Authorizing the Production of Urban Moral Order: Appellate Courts and Their Knowledge Games

Abstract

Using some appellate courts' reviews of projects to maintain moral order in the city as the main source of data, this article shows that an analysis of legal knowledge production that (1) is dynamic and (2) refuses to treat people and texts as totally different entities, one studied by social scientists and the other studied by lawyers, can tell us much about such familiar but seldom theorized legal maneuvers as judicial review and constitutional challenges. Choosing to analyze the dynamics of knowledge processes is inspired by Actor Network Theory (ANT), Bruno Latour's work in particular. This methodological choice is particularly appropriate because judicial review tends to avoid making judgments about the content of impugned laws or ordinances, focusing instead, as Latour does, on form and process. But insofar as legal processes in general privilege form and process to a greater or lesser degree, a more general argument is made about the appropriateness of using tools from ANT to study legal and quasi-legal knowledge networks.

Studying Knowledge Claims in Legal Networks: Method and Its Effects

A central concern of law-and-society scholarship has been to map how extralegal knowledge claims about social relations are introduced into legal processes, either as expert evidence or through other routes, and to document how courts and legislatures then do or don't use these facts and claims. Studies along these lines—which are far too numerous to mention here—generally focus on the content of the knowledge claims that are either accepted or rejected by various legal actors. They raise questions such as the following: Are legal processes increasingly influenced by technical expertise even as some forms of high science lose prestige (Cole 2001; Jasanoff 1995)? What makes certain courts sympathetic or hostile to sociological and cultural evidence (e.g., Haney-Lopez 1996; Valverde 2003a)? And at a more theoretical level, debates have taken place about whether “law” in general is increasingly governing through extralegal “norms,” including scientific knowledge and statistics (Ewald 1991; Hunt & Wickham 1978).

We have learned a great deal from such analyses of the fortunes and misfortunes of technical, scientific, and cultural knowledges within legal contexts. Here, however, I do not seek so much to add to this literature as to stimulate methodological innovation through analyses that ask not about the content of claims but about process and flow—about how actors pick through documents or discourses and cobble together new governing machines that recycle old bits in new ways (see, by comparison, Riles 2000). For this, much can be learned or adapted from studies located at the intersection of science studies and “actor-network” analysis (e.g., Latour & Woolgar 1979; Latour 1987, 1993; Winner 1986; Woolgar 1988; Barry & Slater 2002). Such work is just beginning to have an impact on law-and-society circles.¹ While neither the findings nor the methods of Actor Network Theory (ANT) (Latour 1987; Law 1999; Law & Mol 2002) can be directly applied to legal contexts, borrowing from these resources can reveal processes obscured in more static—and more humanist—analyses of knowledge in law.

In one influential formulation, knowledge is seen as a form of capital embodied in persons and in classes (Bourdieu 1987). The approach outlined here regards legal knowledge not as a capital embodied in people, but as an ever-shifting network in which actors—including things and texts, which I treat here, Latour-style, as actors—deploy legal or quasi-legal tools to creatively recycle knowledge claims generated elsewhere. A key difference between Latour and Bourdieu is that Bourdieu's ultimate aim is to unmask the human relations of appropriation and domination that ground such apparently nonexploitative habits as going to museums. Latour's work, by contrast, arose in a posthumanist intellectual environment. Like Foucault, Latour regards nonhuman relations (lab devices, architectural arrangements, textual formats) as significant actors, alongside humans. Yet he does not privilege the technical effects of things either; he carefully avoids the technological determinism of early science-and-technology work. In Latour's work, as in ANT generally, the analyst remains open-minded about which human and nonhuman actors play which roles with what effects, in the particular case. This is highly relevant to sociolegal scholars who do not wish to limit themselves to studying either “law in the books” with legal tools or “law in action” (people) with social science tools. And this perspective does more than merely exhort us to study both books and people: it provokes us to see what happens when one looks at things and texts as if they were people, and at people as if they were part of a technical assemblage.

The case study of Canadian lawmaking provided below reveals that, at least in this case, lay (nonlawyer) witnesses appearing at parliamentary hearings treated legal categories in a more black-letter manner than senior state lawyers and Supreme Court justices. This finding runs counter to the Bourdieu thesis about professionalization of knowledge and deconstructs the usual opposition of lawyers discussing law versus ordinary people discussing experience. Even if it is unrepresentative, which it may well be, it can help open our eyes to the creativity of knowledge networks. That courts of appeal often reason nonlegalistically is hardly a new finding, but that respect for black-letter texts is not always highest among the professional lawyers is perhaps a more novel insight. If one attempts to capture the dynamics of knowledge processes, rather than categorizing knowledges statically, one sees new things.

But of course every act of seeing also renders other things invisible. Along these lines, Bourdieu would no doubt say that Latour's methodology renders systemic power relations invisible, and he would have a point. We who are neither Latour nor Bourdieu, however, do not have to take sides in this hypothetical debate. Given that, especially in Latour's usage, concepts such as actor and network are merely tools, not parts of a general “theory” (see debate between Latour and John Law in Law 1999), it should be possible for sociolegal scholars concerned about systemic inequality to borrow some tools from actor-network analysis to analyze the workings of a knowledge network while complementing such an analysis with a substantive study of power relations. If Latour's approach is not a theory, as he himself claims, then it does not need to be either accepted or rejected. It can merely be mined. An actor-network analysis is never a complete and total picture. It excludes precisely that which Bourdieu—and Critical Legal Studies, and feminism—tries to highlight. Thus, the case studies below mainly trace the dynamics of knowledge re-creation, but they also signpost some of the systemic governance relations that constitute much of the substance of the claims seen circulating. And a fuller study of the matters involved in the cases below would include a sociological analysis of power relations.²

While much sociology of knowledge focuses strictly on knowledge production—with Bourdieu's studies, by contrast, focusing on consumption and accumulation—actor-network analyses always treat knowledge processes as simultaneously using and producing knowledge. Latour's recent ethnography of the French Conseil d'Etat shows that in the Conseil's arcane process, production and consumption (of knowledge claims) are not separate operations (Latour 2002). This approach is followed here; hence the use of terms such as flows and networks, rather than knowledge production. But because he is borrowing from his own earlier ethnographies of a single lab, Latour is concerned only with the Conseil's own internal productive consumption of knowledge. And as Bourdieu might have pointed out, making one institution into the object of one's study has the inevitable if unintended effect of reinforcing the institutional ideology of autonomy and formalism. Contrary to the self-serving illusions created by formalist techniques, courts and other legal venues are epistemologically derivative and dependent, not only on other legal inputs but also on “extraneous” claims; this dependence is rendered nearly invisible in Latour's rather formalist ethnography.

Once taken up by legal actors, claims about social relations, including those presented formally as “evidence” but also those knowledges that have a less direct influence upon law, get moved around and transformed. In the process, like the explorers' maps and other “immutable mobiles” influentially described by Latour (1987) and by other scholars of knowledge forms (e.g., Cohn 1996), claims or sets of claims (e.g., a document introduced as an appendix to an affidavit) come to have novel uses and hence meanings, even if they retain their physical and textual integrity.

In studying the dynamics of legal knowledge flows, debates regarding urban ills and urban improvement are particularly well-suited to serve as case studies. By contrast with some other, better studied regulatory sites, the problematic space of “the city”' (e.g., Frug 1999; Sennett 1970; Rabinow 1989; Vidlar 1991) has been and remains a particularly heterogeneous space of governance, in which claims made by a variety of types of authority are quickly mixed and reassembled in a rather cavalier manner. Reading about the history of cities and of urban reform projects suggests that both the ills afflicting that conglomerate entity “the city” and the solutions proposed by reformers and by municipalities have generally been ontologically and epistemologically hybrid. For the Romantics, cities were unsanitary for the same reasons that they were ugly, and beautiful parks were regarded as solutions to social and health problems as well as aesthetic projects (Williams 1985). Along the same lines, English late Victorian campaigns around overcrowding also treated the problem as simultaneously moral, economic, and physiological (Wohl 1983). And today's urban regeneration projects also work on the hybrid premise that physical beautification will automatically bring about social salvation, for the middle classes at any rate (Davis 1990; Smith 1996).

This marked hybridity, in which both the illness and its remedy are seen as naturally multitasking, is also found in the relevant legal networks, even today. In the United States, the persistence of rationalities of urban governance mixing moral, economic, and engineering logics has often been noted, but it is usually explained merely as a tool of racism (Delaney 2001; Ford 2001). While not discounting this analysis, it is important to note that in countries with less heavily racialized urban geographies, the same legal tools and many of the same legal rationalities exist. It is thus analytically dangerous to reduce issues of urban vice and virtue to “race,” or to any other single issue. Moral concerns about sexual evils (especially prostitution, that age-old conglomerate of health, moral, and economic urban risks [Valverde 1989]), racist fears about the barbarians within, and economically driven projects to rescue the inner city from decline and “blight” have been and remain intertwined. Arguing about which is most important fails to respect the sincere belief in hybrid solutions for hybrid ills that has driven urban reformers for some centuries now.

