The Centennial of the American Political Science Association is a propitious point at which to reflect on the manner in which the past of the discipline has been imaged—or imagined. Gabriel Almond's Orwellian dictum that “whoever controls the interpretation of the past in our professional history writing has gone a long way toward controlling the future” (1988, 835) may be a bit over the top, but telling the story of political science has always been a crucial element in both the discipline's internal rhetoric of inquiry and its search for identity. And it has been tempting to apply Thomas Kuhn's imagery regarding scientific revolutions (1970 [1962]) to an interpretation of the history of the field and to a description of the conduct of inquiry. During the 1960s, APSA presidential addresses by David Truman (1965) and Almond (1966), as well as various other dimensions of the language of political science, began to employ Kuhn's concept of a paradigm in speaking broadly about the development of the discipline and about the advent of behavioralism. This use of Kuhn can now be construed as somewhat ironic in that these individuals had not yet grasped the extent to which his account of the epistemology and logic of science tended to undermine fundamental tenets of the positivist inspired behavioral image of scientific inquiry and the attending conception of the history of science as linear and cumulative. Moreover, there remains a question of whether the behavioral revolution was, in Kuhn's terms, a revolution as well as the question of whether his concept is applicable to the history of political science or the social sciences in general.

Kuhn found, in his experience, that natural science communities were less contentious than social science communities regarding what constitutes significant problems and the proper methods of inquiry; it was an attempt to explain this difference that led him to his image of a paradigm and his argument about revolutionary scientific change. But although, he concluded, social science was not prone to such hegemonic dogmas, he noted that “both history and acquaintance made me doubt that practitioners of the natural sciences possess firmer or more permanent answers to such questions then their colleagues in social science”(1970, viii). Thus it was, in his view, more the sociology of the natural sciences than their subject matter that rendered them paradigmatic, but the issue of whether or not there were comparable structures and transformations in the social sciences remained unresolved.

Truman ultimately concluded that despite certain periods of “consensus,” such as that represented in the revolt against “formalism” and the move toward “realism” between 1890 and 1930, that the field had not achieved paradigmatic status. Even though, Truman argued, the contemporary study of politics had overcome the prior “unconcern with the system as such” and produced a “new disciplinary consensus” that promised a recommitment to science broadly construed, he was skeptical about the discipline achieving the form of normal science described by Kuhn (Truman 1965, 866). A year later, Almond argued that there had, in fact, been a “coherent theoretical formulation in the American political theory of the eighteenth and nineteenth centuries” and that this paradigm had persisted into the 1950s despite the anomalies that had “shook its validity” after the turn of the century. This paradigm, he claimed, was based on the concepts of the “separation of powers, checks and balances, and the mixed constitution” and, although characterized by a “rank, nontheoretical empiricism,” it constituted an “empirical and normative theory of politics” which, however, by the 1960s, was finally replaced by a new paradigm based on the concept of the “political system” (Almond 1966). I will not attempt to document the subsequent use, and abuse, of Kuhn's formulation in discussions of the state of political science, since it is only one manifestation of the complicated relationship between the philosophy of science and the theory and practice of social science (Gunnell 1998), but I will argue that Truman was incorrect in his assessment that political science was nonparadigmatic and that Almond was mistaken about the logical and historical loci of paradigmatic concepts.

Despite the continuing controversy about exactly what Kuhn initially meant by the term “paradigm,” whether it referred to a general form of scientific theory and practice or concrete “exemplars,” as well as his more recent emendations (Horwich 1993; Kuhn 2000), he was a theoretical realist. Paradigms, according to Kuhn, are not only “constitutive” of the practice of science but of “nature as well, since when paradigms change, the world itself changes with them” (Kuhn 1970, 110–111). He claimed that there was no essential difference between theories and facts and that scientific theories, and the concepts they embodied, were fundamental claims about what existed and the manner of its existence. The unfortunate term “theory-laden” does not convey his position. In this category of theoretic claims, we could place, as a recent example, the transformation in geological theory represented by plate tectonics and the jettisoning of the belief, which still dominated into the 1950s, that the continents were fixed on the earth's crust. In the case of geology as well as the study of politics, new theories entail new histories of both the subject matter and the discipline, since, as Kuhn noted, “each scientific revolution alters the historical perspective of the community” (1970, ix).

