Philosophy of social science

Transcription

1 Philosophy of social science Turner, Stephen. The Social Theory of Practices: Tradition, Tacit Knowledge, and Presuppositions. University of Chicago Press, Pp. x, 145. $39.95 (cloth), $14.95 (paper). Ethics 1996 Stephen Turner proposes to provide a philosophical analysis of a family of concepts falling under the general heading of social practice : the notion of a tacit set of rules, norms, or presuppositions that inform behavior in a given cultural milieu and that are socially authoritative and causally influential. Examples include: skills of the carpenter, passed on from master to apprentice; knowledge of laboratory techniques passed on from senior scientist to graduate student; norms of etiquette and comportment, passed on from adults to children; frameworks of interpretation of literary or religious texts; social practices governing the use and significance of food; and many others. Especially important among such phenomena in Turner s account are epistemic practices: the notion that knowledge and belief are dependent upon schemes, worldviews, or frameworks that are not themselves amenable to rational standards of assessment. Turner makes the important point that we need a better conceptualization of the theory of practice. And he offers detailed and provocative analysis of a host of problems raised in this area. Turner s position is that the concept of practice is seductive but ultimately untenable. The central source of his skepticism has to do with the difficulty of assuring that the same practice is transmitted. Where does a social practice reside? How is it transmitted? How do the young absorb the mores of their elders? How are the rules of these practices represented internally within the individual? What guarantees that two practitioners each equally adept have the same practical knowledge, the same internal grammar? On the basis of skeptical answers to several of these questions, Turner concludes that there is no such thing as a social practice; rather, there are trained similarities of external performance. Turner s skepticism about social practices is ultimately unpersuasive. The reality of social customs, norms, mores, assumptions, worldviews, habits practices, in short is too compelling to be seriously doubted. The challenge, then, is to provide an account of embodiment and transmission of such skills and knowledge that makes sense of the facts of social experience and addresses the concerns that Turner raises. And in fact, the outlines of such an account are not hard to produce. We might postulate that each individual possesses a norm inference engine, on the basis of which the individual observes the social

2 behavior of those around him or her and arrives at a hypothesis about the underlying rule of behavior. As dissonant experience is encountered, the norm inference engine updates its hypotheses. The individual s conduct is generated through application of the current state of the social grammar. A cultural practice is embodied in the ensemble of individual social competences and performances at a given time. Individual social behavior is thus structured by the state of social knowledge embodied in the current generation. These structured behaviors in turn constitute paradigm examples for others grist for the norm inference engines of other members of the community. Through such mechanisms, a population of practitioners successfully transmit their knowledge to the next generation. This ultimate disagreement notwithstanding, Turner s book is a worthwhile contribution. Authors as diverse as Heidegger, Wittgenstein, and Bourdieu have made the notion of practice central to their social theories: language, conceptual schemes, and modes of behavior are to be understood as bearing the imprint of social practice. Turner is certainly right in holding that there is no collective entity underlying a social practice no occult system of mores or traditions that causally interact with individuals. And Turner s careful efforts to unpack the presuppositions of this concept serve well to clarify our thinking about social practices. Zalk, Sue Rosenberg and Janice Gordon-Kelter, eds. Revolutions in Knowledge: Feminism in the Social Sciences. Westview, $44.00 (cloth). Pp. ix, 170. Ethics 1994 This collection attempts to assess the import of recent feminist scholarship for various social science disciplines. A common theme throughout the volume is the importance of distinguishing between two objectives of feminist scholarship: to draw attention to issues pertaining to women s experience that have often been neglected by traditional scholarship, and to reformulate the concepts, methods, and frameworks in terms of which social science research is conducted in such a way as to provide a better basis for understanding women s experience. In line with these concerns, many of the essays attempt to document the need for substantial conceptual and methodological reform in the social sciences. Likewise, many of the essays provide detailed and interesting illustrations of some of the shortcomings of traditional research frameworks in various fields. Finally, a good deal of space is expended on detailing the institutional structure of the respective disciplines. 2

