Julia C. Strauss
School of Oriental and African Studies, London

Words, the grist to the mill of human language and cognition, are notoriously imprecise and slippery instruments.⁽²⁾ Much to the chagrin of nineteenth-century positivists and logicians, words have continued to remain stubbornly dualistic entities, simultaneously possessing both denotative and connotative capacities.⁽³⁾ Denotation literally "notes down," "marks," "indicates," or, in logicians' terminology, is "predicated of what the thing in question objectively `is.'" Connotation "signifies secondarily or in addition," "implies or involves as a consequence, condition or accompaniment," or, in logicians' lingo, "implies or indicates the subject in which an attribute adheres, while primarily signifying or `noting' the attribute itself."⁽⁴⁾

Words do sometimes simply denote, providing a convenient linguistic shorthand to group and classify phenomena that objectively exist and possess concrete attributes that cause them to be so named--a bird, an apple, grain, snow. But more often words also connote, adding or implying attributes to the object denoted. These added attributes, in turn, "charge" the word in question with either positive or negative valences. So the word "chicken" can denote a particular type of fowl that is especially tasty roasted with lemon and garlic, or it can connote faintheartedness and cowardice. Some words are so heavily valenced that one uses an entirely different word to denote exactly the same phenomenon: "politicians" are bad, but "public servants" are good, and "statesmen" are even better.⁽⁵⁾

The connotations (or attributes) that accrete to words over time are often multi-layered and need not necessarily be consistent with each other; words are exceptionally plastic and accommodating to users with differing values and world views. Language is simultaneously instrumental, in that it is a major medium for social communication, as well as a symbolic arena in which different values are thrashed out. As a tool, language may provide a means for control, cultural hegemony, or resistance. As a realm of symbolic interaction, language is often one of the main areas in which different groups articulate their own positions, mobilize followers, come to terms with their heritages, and distinguish themselves from other, competing groups. Given its multiple instrumental and evocative roles, it is little wonder that language (and its basic units of currency, words) offers an attractive field of play for politics.

State administrators are among the more important (if much maligned) actors in the process of statebuilding, as well as in politics in general, for both practical and symbolic reasons. Practically, civil administrators are the tendons and circulatory system of the state. They hold the central state together, carry out its directives, and form the first line of interface between it and society. Without a reasonably committed set of administrators willing to accept hierarchical discipline and work for state goals and interests, a state simply cannot exist. The symbolic importance of state administrators, in my view, flows from their practical indispensability. Since administrators are so undeniably important, the ways in which the state conceives of appropriate roles and tasks for its administrative agents provide a fairly good baseline indication of how the state conceives of itself in terms of ideals, goals, and programs for action. If one looks a little farther afield to consider also the terms for state administrators that groups outside the state use, one can gauge the degree to which there is congruence or radical disjuncture between respective goals and visions of the state and those of important groups in society. Congruence can result from two processes with quite different political implications: either because the state has created a hegemonic vocabulary and set of constructs that social groups accede to or because important groups are coopted into the state, with their vocabulary and set of constructs being reproduced by the state itself.

Important as state administrators are, it usually proves quite difficult to sustain terms to describe them that retain neutral-to-positive connotations over any length of time. English, for example, has at least three commonly used terms to denote state administrators--"official," "bureaucrat," and "civil servant"--as well as two other less frequently used terms--"functionary" and "Mandarin." Of these, only one, "civil servant," retains even a kernel of positive valence, with connotations of diligence, even-handedness, and working for the public good. All the others connote red tape, delay, unresponsiveness, petty detail, and (in the case of "Mandarin") unapproachable, self-contained, generalist elitism. With the exception of "official," which has been in common English usage (originally denoting "one with a clerical or household office") since the twelfth or thirteenth century, most of the other terms came into the English language fairly late, in the early to mid-nineteenth century, usually via French. At around the time that "bureaucracy" and "bureaucrat" (which Carlyle called "the continental nuisance") came into English at mid-century, the previously fairly neutral term "official" began to acquire unsavory and derogatory connotations as new suffixes began to be added, creating the highly negative terms "officialism" and "officialdom."⁽⁶⁾

"Civil servant" and "civil service," on the other hand, came into use towards the end of the eighteenth century, originally in reference to the non-military, covenanted servants of the East India Company, and eventually grew to denote "all non-warlike branches of the public administrative service of the state." As such, "civil service" was, relative to the "continental nuisance of bureaucracy," a home-grown concept, even if it first came into use in the context of overseas colonialism. Thus, even in English, the problem of what to call a state administrator has proven to be a subtle and tricky one: perhaps most terms will eventually tap into an Anglo-American tradition of mistrust of government and monocratic authority and suffer a corresponding decline in connotative valence.

When the focus of analysis is shifted to a place like Republican China, the complexity inherent in the labelling and naming of state administrators increases enormously, because political institutions and political language during this period were, to put it mildly, non-standardized. Political elites in the second quarter of the twentieth century in China were the heirs to a now-crumbled but previously long-existing, indigenous tradition of imperial statecraft that had historically accorded very high status to government officials. Needless to say, the late imperial state's weaknesses in dealing with foreign pressure, its difficulties in accommodating the demands of activist provincial elites for more formal recognition of their political roles, and its inability to put together a strong central state in the first decade of the twentieth century called many of its values and norms into question once the Qing fell in 1911. Unfortunately, the Qing was not replaced by a stable successor regime. The first fifteen years of the Republic were characterized by a crisis of governance: the failure of civilian institutions, increasingly large scale but illegitimate militarization and political fragmentation, elite disenchantment, and increasingly vocal anti-imperialist nationalism that was hard put to find a state on which to pin its nationalist sentiments.

By the mid-1920s, revolutionary mobilization was in full swing and from this time up until 1949 literally all the important questions as to China's future were up for grabs: how China's revolution would be successfully defined, how the state would be reconstituted, and on what basis the state would interact with society. The Communists and the Left insisted that the only revolution worthy of the name had to involve class struggle and social revolution from the bottom up, while the center and right wing of the Guomindang ultimately moved to a minimalist definition of China's revolution as unifying the country, revoking imperialist privileges, and promoting rapid economic development. Initially, the center and the right wing of the Guomindang seemed to prevail over the Communists with the White Terror of 1927. In fact, the purge of the Left in 1927 left two much-weakened parties, each of which considered itself to be revolutionary, to figure out what revolution and state building meant in practice. Not only was there keen political and military competition between the Guomindang and the Chinese Communist Party, but each had to deal with internal divisions as well.⁽⁷⁾

The twin processes of intra-party standardization and inter-party competition motivated both the Guomindang and the Chinese Communist Party to form state (or quasi-state) governments. The Chiang Kai-shek­dominated Guomindang established its state apparatus in Nanjing, while the Chinese Communists--after being pushed out of the cities in 1927--regrouped to establish revolutionary base areas first centered in the marginal hill country of Jiangxi, then, after the Long March, in the equally poor and even more remote far northwest. At various points in time, both revolutionary parties possessed concrete territories that needed administering. Although the Guomindang regime attacked the issue of state reconstitution and civil service with relish while the Chinese Communist Party only very hesitantly and backhandedly admitted the need for administrators at all, both ultimately had to generate vocabularies of state and administrative service. In so doing, the Guomindang and the Chinese Communist Party reflected two main influences: China's own historical legacy of state and administrative service and the respective foreign models that each explicitly turned to for legitimation and support in the 1930s and 1940s. These two sets of factors recombined in unusual ways during the remainder of the Republican period, solidified in the orthodoxies of each side of the Taiwan Straits after 1950, and would experience vigorous historical afterlives during the liberal reform period of the 1980s in the People's Republic of China.

