In general, the scene of 47 Two thirds of the Republican county magistrates served a tenure of less than one year. See Wang Qisheng 王奇生, Geming yu fan geming 革命與反革命 (Beijing: Shehui kexue wenxian, 2020), 364. the local gazetteers' production suggests that the gazetteer of the home locality was a commitment that people across a wide political and cultural spectrum could agree on, and the old pattern of its production was able to absorb the changes of time and move on.
The same dialectics of change and continuity can be seen in the contents of Republican gazetteers. Calls for innovation in the gazetteer were common in the monographs and the articles, with the rationale that the new era provided new technologies that should be harnessed, and new needs that required new categories. 52 Innovations were indeed impressive. New geographical, cartographical, photographical, and printing technologies were widely applied. For example, modern to-scale map and stereotype printing were common among Republican gazetteers. Meanwhile, introduction of new social and political theories, particularly the rise of "New History" (xin shixue), spearheaded by Liang Qichao and ushering in "the national people" (renqun) as historical subject, brought a broader outlook and much keener attention to common people's livelihood. Hundreds of new categories entered the Republican gazetteers, featuring most prominently new terms in economic life, new educational institutions, and scientific concepts 53 (see Table 2 54 ). Amidst these ardent calls for and effective engagement with innovation, however, the most frequently used categories in the Republican gazetteer were still those From Erudition's Zhongguo Fangzhi ku, searched with LoGaRT. Note that these new categories mostly reflect the fresh and keen attention to "people's livelihood" in both history and gazetteer writing of the Republican era. "Dashiji" is not included in his table because, strictly speaking, it was not a new category that had been in use for centuries, such as "relics," "administrative evolution," and "chaste women" (see Table 3 55 ). The mixture of the old and new in content makes sense if we consider that the purposes of the gazetteers, often stated in the gazetteer prefaces by the prominent personages involved in the production, remained largely the same as during the Qing, i.e., to facilitate local governance, promote local public goods, and keep local memory. The innovations hence helped with more accurate recording of the changing local reality and better fulfilled the gazetteers' function. Here, too, change took the form of further growth on the same track, and the new was embraced without renouncing the old.
It is in this ethos that we find the dramatic increase of chronicles in the Republican gazetteers. Of the total 459 chronicles from the whole database, 202 were from the Republican era. In terms of both number and proportion, it is much higher than any previous period (see Figure 1). One obvious cause of this surge is the ordinance of 1929, which required that all local gazetteers provide a "chronicle on major events (dashiji)." Since this was the first official request for a chronicle in the gazetteer, the relevant clauses of the ordinance are quoted here:  From Erudition's Zhongguo Fangzhi ku, searched with LoGaRT. Note that "dashiji" is not included in this table because it does not have a uniform title. Should we include it here, the 202 counts would make it to the top 10. But it will still be smaller than some other, more conventional, categories such as "relics". The moderate "dashiji" counts reveal the enforcement issue of the ordinances of 1929: just like the taboo of gazetteer "ji" in the Qing, the promotion of "dashiji" in the Republic was also a "soft" law that depended on the very people who were supposed to be regulated by it to enforce it. which should be included in a separate section of yange.
[clause 21]: All county and municipal gazetteer projects should be regulated by the provincial gazetteer bureau and guided by the same principles of this ordinance.
[clause 22]: Special municipalities' gazetteer projects are also allowed to follow this ordinance. 56 The distinction between the dashiji and the conventional yange made it clear that the ordinance meant a new section on major political and social events of the locality. By issuing it as a guiding principle for gazetteers at all levels of locality, it was trying to usher in some innovation in the gazetteer. Some chronicle-equipped gazetteers compiled after 1929 clearly stated that they added it because of this ordinance. 57 So, it certainly played a role in the surge. But there are reasons not to explain the surge with the ordinance alone. First, the chronicle clause was not highlighted in the text of the ordinance. Placed as No. 13 in a total of 22 clauses on the gazetteers' content, style, and institutional setting of its production, it seemed that the request was raised by the central government as part of its routine promotion of the gazetteer that the central government in China had been engaging in for centuries. Many of the clauses involved requests for innovations, such as the use of new style maps and photography. Mixed with others, the chronicle request appears new but not strikingly new. Second, there was no strict enforcement of that requirement, and many Republican gazetteers still came out without a chronicle section. The requirement for the chronicle section was, just like the ritual anxiety against it in the late empire, a soft law enforced by the gazetteer compilers themselves. Therefore, some other causes must be sought to explain the increase in number of gazetteers that did include a chronicle. Third and most importantly, the momentum of increase had already started before the issuing of the ordinance. Among the 202 Republican chronicles, 50 were from gazetteers compiled between 1912 and 1929, which makes that short period itself a breakthrough. Discussions about the necessity of gazetteer chronicles, both in the gazetteers' own style statements and in academic journals, also became more common in the early years of the Republic than they had been during the nineteenth century, when many gazetteers, if they came with a chronicle section at all, simply did it without theoretical discussion.
