“Social science is an epistemically privileged discourse that gives us knowledge, albeit always fallible, about the world out there. Poetry, literature, and other humanistic disciplines tell us much about the human condition, but they are not designed to explain global war or Third World poverty…. In my view it cannot be ideas all the way down … because ideas are based on and are regulated by an independently existing physical reality.”
Alexander WendtFootnote 1“I define science as a publicly known set of procedures designed to make and evaluate descriptive and causal inferences on the basis of the self-conscious application of methods that are themselves subject of public evaluation…. Political science is the study of politics through the procedures of science.”
Robert O. KeohaneFootnote 2“I should like to say once again that … political theory … differs from political science or any other empirical enquiry in being concerned with somewhat different fields …”
Isaiah BerlinFootnote 3Statement of the Problem
In the fall of 2007, the department of political science at Penn State University eliminated political theory from the set of subfields in which it would train its graduate students. The event sent shockwaves through the political theory community, and generated two widely circulated letters of protest signed by over 100 political theorists.Footnote 4 As Ben Barber, author of one of the letters, wrote, “We believe that it is essential to the well-trained political scientist and teacher, whether in American, Comparative, IR, or Public Law, that they have a training that includes an underpinning in political theory.” The letter also threatened retaliation: if Penn State went ahead with its decision, the signatories pledged to steer their undergrads away from graduate study at Penn State, and they would “not be able in good faith to recommend Ph.D.s from Penn State to our colleagues for future positions in our departments—again, whatever the subfield.”
As a response of compassion for the faculty and graduate students affected by the decision, the expressions of solidarity were understandable. As arguments, however, the letters left a good deal to be desired. While they might best be understood as acts of containment—were theorists to lose Penn State, political theory programs elsewhere could fall like dominoes—the letters failed to answer the central question posed by Penn State's decision: Why does political theory belong within the discipline of political science; in what way is it proper to categorize this work as “social science” at all?
The answer is not immediately obvious. Theorists are often marginalized within their departments, their contributions to the general disciplinary journals have an “odd man out” quality to them, and their scholarship is often treated as trivial. Many political theorists themselves share their colleagues' belief that political theory is not part of a social scientific understanding of politics.Footnote 5 Some theorists claim to be concerned with interpreting politics rather than explaining its causal mechanisms, a distinction other theorists have claimed makes theory a humanistic, not a scientific, endeavor.Footnote 6 Some theorists “have come to see themselves as engaged in a specialized activity distinct from the rest of political science” primarily because they distance themselves from the broader scientific aspirations of the discipline.Footnote 7 Other theorists are even more skeptical of those aspirations; their complaint is not that political theory is not science, but that so-called “social science” is not science either. As Raymond Geuss claimed, “If one thinks that the mark of an ‘empirical science’ is to yield successful, reasonably precise predictions that go beyond the pronouncements of common sense in complex and changing situations, then it is a genuinely open question whether there are any social sciences.”Footnote 8 And some political theorists, like Sheldon Wolin, believe their vocation should be an engaged and critical activity, involving advocacy, nuanced understanding, and deep appreciation, rather than a systematic attempt to know observer-independent facts about the social world.Footnote 9
If the discipline struggles to understand why theorists belong to it, and theorists seem suspicious of, if not hostile to, the scientific study of politics, then why, exactly, should theorists care to defend their status as political scientists at all? Assuming that there are pragmatic and epistemic reasons to recognize a useful and reasonably permanent division within the university between the social sciences and humanities, why should theorists be classified as the former rather than the latter? In short, what conception of social science is commensurate with the various activities that constitute “political theory?” These are my central questions.Footnote 10
The key to an answer is to note that “political theory” denotes many different kinds of scholarship: conceptual, normative, explanatory, historical, interpretive, and textual. Authors frequently combine two or more of these in their research, and they may also advocate for political and social change. As a whole, political theory is an interdisciplinary field: some of it is social-scientific, some of it is not. The purpose of this article is to explain what kind of political theory is “political science” properly conceived, and what kind is not, and to argue that only that which fits within a broad casting of “political science” ought to remain. Theory that makes political phenomena its primary object of study, that aims to discover observer-independent facts about those phenomena (including what political communities ought to do about them, given certain assumptions), and that makes claims about those phenomena that can be falsified (empirically or otherwise), is political science properly conceived. As such its practitioners have a justifiable claim to disciplinary resources. By contrast, theory that makes authors and texts its primary objects of study, that rejects the existence of an observer-independent world, and that makes claims incapable of being falsified, or that aims primarily at advocacy rather than the generation of knowledge, is not political science properly conceived, and its practitioners have only second-order claims, if any, to the resources of the discipline.
Drawing from history, philosophy, public policy, law, and literature, as well as from political science (properly construed), “political theory” should thus constitute itself more self-consciously as interdisciplinary scholarship. The value of interdisciplinarty scholarship, however, depends largely on the strength of its distinctive, disciplinary contributions rather upon than the creation of some new discipline from earlier strands. The value of political theory as an interdisciplinary activity thus entails reifying and strengthening its disciplinary foundations, which in turn should guide how we think about where political theory gets done in the academy. And the argument here results in its own normative prescriptions: we should strengthen and enhance the role of social scientific political theory within political science, encourage training in methods for humanistic scholarship and political advocacy outside of political science, and combine the two in interdisciplinary centers and programs that coexist with independent and distinct disciplines.Footnote 11
Though we should not fetishize the distinction between the humanities and social sciences, nor its further subdivision by academic departments, neither should we simply reject the distinction as wholly arbitrary. Certainly, academic distinctions are historically contingent—there is a story that causally explains why they are what they are, and demonstrates that they could have been arranged differently. But it does not follow that the organization we now have is capricious or lacks merit. In fact, dividing inquiry into natural science, social science, and the humanities is a practice worth defending (even if it is also true that as a practical matter these distinctions are here to stay anyway).Footnote 12 It advances our knowledge, understanding and appreciation of our world to approach it in scientific and humanistic ways, where “scientific” and “humanistic” denote different sorts of activities. There are good epistemic reasons why people who study zygotes should not be housed (administratively or physically) with people who study Bach, elections, or the Mundurucu people of Brazil. If hermeneutical approaches to the study of texts on the one hand, and empirical analyses of behavior on the other, are very different sorts of activities that contribute in their own ways to the accumulation of knowledge of, or appreciation for, our world, I see no reason to collapse those differences in training and academic organization. I do not deny that knowledge of each is often helpful to the other. I simply presume that these broad disciplinary distinctions have value and that they will be reasonably stable in the foreseeable future.Footnote 13
Why take up this argument now? Because, for one reason, political science may be headed the way recommended by Auguste Comte.Footnote 14 Political theory is now an ancillary subfield in all but the top political science departments, and even in some of those departments, like Cal Tech's and the University of Rochester's, political theory is either devalued or entirely absent. Nor does political theory appear set for a resurgence, as Penn State's decision to eliminate the subfield may reflect. Perhaps the reason many departments still have theory at all, as the chair of a political science department at a small Midwestern university confidentially explained to me, is “not so much that we think political theory is really social science; we just think it's good for students to read.” In an era of downsizing and resource constraints in higher education, these are not particularly good portents.
