Good Policing and the Production of Obscurity

The data on which this article is based were collected through field research conducted between 1999 and 2003 in several neighborhood police “substations” (paichusuo, ) located in Taipei County. This project was aimed at understanding the activities of these policemen in relation to the social and cultural organization of the neighborhoods they policed. As an anthropological study, the primary method of research was participant-observation, which in essence consists of trying to live, as far as possible, with and like the people one wishes to understand. I thus spent most of my time during those years at police stations, in police cars, at police banquets, accompanying police on patrol and household visits, going to staff meetings, observing the mediation of civil disputes, carousing with off-duty policemen and their diverse compatriots, networking with local political figures, riding with volunteer civil patrol groups, and otherwise involved in the order-maintenance sector of the thick fabric of “intermediate institutions” (Reference WellerWeller 1999) that makes up the civic life of contemporary Taiwan.

The working environment of policing is a complex social topography in which a variety of often deeply conflicted relationships between legal instrumentality and intimate solidarity must be carefully managed. A key element in this management is control of information, which has the effect of enveloping the entire institution in a taciturn ethos of secrecy (Reference BittnerBittner 1990; Reference ManningManning 1997).Footnote ¹ It is analytically useful to approach the pervasive qualities of secrecy in police work as not merely the hiding of truth, but also in its positive function as a constructive aspect of social relations; for human relationships are always based in some part on imaginary qualities, “social fictions” (Reference Simmel and WolffSimmel 1950). This kind of approach, concerned more with the production of obscurity than the hiding of truth, allows us to see police information management as an organic aspect of the vocational mandate of the police to facilitate the reproduction of an orderly configuration of social relationships that, in modern nation-states, acquires ultimate validity by reference to a political ideal that is most accurately described as an imagined community (Reference AndersonAnderson 1991).

In accordance with Manning's famous “dramaturgical” characterization of British and American police work, the style of policing I observed in Taiwan had at its core the production of carefully tailored forms of obscurity and misdirection. A beautiful example of this was provided one day in 2002, when I accompanied Patrolman Guo about his duties. Guo was a gregarious young man who had transferred to a neighborhood beat from a riot squad several years earlier. His particular beat contained one of the locality's more renao ( , “hot and noisy”) areas, a street entirely lined with restaurant and market stalls set up illegally on the sidewalks. Police beats in urban Taipei County consist of a couple of city blocks, usually contain about 200–400 households, and are assigned to an individual patrolman who is then held personally responsible for maintaining the area's census records and supplying any other necessary surveillance about local affairs.Footnote ² Whenever spatially located criminal operations (e.g., a brothel or a chop shop for stolen vehicles) are discovered, the patrolman in whose beat they are found is penalized for his failure to have previously detected and reported them. This principle applies to all unlicensed businesses, and thus the entire street market in Guo's beat was technically operating under the largesse of his official liability. This was an intractable situation. There is no way Guo could have single-handedly done anything about the existence of this well-established market; he was simply “on the hook,” as it were, to make sure its existence did not become an irritation to higher levels of the bureaucracy.

Unfortunately for Guo, the market had become an irritation to a new resident of the street, the anonymous author of a concerted campaign of daily e-mails being sent to the city government complaining that when he arrived home from work in the evening he could not get in the front door of his apartment building because it was invariably blocked by crowded restaurant tables. Thus, that afternoon, during the period of his shift dedicated to maintaining census records, Guo stuffed a button-down flannel shirt into his bag and put me on the back of his 125-cc police scooter, and we drove over to the market. He parked down an alley, put the shirt on over his uniform,Footnote ³ and we walked over to one of the restaurant stalls. As we sat down, the proprietor came over and made us comfortable, and Guo ordered two bowls of noodles. The noodles soon arrived, then beer, and soon the table was crowded with unsolicited dishes pressed upon us by the zealously friendly management. We ate and chatted with the husband and wife who ran the operation.

After paying a nominal bill, Guo's informal banter with the management turned substantive. “There have been complaints,” he began apologetically, “about your use of the sidewalk here. Some guy that lives in one of these apartment buildings can't get in his front door when he comes home from work.” “Who is it?” the proprietor immediately asked. Guo replied that he didn't know, but “you guys must know; this kind of situation doesn't just happen overnight.” He went on,

The proprietors agreed to the plan. We then walked all the way down the street, having essentially the same conversation with the other half dozen restaurant owners on the sidewalk in front of the apartment building. In every case the result was the same, a rather stilted acquiescence on the part of the disciplined to a rather convoluted invocation of the law.

As we drove back to the station, Guo volunteered that I had just witnessed a prime example of good policing. The “sentimental feeling,” the ganqing ( ), of the relationship between the substation and the population in its jurisdiction was very good, he said. And it was techniques like the one I had just witnessed that kept it in such good repair. From Guo's perspective, the objectives of policing—maintaining the peace and tranquility of the community, and discharging his immediate orders without generating “a lot of yelling in the street”—were being successfully achieved. I was rather awed by his casual ad hoc orchestration of a streetwide performance of order for the benefit of the state's camera. However, Guo was dismissive of the creativity involved in this technique (and, as we shall see, it was actually a conventional procedure). He was simply killing two birds with one stone, combining the inconvenience of performing order with the issuance of formal sanctions to apply a focused and relatively gentle pressure to targets of civil outrage, motivating them to find the source of the complaint and settle the matter privately. In his narrative to me, he expressed clear personal sympathies with the businessmen who had just fed us: Taiwan is a crowded country, and people selling goods in the streets and sidewalks are working hard to survive, he said. How can you try to keep people from making a living just because they inconvenience you? Nonetheless, he went on, when the order comes down to clear the street, you cannot tell the commanding officers at the county police department, “We've got good ganqing here so I didn't do anything. You have to enforce the law. Dealing with conflicts in the street is simple,” he said. Then he added, “The problems always come in how to write the official reports [gongwen, ].”