3 There is much to admire in this volume. This reader is struck, however, by the negative stance taken throughout toward the value of scientific objectivity. But it is possible to be both feminist and empiricist: to maintain both that the social sciences need to do a substantially better job of investigating and comprehending the experience of women and that social science knowledge depends on standards of empirical adequacy that are reasonably independent of one s prior convictions. And we had better hope that this is possible, if any of our critiques of the status quo are to have any warrant or intellectual conviction. Hollis, Martin. Reason in Action: Essays in the Philosophy of Social Science. Cambridge: Cambridge University Press, Pp. ix, 283. $59.95 (cloth), $17.95 (paper). Ethics 1997 Martin Hollis has made significant contributions over many years to the theory of rationality and the role of this concept within the human sciences. Hollis s signal contribution is his persistent effort to bring these questions back to philosophy, and to locate the points of intersection between the theory of rationality and a host of important philosophical issues. Particularly important has been his continuing engagement with the topic of rationality and relativism roughly, the question of whether there may be profound differences in the mentality of persons within different cultures, and whether the assumption of individual rationality provides a legitimate lever for overcoming the potential incommensurability of systems of concepts and beliefs in different cultures. Readers will therefore find much to interest them in the essays included in Reason in Action. The essays in this volume were written over a period of almost thirty years; nonetheless, they show a remarkable level of consistency and coherence. The volume is organized around three themes: the concept of instrumental rationality as a theory of rationality in action; the relations between roles and reasons in human action; and the problem of other cultures, other minds. The first section discusses foundational issues in the theory of rationality: the problem of rational choice in the face of public goods ( Three Men in a Drought ), the problem of the formation of preferences ( Rational Preferences ), and the problem of strategic rationality ( Moves and Motives ). The problem of preferences is central to the theory of rational action. Given a set of preferences, the rational agent is one who chooses a line of action which maximizes the expected utility of the outcome. But what about the preference system itself? Are there any constraints imposed by reason on one s preferences? Are preferences truly exogenous? Or can one deliberate among 3

4 sets of preferences? Can one aim to change one s preferences? Hollis s general strategy is to recommend a broader conception of rational agency than the standard theory, by permitting rational agents to deliberate about their preferences; to adhere to moral commitments; and to adopt social identities that affect their choices. I want to loosen the usual tie between preferences, actions and outcomes by insisting that a rational agent has rational preferences (p. 54). To this reader the central section of the book is the most challenging and original. Here Hollis interrogates issues which contrast the notion of the rational actor as a transparent preference-maximizer, with the thicker notion of an actor as the bearer of an identity, a role, or a mask. On the first conception, agents have preferences and an ability to compare consequences of alternative courses of action; they then choose that course which has the prospect of producing the greatest expected utility. On the second conception, agents act out of complex identities which derive from their cultures, the roles they assume, and the masks they wear. Further, such actions are only partially deliberative and calculating. An interesting example which Hollis treats in several essays is the notion of a code of honor. Individuals sometimes appear to act out the implicatures of a code of honor which requires acts that are not narrowly prudent. How are we to understand this phenomenon within the context of a general theory of rational agency? Does the agent freely and autonomously choose the imprudent but honorable course? Is the agent programmed by an authoritative normative system? Has the agent adopted a role which compels his/her course of conduct? Hollis s discussion of these issues is richly informed by historical, anthropological, and literary evidence, and his observations take us to a deeper understanding of the complexities involved. In the third and final section of the book Hollis returns to the topic of cultural relativism and the related problem of the interpretation of the meaning of human conduct. Here he addresses the issue of cross-cultural rationality. In Reason and Ritual he confronts the issue of apparently irrational beliefs to which individuals in various cultures adhere. Here the central claim is that the ethnologist can only interpret the beliefs of members of the other culture if he/she is able to attribute shared standards of rationality to the other. The assumptions required for identifying everyday empirical beliefs are common perceptions, common ways of referring to things perceived and a common notion of empirical truth. Unless these assumptions work, the anthropologist cannot get a bridgehead (p. 210). In other words, we cannot maintain a radical relativism of conceptual system and modes of reasoning across cultures, and continue the project of interpreting other cultures. A similar point emerges in Say It With Flowers : in order to interpret an individual s motives in undertaking a given action, we need to know what the person has reason to do (p. 257), and this requires reference to a general and cross-person theory of rationality. The historian starts by trying an intellectual interpretation [of the actor s meaning] 4