In imperial China, state administrators were generally denoted as guan, occasionally as li, and often by a ci that combined the two as guanli. These terms, alone or in combination, are conventionally translated into English simply as "official." Often guan were given the additional modifier of wenguan (or "lettered official"), so that those in the high status civil bureaucracy might be clearly distinguished from wuguan (or military officials), who were much lower in status. As the products of a quite unique interaction between the late imperial agrarian state and society, wenguan in imperial China not only implemented the usual functions typical of administrators in pre-modern states but also carried a set of particular connotations that set them apart from officials and bureaucrats elsewhere in the world.⁽⁸⁾

Guan in imperial China were the agents who held the thinly stretched agrarian empire together. As a general rule, guan were recruited into state service by sitting for horrendously difficult but "open" state-given examinations that tested for the individual's mastery of Confucian texts and composition.⁽⁹⁾ The relative numbers of those who passed these Confucian-based civil service exmas were small. Those who passed, of course, were usually from fairly well-off families who had the spare resources to invest in the requisite years of education and private tutors for their sons; serious preparation for these examinations typically took an astonishing twenty or so long years of rote memorization.

Functionally, this system presupposed that virtue (de) and talent (cai) could be ascertained on the basis of memorizing key texts and composing eight-legged essays. This was often demonstrably not true and, in fact, the contents of the examination system were often singled out for sharp attack as irrelevant for the administrator's subsequent career. But the social uses of the examination system went far beyond its technical functions in ensuring a supply of administrators to the imperial state; the myth of an impartial and open system whereby any studious and conscientious individual could succeed acquired tremendous currency and legitimacy in society at large, as did the correspondingly highly valued "career open to talent."

Guanli stood at the very apex of China's late imperial society. They were certified by the state as having come through a series of examinations all knew to be extraordinarily difficult, and they were on the sole career track that elite society recognized as being of any general prestige. What guanli did when in office, however, was another matter. Especially at local levels, imperial administrators were undersalaried, overburdened with financial and judicial administration for which their Confucian training often ill-prepared them, and heavily dependent on both sub-bureaucracies of even worse salaried local clerks and the cooperation of local gentry.

In this social environment, the term guan carried multiple connotations. Guan comprised a status group of enormous prestige which was, however, validated on the universalistic criteria of examination success. Guan were generalists who, on the basis of their Confucian humanist education, were presumed to have the capacity to "see into the heart of any situation," like the model wise official Baogong. As such they were expected to serve as models of rectitude and proper behavior for the people. As the writings of Huang Liuhong in the late seventeenth century and of the early nineteenth century "statecraft school" in the Huangchao jingshi wenbian make clear, guan at their administrative best quickly acquired on-the-job training and practical experience in "handling affairs." Generalist literati training did not automatically select for incompetent aesthetes, and those who were overly bookish or politically naive did not fare well in the bureaucratic system. On the other hand, guan were poorly regulated and audited, particularly at local levels of government. The lack of adequate salaries combined with enormous administrative demands structurally constrained even the most honest to take extra fees to make up the chronic shortfalls, and the line between "legitimate" fee-taking and "illegitimate" "squeeze" was thin and ill-defined. Despite the fact that many of imperial China's better administrators gravitated towards de facto areas of special competence, the ethic of guanli stoutly resisted overt specialization (i.e., by not incorporating "practical" subjects into the civil service examinations), endlessly repeating Confucianisms that "men are not tools" and that if rule by utility were to replace rule by virtuous example, all would be lost.

Guan in late imperial China were, then, characterized by the following multi-layered connotations: high social status, a career open to talent, walking incarnations of good Confucian norms and values, resolute generalism, and, negatively, corruption and "squeeze." Their cultural and structural position in society was perhaps best summed up by the popular four character expression, shengguan facai ("become an official and get rich"). But the Confucian generalist ideals that undergirded the conception of guan were increasingly called into question in the late nineteenth and early twentieth centuries, made irrelevant by the Qing dynasty's xinzheng reforms in the first decade of the twentieth century, and then by the collapse of the imperial system itself in 1911. Given the close association between guanli and the late imperial state, and the failure of the latter, the connotations of officialdom encapsulated in the popular expression shengguan facai were hardly a set of characteristics to recommend themselves to the new generation of vigorously modernizing, revolutionary elites that emerged out of the chaos of the warlord era in the 1920s.

With the rise of Leninist revolutionary mass parties in the 1920s, the term guan, in all its permutations (guanli, li, and wenguan), began to go out of favor. Guan was fairly consistently used for the remainder of the Republican period as an epithet for status-conscious, lazy, parasitic, and self-serving officials, while guanliao (bureaucracy, bureaucratic) took on connotations of weak, passive, unresponsive, and failed government. In late 1921, Sun Yat-sen equated the early Republic's failure with its "bureaucracy," rhetorically asking "for whom has there been a Republic [literally: "people's government," minguo] for the past ten years? [On the contrary, China has been] a country for bureaucrats: revolutionary thought didn't move forward, the goals of the revolution were not reached, the revolution existed only in name and not in reality."⁽¹⁰⁾ In another speech given in October of 1923, revealingly titled "Party members cannot cherish the intention of becoming officials" (Dangyuan buke cunxin zuoguan), Sun Yat-sen warned against the tendency of Guomindang Party members to pervert the revolution by hankering after the status and opportunities for personal gain inherent to "becoming big officials" (sheng daguan). Sun saw revolutionary spirit and officialdom to be mutually exclusive categories: if one

wanted to do great things, one couldn't want to become a big official: if one cherished the intention of becoming a big official, then the true spirit of Party membership would be lost . . . [for] Party rule [dangzhi] does not mean that Party members [dangyuan] will rule the country: but that Party ideology [dang zhuyi] will rule the country.⁽¹¹⁾

In the early to mid-1920s, however, even seasoned revolutionaries who decried bureaucracy and officialism did not have much in the way of synonyms to positively or neutrally connote state administrators on the rare occasions when addressing the issue of state service and government organization was necessary. In the Jianguo dagang (Fundamentals of national construction) of 1924, in which he outlined his vision of modern China, Sun, having little vocabulary other than guan to denote state administrators, slightly softened the term by adding the suffix yuan, producing the term guanyuan.⁽¹²⁾ Yuan individualized and partially neutralized some of the more negative status connotations of guan, with guanyuan meaning, literally, "official individual" or "one who is an official." And just to make the make the positive connotations of guanyuan absolutely clear, Sun then linked the appointment of guanyuan specifically to the most legitimate aspect of the entire complex of imperial officialdom: the holding of state-sponsored, open civil service examinations as the primary mechanism into state service.

After the mid 1920s, guan, both alone and in combination form, survived for the remainder of the Republican period--often as a term of opprobrium, but sometimes with more neutral connotations. Strangely, the term guan persisted at the very top levels of the Guomindang state and in the most prestigious state organizations as well as at the very bottom of the bureaucratic pyramid, in local administration and low-level military organization, where its usage connoted anything but prestige. In the late 1920s and early 1930s, administrative reformers in the newly constituted Guomindang state attempted, without great success, to institute an American-style hard-and-fast division in the state bureaucracy between political appointees (zhengwuguan), who would invariably be Guomindang Party members, and "career" or "functional" appointees (shiwuguan), who, chosen on the basis of their technical or practical knowledge, might well not be Party members.⁽¹³⁾ Admirers of Western civil service systems who wished to see a depoliticized, technically oriented, examination-qualified set of state administrators also upon occasion translated Western civil service systems as wenguan zhidu, thus subtly linking a prominent feature of the progressive, modern West with a positively valenced political institution indigenous to China.⁽¹⁴⁾

But in general, for the remainder of the Republican period (and indeed after 1950) the term guan continued to be used for specific titles in the Ministry of Foreign Affairs, that most prestigious and most elitist of Guomindang central government ministries. General officers in the Ministry of Foreign Affairs were generally referred to as waijiaoguan, who were internally distinguished from consular officials (lingshiguan).⁽¹⁵⁾ Of all the regular central state organizations, the Ministry of Foreign Affairs exhibited the greatest degree of continuity during the transition from Beiyang to Nationalist rule. The Ministry of Foreign Affairs managed to carry over almost all of its staff from the Beiyang period as well as a personnel system almost entirely separate from the rest of the Nationalist government--complete with its own ranks, grades, promotion procedures, and salary scales. Given this institutional strength, it is no wonder that the Ministry of Foreign Affairs preserved its distinctive system of denotations for different sorts of positions within the organization. In the Ministry of Foreign Affairs, at least, the continued use of guan (via waijiaoguan and lingshiguan) hearkened back to a number of the positive connotations of guan in imperial China: exclusiveness, prestige, status, and resolute generalism.