Sufficient examples indicate that, by 1929, the chronicle was already a trending issue among gazetteer scholars and practitioners. The prominent professor Deng Zhicheng offered courses on local gazetteers in the newly established history department of Peking University; in 1917 Deng published his "Proposed Rules of Style for Contemporary Provincial Gazetteers," which includes that of the "creation of ji chronicles." For Deng the provincial chronicle is analogous to the chronicles of state history in the sense that "while the state-history's ji chronicle is tied to the dynasty, the local gazetteer's ji chronicle is tied to the local place." 58 Liang Qichao's own promotion of Zhang Xuecheng's gazetteer theories, probably the most powerful and influential in the Republican era, was published in the early 1920s. 59 Yu Shaosong's Longyou county gazetteer of 1925, a highly celebrated and widely cited model piece, argued in its "rule of 56 Zhao Gengqi, Xiuzhi wenxian xuanbian, 125-26. 57 For example, see Xu Anyang xianzhi (1933), "fanli". style" section that the ji chronicle should be treated "simply as the backbone to organize all historical events, not just an honorific title for the monarch." A sense of liberation from the ritual taboo of the imperial age characterized Yu's republican pride and gave him a sense of superiority over all his Qing predecessors, as he remarked that "for too long gazetteer writers had been so constrained by the convention of geography … and their own fear of transgression that they dare not to install ji for the gazetteer." 60 The Chengcheng County gazetteer of 1926 probably best demonstrates the Republican gazetteers' position between the genre's tradition and the new political reality: Ever since the Song, many gazetteers followed the example of the Jingding Jiankang zhi and conflated recordings on local events with that on yange. Thus the locality's past, aside from those formatted information [scattered throughout the gazetteer], were reduced to barely more than natural disasters and wars, assigned to such categories as "miscellany" (zazhi) and "scattered information (shiyi)." There was rarely a separate [recording of the past] specifically for the locality. This was so because in the past, the political system was autocracy and the local place simply follow orders from the state; there was little in the local place worth recording. But now the Republic emphasizes local self-governance, local places all handle their own affairs, following their own specific situations, and hence need specific recordings of events." 61 Considering this development in the early Republican era, it is better to view the ordinance of 1929 as a state policy that was itself generated by a surging tide to "chronicalize" the gazetteer, and then in turn greatly fueled the tide. Both the production of gazetteers in general and the inclusion of chronicles in the gazetteer peaked in the 1930s. The two commonly cited motivations for the addition of chronicles in gazetteers of the 1930s-the removal of ritual restraints and the increased local political participation-were the same as those cited in gazetteers compiled during the early years of the Republic. The Yu County gazetteer of 1931 attributed the rarity of chronicles in past gazetteers to the fear of ritual transgression and lamented the timidity revealed thereby: the root chronicle belongs properly to the emperor and none dare to transgress upon it. But it was meant to weave together years and put major events in unity, not just to record the son-of-heaven's daily living and rests; it is understandable if gazetteers avoid the word 'root' in the title, but too much if they avoid the chronicle all together. 62 The Dongming county gazetteer of 1933 attributed the rarity of chronicles in the past gazetteers to the centralization of political power, and celebrated the newly created chronicle section as a symbol of democracy and local self-governance: What is a "major event chronicle"? it records the major events of a county … why were previous gazetteers either extremely brief on this or totally missed it? In the feudal age this was still allowed. But when the prefecture-and-county system extended to the whole All-under-Heaven, authority lay in the monarch, all the commanding power belonged to the state. Consequently state-history became more developed, but the locality did not participate. That is why local gazetteers mostly equated the chronicle with a collection of miscellanies. Today it is very different. The polity is democratic, and governance starts from the local … that's why we install a new chronicle section 63 In general, therefore, the surge of gazetteer chronicles was a trend that spanned the entire Republican era. The central government's request helped the surge but was neither the sole nor the primary cause. The removal of ritual restraint and the increase in local political participation were both related to the 1911 revolution. But if we put this surge in a longer historical perspective, even the revolution appears less a turning point: the ritual restriction was never formally enforced; local self-governance had been talked about by late Qing reformers, implemented before the revolution, and could be traced back to the recurrent discourses of gentry participation in local governance during the late imperial period. 64 Considering that gazetteer chronicles had been promoted by Zhang Xuecheng at the turn of the nineteenth century, and had been implemented every now and then since then, the surge in the Republic was rather an acceleration of a movement long underway in slow motion. It was, just like the other innovations of the Republican gazetteer, a change that enhanced continuity.