The History of Policing in Taiwan

Following Reference FriedmanFriedman's (1977) approach to “legal culture” as the meaningful matrix through which social forces are (or are not) channeled into specifically legal institutions, this article examines the legal culture evident in the contemporary routines of Taiwan's local police. As mentioned, the fact that street patrolmen tend to invoke specifically legal institutions in a relatively circumscribed fashion is by no means distinctive to Taiwan. Modern constabularyFootnote ⁶ policing is universally characterized by a somewhat jaundiced approach to the practical efficacy of legality per se. Thus to better identify the locally distinctive aspects of Taiwanese legal culture, in contrast to institutionally endemic aspects of modern police culture in general, it is useful to begin with some preliminary remarks relating the ethnographic study of police to the broader history of modern policing in Taiwan.

Theoretical discussions of policing address phenomena existing at three relatively distinct levels of abstraction. At their broadest, they take up the issue of social control in general, asking some version of the question Reference ParsonsParsons (1937) named “the problem of order,” i.e., how are normative cultural values operationalized in the practical order of social life? At a more historically particular level, studies of modern police aim to understand how conventional normative modalities of political subjectivity and public order have become established within particular social groups, nation-states, and transnational regimes (an approach that can be pursued from various theoretical vantage points, nicely summarized in Reference GarlandGarland 1990). Finally, at the empirical level, the point of entry into any theoretical discussion is a description of some actually existing police organization, its operations, and the interactive effects (both overt and unintended) that make this institution part of the society in which it is located.

The present article follows the approach to the ethnography of street patrol founded on the work of Bittner, who located the originary mandate of policing in the general problem of order: i.e., police are those authorized to use coercion in answering the call of normative exigency (Reference Bittner and JacobBittner 1974, Reference Bittner1990). Historically, with the rise of nation-states claiming a monopoly on legitimate force within their sovereign territory, a broad spectrum of pre-existing policing institutions was gradually incorporated into, and constrained within, bureaucratically controlled state institutions.Footnote ⁷ Moreover, as champions of modern democratic ideals sought to foster political community based on the liberal mythos of government by consensus, they developed a distinctive organization of police work (first realized in the London Metropolitan Police of 1829) designed to bring the actions of street-level order-maintenance personnel under the influence of prescriptive engineering by administrative agencies and procedural review by judicial institutions, both forms of control motivated toward the goal of reducing policing as much as possible to the pure enforcement of law (Reference KlockarsKlockars 1985).

Nonetheless, this goal has remained chronically out of reach; for where the liberal valorization of law is founded on a mythology of egalitarian consensus, the constitutive mandate of policing has a patriarchal mode of coercive intervention as its defining core (Reference DubberDubber 2005). Thus structurally and genealogically, police and law remain distinct modalities of political power, and achieving the ideal of their integration remains one of the unresolved dilemmas of modern governance (Reference SkolnickSkolnick 1966; compare Reference AgambenAgamben 1998). This situation charges the enterprise of policing democratic society with a set of tensions that are never definitively resolved. They are, instead, managed: by a political division of labor (ranging from the basic distinction between legislative, judicial, and executive powers, to the specifics of administrative procedure and practice) that does its best to minimize tensions while also ensuring that intractable residues of the contradiction between consent and coercion are, as much as possible, obscured from public view (Reference ManningManning 1977). Taking this theoretical framework as the basis for an anthropological investigation into the political culture of democratic Taiwan is a matter of using it to reveal and clarify culturally distinctive aspects of how the intrinsic tensions between legal order and police practice are managed in this particular time and place.

Historically speaking, modern police institutions first arrived in Taiwan as a component of the Japanese colonial regime that replaced the island's Qing Dynasty administration in 1895. By about 1903, the Japanese had established an effective civil police apparatus based on contemporary European standards, its overt goal being the colonial “improvement” of its subject population, and its primary constabulary institution being a dense network of substations (paichusuo, ) located in every rural village and urban neighborhood. These substations functioned as the core institutional node for articulating centralized Japanese political authority into local society, mobilizing for this purpose a subordinate “indigenous” social-control institution known as the Baojia.Footnote ⁸ Over time, as the Japanese regime (quite self-consciously) consolidated its identity as a smoothly functioning “police state,” the substation-Baojia nexus grew in significance as the primary site of interface between state and nonstate political institutions, becoming the backbone of Japanese administration (Reference ChenChen 1975, Reference Chen, Meyers and Peattie1984; Reference LiLi 1996; Reference Ts'aiTs'ai 1990).

When the Chinese Nationalist Party (KMT) assumed control of the island after World War II, it treated the established substation bureaucracy as a turnkey operation, simply replacing the departing Japanese officer corps with a cohort of newly trained Chinese nationalist administrators. The only major structural modification the KMT made to the system was to replace the subordinate Baojia (which had actually been eliminated by the Japanese in early 1945) with the system of police beats (qinqu, ) through which individual police patrolmen were assigned direct jurisdiction over populations of individual families, without additional formal bureaucratic intermediation. This retained the significance of the substations as a key nodal institution in mediating the interface between central political authority and local social order (Y. Reference Huang and WeiyuanhuiHuang 1971; Reference LiuLiu 1990).

During the period of authoritarian, anti-Communist “social mobilization” that lasted from the late 1940s until the early 1990s, the KMT governed Taiwan under a Leninist paradigm that (in some parallels to Japanese colonialism) understood society as an entity possessed of organic political dynamics, which the state should properly capture and cultivate toward political-economic goals. Space here does not permit discussion of the development of the criminal justice system in its role as adjunct to the KMT's aggressive, military-run, party-engineered national-security mission (see Reference ChangChang 1995; Reference MartinMartin 2006; Reference WeiyuanhuiWeiyuanhui 1971; Reference Zhu, Huang and AlZhu 2001). Nonetheless, throughout this period, substations remained defined by their core role as the operational base for a constabulary force responsible for routine local order-maintenance, which endowed them with considerable institutional inertia.Footnote ⁹ Indeed, a layer of something like autonomy is indicated by the classification of substation personnel as “administrative policemen” (xingzheng jingcha, ): the vast majority of their formal interventions in social activities are authorized under codes granting them discretionary authority to apply “administrative penalties” not requiring any further judicial review. Under the martial law regime, the primary such code was the Police Violation Code (PVC), which authorized local police officers to issue summary detentions of up to two weeks in local holding cells, to shut down businesses, and to apply a variety of other forceful sanctions.