If the field of the urban is markedly hybrid—which is not to say that other fields, such as health, are not hybrid, but simply to make a relative claim—epistemological heterogeneity is also visibly greater in legal than in other discourses, as a general rule. For example, as Sunstein has pointed out, analogy is an ancient, prescientific mode of reasoning that is nevertheless openly accepted as a way of making determinations in law (Sunstein 1993) alongside more “rational” modern modes of reasoning. In addition, epistemological heterogeneity is actively encouraged by the technical features of the adversarial system: lawyers pile up totally divergent and even contradictory arguments in the hopes that the court will accept at least one or two of them.

Thus, municipal matters, both because they are legal and because they involve that assemblage of moral-physical risks known as “the urban,” are particularly well-suited to a study documenting the creative mixing and remixing of different kinds of knowledge claims. Nevertheless, the easy blending, without much worry about coherence, of totally divergent knowledge frameworks, is neither unique to law—political debate is also markedly hybrid, with speakers rarely being expected to stick to one mode of reasoning—nor unique to the municipal realm. Indeed, one of the two case studies below, on the Canadian law criminalizing public transactions about prostitution, is not, legally, a municipal affair, since it involved a constitutional challenge to sections of the Criminal Code of Canada. But here too the urban was the site of the “crime” in question. Perhaps because of the location, moral, economic, social, cultural, and gender issues were mixed with great abandon not only by those who participated in the consultations but even at the Supreme Court itself.

My argument about heterogeneity challenges most accepted accounts of modern knowledge. Specifically, Latour's influential book We Have Never Been Modern argues that since the scientific revolution there have been two main knowledge networks, one pursued by the natural sciences, and another, opposite one developed by the cultural and human sciences (Latour 1993). Arguing about whether X is natural or social, the two sides have nevertheless implicitly agreed to never question the underlying epistemological assumption that the role of modern thought is to define everything as either natural or cultural, either physical or moral, or, today, as either real or socially constructed. But as Latour, interestingly, does not explicitly point out in his study of the Conseil d'Etat (Latour 2002), legal networks are an exception to this rule of purity-seeking. In a sociological context, one strives to cleanse one's argument of all vestiges of technological determinism as well as all traces of that dreaded Other of sociology, biologism. By contrast, in a court of law, it seems to add weight to one's argument if one, say, introduces evidence that nude dancing in a strip club is a potential health hazard as well as source of moral blight. Law, it seems, does not wish to be modern.³

Hence, one cannot always pose the question of knowledge in an either-or fashion. Judicial review of municipal ordinances setting out special zoning requirements for sexual businesses can nowadays include planners' or sociologists' studies of property values, crime, traffic, and so on, and sometimes such studies are mandatory: but that does not mean that a scientific logic is driving out moral logics. In some areas of law, science may well drive out other types of claims: but in at least some areas of law, knowledge is not a zero-sum game.

That knowledge is not a zero-sum game is also the beginning point of studies generated by Teubner and other theorists promoting the “autopoeisis” view. The analysis elaborated here acknowledges that legal networks transform knowledges, but it does not support Teubner's claims about the unidirectional absorption of extralegal facts into a separate legal realm (Teubner 1989). Teubner would agree with Latour that claims have no essence, since his analysis, unlike most others, is not static; but his model is unidirectional, with “law” having an intrinsic tendency to become more and more autonomous as “society” becomes, Durkheim-fashion, increasingly differentiated. If we give up on evolutionist assumptions, it becomes more useful to look at the knowledge game not as an upward line but as a constantly shifting circular shape. Hence the usefulness of Latour's term network—despite the fact, noted by Latour himself (in Law 1999), that after the Internet it is difficult to detach the word network from that particular set of practices.

In the urban law network, largely textual entities move along the network but change their character even when they are not excerpted or otherwise manipulated. For instance, for the consultants who produce them, studies of urban problems produced to justify zoning ordinances are money-making commodities; for the city council that uses them, they are legal capital; for the appellate courts, they are mainly procedural entities. The occasional sociologist who reads the study in question for its scientific value is not a superior reader, even if he or she is the only one without a legal ax to grind: he or she is simply an actor in a different network. This article is thus another actor, mainly located in the law-and-society scholarly network but overlapping with “real-life” networks. While some ANT scholars write as if their own accounts are non-networked and are above the fray, I think it is more consistent to regard scholarly work as overlapping with the networks one studies, not in some superior epistemological realm. And this is the case whether or not any appeal court judges read our articles. What is experienced by social scientists, especially expert witnesses, as “being ignored by the courts” is, for the courts, a highly productive knowledge move (Valverde 1996). Bracketing the content of expert evidence allows the judicial network to go on and do what it needs to do.

Expert evidence is of course not the only input that is subjected to often ruthless forms of excerpting, synthesis, decontextualization, and rejection. The network of judicial review would instantly collapse if all the available substantive information relevant to a certain decision or policy were consistently treated by courts as legal content. Extralegal knowledges have to be managed, and this is done in part by reducing documents and testimony to an almost content-free existence. In the case of the U.S. Supreme Court's decision in City of Los Angeles v. Alameda Books (2002) studied below, the Court noted that the city had a study to back up its zoning ordinance and that the study could reasonably be placed within the large universe of studies of urban problems that might support not so much the specifics of the ordinance (which were not even mentioned in the study) but rather the general municipal policy to restrict porn shops. And that was enough.

In this way, an old study comparing crime rates across neighborhoods was turned into an almost wholly formal entity, whose main virtue was that it existed—and whose second virtue was that it dealt with some harm (crime, in this case) that the reasonable person might think has some not necessarily specified relationship to the target of the ordinance (porn shops). This logic is not as curious as it seems. Formalization and excerpting are knowledge moves featuring very largely in many modern information transfers. To give an example: the information contained in a birth certificate does not usually get transferred, holus-bolus, into other documents. One office cares only about the date of birth, another about the country, and the bank does not care about either, needing only the formal knowledge that a proper ID document has been produced. This is what I call “excerpting” (which is a type of reduction). Formalization is also common across bureaucratic systems, as academics know from the experience of reducing the concrete messiness of a particular seminar to some markings on a page titled “final marks.” While specifically engaged in producing law, then, courts also contribute to the more general process of bureaucratic inscription (see Poovey 1998; Ericson & Haggerty 1997; Curtis 2001)—inscription referring to the process by which events and states of affairs are first turned into ink markings on a page and are then homogenized and standardized by the imposition of a format that allows for tabulation, and for inter-individual and intergroup comparison by future paper-pushers in the same or in other agencies.

I continue this analysis of knowledge processes in a more grounded manner, proceeding to show how an assortment of claims underwent transubstantiation as they were placed in a judicially run network. As they move along the network, the effective meaning of knowledge claims is dependent not only on the location and status of the user (as Bourdieu and other sociologists of knowledge have shown) but also—and this has not generally been studied—on the specific features inherent in particular knowledge moves. Legal networks are full of site-specific or field-specific techniques⁴ for using, reducing, formalizing, and synthesizing knowledge claims.

A legal tool to work up knowledge claims, such as judicial review, can thus be seen as a machine made up of certain wheels and cogs—not all of which are used every time—that enable certain kinds of knowledge processes but forbid others. Constitutional challenges, the other kind of legal maneuver studied here, can similarly be seen as a knowledge machine that, while flexible, transforms knowledge claims in characteristic ways. The results are never predictable, since unlike more purist actor-network scholars I stress that the substantive issues involved in a particular case are worth pursuing. But highlighting the dynamics of knowledge processes while backgrounding the content and the politics to some extent—an extent that cannot be theoretically determined in advance—can lead to new insights about the taken-for-granted machinery of law. The machinery of law is often dismissed by critical legal scholars as “technical,” with the contents of the “technical” black box being left to black-letter law professors. The approach presented here is largely derived from European social theory, but it has an old-fashioned virtue, namely, respecting the specificity (which I do not regard as autonomy) of legal complexes (Rose & Valverde 1998).