Yet while there are reasons to be concerned that political theory is in a precarious state—and anxieties about this have motivated an entire sub-genre of work, of which this now is a partFootnote 15—its precariousness is beside the point. Whether or not theory is being shut down or marginalized, the concerns behind these anxieties suggest not only an existential problem—will theory remain?—but a conceptual problem—does theory belong? It is this conceptual problem of inclusion that I believe prior accounts of the role of theory in political science have failed to address, and that this article seeks to address.
Justifying the conceptual bounds of the discipline so that we can determine what should count as “political science” may strike many as an irresolvable problem, if not insufferable navel-gazing. But such a conclusion would overlook the fact that the gatekeepers of political science—of any discipline—regularly use such criteria to distribute professional resources. Its journal editors use them to decide whether to publish articles. Its book review editors use them to decide whether a title should be sent out for review in those journals. Its departments use them to decide whether someone's work counts as political science and thus that they are worth hiring or promoting.Footnote 16 If it were truly impossible to justify one such set of criteria over another, then these judgments would be just so much bullying in the name of preference satisfaction. Even if that describes how these decisions are, in fact, made, it hardly makes them justifiable or defensible. Such imposition of arbitrary rule may be the very definition of injustice itself, and the very heart of the problem of domination. To dismiss a discussion of disciplinary criteria out of hand is to ignore the very real consequences of an arbitrary exercise of power to the lives of those currently doing political theory.Footnote 17 And, of course, it ignores the epistemic consequences to the social-scientific study of politics, should theory be removed from its ken.
Dismissing the discussion because we think disciplinary divisions should not be so hard-and-fast, or that subfields should simply live and let live, leaves us with the situation that political theorists are anxious about, one in which decentralized, uncoordinated decisions at individual department levels are effectively treating theory as so much oncology—something that belongs somewhere else—even if those responsible for these decisions are not trying actively to remove it. At least, we should ask departments, journals and perhaps the discipline as a whole to occasionally struggle with this question and insist that gatekeepers state clearly the reasons they have for the criteria they use.
The consequences of removing all political theory from political science without regard to the differences between its humanistic and social scientific strands would be a loss to political science on its own terms, for it would hamper a complete scientific understanding of politics, just as the absence of theory from economics and psychology has been a loss for those disciplines. To give but one well developed example drawn from other social sciences: We would know more about economic activity if we understood more about value, benefit, and exchange, but that work, being essentially theoretical and hence no longer performed in departments of economics, has been relegated largely to philosophy departments.Footnote 18 Economists who nevertheless continue to speak of “utility” but who cannot cash it out, as it were, except by some appeal to monetary currency are making a conceptual error that limits their understanding of a very real social phenomenon.
On the other hand, the discipline of economics has arguably benefited from not having to spend its resources on humanistic interpretations of Adam Smith, or its faculty's time supervising dissertations on the influence of the Enlightenment ideals on Smith's economic thought. Similarly, psychology is better off for its faculty and students not having to work out Aristotle's view of motivation and depression. And while Aristotle also wrote a good deal about the natural world, it would hardly benefit physics if physics departments started training their students in interpretive techniques of the ancients.Footnote 19 Of course, some will rightly notice that an interpreter of Adam Smith's thought would benefit from training in economics, just as an economist would benefit from knowing more about Adam Smith. And it is also almost certainly true that scholars of religion—whether devoted theists or philosophers of religion—would benefit from substantive training in and knowledge of particle physics, to the extent such training and knowledge gets at the foundations of the universe. Does it then follow that physics departments should expand to include philosophy of religion, and require coursework in it as part of their graduate training of physicists? Shall we support a move to encourage departments of religion to build particle accelerators? The fact that there will often be fruitful cross-disciplinary engagements does not tell us where primary disciplinary training ought to be done. And, as this example illustrates, the arguments some political theorists have made to defend their inclusion in the political science would not be accepted if generalized to other disciplines.Footnote 20
What we need is an offensive argument for what the boundaries of political science ought to be, for why some political theory by right belongs inside them, and for why some political theory does not.
What follows is my own attempt as a theorist to be reflective about these disciplinary boundaries, in the hopes of building a more stable and secure foundation for political theory generally and in political science more particularly. It looks toward a time when an article titled “Why Political Science Needs Political Theory”Footnote 21 would seem as ironic as it was unnecessary; when departmental decisions about resource allocation can follow reasonable departmental views, and not occasion visions of the apocalypse in the minds of colleagues at other institutions; and when our collective efforts might go to doing political theory, rather than explaining why what we do belongs where. By calling for a clearer division of theory into its humanistic and social-scientific modes for training and professionalization, this article may be “offensive” in both senses of the word. But its offensiveness is meant to serve the goal of a better, more secure and productive home for political theory in the social sciences and in the university at large.Footnote 22
The argument continues over four more sections. In the next section, I explain why defensive arguments for remaining in the discipline fail, and why an offensive argument is needed. In the third section, I argue for a demarcation of “political science” based on a broad conception of “the political” (the use or potential use of power over other people) and of basic presumptions of “science” including the systematic discovery of a world that exists independent of our observations of it, and the condition of falsifiability. In the fourth section, I apply these criteria to the various genres of political theory. Finally, in the last section I discuss the implications of this definition for the interdisciplinary practice of political theory in political science and in the university more generally.
Why a Defense of Political Theory Fails
Political theorists have repeatedly and self-consciously explored the place of political theory in political science.Footnote 23 Before turning to the offensive explanation in the remainder of this article, it will be useful to consider why attempts to defend theory's inclusion in political science typically fail.
Consider the arguments in the two letters sent to Penn State that offered reasons to include political theory in political science. In the letter endorsed by the Foundations of Political Theory section of the American Political Science Association, we discover that “[i]n the development of the discipline of Political Science … political theory … has been [its] moral voice”; that political theory usefully raises questions of gender and race; and that it “remind(s)Footnote 24 us that our methodological choices have normative and ethical implications that we cannot ignore.” Further, we learn that more theorists attend organized sections at APSA conferences than do members of other subfields, which “reflect[s] the central place of political theory in the discipline as a whole.” In the more strident epistle written by Ben Barber and signed by 60 other political theorists, we are told that graduate students will “do better in graduate political science programs where they will also get some degree of fundamental training in political theory,” and that political theory is “essential to the well-trained political scientist and teacher.”
The letters display two of the three kinds of pleadings that typify a defensive position: (1) that theory is important (it is “the moral voice” of the discipline, it raises useful questions), and (2) that theory has a claim to membership in the discipline based on its origins and present practice (our panels have better attendance at the APSA than any other subfield's; our articles are published in the top disciplinary journals). A third defense, not reflected in the letters, is (3) that a political (or any social) science modeled on the natural sciences is trivial, if not impossible. Each of these claims may be true—I happen to think that in some form, the first two are true—but they fail to address the conceptual question of why political theory belongs within political science.Footnote 25
Political Theory is Important
The first defense of theory's inclusion in political science is that it is an important intellectual endeavor. The motivation for such a defense may be an understandable response of theorists who feel belittled or marginalized by their non-theory colleagues: empiricists and statisticians who do not read Aristotle or Foucault, who proudly eschew literature, and who do not frequent the symphony or other cultural institutions. In other words, it may be an understandable response to being in professional association with other individuals who apparently have no real interest in leading fully human lives. While their broad intellectual and cultural interests make theorists superior human beings, they do not justify theory's inclusion in the discipline.Footnote 26
There are three variations of the “importance” defense. The first holds that political theory in all of its varieties is important for its own sake, because it fosters deep reflection upon eternal problems of the human condition. I accept this claim as true, and unproblematically so. The second variation holds that because political theory is vitally important to real-world politics, because the world benefits from the promotion of arguments concerning legitimacy, justice, rights, equality, and so on, what theorists do is necessary. (A defense of this sort is associated with the mode of political theory that explicitly endorses critical advocacy and engagement with the political world.Footnote 27) Yet neither theory's intrinsic importance nor its usefulness to politics constitutes a necessary or sufficient condition for including theory with political science. There are many important and meaningful things—like oncology, the music of Bach, and nuclear particle accelerators—and many other things necessary for politics—like teleprompters, public speaking lessons, and money—that no one would seriously argue should be part of political science on account of their importance or necessity.
