This file is also available in Adobe Acrobat PDF format
FEAR OF NOT FLYING
IN MEDIEVAL ENGLAND there was a curious gap between the study and practice of law. From the thirteenth century to the seventeenth, the main language used for pleading in common law courts was Law French. It seems to have developed because Latin, the language of formal records, carried too much historical freight from Roman law for the peculiarities of English circumstances, whereas medieval English was insufficiently standardized for official use. Law French was a hybrid dialect, owing more to Picard and Angevin influences than to Norman French, in which French vocabulary was combined with the rules of English grammar. The common lawyers developed it for their pleadings in the courts, passing it down from generation to generation.1
This evolving vernacular of the common law courts had little, if any, impact on the academic study of law. Latin was the language of jurisprudence in Oxford and Cambridge, and, although Law French could reportedly be learned in ten days or fewer, the professors of jurisprudence appear not to have thought it worth their while. This might have been, as Fortescue said, because Latin was the language of all scientific instruction.2 It might have been, as Blackstone claimed, because the civil law--taught and studied in Latin--was embraced in the universities and monasteries after the Norman Conquest but resisted in the courts.3 It might have been, as contemporary historians such as J. H. Baker maintain, because English Law was thought insufficiently cosmopolitan to merit serious study.4 Whatever the reason, English jurisprudence developed in literal ignorance of the practice of English law.
A comparable disjunction afflicts the human sciences today. In discipline after discipline, the flight from reality has been so complete that the academics have all but lost sight of what they claim is their object of study. This goes for the quantitative and formally oriented social sciences that are principally geared toward causal explanation. Following economics, they have modeled themselves on physics--or at any rate on a stylized version of what is often said to go on in physics. But it also goes for many of the more interpretive endeavors that have been influenced by fashions in the humanities--particularly the linguistic turn in philosophy and developments in literary hermeneutics. Practitioners in these fields often see themselves as engaged in interpretation rather than explanation, thereby perpetuating a false dichotomy. Hence my use of the term human sciences here to encapsulate both endeavors. This book is my attempt to chronicle the extent of their flight from reality, and to combat it.
I should say at the outset that I do not believe the flight from reality has a single source or cause. It results, rather, from various developments that share elective affinities--developments that all too often are mutually reinforcing. Some of their sources are intellectual, having to do with the ebb and flow of academic fashion. Some of them are institutional, reflecting the structure of academic professions and the incentives for advancement in an era of exhausted paradigms and extensive specialization. This can be bolstered by a perverse sense of rigor, where the dread of being thought insufficiently scientific spawns a fear of not flying among young scholars. Some are political in the broadest sense, having to do with the relations between disengaged human sciences and the reproduction of the social and political order. The flight from reality is not without consequences for reality as we will see. At best it marginalizes the potential effects of political and social criticism, and sometimes it contributes to the maintenance of oppressive social relations--however unwittingly.
I begin making this case, with the help of Alexander Wendt, in the opening chapter. We expose the limitations of empiricist and interpretive methods of social research, showing how they bias the enterprise in method-driven ways, and we argue for a realist view in their stead. Rather than do this in the abstract, we pursue it by reference to a concrete phenomenon that has attracted a good deal of attention in the human sciences: the study of consent. Empiricism as we describe it here encompasses two different approaches to social inquiry--both bastard stepchildren of David Hume. The first we dub logicism to call attention to the fact that its proponents embrace the view, made famous by Carl Hempel, that good explanations are sound deductive arguments. For logicists, an hypothesis is scientific only if it is derived from a general theory. Such theories often rest on simplifying assumptions about reality, or even "as if" assumptions that are not valid empirically at all.
