Forthcoming as a “Core Essay” in the International Encyclopedia of Political Science

Sanford Schram teaches social theory and social policy in the Graduate School of Social Work and Social Research at Bryn Mawr College. His research has appeared in the American Political Science Review, American Sociological Review, Perspectives on
Politics, Social Service Review, Theory & Society, and other journals. He is author of four books and co-editor for four additional volumes, including most recently Making Political Science Matter: Debating Knowledge, Research, and Method (New York University Press, 2006).
 Political Science research has remained a contested terrain from its first beginnings with the formation of the discipline in the late 19th Century, the subsequent development as Political Science as a distinct profession, and right down to this very day. As a result, the quest for a unitary paradigm for structuring research about politics has remained unfulfilled.① But the battle lines are being redrawn as we speak.
 In recent decades, Political Science research has been shadowed by a debate between positivists who champion emulating the natural sciences and interpretivists who side with approaching the study of politics along the lines of more humanistic forms of inquiry. Competing positivist and interpretivist epistemologies have spawned distinctive methodologies with separate logics of inquiry, varying preferences for different methods of data collection, and debates about a number of other issues including, most commonly, the value of quantitative versus qualitative data. Most recently, debates between positivists and interpretivists have been complicated by interventions by others who do not situate their investigations in either camp. This group has included a growing number of scholars who refuse to accept that they must limit their research to either a positivist or interpretivist methodology. Mixed-methods researchers have been joined by others who stress the importance of problem-driven over theory-driven research. These researchers want to focus on problems in the real world of politics and then use whatever many different methods of study and forms of data collection necessary to study those topics as best they can. The debates about Political Science research ultimately raise issues about the relationship of Political Science to politics. This essay reviews the major points of contention about political science research today with an eye toward addressing the issue of the discipline's relationship to politics.
Researching the Discipline/Disciplining Research
　　　Beginning with its origins in the late 19th Century, the American incarnation of Political Science as an academic discipline has always been a work in progress, dedicated to the idea of progress. That chiasmus points to the hope the discipline’s early leaders had that Political Science could develop as the scientific study of politics that would redound to the improvement of American democracy and vise versa.② The push for professionalization that came with the initiation of the American Political Science Association (APSA) in the first years of the 20th Century represented an attempt to transform the discipline to more explicitly and successfully address the relationship of science to democracy. John Gunnell has written:
 Between the institution of the APSA and the appearance of the first issue of the American Political Science Review, Max Weber published his 1904 essay on “The ‘Objectivity’ of Knowledge in Social Science and Social Policy” in the newly created Archiv für Sozialwissenschaft und Sozialpolitik, which he shared in editing. There is little to suggest any direct relationship between Weber’s essay and the ideas of American political scientists, but in both theme and context there were a distinct family resemblance and mediated intellectual connections. Weber’s argument was in part a response to the failure of the Verein [für Sozialpolitik] and the ideological and methodological disputes that had characterized its history. Although he presented his essay as an intervention in controversies about the nature of social scientific explanation, he also explicitly addressed it to a wider public audience with the aim of vouchsafing the cognitive authority of academic social science. He stressed the commitment of the journal was to the scientific pursuit of the “the facts of social life,” but it was also concerned with “social policy” and “the training of judgment in respect of practical problems arising from these social circumstances.”…Weber…emphasized various ways in which social scientific knowledge could, in principle, constrain and direct policy decisions as well as the extent to which scientific investigation necessarily proceeded from the perspective of value-laden premises. The authority of social science nevertheless depended, he argued, on acceptance of the autonomy of empirical claims and on the professional status and independence of those who made such claims. The dilemma and solution the Weber articulated bore remarkable similarities to the situation attending the founding of the APSA.③
　　　Frank Goodnow, W. W. Willoughby, Woodrow Wilson, and others formed the APSA based on concerns consistent with Weber's thinking about the connections between science and politics. Goodnow was an especially poignant participant due to his prior involvement in both politics and academic administration, much like Wilson, but also because he helped educate leaders of the next generation who sought to develop a science of politics for the betterment of democracy, including most prominently Charles Merriam.
