1 See Joseph Tanenhaus, Marvin Schick, Matthew Muraskin, & Daniel Rosen, “The Supreme Court's Certiorari Jurisdiction: Cue Theory,” in Glendon Schubert, ed., Judicial Decision Making (New York: Free Press, 1963) (“Tanenhaus et al., ‘The Court's Certiorari Jurisdiction’); or Gregory A. Caldeira & John R. Wright, “Organized Interests and Agenda Setting in the U.S. Supreme Court,” 82 Am. Pol. Sri. Rev. 1109 (1988).Google Scholar

2 See Ulmer, S. Sidney, Hintze, William & Kirklosky, Louise, “The Decision to Grant or Deny Certiorari: Further Consideration of Cue Theory,” 6 Law & Soc'y Rev. 637 (1972); Doris Marie Provine, Case Selection in the United States Supreme Court (Chicago: University of Chicago Press, 1980).Google Scholar

3 In this regard, Perry quotes C. Herman Pritchett's insight that “political scientists who have done so much to put the ‘political’ in ‘political jurisprudence’ need to emphasize that it is still ‘jurisprudence’” (at 3).Google Scholar

4 Perry's analysis would also be controversial from the viewpoint of traditional legal scholars because while he devotes significant attention to the legal issues of jurisdiction and procedure, he does not trace the doctrinal development of these issues.Google Scholar

5 In fact, Perry's approach has a long tradition in the study of the institutions of American politics and, in this sense, is not controversial, except for those social and political scientists who believe that this “traditional” approach should be eschewed in favor of “modern” quantitative and mathematical approaches.Google Scholar

6 I would argue, and do so argue elsewhere (see Herbert M. Kritzer, “Interpretation in Quantitative Analysis,” University of Wisconsin-Madison, 1992) (Kritzer, “Interpretation”)), that with regard to such aspects of the research process as interpretation, quantitative researchers could profitably look to the methods of qualitative research for insights into what is being done.Google Scholar

7 These specific figures are for the 1982 term, and are computed from data reported in Gregory A. Caldeira & John R., Wright, “The Discuss List: Agenda Building in the Supreme Court,” 24 Law & Soc'y Rev. 807, 824 (1990).CrossRefGoogle Scholar

8 One could restate much of Perry's argument in terms of necessary and sufficient conditions. With regard to a vote in favor of cert, there is, with one exception, no sufficient condition, only a series of alternative necessary conditions. What this means is that within a particular decision mode, the absence of any key necessary conditions constitute a sufficient condition for a vote to deny. The one sufficient condition for a vote to grant is, for some Justices, a death penalty case.Google Scholar

9 As of the October 1992 term, all but one Justice participates in the cert pool; see John W. Winkle III & Martha B. Swann, “When Justices Need Lawyers: The U.S. Supreme Court's Legal Office,” 76 Judicature 244, 247 (1993). Thus, Stevens appears to be the last holdout.Google Scholar

10 On an issue that the Court has solidified into a clear and consistent 5-4 split, the minority Justices will decide not to insist on hearing a case knowing that they cannot win. Given the 5-4 vote in Casey v. Planned Parenthood, 112 S. Ct. 2791 (1992), with Justice Thomas voting in the minority, it is interesting to speculate on whether his vote with the majority to deny certiorari a few months later in Ada v. Guam Society of Obstetricians & Gynecologists, 113 S. Ct 633 (1992) (appeal from the Ninth Circuit, 962 F.2d 1366 (1992)), was in fact a “Rule of Five” vote. Perry does not make it clear exactly how the Rule of Five functions in practice, but it would not be surprising if there was a norm for the junior Justice in the minority to change his or her vote so that the published vote was actually 6-3.Google Scholar

11 See Tanenhaus et al., “The Court's Certiorari Jurisdiction” (cited in note 1).Google Scholar

12 Perry specifically indicates that someone suggested to him the work of A. Michael Spence, Market Signaling: Informational Transfer in Hiring and Related Screening Processes (Cambridge, Mass.: Harvard University Press, 1974) (“Spence, Market Signaling”). This led him to Robert Jervis's The Logic of Images in International Relations (Princeton, N.J.: Princeton Uni-Google Scholar

versity Press, 1970) (“Jervis, Logic of Images”), which first introduced the distinction between indices and signals.Google Scholar