But the Ministry of Foreign Affairs was a highly atypical organization in Republican China and could, in view of its extremely high prestige and drawing power, institutionally "afford" to swim against the tide in terms of linguistic denotations as well as a host of other areas. The persistence of guan in a highly exclusive organization like the Ministry of Foreign Affairs was parallelled by its much more widespread usage at the bottom of the administrative hierarchy. When not referred to by a specific title such as xianzhang (county head), local government administrators were typically called difangguan (local officials). Difangguan, in turn, were often linked to the negative attributes of laziness, perfunctoriness, and lack of education.⁽¹⁶⁾ Military officials in the post-1927 Republic, even ones associated with the National Revolution, were invested with titles that often incorporated the character guan as well. In 1934, Chiang Kai-shek defined the "commanding administrative officers of each province" (gesheng xingzheng zhangguan) as "officials close to the people" (qinmin zhiguan) and then, perhaps inadvertently, linked the commanding administrative officers of each province with that most traditional of clichés: "the people's father and mother" (minzhi fumu), which almost exactly replicated the late imperial concept of the district magistrate as a "father-mother official" (fumu guan).⁽¹⁷⁾ Further, the term jiaoguan (literally "teaching official"), denoting the numerous and relatively low-status military instructors in the National Revolutionary Army and in organizations involved with military training, came into widespread usage from the mid-1920s on.

Thus, during the post-1927 Republican period, the term guan bifurcated in both its denotative and connotative aspects. As far as revolutionaries and progressives were concerned, guan tended to connote laxity, status consciousness, unresponsiveness: in short, everything that was regressive and weak from the imperial legacy. As a general rule, state institutions that continued to use guan as a denoter of administrative personnel were either of low status or held office at local levels of government administration. It was only very strong, elite institutions like the Ministry of Foreign Affairs that could continue to use old denotations without fearing a loss of positive connotative valence. Although Sun himself fell back on the uneasy amalgam of guanyuan to denote state administrators in 1924, very shortly thereafter a whole host of other terms would come to be preferred to denote the middle to upper ranks of the state administrative bureaucracy.

Although guan did not fall out of usage entirely, its connotations were sufficiently negative that alternate denotations for administrators in the middle to upper ranks of Guomindang central state organizations proliferated in the late 1920s and early 1930s: zhiyuan (literally "one who is in a post," with the connotation of "official" post), xingzheng renyuan ("administrative personnel"), gongzuo renyuan ("working personnel"), renyuan ("personnel" in general), and gongwuyuan ("public servant"). Which terms were used most varied according to style and personal preference. Guomindang Party documents usually refer to state administrators as zhiyuan, Chiang Kai-shek favored the term xingzheng renyuan, the Ministry of Finance referred to its personnel as caizheng renyuan, and administrative reformers tended to use gongwuyuan.⁽¹⁸⁾ But each of these denotations was invested with a strikingly similar set of connotations, and they were often used interchangeably. First, all these terms were relatively new inventions, created by adding the suffix yuan to the modifier in front. In contrast to guan, which connoted status, position, authority, and (later) corruption and ineffectiveness, yuan was a neutral marker to indicate a person or individual, and in combined form subtly underscored the presumptive neutrality, interchangeability, and technical competence of the state administrator in question. In addition, these terms substantially deflated the status of state administrators by consistently linking them to the concept of "service" (fuwu)--service to the people, service to the Revolution, or service to the country.

Although zhiyuan, xingzheng renyuan, renyuan, and gongwuyuan were all in use from the late 1920s through the 1940s, gongwuyuan (literally "public things person," or, more loosely, "public servant") was the term that acquired the most currency in government and quasi-government circles. Gongwuyuan was particularly, though not exclusively, favored by two important groups deeply involved with the effort to institutionalize a national personnel policy for government administrators: a group of loosely affiliated non-office-holding academics, commentators, and reformers within the Ministry of the Interior, whom I elsewhere call the "Administrative Efficiency School,"⁽¹⁹⁾ and those in the Guomindang's Examination Yuan and Ministry of Personnel, who attempted to establish and enforce a national set of standards for government personnel in the Guomindang state at large.⁽²⁰⁾ Administrative efficiency reformers both inside and outside the state agreed on a program that called for, in effect, the establishment of a strict Weberian bureaucracy: a depersonalized and defactionalized administration separate from party organizations, control over the administrative bureaucracy via the standards and guidelines established by the Ministry of Personnel, recruitment into the state bureaucracy and the promotion of technocracy via Examination Yuan­administered open civil-service examinations, and the promotion of efficiency and effectiveness through the implementation of scientific administration.

Within a discourse that presumed a "public" administration and scientific management to be positive goals in themselves, the term gongwuyuan, with its connotations of service, obedience to political masters, impartiality, and technocratic expertise, was resorted to quite often. But even among administrative reformers, the lack of generally accepted standard terms made for some rather interesting crosscurrents. In addition to using the aforementioned zhiyuan and renyuan, administrative reformers and those in the Ministry of Personnel also used shiwuguan ("functional or career officials") and shiwu renyuan ("functional or career personnel") to distinguish "regular" career administrators from political appointees (zhengwuguan). Further, at least in discussions involving the importance of establishing civil service systems (a topic of much concern to administrative reformers in the 1930s and 1940s), gongwuyuan never completely ousted wenguan: when administrative reformers resorted to citing the examples of Western civil service systems as partial models for China, both wenguan zhidu and gongwuyuan zhidu were used to denote foreign civil service systems.

But some felt compelled to make a clear distinction between wenguan and gongwuyuan. To cite just one example, in Xingzheng yanjiu yuekan (Administrative research monthly), an administrative reformer by the name of Xie Tingshi wrote in 1936:

In modern China, guan are called gongwuyuan. Although gongwuyuan [public servants] are the equivalent of guan [officials], they [gongwuyuan] are different because of their set functions and responsibilities. . . . [G]ongwuyuan exist on the strength of their vocation [yewu], while fengjian guan [feudal officials] exist on the strength of their power and position [quanli]. . . . [I]n order to implement a modernized and professionalized system of officials we must implement classification on the basis of function, work, responsibilities, and qualifications.⁽²¹⁾

Gongwuyuan prevailed in other ways as well. Of all the terms in common usage from the late 1920s through the late 1940s to denote state administrators, gongwuyuan was the one that was eventually incorporated into the Guomindang state's own formal representation of itself. The two most important laws on public personnel to come out of the 1930s were the Ministry of Personnel's Gongwuyuan renyongfa (Law on the appointment of public servants) of 1933 and the Gongwuyuan kaojifa (Law on the assessment of public servants), promulgated in 1935.⁽²²⁾ The former law attempted to establish minimal criteria for standardization of ranks and quality control; the latter tried to depersonalize and defactionalize inner Ministry workings by instituting "objective," "scientific," numerically assessable criteria for work performance as the basis for promotions. Although both of these laws were but weakly implemented, the language they employed reveals a great deal about the wing of the party-state that captured policy on national administrative personnel: virtually all state administrators of any prestige (with the exception of waijiaoguan and lingshiguan, who were given their own separate section) were generally denoted as gongwuyuan--thus connoting science, objectivity, functional application, and service as the bases upon which the Guomindang state wished, at least in theory, to reconstitute itself.