The third variation of claims about the importance of political theory is that it is useful to those who do political science. I will first note that claim is exaggerated as a description of what actually happens in the study of politics. I would venture to say in fact that many more political theorists find what empirical political scientists do useful than the other way around.Footnote 28 Empirical political science articles may occasionally start with gems of wisdom from some dead white guy (it seems most often to be Aristotle or Tocqueville) to signal the theoretical heft of the question being studied, but the subsequent article rarely engages the problems political theory has identified. One might more substantively claim that political theory is useful because it generates empirical questions for study, or useful because it explains why some questions—for example, questions about race and gender, as asserted in the Foundations letter to Penn State—are important to ask.Footnote 29
These claims are sometimes collapsed together, but their separation is critical. First, the fact that empirical political scientists rely on conceptual clarification to help them ask the right kinds of questions helps us see why such conceptual work is itself political science, not merely useful to political scientists. As Brian Barry observed, when a normative theorist notes that the concept of “social exclusion” should include both voluntary and involuntary exclusion, that is a claim about the proper object of study, and it contributes to our discovery of facts about the political world.Footnote 30 Or, for another example, when empirical political scientists equate “legitimacy” with “public support” or “public approval,” they are either using language in a very funny way, or they are simply making a mistake. For were we to treat “legitimacy” as synonymous with “public support”, we would be forced to admit that, to the extent he enjoyed mass public approval, Hitler was a legitimate ruler. (Uh-oh.) Failing to recognize the conceptual difference between legitimacy and approval is failing to understand the world as it actually is and to provide a complete social-scientific rendering of that world. (I will return to this point in greater detail in the fourth section.)
There is second argument that political theory is useful to political science not because of the concepts it identifies, but because it helps direct the work of empirical political scientists towards important topics. We might call this claim, (made by Brian Barry and described in the attached letters as the “moral voice of Political Science”) an appeal to the usefulness of “agenda-setting by reference to moral and ethical norms.”Footnote 31 It is an ominous assertion to be sure, that, if not for theorists, other methodologically driven political scientists would be in some kind of moral wasteland asking unimportant and trivial questions, ignoring questions of race and gender, and, one must add, oppression more generally. With their moral senses so dulled one might also fear they are engaging more generally in acts of moral turpitude, so it may certainly be a good thing that theorists are here to help.
Nevertheless, even if it were true that other political scientists were unable to choose morally important topics, or understand the ethical implications of their work, engaging in this kind of agenda setting does not qualify someone for membership in political science, or any corresponding discipline. Consider an analogy to physics. In the 1940s some physicists learned how to split the atom in large part because defeating the Nazis was seen as a moral imperative. Even though that agenda-setting decision caused physicists to turn their attention to the development of nuclear weapons, the decision itself was in no way an instance of doing physics. The choice of political scientists to study race and gender, or more broadly, the choice of any scholar to study what he or she does, is the result of many different factors, including the influence of government agenda-setting, the pressure from social relationships, and the desire for recognition from one's peers, none of which become political science for their influence on what political scientists study.Footnote 32 It may be that since politics as well as ethics are involved in any agenda-setting, it is easier to confuse such agenda-setting in political science for the practice of political science in a manner that we would obviously reject in other disciplines.
More generally, there are many things of use to those who do political science, but use-value to others is neither a necessary nor a sufficient qualification for inclusion in a discipline. Particle accelerators are useful for doing physics, but those who design and build them are not physicists for having done so.Footnote 33 Audio equipment is useful for studying the music of Brahms, but the study of audio technology does not belong in a department of music history. Use-value might in practice explain a subfield's inclusion in a discipline, but that would always be a second-order justification, dependent upon the continued goodwill of first-order political scientists who find their work useful. Strategically it leaves members of such subfields vulnerable to precisely the kinds of problems currently faced by political theory when political scientists do not appreciate the value of their work. The task for political theorists is to explain why what they do is political science properly conceived, no matter its use to others.
Political Theory is and Always has been Part of the Discipline
A second variation of defensive argument justifies the inclusion of political theory in political science on the basis of history and present practice: Political science emerged historically out of political theory,Footnote 34 and if you define political science by its current practice (what gets done in its departments, what gets published in its journals, what groups are most active at its professional meetings) you will indeed find political theorists among its practitioners.Footnote 35 Unlike defenses based on importance or usefulness, which can only offer second-order justifications for inclusion, this defense qualifies as a first-order argument for membership: a “political scientist” is anyone doing what political scientists have historically done, and anyone currently working in a political science department and publishing in political science journals. This defense can explain why political science departments do not advertise for oncologists or music theorists: oncologists have never been part of the discipline in the past, and sane music theorists do not attempt to publish their work in the American Political Science Review. Political theorists, on the other hand, belong in political science because we have always been here, and we continue to be here.
Unfortunately, history and present practice provide no more of a basis for political theory's inclusion in political science than does any longstanding practice provide a justification for any kind of entrenched institutional arrangement. Yes, political theorists were instrumental in the creation of the discipline, and this fact may help us to causally explain why we are here now. Yes, the most esteemed journals in the discipline publish articles on Xenophon, and this fact may help us to causally explain why their authors advance in their careers. And yes, individuals do hold advanced degrees in political science having written dissertations interpreting the meaning of American exceptionalism. But these facts do not address the central question of why we ought to maintain current and historical practices.
Complexity of the Social World
A final variation of defensive argument for the inclusion of political theory in political science rejects the scientific aspirations of political science (and social science more generally). Since “political science” is not really science anyway, there is no harm in admitting political theory, which is also not science, to its fold. In large part, this argument depends on seeing social complexity as a unique kind of difficulty for inquiry to overcome, either because human free will might interfere at any moment, or because of the limits to using human experiments to yield results. Thus, social science can yield at best trivial information about simplistic causal questions of political life—and then only about the past.Footnote 36
The first strand of this defense is an empirical claim that we should not rely on political science because the social sciences have not had the kind of success that the natural sciences have. As Bent Flyvbjerg put it, “the study of social phenomena is not, never has been, and probably never can be, scientific in the conventional meaning of the word ‘science’; that is, in its epistemic meaning.”Footnote 37 Yet, if the social world admits even of the same complexity as natural science, there is good reason to think that the social sciences would take longer and require more technical sophistication to produce results than the physical sciences did; there is no basis yet for thinking that such inquiry is impossible. In the seventeenth century, Galileo could tell us something about the stars using a telescope and his wits. Two hundred years later, Durkheim needed the resources of national governments to collect rudimentary data on suicide rates to propose rather crude (by contemporary standards) explanatory hypotheses about suicide; he would need far greater computing power and data collection operations to generate anything close to the kinds of causal tests that have been conducted for centuries in laboratories of the natural sciences. As recently as 1966, Gabriel Almond argued in political science's flagship journal that “in the last decade or two the elements of a new, more surely scientific paradigm seem to be manifesting themselves rapidly.”Footnote 38 Such a view arguably dominated the natural-scientific world already by the seventeenth century, if not before. Given such constraints, Flyvbjerg's conclusion seems premature.