It is conventional to defend this practice on the grounds that these theories do a good predictive job in accounting for empirical reality. This might sound reasonable in principle, but in practice logicists often formulate their claims so generally that they turn out to be compatible with all possible empirical results--in effect rendering the empirical world epiphenomenal to the theory. We show how this vitiates the study of consent in practice, when theorists of rational consent have sought to explain away apparent anomalies by concocting redescriptions that render them compatible with their preferred theories. This was understandable with theorists of the early Enlightenment because, as I have argued elsewhere, they embraced the view that certainty is the hallmark of science--making the toleration of counterexamples unacceptable.5 However, mature Enlightenment views of science assume knowledge generally to be corrigible, with the implication that scientists will not take hypotheses seriously if they cannot be falsified--and, indeed, if the conditions under which they will be rejected cannot be specified in advance. The irony, then, is that although contemporary logicists often like to think of themselves as the only genuinely rigorous practitioners of social science--"if you ain't got a theorem you ain't got shit!" as a partisan of this view once put it to me--in reality they are wedded a view of science that most practicing scientists have not taken seriously for centuries.
Hume's other bastard stepchildren discussed in chapter 1 are empiricists of a particular stamp--those who became known in the 1950s and 1960s as behavioralists. Often skeptical of the breathtaking theoretical ambition characteristic of logicists, these empiricists were partisans of Hume's insistence that knowledge is grounded in observation of events and that causal knowledge inheres in observing their constant conjunction. Whereas the logicist derives comfort from the certainty that seems to inhere in the deductive relations between premises and conclusions, the Humean empiricist looks to observation for reassurance. We show how, in the political science power literature, this biased research away from attending to factors that coerce people into apparently consensual behavior--whether by surreptitious manipulation of agendas, structuring people's perceptions of alternatives, or even shaping their preferences. The focus on observed behavior was conceived as a corrective to elite theories of politics put forward by Gaetano Mosca, Robert Michels, Vilfredo Pareto, and C. Wright Mills. They had been cavalier in their treatment of observable behavior, ignoring it or explaining away any tensions between it and what Michels described as the "iron law" of oligarchy. Ironically, the behavioralists ended up with a different kind of method-drivenness, one in which the realm of observed behavior was assumed to be the only pertinent realm in accounting for consent.
If behavioralists bias research in favor of the phenomenal realm in one way, partisans of interpretation do it in another. By interpretivism I mean the research agenda that came into vogue in the human sciences during the 1970s, largely because of dissatisfaction with various failed reductionist enterprises. Prominent among these was Marxism, by then famous for its inability to account for the major political developments of the twentieth century. Instead of spawning revolutionary socialist proletariats, the advanced capitalist countries had experienced tenacious nationalism and working-class conservatism--not to mention the rise of Nazism in Germany and Fascism in Italy. Communists who did come to power either did so in peasant societies such as Russia and China contra Marx's prognostications, or they were forcibly imposed by the Soviets after World War II. In any case, by the 1950s it was obvious that, at best, communism as practiced rested on grotesque distortions of Marx's principles.
Attempts to rescue Marxism from an unhelpfully recalcitrant reality produced more noise than light over the succeeding decades. One motivation for those who found the interpretive turn attractive was to get away from the sectarian bickering over how to save Marx's materialism--even when it was conceded that this could operate only "in the last instance."6 The interpretive turn involves treating articulated beliefs and ideas as elemental to human interaction. They are seen not as part of an epiphenomenal superstructure, to be understood, however circuitously, by reference to its links with the "underlying" material base. Rather, to use one of the buzz words of the day, they constitute reality--or at least human social reality--through language. Social reality is linguistic reality on this view. When human beings do things like create obligations or social contracts they do this through language, not by some other means that is then described by language. Understanding social reality means understanding the linguistic processes that give rise to it.