Did this ideal of functional, depersonalized, "scientific" administration ever translate into reality? Certainly, most contemporary observers thought not. Guomindang state administration was roundly criticized for being faction riven, corrupt, and ineffective--and the administrative efficiency reformers were among the most vociferous in these complaints. But at least in pockets of the Guomindang state, the ideas of objectivity, technical competence, and de-politicization took hold and began to be implemented in the mid-1930s. Technically oriented ministries and commissions, such as the Ministry of Finance, voluntarily availed themselves of the Examination Yuan­held examinations to staff new departments and seem quietly to have institutionalized norms of technical competence and on-the-job performance for personnel up until the outbreak of the Sino-Japanese War.⁽²³⁾

By the end of the 1930s, those administrators in the Guomindang state with any claim at all to some sort of technical or functional expertise were invariably denoted by a label that stood in implicit opposition to "politics" and partification: gongwuyuan, gongzhi renyuan, zhiyuan, shiwu renyuan, xingzheng renyuan. Although it was preferred for legal usage, gongwuyuan did not establish hegemony over these other, similar terms. But when taken as a set of relatively new terms with nearly identical connotations, gongwuyuan and related terms effectively replaced guan and its derivatives for a very important tier of the Guomindang state: those in middle to upper middle positions in the fairly prestigious central government organs. Although certainly serving that function, gongwuyuan et al. did not merely connote a wish list of positive attributes for Weberian, depersonalized bureaucracy bandied about by a small group of administrative reformers. New recruits into the Guomindang state organizations in the late 1930s were very much aware of a real gap between generalists and politicos (wenguan and zhengwuguan) on the one hand, and technical or functional specialists (gongzhi renyuan and shiwu renyuan) on the other.⁽²⁴⁾ Thus, these new, still non-standardized terms also denoted a new breed of technocratic administrator that was increasingly important in the central state administration of the Guomindang state in the 1930s and 1940s, one that quietly staked a claim to a quasi-autonomous, professional sphere of expertise.

Since the term ganbu (cadre) later acquired such clear hegemony over all other terms to denote party and state administrators in the People's Republic of China, there is a strong tendency anachronistically to project its dominance in leftist and revolutionary circles backwards in time to at least the early 1920s, if not before. In fact, there is a great deal that is unknown and elusive about ganbu: there is no agreement as to the exact route by which the term came to China, and, even more surprisingly, its usage in the 1920s was rare. Virtually all sources agree that the word ganbu originates with the French cadre, but there the consensus ends. The Hanyu wailai ci cidian, the standard reference on foreign terms that found their way into Chinese, states quite clearly that ganbu came into Chinese via the Japanese term kanbu (a word in which the kanji is identical to the Chinese characters); the Japanese term kanbu, in turn, had its origins in the French "cadre." A number of other sources, however, insist equally adamantly that the term came into China through the Soviet Union.⁽²⁵⁾

The connotations of ganbu are equally ambiguous, ranging from the quite restricted to the extremely broad. In the Chinese Communist lexicon, ganbu has been variously used to denote the "backbone" (or core) of the Party (usually in the context of revolutionary activism), "leader(s)" (in the Party: with or without administrative position), and, in its broadest and most common usage, "all those who are working personnel in Party, State, or military units with a definite culture, a level of specialization and capability . . . exclusive of soldiers, handymen, and workers of that sort."⁽²⁶⁾ Ganbu's shift in denotation from minimalist (the "backbone" of the Party) to maximalist (encompassing virtually all personnel with authority or expertise in Party, state, army, or units of economic production, be they Party members or no) was a very large denotative leap indeed, and the evidence indicates that the Chinese Communist Party only reluctantly came to the broadened connotations of the term over the course of the 1930s and early 1940s.

Surprisingly, ganbu was seldom used in the 1920s, nor does it appear to have been a term primarily associated with the Chinese Communist Party. Sun Yat-sen used the term once or twice in the early 1920s, and those in both the CCP and the GMD, as the respective heirs to Sun Yat-sen's legacy of nationalism and revolution, occasionally called their party members and activists ganbu. But ganbu appeared to be closely associated with no particular side during the 1920s, and it was used but infrequently by both. Sun, the Guomindang, and the Chinese Communist Party much more regularly used other terms to refer to party members and activists during this decade of revolution and counterrevolution. The Guomindang's official documents issued at the conclusion of Plenary Sessions of Party conferences from January 1924 through March 1938 without fail employ the term dangyuan (literally, "party member") to refer to those in the Guomindang,⁽²⁷⁾ and until the early 1930s, dangyuan and tongzhi (comrade) were the preferred terms in the Chinese Communist Party as well. On the extremely rare occasions during the 1920s when it was necessary to refer to state administrative personnel (as opposed to party members and activists), the terms employed by the Chinese Communist Party and the Guomindang were remarkably similar: renyuan, zhiyuan, and, in the case of Mao's writings from the late 1920s and early 1930s, gongzuo renyuan ("working personnel") and zhengfu gongzuo renyuan ("government working personnel").

As a revolutionary party locked in mortal combat with a rival party that had gone on to establish a party-state (complete with organs of civil administration), the CCP was clearly uncomfortable with the concept of government administration, except perhaps as an epithet to hurl at the Guomindang, and for as long as possible it avoided directly incorporating the concept into revolutionary language as a neutral or positive term. The Chinese Communists' distaste for anything that smacked of the status and institutionalized inequality of "officialdom" and "bureaucracy" went even further than that of Sun Yat-sen. Mao Zedong was perhaps extreme in his condemnations of bureaucracy, but nonetheless extraordinarily influential:

We must not be bureaucratic in our method of work in mobilizing the masses. . . . This great evil, bureaucracy [guanliao zhuyi], must be thrown into the cesspool, because no comrade likes it. . . . [O]ne of the manifestations of bureaucracy is slackness in work due to indifference or perfunctoriness: another manifestation is authoritarianism. Superficially, authoritarians are not slack in their work and seem to be hard working. Actually, to develop co-operatives in an authoritarian way will bring no success; even if they are apparently developed for the time being, they cannot be consolidated. . . . [W]hat we need is energetic agitation to convince the masses and, according to specific circumstances and the real feelings of the masses, to develop the co-operatives, promote the subscription to the bonds, and do all kinds of work for economic mobilization.⁽²⁸⁾

Although the long years of struggle tended to collapse revolutionary and military roles, the political necessity of consolidating the Jiangxi base area in the early 1930s called for economic and civil construction. Economic and civil construction, in turn, required that Party members, comrades, and fighters in the Red Army also begin to take on a de facto third role of administrator, which required new labels and explicit ideological and practical training (xunlian). It was at about this time that the term ganbu began to appear regularly along with tongzhi (comrade) or dangyuan (party member) to denote those in the Chinese Communist Party. But during the Jiangxi Soviet period, there was still a great deal of confusion about who exactly was a ganbu and what ganbu-dom involved. Significantly, in the early 1930s the CCP was much clearer about its military organization than its nascent civil administration. When Party training courses for ganbu were set up in the early 1930s, the term ganbu was still so vague that draft training manuals had to classify ganbu grades (as upper, middle, or lower) by using military rank as the common frame of reference. Middle-level cadres were the equivalent of platoon commander (paizhang) to assistant company commander (fuyingzhang), while upper-level cadres were equated with the ranks of company commander (yingzhang) up to assistant regiment commander (futuanzhang).⁽²⁹⁾ Even while taking on administrative functions, ganbu continued to be linked very closely (if not conflated entirely) with heroic leadership in both the Party and the Red Army.