5 and, even where only partial explanation results, will have determined what is missing. This is what is involved in making the actors out as rational as possible (p. 258). The final essay in the book combines themes in the second and third parts, by interpreting the fascinating example of Marcus Atilius Regulus, a Roman general in the First Punic War. Did Regulus act rationally in his decision to keep his oath and return to Carthage to face death by torture? How do reasons of honor figure within a general theory of rationality? And how are we to interpret the actions of persons motivated by honor? Hollis s conclusion is that we must find a way of incorporating principled action into our theory of practical rationality. In short, Reason in Action is a highly successful volume which will reward the attention of a wide audience of philosophers and social scientists. Hollis sheds important light on foundational problems in the human sciences through his developed discussion of issues of individual rationality, cross-cultural standards of rationality, and standards of interpretation of human action. Mohr, Lawrence B. The Causes of Human Behavior: Implications for Theory and Method in the Social Sciences. Ann Arbor: University of Michigan Press. Pp. ix, 183. $37.50 (cloth). American Journal of Sociology 1997 Lawrence Mohr, a political scientist with empirical research interests in organizations and organizational behavior, offers a practitioner s analysis of some fundamental philosophical issues in the foundations of the social sciences. Two questions loom particularly large for Mohr: What is involved in providing a good causal explanation of social phenomena, and in what sense are intentional human actions caused? These are important questions for those of us interested in the philosophy and methodology of the social sciences. Mohr s treatment, however, adds less to our understanding of these issues than one might hope. Causal explanation is the central topic. But in addition to an extended discussion of causation, the author provides an analysis of the causes of intentional behavior (do thoughts cause behavior?), an examination of some possible implications of the author s findings for research methodology, and a discussion of the relationship between explanation and law (chapter 5). Mohr argues that there are strong reasons to doubt in principle whether there could be strong law-like generalizations in the social sciences. This conclusion depends on the fact that social outcomes are the result both of causal processes and encounters chance circumstances that do not fall under lawlike generalizations (pp ). 5

6 Consider first Mohr s analysis of causation. Mohr argues that current philosophical treatments of causation are unsatisfactory, particularly for the social sciences. He summarizes these theories under a Humean regularity view and a necessary conditions view. In place of these theories, he argues for a pair of conceptions of causation: physical causation and factual causation. Physical causation, he believes, is the fundamental notion; it involves a relation between a force on an object and a motion of the same object (p. 44). More generally, physical causation is mechanical causation, in which the physical characteristics of the causing factor produce or bring about the characteristics of the effect. Mohr regards this as the bedrock causal concept, but observes that many causal ascriptions do not in fact refer to physical cause couples. Instead, many causal ascriptions refer to relations among statements or facts: the fact that the road was slick caused the fact that the accident occurred. This is what Mohr refers to as factual causation. Mohr defines factual causation in the following terms: X was the factual cause of Y if and only if X and Y both occurred and X occupied a necessary slot in the physical causal scenario pertinent to Y (p. 27). (Parenthetically, it does seem odd to call X the factual cause if it is merely one out of many necessary slots in the physical causal scenario.) In this reader s view, this analysis of causation fails to illuminate. There are very hard questions to be answered if we are to have a satisfactory analysis of social causation, but the retreat to physical causation seems not to be helpful in attempting to answer such questions. (Does it make sense to ask what was the physical cause of the fall of the Roman Empire?) There is a substantial literature within the philosophy of science on the issues of causation and causal explanation (for example, work by Wesley Salmon, Nancy Cartwright, or Jon Elster). Within the context of the current state of debate, Mohr s analysis of physical and factual causation appears unhelpful. Turn now to Mohr s treatment of the causation of behavior. He asks a simple question do thoughts (purposes, reasons, wishes, desires) cause behavior (physical movements of the body)? His answer is negative, because we sometimes decide to do X but fail to do so (the irregularity problem). The connection between our thought categories and the appropriate behaviors does not seem to be necessary or lawlike (p. 63). He argues that we should here take our cue from the analysis of physical causation that preceded, and look for a set of physical circumstances that could be said to cause behavior. Here the likely suspect is neurophysiology. Agents have brains, and brains are the causal background of behavior. Thus Mohr believes that it is necessary to identify the physiological mechanism that underlies the reason-behavior nexus if we are to have adequate causal explanations of behavior at all (p. 58), and he offers the affect-object paradigm as the basis of further theorizing in this area (p. 70). This is a highly speculative account of how desire and thought might be embodied within the central nervous system. Here we are in deep and troubled 6