The interpretive turn thus went hand in glove with the ascent of ordinary language philosophy associated with the later Wittgenstein and J. L. Austin in the 1950s and with developments in literary hermeneutics in which understanding social processes was modeled on the interpretation of texts.7 It was but a small step from this to the view that society should be conceived of as a text, whose meaning is best recovered by exploring the web of linguistic conventions within which social agents operate as collective authors. We are locked within a prison-house of language, as Frederick Jameson colorfully put it, the implication being that it is better to try to understand linguistic reality from the inside than to indulge vain fantasies of escape.8 Different theorists had different views of how such understanding is best achieved, but they all agreed that the point of the exercise is to elucidate social meanings by exploring the linguistic conventions--the language games, as Wittgenstein had it--within which people inevitably operate. Social reality arises out of conventional linguistic usage, and the key to understanding it lies in recovering the conventions so as to see how people use them to act in the social world.
Elsewhere I have discussed the interpretative turn's impact on the historical study of political theory by examining the contextual theories of the Cambridge school--John Dunn, J.G.A. Pocock, and Quentin Skinner.9 There is much to commend their approach to the study of the texts in the history of ideas. In particular, their insistence that contextual knowledge is essential to recover what an author meant to do in writing a text was an important corrective to prevalent methodologies that had assumed reading the text "over and over again" to be sufficient. Some of their contextual rereadings of particular authors are debatable and have been debated, but one would be hard pressed to dispute that important correctives to received interpretations have resulted from this scholarship. Skinner's rereadings of Hobbes have stood the test of time especially well--displacing a tired stereotype of him as "the monster of Malmesbury."10 Dunn's relocation of Locke's political writings in the theological disputes that were his lifelong preoccupation have revolutionized Locke scholarship for a generation, and the careful contextual researches of Peter Laslett and Richard Ashcraft have established that the Two Treatises of Government were written the better part of a decade before the Glorious Revolution of 1688--rubbishing an older conventional wisdom that they were written to justify it.11 Locke's contradictory views on slavery have received definitive illumination through the contextual analysis of James Farr.12 Pocock's magisterial recovery of the civic humanist tradition has spawned a revival of interest in republican ideas, complicating, at least, our picture of liberalism's emergence and evolution.13 This is to say nothing of the revisions of received interpretations of medieval and early modern natural law theory at the hands of Richard Tuck and James Tully, or the accounts of Adam Smith and David Ricardo's politics from Donald Winch, Shannon Stimson, and Murray Milgate.14
It is one thing to say that understanding what an author was trying to do depends critically on recovering the context in which he was writing; quite another to turn this into an a theory of politics and political change. It is this vastly more ambitious agenda, most self-consciously articulated by Quentin Skinner, with which I take issue. I agree with Skinner that any plausible account of political reality must take account of the role political ideas play in shaping it. But making this move inevitably puts large causal questions on the table about what ideologies are, how they shape and are shaped by political conflict and change, and how--if at all--they might be related to the ideas of political theorists.
Skinner ducks these questions by eschewing all causal analysis in favor of the "interpretation," but I argue that in effect this means he does his causal analysis behind his back, which he insists that we should not. He equates the meaning of a text with what an author intended to convey, and he gets at this by seeing how the author's ideas were received by his intended audience. But this overlooks gamuts of relevant possibilities once we are studying their ideas as ideologies. What people overlook might be more important, ideologically, than what they discern. People might be misled, whether for malevolent or accidental reasons. They might supply inadvertent legitimation for practices that they perceive dimly, if at all. How people's ideas are appropriated or misappropriated by subsequent generations might be more important than their intentions as communicated to contemporaries. By assuming that an "internal" reading, geared to recovering authorial intention, is synonymous with studying the history of ideas as the history of ideologies, Skinner affirms a new--rather whiggish--reductionism without ever acknowledging it.