What have been, and are, commonly called theories in political science (systems analysis, rational choice, decision theory, institutionalism, realism, and the like) are not constructions of the kind Kuhn posited. These conceptual schemes of political science are more properly described as analytical frameworks, approaches, or models. They are not basic reality claims that could ultimately be judged untrue in the manner in which, for example, the theory of phlogiston was discarded. Such schemes are predicated on the assumption that “facts” are in some way given to experience and that “theories” are tools or instruments for ordering and explaining data. We can assume, however, that there are theories and paradigmatic concepts conforming to Kuhn's characterization which are implicitly embedded in the “facts” that political scientists adduce and which form part of their “tacit knowledge.” The question I am posing is whether there have been paradigmatic shifts in political science that are logically comparable to what Kuhn took to be represented in the history of natural science, that is, fundamental theoretical revolutions that constitute new paradigms and incommensurable domains of facticity, as well as, in the case of political science, a new normative vision of its subject matter. My answer is that in addition to various innovations and continuities in the “disciplinary matrix,” there has been one such “revolution” in the roughly 150-year career of political science as a distinct discourse and discipline. This limited degree of paradigmatic change should not seem unusual, since most scientific theories have had a longer life than 75 years. This revolution involved a basic transformation in what was considered to be the very “stuff” of politics and, consequently, a transformation in the methods of studying it, and it entailed a normative reconstitution of the idea of democracy and the criteria for assessing the United States as a democractic polity—a congenital and defining concern of political science from its inception and conception (Gunnell 2004). The transition that I am referring to is that between the nineteenth-century theory of the state, initiated by Francis Lieber and more fully elaborated by John W. Burgess and a range of other political scientists in the last half of the nineteenth century, and the theory of democratic pluralism which succeeded it and which was fully articulated by the late 1920s and early 1930s and which persisted through the twentieth century.

Caught, as many reflections on the past of political science have been, in the spell of work such as Morton White's account of the “revolt against formalism” (1957), it is common to characterize the theory of the state as an obsolete archaic legalistic formulation. But while we may today find that theory as implausible as the idea of static continents, there are no independent or external “factual” or philosophical grounds that can be advanced to falsify it and no terms in which we can speak of its successor as representing “progress.” It was a theory of both politics and democracy that constituted the facts of American politics and accounted, structurally and temporally, for the United States as a democratic regime. In this period, what the term “state” referred to was not the institutions of government but the concept of an organic people or community that preceded both the constitution and government and that gave meaning to the ideas of popular sovereignty and democracy. This public, this people, was, to be sure, invisible—as in the case of many theoretical entities, but it was the basis of a vision and explanation of American history as well as of the particularities and character of politics. It provided the framework of values for the practice of political science, and it governed the conduct of empirical inquiry.