Mao himself used ganbu with regard to economic construction during the Jiangxi period: after warning against the pernicious effects of bureaucracy and authoritarianism, in the essay cited above, he then went on to note briefly the importance of ganbu as well as to stress that with proper values and training, anyone could become a cadre:

They [the cadres] are the commanders on the economic front, while the broad masses are the soldiers. People often sigh over the shortage of cadres. Comrades, is there really a shortage of cadres? . . . Give up your erroneous viewpoint and you will find cadres standing right in front of you.⁽³⁰⁾

But even as ganbu acquired currency during the early 1930s, it was not used exclusively for Soviet administrators, and still less did it replace tongzhi as a general term of address or reference to the Party faithful. Even in the arena of economic construction, zhengfu gongzuo renyuan was used as late as January 1934 in Mao's essay "Our Economic Policy": "Financial expenditure must be governed by the principle of economy. . . . [W]e should make it clear to all government personnel [zhengfu gongzuo renyuan] that corruption and waste are the greatest crimes."⁽³¹⁾

After the Long March and the establishment of the Shaan-Gan-Ning base area, ganbu increased in usage and was deemed important enough as a category in the Communist lexicon to have long sections of speeches devoted to its clarification in the late 1930s and early 1940s. Ironically, as the CCP explicitly turned to Stalin's oft-cited dictum that in a revolution "cadres are everything" to establish the term's revolutionary pedigree and impeccability, it sinified and expanded the connotations of ganbu far beyond its original connotations in the Russian.⁽³²⁾ In "The Question of Cadres" from the speech "Strive to Win Over Millions" of May 1937, Mao clearly laid out the specific connotations of ganbu-dom:

To guide a great revolution, there must be a great party, and many excellent cadres. . . . [T]hese cadres must understand Marxism-Leninism, they must have political insight and the ability to work, they must be full of the spirit of self sacrifice, capable of solving problems independently. They must remain firm in the midst of difficulties and work loyally and devotedly for the nation, the class and the Party. . . . [T]hese people must not be tainted with selfishness, individual heroism or vaingloriousness, indolence or passivity or arrogant sectarianism: they must be the selfless heroes of the nation and the class. . . . [B]eyond a doubt we ought to acquire these qualities in order to remould ourselves into better people and raise ourselves into better people and raise ourselves to a higher revolutionary level. . . . [O]ur revolution depends on the cadres, just as Stalin has said, "cadres are everything."⁽³³⁾

A year later, as the Sino-Japanese War widened and the Communist leadership continued to stand by the ideal of a broad anti-Japanese united front, Mao further widened the concept of ganbu to encompass those outside the Party and to warn that the line of "employing only the worthy" ought to prevail over the erroneous line of "employing only the near and dear."⁽³⁴⁾ But even while using Stalin as to justify the importance of ganbu, Mao ironically re-invested the term with several of the key norms that had animated the imperial ideal of guan: lack of overt functional specialization, the implicit belief that understanding of key texts would enable the individual to solve problems independently, and the assumption that individual ganbu would stand as walking models of virtue and correct thoughts. In other respects, of course, ganbu stood in stark contrast to imperial guanli: the content of the sacred texts of Confucianism and Marxism-Leninism could not have been more dissimilar, and the Maoist variant of Marxism-Leninism remained deeply suspicious of hierarchy and status differentials.

But resistance to overt specialization and functional divisions of labor and status and a corresponding tendency to collapse all revolutionary roles and actions into one term (or one individual, since ganbu referred to particular individuals as well as to a collective group) paradoxically hobbled the Chinese Communist Party after it actually took power in the early 1950s. Throughout the 1940s and into the 1950s, ganbu had expanded enormously in denotative scope, from a small group of "backbone" Party activists to well over half of those employed by the Party, state, army, or CCP-dominated units. But ganbu's set of revolutionary connotations remained stubbornly unchanged. The concept of ganbu continued to require purity of thought, knowledge of Marxism-Leninism, practical application, and continuous revolutionary heroism. The unchanged demand that ganbu be simultaneously heroic revolutionaries and practically oriented administrators (or "Red" and "Expert," in Schurmann's formulation) without accompanying indications of how two such disparate roles might be combined once the revolutionary tasks of winning the civil war and consolidation were accomplished put ganbu under a tremendous amount of uncertainty and stress in the late 1950s and again in the decade of the Cultural Revolution.

Interestingly, during exactly the same period that ganbu was being standardized and expanded in the CCP (the late 1930s and early 1940s), the ideological opposites of the CCP on the far right wing of the Guomindang attempted to bring the term back into regular usage through the xunlian (training) movement to partify and reinvigorate Guomindang Party, state, and military organizations. After the shocking defeats of the summer and fall of 1937, the Guomindang's subsequent retreat to the far interior, and the accompanying loss of its economic heartland and political base in the Jiangnan, the domestic political balance of power in China was decisively altered, and the Guomindang found itself in an infinitely weakened and highly stressed position vis-à-vis both semi-independent military commanders and the Chinese Communist Party. Given this position of relative weakness, the Guomindang Right tried to ensure the loyalty of questionable military units through short- to medium-term Party training courses (xunlianban) that combined Party indoctrination with some amount of military and physical training. The xunlian program, however, did not end with the military. It swiftly expanded from military units to Party organizations and then eventually passed to state units as well.

The attempt to bring xunlian into the Guomindang state in the late 1930s and 1940s sharply reversed the norms of technical competence, depoliticization, and Weberian bureaucracy that had been quietly institutionalizing in parts of the Guomindang state over the course of the previous decade. The xunlian program--which conflated Party loyalty and discipline, nationalist resistance, "practical" application, physical training, and military mobilization--was astonishingly similar to the rhetoric and policies of the Chinese Communist Party during the same period. Like most language during the Republican period, the vocabulary of xunlian was not standardized: the targets of xunlian were variously referred to as both shou xunlian renyuan ("personnel receiving training") and xunlian ganbu. However, the language surrounding the Guomindang Right's xunlian was remarkable for its revival of ganbu, which took on a set of connotations very close to the contemporary CCP usage of the term. Thus, Wang Dongyuan, the chief figure associated with central Party training for high-level military and civil administrators during the Sino-Japanese War, wrote in a 1939 pamphlet titled Ganbu xunlian wenti ("Questions on cadre training"):

Why do we need cadres [ganbu]? Upon whom, then, will our mission of reviving the race and building up a new country depend [if not cadres]? Without a doubt, it is necessary to have a large number of leading cadres . . . to provide leadership, to unite our forces, to rid ourselves of difficulties and obstructions, and to carry out national construction. . . . [C]adres are the leaders of the revolution . . . and although we currently lack cadres, we must look for cadres from party, state and army . . . to lead and organize the broad masses of people.⁽³⁵⁾

Such sentiments were virtually indistinguishable from those of Mao Zedong at nearly the same point in time.

Although the Guomindang's xunlian program has not yet been fully investigated and evaluated, it appears that its exercises in partification and indoctrination fell far short of their intent. Although some upper-level administrators in party, state, and army units were temporarily detached from their organizations to attend centrally run xunlianban, the vast majority of the xunlian programs were devolved to the units themselves, thus allowing for party, army, and state organizations, if so inclined, to dilute or deflect large parts of xunlian, to conform with the letter of the order to carry out xunlian while minimizing its spirit, or even to turn the xunlian process into an arena for factional politics. The highly uneven implementation of xunlian in the assorted units of the Guomindang party-state apparatus is at least in part reflected in the different names adopted by organizations undergoing xunlian.