 With time, a number of exemplars of the attempt to fuse science and democracy to their mutual benefit appear in the histories of the profession. Prominent among them was Merriam’s colleague at Chicago Harold Lasswell who would eventually turn to attempting to forge what he called a “policy science.”④ Perhaps because of the promise of these efforts has seemed forever deferred, periodic challenges to the idea that a science of politics could serve democracy have occurred throughout the discipline’s history.⑤ Yet, the idea is resilient and continues to inspire political scientists. Like moths to a flame, they keep coming back revisiting the debate, tempted to push for either more science or democracy, and seeking to right the balance once and for all. Most recently, the challenge has come from the Perestroika movement, which has included arguments that the preoccupation with becoming a science has led the discipline away from improving politics.⑥ Rogers Smith and Ian Shapiro have led the call to move away from method-driven research toward problem-driven research that is framed around addressing fundamental political concerns in the real world of politics.⑦ For Smith and Shapiro, it is less important whether Political Science research is grounded in one or another distinct methodology underwritten by a particular philosophy of science and more important that it be premised first and foremost on addressing real problems in the world of politics.
　　　The Perestroika challenge has also included a call for methodological pluralism.⑧ The call for methodological pluralism has often been couched in terms of ending the supposed hegemony of positivistic approaches to the exclusion of interpretive research, particularly in the leading journals. John Dryzek however has encouraged a modification of the call for methodological pluralism, arguing against an “empty pluralism” that allows for a thoughtless “anything goes” attitude. Instead, Dryzek calls for a “critical pluralism” that would involve critical engagement of competing research traditions so that they can better learn from each other.⑨ Keith Topper has seconded that critique by emphasizing the need for a critical pluralism that combines research methodologies to create a more robust knowledge of politics.⑩ Topper joins Dryzek in pushing for a critical pluralism within research projects so that the interpretive moments in all analyses can inform positivistic ones and vise versa.
　　　Political Science research is now arguably increasingly post-paradigmatic in the sense of that the field is moving away from the idea that research should be according to one overarching philosophy of science, overall methodology or distinctive logic of inquiry. In fact, the changing currents associated with Political Science research indicate that perhaps the field was always best seen as non-paradigmatic both in terms of what it ideally should be in theory as well as what it actually was in practice. It is perhaps only now with Minerva’s Owl that we can look back to see the value of the diversity that was present in Political Science research throughout its history.
Against Methodological Purity and for Methodological Pluralism
 The current debate about methodological pluralism has been a long time coming, arriving at the end of an arduous journey across the methodological minefields that have served as the ground for Political Science research. The pressures for a unitary paradigm have been real and persistent temptation. There are understandable reasons for the persistent temptation to insist on methodological purity as opposed to methodological pluralism. Political Science is an academic discipline; and like all academic disciplines, it is organized to promote scholarship according to the highest standards of what counts as good research. The temptation to discipline research is therefore implicit in the very idea of organizing an academic discipline. Gatekeeping to keep out bad research and include only good research is an unavoidable corollary.
 While this sort of policing of research has always been practiced in the discipline, it became much discussed during the highwater mark of the behavioralism movement in the 1960s. At this time, explicit discussion of methodology, as in an underlying logic of inquiry, overtook debates about the value of specific methods, as in particular forms of empirical observation and data collection. The behavioral movement very much sought to deliver on the longstanding promise that Political Science could be become a scientific discipline if research in the field were conducted by emulating the natural sciences. Yet, in spite of desire to emulate the natural sciences, proponents of behavioralism (like Robert Dahl, David Easton, Heinz Eulau, David Truman and William Riker to name just a few of the most prominent behavioralists of the time) turned not to natural science but to the philosophy of science for models of how to conduct scientific research on politics.11
 The philosophy of science provided an explanation of the logical structure of scientific explanation known variously as logical positivism or logical empiricism and is now commonly referred to as just plain “positivism.” Positivism served to create a methodological foundation for Political Science as a science. Political scientists would be doing science if they structured their research projects consistent with the positivistic methodology. In other words, regardless of the specific methods of empirical observation or data collection used, research needed to be framed to support explanations consistent with the tenets of positivism. Research needed to produce the basis for testing causal theories that could explain why political phenomena, relationships and processes were the way they were.
 The emphasis on positivism eventually led to more a generalized understanding of how research was organized to contribute to the scientific discipline of Political Science. This model could be said to include the following hierarchy of assumptions:
 (1) Political Science exists to help promote understanding of the truth about politics;
 (2) Political Science research contributes to this quest by adding to the accumulation of an expanding base of objective knowledge about politics; 
 (3) the growth of this knowledge base is contingent upon the building of theory that offers explanations of politics;
 (4) the building of theory is dependent on the development of universal generalizations regarding the behavior of  political actors;
 (5) the development of a growing body of generalizations occurs by testing falsifiable, causal hypotheses that demonstrate their success in making predictions;
 (6) the accumulation of a growing body of predictions about political behavior comes from the study of variables in samples involving large numbers of cases; and
 (7) this growing body of objective, causal knowledge can be put in service of society, particularly by influencing public policy makers and the stewards of the state.