In contrast, I argue for openness to "external" readings. These are geared to locating subjective accounts in larger causal processes without prejudging what those processes might consist in, without deciding in advance whether and how much they might shape or be shaped by political interests, agendas, and events, and without assuming anything a priori about how--if at all--they might be subsumable into a general theory of politics. These are all subjects for research that cannot be settled before it begins. The scientific outlook requires a commitment to discovering what is actually going on in a given situation without prejudging what that is. Opting for the recovery of what a particular political theorist meant to say involves one of many possible cuts at accounting for ideology's role in politics. It has to be justified by comparison with the going plausible alternatives, not smuggled in by the backdoor under the guise of eschewing the world of causation for that of interpretation. Partisans of interpretation often see themselves as fundamentally at odds with behavioral social scientists. So it is ironic that they end up embracing a reductive view that makes them cousins of the behavioralists. Both rule out looking behind the world of appearances. This biases the study of consent by taking some of the most significant possibilities off the table before research begins.
My call for attention to the "external" dimensions of political action as well as the "internal" ones is embedded a realist view of science. It owes much to the work of Rom Harré, Roy Bhaskar, Richard Miller, and others, and I harbor no ambition to offer a full-blown defense of it here. Rather, my goal is to underscore the commitments that are embodied in the realist outlook, explain their significance for the conduct of social inquiry, and show why they should be expected to lead to better results than the going alternatives. To be sure, I mean to portray the realist outlook in an attractive light in these pages, but I try to do this more by illustrating its felicitous consequences for social and political inquiry than by arguing for it from the ground up. Some prefatory remarks are nonetheless in order here to indicate what I take to be involved in the commitment to scientific realism, and to differentiate it from doctrines with which it is sometimes confused.15
I take the core commitment of scientific realism to consist in the twofold conviction that the world consists of causal mechanisms that exist independently of our study--or even awareness--of them, and that the methods of science hold out the best possibility of our grasping their true character. Adherents to this view are sometimes characterized as "transcendental" realists.16 This cumbersome and loaded term perhaps obscures more than it illuminates. I take it to underscore the fact that the realist commitment is implicit in the conduct of science, not a product of it. Unless scientists assumed it to be valid, as they generally do, they would have no good reason to see their enterprise as superior to religion, superstition, tradition, and other pretenders to authority in accounting for reality. This is not to say that the realist commitment implies fidelity to any particular theories or hypotheses about reality's causal structure. Rather, embracing the commitment is necessary for thinking it worthwhile to develop theories and hypotheses, and to evaluate them by reference to the methods of science.
Wendt and I show how, in the study of consent, a realist commitment opens up research agendas to the study of causal questions that are ruled out of court by the behavioral and interpretivist schools. Yet it does this without dismissing behavior and subjective understanding as epiphenomenal, or affirming a reductionist view that is impervious to the demands of evidence so characteristic of logicist ventures. We discuss John Gaventa's Power and Powerlessness as exemplifying social science conducted in a realist spirit, both in its attempt to illuminate opaque causal mechanisms that produce consent--"quiescence" is Gaventa's term--in circumstances of domination, and in explaining the relations between those mechanisms and the realms of subjective perception and behavior. From a realist perspective it thus becomes plain that behavioral and interpretive methods exemplify the flight from reality, even though their proponents often resist the ambition to develop general theory.
Perhaps as a reaction against the behavioral and interpretive hostility to general theory, logicist enterprises have won a new lease on life in recent decades. The main vehicle has been the import of microeconomic models into the noneconomic human sciences--notably to political science, sociology, and law--under the banner of rational choice theory. Donald Green and I explore this development as it relates to political science in chapter 2, by responding to critics of our book Pathologies of Rational Choice Theory.17 Although logicist ventures are wanting from a realist perspective, we show here that they also collapse under their own weight--largely because of their quixotic theoretical ambition. Taken on their own terms, rational choice theories have, for the most part, degenerated into elaborate exercises geared toward saving universalist theory from discordant encounters with reality. Belying the fanfare about theoretical rigor that often accompanies their claims, we show how rational choice theorists play fast and loose with the definition of rationality in developing hypotheses, in specifying their empirical implications, and in testing them against the evidence.