In 1991, the PVC was replaced by the current Social Order Maintenance Law (SOML), greatly restricting the scope of “police offenses” in Taiwan. This event is considered one of the constitutive indexes of the regime's shift to democratic practice, and it is from this point that we can turn our focus to the conduct of police work in the current period. The overall bureaucratic organization of police in the Republic of China is established by its Police Law, which places the entire bureaucracy, roughly 80,000 persons, under command of the National Police Administration (NPA), a bureau of the Ministry of the Interior within the central government's executive branch. About half of the personnel employed by the NPA are involved in local administrative policing, working in a centralized national hierarchy that descends from the NPA to the county-level departments, the regional (generally municipal) precincts, and finally to the local substations.Footnote ¹⁰ Under the democratic idiom of governance, the front-line officers working in these substations are clearly identified as servants of the public, expected to provide the service of assisting in the resolution of any situation that transgresses public sensibilities of order and security in such a way as to compel state intervention. The primary resource they are given to achieve this task is the set of relatively modest formal powers specified by legal codes such as the SOML. Thus, following Bittner, we take the work of neighborhood patrolmen as the craft in which legal authority is deployed as a tool (one of many) in the task of resolving emergent moments of disjuncture between the imagined and practical order of civic life.

The Idea of Police in Contemporary Taiwan

A useful trope for relating observations of this sort of on-the-ground craftwork to larger concerns about the overall substantive content of political culture is provided in a recent paper by Reference Junior and MunizJunior and Muniz on the “idea of police” (2006). Briefly summarized, for Junior and Muniz the idea of police is a measure of overall public faith that the state police force will reliably resolve all exigent crises. When the idea of police is strong, people find it unnecessary to “make other arrangements” for the preservation of their personal security. Conversely, the absence of a well-established idea of police is evidenced in the expansion of parochial/informal/sub-public security-maintenance institutions. The major empirical example of the cited article is a differential response to the Brazilian police strike of 1997: in a locality (Minas Gerais) where the striking police remained at the ready in their barracks and ideologically available to the public, there was relatively minimal response to the strike, despite the fact that the police did not in practice do anything. However in another locality (Pernambuco) where the police disbanded publicly and declared their resolve not to respond to any calls, “The idea of police … ceased to exist, and the public responded with remarkable speed to the situation. Within 24 hours, those who could afford it had hired armed guards. Curfew and convoying came into being. …. In ten days, the open organization of private militia had begun” (Reference Junior and MunizJunior & Muniz 2006:253).

As we here take up the idea of police in the Taiwanese context, we are less concerned with measuring changes in degree entailed by political liberalization than we are with understanding the substantive contents of current popular ideology about what police are, and cultural sensibilities of the limits of what they can and cannot be reliably expected to do under their current democratic mandate. To be sure, in respect to measurements of degree, Taiwan's idea of police appears to have declined slightly in the course of liberalization. That is, while Taiwan's democratic transformation is widely considered to have been a “velvet revolution” exemplary in its stability and orderliness (Reference Chao and MeyersChao & Meyers 1998; Reference WuWu 1995), nonetheless, the moment in 1989 at which the military publicly withdrew from its first domestic order-keeping function (border patrol) was immediately followed by the emergence of public anxieties about the “degradation of social order,” which expanded steadily until they reached the scale of a full-blown social movement in 1997 (Reference ChuangChuang 2004; Reference LinLin 1991).

As Junior and Muniz would predict, this erosion of the idea of police has been practically substantiated in an expansive distribution of non-public order-maintenance institutions. It should be noted here that most sociological and anthropological studies of Taiwanese society conducted since World War II have found “parochial,” “particularistic,” or “intermediate” institutions to be structurally central to political-economic organization (e.g., Reference BoscoBosco 1992; Reference Gallin, Gallin, Fogelson and AdamsGallin & Gallin 1977; Reference SkoggardSkoggard 1996; Reference WellerWeller 1999; Reference Winckler, Ahern and GatesWinckler 1981; Reference Winn and RubensteinWinn 1994a). The democratic period, however, has made the contrast between parochial and public idioms of sociopolitical virtue increasingly problematic (compare Reference Hunter, Suttles and ZaldHunter 1985). The primary example is the emergence of a set of political-economic phenomena referred to locally as “Black Gold,” a sort of fusion between politics, business, and organized crime that seems to have been an unintended consequence of political liberalization (Reference ChinChin 2003). From the perspective of Junior and Muniz, we can see Black Gold as organically related to decline in the idea of police. An important corollary of this perspective is its implication that we interpret the institutions of Black Gold as involved in reproducing their own “parochial” forms of security and order (on this point, see especially Reference Winn and RubensteinWinn 1994a).

The ubiquity of informal order-keeping institutions in Taiwan (the majority not stigmatized as “black,” i.e., illegitimate) implies that large sectors of the social order are being maintained outside any ostensible monopoly by the state on the use of legitimate force. This leads us to take a closer look at the conceptions of “legitimacy” in active operation, especially in what they show of popular sensibilities about the proper practice of democratic government. Indeed, I believe this is a site in which we will find some of the most satisfying explanations for the paradoxical situation noted at the outset of the article, i.e., the consolidation of “rule of law” at the level of state institutions paired with the persistently marginal prestige of legality per se within normal social interaction.