Nuisance or Crime? Street Soliciting and Lay Knowledges of Jurisdiction

The word ‘harm’ has been used more frequently than the word ‘nuisance’ in the context of the discussion concerning the problems related to street prostitution. While the term ‘nuisance’ might be more appropriate from a legal perspective, the affected public felt that it was not broad enough to cover the real impact of street prostitution on their lives. (Canada, Department of Justice, Executive Summary, 1998)

This statement, drawn from the Report of the Canadian Federal-Provincial Working Group on Prostitution, encapsulates the key legal knowledge dilemma in the long-running discussions about what is to be done about the annoyances caused by street prostitution. Is the problem criminal harm? Or is it merely a (municipal) nuisance? In Canada, provinces have their own family and civil law but criminal law is federal; hence, the legal distinction between the tort of nuisance (a provincial offense that municipalities enforce) and the Criminal Code of Canada signals a yawning political and geographical split as well as a jurisdictional matter.

A clever lawyer might note that in addition to the much-prosecuted provincial offense of nuisance, Canada does have an obscure crime of “common nuisance.”Bilson's treatise (1991) notes that this very rarely used section could be used against risks or problems that affect “all of the subjects of Her Majesty in Canada” (as the section puts it) or, more realistically, that affect people more diffusely than garden-variety local nuisances. But she adds that since it has not been so used, it is practically redundant, given the existence of provincial nuisance law. Interestingly, none of the legally learned briefs submitted by provincial attorneys general and other participants to the parliamentary committee holding hearings on the proposed street prostitution law mentioned this potential compromise solution, i.e., the crime of “common” (not public) nuisance.

Instead, two other things happened. At the political level, the debate on the proposed law took the simple form described in the quote above. Those against criminalization argued that soliciting is a mere nuisance, hence not a crime, whereas police chiefs and other nonlawyer participants who wanted to criminalize argued that it was more than a nuisance, hence not a nuisance but a crime. Both sides thus took the existing public-nuisance jurisprudence for granted and respected the traditional jurisdictional divide between the federal criminal law and provincial civil and regulatory law. But at higher legal levels, including at the Supreme Court, street soliciting was not categorized as either a (public) nuisance or a crime: instead, it was labeled as something previously unknown to law, namely, a “social nuisance”—which the majority of the Supreme Court thought constitutes a crime and hence exclusively a federal responsibility.

The political plot went as follows. In the wake of some lower court decisions striking down, as ultra vires, municipal efforts to regulate the nuisance aspects of street soliciting, the federal government introduced Bill C-49, criminalizing all communication in public spaces for the purposes of prostitution, in 1985. “Communication”—rather than “soliciting”—was put in to include (male) johns in the police sweeps of hookers and johns, thus responding to feminist criticisms about the use of prostitution laws to target women (Brock 1998). This bill became law on January 1, 1986, after parliamentary hearings featuring representatives from legal groups, women's groups, two prostitute organizations, police chiefs, mayors, the Minister of Justice, and representatives from affected urban communities, as well as the very vocal Members of Parliament (MPs) serving on the parliamentary Justice and Legal Affairs Committee (Act to amend the Criminal Code (prostitution), S.C. 1985, c. 50).

The new law, which had come into effect shortly after the new Charter of Rights and Freedoms codified and expanded individual rights, was challenged in several unrelated lower court cases. With law enforcement in disarray, several provincial governments collaborated in presenting a question to the Supreme Court regarding the constitutionality of the new antisoliciting law (as well as that of the much older statute criminalizing “bawdy houses,” though little attention was devoted to this). Direct questions from governments to the Supreme Court result in decisions known as References. The Supreme Court issued, in 1990, what was tellingly known as “the prostitution Reference,” even though the criminalization of prostitution itself was not contemplated (“Reference re ss. 193 and 195 of the Criminal Code,” C.C.C. 1990 [3d], at 65).

The federal government and those provinces that intervened in the Reference used, as their main stock of knowledge claims about street prostitution, the transcripts of the 1985 parliamentary hearings on Bill C-49. There was of course a huge outpouring of media and other discourse, but for the sake of tracing a network that includes the Supreme Court, I have relied mainly on the thousand or so pages of those transcripts, plus the briefs submitted by governments and by intervenor groups, and of course the decisions themselves.

Examining the wide range of claims about street soliciting and its risks disseminated along this network reveals that the contrast between “harm” versus “nuisance” set out in the quotation prefacing this section repeatedly appeared as a crucial tool. The point of view that eventually prevailed both in Parliament and at the Supreme Court—that street soliciting is not a “mere” nuisance, being more of a “harm” and thus warranting criminalization—consistently used the nuisance/crime distinction as a decisive switch point; but, interestingly, nuisance was never defined by the pro-criminalization speakers, being used merely as a negative term. Although both provincial civil law and the federal Criminal Code make brave efforts to define nuisance, and case law has articulated some principles, chief among them the quiet enjoyment of one's property (Bilson 1991; Buckley 1996; Cooper 2002), none of the lawyers and judges participating in the debates bothered to define or even describe nuisance. (Two participants did define or describe nuisance, as we shall see, but neither was a legal professional.)

At the parliamentary hearings, neither citizens nor politicians imagined that one could label something a nuisance and still criminalize it. A representative of a Vancouver neighborhood appeared before the parliamentary committee to say, “In our community the word ‘nuisance’ is ridiculous. A nuisance is the person who squealed his Tires … A nuisance is not 400 johns driving around. It is a disaster” (House of Commons [24 October 1985], vol. 4, p. 47). The MP for that neighborhood, present at the same hearing, supported this:

I have had many, many phone calls and would certainly confirm the kind of neighbourhood … [ellipsis in original] not nuisances, as the witness said, but real harassment that has been going on. There is no question the rights of the neighbourhood have to be a very major concern. (House of Commons [24 October 1985], vol. 4, p. 52)

The representative's claim was thus that her own neighborhood's problems were such as to warrant a nationwide new criminal law, but in her quite legally correct view, in order to get Parliament to create a new crime one had to define the targeted activity as a non-nuisance. A Vancouver city official, speaking a few days later, concurred—and helped nationalize (and thus criminalize) the problem by quietly shifting the scale from the neighborhood to the nation-state. Referring to a speech given earlier by the mayor of Niagara Falls, Ontario, floridly describing the negative impact of prostitutes on Niagara Falls' honeymoon business, the Vancouver official stated, “First thing, I think soliciting goes beyond a nuisance nature. The mayor of Niagara Falls and other representatives here addressed that” (House of Commons [29 October 1985], vol. 5, p. 37).

The separation of crime from nuisance was also performed by all the critics of the new law. Like their opponents, they too assumed the federal/provincial, crime/nuisance legal-political binary to be fixed, and simply argued the other side of the same coin. A speaker who, exceptionally, cited common-law principles of nuisance was the mayor of the city of Ottawa, Marion Dewar, the only mayor appearing at the hearings to oppose the criminalization of public soliciting. She argued that criminalization was inappropriate because “it seems to me from all the evidence that the results of street soliciting are of a nuisance nature and relate to the enjoyment of public streets and private property, which can be addressed by municipal regulation” (House of Commons [29 October 1985], vol. 5, p. 7). Similarly, Lucie Pepin, Liberal MP for the Montreal neighborhood of Outremont, also attempted to use legal tools to narrow the scope of the bill during the clause-by-clause discussion in the Commons Committee. The wording of the bill (“… stops or attempts to stop any person or in any manner communicates or attempts to communicate for the purposes of …”) was too broad an attack on freedom of expression and movement as well as on the prostitutes' right to make a living, Pepin argued. She proposed an amendment that would deploy existing nuisance principles to generate a narrowly tailored bill. Her substitute clause read: “… interferes with the use or enjoyment by other persons of that place or of premises adjacent to that place by in any manner communicating or attempting to communicate. …” (House of Commons [7 November 1985], vol. 8, p. 18). This amendment, however, was duly voted down, and thus Canada ended up with a law criminalizing a type of expression characterized not by its harmful effects but merely by its content (for the purposes of prostitution).

Thus, the residents, mayors, city officials, and other speakers at the hearing about the proposed law seemed to know the law better than the Supreme Court: they took it absolutely for granted that if they defined street soliciting as a nuisance, then they were automatically removing it from federal jurisdiction. The Supreme Court, by contrast, proceeded to deconstruct the nuisance/crime legal binary and to thus blur or even erase ancient jurisdictional lines. The knowledge claims deployed to do this were not made wholly explicit in the judgment, but one can detect their iceberg-like presence by paying close attention to some of the textual detail of the judgment and of a few texts that were part of the Court's network. Chief among the knowledges of urban disorder invoked—or, more accurately, evoked, without actually naming it—by the majority of the Supreme Court was the broken-windows theory of disorder and crime.

Nuisance as Crime: Broken-Windows Urban Sociology at the Supreme Court of Canada

Very few witnesses appearing before the parliamentary committee cited scholarly studies; by and large, they took the standpoint of “experience”—even if the witnesses were police chiefs and mayors. But once the local experiences of nighttime noise, harassment of local women by johns, and so forth entered into the strictly legal network, being either synthesized or, more commonly, indirectly alluded to, the neighbors' experience turned into a different epistemological object. Alluding to other, unnamed people's experience generates abstract conceptual objects (urban blight, incivilities, disorder)⁵ whose abstractness is obscured by the constant references to urban “experience.”