The litany of failures that we identify includes elaborating sufficient accounts for political phenomena without showing how or why they should be preferred to the going alternatives; "explaining" stylized facts that turn out on close inspection not to bear much relationship to any political reality; post-hoc fiddling with theories in ways that amount to little more than thinly disguised curve-fitting; specifying theories so vaguely that they turn out to be compatible with all empirical outcomes; scouring the political landscape for confirming illustrations of the preferred theory while ignoring the rest of the data; and projecting evidence from the theory by coming up with tendentious descriptions of the political world. Even when rational choice theorists back away from pure universalist claims, they do so in ad-hoc and in unconvincing ways that reinforce their reluctance to entertain the possibility that their theory is incorrect. It is as if someone, on observing one day that red apples no longer fell toward the ground when dropped, asserted that the theory of gravity is fine; we must just accept that it does not apply to red apples.
As this example might suggest, when all else fails the universalist impulse leads rational choice theorists to take refuge in the philosophy of science. Accusing their critics of being "naïve falsificationists," they appeal to the arguments of Thomas Kuhn and Imre Lakatos in vindication of their procedures. Those who appeal to Kuhn seem innocent of his notorious inability to distinguish developing research paradigms, when knowledge is advancing, from decaying ones, when it is not. Given the failure of rational choice theorists to identify unambiguous advances in empirical knowledge, this is a serious worry. Indeed, as Kuhn was careful to insist, the human sciences have yet to reach the stage where there is a dominant paradigm within which normal science can proceed by puzzle-solving. Hence his description of them as "pre-paradigmatic." Those rational choice theorists who acknowledge that this is the true state of affairs sometimes see it as their mission to provide a remedy--as my discussion of David Laitin's agenda for political science in chapter 6 reveals. They would be well advised to take note of Kuhn's insistence that "I claim no therapy to assist the transformation of a proto-science to a science, nor do I suppose that anything of the sort is to be had."18
Rational choice theorists who appeal instead to Lakatos make heavy weather of the claim that you can't beat something with nothing; that theories are not falsified by being tested against "the facts," but rather when a better theory comes along--one that explains what was known before and then some. This is concededly a good account of what often happens in science, yet the difficulty we identify is that in political science rational choice theorists operate is if it applies to their critics but not to themselves. Their characteristic proclivity is to ignore previous scholarship on the topics they study, to create trivial null hypotheses when any are considered at all, and to translate existing knowledge into their preferred terminology rather than add to it. Rational choice theorists operate more like brief-writers for their universal theories, not Lakatosian scientists who try to add to the inherited stock of knowledge by developing theories that perform better empirically than those that have been tried before. If rational choice theorists took their Lakatosian protestations seriously, they would abandon the logicist impulses that flow from Hempel's theory of science and engage seriously in comparative empirical evaluation of their arguments against the most plausible going alternatives. We note that the few of them who have done this have, indeed, contributed to the study of politics. For most, however, the impulse to flee from problem-driven theory to method-driven theorem proves irresistible.
In chapters 3 and 4, I turn from the explanatory to the normative dimensions of the flight from reality. One way in which commentators have sought to link positive and normative theory is to try to derive principles for action and policy from their accounts of how the world works. Another is to try to intervene in the world to get it to conform to one's normative ideal better than presently it does. Richard Posner, a United States appellate court judge since his appointment to the Seventh Circuit by Ronald Reagan in 1981, engages in both. I provide an assessment of his efforts in chapter 3.
Getting from explanation to prescription means getting from is to ought. Posner's attempt to do this is elusive and ultimately unsuccessful, but it reveals one of the ways in which the flight from reality can have an impact on reality. Posner posits a functionalist account of the common law's evolution, according to which it operates over time to maximize economic efficiency. His is an invisible-hand account in that this process is alleged to occur beyond the ken of common-law judges, who typically do not perceive--let alone understand--the ways in which their decisions contribute to this result. Indeed, as if to underscore this point, Posner upbraids some judges who have sought to apply his theory of wealth-maximization in the course of adjudication. Yet Posner nonetheless delivers a series of nostrums about how judges should behave and proposes various reforms to the administration of courts in the United States based on appeals to his efficiency arguments.