Between the late 1890s and early 1920s, the theory of the state was severely strained and finally, in the face of growing anomalies or incipient theoretical challenges, put into crisis. There is a persistent and prevailing belief that it was supplanted by Progressive political science as defined by individuals such as Charles Merriam, but this assumption requires considerable qualification. There is no doubt that the new generation of political scientists who succeeded theorists such as Burgess at Columbia and Herbert Baxter Adams at Johns Hopkins, and who were, as in the case of Merriam and Woodrow Wilson, often their students, embraced a different ideology and championed new methodologies, but there was often an essential continuity in theoretical commitments. As Kuhn claimed, “novelty only emerges with difficulty” and against the background of deeply sedimented beliefs (1970, 64). Some Progressives maintained that the American people were merely dormant and needed to be awakened, while others, such as Merriam and public intellectuals such as Herbert Croly and Walter Lippmann, took the position that if a democratic public did not exist, the task was to create it or its functional equivalent. They came to accept the idea that the state was the government, but they retained the image of an objective public interest if not a public. What slowly evolved alongside the Progressive vision, however, and in response to the crisis precipitated by the demise of the theory of the state, was a new empirical account of American society as pluralistic and lacking any general communal existence. The methodological “revolution” often associated with this period, and sometimes viewed as a proto-behavioral movement, was entailed by this new vision of political reality. Through the 1920s, Progressives, such as John Dewey, and conservatives, such as William Yandell Elliott, clung to the remnants of the old theory, but by the mid-1930s the revolution was complete within the theory and practice of political science (Gunnell 2004). As Kuhn indicated, a theory becomes invalid only when there is another to take its place, and what transpires is more a “conversion” and “agreement” and the “assent of the relevant community” than a “discovery” (1970, 94). The change involved the creation of a new “language community” as the word “state” came, in the work of individuals such as W. W. Willoughby, to refer no longer to the concept of the people but to the institutions of government as surely as the term “continent” came to refer to a different kind of thing than it did before the theory of continental drift—and the word “democracy” meant something as different as the term “motion” after Newton.

Although a pluralist description of politics and a nascent image of how pluralism might add up to democracy was beginning to appear as early as 1914, a full-blown normative and empirical account of democratic pluralism did not emerge in coherent form until the late 1920s. The empirical work of Pendelton Herring and the conceptual innovations of Arthur Bentley were important contributions, but the actual principal authors of pluralism as a coherent and integrated theory, such as John Dickinson and Walter Shepard, are now largely forgotten. And the manner in which they built upon the work of English pluralists such as Harold Laski, and students of Graham Wallas such as G. E. G. Catlin, has been obscured. Their formulation, and their rejection of what they came to call the “democratic dogma,” became, however, the basis of the image of liberal democracy that informed mainstream political science during the 1930s and that was recapitulated before World War II by Herring (1940) and once again after the War by individuals such as David Truman (1951) and Robert Dahl (1956).

The behavioral revolution in political science, despite the claims of its self-ascribed participants, was far from a revolution in the Kuhnian sense. It was actually more a reformation that vouchsafed normative principles and empirical methods that had defined American political science for a generation but which, by the 1950s, had been frontally challenged by a new generation of scholars who had emigrated to the United States from Europe and who, adopting and adapting to the subfield of political theory, had called into question both the commitment to science and the facts and norms of democratic pluralism (Gunnell 1993). The behavioral “revolution” was, in many respects, a conservative movement, a rear guard action, in defense of the reigning theory of democratic politics, and the historical and philosophical vision attending it, that had become deeply embedded in the language and perceptions of the discipline. It was an example of “normal science” solidifying itself under the pressure of challenges rather than a frontal rejection of a previous theory of political reality and democratic norm. What it rejected, if anything, from the past of the discipline, was still an echo of the old paradigm. There is, for example, no theoretical element of Dahl's formulation of pluralism in the 1950s and 1960s that had not been articulated by the early 1930s, and despite the embrace of new philosophical grounds for the idea of science, there was little change in the basic image and goals of science. The introduction of the concept of system by David Easton (1953) and the recommitment to quantitative methods may have, in several senses, revolutionized the practice of political science, but it was not a paradigmatic revolution. It changed neither the basic conception of democracy nor the dominant image of what constituted scientific inquiry and its purpose as articulated by Merriam, Catlin (1927), Harold Lasswell (1950), and others. The reaffirmation of empiricism was in part a defense against the new and unprecedented attack on the idea of a science of politics, but it was also a response to a context that valorized pure science and funded quantitative research. What in part had been involved in the pluralist revolution was a rejection of the notion that there was a substantive domain, such as the state, that constituted the subject matter of political inquiry. Politics was ever after to be defined analytically by formulations that would be termed “theories” but which would bear little resemblance to theories in natural science and which would be still tied to the practical motives and motifs of the discipline (Gunnell 1997). There were also many reasons for the recodification of pluralist theory after the war: sedimented as it was in the practice of political science, it was not immediately visible; it was challenged by claims such as that of C. W. Mills and various Marxists; and there was again, as in the early 1930s, a need to distinguish democracy, and American democracy, from foreign regimes. But, above all, an attack on liberal pluralism, and the science of politics which was grounded upon it, was mounted within political science itself by the new generation of émigré political theorists such as Leo Strauss and their progeny, as well as by certain indigenous thinkers, such as Theodore Lowi (1969), who still subscribed to the remnants of the Progressive vision. It was this confrontation with the basic values of American political science that served to precipitate and galvanize the behavioral movement.