The allusive deployment of other people's urban experience—a technique that law shares with politics, but which is sharply frowned upon in the social sciences—was used in several of the briefs submitted by attorneys general⁶ in which government lawyers echoed, without citing it, the theory of urban decay that had become popular in the 1980s: the broken-windows theory of crime and disorder. The broken-windows theory (or more accurately, metaphor), developed by James Q. Wilson and others, was very useful in the Canadian government's campaign to criminalize soliciting (Wilson & Kelling 1982), and hence useful to the Supreme Court judges who wanted to uphold the new law.

From the constitutional law standpoint, the 1985 “communicating” law had serious flaws. By banning communication throughout the public realm, it was on its face overinclusive; this was compounded by the lack of any requirement of evidence of actual harm or even annoyance. The broken-windows thesis, which argues that litter on sidewalks, broken-down cars, and disorderly people are reliable indicators of urban decay even if they do not cause any actual harm, was hence a powerful debating weapon. The broken-windows thesis was not introduced via expert studies. But the arguments about preventive policing and about the cumulative impact of small disorders were imported inside a package with a legal label, namely, the term nuisance, severed, however, from its actual context in Canadian law. One could thus say that the “communicating” statute, which was at risk of failing within the purely legal network of constitutional law, succeeded by being linked to the nonlegal but politically very powerful network of urban disorder broken-windows talk. But the link constructed in the judicial text was not “expert studies” of disorder but rather legal terms—nuisance, mainly, and blight, secondarily—borrowed from urban law but torn out from this context without any explanation of their new meaning in the context of a constitutional challenge involving not zoning law but a criminal statute. A term such as nuisance, if transported into a novel legal context or network, loses much of the determinate meanings it has acquired in the established network of the common law of public nuisance.

A superficial look at the court documents might make one think that extralegal social evidence was indeed used. The briefs by the attorneys general of British Columbia, Ontario, and Canada did footnote some of their claims about the inevitable association of street prostitution with urban decay and degeneration. However, looking more closely, it becomes clear that the footnotes consisted mainly of references to the parliamentary hearings sampled above (which featured no sociologists) and to government reports that contained at least as much unsupported opinion as factual evidence.

Anecdote and opinion from citizens living in areas in which prostitutes work are of course a valid political/legal commodity, even if they are not social science. But the experiences and opinions in question could have served equally well to support a regulatory local strategy, since the experiences were not only confined to a handful of cities, but were actually located only on a few particular street corners. Indeed, the experiences recounted by citizens and city officials are textbook examples of just what enjoyment of one's property (Bilson 1991; Buckley 1996; Cooper 2002), the key common-law principle of nuisance, covers. And perhaps because to lawyers, discussions about street noise would naturally suggest regulation, not criminalization, even high-status legal speakers sometimes forgot that they were supposed to be supporting a crime, not a new nuisance bylaw. The Attorney General of Canada's brief to the Supreme Court thus argued that the impugned statute was constitutional because it was “at worst”“a ‘time and place regulation’”—an obviously false as well as legally counterproductive statement, since there is no time when soliciting is allowed, and the space on which it is banned is the whole of the public sphere, including cars. The fact that criminal statutes are in general distinguishable from local regulation precisely because the criminal code does not deal in time-and-place requirements seems to have escaped the writer, who generated a text that, like the Supreme Court majority decision, completely deconstructed its own key binaries (harm versus nuisance, crime versus regulation).

Similarly, when introducing the bill in the House of Commons, Justice Minister John Crosbie slipped into the (municipal) language of streets and neighborhoods:

The purpose of this Bill is to help the citizens of this country who live in certain of our major urban areas and the police forces of this country to regain the streets because they have lost control of the streets and neighbourhoods in certain urban areas of this country …. (cited in Factum of the Attorney General of Canada, Par. 5; emphases added)

The neighborhood, of course, is not an element of the criminal law: it is, instead, the key figure or object of municipal zoning and planning.

Neighborhood versus nation; municipality versus central state; nuisance versus the criminal code—these pairs were, as we saw earlier, deployed as opposites in order to produce the “nuisance versus crime” argument around which the anti- and pro-criminalization forces arranged themselves. But the pairs of socio-political-legal opposites were constantly being brought back together again, in a move that logically contradicted the first but seems to have worked well to get the law passed. Crosbie's speech embodied the “nuisance is crime” strategy perfectly, since, although it was part of a speech calling on Parliament to create a new crime, it could just as well have functioned as the “whereas” clause justifying a municipal zoning bylaw.

When all of this reached the Supreme Court, there was clearly an embarrassment of legal as well as political and symbolic riches. The Court dealt with numerous arguments, such as discussions about freedom of commercial speech, but here we focus only on the network of nuisance/crime.

We saw earlier that the Attorney General of Canada's brief claimed that the new crime of communicating for the purposes of prostitution was a mere “time and place regulation,” and was hence not out of keeping with the Charter of Rights. Elsewhere the brief described the harms targeted by the new criminal law as a “public nuisance” (Par. 3)—a curious maneuver, given that public nuisance is not within federal jurisdiction. But then, in the next paragraph, the writer hit on a novel term that is not quite as reminiscent of municipal governance: “social nuisance” (emphasis added). Prostitution itself is legal, the writer stated, but the transactions around it cause a “social” nuisance (Par. 4). (That the transactions have to be conducted in public because brothels are illegal in Canada is of course not mentioned.)

The Ontario government's brief did not employ the term public nuisance, but it otherwise made the same moves as the federal brief, even using the same nonlegal term:

By way of conclusion, it is respectfully submitted that when one weighs the serious social nuisance caused by prostitution and the valid and real objective in Parliament in enacting s.195.1, against what at best can be considered a limited interference with commercial conduct, the scales of justice clearly tip substantially in favour of the propriety of s. 195.1. (Factum of the Attorney General of Ontario in Reference, Par. 85; emphasis added)

Taking his cue from these briefs, Supreme Court Chief Justice Brian Dickson's majority decision first of all noted that the statute itself was a response to broken-windows sensibilities: “[t]he Criminal Code provision subject to attack in these proceedings clearly responds to the concerns of homeowners, businesses, and the residents of urban neighborhoods …” (Reference 1990:73). That the concerns of homeowners living near certain notorious street corners could be addressed with less drastic municipal regulatory measures was not explicitly discussed in Dickson's text—his decision relied on an indirect argument against the counterproposal that had been put forward by Ottawa Mayor Marion Dewar and others. Deploying the broken-windows thesis—that one broken window or one prostitute may not constitute a harm, but the proliferation of many non-harms results in a supra-individual harm—he argued that it was possible for Parliament to add a series of merely possible nuisances and get an actual crime:

While it is the cumulative impact of individual transactions concentrated in a public area that effectively produces the social nuisance at which the legislation in part aims, Parliament can only act by focussing on individual transactions. (Reference 1990:74; emphasis added)

As the Chief Justice of Canada knows, the Parliament of Canada can only act on individual transactions because the Criminal Code is its only legal tool: legal tools providing for time-and-place regulation are provincial. By ignoring the central issue of jurisdiction, Dickson wove a magic carpet that took him—and the legal network—from “the federal Parliament” to “law as such,” instantaneously.

The argument about many small inconveniences adding up to one big harm justifying the prosecution of all participants in each annoying act was borrowed from the materials presented to the Court, and it embodied broken-windows-type knowledge of the laws of urban disorder as well. In a related appeal of a Nova Scotia prostitution-related case, the Ontario attorney general intervened—with a brief that was included in the 1990 Reference materials—to argue for the existence of another legal creation covering the same ground as “social nuisance,” namely, “cumulative harm” (Brief of Attorney General of Ontario in R. v. Skinner (1990), Appendix to Attorney General of Ontario's submission in the Reference). The word cumulative reappeared in the Dickson passage just cited.

The Federal Department of Justice's point man on the statute, Rick Mosley, had made a similar claim at the parliamentary hearings held several years earlier, although as a pragmatic law enforcement argument, not as a substantive claim about social evils. Mosley explained to nonlawyer parliamentarians why it was necessary to criminalize communicating for the purposes of prostitution.