　　　Critics from the time of behavioralism and since have been quick to note that this paradigm excludes much valuable research. For instance, it assumes that the study of a single case is “unscientific,” provides no basis for generalizing, does not build theory, cannot contribute to the growth of political knowledge, and, as a result, is not even to be considered for publication in the leading journals and is to be discouraged as a legitimate doctoral dissertation project.12 While there have always been dissenters to the drift toward “large-n,” quantitative research in service of objective, decontextualized and universally generalizable truth about politics, there is a good case to be made that the dissenters have increasingly been marginalized as the center of gravity of the discipline has drifted more and more towards reflecting these core assumptions about political knowledge.13
 As fate would have it, with the continued quest for a scientific discipline increased challenges were put forward by proponents of alternative logics of inquiry. In other words, rather than move from the methodological purity of the positivist paradigm to a methodological pluralism instead the most common response until the Perestroika movement of recent years has been to push for a more “interpretive” approach to the study of politics.
 The interpretive approach is really a loose collection of many different approaches including such disparate approaches as political ethnography, constructivism, discourse analysis and many others. What these approaches share in common is an emphasis on the interpretive dimensions political analysis, stressing the importance of accounting for how political phenomena, relationships and processes are not so much pre-existing objective facts of the social world as they are subjectively experienced and interpreted phenomena.14 This distinction revisits the debates that preoccupied Weber and his colleagues over whether social science ought to be more about erklaren (explanation) or verstehen (understanding). The interpretive approaches emphasize it is more important of try to arrive at understanding how the social world is subjectively experienced and interpreted by people than it is to provide an explanation of what caused social phenomena to happen. Most interpretive approaches therefore do not look to the natural sciences for a model of how to conduct research on politics because they see an asymmetry between social sciences and the natural sciences stemming from what Anthony Giddens and others have called the double hermeneutic.15 From this perspective, the natural sciences are interpretive in that natural science research is framed through interpretive lenses for constructing the facts that are observed whether they are quarks within atoms or the black holes in the cosmology; however, the social sciences are doubly hermeneutic in that research on social phenomena involves interpreting the interpretations social actors make of their experiences. Social science research is doubly hermeneutic because it involves researchers’ interpretations of other people’s interpretations.
 The “interpretive turn,” as it came to be called in the social sciences, had many sources, including, perhaps, most prominently Clifford Geertz and his leadership in the School of Social Science at the Institute for Advanced Study in Princeton, New Jersey.16 Geertz, to be sure, saw interpretive approaches as providing important perspectives for understanding whatever was being studied. Yet, Geertz resisted the idea that researchers had to choose either an interpretive or positivist approach as a distinct logic of inquiry.17 Nonetheless, over time, the main thrust of the interpretive turn has been to insist that interpretive social science implied a distinct logic of inquiry that prevented mixing methodologies.18 Positivism and interpretivism became the oil and water of Political Science research. While researches might be able to mix different methods of data collection, they increasingly were discouraged from mixing the methodologies on the grounds that that positivism and interpretivism implied distinct logics of inquiry that could not be sensibly combined in the same analysis. Over time, the separate of logic of inquiry argument has undoubtedly contributed to the idea that Political Science is a fractured discipline where different researchers employing different approaches talk less and less to each other.19
　　　Yet, Joe Soss, among a growing number of political scientists, has challenged the idea that the separate logics of inquiry forces researchers to choose to pursue positivist or interpretive approaches to the exclusion of the other.20 Recovering the insights of Geertz, Howard Becker and other leaders of the interpretive turn of the last generation, Soss and others have encouraged mixing methodologies while conducting research projects that strive to studying specific topics as fully and thoroughly as possible.21
　　　There is now a growing interest in mixed-method research in the social sciences today, including books and whole journals dedicated to promoting the idea.22 While the idea of mixed-methods research itself can be contained within either the positivist or interpretivist paradigm, other versions cross those boundaries more freely, and still end up appearing even in the pages of the American Political Science Review.23
Toward Critical Pluralism Within Political Science Research
 Even though mixed-methods research is a growing trend in Political Science and related fields, such research comprises still the minority of projects and is even less represented in journal articles. Given the format and space constraints of most academic journals, the tendency is to present findings associated with one particular data collection effort even if it was part of a mixed-methods project.24 These constraints may need to change if mixed-methods research is to get a full hearing. The possibilities that come with electronic publishing of journals over the internet may well help in this regard.