7 waters, for it seems fairly clear that cognitive science and neurophysiology are still very far from having convincing answers to questions like what is the physiological basis of thought?. Fortunately, most other observers have concluded that the sort of inter-theoretic reduction that Mohr aspires to is in fact unnecessary. We can give satisfactory answers to questions like how do falling interest rates cause rising levels of investment? without having a neurophysiology of investor decision making. More generally, it is not necessary to reduce causal assertions at one level of description to causal relations at a lower level of description. In summary, Mohr deserves credit for focusing on a series of issues that are fundamental to our understanding of social research. Mohr s discussions of many of these issues ask some of the right questions, but his answers are often less than satisfying. Gilbert, Margaret. On Social Facts. Princeton University Press, Pp. x, 521, index. Cloth $69.50; paper $ International Studies in Philosophy 1995 This is an intelligent, closely argued and extensively analyzed treatment of the problem of social collectivity. What is a social group? What distinguishes a group from a random set of individuals e.g. the set consisting of W. V. O. Quine, Madonna, and Napoleon? Is a social class e.g. the English working class in the 1880s a social group? Gilbert s primary contention is that the notion of a collectivitity individuals constituting a group is the central feature of social ontology and the chief focus of empirical social science. And she maintains that this concept can best be analyzed by the idea of a plural subject the referent of the first-person plural pronoun, we. The core of Gilbert s theory of social groups involves the idea of the mutual recognition by a set of persons that they are engaged in some joint actions or beliefs. A set of people constitute a social group if and only if they constitute a plural subject ; and a plural subject is a set of people each of whom shares with oneself in some action, belief, attitude, or similar attribute (p. 204). Gilbert argues that the pronouns us and we are the linguistic elements through which we refer to plural subjects in English. And she believes that plural subjects exist; they are not fictions or constructions, but agents which have beliefs, perform actions, and succeed or fail in carrying out their intentions. In later chapters Gilbert extends her conception of collectivities and plural subjects by considering several other important social notions: the idea of a social fact in Durkheim s sense, the idea of a collective belief, and the idea of a social convention. In each case Gilbert argues that the concept of a plural subject 7

