¹ Reviewing law and society research a decade ago, Richard Abel (1980) argued that concern for legal effectiveness, a concern embedded in the ideological premises of the legal system itself, so dominated theories about law and social change that research was limited to impact studies—studies that examined the law's effectiveness in producing behavior conforming with legal norms. In a related criticism, Austin Sarat and Susan Silbey (1988) have suggested that the field's fascination with policy questions focused law and society research on the gap between legal ideals and institutional practices. Studies of law and social change, they argued, ignored both the law's displacement of alternative conceptions of order and its constitutive role in maintaining the everyday and routine patterns of social life. Both criticisms have suggested that research on law and social change has often served to reinforce the ideological premises for the legal system itself.

² Although Abel uses the term “function” to describe the role of legal institutions in society, feedback mechanisms seldom exist to link the effects of law with the behavior of participants in legal institutions. Thus, “functional” is inappropriate as a formal theoretical concept describing law's causal role. Nor do I believe that Abel uses the term in its formal sense. As I discuss further in note 7, the purposive and instrumental quality of legal behavior is captured better by another theoretical concept, “intentional” causation.

³ I disagree with Friedman's characterization of most longitudinal litigation research as court-centered, because the goal of most researchers is in fact to explain the origin and processing of disputes, and the court as a separate entity is treated as mere recordkeeper, playing little or no role in the explanation of disputing or litigation.

⁴ This perhaps explains why there has been surprisingly little interest in formal litigation outcomes in studies of the relationship between community change and trial courts (Friedman, 1989b), since community and courts are intertwined in so many other significant ways. In a different theoretical frame of reference outcomes have been of great interest to those attempting to understand the internal working of courts as organizations because formal outcomes bear a much more direct relationship to the formal and informal processes within trial courts than they do to the external environment of courts (Padgett, 1990; Seron, 1990).

⁵ Critics often point to the gap between the concepts employed and the measures constructed from statistical data or litigation rates to represent them, the often inappropriate use of statistical techniques, and the absence of detailed data about the decisions of litigants and courts. But most critics have argued that the problems of the field lie deeper, in weaknesses of conceptualization and theory. For example, critics have often begun by examining the units of analysis that docket studies have often taken for granted, “case,” “court,” and “dispute” (see Engel, 1980, 1990; Kidder, 1980-81; Yngvesson, 1988), arguing that the meaning of each unit of analysis depends on the perspective chosen from which to view it. Thus, a “case” means something different to a potential litigant and a court clerk; the “court” has different boundaries for the litigant, the court clerk, and the observer employing a theory of complex organizations.

⁶ A theory formulates questions for research. The mere fact that research is theory-driven has no bearing on the choice of methods for research, for example, whether they are quantitative or qualitative, observational or archivally based, or whether still other techniques or some combination of those just mentioned are selected. But the adoption of a particular theory may have important implications for the particular methods that one might use to answer questions, since some ways of gathering or studying information seem better suited to answering certain kinds of questions, and because some theories imply a particular relationship between this researcher and the subject of research, as, for example, in the case of Bourdieu's claim (1977: 8–9, 171) that a mere observer cannot understand the whole process of contract formation.

⁷ In some obvious ways the “gap” or “legalist” paradigm of law and social change still poses important empirical questions that must not be overlooked as we move toward a broader, more inclusive understanding of law and change. First, the primary question posed by the legalist perspective, namely, what are the symbolic and instrumental effects (Lempert, 1966) of decisions and other trial court actions, is not uninteresting or unimportant. Second, even if the answer is that these effects are small or irrelevant to the role played by law and by trial courts, or conversely, marginal to the frame of reference of some participants in conflict or dispute resolution outside of courts, the theory that they do have an effect strongly influences the behavior of many of the actors in settings studied by longitudinal trial court research. Indeed, the paradigm of “intentional” causation, in which an actor is influenced by the anticipated effects (Elster, 1985) is often a more accurate description of the relationship between law and social action than a “functional” causation paradigm in which an actor is said to influenced by the effects of action through a further feedback effect. Thus, what has been called a “gap research” paradigm may be pursued with other objectives in mind, for example, with the objective of examining beliefs about the legal system that continue to make the goal of effectiveness central to the maintenance of the behavior of trial courts and other legal institutions.

⁸ I am indebted to Mayer Zald for this observation on the importance of historical models of change and the frequent misapplication of ahistorical causal models in longitudinal research. The comments were made at the Conference on Longitudinal Research on Trial Courts at SUNY Buffalo, August 1987.

⁹ This report by the Committee on National Statistics of the National Academy of Sciences addresses the need for and technical aspects of data sharing as well as the benefits for a discipline and its policy applications. I am also concerned, as this report is not, about the problems and limits of attempting to specify theoretically substantive goals for a data collection as a matter of public policy.

¹⁰ These may usefully be thought of as alternative models of collaboration. See below.

¹¹ What data should be included in a particular database for longitudinal study of trial courts seems a less difficult problem than the more important issue of what data are required to answer particular questions (see below). For retrospective research the options for data collection will be limited in most cases by the nature of the surviving court records and the likelihood that other types of information will be unavailable. These limitations make specification of “core” data easier by default but will make theoretically driven research more difficult because of the lack of information about theoretically significant units of analysis or processes. A retrospective study of trial courts employing docket data should attempt to include the basic information that exists in most official records together with usually available supplementary information that will render the docket data useful to a broad range of perspectives and theories. For example, the investigator should gather information about the number and types of cases, types of litigants, presence of lawyers, intermediate processing (such as jury trial), and formal disposition of cases. The investigator should also include basic data about the court structure and personnel, e.g., numbers and backgrounds of judges if available, number and types of other court personnel, and details of the court's jurisdiction, organization, and control by other governmental bodies. Many existing docket studies have not systematically gathered these data, even though the data were available. Gathering such data does not guarantee comparability or usefulness for all research purposes, of course, since there are no standardized categories or measures of court activity. These data, however, recorded in the terms and categories created by the courts, are directly relevant for research on the behavior of the courts themselves and so provide one important starting point for other types of data collection about disputes, cases or litigants.

¹² As one participant remarked at the SUNY-Buffalo conference, “The question of what question to ask is the most fundamental, almost political question.” This comment was made still more meaningful by the one which preceded it, in which the same participant observed that funds for research are limited.

¹³ With respect to the political functions of courts, at the conference Lawrence Friedman observed that North Americans think of courts as part of their “government” while many other cultures, notably European, think in terms of a “state.” “Government” and “state” have quite distinct connotations. Thinking of courts as part of government suggests that they are like administrative offices of government and exercise what is fundamentally perceived as nonpolitical authority. The concept of the state is more suggestive of power relations and the political uses of power, and societies with different political cultures are quite comfortable with the view that courts are a part of such a state. This difference in political cultures has, quite clearly, also influenced the approaches to courts taken by North American social science.