Central state ministries, particularly those with large numbers of technical personnel, tended to keep some version of renyuan or gongwuyuan in the titles of their xunlian courses, while provincial and local party and military organizations often incorporated ganbu into their respective designations for xunlian. The Ministry of Communications convened a Jiaotong jishu renyuan xunliansuo (Training group for communications technical staff), the Ministry of the Interior a Weisheng renyuan xunliansuo (Training group for hygiene personnel), the Army and Government Ministry (Junzheng bu) a Kuaiji renyuan xunliansuo (Training group for accounting personnel). The Ministry of Finance organized the Quanguo caiwu renshi xunliansuo (National training group for finance personnel), for which it ultimately produced its own set of guidelines, titled the Caiwu renyuan xunlian dagang, suggesting that at least the Ministry of Finance fought hard and successfully to maintain control over the process of xunlian. Military and party units at provincial and local levels, on the other hand, exhibited no such consistency in terminology. A fair number brought the term ganbu into their xunlian organizations: for example, the Xi'nan xibei youji ganbu xunlianban (Training group for southwest and northwest guerilla cadres), the Yunnan dangwu ganbu renyuan xunlianban (Training group for Yunnan Party affairs cadre personnel), the Guangdong difang xingzheng ganbu xunliantuan (The training league for Guangdong local administrative cadres), and the Anhui zhengzhi junshi ganbu xunlianban (The training group for Anhui political and military affairs cadres). Other local and provincial military and party organizations stuck with the more neutral-sounding renyuan or gongwuyuan designation: the Guangxi xianzheng gongwuyuan xunlianban (The training group for Guangxi County government public servants), which was closely identified with the semi-independent and progressive general, Bai Chongxi, and the Fujian xianzheng renyuan xunliansuo (Training group for Fujian County government personnel).⁽³⁶⁾

In periods of extreme stress, when the right wing of the Guomindang was able to move out from its base in propaganda units to capture the national personnel agenda for party, army, and state, there was little that state organizations could openly do to stem the onslaught. But the weak institutional capacity of the Guomindang Party to implement vigorously either xunlian or ganbu-ification throughout all the units nominally under its hegemonic control meant that at least some technocratically inclined state organizations like the Ministry of Finance were able quietly to insulate the organization from the worst excesses of partification and xunlian. Regardless of how much the far right of the Guomindang would have liked to "cadre-ify" its military, party, and state organizations during the politicized 1940s, ganbu (with its accompanying set of connotations) never succeeded in fully replacing renyuan, gongwuyuan, and zhiyuan, and in many units it made insubstantial headway. The parallel evolution and usage of ganbu and gongwuyuan in the Republican period mirrored the organizational strengths and weaknesses of the Guomindang and Chinese Communist Party in the 1930s and 1940s and encompassed double paradoxes that would only become clear in the post-1949 era. Strangely, the weakness that led to the Guomindang's downfall in the civil war of the late 1940s also permitted the survival of concepts and terms allowing for professional quasi-autonomy of technocratic state administrators that would prove to be instrumental in the consolidation of the Guomindang party-state on Taiwan. And the organizational strength and cohesion that enabled the Chinese Communist Party to collapse so many uneasily reconciled connotations and values into ganbu as it made that term the standard for administrators hamstrung the post-1949 People's Republic into demanding simultaneous high performance in the oil-and-water roles of revolutionary hero and objectively competent, technically proficient administrator.

Although it is tempting to consign the politics of labelling state administrators in Republican China to the category of the historically interesting but contemporarily irrelevant, the reform era of the 1980s in the People's Republic of China unleashed a series of debates about the nature of the state, the efficacy of state administration, and the potential reform of state personnel systems that echoed a similar set of discussions in the 1930s and 1940s. In the increasingly open climate of the mid- to late 1980s, as political and structural reform began to be discussed openly, academics, advisors, and even working groups within the Chinese Communist Party turned to topics that had been taboo since 1949: the ineffectiveness of the existing cadre system, the need to implement a genuinely "public" administration, bringing scientific management into state organs and enterprises, and the unlinking of state (administration) from party (politics).

Journals, handbooks, policy documents, and monographs on public administration and reform of state personnel policy suddenly began to be published in large numbers in the late 1980s.⁽³⁷⁾ Much of this new wave of publication was uncannily reminiscent of the debates surrounding state service and personnel policy in the Republican period. Administrative reformers in the 1980s revived language, terms, and concepts that had lain dormant since the late 1940s. The sheer number of publications from the late 1980s that contain the term gongwuyuan on the title page indicates the revival of gongwuyuan as a positively valenced denoter for state administrator (in subtle opposition to ganbu). Wenguan was also brought back in and used variously to refer to civil service systems abroad (its meaning during the Republican period) as well as for historical discussions of the Guomindang personnel system during the 1930s and 1940s. Drawing on a mix of foreign models and the glories of China's indigenous tradition of statecraft, administrative reformers in the 1980s openly pushed for programs for civil service reform that, while radical in the context of politics and administration in the People's Republic of China, exactly replicated those of their Republican predecessors in the 1930s and 1940s. The themes thus raised included the need to ensure fairness and de-politicization in state service, security of tenure, unbiased and "scientific" methods of evaluation and promotion, and the desirability of instituting open and public civil service examinations as the prime method of entry into state service in the first place.

As was the case in the Republican period, the mid- to late 1980s in the People's Republic of China saw language and terminology in a tremendous state of flux in which old, previously standard terms for administrators (ganbu in the 1980s, guan and its derivatives in the 1930s) were still used, but increasingly took on negative connotations of status differentials, obstruction, corruption, and lack of responsiveness. The frequent use of gongwuyuan on the part of administrative reformers in the mid- to late 1980s in the People's Republic of China was not accidental. This group subtly challenged the viability of the whole concept of ganbu (and, by extension, the entire political system of the People's Republic of China) by consistently using an alternate term in conjunction with programs for reform which all could agree were badly needed. But ganbu and the authority (and ultimately, the military) structure that stood behind it were still powerful enough in the 1980s that the term was not usually taken on by the administrative reformers directly. In fact, in the new dictionaries of terms, ganbu and its combinations were allotted as much in terms of space and explication as was gongwuyuan. Further, Party conservatives showed themselves entirely capable of fighting back and defending ganbu by sticking with the Dengist line of the early 1980s: conceding the need to reform the ganbu system itself through repeated recourse to the rhetoric of "younger, better educated, and more technically proficient cadres," thus appropriating the program the administrative reformers wished to apply to a new personnel structure of gongwuyuan outside the status quo system of cadres.⁽³⁸⁾

In the aftermath of 1989, discussions and materials that openly suggested the replacement of cadre reform with public servant reform were shelved along with all other open discussions of political and structural change. But gongwuyuan, with all of its connotations, lives on even in post-1989 editions of dictionaries of terms, and it remains to be seen how it will fare in its post-Deng afterlife. Given the stubbornly intractable problems associated with state service and administration and the Chinese Communist Party's inability to deal systemically with those problems through ganbu, it is a fair guess that gongwuyuan and associated terms will re-ascend and be an important conceptual component of any structural and administrative reform that China undergoes in the future.

1. The key ideas in this opening section--that words are denotative and connotative, and that words are positively or negatively valenced--were stimulated in discussions with Martin Landau in a seminar on politics and theory he taught at the University of California, Berkeley in the autumn of 1986. At the time, I little imagined that I would ever be actively utilizing any of this material in my own research, and it is a testament to Landau's powerful insights and unforgettable teaching methods that his ideas resonate so strongly in a field far removed from his own so many years later.