Attention to Posner's invisible-hand account reveals good reasons for skepticism, but even if we bracket them it is far from clear that we should regard it as benign. Consider Posner's efficiency-based account of why the criminal law disproportionately punishes harmful acts committed by poor people.19 Criminal sanctions become necessary, on his account, when the threat of compensation through the torts system fails to deter potential wrongdoers. Because poor people typically lack the resources to compensate victims that would be sufficient to deter potential perpetrators, the threat of incarceration is needed instead. This "efficiency" based account of a system that discriminates against the poor entails nothing about its moral attractiveness, unless supplemented by an argument showing efficiency to be more desirable than equitable treatment. This becomes obvious when we recall thatscholars in the critical legal studies movement have argued for the same evolutionary thesis as Posner's, but to make a rather different normative point.20 Claims about neutrality to the contrary notwithstanding, they argue, the law operates to sustain capitalist market relations. Oliver Wendell Holmes might have been right that the Fourteenth Amendment "does not enact Mr. Herbert Spencer's Social Statics," but they think that it creeps into the common law nonetheless.
The functionalist case does not get Posner from is to ought any more than it gets Unger and his colleagues from is to ought not. Nor does it exhaust the possibilities for pressing explanatory theory into the service of normative argument. Another strategy Posner deploys is to point to gaps between purposes he alleges to be immanent in the law and the reality on the ground, thereby supplying impetus to the suggestion that reality stands in need of reform. This is strategy is at least as old as Jeremy Bentham's market-failure theory of the need for government because of the possibility of free riding, tragedy-of-the-commons problems, and other by-products of selfish individual behavior.21 A different variant was pursued by the legal realists, who pointed to yawning gaps between the professed ideals of American law and the brutal realities of criminal prosecution during the 1930s and after. This eventually spawned a wide range of reforms to criminal procedure and defendants' rights by the Warren Court, designed to bring reality into better conformity with professed constitutional ideals.
Posner's modus operandi is to identify inefficiencies in the law and the administration of the courts, thereby generating impetus for reforms oriented to bringing them into better conformity with his efficiency ideal. Passing over the difficulty, already noted, that it is far from clear that his efficiency ideal is in fact immanent in American law, I show how his appeals to efficiency all involve reifying contestable economic theories as "the" economic theory or the law. This enables Posner to mask a particular ideological agenda in the garb of abstract theory. For instance, close inspection reveals that Posner's unhappiness with the increase in litigation in federal courts in recent decades does not depend on his--or any--general theory of the optimal level of litigation or judicial services. Rather, it reflects a conservative antipathy for government.
These proclivities extend to Posner's activities on the bench. Through an analysis of his labor law and antitrust opinions in his first five years as a federal appellate court judge, I show that he imports his version of the Social Statics into judicial opinions in predictable ways, all the while portraying them as uncontroversial economic theory--if not unassailable common sense. The truth, as I show, is that different economic theories than the ones Posner invokes would produce different results in these cases; yet he never supplies us with principled reasons for preferring the ones he advocates. The logicist's variant of the flight from reality involves misrepresenting the world by confusing it with dubious models that flow from a pet general theory. Posner's academic writing exemplifies this pathology at every turn. More troubling is that his position on the bench gives him the power to try to reshape reality in accordance with his models--to the arbitrary detriment or benefit of litigants who happen to have their cases appealed to his court.