Throughout the nineteenth century, the call to include chronicles in the gazetteer had been intertwined with the effort to place the gazetteer on the seat of "history"; both bore the impact of Zhang Xuecheng, and both progressed slowly. In contrast, the dramatic increase of gazetteer chronicles in the Republic, together with the booming gazetteer production, was a clear triumph of the genre. It was also a triumph of Zhang Xuecheng: acceptance of his "gazetteer is history" thesis was almost universal in discussions on the gazetteer's nature during the Republic; he was generally accepted by modern gazetteer scholars as the founder and foremost theorist of the gazetteer field; and his name was referred to 268 times among the 776 Republican gazetteers in the database. (Considering that many of the gazetteers were compiled by non-specialists, that is, people not necessarily familiar with the gazetteer theory or interested in its history, such a rate of reference is remarkably high.) But the gazetteer's triumph should not be attributed all to Zhang himself. After all, there was the century-long low period between his death and the thriving of the cause. Other factors, such as the rise of the "New History," must have also helped to shift the subject of history from the monarch to the people, and to generate energy among the latter to write their own history. 65 Dynamics of republicanism and nationalism, however, are different from dynamics of gazetteer production. There must be other factors involved for people to turn to the local and the gazetteer in the first place, factors that made the local stand more prominently for the national.
This brings us back to locality itself. The movement that Zhang launched was deeply imbricated in the late imperial social and spatial order. While its stagnancy in the immediate aftermath of Zhang's death could be explained by the lack of urgency, the very locality-polity balance which enabled it also contained the possibility that, 63 Dongming xian xinzhi 東明縣新志 (1933), juan 22. 64 Kuhn, "Local Self-Government"; Min Tu-Ki, National Polity and Local Power: The Transformation of Late Imperial China, edited by Philip Kuhn and Timothy Brook (Cambridge, MA: Harvard-Yenching Institute, 1989). under certain social and political circumstances, either the polity's whole or the locality's parts could be pushed to the fore and flare up. Recent scholarship has in fact identified periodic "localist turns," or high waves of local activism, within the stable localitypolity balance during the late imperial period. 66 What broke the stalemate and caused the drastic acceleration of the gazetteer's movement in the Republic, as the next section tries to demonstrate, was exactly such a conjuncture: amidst the unprecedented challenges facing the polity since the mid-nineteenth century, decades-long attempts to solve the problem bottom-up pushed locality to the foreground, creating another "localist turn," though with definitely new features.

The Troubled Polity and the Foregrounded Locality
The localist turn since the mid-nineteenth century has been a well-studied subject. 67 Based on available studies but slightly changing the usual perspective of the state-society relationship, this section focuses instead on people's understanding of the localitypolity relationship during this strong and enduring wave of localism, compares it with the late imperial view, and uses this as a way to see what happened in the localities that modern China inherited from its late imperial past.