The claim that a political science modeled on the natural sciences is impossible because it has not had much success also ignores many core regularities that political science has discovered, like the median-voter-theorem, or the way issue framing and agenda-setting lead to sub-optimal choices. As a tactical matter, the claim further marginalizes theorists from their colleagues and is at the heart of the ill-fit between some theorists and their empirical colleagues. And using an “impossibility” argument to trivialize or discourage social science bears a striking similarity to early religious suspicions that knowledge of the natural universe was not accessible to the human mind. These arguments became self fulfilling prophecies that retarded the development of natural science and technological progress, and would do the same in the social sciences if they were widely accepted.
The second strand of this defensive argument is more substantial and philosophically sophisticated. It hinges its rejection of social science on the nature of the object being studied, not merely on its presumed lack of success. Human society, it is supposed, is too complex to be comprehensible by objective observation; it can only be fully comprehended by subjective interpretation. Theory thus belongs in political science because it is a mode of inquiry that can make sense of “the human,” the only mode of inquiry that can make sense of “the political.”Footnote 39 Yet if political theorists believe that social science cannot be done in the mode of natural science, for whatever reason, it is unclear why they would want to remain in a discipline with so many scholars who believe it can be and who are dedicated to purusing that aim. So even if the critics are right, and social science is not “science” at all, the complexity argument does not defend political theory's inclusion in a discipline dedicated to a scientific study of politics; rather, it justifies its emigration out of it (if not activism to shut down the billions of dollars that goes towards such “impossible” activity).Footnote 40
What is “Political Science”?
Establishing Criteria for a Membership in the Discipline
So long as political theory's inclusion in political science is only defended by arguments that do not get at the conceptual core of the question, does theory belong?, its position in political science will remain tenuous. By contrast, the first step of an offensive approach would be to establish what Karl Popper called criteria of demarcation for “political science” that can be used to evaluate what kinds of work would qualify as such.Footnote 41 The second step, taken up in the fourth section, would be to use those criteria to demonstrate why some kinds of political theory belong in the discipline and why others do not.
The demarcation criteria for which I will argue are derived from this preliminary formulation: political science properly conceived is research that takes political phenomena as its object of study, using a method that does not violate the assumptions of science. There are three important elements of this formulation: an activity (“research”), an object of study (“political phenomena”), and a method (“the assumptions of science”). The claim that political science is research may seem obvious; it explains why playing basketball is not political science.Footnote 42 But it also explains why advocacy as such is not political science (even if political science is extremely useful to advocacy, a point I will return to in the fourth section).
In this section of the paper I will develop the second and third elements, since they are doing most of the analytical work. But first, I should address two worries and explain two constraints on the analysis.
The first worry is that any answer to the question what constitutes political science? is open to infinite, irresolvable debate, because the boundaries of political science are “essentially contestable.”Footnote 43 Although any proposed boundary will be contested, there is no reason that a project of demarcation must therefore fail. No one reasonably believes that oncology or music theory should be classified as “political science”, or that political science journals should publish their research, or that political science departments should train people in their methods. And the reason it is unreasonable to believe these things is because oncology and music theory lie outside even the fuzziest boundary of what “political science” rightly denotes. Nor are these judgments merely pragmatic ones of this kind: “oncologists should not be trained in political science departments because people currently in those departments would not train them well.” If the discipline of political science were to reconfigure itself so that it could train oncologists, it would cease to be political science, even though we might continue to use the label “political science” to refer to it. The study of cancer, or of music, does not constitute political science rightly understood. We all have some rough idea of what should not count as “political science,” which means there is indeed agreement at a certain level of generality about what political science denotes, and that demarcation work can be and is being done. Indeed, to remind the reader of what I said at the introduction, gatekeepers of the profession regularly use such criteria to decide whose work should receive disciplinary resources, whose careers thus are successful, and, ultimately, whose lives as political scientists go well. Ignoring or dismissing such work as navel-gazing evades the justificatory project at the heart of many if not all theories of justice.
A second worry is that an argument about criteria that derive from the terms “politics” and “science” arbitrarily proceeds from the name given to most departments in the discipline, “Political Science”—so that my argument would seem to be different were the discipline to be named “Politics,” or “Government,” as some departments in the discipline are.Footnote 44 However, I do not alight upon “politics” and “science” as key elements of my criteria because of the name of the discipline. The name, as a signifier, is doing no analytical work whatsoever: were we to rename the discipline and all its departments “Politics,” “Government,” or “Pinky Tuscadero,” the analysis that follows would remain the same. I presume that the proper aspirations of the social sciences are to come to know things about the social world as it actually is, and that if one believes there is no “as it actually is” there can be no science at all, social or otherwise. So, for epistemic purposes, disciplines should be formed around their objects of study (e.g., power, exchange, culture) in large part because different objects of study will require different specific methods in order to know things about them. These two elements of my criteria of demarcation—political phenomena and the assumptions of science—explain why most departments and the discipline itself are called what they are, not the other way around.
The fact that different objects of study require different kinds of methods explains why the subdivisions of the social sciences (as well as the natural sciences and humanities) are useful and come to have the names they have. We presume that different methods may be required to study culture than would be used to study of politics, precisely because culture and politics are different sorts of entities. But we should be very careful to avoid what I will call “political science exceptionalism,” the view that the study of politics is so completely different from other social phenomena that it requires wholly different methods of study. We can avoid political science exceptionalism and judge the strength of our claims about what should and should not be included in political science by generalizing those claims to the study of other social phenomena, to see how well they fit there. For example, if conceptual analyses of power belong in a political science department because conceptual analyses of power are a part of a complete understanding of the political world, then presumably conceptual analyses of culture, exchange, and society belong in departments of anthropology, economics, and sociology respectively. Similarly, if textual analysis of Plato's Republic, or hermeneutical analysis of an election, belong in political science, because textual and hermeneutical analyses are a part of a complete understanding of the political world, then presumably a textual analysis of Adam Smith's Wealth of Nations and a hermeneutical analysis of a nation's GDP belong in an economics department. If we find we cannot generalize our claims in these analogous ways to the study of other social phenomena, then we will have to justify the exception we are making for the study of politics. This test is meant to keep us honest by separating principled claims for inclusion in political science from merely pragmatic claims that we make to protect or reject a particular field of inquiry.
Finally, a narrow definition of “political science” would be contentious, and unlikely to generate much agreement, as the disciplinary debates of the last decade that emerged in the so-called “Perestroika” controversy demonstrated.Footnote 45 Although there are clearly connections between my argument and that controversy, the methodological component of the Perestroika debate was almost exclusively one of what constituted “science.”Footnote 46 In what follows, I will discuss disciplinary boundaries at a broader level and in ways that mostly leave that debate intact. I take no position on whether small-n, qualitative and descriptive studies of the sort for which many self-described Perestroikans advocated are a good way of doing science. They meet the assumptions of science, and thus cannot be excluded ex ante. Footnote 47
Certainly, as the Perestroika debate illustrated, modeling a study of society on the methods of the natural sciences presents numerous obstacles to achieving success, but as many, many others have written before, such obstacles are complicated, nuanced, demanding and difficult to overcome precisely because the object of study is complicated, nuanced, demanding and difficult.Footnote 48 And so long as the aspirations of the social sciences are to come to know things about the social world, a number of useful distinctions can still be made concerning membership in their disciplines. At that level of generality we can proceed on the basis of minimal assumptions both about what science requires and about the scope of political phenomena, and avoid demands about the possibility or validity of using specific methods. I thus want to aim at overly broad criteria for political science that will err on the side of including too much and leave room for a good deal of contestation. As we will see, even overly broad criteria that leave a good deal of methodological pluralism within political science will still provide a useful tool by which to determine what kind of political theory is political science and what kind is not.