8 supports a plausible and intuitively convincing analysis of the social concept in question. According to Gilbert, social groups are plural subjects, collective beliefs are the beliefs of plural subjects, and social conventions are the fiats of plural subjects (p. 408). Gilbert s account of social conventions is developed through extensive discussion of David Lewis s influential formulation of this concept. Gilbert argues against the individualism of Max Weber (and by implication, the premises of rational choice theory), by arguing that collectivities are the central subject of the social sciences, and that collectivities cannot be subsumed under (narrowly) individualist concepts. Thus Gilbert suggests that her theory offers support for holism over individualism (p. 3). Does it? I think not. An individualist is free to acknowledge that individuals have beliefs that refer to other persons and groups of persons; the position permits reference to shared purposes and actions involving a collection of persons deliberately orienting their actions towards a shared purpose. What individualism requires is simply that these are all the aggregate results of individual states of mind, and that the behavior of the ensemble is to be explained by reference to the beliefs and intentions of the participants. An important test case for Gilbert s account is the problem of collective action. Rational choice theory places much emphasis on public goods problems and the phenomenon of free-riding. How does Gilbert s conception of plural subjects treat the problem? It appears to this reader that Gilbert makes collective action too easy. Plural subjects (groups) have purposes; individuals within these groups express quasi-readiness to perform their part of the shared action; and when circumstances are right the group acts collectively to bring about its collective goals. The people concerned would be jointly ready jointly to perform a certain action in certain circumstances (p. 409). She speaks of group will or communal will (p. 410). But the actions of a group are still the result of the choices made by constituent individuals. And however much the individual may align him- or herself with the collective project, the collective behavior is still no more than the sum of the actions taken by particular individuals. Moreover, it is necessary to acknowledge the endurance of private, individual interests that remain prominent for individual agents with the result that we should expect individuals actions to sometimes involve free-riding, defection, and favoring of private over collective interests. It seems to this reader, then, that Gilbert leans too far in the direction of the Rousseauvian general will interpretation of social action. How important for the social sciences is the notion of a social group or collectivity? Gilbert s view is that this concept is foundational; it is the basis for a unitary definition of the subject matter of the social sciences. This overstates the importance of collectivities, it seems to this reader: there are important instances of social explanation that do not involve analysis of groups in Gilbert s 8

9 sense, and whose explanatory frameworks do not refer to groups, their behavior, their shared beliefs, or their collective intentions and self-understandings. A few examples might include neo-malthusian analysis of the relation between economic change and demographic variables; analysis of the effects of changes of the transport system on patterns of settlement and economic activity; and explanation of patterns of historical processes of urbanization in terms of changing economic and political institutions. These examples explain social phenomena as the aggregate result of large numbers of rational individual actions. They commonly refer to impersonal social structures and circumstances that function as constraints and opportunities for individuals. And they make no inherent reference to the forms of group collectivity to which Gilbert refers. This is a rich book, and one that repays careful reading. It will be of particular interest to philosophers of social science and social philosophers, and the level of philosophical rigor will interest philosophers in other fields as well. Bunge, Mario. Finding Philosophy in Social Science. Yale University Press, Pp. xii, 432. $45.00 (cloth). American Journal of Sociology 1999 This is an ambitious book. Mario Bunge is an accomplished and learned philosopher who has written for decades on topics in the philosophy of science, metaphysics, and epistemology. In this work Bunge turns this erudition to the task of constructing a systematic philosophy of the social sciences from first principles. Moreover, he aims to produce a system that will be recognizable to practitioners as addressing conceptual and methodological issues that arise in their work. The results should be of substantial interest to philosophers and philosophically minded social scientists. Bunge offers a new, science-oriented philosophical system (12), within the context of which he looks to redesign the philosophy of social science. Philosophy of social science, according to Bunge, includes answers to questions in seven areas: logic, semantics, epistemology, methodology, ontology, axiology, and ethics. And the standards of evaluation for a proposed philosophy of social science are equally explicit: relevance, intelligibility, exactifiability, internal consistency, external consistency, size and depth, truth, universality, fertility, and originality (11). As this treatment suggests, Bunge brings a Cartesian exactness to his inquiry. His theorizing follows from an exacting analysis and classification of the problems to be considered. The book is organized into three parts: From Fact to Theory, From Explanation to Justification, and General Philosophical Problems in Social Science. The first part focuses on problems of explanation and theory. What is 9