13 This does not mean that strategic maneuvering that looks something like bargaining never occurs. Documents in the papers of the late Justice Thurgood Marshall suggest such maneuvering in the case that became known as Bowers v. Hardwick, 478 U.S. 186 (1986); one description (N.Y. Times, 25 May 1993) labeled what happened as “poker-playing strategies.”Google Scholar

14 418 U.S. 683 (1974).Google Scholar

15 347 U.S. 483 (1954).Google Scholar

16 410 U.S. 113 (1973).Google Scholar

17 367 U.S. 643 (1961).Google Scholar

18 372 U.S. 335 (1963).Google Scholar

19 Of course, this is a simplification. I could, and have, argued that the process of selecting and coding data for statistical analysis represents an important aspect of interprets-Google Scholar

tion because what you select and how you code constrain the types of interpretations that one might ultimately arrive at. However, the nature of this process is such that it is often delegated to someone divorced from the later analysis in a way that the interpretive aspects are at least hidden from the researcher (this is clearest in the case of secondary analysis of data obtained from an archive like ICPSR).Google Scholar

20 One clear example from my own research involved an interview project in Toronto. The question motivating the research was whether rules governing litigation costs and the availability of review of legal bills by an officer of the court modified the lawyer-client relationship. It became evident after a half-dozen interviews with corporate lawyers and their clients that cost rules and review of legal bills were absolutely irrelevant to lawyer-client relationships at the corporate level. See Herbert M. Kritzer, “The Dimensions of Lawyer-Client Relations: Notes toward a Theory and a Field Study,” 1984 A.B.F. Res. ). 409.Google Scholar

21 Jerome Kirk & Marc L. Miller, Reliability and Validity in Qualitative Research 29-30 (Beverly Hills, Cal.: Sage Publications, 1986) (“Kirk & Miller, Reliability and Validity”).CrossRefGoogle Scholar

22 This attribution is from id. Google Scholar

23 Norwood Russell Hanson attributes this translation (and an alternative “abduction”) to the philosopher Charles S. Peirce; see Hanson, Patterns of Discovery: An Inquiry into the Conceptual Foundations of Science 85 (Cambridge: Cambridge University Press, 1965). In the third edition of Research Methods in Social Relations (New York: Holt, Rinehart & Winston, 1976), Claire Selltiz, Lawrence S. Wrightsman, and Stuart W. Cook devote a brief section to rettoduction (at 32-35), describing it as “reasonling] from conclusions to reasons for conclusions.” Interestingly, this specific edition of this popular methods book was the only “how-to” book I found that makes specific reference to retroduction. Neithet the edition preceding or the one following the third edition, nor any of the other dozen of so texts I checked, included a discussion of this concept.Google Scholar

24 See David R. Maines, William Shaffir, & Allan Turowett, “Leaving the Field in Ethnographic Research: Reflections on the Entrance-Exit Hypothesis,” in William B. Shaffir, Robert A. Stebbins, & Allan Turowetz, eds., Fieldwork Experience: Qualitative Approaches to Social Research (New York: St. Martins's, 1980); Steven]. Taylor, “Leaving the Field: Research Relationships and Responsibilities,” m William B. Shaffir & Robert A. Stebbins, eds., Experiencing FieldtwrJc: An Inside View of Qualitative Research (Newbury Park, Cal.: Sage Publications, 1991).Google Scholar

25 Note that this distinction differs from the more common “Before Data” and “After Data” mentioned by Kirk & Miller, Reliability and Validity 61-62.Google Scholar

26 One could make the more general argument that the data collection/data analysis distinction is an artifact of one particular research genre. The more general concepts should be something akin to data acquisition and data review.Google Scholar

27 See Tanenhaus et al., “The Court's Certiorari Jurisdiction” (cited in note 1); Ulmer et al., 6 Law & Soc'y Rev. (cited in note 1).Google Scholar

28 Here I am particularly referring to the key question of conflict in circuits. The original research on cue theory omitted this cue (in large part because of the often subjective nature of judging whether a conflict exists). Some of the most recent research in the cue theory tradition does include this element; See Caldeira & Wright, 82 Am. Pol. Sri. Rev. (cited in note 1), and 24 Law & Soc'y Rev. (cited in note 7).Google Scholar

29 As Perry explained it, the theories about decision making by members of Congress make sense only in the context that we have fairly detailed knowledge of “how a bill becomes a law.” Theorizing about Supreme Court decision making about cert would make sense only when we had comparable information about how a petition becomes a case.Google Scholar