2. The arguments that follow about the changing meaning of specific terms for state administrators in Republican China is based on several years of research on bureaucratic institutions, particularly on the Guomindang's Examination Yuan, Ministry of Finance, and Ministry of Foreign Affairs. The sources for this research have included publicly available handbooks and documents produced by the bureaucratic organizations themselves, archives available in the No. 2 Historical Archives in Nanjing and in the Guoshiguan in Taipei, publications from the Republican period such as Xingzheng xiaolü, memoirs by former officials, and extensive interviews with former GMD bureaucrats in both the People's Republic and Taiwan. In some cases I offer complete citations to back up my claims. In others I will refer the reader to either my dissertation or my article on the Examination Yuan previously published in Modern China. In still other cases specific documentation will be provided in the manuscript I am currently writing entitled "Bureaucracy, Civil Service, and the Twentieth-Century Chinese State."

3. The definitions of these terms that follow are drawn from the Oxford English Dictionary, which states that the proper adjectival forms of the verbs "denote" and "connote" are "denotive" and "connotive." However, throughout this essay, I will use the alternative forms "denotative" and "connotative," which are more commonly used in the United States.

4. These paraphrased definitions are all drawn from the compact version of the Oxford English Dictionary (1971). The specific citations of usage in logic are drawn from John Stuart Mill's mid­nineteenth-century formulation, reproduced in the Oxford English Dictionary.

5. It was Harry S. Truman who stated somewhat acerbically that "a statesman was a politician who had been dead for ten years." Interestingly, given the current debates on appropriate use of gender in language, many groups now shun the term "statesman" altogether.

6. See the Compact Edition of the Oxford English Dictionary for examples.

7. The Chinese Communist Party was a little less obviously internally divided than was the Guomindang during most of the Republican period.

8. The literature in Chinese and English on officialdom and examination systems in imperial China is vast. For the following discussion I draw particularly on Ichisada Miyazaki, China's Examination Hell (New Haven: Yale University Press, 1981); Djang Chu, A Complete Book Concerning Happiness and Benevolence: A Manual for Local Magistrates in Seventeenth Century China, trans. Huang Liu-hung (Tucson: University of Arizona Press, 1984); and John Watt, The District Magistrate in Late Imperial China (New York: Columbia University Press, 1972).

9. 10. Sun Zhongshan, "Dangyuan xu xuanchuan geming zhuyi" [Party members ought to propagandize revolutionary ideology], Guofu Quanji, vol. 3, ed. Yu Youren, (Taipei: Zhongyang wenwu gongying chubanshe, 1957), p. 204.

11. Sun Zhongshan, Guofu Quanji, vol. 3, p. 263.

12. Sun Yat-sen, The Fundamentals of National Construction (Taipei: China Cultural Service, 1953). Published as a Chinese and English text.

13. The attempted institutionalization of a distinction between zhengwuguan and shiwuguan was felt by many to be an important component of the institutionalization of the Guomindang state in the 1930s, including the Guomindang's own Examination Yuan, administrative reformers, and opposition academics such as Quan Duansheng. The Examination Yuan's Ministry of Personnel repeatedly drew distinctions between political appointees and "regular" bureaucrats. For examples, see assorted regulations in the Quanxu bu nianjian (1931) and the Ministry of Personnel's Law on civil service appointments (1933), reproduced in English in Kwei Chungshu, ed., The Chinese Yearbook (Shanghai: Commercial Press, 1936), pp. 268-69. In addition, the 1930s saw the coalescence of an informal group of individuals, both in and out of government service, whom I dub the "Administrative Efficiency School." This loose coalition of individuals, composed of Gan Naiguang, his immediate protégés, an ever-widening circle of academics and specialists on public administration, and a few members of the Examination Yuan itself, wrote for a series of publications in the mid-1930s. Revealingly titled Xingzheng xiaolü [Administrative efficiency], the series appeared bi-weekly in 1934-36 with a companion volume in English called The Chinese Administrator, which was published irregularly in the mid-1930s. In mid-1936, the publication switched its name and format to Xingzheng yanjui yuekan [Administrative studies monthly] and came out as a monthly between mid-1936 and the outbreak of the Sino-Japanese War. The "Administrative Efficiency School" was very much influenced by prevailing currents in American public administration and was therefore, among other things, very much concerned with the institutionalization of a separation between "politics" (and political officials, zhengwuguan) and "administration" (xingzheng, to be implemented by shiwuguan). These sentiments can be found in most of the articles in these publications. For a more detailed review of the activities of the "Administrative Efficiency School," see my paper "The Cult of Administrative Efficiency: Myth and Statecraft in the Chinese Republic, 1912-37," unpublished. See also Qian Duansheng's (Chi'en Tuan-sheng) acerbic comments on the lack of a separation between political and administrative offices in The Government and Politics of China (Cambridge: Harvard University Press, 1961), pp. 220-21 and 243-44.

14. The use of wenguan zhidu as the term to translate Western civil service and examination systems was by no means standard, either in the Republican period or in the 1980s in the People's Republic of China. The term gongwuyuan zhidu ("public servant system") was used at least as often--sometimes consciously, and often subconsciously, as the terms have never been totally standardized with respect to their denotations. For specific examples of the usage of wenguan zhidu and gongwuyuan zhidu to denote Western civil service, see multiple articles in Xingzheng xiaolü and Xingzheng yanjiu yuekan, passim.

15. The Guomindang state's own acceptance of the differences between its "regular" bureaucracy and the separate (and elitist) Ministry of Foreign Affairs was formalized in the Guomindang's incorporation of the Ministry's own denotations of its members as lingshiguan and waijiaoguan in the Ministry of Personnel's Law on civil service appointments (1933), which replaced these terms. See also Chen Tiqiang, Zhongguo waijiao xingzheng (Kunming: Shangwu Press, 1942), pp. 141-61 and Julia Strauss, chap. 8 in "Bureaucratic Reconstitution and Institution Building in the Post-Imperial Chinese State: The Dynamics of Personnel Policy, 1912-45" (Ph.D. diss., University of California, Berkeley, 1991).

16. These sentiments were shared by administrative reformers and Chiang Kai-shek alike. Virtually every issue of Xingzheng xiaolü contained at least one article that discussed at some length the poor training and lack of commitment of difangguan as one of the major factors in the weakness of local government. Chiang Kai-shek, for his part, resorted to morally based exhortations for difangguan to clean up their acts. See Chiang Kai-shek's speech of July 12, 1932, "Xianzhang shi zhengfude jiben liliang" [County chiefs are the base strength of government], in Zhang Qijun, ed., Xianzongtong Jianggong quanji, vol. 1 (Taipei: Zhongguo wenhua daxue, zhonghua xueshuyuan, 1984), pp. 646-52.

17. Chiang Kai-shek, "Xiandai xingzheng renyuan xuzhi" [What modern administrative personnel need to know], speech given on 20 March, 1934. In Xianzongtong Jianggong quanji, vol. 1, p. 827. The images of officials, be they military or civil, as "the father and mother of the people" was further elaborated by Chiang a year later in a speech given on 7 September 1935, titled "Xingzheng renyuan tuanjing ganbu zhi diwei yu zeren" [The positions and responsibilities of administrative personnel and police corps cadres]. In this speech, administrative personnel were explicitly called renmin de fumu, while cadres of the police corps were denoted as minzhongde daoshi (leading teachers for the masses). In Xianzongtong Jianggong quanji, vol. 1, p. 998.