It is very difficult, senator, to establish the causal nexus between the conduct of an individual and the resulting nuisance. The problem in most of the communities that are affected by street prostitution is that there is a cumulative effect. It is the result of so many prostitutes being present, so many prospective customers and so many individuals who are simply cruising to watch the action. To link the nuisance occasioned to the residents of that district to the conduct of one individual would be virtually impossible in a criminal prosecution. … (House of Commons [3 December 1985], vol. 30, p. 19; emphasis added)

Some of the evidence presented along the network suggests that some police forces, in the years leading up to the new statute, had refrained from vigorously enforcing various public-order laws that could have been used to mitigate the problems, precisely in order to promote the sort of chaos that could be expected to lead to a new criminal offense. And during the appeal of the Halifax Skinner case (R. v. Skinner 1990), it became evident that soliciting-related disorder had increased in Halifax only after the police had, for unrelated reasons, gone on strike.

But the political and legal activism of the police, mentioned a few times by prostitute groups and their allies at the hearings, were cleansed from the strictly legal components of the knowledge network. Attorney general affidavits and court judgments contained no references whatsoever to the police as active participants in the lawmaking process.

In a judgment concurring with Chief Justice Dickson, Supreme Court Justice Antonio Lamer agreed that the communication law was a piece of social engineering. According to him, it targeted not the usual object of criminal law—the individual's will—but rather a sociological entity, a risk entity to be exact, also described as “cumulative.” And while happily using the term nuisance to describe a crime, he added another term, not used by other participants, taken straight from the field of urban planning and zoning: blight. Once more drawing on the powerful cultural resources of the broken-windows school, and on the equally powerful, convergent legal tradition of preventive police powers (Dubber 2004), Lamer wrote,

I do pause to note, however, that the appellants correctly point out that the act of soliciting by a single prostitute may not by itself produce a nuisance. But this argument, with respect, misses the point that the legislation is designed to prevent the congregation of prostitutes and customers in the streets. It is the cumulative effect of this congregation that produces nuisance and blight. (Reference 1990:120; emphasis added)

The metonymic slippage between the physical disrepair of buildings on seedy downtown corners and the moral risks posed by prostitution had already been performed (like so many other of the Supreme Court's knowledge moves) by the attorney general of Ontario's brief, which had stated that the cumulative effect of soliciting is to cause “blight” (Par. 26). Blight, here as in urban law, designates both a present harm and a future risk—or, more accurately, the term blight actually travels the considerable distance between present, documented harm and future risk. Even if there is no actual, present harm, the Supreme Court joined James Q. Wilson in telling us, it does not matter, since these non-harms are sure indicators of future risks, of future decay. And the Court thus need not inquire into the actual effects of prostitution, nor demand that prosecutors introduce evidence of actual harm to specified parties.

Importing urban law terms into the inquiry into the constitutionality of a criminal statute, while legally dubious, turned out to be a highly productive knowledge move. Using expert studies of the empirical links between disorder and crime—a highly plausible alternative—would have drawn too much attention to the specific causal claims made by the experts and by the courts' summaries of the experts. Using urban law terms instead had the effect of making controversies about scientific claims irrelevant and unnecessary. That the terms were not used in their actual legal meaning, that they were used more as analogies or metaphors, was only noticeable if one paid very close attention to the (black-letter) details of nuisance law, something that only a couple of nonlawyer politicians (Dewar and Pepin) bothered to do.

The final point that needs to be made in this case study is that social nuisance was also invoked, again without either definition or authority, in the dissenting judgment by Supreme Court Justice Bertha Wilson, also signed by Supreme Court Justice Claire L'Hereux-Dubé. They argued that street soliciting need not be criminalized, since it does not cause the John Stuart Mill type of harm that, in their view, the Criminal Code is meant to target. They did admit that soliciting disrupts neighborhoods; but disruption, even disruption that is more than an individual problem and has assumed the proportions of a “social” nuisance, is not necessarily a criminal harm. The criminal law is “not a proportionate way of dealing with the public or social nuisanc …” (Reference 1990:141).

Suddenly separating the two entities that her colleagues had joined together, Wilson concluded that the legislative target was “a social rather than a legal nuisance” (Reference 1990:130; emphasis added). In both majority decisions, the urban sociolegal network of nuisance and blight was made to converge with the constitutional legal network in such a way as to justify the creation of a new crime—a crime whose harm does not lie in the actus reus or the mens rea but in the sociological, aggregate impact of a certain type of behavior. In the dissent, however, the social network diverges from law. This is the same move that, in the 1960s, enabled courts and politicians to decriminalize fields of conduct that were deemed as socially questionable or undesirable, but as not within the scope of the criminal law. The two majority decisions, by contrast, reproduce the shape and the content of the knowledge network of broken-windows criminology, whose political effect is to justify the use of coercive law to govern “social issues,” i.e., aggregate risk entities.

The City as an Ethical Machine: Hybrid Knowledges of Urban Vice in U.S. “Adult” Zoning

The above case study has highlighted some of the knowledge moves through which the Canadian government gave urban police forces and moral reform mayors renewed powers by means of a new crime, and has described how these knowledge claims were converted by the Supreme Court into the power to use federal criminal law to govern localized nuisances on a few street corners of urban Canada. By contrast, the efforts made in the United States to use zoning and other minor legal practices of urban regulation, e.g., to clean up Manhattan, driving out porn shops as well as panhandlers, drew only on existing tools of urban law—tools that, especially in the United States, with its powerful First Amendment, are better than criminal statutes as weapons in urban-cleanup campaigns. National differences in law, however, should not be exaggerated. It is true that in the United States today it would be very difficult to criminalize any form of “communication” not associated with the specter of terrorism; but nevertheless, it has not proved impossible to use the criminal law against all manner of moral risks that may or may not turn into actual harms in the future (Dubber 2004).

Some commentators believe that the use of zoning bylaws for social/moral ordering purposes is a new tactic used by the New Right to effect their puritanical plans (e.g., Papayanis 2000). But when put in the context of the history of U.S. zoning and urban law generally (e.g., Frug 1999; Novak 1996; Smith 1996), the adult-zoning bylaws do not appear exceptionally draconian. In fact, the opposite is the case: businesses that happen to enjoy some First Amendment protection (and stripping as well as reading porn have in recent years come to be considered as expressive activities) have a much greater ability to challenge urban regulations than do the businesses that see a lot of popular speech inside their walls (hair salons, billiard halls, etc.) but whose age-old targeting by licensing and zoning techniques is rarely subject to judicial review. This is why we have appellate court cases on adult zoning and on strip club regulation (and on billboards, too), whereas we have virtually no appellate court cases through which one can investigate how municipal or state authorities justify onerous regulatory provisions applying to other establishments traditionally subject to special regulations.

This case study focuses primarily on the May 2002 decision of the U.S. Supreme Court in the Los Angeles porn shop zoning case, City of Los Angeles v. Alameda Books (2002). The case and its predecessors are not analyzed for their legal or local-politics peculiarities but are rather used to highlight certain continuities—at the level of knowledge processes—running throughout urban zoning law. Although a comparative study is not possible in the strict sense because other common-law countries do not have First Amendments, nevertheless, the knowledge moves that are made in judicial review by U.S. appellate courts monitoring municipal action against disreputable businesses are by no means unique.

Two moves in knowledge production that were key bricks in the constitutional wall built to protect adult-zoning ordinances are worth studying, since they recur often in the epistemology of urban law. Both involve the happy-go-lucky mixing of values and facts from different realms. First, there is a marked ontological heterogeneity that refuses to draw any line separating moral, economic, and physical risks and dangers. That the architectural design of a public housing project can incite tenants to virtue and achieve engineering excellence is the sort of hybrid hope, held by urban reformers since the eighteenth century (Joyce 2003; Williams 1985), that necessarily produces ontological heterogeneity. This mixing of moral and economic values, of cultural and material rationalities, has certainly been noticed by scholars, who, assuming that scientific purity is a Good Thing, often chastise urban planners as well as appellate courts for “contaminating” practical urban reform projects with moral and aesthetic desires.

The second kind of heterogeneity, by contrast, has not been thus far noticed in the literature. This is a temporal heterogeneity. We already saw, in the prostitution case study above, that political and legal discourses on street soliciting were remarkably cavalier about causality, not usually distinguishing the claim that soliciting has already caused identifiable harm from the temporally quite distinct claim that the habitual use of urban spaces for purposes of sexual transactions might at some future point bring about social and/or economic evils. In Canadian jurisprudence this can be easily identified as the slippage from “harm” to “risk” performed by and in the ubiquitous phrase “risk of harm” (compare with Harcourt 1999).