30 Compare this to the probably mythical graduate student who had traveled to a remote country to do his dissertation field research on the local party apparatus, only to wake up the morning after arriving to see the street filled with the soldiers who had, overnight, carried out a coup; a cable to his dissertation advisor pleading for advice on what to do now that the dissertation plan was in shambles, produced a two-word reply: “Take notes.”Google Scholar

31 One of the major texts shows these as aspects of the analysis of qualitative data; see Matthew B. Miles & A. Michael Huberman, Qualitative Data Analysis: A Sourcebook of New Methods 67-68, 216, & passim (Beverly Hills, Cal.: Sage Publications, 1984) (“Miles & Huberman, Qualitative Data Analysis”). My position is that they represent the core aspects of analysis; this is essentially consistent with other works on qualitative analysis see in particular Anselm Strauss, Qualitative Analysis for Social Scientists (New York: Cambridge University Press, 1987) (Strauss uses the term “pattern uncovering” to refer to a second level of pattern identification, with the first level referred to as “axial coding”); Robert K. Yin, Case Study Research: Design and Methods 109-15 (rev. ed. Newbury Park, Cal.: Sage Publications, 1989) (“Yin, Case Study Research”); and James P. Spradley, Participant Observation (New York: Holt, Rinehart & Winston, 1979).Google Scholar

32 Traditional tropological analysis is a form of interpretation that relies on the identification and explication of tropes found in a text: “Tropes are deviations from literal, conventional, or ‘proper’ language use, swerves in locution sanctioned neither by custom nor logic”; see Hayden White, Tropics of Discourse: Essays in Cultural Criticism 2 (Baltimore: Johns Hopkins University Press, 1978) (“White, Tropics of Discourse”). In the most general sense tropes are those forms or elements of communication that have come to be accepted norms of expression and explication; they are the kinds of usages that we have come to expect and that we are able to readily interpret.Google Scholar

33 Kenneth Burke refers to these as the “four master tropes”; see A Grammar of Motive 503-17 (New York: Prentice-Hall, 1945). The identification of these four tropes as central is by no means a 20th-century phenomenon. White, Tropics of Discourse 204, points out that the 18th-century thinker Giambattista Vico argued in his treatise Principi di una Scienza Nuova [The New Science] (trans. 1968) that all figures of speech could be reduced to the four tropes listed above (§§404-9).Google Scholar

34 Dorinne K. Kondo, Crafting Selves: Power, Gender, and Discourses of Identity in a Japanese Workplace 7-9 (Chicago: University of Chicago Press, 1990) (“Kondo, Crafting Selves”).Google Scholar

35 Donald N. McCloskey, The Rhetoric of Economics 154-59 (Madison: University of Wisconsin Press, 19).Google Scholar

36 Of course, there are differences in the way that metaphors (and other tropes) are used in literary and scientific analysis; see W. H. Leatherdale, The Role of Analogy, Model and Metaphor in Science 207 (Amsterdam: North-Holland Publishing Co., 1974) (“Leatherdale, Role of Analogy, Model and Metaphor”):Google Scholar

In both literature and science metaphor is a means to an unfolding, a revelation, a new perspective or focus, but while in literature metaphor seeks to shake off the routine and familiar and reflect and refract in new lights so as to vivify and colour a world grown grey and featureless from too much familiarity, in science metaphor seeks to identify, to make the strange and unknown familiar, the complex and anarchic simple and ordered, and by contrast with literature and art, to disencumber our experience from the sensual and the concrete and the particular circumstances of time and place.Google Scholar

The commonality of usage is nicely captured in George Lakoff's observation that metaphors “are general mappings across conceptual domains.” Lakoff, “The Contemporary Theory of Metaphor,” in Andrew Ortony, ed., Metaphor and Thought 203 (2d ed. New York: Cambridge University Press, 1993).Google Scholar

37 Richard Harvey Brown, A Poetic for Sociology: Toward a Logic of Discovery for the Human Sciences 77 (Chicago: University of Chicago Press, 1977).Google Scholar

38 Some philosophers of science have made an argument that scientific explanation is itself fundamentally metaphorical; see Max Black, Models and Metaphors (Ithaca, N.Y.: Cornell University Press, 1962); Mary Hesse, Models and Analogies in Science 157-77 (Notre Dame, Ind.: University of Notre Dame Press, 1966); Scott Greer, The Logic of Social Inquiry 138-49 (Chicago: Aldine Publishing Co., 1969); Leatherdale, Role of Analogy, Model and Metaphor. Google Scholar