Western theorists have, since at least Aristotle, tended to approach legitimacy in terms of the subjective orientations that transform power into authority (Reference ArendtArendt 1969; Reference TylerTyler 2006), and modern theorists generally work from Weber's classification of distinct traditional, charismatic, and rational-procedural bases for the perception that acts of domination are legitimate (with some authors adding “efficacy” as a fourth modality, e.g., Reference LipsetLipset 1994). Police operations are often taken as a relatively direct empirical indicator of the legitimacy of the particular regime they serve, i.e., it is assumed that there will be an inverse relationship between the level at which state authority per se is sufficient to foster orderly reproduction of the status quo, and the amount of naked coercion necessary in the conduct of police order-maintenance.

The situation described in this article shows a potential problem with assuming that low levels of police coercion can serve as a direct index of state legitimacy, to wit: lack of coercion may be explained as accurately by the weakness of state authority as by its strength. Indeed, my overall observations of constabulary policing in Taiwan have been characterized by relatively low levels of overt coercion, but this often appears as reluctance more than restraint (i.e., waiting for a street fight to play itself out before moving in, allowing people to avoid interrogation without pursuit, etc.).Footnote ¹¹ When I have asked about such apparent reticence, I have been answered with stories of unfortunate policemen who fell prey to civil lawsuits or to the other kinds of influence that regularly prove themselves superior to the weak bureaucratic proceduralism supplying the official backing to any individual policeman trying to justify naked antagonism solely by formal license. The moral of these stories is summed up in the pithy aphorism “More doing, more trouble; nothing doing, no problem ( , ).”

In sum, Taiwanese street patrolmen carefully modulate their interventions so as to avoid aggravating social forces that exceed their powers of control, and they must do this because their actions are embedded within the reproduction of a social “order” that is constituted in a balance between contradictory foundations of power and authority. Looking at how legitimacy is figured in this situation, I argue next, reveals the practical logic that endows the rule of law in contemporary Taiwan with its paradoxical qualities.

Invoking State Authority

A dissatisfied resident of Taiwan hoping to enlist state agency in his or her cause has a choice between two broadly distinguishable strategies: either staging a public chenqing, “complaint” ( , lit. “to display the situation”), or mobilizing a particularistic network through the “back door” of personal connections. These two channels are by no means exclusive; in fact, it is generally understood that successful invocation of higher authority will likely require some degree of involvement in both. Nonetheless, even as both avenues may simultaneously be at work in a given situation, they remain superficially antithetical, and their prima facie contradiction exemplifies the tensions that characterize police action in contemporary Taiwan.

To begin with, the back door is very wide; clientage networks are ubiquitous, and most people have at least some friends or family with political connections (Reference BoscoBosco 1992; Reference Winckler, Ahern and GatesWinckler 1981). Moreover, as the democratic transition is understood as a shift in the definitive mission of government toward “service” (fuwu, ), it is de rigueur for politicians within the category of “popular representative” (including borough chiefs, city and county councilors, and national legislators) to maintain outreach offices through which their constituency can bring matters to their attention in a quasi-informal capacity. These offices function as campaign headquarters during elections, and the gathering of electoral support is self-consciously understood as the formation of a patronage network, with campaign materials stressing the responsiveness of the aspiring representative to his or her clients.

At the same time, by virtue of the competitive aspects of the democratic process and the complexity of the administrative bureaucracy, back channels are not universally accessible or effective, and so it also frequently becomes strategic in the pursuit of a particular goal to deploy the forces of public attention. Public complaint shows up in a variety of media, from graffiti or the “white cloth/black letter” banners hung in protests through the entire spectrum of advertising and right into the content of the news itself. There is an art to framing effective complaints, and the conventions of this genre impose a structure on the content of public statements of dissatisfaction. As illustration, consider the following complaints taken from the Taipei County ombudsman's Internet message board. These two complaints, concerned with an incident of illegal construction, provide an example of the general voicing and conventional scripts used for the public expression of dissatisfaction in contemporary Taiwan and are especially revealing of the slippery boundaries between public legitimacy and particularistic efficacy.

The basic form of these complaints derives from the complainants' attempt to align their position as closely as possible with foundational values in Taiwan's contemporary political mythology, most particularly “the people's” entitlement to effective enforcement of “true/impartial law” (zhengfa, ). Motivated by irritation with the illegal construction of a garage, they present a plea to legitimate authority in the abstract, calling for disaster-averting intervention on behalf of a humble and hapless subject population threatened by the “mean hands” of a local bully. The undifferentiated category of “the people” is held up as the definitive legitimate political subject, specifically opposed to the unjust manipulation of political power through a factional patronage network. The illegal building is made to signify a “flagrantly” unjust world, animated by individual selfishness and produced through the machinations of a political system that operates effectively only when manipulated by the hidden networks of personal connection lurking within the representative bodies of the local and national government. In this case, the complainant exposes the network behind the garage to be the political machine of the “Blue Party” (i.e., the KMT; it is relevant that the county executive was at the time of this complaint a member of the “Green Party,” i.e., the Democratic Progressive Party, while the mayor of Weixing City was a KMT member).

The crux of these particular complaints is the “eating of a case” (chian, ), the disappearance of a case midstream in its bureaucratic processing. “Eating cases” is a chronic dimension of government operations in Taiwan, a technique used for diverse purposes including easing overwork, manipulating crime statistics, and integrating the demands of personal networks with the bureaucratic process. It is, accordingly, one of the most common concrete concerns expressed by Taiwanese citizens frustrated with the incapacity of the government to do what they want.

Finally, despite the centrality of the accusation of “influence peddling” to the complaint, it is not, however, an appeal to rule of law in the abstract. The positive aspect of the appeal is framed by reference to the vote as a personal transaction between subject and elected official, and it is directed to the personal virtue of the office holder, suggesting that the office holder's “courage” in the specific action at issue will be decisive in restoring a just world ordered by the “king's law” in place of the existing situation ruled by power and connections.Footnote ¹² This is key: even as the machinations of back-door connections are scandalous from the perspective of a public chenqing plea, politics remains nonetheless understood in personal terms.