With these preliminaries out of the way, I propose that to count as political science, inquiry must meet these three criteria:
(1) it takes political phenomena to be its object of study;
(2) it assumes that there is an observer-independent world about which its aim is to discover facts; and
(3) it issues claims that are falsifiable (empirically or otherwise).
I emphasize as strongly as I can that (2) and (3) are assumptions of science and not ontological assertions. Whether or not they are true, in order to do science, one must assume that they are.
Political Phenomena
The first of our offensive demarcation criteria is that, to count properly as political science, research must take political phenomena as its object of study.Footnote 49 It is not easy to define precisely what counts as “political phenomena,” in part because many seemingly non-political events have political causes or political implications. A study of Richard Wagner might either be a study of music or of politics, or quite likely both, since it is a matter of debate whether Hitler's promotion of Wagner's music as authentically German aided in the devastation of European Jewry during WWII.Footnote 50 It is difficult also because political phenomena are tied to other objects of culture, society, and exchange. Yet the fact that non-political phenomena can have political causes and implications, and the fact that there may be no firm boundaries between different sorts of social phenomena does not mean that it is impossible to define “the political.”
Let me propose that for something to count as a political phenomenon it must involve the use, or potential use, of power over people. This unoriginalFootnote 51 and merely necessary condition explains why the structure of a musical score is not a political phenomenon, even if the decision to perform it might be. This condition also excludes as political the study of states of consciousness and solitary acts done in private in and of themselves. Footnote 52 As with the music example, these states of consciousness may have political origins or implications, and to the extent a particular study concerns such origins and implications, it would meet this condition. But the personal is only political to the extent it has consequences for other people, or is a consequence of the power that others have imposed. In short, I don't think there is a single phenomenon that we would rightly call “political” that does not involve the use, or potential use, of power over people.
Being a necessary but insufficient condition, this formulation errs on the side of over-inclusion. For example, by this view a parent's relationship with a child would count as a “political phenomenon,” as would a mountain climber leading an expedition, an airline pilot flying a plane full of passengers, or a parent driving a carpool, since they all involve the use (or potential use) of power over other people—even though I do not believe that these are normally “political acts,” in any full account of that concept.Footnote 53
Discovering Facts about an Observer-Independent World
Judging from the forests destroyed mounting various defenses of this or that conception, arriving at a non-controversial definition of “science” would be even more difficult than arriving at one of “political phenomena.”Footnote 54 Even before getting to a definition, some would deny that humans can perform any objective (or non-subjective) inquiry that reveals observer-independent truths about the world. Such a claim may be based on the inconsistent or at least solipsistic claim that no world exists apart from the observer, and thus that we cannot study something that does not exist. Or it may be based on a stronger, and more plausible, epistemic skepticism: objectivity is impossible in practice, because there is no way to access the world without altering what we find. Perhaps not surprisingly, the debate surrounding epistemic skepticism quickly bleeds into questions of power. Since we cannot arrive at objective knowledge about a world that exists apart from our observations of it, it is said, any description of that world is a mere story we tell each other in order to impose our will on the world.Footnote 55 It is a short step from there to the now-familiar critique of science as merely a hegemonic enterprise, fueled by those in power to secure their own privileged status and to oppress anyone who offers alternative “epistemologies” or “ways of knowing.”Footnote 56
A different kind of critique accepts the usefulness of science to study the physical world, but questions its usefulness when applied to the social world. A stick of butter will hold still while you observe what happens if you heat it to 150 degrees, but a human being will simply leave the room.Footnote 57 One can study the regular motions of the planets or the probabilistic motions of atomic particles; but human beings can change their minds.Footnote 58 In its most moderate form, the objection is that a study of society must at least incorporate both scientific and non-scientific methods to fully know and understand our social condition.
I do not dispute the fact that the social world is different in kind from the physical world, and, for that reason, that the social sciences may require a different set of methods than the natural sciences. But they are linked by one common aspiration: they are, and in order to be “science” must be, structured to discover facts about an observer-independent world. Thus to be counted as “science,” a method must assume that there exists an observer-independent world “out there” about which a researcher may systematically discover true facts and reject false ones. Research that does not share these broad assumptions is outside of what is and ought to be distinctive about any social science as a science.
If these fundamental methodological and epistemic objections to science are true—if there is no reality “out there” or if we really can never know anything about it—then science cannot be done. But that conclusion would call for a far more radical reorganization of all our disciplines beyond any argument about whether what anyone does belongs as part of a scientific enterprise. In any case, we need not take a position on whether or not science is possible; whether, that is, there does in fact exist a world “out there” about which we are trying to know things. The activity of science proceeds as if an observer-independent world exists. If a scholar's work is not one of discovery of that world, but creation or imposition upon it, then it is not science, whatever else it may be.
Falsification
If science presumes the existence of an observer-independent world about which we are seeking to discover true things, any scientific claim we might make about that world should be falsifiable by contradictory evidence gained from reason or experience.Footnote 59 This is also an assumption of science, one that follows from there being an observer-independent world. I am not defending falsification as good, bad, or even coherent as a procedure. And, as Popper noted long ago, “falsifiable” does not mean that any particular claim can currently be tested or falsified, just that it is in principle possible to do so. Because of the technical apparatus required for testing, some of Einstein's most provocative claims were falsifiable but untestable until almost a century after he made them.Footnote 60 What this means is that, for any claim to count as scientific, we can specify what would demonstrate that the claim is wrong.
The emphasis on observer-independence, and the requirement that claims be falsifiable, as assumptions of science may strike theorists as simply old-fashioned positivism that many thought was discredited.Footnote 61 But this is not the case. While empirical positivism begins by assuming an observer-independent world about which we are seeking knowledge through falsification, it limits how we go about gaining that knowledge to empirical means alone. But logical positivism extends to the logic we use to falsify our hypotheses. If I claim X, whatever X is, and it necessarily implies its opposite, I need not make a single empirical observation to have falsified the claim. Whether or not this caricature of positivism is coherent (as even Carl Hempel acknowledged, non-empirical “theory” structures and categorizes our empirical observations (Hempel Reference Hempel1966)), it asserts too much for our present purpose. For example, logical positivism excluded the study of moral values from a proper scientific study. Yet if moral values, like, say “natural right” exist independent of their observer, and if the researcher's aim is to discover (not create) something that is true about those rights, then such inquiry would be consistent with the presumptions, if not the current practice, of science, so long as they were subjected to principles of logical analysis, and thus should be included.
The centrality of falsification for doing science should also not be overstated in the way that Popper did by rejecting supporting evidence as being of no consequence to a theory's being true. Like David Hume, Popper argued that the problem of induction was epistemically insurmountable; his view (and Hume's) was that we have no basis to presume that laws governing nature will continue to operate in the future as they did in the past. After all, the only reason we could have for presuming that they will is that in the past, laws that operated in the past continued to operate in the future—yet this assumes the conclusion we would be seeking to explain. But my view here is simply that a theory or claim that cannot be falsified in some way fails to be a scientific theory or claim, whatever else it might be, or whatever else is required for it to count as scientific.