Dramatically arresting as Posner's reification of efficiency might be, I argue in chapter 4 that he is scarcely alone in his general approach. Political theorists often fail to appreciate that any claim about how politics is to be organized is bound to be a relational claim involving agents, actions, legitimacy, and ends. If they did, they would resist many of the standard contending views in the controversies that preoccupy them. I demonstrate this by reference to contemporary debates concerning the nature of right, law, autonomy, utility, freedom, virtue, and justice. Rather than confront the complexities implicit in the relational logics of these and other political ideals, all too often political theorists appeal to gross concepts--my term for ideas that feed into and promote misleading dichotomies. By appealing to gross concepts, political theorists reduce what are actually relational claims to claims about one or another of the terms in a relational argument. I explain how this systematically obscures the phenomena they purport to analyze, diverting attention from first-order questions about the world to second-order conceptual debates that can never be settled because they rest on category mistakes. Gerald MacCallum Jr. pointed this out a long time ago in connection with the arguments about liberty. He showed that debates between "negative" and "positive" libertarians are not really about kinds of freedom, as protagonists since Isaiah Berlin have supposed, but rather about who is free, from what constraints or because of what enabling conditions, to perform which actions.
Despite a good deal of ritualistic genuflecting to MacCallum in the literature, my analysis reveals that his reasoning has penetrated little. This is partly because taking it seriously would entail coming to grips with contentious empirical debates about human psychology and the causal structure of the social world--which theorists are loath to do. In arguments about freedom, as in numerous other controversies that exhibit a similar structure, gross concepts are instead presented as dichotomous alternatives. Protagonists on both sides defend their views mainly by pointing to the demerits of the supposed alternative--about which they are invariably right. The result is to perpetuate debates that lurch among gross concepts without getting anywhere.
The activity thrives nonetheless, partly because it is rewarded by professional incentives in the academy but partly, I argue, because many people find gross concepts appealing. They are comforting and simplifying devices that function, as ideologies, to legitimate things that people want to believe. Recognizing the alluring power of gross concepts, I defend an account of the political theorist's vocation that revolves around resisting it. Our job is to reel in gross concepts, not to traffic in them. Rather than try to find the right gross concept to champion, we do better to operate as principled social critics whose goal is embellish political argument with political reality. We should be roving ombudsmen for the truth rather than partisans of any particular message. What this means for the conduct of political theorists within the political science discipline is explored in the final two chapters.
This question is taken up in chapter 5 in the context of a critique of the impulse toward, and rewards for, reductionist explanations in political science. My point of departure is the observation--often leveled at my argument with Green in Pathologies--that all observation is theory-laden. Because there is no theoretically unsullied account of "the facts," so the argument goes, we were naïve in supposing that any particular explanation can be evaluated simply by how well it does empirically. Agreeing that Green and I attended insufficiently to this question, I make the case here that it does not stand in the way of a view of the social-scientific enterprise geared to getting at the truth, that this is unlikely to be achieved by any of the going reductionist ventures in political science, and that the endemic availability of alternative descriptions of political reality creates an important ongoing responsibility for political theorists in the division of labor within political science.
The assertion that all observation is theory-laden turns out on close inspection to merit the retort that some types of theory-ladenness are nonetheless more plausible than others. Once the headlights are turned up on particular cases, it becomes plain that there is a world of difference between theory-driven description, in which tendentious accounts are projected onto the problems in order to vindicate pet theories, and situations where there are persuasive reasons, while taking account of previous attempts to study a problem and their limitations, for preferring an alternative construction of it. This is to say nothing of method-driven work, where the construction of the problem is contaminated by the methods available to the researcher. A study in which John Huber and Charles Shipan try to measure Congress's desire to confer discretion on administrative agencies by the number of words in the relevant statutes is easier to comprehend once one is reminded of the "word count" feature of modern word-processing programs.22 No matter that it would seem to suggest that the countless reams of the Internal Revenue Code are there to limit the discretion of IRS agents. Try telling that to your accountant.
Moreover, whereas some disagreements about explanation are really disagreements about competing descriptions in drag, many are not. When a particular description is recognized as apt, there is then a truth of the matter as far as explanation is concerned--even if no one yet knows what it is. That is the realist presumption on which the conduct of all science, on my account, rests. The mistake is to suppose that explanations are scientific only when they flow from a single general theory, which then leads scholars to shoehorn the construction of every problem into terms that are compatible with it. Antireductionist priors, that build toward feasible generality rather than take it for granted prior to empirical research, may well involve descriptions that are theory-influenced. But they will not be theory-driven.