10 a fact? What is a generalization? What is an explanation? Bunge puts it forward that the heart of explanation is systematization: the effort of the scientist to place a given fact within an orderly system or pattern (108 ff.). Systematization may proceed through classification (ranging individuals into groups or classes); more fundamentally, it may proceed through the construction of a theory whose concepts and hypotheses allow the scientist to place the given fact or state within a broader context of law-governed order. The treatment places the notion of nomothetic order at the center of the account. A fact is explained when it is placed within the context of a hypothetico-deductive system of laws and boundary conditions. It is worth noting that Bunge s account does not give much weight to the concept of social causation; his theory of explanation does not emphasize causal relations as a central basis of social explanation. Another central concern in Part A is the logic of scientific inquiry. What is involved in formulating a scientific research hypothesis? What is involved in asserting, or assessing, the truth of a hypothesis? How are scientific theories corrected and criticized? Bunge describes what he calls the methodics of science in these terms: the methodics of a scientific approach is constituted by all the relevant scrutable procedures: that is, all the rule-directed procedures that yield objective knowledge and can be justified theoretically as well as empirically (p. 80). Bunge regards scientific research as a process that is partially rule guided, but ultimately underdetermined by empirical constraints: data filtering and hypothesis formation depend on background assumptions that are not themselves empirically constrained at a particular moment in the history of science. Part B turns to a treatment of the epistemology of social science: the empirical and logical procedures through which scientific hypotheses and theories are tested or validated. Bunge s analysis displays an admirable confidence in the objectivity of science and the possibility of rationality and method in the social sciences. Throughout the author endeavors to establish the rational grounds of social science knowledge. A noteworthy discussion of upward and downward explanation of social facts occurs in Chapter 5. Here Bunge addresses issues that fall within the general framework of reductionism and methodological individualism; but he offers a clear, logically articulate account of the relations that obtain between higher-order theory and micro-level events. Bunge distinguishes between macro- and micro-facts in the following terms: A macro-fact is a fact occurring in the system as a whole. A micro-fact is a fact occurring in or to some or all of the members of the system at the given level (p. 147). He then distinguishes between top-down and bottom-up explanations: A top-down (or micro) explanation of a macro-fact is the deduction of the proposition(s) describing it from propositions describing (micro-) facts in components of the system where 10

11 the macro-fact occurs (147). A methodologically individualist explanation, then, is an instance of a top-down explanation in Bunge s parlance, as is an explanation of the properties of a cloud in terms of the properties of the water drops which compose it. Against both individualism and holism, Bunge maintains that adequate social explanation requires both kinds of systematization. Individuals constitute social facts; but individuals act within social contexts. Bunge refers to this alternative as systemism an approach to social explanation that identifies agents within structures as the unit of analysis (264). There are few traditional topics in the philosophy of the social sciences that do not find at least a few pages of discussion in this book. The topic of verstehen, for example apparently very distant from the logical positivist tone of much of the book receives brief but helpful treatment in Chapter 5. Bunge argues that there are two strands of thought underlying the verstehen methodology Dilthey and Weber and that Weber s rational motivation interpretation is the superior (154). An especially original contribution in this part of the book is Bunge s extensive and useful account of the epistemology of social indicators e.g. longevity as an indicator of quality of life. The discussion is interesting and original, and raises a topic not often discussed by philosophers, or by social scientists with the philosophical acuity which Bunge brings to the discussion (170 ff.). The general question is this: what are the epistemic features of an adequate social indicator? What conditions must a given indicator satisfy in order for us to be confident that the measure serves to indicate an underlying social characteristic? The topic is of importance because it constitutes the empirical link between theory and the social world in most instances: empirical observation in the social sciences depends on construction of indicators that permit the researcher to observe social phenomena poverty, social unrest, prejudice. But it is all too common in social science research that an investigator sets out to probe theoretical property X; asserts that observable property P indicates X; finds that P behaves thus-and-so; and concludes uncritically that X behaves thus-and-so. Before drawing such conclusions, however, it is crucial to establish the conditions of validity and veridicality which pertain to the relation between the indicator and the underlying social characteristic. In Part C Bunge turns to topics which have often been discussed under the rubric of philosophical problems raised by the social sciences (as opposed to philosophical frameworks which can be applied to the social sciences). Central among these problems are individualism and holism; idealism and materialism; subjectivism and realism; and the status of the assumption of rationality in social science explanations. Here again the careful reader will be repaid with ample philosophical insight into the problems Bunge treats. 11