In U.S. zoning law, one common term that both embodies and performs this temporal hybridity is blight, a word that, in an excellent example of the convertibility of words and things in legal networks, became a wrecking ball by being used to justify the tearing down of whole neighborhoods in the 1950s and 1960s. If a certain block was declared a “blighted area,” this meant that it could be torn down, since—the argument went—if it was not actually dangerous now, it would inevitably breed economic, architectural, and social decay sooner or later. In the context of adult zoning, in the 1986 Renton decision about porn theatre zoning (largely upheld in the 2002 Alameda Books decision), Justice Rehnquist cited precedents (i.e., legal words), rather than either Renton (a suburb of Seattle) citizens' experience or the reports of experts, to conclude that “the location of adult theatres has a harmful effect on the area and contributes to neighborhood blight” (Renton v. Playtime Theatres, 475 US 41, 51 [1986]). Here as elsewhere, “blight” is both a physical, presently existing harm and a future risk.

This temporal ambiguity or hybridity replicates the curious temporality of the “blighted lives” of Victorian social novels, faithfully reflected in late Victorian cases involving bankruptcy, libel, and breaches of promise of marriage. In the jurisprudence of “blighted hopes,” it was not necessary to prove that the plaintiff had already suffered measurable harm; courts took it as common knowledge that the plaintiff's future would be dim indeed. Ill effects could thus be presumed. Since the blighted hopes jurisprudence used exactly the same temporal logic as the broken-windows theory, it is perhaps not too digressionary to quote just one among a myriad of blighted hopes tort decisions⁷:

In contemplation of law, reputation is a delicate plant, withered by the breath of scandal. Any publication which imputes to another conduct which right-thinking men condemn … is presumed in law to have injuriously affected the reputation of the person so assailed, and, by such injury, to have caused him some damage. … Every one recognizes the blighting effect of scandalous utterances. … Every one recognizes that such assaults, publicly made, tend injuriously to affect the reputation and standing of the one so assailed among his fellows. It is from the recognition of this that the law implies damages, without allegation or proof of special damages. (Hughes v. Samuels Bros., 159 N.W. 589 [Iowa 1916])

This is the same logic as that of the original “broken windows” article (Wilson & Kelling 1982), which claimed that even if panhandling, public drinking, and broken windows do not actually lead to crime or other serious harms, nevertheless, they damage the community's reputation, and this damage to perceptions and feelings automatically turns into actual harm and decay. Presented by Kelling and Wilson as a new thought, this blight logic goes back to the days when zoning law emerged out of nuisance law and had been common currency in urban reform circles for decades. Thus, Mel Ravitz, an influential urban sociologist, provided the sole expert witness affidavit in the landmark 1970s Detroit adult-zoning case, stating:

That the ordinance in question seeks to restrict the concentration of specified uses on the ground that any such concentration is an actual danger sign in a residential neighborhood. It is such a threat in that it leads people to believe that the neighborhood is declining and more of the kind of people who accept a declining neighborhood will flock there. This process hastens the reality of neighborhood decline. (Exhibit 1 in Young v. American Mini Theatres 1976:5)

The temporal hybridity of discourses about urban vice that is performed through and in the term blight is thus but the tip of a powerful urban law discourse. Within the “blight” network, the reputation of a neighborhood is taken as something one can presume to be a social fact—not in sociological studies, whose logic would eschew mixing opinion with fact, and would distinguish present from future, cause from effect, but in legal networks that wish to invoke urban sociology without deferring to social science logic. The legal network thus assembled admits and enables temporal hybridity in a way that contrasts sharply with the insistence on separating cause from effect, past from future, found in most science. But, usefully for law, in urban sociology—at least in the version of it that promotes law-and-order measures to preventively control disorder—the temporal hybridity is embraced (though often somewhat awkwardly).

The Detroit ordinance upheld in the case in which Ravitz testified (Young v. American Mini Theatres 1976), afterward used as a model by municipalities around the country, contained a passage approvingly cited by Justice Stevens in his justification of the ordinance:

There are some [land] uses which, because of their very nature, are recognized as having serious objectionable operational characteristics, particularly when several of them are concentrated … special regulation of these uses is necessary to insure [sic] that these adverse effects will not contribute to the blighting or downgrading of the surrounding neighborhood. (Young v. American Mini Theatres, 427 US [1976])

If blight is inherent in some land uses, rather than being a potential risk that could be averted through precautionary regulatory requirements or through prudence on the part of the proprietor, blight exists already as soon as the porn store opens. Which means, conveniently for the municipalities and for the appellate courts, that few if any inquiries need to be undertaken to determine the actual effects of such shops, in the present or in the (scientifically predictable) future. The future effects can be presumed to be negative because the land use in question is “inherently” objectionable.

It is already evident from the foregoing that the ontologically and temporally hybrid objects of twentieth-century urban law and regulation (blight, renewal, downgrading, etc.) persist today. But current cases exhibit some new knowledge moves not found in older zoning decisions, such as Berman v. Parker (1954) and Village of Belle Terre v. Boraas (1974). In the older cases, the desire of white middle-class patriarchal families to isolate themselves from all others in leafy suburbs was taken as auto-effective. In the newer cases, by contrast, two features relevant to the question of knowledge production emerge. The first is an expectation that the municipality will produce some evidence of harm/risk. The second is a demand that the regulation in question be described and justified as “narrowly tailored” or “targeted.” Just what counts as “narrow” is rarely specified, but there is an indeterminate expectation that municipal regulations be both legally targeted (by having some kind of list of what counts as an objectionable activity or establishment) and spatially targeted, in the sense of acting only on some portion of urban space, and not—as in the Canadian criminal statute analyzed above—over the whole of the public sphere. Let us take each of these features in turn.

Indeterminate Utilitarianism: Producing Evidence of Harm to “Communities”

In 1986, the municipality of Renton had sought a Supreme Court decision that would allow it to zone porn theatres out altogether, and part of its evidence was a study of the negative effects or associations of such businesses. However, Renton had no such businesses when the ordinance was drawn up. Thus, the study was necessarily borrowed. That studies of much larger and more diverse cities would not be factually relevant in the Renton situation would be the first objection made by a sociologist reading the Renton decision, but of course sociologists were not the intended audience. And the intended audience, the Supreme Court, ruled that any municipality in the United States could use virtually any urban study for the purpose of showing that it had acted legally in drawing up special adult zoning ordinances. Understanding this apparently cavalier knowledge move requires peering into the legal black box of judicial review. To speed us along the network, we proceed to examine Justice O'Connor's summary—in the 2002 City of Los Angeles v. Alameda Books decision—of the Renton precedent, since it is in this later form that the Renton decision now does its work.

A key bone of contention among the Supreme Court Justices in the 2002 case was the role of a study that had been carried out by Los Angeles city officials (not independent experts) in 1977 and introduced in evidence in the 2002 case. This had compared crime rates in some areas of Los Angeles that had many porn shops with crime rates elsewhere in Los Angeles.⁸ Justice Souter argued that this study hardly constituted the kind of evidence of harm that needs to be produced if a municipality is going to interfere with the First Amendment rights of porn sellers and porn consumers. But, speaking for the majority, Justice O'Connor critiqued Justice Souter's more utilitarian/realist dissent in the following terms: “Justice Souter asks the city to demonstrate, not merely by appeal to common sense but also with empirical data, that its ordinance [about porn shops] will successfully lower crime. Our cases have never required that municipalities make such a showing” (City of Los Angeles v. Alameda Books, 535 US 425, 439 [2002]).

Souter never wandered so far from established zoning jurisprudence as to propose that the appellate courts should adjudicate the question of which policy is best for lowering crime. Souter is not Roscoe Pound. He merely asked for more relevant studies to be introduced in evidence (for one thing, the only city regulation challenged by the porn shops was a rule about not having two porn shops under a single roof, a question not contemplated in the 1977 study). O'Connor's unfair summary obscures from view the fact that the mechanism of judicial review offers appellate judges a handy tool for managing thorny knowledge issues by bracketing and rendering irrelevant many of the issues that would greatly preoccupy scientific actors. Souter merely wanted a different, more precise kind of bracketing.

The study's contents were rendered irrelevant through a specifically legal knowledge move that shifted attention from substantive claims about the city and focused it on the beliefs of the party submitting the study in evidence. The key dynamic or kinetic property of a city study deemed to be appropriate evidence for judicial review purposes was that, whether it was actually relevant or not, a municipality had to have a reasonable belief in its relevance: “[i]n Renton, we specifically refused to set such a high [evidentiary] bar for municipalities that want to address merely the secondary effects of protected speech. We held that a municipality may rely on any evidence that is ‘reasonably believed to be relevant’…” (City of Los Angeles v. Alameda Books, 535 US 439).

Thus, in Renton, the question was as follows: Do municipalities act reasonably in believing that a decades-old study about crime rates is perhaps not determinative but at least “relevant” to the (new) zoning requirement that porn shops be 500 feet from schools and churches and 1,000 feet apart from one another?⁹ However odd the chain of reasoning just laid out sounds to a social scientist, given the powers of the judicial review black box to generate valid legal currency out of epistemological odds and ends by sidelining questions of objective “fact,” the normal judicial answer to the constitutionality question was yes. In the Alameda Books case, the impugned ordinance was somewhat different, involving a prohibition on operating two porn shops under the same roof. But since the facts about porn shops and their effects had been successfully sidelined already in Renton by shifting from objective facts to corporate belief, invoking Renton in 2002 worked to repeat the same knowledge move in a different context, thus efficiently effecting the sidelining of quite different facts.