18. See previous references to Chiang Kai-shek's speeches, especially "Xiandai xingzheng renyuan xuzhi," and various drafts of the Quanzu bu renyongfa (The Ministry of Personnel's Law on civil service appointments). For Guomindang Party usage of these terms, refer to Zhongguo Guomindang lici huiyi xuanyuan ji zhongyao jueyi huibian [A selected compilation of the Guomindang Party plenary speeches and important resolutions] (Chongqing: n.d. [1939?]). For the Ministry of Finance, see Kong Xiangxi, Caizheng bu shinianlai zhi caiwu renyuan [Finance personnel in the Ministry of Finance over the past ten years] (Chongqing: Ciazheng bu, 1943), references throughout.

19. See "The Cult of Administrative Efficiency."

20. For a fuller description of the Examination Yuan and the Ministry of Personnel, see my "Symbol and Reflection of the Reconstituting State: The Examination Yuan in the 1930s," Modern China 20 (April 1994): 211-38.

21. Xie Tingshi, "Lun wenguan guandeng guanfengde gaiding wenti" [On the problems of reforming civil service grades and salaries], Xingzheng yanjiu yuekan 1 (1936).

22. These laws are reproduced on virtually every scrap of paper that the Ministry of Personnel issued from the mid-1930s on. For a fairly recent compilation, see Quanxu faigui huiban (Taipei: Kaoshi yuan, 1979), pp. 329-32 and pp. 719-22.

23. See chap. 5, "The Ministry of Finance: The Range of Institution Building Strategies," in my manuscript "Strong Institutions in Weak Polities: Personnel Policies and State Building in China," for a much fuller discussion of these trends. An unrevised version of this argument is also available in chapter 6 of my dissertation.

24. These specific terms, and their opposition to each other were cited by Wu Chengming in my interview with him on 21 May 1988, Beijing.

25. Yang Youwu and Wang Zhenchuan, eds., Zhongguo gongwuyuan baike cidian [A dictionary of terms for China's public servants] (Beijing: Guoji wenhua chubanshe gongsi, 1988), p. 191; Gong Jianhua, Xiandai ganbuxue (Guangzhou: Guangzhou wenhua chubanshe, 1988), p. 11.

26. This set of connotations is drawn from the definitions of ganbu as listed in Yang Youwu and Wang Zhenchuan, Zhongguo gongwuyuan baike cidian, Xinmingci cidian (Shanghai: Shanghai chunming chubanshe, 1955), and Zhongguo wailaici cidian.

27. I base this supposition on the contents of Zhongguo Guomindang lici huiyi xuanyuan ji zhongyao jueyi huibian. This three-volume set contains all of the official conclusions of GMD Party plenary sessions from January 1924 through March 1938.

28. Mao Zedong, "Bixuzhuyi jingji gongzuo" [We must attend to economic work], Mao Zedong xuanji, vol. 1 (Beijing: Renmin chubanshe, 1961), pp. 118-19 and Selected works of Mao Tse-tung, vol. 1 (London: Lawrence and Wishart, 1954), pp. 135-36.

29. These distinctions are laid out in a short document from the Jiangxi Soviet period titled "Ganbu zhengzhi jiaoyu jihua cao'an" [A draft plan for cadre political education] (n.d. [early 1930s]), Chen Cheng Collection Reel 8, Document 31.

30. Mao Zedong xuanji, vol. 1, p. 119 and Selected Works of Mao Tse-tung, vol. 1, p. 135.

31. "Womende jingji zhengci" [Our economic policy] in Mao Zedong xuanji, vol. 1, p. 129 and Selected Works of Mao Tse-tung, vol. 1, p. 145.

32. In Russian, the term for "cadre" connotes "core group" and also "personnel" in the general plural (as opposed to an individual). The speech of Stalin's most often cited as justification for the importance of cadres in a Communist system was the "Address to the Graduates from the Red Army Academies," delivered in May 1935. The second most cited of Stalin's references to cadres was a brief mention in the "Report on the Work of the First Central Committee to the 18th Congress of the Communist Party of the Soviet Union," delivered in March of 1939. These speeches can be found in The Complete Works of Josef Stalin. I suspect that these remarks of Stalin's were at least in part taken out of context: beyond these two relatively minor mentions, Stalin does not usually wax poetic about cadres or, in fact, refer to them much at all. However, Stalin's precedent in using the term was extremely important for the CCP in the late 1930s and early 1940s, and enabled the Chinese Communist leadership to begin to standardize the term around a much broader set of connotations than existed in Russian.

33. Selected Works of Mao Tse-tung, vol. 1, pp. 279-80 and Mao Zedong xuanji, vol. 1, pp. 267-68.

34. Even during the Yan'an period that the CCP later perceived as its "Golden Age," the CCP had to contend with a basic contradiction that plagued both the imperial state and the Guomindang: the age-old conflicting imperatives of a state that required norms of universalistic merit vs. a society grounded in particularistic norms of ganqing (human feeling) and giving absolute priority to caring for those near and dear. In "The Role of the Chinese Communist Party in the National War," a report given to the Central Committee in October, 1938. Selected Works of Mao Tse-tung, vol. 1, p. 252 and Mao Zedong xuangi, vol. 1, pp. 514-15.

35. Wang Dongyuan, Ganbu xunlian wenti (Chongqing[?]: Zhongyang xunliantuan dangzheng xunlianban diwuqi jiangyanlu, 1939), pp. 4-5.

36. This list of xunlian organizations was drawn from several sources, including Ganbu xunlian wenti and Shinianlaizhi caiwu renyuan.

37. This sudden proliferation of published material is far too vast to allow a complete bibliography. Some representative publications include: Zhongguo xingzheng guanli [China administration management], a periodical which began publication in early 1986. For handbooks and reference dictionaries, see Zhongguo gongwuyuan baike cidian; Li Shengping, ed., Gongwuyuan baike cidian [A dictionary of terms for public servants] (Beijing: Guangming ribao chubanshe, 1989); Sun Weiben and Shao Qihui, eds., Zhonghua renmin gongheguo zhengwu da cidian [A great dictionary of political affairs for the People's Republic of China] (Beijing: 1988), which contains a long section on personnel (renshi) terminology; Zhang Feng, ed., Dangdai Zhongguo baike da cidian [A great dictionary of terms for contemporary China] (Beijing: Dang'an chubanshe, 1991). Historical description, scholarly investigation and public policy proscriptions are blended in Su Meifeng and Zhuo Songcheng, eds., Zhongguo gongwuyuan zhidu [China's public servant system] (Wuchang: Huazhong ligong daxue chubanshe, 1988); Zhao Lukuan and Wang Ziping, Renshi guanli he shehuixue [Personnel management and sociology] (Beijing: Zhongguo zhandang chubanshe, 1986); Lin Daizhao, ed., Zhongguo jinxiandai renshi zhidu [China's modern and current personnel system] (Beijing: Laodong renshi chubanshe, 1988), which offers a historical overview of China's personnel systems from 1840 to the present; Liu Junlin and Dai Guangqian, eds., Zhongguo gongwuyuan zhidu jianghua [Discussions on China's public servant system] (Beijing: Nengyuan chubanshe, 1989); and Gong Jianhua, Xiandai ganbuxue [Contemporary cadre studies] (Guangzhou: Guangzhou wenhua chubanshe, 1988).

38. See particularly Xiandai ganbuzue, for the conservative, Dengist position on reform within the cadre system and, Zhongguo gongwuyuan zhidu jianghua, for as close to an open challenge as one could get on the need to reorganize the existing cadre system into public servant system (gongwuyuan zhidu)--involving such unheard of elements as open examinations, position classification, evaluations and bonuses on the basis of legally established standards, and promotion on the basis of evaluated achievement.