39 Analysts use tropes such as metaphors both for arriving at an understanding and in communicating that understanding to others. The latter process presumes that the audience shares the communicators interpretation of the trope. For example, the now common metaphor of “the shadow of the law” used by sociolegal scholars could easily be misunderstood by others to reflect not the casting of a shadow but as something meant to convey a sense of the sinister (i.e., the darkness and sinistemess associated with a shadow). Someone not literate or experienced in the terminology of the metaphor, whether mathematical or textual, might not grasp the patterns being identified.Google Scholar

40 See, e.g., John Allen Paulos, Innumeracy: Mathematical Illiteracy and Its Consequences (New York: Hill & Wang, 1988).Google Scholar

41 James P. Spradley, The Ethnographic Interview 86-88 (New York: Holt, Rinehart & Winston, 1979).Google Scholar

42 One way of describing this was that after an interview provided “revelatory” information, Perry used later interviews for confirmatory purposes.Google Scholar

43 Kirk & Miller, Reliability and Validity 24 (cited in note 21).Google Scholar

44 See Miles & Huberman, Qualitative Data Analysis 234-35 (cited in note 31). This notion is essentially a metaphor drawn from the process of location finding that involves drawing a triangle by spotting the point in question from two other points.Google Scholar

45 One might wonder why Perry did not use some type of computer-based retrieval software. In describing how he used a procedure like Perry's, Elliot Liebow, Tell Them Who I Am: The Lives of Homeless Women (New York: Free Press, 1993) (“Liebow, Tell Them Who I Am”), notes that “an anonymous reviewer describe[d] this ‘fussing’ with my notes as ‘dinosaurlike’ in view of all the computer software available for dealing with qualitative data.” Liebow goes on to observe: “My problem, however, was not retrieval or even organization of my notes but rather seeing how they related to one another and integrating them, and I did not know how to make the software do this for me.” In fact, I am not sure that computer software could significantly modify this task, albeit one might use a “card-file” program and do the sorting on screen.Google Scholar

46 More generally one might refer to much of this as the internal “lingo” of the Court, but much of that lingo was heavily imbued with metaphoric images.Google Scholar

47 Actually, there is another vote option in some cases, “Call for a Response”; i.e., ask the respondent who did not reply to the petition to file a reply (at 168).Google Scholar

48 In addition to casting a “join three” vote, Justices who vote to “grant” but do not garner four votes can threaten to dissent publicly from the denial.Google Scholar

49 Barney G. Glaser & Anselm L. Strauss, The Discovery of Grounded Theory: Strategies for Qualitative Research (Chicago: Aldine, 1967); see also Anselm Strauss & Juliet Corbin, Basics of Qualitative Research: Grounded Theory Procedures and Techniques (Newbury Park, Cal.: Sage Publications, 1990).Google Scholar

50 See Spence, Market Signaling (cited in note 12).Google Scholar

51 See Jervis, Logic of Images (cited in note 12).Google Scholar

52 As Perry acknowledges (at 122), he was introduced to the indices/signals distinction by Evelyn Fink, who also referred him to the work of Michael Spence.Google Scholar

53 At one point in his research, Perry thought he might be able to obtain copies of discuss lists, but that did not work out. If he had obtained that information, he might well have pursued the discuss list issue more systematically. The research that has been done regarding cases on the discuss list has relied on individual Justices' docket books; see, e.g., Caldeira & Wright, 24 Law & Soc'y Rev. (cited in note 7); Ulmer et al., 6 Law & Soc'y Rev. (cited in note 2); Provine, Case Selection (cited in note 2).Google Scholar

54 The role of writing as part of the process of analysis is well known (see, e.g., Sharon Friedman & Stephen Steinberg, Writing and Thinking in the Social Sciences (Englewood Cliffs, N.J.: Prentice-Hall, 1989) (“Friedman & Steinberg, Writing and Thinking in the Social Sciences”). Undoubtedly, the centrality of writing in analysis varies from researcher to researcher, with some fully working out what they want to say and how they are going to say it before beginning to write and some others developing their analysis pretty much as they write; most likely, the majority of researchers fall somewhere in between, starting with a framework and line of argument but only coming to understand fully the strengths, weaknesses, and gaps as the writing progresses.Google Scholar