Applying State Authority to Problems of Public Order

As a successful chenqing is accepted as legitimate by its addressee, it is publicly validated by a response,Footnote ¹³ whereupon it disappears into the bureaucratic machinery to be converted into the procedural requirements of the formal division of political labor. Complaints such as those described in the previous section concerned with illegal construction in public space, regardless of the channels through which they were received and acknowledged, generate their first practical response within two offices of the county government: the Department of Public Works (responsible for demolition of illegally built permanent structures) and the Department of Environmental Protection (responsible for dealing with temporary constructions, debris, and pollution). These departments are responsible for evaluating the actual situation on the ground and then making or recommending remedies according to relevant legal codes (potentially including the mobilization of a demolition team to clear the space). The implementation of these responsibilities generally takes place through subordinate bureaus of the county-level departments, housed within city government offices.

Police involvement in street clearing operations is standard on the pretext that forcing people to relinquish their control of space tends to cause conflict, and managing conflict is the police mission. It is quite uncommon to see demolition of “permanent” structures of the kind mentioned in the above complaints, probably because few people are foolish enough to invest resources in building them without first “smoothing the way” (baping, i.e., marshalling the connections required for a given project); when buildings are torn down, this takes place in the context of pitched political battles that are qualitatively distinct from the routine invocation of police authority at issue in this article. However, throughout 2002, the Environmental Protection Department sponsored a large-scale street clearing campaign aimed at nonpermanent structures, and under these auspices the city government kept a forklift and dump truck in continuous circulation throughout the city, cooperating with individual police substations on a weekly basis. Under this initiative, I was able to observe many instances of police action in clearing public space.

These events would begin with an employee of the city's Environmental Protection office (the forklift driver) arriving at a substation with a stack of books filled with thousands of official reports of “street occupation” (luba, ). Covering several months of investigative activity by this office, these books were filled with entries recording the sites of such things as “motorcycle repair operation causing sidewalk blockage,” “discarded vehicle,” “flowerpots.” Under each heading was a list of dates of observation, each followed by the notation “Checked. No improvement.” This documentation was carried into the field as an immediately accessible formal justification for a raid on the day's targeted street. The actual work of clearing the street was executed by the forklift and a dump truck, augmented by a troupe of a half dozen or so older women who volunteered at the city Environmental Protection Department and were thus called the “Environmental Protection Mamas.” Working together, this team was capable of indiscriminately scraping the sidewalks and streets clear of all nonpermanent structures, loading everything from restaurant operations to piles of garbage to potted trees to boards covering potholes into the truck to be taken off to the dump. While doing this, they were accompanied by two policemen in their official “guardian” function, who observed the goings-on, wrote tickets to selected violators, and documented the results of the cleanup effort. A typical outing I accompanied is described below:

Again, like Patrolman Guo's intervention described earlier, the practical logic of the operation focused on the production of a photograph of a clean street, a situation that lasted at most a few hours and generally only minutes. This exemplifies a practical approach to law enforcement in which the “thinnest” procedural requirements are used to establish a de facto common goal between the enforcer and the targets of enforcement, which then allows larger contradictions between the official intentions of the street cleaning team and the interests of the street merchants to be bracketed within a space of mutual nonacknowledgement while they practically cooperate in the production of photographic proof that the mission has been “accomplished.” The flexibility with which “thin” procedural requirements can be manipulated to avoid implication in contested dimensions of the contradictory “thicker” interpretations held by various participants in the event provides a crucial resource for police management of public order problems in contemporary Taiwan (Reference ThacherThacher 2001). The messy and potentially volatile street-level conflicts are thus converted into the more abstract and tractable problem of, as Guo put it, “how to write the official reports.”

Sentiment, Reason, and Law: Police Work From the Patrolman's Point of View

Taking some poetic license with translation, the term Guo used for “official reports” (gongwen, ) could be rendered as “public text,”Footnote ¹⁴ and we can take these examples of good police work to be exemplary by virtue of their successful management of the tensions that emerge where “public texts” composed in the language of the law intersect with the fabric of particularistic “sentimental feelings” that define the political textures of local community life. This does indeed appear to be an accurate evocation of the most conventional orientation Taiwanese patrolmen adopt in evaluating the quality of their own work. Throughout my research, I have found patrolmen's discussions of their work to invoke framing in terms of the pursuit of a balance between law (fa, ) and sentiment (qing, ), as mediated by a third term, reason (li, ).

These three terms define a well-established trope in the wider culture; their invocation and discussion of their various interrelations has a venerable pedigree in Chinese philosophy and remains common in casual discussion of sociopolitical phenomena throughout contemporary Taiwan. In current Taiwanese popular discourse, their most prevalent invocation is in the form of a contrast between two hierarchical sequences of precedence: “sentiment-reason-law” (qing-li-fa, ) vis-à-vis “law-reason-sentiment” (fa-li-qing, ). In the qing-li-fa sequence (which is often identified as traditional, Chinese, or the “rule of man”) sentiment precedes law, i.e., sentiment overdetermines law, and so the thick webs of guanxi ( )—individualistic relations between kin, patrons, friends, or other varieties of nonbureaucratized solidarity or patronage—undermine the systemic integrity of formal organizations such as labor unions, professional associations, or state bureaucracies. The reverse sequence, fa-li-qing (which, by contrast to the above is often associated with modernity, the West, or the “rule of law”), indicates a cultural milieu in which proper procedure always takes precedence over personalistic ties, and efficient bureaucratic machines assemble almost spontaneously in the aggregation of social individuals such as those Reference WhyteWhyte called “organization man” (1956). Thus, editorials in the popular press frequently contain statements such as, “[i]n order to improve the contemporary degeneration of Taiwan's political environment and its pervasive political culture of influence-brokering, the Chinese must change this enduring valorization of ‘sentiment-reason-law’ which forms the ideal of the ‘rule of man’ in order to develop the valorization of ‘law-reason-sentiment’ which forms the ideal of the ‘rule of law.’”Footnote ¹⁵ Within this reasoning, there is a sense of zero-sum opposition, where the imposition of law indicates the rupture of sentiment, e.g., “Where sentiments are thick, the law retreats. When sentiments are broken, the law advances ( , ).” This enables as counterpoint to arguments for the rule of law a romantic valorization of the idealized harmony of sentimental community over the impersonal mechanisms of legal authority. Thus where law sees sentiment as partiality, sentiment sees law as arbitrary. And the debate goes on.