12 As this survey indicates, Finding Philosophy in Social Science is an original and fruitful work. Bunge offers a systematic effort to apply contemporary philosophy of science to the logic of research in the social sciences, and he successfully identifies a handful of intriguing philosophical problems raised by contemporary social science research. Barnes, Barry. The Elements of Social Theory. Barry Barnes. Princeton University Press, p. $29.95 cloth. Giddens, Anthony. Politics, Sociology and Social Theory: Encounters with Classical and Contemporary Social Thought. Stanford: Stanford University Press, p. $45.00 cloth, $16.95 paper. American Political Science Review 1996 These two recent books, both by respected social theorists, offer a valuable look at the development of classical social theory since the middle nineteenth century and its relevance at the end of the twentieth century. In the final years of the twentieth century, it is highly pertinent to ask whether social theorizing largely produced before the end of the nineteenth century can possibly serve to illuminate contemporary society. (By classical social theory these authors mean roughly the theories of Marx, Durkheim, and Weber, along with subsequent elaborations and criticisms of these theories.) Are the classics of social theory of continuing relevance for our own attempts to understand and explain important social processes? Or have events and ideas superseded them? Giddens and Barnes have each done deep and important work on the traditions of social theory, so one reads these books with interest and high expectation. The books are highly complementary, in that Giddens provides a thoughtful historical discussion of the traditions of social theory, while Barnes offers a developed account of how the currents of this tradition ought to be deployed in contemporary social theorizing. Barnes wishes to provide a synthesis of an adequate social theory, taking account of the theoretical insights of classical social theory as well as well-developed criticisms of those theories. There are several foundational questions to raise in the context of the topic of these books. What is the intellectual role of social theory? And how should the micro-macro problem be handled in good social research what is the relationship between the individual level and the social level in social processes, structures, and functions? Barnes offers one possible answer to the first question in the first sentence of his book: The most widely read theories are those that serve as the basis for commentary on the nature of present-day industrial societies and on how they are likely to develop and change (1). Later he writes, 12

13 The formulation of links and connections between particulars is what theorizing is (2). This is to say, evidently, that theory serves to identify the generalized processes which underlie specific social occurrences. Barnes takes explicit issue with what is sometimes referred to as the toolbox model of social theory: that social theory identifies a range of distinct social processes and mechanisms, and that the empirical researcher should be broadly acquainted with these theoretical alternatives and choose the theory that best fits the empirical problem that she is currently studying. Against this view, Barnes argues for theoretical unity (33). It bears saying, however, that it may be that social phenomena are in fact diverse, deriving from a number of non-reducible causal processes. In this case, one would be best advised to the more eclectic or pluralistic view of the role of theory. If the social world is causally diverse, then there will be no grand unification theory possible in the social sciences. A similarly foundational question focuses on the social in social theory: what is involved in asserting that human beings are social beings? Barnes is at pains to question a broadly individualist answer to this question, according to which social institutions constrain and impel rational individuals who are defined independent from social facts. He argues instead for the irreducibility of the social. (It might be noted that Barnes s critique does not merely apply to rational choice theorists; he believes that there is a pernicious individualism lying at the foundation of Parsonian sociology as well; pp. 54 ff.) Against these individualistic social theories, Barnes stresses that we must understand individuals as thoroughly socialized: their motives, beliefs, norms, inhibitions, and conceptions of themselves and their relations to others, are thoroughly imprecated in social facts and circumstances. As Barnes puts it, to insist that we are social creatures and to treat this as a matter of profound and pervasive significance is to announce a strongly anti-individualistic sociological theory (4). He gives substantial attention to a number of social concepts: norms, groups, movements, practices, knowledge, class, and hierarchy. And he aims to persuade readers that an acceptable social theory will be one which gives full recognition to the supra-individual properties that such concepts invoke. Barnes appears to believe that familiar problems of suboptimality of rational individual choice in circumstances of interactive action freerider problems and prisoner s dilemma problems indicate a defect in the theory of individual rationality. But it is surely no critique of evolutionary theory to demonstrate that suboptimal adaptations may emerge through natural selection; it is rather a feature of the world that evolutionary theory discerns and explains. Likewise, the fact that rational individuals when placed in a one-time prisoner s dilemma situation will choose strategies that lead to a predictable suboptimal equilibrium does not refute the theory of economic rationality; it indicates only that individual rationality does not always lead to collective optimality. 13