55 United States v. Nixon, 418 U.S. 683 (1974).Google Scholar

56 Dames & Moore v. Regan, 453 U.S. 654 (1981).Google Scholar

57 112 S. Ct. 2791 (1992).Google Scholar

58 Buckley v. Valeo, 424 U.S. 1 (1976).CrossRefGoogle Scholar

59 Oregon v. Mitchell, 400 U.S. 112 (1970).Google Scholar

60 My own classification of cases deciding during the October 1992 term, based on the case summaries published in the National Law Journal (23 Aug. 1993) shows that 51 cases were probably based on jurisprudential concerns and 39 were possible based on outcome concerns (albeit 17 of those were decided unanimously); I have omitted from my count those cases that were dismissed as moot or as improvidently granted (“DIGged”). For more discussion, see table 1 and the accompanying text.Google Scholar

61 Even for cases which are likely to involve a great deal of interest about outcome, jurisprudential concerns can play an important role; see Lee Epstein & Joseph F. Kobylka, The Supreme Court and Legal Change: Abortion and the Death Penalty 302 (Chapel Hill: University of North Carolina Press, 1992) (“the law and the legal arguments grounded in law matter, and they matter dearly”).Google Scholar

62 I do not want to make too much of the unanimous cases because the 1992 term had a notably higher proportion of unanimous cases than did terms just prior to 1992.Google Scholar

63 I classified these decisions as the National Law Journal reported the decisions in the late spring and early summer of 1993. Late in the summer, NLJ published its annual “Supreme Court Review,” which contained summaries for all cases decided during the October 1992 term. Using that publication, I coded all cases decided (including those I had previously coded). The intracoder reliability of those I coded twice was 88%. Also of interest is that the proportion coded as reflecting outcome preferences was higher for the end-of-term cases for the term as a whole: 22 nonunanimous outcome, 17 unanimous outcome, and 51 jurisprudential (omitting those cases dismissed as improvidently granted or moot). For a brief discussion of the decisions of the 1990 term, see Herman Schwartz, “The New East European Constitutional Courts,” 13 Mich. J. Int'l L. 748 n.30 (1992).Google Scholar

64 For an extensive discussion of the role of side information in quantitative analysis, including examples drawn from the political science literature, see my essay, “Interpretation in Quantitative Analysis” (cited in note 6).Google Scholar

65 See Miles & Huberman, Qualitative Data Analysis 221-22 (cited in note 31).Google Scholar

66 To take a totally different sociolegal context, it was the decoding of internal tropes used by criminal defendants (e.g., “street lawyers”) that led Jonathan Casper to many of his important insights about how criminal defendants perceive the American justice system; see his American Criminal Justice: The Defendant's Perspective (Englewood Cliffs, N.J.: Prentice-Hall, 1972).Google Scholar

67 See Kritzer, “Interpretation.”Google Scholar

68 See Kondo, Crafting Selves (cited in note 34). For an example of this trope in use, see Liebow, Tell Them Who I Am 1-16 (cited in note 45).Google Scholar

69 Recent writing on comparative research has argued that “studies based on a small number of cases have difficulty in evaluating probabilistic theories”; Lieberson, Stanley, “Small N's and Big Conclusions: An Examination of the Reasoning in Comparative Studies Based on a Small Number of Cases,” 70 Soc. Forces. 307 (1991). Even if this is true, one must not presume that qualitative research cannot deal with probabilistic theories.CrossRefGoogle Scholar

70 See Caldeira & Wright, 82 Am. Pol. Sci. Rev. (cited in note 1), and id., 24 Law & Soc'y Rev. (cited in note 7).Google Scholar

71 There are some social science genres for which this may not be as true. I am thinking in particular of reports from experimental research, particularly laboratory-based experiments in fields such as psychology. See Charles Bazerman, “Codifying the Social Scientific Style: The APA Publication Manual as a Behaviorist Rhetoric,” in John S. Nelson, Allen Megill, & Donald N. McCloskey, eds., The Rhetoric of the Human Sciences (Madison: University of Wisconsin Press, 1987).Google Scholar

72 The one clear exception to this is discipline of anthropology, which in recent years has paid increasing attention to the role of writing; see John Van Maanen, Tales of the Field: On Writing Ethnography (Chicago: University of Chicago Press, 1988); James Clifford & George E. Marcus, eds., Writing Culture: The Poetics and Politics of Ethnography (Berkeley: University of California Press, 1986).Google Scholar