Late imperial local activism, as recent local history scholarship has demonstrated, could take various forms, ranging from celebrating the local place's excellence in pursuing canonic virtues to exalting the local particularities or engaging in local welfare projects; local identity formation could crystalize on many different levels from the sub-county to the province; and "localist turns" often came after a high wave of state-activism and sometimes coincided with dynastic decline. 68 In the mid-nineteenth century, however, the decline of the Qing converged with the coming of the modern West, and eventually "transformed a dynastic decline of largely traditional type into a social and intellectual revolution." 69 This transformation, it appears, did not break down the old locality-polity balance. Rather, it brought new forms to express local pride and new theories to explain the significance of the local. Locality was now pushed to the foreground with its various scales as in the past, but it carried newer meanings, greater expectations, and higher urgency. Mid-century administrative reformers in the aftermath of the Taiping rebellion, such as Feng Guifen, usually resorted to the fengjian theory in their push for enlisting energies of local people to revitalize the state. Feng was inspired by Gu Yanwu; for both, locality's plight is the symptom of the polity's problem, and reinvigorating the former leads to the strengthening of the later. Underlying their proposals was a view of the polity described by both Wang Yangming and Zhang Xuecheng, the view that locality is the microcosm of the polity and the site where good governance of the polity starts. Philip Kuhn described this view of the polity as "immanence," i.e., the essence of the whole collectivity to be found in every unit of the local component. 70 Constitutional reformers toward the century's end often shared with the fengjian tradition a bottom-up approach to solve problems of the whole. 71 Huang Zunxian's lectures to the Southern Study Society in Hunan, for example, mixed the fengjian idea with the new term "self-governance" he borrowed from Meiji Japan, and appealed to the Hunanese local elites to solve their own problems by assuming responsibility for the need of their home districts (xiang). 72 Huang's overall goal of economic development and national strength was already different from such traditional concerns of the local elites as water control and banditry, but the underlying view of the locality-polity relation remains the same: the local parts held the key to the salvation of the collective whole. This view was most dramatically revealed by Huang's young collaborator Liang Qichao, who cited the serious situation facing the state and the real possibility that the polity may be carved up by foreign powers. The meaning of their society and the local selfgovernance efforts in Hunan, Liang explained, was to prepare for "the struggle of Hunan after the fall of the polity." 73 That Hunan has a future after the fall of Chinathe goal of its future struggle, though not specified, could be nothing but the continuation and restoration of the polity-is conceivable only if Hunan contains all the qualities and potentials of the polity. It is, to use the phrase of Zhang Xuecheng, the "same structure in small size." Liang made this remark passingly; yet for this very reason it appears a more effective illustration of how profoundly Chinese political thought at the turn of the twentieth century were conditioned by the "immanence" view of the polity.
Following the political catastrophe of 1898, both Kang Youwei and Liang Qichao started to promote the cause of "self-governance." Kang linked the fengjian line of thought to the social-Darwinian concept of the survival of a "fit" people: enfeoff the people, energize the local, yet organize them within the national system of rules. Liang Qichao, on the other hand, distanced himself from the fengjian tradition, which itself was being discredited with the rise of statism in the early twentieth century, and in its stead sought inspiration from western political philosophies he was guzzling in while in exile in Japan. 74 Yet the immanence concept of the polity is obvious in his treatise "On Self-governance": 70 Philip Kuhn, "Late Ch'ing Views of the Polity," in Selected Papers from the Center for Far Eastern Studies (Chicago: University of Chicago Press, 1981), 1-18; here 11. 71 On the forerunners of the constitutional movement and their ideas, see works by Zheng Guanying 鄭 觀應, He Kai 賀凱, Chen Qiu 陳虬, and others. See also Kuhn," 269;Min,National Polity and Local Power,133. 72 Huan Zunxian, "Huang Zunxian nanxue hui diyi ci jiangyi," in Liang Qichao, Wuxu zhengbian ji, juan 8: 13a. In 1908, the idea of bottom-up mobilization finally prevailed, and local selfgovernance became the law of the land. A hierarchy of assemblies was established at the sub-county, county, provincial, and national levels. John Fincher's study of these assemblies finds that "loyalty to a grouping at one level was compatible with loyalty to a grouping of another level," and that layers of loyalties seemed to have linked together to form the foundation of nationalism. 83 In the implementation of local selfgovernance, the Qing court's quest for control and the local elite's quest for mobilization were often discrepant. While the state sought to co-opt and utilize the local elites, sometimes local elites were ready to go to the extreme on the idea that the realm is a collection of local places, and ignore the state and the center altogether in their discussion of national affairs. 84 Inside local societies, institutional change incurred new strife among different social groups, for example over the drawing of constituency boundaries at the sub-county level. 