Although it is not widely discussed in the democratic era (indeed, it is to some degree suppressed as an embarrassing relic of the bad old days), there is actually a recent historical layer of ideological doctrine beneath the rhetorical conventions of this contemporary popular discourse. To wit, as a triad the categories of sentiment, reason, and law figure centrally in the Nationalist Party's official doctrine of the Three Principles of the People (Sanminzhuyi, ), in fact purporting to provide a sort of foundational basis for the three principles themselves.Footnote ¹⁶ It is thus not surprising that police patrolmen, whose graduation from the Police Academy was until quite recently predicated on passing exams in Sunnist doctrine, would find in these basic categories a ready scheme for rationalizing their work. In patrolmen's discourse the qualities of sentiment, reason, and law are not described as exhibiting a fixed hierarchy in which any one is clearly dominant over the others (i.e., the binary qing-li-fa/fa-li-qing opposition of popular discourse). Rather, like the older Sunnist doctrine, they are presented as independent dimensions of causality, with the successful resolution of conflicted situations demanding an orientation toward balancing the three.Footnote ¹⁷

As mentioned previously, a primary institutional determinant of modern police culture arises from the nature of the patrolmen's position as state-authorized agents of conflict management; that is, patrolmen spend their time answering summons to situations where something needs to be done and state agency has been invoked to do it (Reference Bittner and JacobBittner 1974). Accordingly, Taiwanese patrolmen arrive at the scenes of their professional performance as formal representatives of state authority, and the most common self-identification by the patrolmen of their social role is as a “law executor” (zhifazhe, ). This is a difficult role to play; to a Taiwanese neighborhood patrolman, the law is a dangerous thing. As one patrolman put it, the law is a knife and a policeman must be a skillful surgeon if he himself is not to lose control of the blade. To wit:

Several significant themes appear in this comment. First, it is a thin invocation of the principle of legality; it frames law as purely instrumental. From the patrolman's perspective, law is simply a thing used to create a link between misdeed and punishment: “If you aren't wearing your motorcycle helmet, you get a ticket, there is nothing else to say,” i.e., that's just the law (quote taken from the same interview). The mechanics of this connection are understood to be “rigid and harsh … violation of the law is just a fact,” (quote taken from the same interview). Such arbitrary authority is, on its own, understood by the patrolman to be unreasonably absolute; the law carries no connotation of anything inherently legitimate. Indeed, the kinds of logical or sympathetic qualities involved in legitimating legal action are understood to exist as explicitly distinct categories: the harmfulness or helpfulness of a given invocation of law is entirely dependent on “external” criteria, the qualities of reason and sentiment being most significant among them.

A second theme in the cited fragment is the necessity of consciously appreciating the fact that law, as the formal bureaucratic power of the state, is not the only game in town. Self-preservation demands that street-level law enforcers be highly sensitive to all forms of power capable of structuring the course of events in their working environment. Police work is understood to be intrinsically “complicated” (this being a sort of all-purpose euphemism for anything from routine happenstance to deep political-economic intrigue). In particular, as mentioned above, post-authoritarian Taiwan has come under sway of a “Black Gold” regime, a phrasing indicating the pursuit of financial interest (i.e., “gold”) through hidden (or “black”) networks of collusion, and referring to a ubiquitous, even structural, intersection of organized crime with the political infrastructure (Reference ChinChin 2003). At the face-to-face level of ethnographic reality, this kind of political-economy involves a sort of covert parallel universe of patronage, collusion, and conspiracy submerged within the conventional routines of everyday interaction. As cultural forms, these relationships take the shape of particularistic bonds, more or less euphemistically described and performed in the terms of sentimental attachment, notably kinship terms of address and the motif of brotherhood. It is worth noting here that the use of forms of real and fictive kinship as a basis for economic organization is a human universal, and social anthropologists have long been impressed by the capacity of southern Chinese idioms of kinship to function as the basis of highly elaborate economic formations (e.g., Reference Freedman and FreedmanFreedman 1979; Reference WinnWinn 1994b). So it is not surprising that the sentimental dimensions of life in the substation would serve as the arena in which the shady side of the job is taken into account.

In this respect, we should consider that the average patrolman's middle-class salary of about 1,500 US$ per month is potentially augmented by an undocumentable but by all estimates roughly equivalent amount in various informal “regulatory fees” (guifei, ), “grease water”(youshui, ), or “A-money” (A-qian, A- , sometimes “A-Food,”A-cai, A- ), which flow in through the substation and up through circuits of administrative oversight into the wider regulatory bureaucracy. An organizing principle for the practical integration of the formal regulatory apparatus with the informal neighborhood economy is provided by the fixed responsibility assigned to individual policemen for managing the affairs of their beat—in particular, the way their individual accountability for illegal enterprise discovered to be operating there creates an interest in these enterprises (which constitute a significant proportion of the local economy) remaining “undiscovered” (i.e., unremarked upon). The prototypical situation is an unlicensed establishment—paradigmatically the so-called special industries involved in the minor vices of drinking, gambling, and prostitution—passing a “red envelope”Footnote ¹⁸ containing the equivalent of several hundred US dollars to the local beat officer each month, a sum that is split between the patrolman and certain other cooperating figures. As this money flows into the substation, it is accumulated by the working personnel not simply as personal profit, but also (even primarily) as a repository from which to supply the funds that must be passed up the regulatory food chain during periods of administrative review and consideration for promotion.