The behavorial era was admittedly marked by methodological innovation and refinement, but it largely advanced what was the “normal” science of politics, and, above all, it was predicated on the pluralist theory of democracy which it in turn sought to validate. The estrangement between mainstream political science and the subfield of political theory, which was rooted in debates that began to appear in the 1950s, was really a function of a fundamental disagreement about democratic theory and political reality. The post-behavioral era was not characterized by any final resolution of the basic issue but rather by two academic communities going their separate ways. The theory of democratic pluralism remained the foundation of mainstream political science, while much of the conversation of political theory, from the 1960s through most of the 1980s, was dominated, and defined, by a persistent rejection of the empirical and normative dimensions of that theory.

Where are we today? I would suggest that political theory, as the “vocation” conceived by individuals such as Sheldon Wolin (1969), was, in many ways, an attempt to return to vestiges of the old paradigm as a way of dealing with the anomalies that increasingly seemed to plague the theory of pluralist democracy, and Easton's call for a “post-behavioral revolution” (1969) harked back to the practical concerns that had traditionally animated the field of political science. The search for “the political” and for what Wolin referred to as “fugitive democracy” as well as Lowi's vision of “juridical democracy” were still grounded on an image of what Lippmann had, two generations earlier, termed a “phantom public,” and both also reflected the practical commitments that had characteristically defined the discipline of political science. By the mid-1990s, the apostate “calling” of political theory had, for a variety of reasons, largely collapsed as something other than a separate professional identity. Its theoretical challenge to behavioralism evaporated, and there was a relatively sudden wholesale return to the fold of pluralism—even by some of those who had been its most passionate critics. Albeit, the philosophical justifications were often new and diverse, but by the beginning of the twenty-first century, it was difficult to find any major persuasion in political theory that did not assume the reality of pluralism and seek to accommodate and build upon that reality the requisites and possibilities of democracy. The pluralist turn was reflected in a wide range of claims about deliberative democracy, and whether it was called associational democracy or the recognition of difference, pluralism was both the “fact and norm” for everyone from John Rawls to Jürgen Habermas. The new pluralism is, in many respects, not the same as the old pluralism—any more than contemporary evolutionary theory in biology is the same as Darwin's account, but it is, at bottom, the same theory. What is involved is change within a theory as opposed to change between different theories. For better or worse, pluralism evolved and survived the challenges of the midtwentieth century.

The “real” revolution in political science took place four score years ago, and, in the end, we have not been able to think effectively about political reality and democracy outside the theoretical terms of that revolution. Neither the behavioral revolution nor the post-behavioral revolution were Kuhnian revolutions. Even communitarian liberals and those who equate democracy with social capital accept pluralism as a sociological reality and hardly see it as the threat to democracy that Hannah Arendt, Wolin, or Habermas had suggested in their erstwhile defenses of the public realm. There may be worries about the facts and values of multiculturalism and identity politics, but few dare suggest any longer, as many had in the 1960s, that social pluralism is the congenital disease of democratic society. What is worth considering, however, is whether the anomalies, and pathologies, once attributed to pluralism as a theory of democracy have been successfully resolved by the contemporary formulations or whether recent pluralist theory represents what Kuhn designated as the tendency of debilitated paradigms to maintain themselves through the medium of ad hoc accretions.