Methodologically and culturally, the presumption of falsification marks a rough and ready distinction between the humanities and sciences with regard to the kinds of errors their practitioners are willing to make. Because of the process of testing, science resists novel claims about the world. By contrast, the humanities are primarily concerned with creative expressions and interpretations that construct an appreciation for the distinctively human features of our world. We might even say that science is designed to systematically err on the side of rejecting claims that turn out to be true, and the humanities are designed to systematically err on the side of accepting claims that turn out to be false. Thus do the humanities often seek evidence to support an argument, and emphasize its coherence and persuasiveness, rather than trying to falsify it, or articulating what the conditions of coherence or persuasiveness are or should be.Footnote 62 The difference in method corresponds to whether a person would, at the margins, prefer to avoid believing false claims, or prefer to believe true ones.Footnote 63
Thus, the humanities are distinctive by their use of subjective interpretation to bring meaning to the world rather than to discover objective truths about it.Footnote 64 This humanistic orientation may arise out of the view that observer-independence is impossible, or that falsification is “verification” by another name, or that pretending to either one introduces systematic biases. But I do not here take a position regarding the merits of humanistic and scientific research. I would not be surprised if a complete understanding and appreciation of the social world requires both social-scientific and humanistic research.Footnote 65 But I presume that these are different sorts of activities, and it is a presupposition of social science that they are not the same either.Footnote 66 To adopt an observation Rousseau made in a different context, the humanities may persuade without convincing, even as the sciences convince without persuading.
Drawing Lines
We are now in a position to use our criteria of demarcation to test what sorts of research properly belongs in political science, asking the following questions:
(1) Does the main subject of the research concern the use of power over people?
(2) Does the research assume that there are facts about a world “out there” to be discovered, and is it part of an enterprise to systematically discover these facts?
(3) Can the findings of the research be shown to be mistaken in some way?
Given these questions, where does political theory fit?
One way to proceed would be to draw examples from the literature and demonstrate why this study would count as political science, and that one would not. Instead, I have chosen a more categorical approach, speaking about general types of political theory rather than specific examples, in order to focus on the conceptual core of the argument. Naming names would be provocative and distracting at the same time, for to proceed that way we would not only be asking if a type of work belongs, we would simultaneously be asking whether a particular example used belonged to that type. I admit the failure to offer examples throughout what follows may lead to the impression I am creating so many straw men, but I could not justify any particular case without far more development than I have the space for here.Footnote 67
Turning then to a categorical approach, we begin by noting that “political theory” properly denotes a wide range of very different types of scholarship:
Conceptual political theory: research about political concepts.
Normative political theory: research about the norms we ought to endorse about the use of power, and/or the way that power and resources ought to be distributed based on those norms.
Advocacy: research that promotes social and political change.
Explanatory political theory: research that offers causal accounts of political events.
Interpretive political theory: research that makes claims about the meaning of political events to those who participated in them, and/or that offers the researcher's interpretation of those events.
History of political thought: research that traces the development of an idea through time, and/or that offers an interpretation of what an author meant in writing what he did.Footnote 68
We will treat each in turn, grouping them when appropriate.Footnote 69
Conceptual and Normative Theory
To the extent conceptual and normative researchers seek to discover something about how power is, or ought to be, used over people, in a way that presumes discovery of a world independent of the observer, animated by a method that systematically favors the rejection of false claims, this kind of research falls well within the bounds of political science.Footnote 70 For example, conceptual analysis that focuses on core political concepts like power, domination, citizenship, etc—not the words “power,” “domination,” and “citizenship,” but the underlying ideas that they denote—would fit to the extent such research seeks to discover observer-independent facts about the political world, in just the same way that similar conceptual analysis is critical to the study of physics and other natural sciences, because these concepts are central to the objects of study. Similarly, normative theory that explains which institutions and practices are consistent with widely shared norms, or treatments that attempt to discover moral truth (rather than create and impose it) are also consistent with the broad aims of the social sciences and, as I noted above, are regularly falsified through logical errors or the use of counter-examples.
The importance of normative and conceptual analysis on their own terms to a scientific knowledge of politics can best be illustrated in the breach, where a failure to get a concept or norm correct leads to mistakes about our understanding of the political world. To extend an earlier example, equating “legitimacy” with “public approval” presents an inaccurate description of the political world. The concept of legitimacy—again, not the word itself, but the underlying idea—is a normative concept which concerns whether a regime, institution or individual has the right to the power they may possess. As such, public approval may well be necessary for legitimacy, but it is hardly sufficient.Footnote 71 We might nevertheless define “legitimacy” as public approval. We might also define “legitimacy” as the successful mating practices of bears, and go about offering analysis of the legitimacy of institutions accordingly. But if we did, we would misdescribe the political world (even if doing so might be the premise of a terrific new reality television show). Our understanding of the political world in this case would be limited not by the empirical analysis of the terms as described, but by the fact that our conceptualization of the underlying political phenomenon is simply mistaken. Getting political concepts right is thus part of political science properly conceived.Footnote 72
Now of course the fact that one view of the concept of legitimacy, let alone of its operationalization, conflicts with another view does not tell us which is the correct one, even if we are pretty sure that identifying legitimacy with the mating practices of bears is just plain wrong. But we can imagine a critical test: if public approval were conceptually the same as legitimacy, then Hitler and Saddam Hussein would have in fact been legitimate rulers; slavery would have in fact been a legitimate institution; the colonization of a native people anywhereFootnote 73 would have in fact been legitimate. This is where normative theory is helpful in understanding the political world as it is: such cases illustrate that the way things are (“the public approves”) does not necessarily correspond with the way things ought to be (“the right to rule.”)
When we find such spectacular and revolutionary results as “slavery was a legitimate institution,” we might indeed have to revise our identification of “public approval” with “legitimacy,” because we come to realize this is not an accurate reflection of the political world. Or, we might go the other way and realize that it does reflect the political world correctly, and our concept of legitimacy is in need of revision. Most people would agree that any analysis in which American chattel slavery was legitimateFootnote 74—not just “publicly approved” but legitimate—would be odd, and a confusion of the very object that is purportedly being studied.Footnote 75
In both of these cases the normative and conceptual accounts of legitimacy might be wrong, and their empirical treatments right. But without any account of how the normative aspects of the concept obtain our scientific understanding of the object of study is mistaken. Concepts and norms are thus not merely useful to our understanding of politics, they are themselves part of the political world and the phenomena that political science, rightly conceived, ought to study. So long as the research is done in a mode of discovery and in a manner that can be falsified, it would count offensively as a member of the discipline of political science. This conclusion also survives the political science exceptionalism test, since it applies with no more or less force to research into concepts and norms in the cultural and economic worlds: to use an earlier example, a study that simplistically equates “value” with “economic gain” is measuring something, but it obstructs a complete scientific understanding of the social phenomenon of exchange.