The operation of this financial system (which, of course, exists as an objective social fact generating the conditions under which individual officers navigate their careers) has the effect of maintaining an alignment between the interests of the local regulatory apparatus and those of the local unlicensed economy, ensuring that discretionary slack in the bureaucratic machine is oriented toward the continued smooth operation of the total system. Not every unlicensed enterprise chooses to participate in this system; only those that understand their own financial interests as served by making a long-term investment in the regulation of their local neighborhood, i.e., the good citizens of the informal economic order. And since the economic health of the neighborhood is, like that of the island as a whole, inseparable from its informal dimension, such contributions are locally understood (by those involved at least) as ultimately no less legitimate than the formal donations made to the police by the local chamber of commerce, Lions Club, or “Policemen's Friends Club.”

What chiefly distinguishes these informal economic relationships from more formal taxation is the regime of social relations in which they exist, and that their operations maintain as institutionalized relations of solidarity. That is to say, the day-to-day management of informal economic cooperation is conducted within the idiom of qing, i.e., as “sentimental” and particularistic relationships defining a space of intimacy explicitly insulated from the formalized distances of public interaction and generally marked by the strategic inversion of selected “public” values (Reference HerzfeldHerzfeld 2005). The ostensible “intimacy” of this regime does not, however, set it outside the scope of the order that police authority is deployed to maintain. Quite the opposite, in fact; when talking about the quality of the substation's jurisdiction or their personal beat, patrolmen exhibit their strongest primary orientation to ganqing, the “sentimental” dimension. Indeed, most of the routines of working life in the substation take place nestled in metaphors of brotherhood and friendship (or, in instances of disharmony, a discourse of more or less vitriolic personal conflict), operating on a rationale that takes “mutual assistance” as the highest human social ideal. This accords with a valorization of social relations based on “compassion,” (renqingwei, ), which is one of the most widespread affirmations of what is distinctively good about Taiwanese society. The order of qing is, in other words, a rational moral universe defining a normative ideal of the good society, in which the amalgamation of formal and informal money and power makes sense. Grounded in the sentimental values of righteousness and loyalty, patrolmen are just average guys, good family men striving to do right by their intimates as they collectively struggle to survive in a treacherously complicated environment.

No small part of the complication in police life arises from the fact that, when framed by the expectations of bureaucratic rationality, these intimate circuits of exchange of money and favor are, quite simply, illegal. To be sure, the tension between “cooperation” and “corruption” always contains a degree of ambivalence (itself a crucial element in the routine operations of any criminal justice system). But there are occasions when the ambiguity must be resolved into a definite value judgment. And when this time comes, the choice of which epithet or euphemism is applied to a particular case depends, ultimately, on the standpoint of evaluation. Public discussion of police corruption in Taiwan is generally unforgiving. The process of political liberalization has fostered public sensibility that the progressive future lies squarely on the side of rule of law in opposition to the retrograde past of rule by man and, accordingly, public sphere discussion increasingly invokes liberal-democratic discourse about the ultimate value of legality as the defining factor in evaluations of illegitimacy. This is a positive factor in consolidating support for rule of law within state institutions. But at the local level, where the lines between state and nonstate authority are hazier, this shift in the terms of public culture exerts a destabilizing effect on the established balance between sentimental community and local public order. In other words, from the perspective of the average policeman on the street, democratization has intensified the political tensions saturating the arena of their discretionary practice, making their practical reliance on intimate networks more crucial than ever.

These tensions are described by patrolmen in terms of an acute confrontation between fa and qing. To understand this discourse, we must recognize that the moral bonds of sentimental alliance are not always valorized as good. People who perceive their social position as exploited through entrenched and inaccessible networks of other people's solidarity interpret the order of sentiment as a tragic element of Taiwanese culture. The “free and easy” qualities of community in which problems are worked out through the deployment of informal social resources is to them an insidious mythology from which the “true law” (to use the rhetoric from “The Idea of Police in Contemporary Taiwan”) should properly rescue them.

There is a broad contradiction in social values here, and the police, as professional managers of social conflict, cannot comfortably take a firm stand on either side. Thus for them, a delicate balancing act arises in every direct confrontation of the mobilized authority of the state's fa with the institutionalized political-economy of the local “sentimental society” ordered by qing. Consider, in this respect, that when a citizen calls the police the emergency 119 system automatically routes the call to the county police station, where it becomes a matter of formal record. It is then relayed to the appropriate substation's front-desk telephone line, from which it is radioed to the policeman on patrol duty at the time. Receiving such a dispatch, the patrolman knows that the county station's record of the complaint combined with the substation duty roster have created a paper trail identifying him personally as legally accountable for the management of the problematic situation.

The policeman thus arrives at the scene to which state authority has been called as an individual embodiment of that authority. In this situation, the image of “standing on the edge of a knife” becomes compelling. Whenever called into a situation where informal privilege is being challenged by an outraged citizenry holding formally valid expectations that illegal activities will be disciplined by the state, a policeman finds himself at the cutting edge of an intersection between two distinct regimes of power and authority, to both of which he is beholden. In the course of their confrontation, both forms of power—the formal legal system and the informal political-economic system—threaten to become further mobilized. And escalation in either of these realms will immediately go over the head of the lowly patrolman, circumscribing his discretionary command of the situation while leaving him to account for the actions of his superiors across the awkward disjuncture of two forms of power that mutually recognize one another as illegitimate.

The personal interests and, hence, discretionary resources of the policeman are therefore primarily invested in ensuring that the conflict does not escalate beyond his control so long as he remains implicated in its outcome. In this balancing act, reason serves as the decisive fulcrum. This is obvious to an observer of police-mediated negotiation: the policeman's constant refrain is an exhortation to the involved parties to “be reasonable.” It is also central to the way policemen reflect on the skills required for their work. By contrast to the harsh cutting edge of law, reason is described as a circular, rounding, and smoothing quality that fits the law to context, the “soft” consensus with which everyone (at least, all reasonable people) will “self-identify.” And where sentiment is understood in terms of inherently particularizing allegiances, reason is the basis for a universalism in which all reasonable people can be expected to participate.