This is not to deny that significant challenges are posed for political science by the theory-ladenness literature. Different ways of characterizing social phenomena predispose researchers to reach for different explanatory arguments, raising questions about how appropriately to choose among them. The conventional answer is prediction. If we can better predict the ways in which judges will decide cases by looking at which outcome maximizes wealth rather than at their jurisprudential commitments, then we should opt for the efficiency-based characterization. But prediction is not all that it is cracked up to be. It can lead us to the misconception that playing basketball makes people taller. More consequentially, I show that because prediction is so difficult in the human sciences, excessive preoccupation with it can drive researchers to focus on trivial but tractable questions--three points to the right of the decimal. Alternatively, it can lead them endlessly to refine predictive instruments that are never going to work. The result? Immensely complicated clocks that neither tick nor tell the time.
Political theorists have important roles to play in political science just because there is no algorithm that dictates the correct descriptive cut at the social world. Among our central tasks is to identify, criticize, and suggest plausible alternatives to the theoretical assumptions, interpretations of political conditions, and above all specification of problems that underlie prevailing empirical accounts and research programs, and to do it in ways that can spark novel and promising problem-driven research agendas. And, especially when esoteric forms of redescription are involved, they must elucidate the links to more familiar understandings of politics.
If the study of politics is inherently contentious, how should it be taught to the next generation of students? This is the subject of my final chapter, a critique of David Laitin's recent proposal to standardize political science teaching to undergraduates. His goal is to do for political science what Samuelson did for economics by creating an orthodox curriculum, and he proposes a model syllabus for its introductory course. I make the case that Laitin's proposal is wrongheaded in ways that both reflect and encourage the flight from reality.
Laitin's proposal reflects the flight from reality by conflating the discipline with its object of study, calling to mind the economist's reputed answer to a student query that economics is "what economists do." Undergraduates typically become political science majors because they are interested in politics, not the political science discipline. It should be troubling if these students are presented, as Laitin believes that they should be, with an account of political science that is sanitized to obscure its controversial character. The impulse to do this has more to do with professional incentives to get grants and public recognition for the discipline, as Laitin comes close to conceding, than with any plausible account of how to teach students to think systematically about the fundamentals of politics. Even when scholars believe they have the One True Theory or the One Best Method--indeed, particularly in that case--they do students a disservice not to tell them about the others.
Laitin's proposal is liable to encourage the impulse to flee reality by detaching scholars from a particular kind of discipline that undergraduate teaching offers. His reasoning depends on the illusion that an initial political science course can get students up to speed in the same way that essentials are covered in introductions to physics, chemistry, and math. These disciplines differ from the social sciences in that there is little, if any, disagreement among their practitioners on what the appropriate descriptive cut on the world actually is. Against Laitin, I argue that the economists' decision to standardize their discipline as he recommends and would emulate has had the disastrous consequence of divorcing introductory teaching from controversies at research frontiers of their discipline. This famously alienates undergraduates from what they are required to study in introductory economics courses. It also reinforces the powerful disincentives for active research scholars to teach the introductory courses, turning them over instead to adjunct faculty and flunkies.
This is a loss for the students, but also for the researchers. Bright undergraduates often have noses for important problems and a refreshing desire to cut to the chase about what difference the theories they are being asked to understand make in the actual world. They are not hostage to academic literatures that may be little more than rotten boroughs in the ways that graduate students so often are. If scholars in political science were to free themselves from the discipline that comes with having to teach undergraduates as completely as economists have done in recent decades, they would have yet one more reason to succumb to their--already overdeveloped--fear of not flying.
Return to Book Description
File created: 8/7/2007
Questions and comments to: email@example.com
Princeton University Press