A related and more common species of normative political theory takes up the implications of a set of norms (usually, but not necessarily, “widely endorsed norms”) in order to generate conditional norms of the kind “if equality means x, then institutions should be designed the following way …” So long as our focus is on the use of power, and those implications from first premises could be shown to be false (through logical contradiction or empirical evidence), this kind of conditional normative theory fully qualifies as political science. Indeed, it proceeds precisely the same way that mathematics does, with the added benefit (from the standpoint of membership in political science) that it takes political phenomena as its object of study.Footnote 76
Advocacy
We turn now to “advocacy” as a mode of political theory closely related to, but not the same as, normative analysis. Advocacy, as I conceive of it here, is concerned with using power, and advocacy research is primarily concerned with effecting change, rather than seeking to know things about the world as such. Advocacy as a mode of political inquiry has a history that goes back at least as far back as Marx; more recently, it was at the center of Sheldon Wolin's seminal “Political Theory as a Vocation,”Footnote 77 and it continues to be the core of New Political Science's goal to restructure the discipline. Consider this claim by Sanford F. Schram:
A political science that forgoes the dream of a science of politics in order to dedicate itself to the enhancing the critical capacity of people to practice a politics is, for me, an exciting prospect. A political science that does this to enhance the capacity to challenge power from below is all the more exciting. I would argue that the new political science would be not just more politically efficacious but also more intelligent, offering more robust forms of political knowledge.Footnote 78
Schram here explicitly connects advocacy to the pursuit of knowledge, acknowledging that were political science to turn fully towards this sort of advocacy, it would offer “more robust forms of political knowledge,” an end that would fit within political science as I have depicted it.
The problem, though, is that science as a distinct and worthwhile enterprise is and ought to be primarily concerned with the acquisition of knowledge, not the promotion of change or preservation.Footnote 79 There may be a close relationship between advocacy and normative research: advocates might rely on normative research to guide how they change the world, to issue calls to action, and promote activity to change entrenched and unjustified political structures; normative theorists may also be motivated by a real desire to promote, say, justice and social change, but their work is dedicated to knowledge of what to do, not the actual doing. To the extent a work qualifies as “advocacy” in this latter sense, it is not political science, because it is not seeking to know things about the political world, it is seeking to change that world.
The point that the acquisition of knowledge is different from the use one makes of it should not be mistaken for the view that social and political problems have, or should have, no bearing on what goes on within science, particularly in the choice of what problems one takes to be important. This is a point about agenda setting that I took up in the second section above. Of course there might be very good reasons to spend one's life coming to know about how power is systematically abused, rather than focusing on some trivial problem, and to endorse a problem-centered approach to the study of politics. But our concern here is one of membership: whether a question is trivial, useful, or meaningful does not speak to whether it should count within the discipline as political science per se. Neither the choice of whether a problem is good or important to study, nor the use one makes of that knowledge, is political science properly conceived.
Explanatory and Interpretive Theory
We move now to explanatory political theory, research that offers an explanation of the causes and mechanisms of political phenomena. So long as it is consistent with the two assumptions of science about an observer-independent world and falsifiability, this work is unproblematically political science. Most empirical studies are employed on behalf of some explanatory theory, but perhaps the most familiar form of this genre is positive political theory—which ceases to be political science when the accounts it offers cannot be falsified.Footnote 80 Similarly, interpretive theory that focuses on the social significance of events or ideas to individual actors as those actors view them is political science as here conceived, because it interprets what happened in the political world by reference to the significance and meaning of events or ideas to a set of people. This is work that can be falsified, because the claim that symbol x meant y to some person z who therefore performed some action a may turn out to be false, by analysis of the data.
It would indeed be very helpful to know whether certain words, signs and other signifiers (like complex events) meant something to those who encountered them, because these signifiers presumably have some causal influence on their observers. Interpreting a political advertisement as an expression of racism because it shows a shady and disheveled African-American man leaving a prison cannot proceed simply by listing the features of the ad. What is necessary is an account of why those images have the power they do over the observer: they play into conventional stereotypes and thus may have the causal effect of influencing voters for presumably bad reasons (a hypothesis that is falsifiable). The fact that social events require explanation by reference to how participants interpret them should be obvious, if only because they are inexplicable by a mere cataloging of their brute features.Footnote 81 Thus, interpretative accounts of the underlying significance of events to participants are fully within the ken of social science.Footnote 82
By contrast to interpretative theory that explains the meaning of events to participants, a second form of interpretive theory purports to read social events as texts, explaining the meaning of them to the researcher, much as someone offers a literary analysis of a work of fiction. Interpretation of this kind does not belong in political science because it is a creative imposition of a view onto the world that cannot be falsified in any way, it does not belong with political science.
Defenders of the role of interpretation have perhaps been among the most vocal in rejecting a social science based on the model of the natural sciences because, it is said, human agents, unlike material objects in the natural world, bring meaning to the activities in which they engage. And on the view of social science I present here, if an analysis does not consider what an action meant to those who engage in it, then it is incomplete.Footnote 83 This is where interpretive social science needs to offensively state its case for being included in political science, and, so long as we are explaining what an event in fact meant to others (and not merely reporting our own view) it is a very strong case indeed. But when an interpretation reduces social events to texts waiting for the interpreter to impose meaning on them, it ceases to be science, however meaningful and important it might be otherwise.
Textual and Historical Political Theory
Another prominent mode of political theory takes up written texts and interprets them to demonstrate what an author's intention was, or to interpret what the text really means. By the criteria here advanced, this kind of work does not qualify as political science, for two reasons. First, what a text really means, or what an author's intentions were, as such—the concerns of a good deal of this literature—has nothing with the use of power over people. Second, these treatments often do not presume observer-independence or falsifiability. Indeed, as I suggested above, it is telling that such interpretive treatments are sometimes praised for being “convincing” or “persuasive,” rather than being “true” or “correct.” Work of this sort, even if it does concern the use of power over people, thus fails to meet the second and third criteria for inclusion. It may be of great value in and of itself to an appreciation of all that is human, but it belongs in the humanities.
If one still believes that interpretive work that takes up questions like “What did Sophocles intend to do by writing Antigone?” belongs in the social sciences, we must subject that claim to the test of political science exceptionalism. Do we think that it would be epistemically appropriate for economics as a discipline to take up the question, “What does The Republic tell us about what Plato thought about the exchange of property?” Do we think that it would be epistemically appropriate for anthropology departments to sponsor research on Aristotle's view of culture? These are not necessarily decisive questions—maybe we think that this kind of work would benefit from placement in their corresponding social-scientific disciplines, rather than in disciplines that would provide better tools in historical and textual analysis. And since such work may be of great use to those who actually study these core ideas of property and culture, maybe as a pragmatic matter it would be useful on a case-by-case basis to sponsor and incentivize this kind of research (a point we will return to in the next section). But these examples are meant to keep us honest: we cannot in principle defend the study of what Plato thought about politics as belonging in political science without endorsing such work in anthropology or economics when it concerns an author's conception of culture or exchange.