In some ways, the patrolman's valorization of li—taking a natural human capacity for reason as the foundation on which to understand the practical possibility of finding harmony in social relations—resonates with the kinds of liberal political philosophy that undergird modern democratic theory. This similarity ends, however, at the patrolman's understanding of inherent contextual limitations to the effective authority of reason. That is, the discretionary form of reason used by Taiwanese policemen to mediate the disjoint orders of law and sentiment is not a properly “public” quality. It is provisional, ad hoc, and radically situated in its immediate context.

That is to say, the appeal to reason made by a patrolman in the course of managing conflict takes the form of a call to reflection on the part of the involved parties concerning their shared interest in resolving the conflict. Such reflection is, on its own, inadequate to challenge or resist the compelling forces attached to state law or sentimental allegiance. To be sure, in the routine situation of a standoff between the powers of fa and qing (i.e., the kinds of chronic “complications” of local political-economy that we have seen herein), a mediator can effectively orchestrate resolution of conflict from the detached standpoint of a fair broker. But the slightest shift in the underlying dimensions of the conflict will quickly efface the practical standpoint for this kind of intervention. Thus, from the patrolman's vantage point, there is an antipathy between reason and publicity. Publicity—the mediation of conflict into wider spaces of attention—expands the number of distinct interests that must be taken into account in achieving a resolution, attenuating the space of their common conjunction. Compromise is possible behind closed doors, but handling conflict in the public eye utterly transforms the kinds of stakes that are attached to the act of relinquishing one's position; once an event has been pulled into the realm of unbounded public spectacle, there is little likelihood of a reasonable resolution. In other words, keeping contradictions in order requires obscurity, a space in which conflicted “principles” can be compromised without causing further repercussion.

This brings us to a final observation. When describing to me the skills of good policing, Taiwanese patrolmen occasionally invoked the ideal of touming ( ), “transparency.” However, in the context of their engagement with the sentimental fabric of local order, they used this term to indicate a relationship of intimacy, a relationship of trusting openness between a patrolman and the individual inhabitants of his beat. Implicitly framed by the normative order of qing, this idiom of intimate transparency is entirely contrary to the standard association of (public) transparency with the rule of law. Intimate transparency is not a transitive relationship: if A is intimately transparent to B, and B is intimately transparent to C, this does not then imply that A is transparent to C. Quite the opposite, in fact. As intermediary, B finds himself in a delicate position where the slightest indiscretion on his part can have potentially dramatic consequences for the relationship between A and C. Indeed, the fact that A and C rely on B to keep a mediated distance between themselves is less often an accident of circumstance than an intentional and integral aspect of their relationship, e.g., B is the patrolman, A is a local businessman, and C is a district prosecutor. The patrolman's crucial function in maintaining a harmoniously integrated balance of powers rests on his skills in strategic translation of the meaningful connections (aka “information”) shared within the intimate order of ganqing into those demanded by the operations of the legal system, concealing and revealing information in such a way that none of the involved parties feels that their reasonable expectations of orderly give-and-take have been egregiously violated.

Conclusion

The term legitimacy sometimes appears as a theoretical thread suggestively knitting the normative orientations of individual subjects into a “general theory of association” (in de Tocqueville's terms) that animates civil society and provides a substantive ideological foundation for the sovereignty of the state. But things are seldom so simple; the processes of legitimation active at these different scales are categorically distinct, and their interaction is as often as not cross-cutting, contradictory, and chaotic. Ethnographic studies of bureaucratic practice reveal the tremendous amount of work necessary to manifest ostensibly universal normative principles in actual events, showing how tenuous and provisional these manifestations are and how profoundly their formal institutional supports are saturated by, and articulated with, more diffuse but no less determinate cultural structures of sociological order.

The recent emergence of a relatively autonomous sphere of judicial institutions dedicated to the principles of a “liberal democratic ‘thick’ version of the rule of law” is something new in Taiwan's historical experience. It is a phenomenon of profound significance to the island's future hopes, as well as a core feature of whatever “lessons” Taiwan may be said to offer others' experiments with political liberalization. Nonetheless, from an ethnographic perspective, the actual significance of judicial institutions to the lived order of social life in contemporary Taiwan remains obscure. This article has described some of the social dynamics that motivate and maintain this obscure relationship. In examining how Taiwanese neighborhood patrolmen confront and manage their particular engagement with the tension between law and social order (intrinsic to modern constabulary practice) we have seen them serving, quite self-consciously, as caretakers of an obviously conflicted intersection between ideals of order embodied in law (and invoked in public complaint) and the pragmatic order of a local society organized through particularistic connections rationalized in terms of sentiment. Practically, they use a strategy of reducing law to its thinnest procedural requirements (exemplified in staging a photograph to serve as “proof” of a clean street) in order to creatively utilize these minimal requirements as resources for the task of immediate conflict management. And they discuss this sort of practice in terms of the “reasonable” management of an intersection between two entirely distinct spheres of social logic, i.e., “law” and “sentiment.” Summarizing these results, we can conclude that police in Taiwan take a trope of the reasonable person to be a legitimate standpoint for engaging in the active manipulation of other, contradictory claims about the character of legitimate authority, a manipulation necessary to successfully “muddling through” the historical circumstances of their job.

To the degree that this kind of narrative is accepted by the wider society (a question for further research), we can say that the idea of police in Taiwan is best described as a general public faith that the police will, when called, deploy their quotient of state authority in the role of facilitating the reasonable management of the problem at hand. This ideal accepts the underlying persistence of structural contradictions (between, for example, hegemonic ideals of public order and the dense fabric of particularistic relationships that constitute the substantive order of local life), making a relatively minimalist demand that these contradictions be kept within reasonable bounds. An immediate corollary of this idea of police is that interested participants in conflict can be expected to routinely invoke “other” authorities to the degree that such invocations can be sustained as reasonable. It is, in other words, a cultural space of legitimacy in which the solidification of the rule of law within state institutions is kept within boundaries of a social sensibility that does not take law as the last word.