 Were the conditions to be made right, the possibilities increase for a critical pluralism to flourish not just across Political Science research efforts but within specific political science research projects. These are projects that resist the Manichean debates between positivists and interpretivists that pose the false choice of the separate logics of inquiry. These are projects that often find their frame as primarily what Shapiro has called problem-driven research where a researcher starts with a specific problem in the world of politics and then employs multiple approaches to study it as best he or she can.25 Yet, these are not simply mixed-methods, problem-driven projects but instead are mixed-methods, problem-driven investigations that explicitly seek to gain an analytical advantage by highlighting the interpretive moments in their positivistic analyses and vise versa. This is more than triangulating findings by collecting several different types of data about a particular topic. In joining Dryzek’s call for critical pluralism in research, Topper has provided the most sustained philosophical argument on behalf of such Political Science research.26 His argument can be summarized without too much violence as highlighting the analytical power that comes with constructing a critical pluralism within specific Political Science research investigations.
 The power of such combined approaches needs to be assessed by researchers on both sides of the positivistic and interpretive divide. Should the claims made for a critical pluralism within political science research hold up when evaluated by political researchers of various stripes, we might start to see our way beyond some of the most fractious and debilitating divisions that afflict the discipline. Then, mixed-methods, problem-driven research might be able to better deliver on the founding promise of the field to connect science and democracy, leading to the improvement of both, not just in theory but in practice as well.

-----------------
① On the idea of a unitary paradigm for Political Science, see David D. Laitin, “Disciplining Political Science,” American Political Science Review 89, 2 (June 1995): 454-456.
② See John G. Gunnell, “The Founding of the American Political Science Association: Discipline, Profession, Political Theory, and Politics,” American Political Science Review 100, 4 (November 2006): 479-486.
③ Gunnell, “The Founding of the American Political Science Association,” pp. 480-481.
④ See James Farr, Jacob S. Hacker, and Nicole Kazee, “The Policy Scientist of Democracy: The Discipline of Harold D. Lasswell,” American Political Science Review 100, 4 (November 2006): 579-588; and John S. Dryzek, “A Pox on Perestroika, A Hex on Hegemony: Toward a Critical Political Science,” In Kristen Renwick Monroe, ed., Perestroika! The Raucous Rebellion in Political Science (New Haven: Yale University Press, 2005), pp. 509-525.
⑤ See Bernard Crick, The American Science of Politics: Its Origins and Conditions (London: Routledge & Keegan Paul); David M. Ricci, The Tragedy of Political Science: Politics, Scholarship, and Democracy (New Haven: Yale University Press, 1984); and John G. Gunnell, Imagining the American Polity: Political Science and the Discourse of Democracy (University Park: Pennsylvania State University Press, 2004).
⑥ Rogers M. Smith, “Should We Make Political Science More of a Science or More about Politics?” PS: Political Science and Politics 35 (2002): 199-201. Also see the essays in Kristen Renwick Monroe, ed., Perestroika! The Raucous Rebellion in Political Science (New Haven: Yale University Press, 2005).
⑦ Smith, “Should We Make Political Science More of a Science or More about Politics?” pp. 199-201; and Ian Shapiro, The Flight from Reality in the Human Sciences (Princeton: Princeton University Press, 2005).
⑧ See Sanford F. Schram, “Return to Politics: Perestroika and Postparadigmatic Political Science,” Political Theory 31 (December 2003): 831-851.
⑨ Dryzek, “A Pox on Perestroika, A Hex on Hegemony,” pp. 509-525.
⑩ Keith Topper, The Disorder of Political Inquiry (Cambridge: Harvard University Press, 2005).
11 John G. Gunnell, Philosophy, Science, and Political Inquiry (Morristown: General Learning Press, 1975), pp. 11-28.
12 The Perestroika listserv is replete with examples of dissertation advisers and journal editors who as a rule will not consider case studies. The archives of the listserv can be accessed by emailing perestroika_glasnost_warmhome@yahoo.com.