Another variety of the history of political thought traces how ideas have shaped the political world through time. This sort of history, so long as it proceeds in a manner consistent with the assumptions of science, is unproblematically political science. By this view, scholarship that demonstrates the real causal influence of Locke's ideas on the use of power in the seventeenth or eighteenth century, or the influence of behavioral economics on the policies of the Obama Administration, would clearly count, so long as it focused on how ideas shape the manner in which power is used. And our test of political science exceptionalism is again useful here, since similar work, to the extent it focuses on getting at our understanding of exchange, would fit well within economics. Indeed, this kind of history of political thought is best categorized as another species of explanatory political theory, using ideas as the causal mechanism that explains political events. Still, we should be very wary of the dogmatic assertion that “ideas matter!” as a replacement for the connective tissue that would form the actual explanation. It is not that ideas don't matter; it is just that whether and how they matter is precisely what needs to be demonstrated and explained, not asserted, in order for that work to be political science, properly conceived, rather than a form of intellectual history.Footnote 84
Finally, what of the argument that the history of political thought helps to resurrect very different ways to think about politics, from institutional arrangements no longer supported to normative ideals long out of fashion? Such work can help others understand contemporary politics, but that, as we saw earlier, is only a second-order justification for inclusion in a discipline—and a rather tenuous one at that. Political scientists will and ought to engage the history of political thought when appropriate, and there is value in bringing political science and history together, each on its own disciplinary foundation. But that does not justify including history of political thought as a field of political science properly conceived.Footnote 85
Conclusion and Practical Applications
This article is motivated by the ongoing lack of engagement, suspicion, and even animosity between theorists and non-theorists within political science, most recently evidenced by the turn of events at Penn State. Whether the protest against the elimination of political theory was motivated by a sense of shared community, a worry that other departments would soon follow Penn's lead, or something else entirely is hard to say. But instead of offensively explaining why political theory is political science properly conceived, the letters threatened counter action, and defended theory's inclusion without getting to what I believe was the central question posed by Penn State's actions: why does political theory belong?
I have offered one answer to that question: political theory belongs in political science, because it is political science. I do not suspect readers will quickly accept my view of what the proper criteria of political science are or ought to be, or my classification of different kinds of theory stylized as they were. But this is the kind of work that I believe is necessary to explain why and what kind of political theory belongs with political science.
Although I think my proposed criteria of demarcation for political science are justifiable, they are not the only possible ones, and they may well be incorrect or of no use. I know, in fact, that many theorists do not agree with them, for reasons that are well-developed and defended in the philosophy of social science, and to some of which I have referred above. I have tried to avoid such debates by arguing conditionally—if we are going to do science, then we must accept its assumptions—rather than from a view that such assumptions are in fact ontologically correct. Still, this argument about assumptions is itself an argument others may not accept.
I want to be clear that I do not think that some centralized body should try to enforce these or any other criteria in allocating disciplinary resources. For one thing, the categorization of individual research will, at the margins, be contentious (this is why I did not develop specific examples or name “names” in the fourth section). Retaining local application of various criteria would also provide a good pragmatic check on the correctness of any one set of them. Departments should always be able to judge on a case-by-case basis whom to hire, and indeed some of the most interesting departments in the discipline have achieved their status by hiring without reference to firm boundaries, explicitly preferring to foster cross-fertilization.Footnote 86
But the fact that the development and application of criteria should proceed on a local level does not mean that we should abandon justification altogether. As I said at the start, the gatekeepers of the discipline in whose hands rests so much power to shape individual lives ought to be more reflective and explicit about how their own often implicit and unreflective criteria of demarcation guide their decisions. As a matter of justice, we should expect a consistent application of criteria over cases within their domain, and explanations when they choose to deviate from them. And just as we avoided “political science exceptionalism” in our determination of the proper conceptual bounds of the discipline, we must avoid “political theory exceptionalism” in their application. When members of the discipline make the claim, “That's not political science,” we ought to ask them first to back it up by stating their own criteria of demarcation, whether or not they are the ones offered here, and then use their own subfields to test their conclusions about what would and would not fit.
As I have said throughout, the argument here poses a problem for including political methodology as a first-order member of the discipline. By the arguments here, graduate students in political science would likely do better studying with first order statisticians than second-order political scientists, just as many theorists who study Tocqueville would not learn well how to read French from a political scientist. But if on occasion, political science departments need to employ “in house” statisticians (or French linguists) in order to keep their focus on methods relevant to the political problems they study, they might justify using their resources to be sure that work gets done. We just should not confuse that for the argument that such work is political science; and if others think it is, then political theorists are on much stronger ground than they imagine.
Given that I have argued for a strengthening of separate disciplinary strands within political theory, readers might understandably conclude that I oppose the recent trend to expand interdisciplinarity within the university. That is not the case. But there are two ways to understand “interdisciplinarity”: The first treats it as a way to get rid of, or around, current disciplinary training and replace it with some new set of disciplines. This is actually not so much “interdisciplinarity” but rather “new-disciplinarity,” in which, for example, philosophers, neuroscientists and psychologists (PNP) have gathered to work out overlapping problems of interest, and then establish distinct training and faculty lines as they start new programs around those problems. Such an approach simply replaces old disciplinary boundaries with new ones, and, though at its inception it may involve interaction between different disciplines, it would lose that interdisciplinarity over time.
The second, and I think far more promising, way of understanding and valuing interdisciplinarity is as a way to strengthen existing disciplinary distinctions, bringing people who are trained in very different core disciplines together for cross-fertilization.Footnote 87 This kind of interdisciplinarity is of great value, because each participant brings to the table a different set of methods and questions that arise in their core discipline. But it presupposes core disciplinary training among its participants to create something of greater value. Indeed, a commitment to this kind of interdisciplinarity requires the kind of offensive argument I am offering here to strengthen and focus the development of the core disciplinary work upon which it depends.
Furthermore, we do no service to political theorists by training them poorly in methods that they need to do their work. We can note anecdotally, and with the risk of “naming names,” that among the best practitioners of various modes of political theory, those who are at the top of their sub-subfield received their initial training in disciplines appropriate to their task: John Rawls, Charles Mills, Charles Taylor, and Iris Marion Young in philosophy; Jürgen Habermas in philosophy and sociology; William Riker in political science; J. G. A. Pocock and Quentin Skinner in history; and Martha Nussbaum in classics. This is as it should be: Historians should get proper training in historical methods. Literary critics ought to be trained in the most sophisticated tools of textual analysis and interpretation. Theorists, whose work is firmly grounded in discovering observer-independent truths about politics, ought to be trained in the scientific study of politics.
Interdisciplinary programs and activities should thus be structured not to obscure disciplinary foundations, but to strengthen and build upon them. The recently-formed Association for Political Theory, that aims to include political scientists, philosophers and historians, or the older American Society for Legal and Political Philosophy, that includes theorists from political science, philosophy and law, are exemplars of this kind of activity.Footnote 88 There are formal and informal interdisciplinary political theory programs that seek to build on rather than eradicate disciplinary training at a range of institutions that provide different examples of how this might be arranged institutionally. Given the strong tradition of advocacy and social change in political theory, these kinds of interdisciplinary activities might be most appropriately housed in professional schools of public policy, law, or social work—schools that have long traditions of hiring across disciplines and that often have aspirations towards advocacy as well.Footnote 89
One of the great intellectual attractions of political theory is its broad range. A theorist can read Hobbes in the morning, make conceptual distinctions in the afternoon and issue a call for action before retiring for bed. I do not think all of this is political science properly conceived. But with the criteria of demarcation I am proposing, theorists would have a strong argument to explain why some of it is, and for why, as a matter of justice, disciplinary gatekeepers should reflect on the criteria that they use to allocate the resources they have at their command. Until theorists stop defending what they do, cease reacting to events with poor arguments for inclusion, and start making offensive claims to membership, our future in the discipline will continue to feel precarious, whether or not it actually is.
Supplementary Materials
Foundations of Political Theory Letter http://journals.cambridge.org/pps2010013
Written by Ben Barber and signed by Sixty Others http://journals.cambridge.org/pps2010014