85 Despite all these, the deeply entrenched "immanence" view of the polity seems to be a common ground. This explains why political thinkers of China at the turn of the twentieth century so keenly examined the emerging new polity through its manifestations at the local level. In Kuhn's words, "if the power of the total society was immanent in every part of it, then presumably the place to revitalize the society was at its grassroots. … to revitalize the nation, it was local society that had first to be revitalized." 86 "The local," it seems, had become the catchword for the rising nationalism and the emerging nation-state on the eve of the Republic's founding. The philosopher Feng Youlan once recalled that his father, while an official serving in Wuchang during the final years of the Qing, was advised to quit his official post and go home, for the career of a local gentry brings a brighter future than serving the state. 87 After 1911, the new republic largely retained the Qing structure of field administration system in China Proper, except for the abolishing of the prefecture. 88 Thus, in general, the same provinces and counties of the Qing now stood for localities of the Republic. 89 But decades of debates, experiments, reforms, and revolutionary activities that had pushed the local to the foreground in the late Qing helped to set it on a new footing under the Republic. In terms of institutions, election-based legislative bodies of the province and the county all became constitutional. Concepts, if not smooth practices in all places and all time, of local budget and local tax at both the province 83 John Fincher, Chinese Democracy: The Self-Government Movement in Local, Provincial, and National Politics, 1905-1914(New York: St. Martin's Press, 1981 For example, in order to cooperate among themselves and pressure the Qing court to shorten the preparation period before the adoption of a constitution, provincial assemblies sent delegates to meet and discuss national affairs. The location of the union meeting was selected in Shanghai, avoiding the imperial capital Beijing. 85 Satō Yoshifumi, Jindai Zhongguo de xiangtu yishi, 28-71. 86 Kuhn, "Late Qing Views of the Polity," 9. The new level of Dao between the province and the county did not necessarily conform to the territorial shapes of the old prefecture, and hence they were not able to become a territorial entity that people could identify with in a short period. The ward (qu) level of government beneath the county had started before the founding of the Republic but was standardized during the Republican era. Since few qu-level gazetteers were produced during the Republican period, most likely due to the very newness of the qu entity, the qu is not included in the discussion of locality here. On the Republican local system, see Lin Jiongru, et. al, Zhonghua minguo. and county levels (which, strictly speaking, did not exist at all in the Qing until 1908) come into being from the very beginning of the Republic. 90 In theory at least, such institutional changes made it possible to view localities as distinctive social and political entities in their own rights. In a sense, they did to locality what the chronicle ordinance of 1929 did to the gazetteer: formal recognition of a status that was often already in de facto existence. In real politics, the theme of building the nation through empowering the locality had gathered enough momentum to survive the attempts to rein it in. Yuan Shikai's decree to abolish local assemblies in 1914 dared not to completely remove the trappings of local self-government, and was soon rescinded after Yuan's death. 91 The Beijing government after Yuan issued new laws on this matter, and established schools to train local self-governance cadres for the provinces. 92 Sun Yat-sun had long since been interested in local self-governance, and made it clear in public speeches during the years of Yuan's presidency that "local self-governance is the basis of nation-building." 93 In the early 1920s, with the center weak and provincial identity strong, and social Darwinism outliving its appeal (now that the country had achieved independence from the Manchus), Western political theories of federalism became particularly attractive, and the dynamics of localism soon took the form of province-based federalism. 94 The most dramatic component of this movement was the Hunan nationalism that involved none other than the young Mao. But the overall view of the locality-polity relation stayed unchanged. Mao's own promotion of Hunan nationalism was never unmoored from an imagined China in the background. He wrote that "we must strive first for a 'Republic of Hunan,' to implement the new ideal … and to become the leader of 27 small Chinas." Eventually, Mao admitted, China would be reunited, if only from the bottom (province) upward. 95 Sun Yat-sen did, for a moment, object to the province-dominated approach of the federalist movement. But this does not conflict with his overall engagement in local selfgovernance. Sun's difference with the federalist movement was rather that of military vs. civil road to unification, as Sun insisted that "real self-governance of the provinces is possible only after the accomplishment of the National Revolution." 96 Until the end of his life, Sun remained a believer in the bottom-up approach to nation-building, as he settled on the county as the starting point of democracy in the Outline of National Reconstruction (1924), and stated in the lectures on the Three People's Principle (1924) that "the native place sentiment of the Chinese is very deep-rooted; it is especially easy to unite people from the same province, prefecture or village." It 90 is clear that he viewed local self-governance as prerequisites for national integration and strength, and was hence perfectly consistent with the age-old view of the locality-polity relationship. 