13 The second-coming for positivism in Political Science was marked by the publication of Gary King, Robert O. Keohane, and Sidney Verba’s Designing Social Inquiry: Scientific Inference in Qualitative Research (Princeton: Princeton University Press, 1994) which argues for how qualitative research can be made to be consistent with the positivist paradigm that tends more often to lead to large-n quantitative studies.
14 See Dvora Yanow and Peregrine Schwartz-Shea, ed., Interpretation and Method: Empirical Research Methods and the Interpretive Turn (Armonk: M.E. Sharpe, 2006).
15 Anthony Giddens, New Rules of Sociological Method: A Positive Critique of Interpretative Sociologies (London: Hutchinson, 1976).
16 See Joan W. Scott and Debra Keates, ed., Schools of Thought: Twenty-Five Years of Interpretive Social Science (Princeton: Princeton University Press, 2001).
17 Clifford Geertz, Available Light: Anthropological Reflections on Philosophical Topics (Princeton: Princeton University Press, 2000). Also see Richard A. Shweder’s review in Science, November 24, 2000.
18 See Peregrine Schwartz-Shea, “Conundrums in the Practice of Pluralism,” In Sanford F. Schram and Brian Caterino, eds., Making Political Science Matter: Debating Knowledge, Research and Method (New York: New York University Press, 2006), pp. 209-221; and Edward Schatz, “Methods are not Tools Ethnography and the Limits of Multiple-Methods Research,” Working Paper, Committee on Concepts and Methods, International Studies Association, January 2007: http://www.concepts-methods.org/working_papers/20070123_26_PM%2012%20Schatz.pdf.
19 See Gabriel A. Almond, “Separate Tables: Schools and Sects in Political Science,” PS: Political Science and Politics 21, 4 (Autumn, 1988): 828-842
20 Joe Soss, “Talking Our Way to Meaningful Explanations: A Practice-Centered View of Interviewing for Interpretive Research,” In Dvora Yanow and Peregrine Schwartz-Shea, eds., Interpretation and Method: Empirical Research Methods and the Interpretive Turn (Armonk: M.E. Sharpe, 2006), pp. 127-149. A good example that combines positivistic analysis of survey data with interpretive in-depth interviewing of selected informants, see Joe Soss, “Lessons of Welfare: Policy Design, Political Learning, and Political Action,” The American Political Science Review 93,?2 (June 1999):?363-380.
21 The sociologist Howard Becker’s contributions to the interpretive turn rivaled Geertz’s in importance for the social sciences. And like Geertz, he resisted the separate logics of inquiry insistence that followed his path-breaking work. See Howard S. Becker, “The Epistemology of Qualitative Research,” In Richard Jessor, Anne Colby, and Richard Schweder, eds., Essays on Ethnography and Human Development (Chicago: University of Chicago Press, 1996), pp. 53-71.
22 For books on mixed methods research, see Abbas Tashakkori and Charles B. Teddlie, Mixed Methodology: Combining Qualitative and Quantitative Approaches (Thousand Oaks: Sage Publications, 1998); and John W. Creswell and Vicki L. Plano Clark, Designing and Conducting Mixed Methods Research (Thousand Oaks: Sage Publications, 2006). For journals, see the Journal of Mixed Methods Research, http://mmr.sagepub.com/.
23 For one example of problem-driven research that employs both interpretive and positivistic approaches, see Joe Soss and Sanford F. Schram, “A Public Transformed: Welfare Reform as Policy Feedback,” American Political Science Review 101, 1 (February 2006): 111-127. This article examines the question of whether Bill Clinton was right that by enacting welfare reform he would be leading the public to become more supportive of helping the poor and welfare recipients, who were seen being comprised mostly of African Americans.  First presented is a positivistic multivariate analyses of public opinion data, which is followed by an interpretive analysis of the symbolism of welfare to understand why welfare reform did not lead to the public seeing the poor, welfare recipients and low-income blacks as more deserving. The multivariate analyses of public opinion data provide a basis for asserting the welfare reform failed as a form of policy feedback to cause change in the public’s reluctance to assist welfare recipients.  The interpretive analysis of the political symbolism of welfare helps understand what welfare means to people so as to better explain why the change in opinion did occur.
24 A recent noteworthy exception is Soss, “The Lessons of Welfare,” pp. 363-380.
25 Shapiro, The Flight from Reality in the Human Sciences.
26 Topper, The Disorder of Political Inquiry.