Abridged from Ragin Charles and Becker Howard (eds), "What is A Case?, Exploring the Foundation of Social Inquiry, Cambridge University Press, pp. 1-18.

Social science methodology is anchored by a number of basic precepts that are rarely questioned by practitioners. One precept that is central to the logic of analysis is the idea of having cases. Social scientists use terms like "N of cases," "case study," and "sample of cases" with relatively little consideration of the possible theories and meta-theories embedded in these terms or in the methods that use cases and make conventional forms of analysis possible. For example: A study that uses interviews of employees to construct a picture of the informal organization of a firm looks superficially like one that uses interviews of employees to address variation in job satisfaction. Both studies use interviews of employees as the primary data source, but the first is about the firm as a whole, while the second is about employees' subjective states. It is argued here that the term "case" and the various terms linked to the idea of case analysis are not well defined in social science, despite their widespread usage and their centrality to social scientific discourse. To the question "What is a case?" most social scientists would have to give multiple answers. A case may be theoretical or empirical or both; it may be a relatively bounded object or a process; and it may be generic and universal or specific in some way. Asking "What is a case?" questions many different aspects of empirical social science.

The conversation about the term "case" presented in this volume had its origins in other conversations. The issue of cases came up often in a logic-of-analysis workshop that Howard Becker and I conducted in the winter and spring of 1988. This particular workshop had its origins in still other exchanges, based on our shared reaction to the unexamined status of the case in social science methodology.

This peculiar status of the "case" was clear to me in my work The Comparative Method (Ragin 1987). In that work I showed how conventional variable-oriented comparative work (e.g., quantitative cross-national research), as compared with case-oriented comparative work, disembodies and obscures cases. In most variable-oriented work, investigators begin by defining the problem in a way that allows examination of many cases (conceived as empirical units or observations); then they specify the relevant variables, matched to theoretical concepts; and finally they collect information on these variables, usually one variable at a time - not one case at a time. From that point on, the language of variables and the relations among them dominate the research process. The resulting understanding of these relations is shaped by examining patterns of covariation in the data set, observed and averaged across many cases, not by studying how different features or causes fit together in individual cases.

The alternative, case-oriented approach places cases, not variables, center stage. But what is a case? Comparative social science has a ready-made, conventionalized answer to this question: Boundaries around places and time periods define cases (e.g., Italy after World War II). In comparative and historical social science, there is a long tradition of studying individual countries or sets of theoretically or empirically related countries conceived as comparable cases. The conventionalized nature of the answer in macrosocial inquiry made it simple to contrast variable-oriented and case-oriented approaches. It could just as easily be argued, however, that not countries but rather parallel and contrasting event sequences are cases, or that generic macrosocial processes, or historical outcomes, or macro-level narratives are cases. "What is a case?" is problematic even where researchers are confronted at every turn by big, enduring, formally constituted macrosocial units such as countries.

The problem of "What is a case?" is even more crucial when the contrast between variable-oriented and case-oriented approaches is transferred to other research domains, because in most research areas the answers are less conventionalized. Is a social class a case or a variable?. This is not a trivial question for scholars interested in social movements and the future of inequality. Is an analysis of United States census data a study of many cases (individuals) or one case (the United States)? As I pushed my ideas about case-oriented research into new substantive areas, I found that I had a lot of new questions about cases.

There are different approaches to answer the question "what is a case"?. To understand them, consider two key dichotomies in how cases are conceived:

(2) whether these, in turn, are understood as general or specific.

The first dichotomy (whether the question of cases involves empirical units or theoretical categories) is common in discussions of social science methodology and overlaps with the philosophical distinction between realism and nominalism. Realists believe that there are cases (more or less empirically verifiable as such) "out there." Nominalists think cases are theoretical constructs that exist primarily to serve the interests of investigators. A realist sees cases as either given or empirically discoverable. A nominalist sees cases as the consequences of theories or of conventions.