97 Like Sun, many in the 1920s who objected to province-based federalism actually supported sub-province-level local self-governance. 98 In 1928, the Nationalist Party (GMT) came to power through military unification. In terms of Sun's agenda, the enterprise of nation-building now entered the tutelage stage. By then the idea of bottom-up mobilization and the general guideline of local self-governance, i.e., starting from the county, gradually reaching the province, and culminating in national level constitutional rule as envisioned by Sun, had become the mainstream of the GMT political theory. One of the party-state's most proclaimed commitments during this stage was to implement local self-governance according Sun's vision. The County Organization Law, which defined the county as both an administrative and a self-governance unit, was drafted in 1928 and issued in 1929. A nation-wide schedule was made to reorganize county government in 1930 and accomplish county self-governance by the end of 1934. 99 In actual practice, the needs to consolidate power and fight the communist rebels pushed the GMT decisively away from the idea of mobilization and toward more rigorous bureaucratic control. But Chiang Kai-shek never abandoned, nor could he afford to abandon, the idea of popular mobilization from the bottom up, which was a signature program of the party's founder. Throughout the GMT rule, major publications on the country's political system consistently featured the tutelage goal of local self-governance. This was the case also in the internal speeches during GMT's highest level training sessions for its cadre. 100 When the effort to tighten bureaucratic control of the county was proved unsatisfactory, Chiang promulgated a "New County System" in 1939 to reinstitute certain aspects of local self-government. 101 In general, one may argue that local self-governance continued its oscillation between control and mobilization as it had under the Qing. The county elites were often frustrated in their hopes to transform China's politics from the bottom up. But the state power Chiang dreamed of never became a reality, and the surging tide of localism originating in the mid-nineteenth century was never turned around. 102 In this circumstance, compiling and publishing local (mostly county) gazetteers was probably the least they could do, and did, engage in and accomplish. 97 Keith Schoppa's study of federalist movement in Zhejiang has demonstrated that provincialism in the 1920s could be the genuine vehicle of provincial interest, and he argued that "provincialism could be a fullfledge alternative to nationalism as well as a vehicle for it." See Schoppa, "Province and Nation." My take on this is that representing genuine provincial interest doesn't necessarily make it an enemy of nationalism, though it does make it an alternative to statism.  101 Kuhn, "Local Self-Government," 297. 102 On the shift of power toward the local since the mid-nineteenth century, see Kuhn, Rebellion and its Enemies, Frederic Wakeman Jr., Strangers at the Gate, and Joseph Esherick, Reform and revolution in China, among others.
As a whole, the several decades after the mid-nineteenth century crisis presents a picture in which the local stayed in the foreground of political discourses and actions, from saving the state to building the nation. The situations in which the local was called on varied, but the underlying rationale had always matched the entrenched view of the locality-polity balance. It was against such background that local people's sense of agency rose high, culminating in the Republican era, and energizing them to engage the gazetteer more often and more assertively. The statement proclaiming the installation of a chronicle for the Zhongxiang county gazetteer (1937), while directly addressing local people's relationship with officialdom, also reveals their faith and hope in locality if we consider its venue of publication: Ever since antiquity, officialdom was the ruler. The people were just passively ruled, so the shifts between chaos and order were often determined by officialdom, not by the people … but don't the people who live on this soil also share some responsibility over its state of order or chaos? Since the revolution the polity has fundamentally changed. What used to depend on officialdom to sustain are now perhaps: decided by the people. The shifts between order and chaos therefore are now determined by the people, not officialdom. 103 Such high-profile celebrations of locality, common in Republican gazetteers, attest to a correlation between locality and the gazetteer, the kind of correlation that was also at work during Zhang Xuecheng's intervention into the gazetteer genre in the late eighteenth century. If Zhang's intervention was a case whereby the growth of the gazetteer requested the "rectification of name" for its subject, the triumph of the genre in the Republic can be viewed as a scenario in which the dynamics of locality pushed its gazetteer ahead. 104

Conclusion: The Local Across Two Worlds