73 Kirk & Miller, Reliability and Validity 21 (cited in note 21). The term “validity” is used somewhat more narrowly in the context of measurement theory to refer to whether one is actually measuring what one thinks one is measuring.Google Scholar

74 See, e.g., Donald T. Campbell & Julian C. Stanley, Experimental and Quasi-experimental Designs for Research (Chicago: Rand McNally, 1963); or Thomas D. Cook & Donald T. Campbell, Quasi-Experimentation: Design and Analysis Issues for Field Settings (Chicago: Rand McNally, 1979). Standard discussions draw such distinctions as that between internal validity (i.e., Does the research design include appropriate controls to allow the researcher to reach appropriate conclusions?) and external validity (Can the results be generalized beyond the specifically observed situation?).Google Scholar

75 Some writers have tried to distinguish these issues in qualitative research by use of the terms credibility, transferability, dependability, and confirmability in place of internal validity, external validity, reliability, and objectivity; see Yvonne S. Lincoln & Egon G. Guba, Naturalistic Inquiry 301-27 (Beverly Hills, Cal.: Sage Publications, 1985). Catherine Marshall & Gretchen B. Rossman, Designing Qualitative Research 144-49 (Newbury Park, Cal.: Sage Publications, 1989), refer to these under the rubric of “criteria of soundness.” I choose to use the traditional terms because my ultimate goal is to argue for the similarity between quantitative and qualitative research rather than the distinctiveness; quantitative researchers would benefit from thinking about validity, reliability, and objectivity under alternate labels.Google Scholar

76 Kirk 6k Miller, Reliability and Validity 29-30.Google Scholar

77 See, e.g., Hubert M. Blalock, An Introduction to Social Research 45-46 (Englewood Cliffs, N.J.: Prentice-Hall, 1970); Louise H. Kidder, Selltiz, Wrightsman and Cook's Research Methods in Social Relations 105 (4th ed. New York: Holt, Rinehart & Winston, 1980); Robert A. Bernstein & James A. Dyer, An Introduction to Political Science Methods 67 (Englewood Cliffs, N.J.: Prentice-Hall, 1979); Friedman & Steinberg, Writing and Thinking in the Social Sciences 79 (cited in note 54). The strongest critiques of qualitative research, by quantitative researchers, tend to come not in published writings but in late-night conversations in hotel bars at professional conventions (as do the strongest critiques of quantitative research by qualitative researchers); often these critiques argue that qualitative research is more akin to journalism than it is like quantitative social science.Google Scholar

78 There are many examples of published ethnographies serving as the basis of further analysis (or reanalysis); an archive of published ethnographic material, the Human Relations Area Files, was designed specifically for this purpose. The only example that I have been able to locate of raw textual data collected by researchers being reanalyzed by another researcher is one chapter in Jack Katz's Seductions of Crime: Moral and Sensual Attractions in Doing Evil 164-94 (New York: Basic Books, 1988). Interestingly, the collectors of the data, which consists of narrative accounts of armed robberies constructed from police records, coded those narratives for quantitative analysis; see Zimring, Franklin E. & Zuehl, James, “Victim Injury and Death in Urban Robbery: A Chicago Study,” 15 J. Legal Stud. 1 (1986).CrossRefGoogle Scholar

79 See Yin, Case Study Research 43-44 (cited in note 31).Google Scholar

80 For an interesting effort to do precisely this, see Gary King, Robert O. Keohane, & Sidney Verba, Designing Social Inquiry: Scientific Inference in Qualitative Research (Princeton, N.J.: Princeton University Press, forthcoming 1994).Google Scholar

81 My emphasis here has been largely on the “process” of social science; another line of argument about fundamental similarities could be made for the “practice” of science more generally. Recent work on the sociology of science has emphasized the socially constructed nature of scientific theory and scientific fact; for a very brief summary of this view, see Jasanoff, Sheila, “What Judges Should Know about the Sociology of Science,” 77 Judicature. 77 (1993).Google Scholar

82 This is similar to Yin's idea of “analytic generalization”; see his Case Study Research. Google Scholar

83 For a discussion of the process by which something becomes accepted as a tropological constructs, see Stephan Lewandowsky & Ian Spence, “The Perception of Statistical Graphs,” 18 Soc. Methods & Res. 200 (1989-90).CrossRefGoogle Scholar