The second dichotomy concerns the generality of case categories. Are case designations specific (e.g the "authoritarian personality" or the "anti-neocolonial revolution") and developed in the course of research (e.g., through in-depth interviews or historical research) or are they general (e.g., individuals, families, cities, firms) and relatively external to the conduct of research? In many areas of research, generic units are conventionally treated as cases, and case categories are neither found nor derived in the course of research. They exist prior to research and

are collectively recognized as valid units by at least a subset of social scientists. Specific case categories, by contest, emerge or are delineated in the course of the research itself. What the research subject is a "case of" may not be known until after most of the empirical part of the project is completed. To a limited extent, this second dichotomy overlaps with the qualitative-quantitative divide in social science. The cases of quantitative research tend to exist as conventionalized, generic categories independent of any particular research effort. The cases of qualitative research tend to coalesce as specific categories in the course of the research. "What is this - the research subject - a case of?" is a question that is best asked in qualitative social science.

The cross-tabulation of these two dichotomies (Table 1) yields four possible starting points for answering the question "What is a case?" Consider the nature of "cases" from the perspective of each cell of the cross-tabulation:

Cell 1: Cases are found. In the first quadrant, researchers see cases as empirically real and bounded, but specific. They must be identified and established as cases in the course of the research process. A researcher may believe that "world systems" (networks of interacting and interdependent human societies) are fundamentally important empirical units for understanding the history of human social organization and therefore may seek to determine the empirical boundaries of various historical world systems (verifiable, e.g., through evidence of trade in bulk goods between peoples of differing cultures). Researchers who approach cases in this way see assessment of the empirical boundingof cases as an integralpart of the research process.

Cell 2: Cases are objects. In the second quadrant, researchers also view cases as empirically real and bounded, but feel no need to verify their existence or establish their empirical boundaries in the course of the research process, because cases are general and conventionalized. These researchers usually base their case designations on existing definitions present in research literatures. A researcher interested in explaining contemporary international inequality, for example, would accept nation-states (as conventionally defined) as appropriate cases for his or her analysis. Often coupled with this view is an instrumental attitude toward cases - they exist to be manipulated by investigators.

Cell 3: Cases are made. Researchers in this quadrant see cases as specific theoretical constructs which coalesce in the course of the research. Neither empirical nor given, they are gradually imposed on empirical evidence as they take shape in the course of the research. A cell-3 investigator interested in tyranny, for example, would study man possible instances of tyranny. This investigation might lead to an identification of an important subset of instances with many common characteristics, which might be conceived, in turn, as cases of the same thing (e.g., as cases of "patrimonial praetorianism" or as cases of "modern tyranny"). Interaction between ideas and evidence results in a progressive refinement of the case conceived as a theoretical construct. At the start of the research, it may not be at all clear that a case can or will be discerned. Constructing cases does not entail determining their empirical limits, as in cell 1, but rather pinpointing and then demonstrating their theoretical significance.

Cell 4: Cases are conventions. Finally, in the fourth quadrant, researchers see cases as general theoretical constructs, but nevertheless view these constructions as the products of collective scholarly work and interaction and therefore as external to any particular research effort. A researcher, for example, might conduct research on "industrial societies," recognizing that the assignment of empirical cases to this theoretical category is problematic-and that the theoretical category itself exists primarily because of collective scholarly interest. In this view, cases are general theoretical constructs that structure ways of seeing social life and doing social science. They are the collective products of the social scientific community and thus shape and constrain the practice of social science.

This fourfold division of case conceptions is not absolute. A researcher could both use conventionalized empirical units, accepting them as empirically valid (cell 2), and try to generate new theoretical categories or case constructs (cell 3) in the course of his or her research. Frustrations with conventional case definitions and practices (cell 4) could lead researchers to intensify their empirical efforts and to define cases and their boundaries in a more inductive manner (cell 1). In fact, most research involves multiple uses of cases, as specific or general theoretical categories and as specific or general empirical units. These multiple uses occur because research combines theoretical and empirical analysis, and the two kinds of analyses need not use parallel cases or units. The point of Table 1 is not to establish boundaries between different kinds of research, but to establish a conceptual map for linking different approaches to the question of cases.

Answer the following questions: 1. Explain the two key dichotomies that help explaining how case are conceived. 2. Explain each of the four types that are presented in table 1.