There is no doubt that the foregrounding of the local was part of a strategy that had the whole as its end concern. But that does not mean the local was reduced to mere means or that it lost its dignity. The advocates of this strategy were either retaining the local roots of their social life (such as Feng Guifen) or taking a local perspective on the whole (such as Kang Youwei or Sun Yat-sen). In this strategy, the mindfulness of the whole was compatible with a local viewpoint, for its logic is that the whole is good if and only if the constituent local parts are good. This leaves room for the local to be active and distinctive. It is no accident that, in Zhejiang, for example, after decades of promotion, this new "localist orientation" generated a provincialism during the federalist movement of the 1920s that served as a genuine vehicle of provincial interests, yet 103 Zhongxiang Xianzhi 鍾祥縣志 (1937), juan 1 104 As a sidenote, it is probably worth noting that up to December 1929, offices of the Nationalist government involved in the issuing of the ordinance on gazetteers were all headed by believers in local selfgovernance, including the Executive Yuan premier Tan Yankai 譚延闓, and the Inner Ministry ministers Yan Xishan 閻錫山, Zhao Daiwen 趙戴文, and Yang Zhaotai 楊兆泰. Some were also deeply involved in the practice of it. Tan was the governor of Hunan during the Huanan Constitutional movement; Yan famously ruled over Shanxi, his home province, for decades; Zhao and Yang were both Yan's proteges and played crucial role in the Shanxi local self-governance experiments. See Donald Gillin, Warlord: Yen Hsi-shan in Shansi Province 1911-1949(Princeton: 1967; see also Li Dajia, Minduo chunian de liansheng zizhi yundong. still operated within the parameter of an imagined nation. 105 In terms of resisting the arbitrariness of the center, this modern provincialism is as genuinely local as the localism of the late empire, though its political power in an age of state weakness was real enough to cause worry in an aspiring state-maker such as Sun Yat-sen. 106 Yet the commitment to bottom-up mobilization and local self-governance by Sun, and through Sun the GMT, was strong nonetheless. It is fair to say that the prevalent view of the localitypolity relationship in modern China up to the CCP takeover was comparable to that of the late imperial literati such as Wang Yangming and Zhang Xuecheng.
The flourishing of local gazetteers in the Republic must be understood against this background. A genre serves its subject matter and changes with it. If the late imperial locality survived into the Republic, it was only reasonable for the gazetteer to follow along, with some technical changes if necessary. As with its late imperial predecessors, the vitality of the Republican local gazetteer was based on and reflected the locality's status as the dignified building blocks of the polity and its irreducible quality as the microcosm of the latter. To treat the gazetteer as "history" or to grant it a chronicle are both efforts to elevate the status of the genre, justifiable only on the ground of a conscious recognition of this locality-polity relationship. In the late eighteenth century the recognition was wide enough to allow Zhang Xuecheng's intervention in this direction a smooth start, but the situation changed only slowly in the nineteenth century, because the requests were mainly driven by the enthusiasm of gazetteer compilers and gazetteer scholars. It was only in the Republic that the convergence of several factors, including the fall of the monarchy and the locality's position on the frontline of political discourses and actions, led to a change of the dynamic. The Nationalist government's ordinance in 1929 finally hammered this down. The campaign to elevate the local gazetteer's status finally succeeded one and a half centuries after it started.
Much, of course, had happened in between. The polity in the time of Zhang Xuecheng was still the tianxia centered on Beijing, its history was "the history," coinciding with "philosophy" ( jing 經) and connoting universal truth. The polity of the 1930s, when both the gazetteer and its chronicle witnessed their golden age, had largely abandoned the idea of "heaven," and completely lost its centrality. Its history was just "a history," and the "ji" chronicle lost all its ritual trappings. It was then that, formally and officially, the gazetteer received its long overdue recognition as "history," and locality gained its voice as the subject of historical narrative. Thus, ironically, we must say that when China was no longer the "All-under-Heaven" but a nation-state, its relationship with its own building blocks was not only retained but also clarified and formalized. Modernity certainly changed a lot of things in China, but not its internal spatial order-that is, not until 1949.
What happened after the Communist takeover is too complicated to be discussed here. Speaking from the well-known and the indisputable, the Tiananmen Tower and the socialist motherland replaced the native soil as the beginning point and the end goal of history and geography education for school children. There were still talks about giving the locality more leeway and allowing them to have more initiative, as Mao himself famously did in 1956. 107 But there was a world of difference between 105 On this subject, see Schoppa, "Province and Nation," and Duara, Rescuing History, 177-204. See also Li Dajia, Minduo chunian de liansheng zizhi yundong, and Hu Chunhui, Minchu de defang zhuyi yu liangsheng zizhi. this "centralized decentralization" ordained from the center, and that "decentralization as a pre-nationalist 'fact of nature.'" Mao's was the confidence of a conqueror that had completely subdued the locality and made it docile. The locality-polity balance that had been there for centuries was gone. 108 The first toll of this fatal transformation was probably the local gazetteer. They were still compiled. But with the destruction of the local elites it had become purely a function of the socialist bureaucracy. Zhang Xuecheng's questions were no longer relevant and the gazetteer lost all its charm.