However, O'Connor did not rely only on the mere, formal existence of a study and on the city's corporate reasonable belief in relevance to arrive at the conclusion that the city of Los Angeles had acted reasonably. Grabbing the most venerable object within the common-law network that a judge could find—“experience”—the key passage in her decision stated, “[t]he proposition to be shown [by the city] is supported by common experience and a study” (City of Los Angeles v. Alameda Books, 535 US 425; emphasis added). Blackstone's common experience having suffered blows from multiculturalism and feminism as well as attacks by social science, O'Connor's text was careful not to simply take the side of the common experience as against science. Just as Souter is not Roscoe Pound, so too is O'Connor not a pure legal formalist. She thus added that the “reasonably believed to be relevant” standard “is not to say that a municipality can get away with shoddy data or reasoning” (City of Los Angeles v. Alameda Books, 535 US 438).

Just what would count as “shoddy data” was left unspecified, but a clue was given in an apparently extraneous comment made at the end of O'Connor's text. The clue concerned not scientific criteria of validity but rather the strictly legal questions of jurisdiction. O'Connor stated that zoning law consists of an “elaborate web of land-use regulations intended to promote the social value of the land as a whole” (City of Los Angeles v. Alameda Books, 535 US 458; emphasis in original), and that evaluating the social value of land and land uses (as distinct from economic value, which is presumably within the jurisdiction of appraisers and real estate experts) is something that municipal councils do. Regarding the social value of land,¹⁰ she stated with finality, “[t]he Los Angeles city council is in a better position than the judiciary to gather and evaluate data on local problems” (City of Los Angeles v. Alameda Books, 535 US 458). We still don't know what counts as shoddy data, but we know that city councilors are authorized producers of such data.

This repeats, almost word for word, a passage in the related nude-dancing ordinance case of City of Erie v. Pap's AM (529 US 277 [2000]). Here too the Supreme Court managed to put city councilors on an epistemological pedestal, not by evaluating either their credentials or their filing cabinets, but by quietly slipping from legal authority to epistemological authority. Contrary to what one sees in the cases—e.g., toxic torts—in which experts' voices trump those of local elected officials, in relation to adult zoning, the fathers of the city know best. Of course this slippage is a venerable feature of the common law: in the Blackstonian paradigm, the sovereign knows best in part because he has jurisdiction—or perhaps he has jurisdiction because he always already knew best, the logic being implicit and hence fuzzy. But rather than quoting Blackstone, the Supreme Court preferred to invoke the more contemporary evidentiary commodity of personal experience, “first-hand knowledge” (City of Erie v. Pap's AM, 529 US 298 [2000]).

Now, firsthand knowledge—eyewitness evidence—is generally contrasted to expert evidence. However, in this text, the councilors' knowledge, while firsthand, was not exactly eyewitness knowledge. First, none of the councilors needed to have actually visited the establishments in question. Second, the knowledge the Court was imputing to them was regulatory, not experiential: they were presumed to always be wearing their city-father hats when prowling the tenderloin districts, not participating as willing consumers. Gilding the knowledge lily, the Court then claimed that firsthand knowledge, of sexual vice at any rate, qualified councillors as experts:

The council members, familiar with commercial downtown Erie, are the individuals who would likely have first-hand knowledge of what took place at and around nude dancing establishments there, and can make particularized, expert judgments about the resulting harmful secondary effects. (City of Erie v. Pap's AM, 529 US 298)

In this decision, then, which most lawyers would simply label as deferential to municipalities, we can see that what goes on inside the black box of “deference” is that courts construct a epistemological hybrid made up of (1) jurisdiction, (2) firsthand knowledge, and (3) expert knowledge. And to complicate the recipe even further, none of these, except perhaps jurisdiction, occurs in its “pure” form, since as we have seen the city councilors are not said to have direct experience as consumers of the sexual pleasures on offer, only experience of the resulting problems; and neither are they “experts” in the usual sense, since their expertise derives solely from their election as councilors.

Justice Kennedy's concurring judgment in City of Los Angeles v. Alameda Books (2002) was less deferential to municipal cleanup campaigns. Although he agreed with O'Connor that the city met the reasonableness standard by producing a study that was consistent with “common experience,” he put the study first and common sense second. Nevertheless, he concurred with O'Connor's plurality that when it comes to knowledges of what he called the “sordidness” (another moral-physical hybrid) created by adult establishments, “the Los Angeles City Council knows the streets of Los Angeles better than we do” (City of Los Angeles v. Alameda Books, 535 US 458).

Here as in many other networks, the object of the councilors' firsthand/expert knowledge, “the streets,” is not coterminous with any set of addresses. “The streets” exist here somewhere between the physicality of the built urban environment and the purely legal space that is created by jurisdictional rules assigning streets to local government, highways to states, and air corridors to the federal government. That “the streets” invoked in this and other decisions are not quite the ones upon which the rest of us walk or drive is clearly evident in the fact that Kennedy's sentence, while ostensibly referring to Los Angeles, would not work if he had said “the freeways” instead of “the streets.”“The street” is a cornucopia of cultural, legal, and symbolic capital: “the man on the street”; “streetwise”; “the Arab street.” Because municipalities have jurisdiction over streets, as any citizen complaining about potholes knows, municipal governments can borrow the rich symbolic capital of “the streets” without any mediation and instantly turn it into municipal legal capital. Jurisdiction can be magically turned into authorized knowledge of the urban.

The dissent written by Justice Souter at first distanced itself from the judicial conflation of jurisdiction with knowledge of urban vice, since its key argument was that the city did not provide sufficient relevant evidence of harm.

Requiring empirical justification of claims about property value or crime is not demanding anything Herculean. … These harms can be shown by police reports, crime statistics, and studies of market value, all of which are within a municipality's capacity or available from the distilled experiences of comparable communities. … And precisely because this sort of evidence is readily available, reviewing courts need to be wary when the government appeals not to evidence but to an uncritical common sense in an effort to justify such zoning restrictions. (City of Los Angeles v. Alameda Books, 535 US 458)

While sharing social science's disparagement of common sense, Souter nevertheless did not reach the social science conclusion, namely, that if common sense is to be interrogated, independent experts ought to be asked to produce expert evidence. Souter wanted evidence, but not from social scientists:

Increased crime, like prostitution and mugging, and declining property values in areas surrounding adult businesses, are readily observable, often to the untrained eye and certainly to the police officer and urban planner. (City of Los Angeles v. Alameda Books, 535 US 458)

What Souter did was thus not taking the side of the social, but rather deploying one dimension of old police science/police powers (administrative knowledge; Levi & Valverde 2001) as against another dimension (the power/knowledge of elected city councilors). Souter wanted the city staff to produce its own reports instead of letting city councilors use their firsthand/expert knowledge of urban vices.

The Porn Shop as a Moving Target: The Aporias of “Narrowly Tailored” Ordinances

Apart from using—or rather invoking—studies, the other knowledge move that distinguishes the more recent adult-zoning cases from traditional zoning law is the demand that zoning ordinances (at least those respecting businesses and persons protected by the First Amendment) be “narrowly tailored.” Thus, municipalities cannot prohibit panhandling throughout the city, but they can ban panhandlers from certain locations. In the case of adult zoning, current law has it that municipalities cannot ban all pornography sales; but bookstores and video arcades that count as “adult” establishments can be subject to special regulations such as the requirement that these businesses be more than 1,000 feet from one another. As a matter of urban experience it is not difficult to distinguish a convenience store that sells porn from a porn bookstore, but law is a very blunt tool for the regulation of consumption.

In fact, adult-zoning ordinances do not even try to define a pornography business. Instead, they do something that has defeated generations of judges: namely, they define pornography itself, that is, the kind of publication that constitutes “adult” material. In the definition recycled in countless municipal ordinances across the United States, including Los Angeles, adult material is defined as featuring “specified sexual activities” and/or “specified anatomical areas.”

Because of its length and pornographic attention to detail, the list of acts and body parts gives an impression of great precision and hence appears as an appropriate tool for modern, tailored, “smart,” targeted governance (Valverde 2003b). Unfortunately for judges, its precision is legally beside the point. This is because the adult-zoning ordinances are not like film classification schemes. The publication's character is merely a clue to the character of the business in question, since municipal licensing and zoning concern establishments, not publications. And great imprecision is found at the key switch point linking the publications to the business. An adult business is defined merely as one having a “substantial” or “significant” part of its stock in adult publications.

Let us thus turn to the business classification work of the Los Angeles zoning inspectors, the personnel who in their practice construct the meaning of “adult.” This is documented in the brief submitted by the porn bookstores in the Alameda Books case (2002). The lawyer for the bookstores first described the preliminary phone calls made to ensure compliance with the zoning regulation, going on to express a no doubt disingenuous shock about how administrative discretion operates.

In not one of the proceedings in which I [the lawyer] was involved did any of the City officials ever give any objective guidelines or criteria … the City officials candidly admitted that it was a completely subjective text and it depended upon whether the particular location struck the particular City zoning inspector as an ‘adult bookstore’ or an ‘adult arcade.’ (Petitioners' brief, Alameda Books, 2000 US Briefs 799)

With the brief, the lawyer for the businesses submitted excerpts from the depositions made by zoning inspectors.

Q. [Lawyer] What if it [adult content] were only 1 percent of the location, could it still be an adult bookstore?

A. [Inspector] You're asking me to speculate. It's conceivable. Again, every situation is different and should be weighed on its own. (Petitioners' brief, Alameda Books, 2000 US Briefs 799)

The bookstore's lawyer asked a different official to confirm or deny the existence of this huge discretion:

Q. When Mr S was using the word ‘quantity’ in terms of his questions to you regarding the standard to use in evaluating whether a bookstore was an adult bookstore or not, did you construe those questions to be related to the percentage of material involved?

A. Yes, I do.

Q. So it is your belief that there is no particular percentage of inventory of a bookstore that would qualify it as an adult bookstore under the Los Angeles municipal code?

A. To the best of my knowledge, there is no definitive percentage within the municipal code. (Petitioners' brief, Alameda Books, 2000 US Briefs 799)

The Los Angeles zoning inspectors are not unique.¹¹ Like other officials provided with checklists or “best practices” guides or other tools that act as norms but not as hard and fast rules, they are hard pressed to explain just how they make their determinations. Some common knowledge, some administrative traditions and precedents, and some knowledge of what a judge might say if there was a court challenge are elements in the rich epistemological mix deployed by administrative officials in their daily work. But the key point here is that listing every conceivable form of sexual activity that makes a publication pornographic—which is what the definition section of the ordinance does—does not work to make the ordinance “narrowly tailored.” That is because the list of sexual activities and zones can only, at best, identify a publication, not an establishment. Zoning inspectors (who are not plentiful in cash-strapped twenty-first-century cities) can hardly undertake a full content analysis of every item, and so must make global, gestalt-like judgments.

The attempt to rationalize, modernize, and narrowly tailor the governance of sexual representations thus founders on a specifically legal shoal: a conflict of ontologies. “Establishments” are the entities regulated by municipalities, not magazines or videos, but publications are the entities relevant to the very different legal network of First Amendment law. The two sets of entities are brought together by the magic of the ordinance's text, but the magical operation is quickly undone as the defense lawyer cross-examines the hapless inspectors. Like the appellate judges who look for ways to turn factual information into more formal objects, the front-line officials too have to manage the vast amount of information that is available to them by performing certain common bureaucratic moves—including using a global judgment rather than a detailed analysis of each publication to decide whether a bookstore is a porn shop. In their everyday work they easily black-box the fragile link between “specified sexual activities” marking a publication as pornographic and the designation of the entire establishment containing the publication as “adult.” When the black box is opened at the deposition by cross-examination, the result, for the participants, is to make the inspector look foolish. The more important insight at the level of network analysis, however, is that the networks of urban zoning inspection are very difficult to merge into or even combine with the network of First Amendment jurisprudence, not least because there is a significant ontological break separating them. Publications and establishments don't quite seem to belong in the same drawer, whether looked at legally or from the point of view of everyday experience.

The necessary failure of the attempt to contain damage to freedom of expression by “tailoring” zoning regulations is hence clearly rooted in broader clashes of networks. The zoning powers of municipalities are, by definition and by the history of the common law, extremely broad and extremely intrusive (Novak 1996; Frug 1999). The clash between police power and constitutional rights is sometimes addressed through bureaucratic efforts to make regulations “smart,” as they say in pharmacology and in the military. But serious legitimation problems arise when narrowing and targeting and tailoring work is done not by changing the text of the regulation but rather by waving the First Amendment across the courtroom. The First Amendment appears, within zoning law, as an extraneous actor whose protection can only extend to individual publications—not to places or businesses—and who works exclusively through the logic of rights. The First Amendment is necessarily at odds with the risk logic of urban regulation, which is concerned only with “establishments,” places, and times, and whose logic is preventive and aggregate and thus at odds with individual-rights logics. These underlying conflicts remain black-boxed most of the time but are sometimes opened up, especially in the course of challenges to the state's police power. Then, appellate courts, faced with a sort of unraveled Penelope's cloth, are forced to try to knit it back together or, to mix metaphors, to close the black box again, since it is their job to try to produce as smooth a legal fabric as possible.

Conclusion

Let us review the main findings before concluding with a methodological comment. We first saw in the Canadian study that ordinary witnesses discussing their experience of living alongside prostitutes tended to use the crucial terms nuisance and crime more legalistically than the Supreme Court: they thought they had to prove that prostitution was more than a nuisance, hence not a nuisance, in order to argue for its criminalization. The lay witnesses were thus taking the legal network for granted and trying to use it without changing it. The Court, however, was happy to (unanimously) claim that street soliciting constituted something previously unknown to law: a social nuisance. While the residents and mayors and other witnesses tried hard to argue along established legal lines, then, the Supreme Court preferred to rely on a secondhand version of a certain sociological knowledge: namely, the broken-windows theory of urban disorder and crime, to justify a criminal statute that, if considered strictly within the network of constitutional law, would be difficult to justify given the new Charter of Rights. But the broken-windows theory did not appear under its own name, being partly hidden by the apparently legal terms nuisance and blight, used by the Court in a nonlegal manner. Thus, the Supreme Court did not take established jurisdiction seriously, deliberately blurring the line between nuisance and crime, whereas other actors took jurisdiction as a fixed given and proceeded to argue within the parameters of that framework.

In the judicial review of ordinances undertaken by the U.S. Supreme Court, by contrast, municipal jurisdiction appeared as a taken-for-granted black box to all parties: the criminalization of “adult” establishments by federal or state governments was never considered. Precisely because U.S. courts have by and large authorized municipalities to govern urban vice with a great deal of discretionary power, appellate courts in the United States have not, in the adult zoning cases at any rate, made their own findings about whether this or that nuisance amounts to a crime; rather, they have simply endowed city councils with rather fantastic epistemological powers. By the same decisions that affirm their ancient jurisdiction over streets and over urban disorder (as opposed to crime), municipal councils are also endowed with more epistemological authority than they asked for (city councilors did not claim to appear as expert witnesses). Reliable knowledge of urban ills and their solutions is imputed to them by courts. This knowledge turns out to be a curious hybrid of (1) jurisdiction itself, (2) “firsthand” local experience of local problems, and (3) “expert” authority to solve those problems.

Methodologically, it is important to note that while specific knowledge claims about porn shops or about prostitutes made at lower levels may be “leached out” as the cases proceed to the appellate level (Latour 2002), in part through the mechanism of turning studies and other evidence into merely formal entities, nevertheless, in a general way, the content of cases—the political and social content—does matter, more than Latour might allow. The same courts that happily impute authority over adult zoning policy to city councilors would hardly defer to either their firsthand or their expert knowledge if the city water supply were at issue.

This analysis of the way in which participants, from ordinary witnesses to supreme courts, borrow, repackage, and send knowledge claims along the network, while emphasizing legal nuts and bolts more than is common in critical sociolegal work, is not out of keeping with the fundamental insight of law-and-society scholarship about the mutually constitutive relations that obtain between legal networks and extralegal networks. That popular and sociological knowledges of urban ills are freely borrowed, usually without attribution, by legal networks, and that ordinary citizens deploy highly legalistic terms to describe their “everyday” experience, is a fundamental point about knowledge processes made by many scholars before me. That ordinary people often take law's categories more seriously than judges, and that judges often resort to nonlegal reasoning, are well-known facts. But previous analyses of legal knowledges have tended to develop somewhat static typologies of knowledges. The distinction of this study lies first in showing that legal networks do not make the ontological distinction between “persons” and “texts” or things that is taken for granted by sociologists who insist on studying people rather than papers,¹² and second in emphasizing that, since the world we live in is always in flux, as philosophers from Heraclitus to Nietzsche have told us, we should try to develop analytic tools that are appropriate to understand the movement, not just the static relations, of the hybrid social-legal claims that are the raw material of “law.”