14 Barnes believes that an appropriately constructed theory of the social individual permits a solution to the collective action problem (4, 76-85). Barnes is right in focusing on collective action as a theoretically central problem for the social sciences. But it is a problem for all stripes of social theory, not merely rational choice theory. It is true that unadorned rational choice theory predicts too little collective action. But a lavishly social theory of collective action surely predicts too much collective action. The assumption of social behavior (that is, behavior oriented toward group purposes) makes the problem of collective action too easy. If individuals always choose to act according to their assessment of collective interests, strikes should never collapse, revolutions should never fail, and ground water sources should never be polluted. Barnes gives substantial attention to the topic of social norms: What is involved in an individual following a norm? He writes, Conforming to norms is a collective activity (57). More broadly, Barnes maintains that the phenomenon of normative action gives strong support to the anti-individualist methodology that he favors. This position is unpersuasive, however, in that it appears relatively simple to provide microfoundations for the fact that individuals are influenced by norms, practices, and worldviews that extend beyond their own psychological states. To put it crudely, the Parsonian model of the internalization of norms that Barnes criticizes appears to be more credible than Barnes s own account. In general, this reader judges that Barnes overestimates the difficulties of framing an empirically adequate individualist (or microfoundational) theory of norms and institutions. In sum, Barnes s book repays close study. He is a thoughtful commentator on foundational questions in the social sciences, and even if one ultimately disagrees with his position, the reader will learn much from his discussion of these topics. Turn now to the Giddens volume. This collection of previously published essays provides a fine discussion and commentary on the main contributors to classical social theory Marx, Weber, Durkheim, Comte, and Parsons, and a suggestive reflection on more recent social theorists Garfinkel, Habermas, Foucault, and Nietzsche. Giddens has much of interest to say about the development of social theory and sociology over the past century and a half. Giddens himself is a distinguished contributor to this tradition, so his discussion will be sure to interest many readers. The heart of classical social theory is the trio of Marx, Durkheim, and Weber. Giddens takes pains to scrutinize the routes by which the thought of these thinkers made its way into the English-speaking world of social theory and sociology. Giddens treatment of Weber is particularly thorough and useful; he makes a sustained effort to explore the connections between Weber s intellectual development and the social and political circumstances in which he found himself in Wilhelmine Germany, through an analysis of Weber s political 14

15 writings. Giddens also provides a nuanced discussion of the relationship between Marx and Weber (chapter 2). Here the issue largely concerns the status of historical materialism as an explanatory framework. Giddens argues that Marx and Weber in fact shared a good deal of common ground concerning their interpretations of Roman history and pre-capitalist forms of economy and society (pp ). (Refreshingly, Giddens also defends the view that Marx remains a classical social theorist, with important contributions to make to our understanding of the dynamics of modern capitalist society and economy, in spite of the collapse of communism.) Durkheim s thought is reconsidered through the lens of his political sociology. Giddens takes the view that other readers have given too little attention to Durkheim s efforts to make sense of forms of authority in modern society. This concept cuts across typically sociological discussions of order with political discussions of power. Durkheim is most commonly associated with the regulative authority of norms and of organizations of civil society, but Giddens argues that there is equally a concern for the exercise of the coercive power of the state ( ). Giddens offers the summary assessment: Durkheim s sociology had its origins in an attempt to reinterpret the claims of political liberalism in the face of a twin challenge: from an anti-rationalist conservatism on the one hand, and from socialism on the other (115). Giddens also provides an extended discussion of Durkheim s relationship to individualism. Durkheim is plainly hostile to methodological individualism as a principle of sociological explanation; less well understood is his relation to moral individualism. In Giddens telling of the tale, moral individualism is, according to Durkheim, a defining characteristic of modern Western social life. Respect for the individual, and the concomitant demand for equality, became moral imperatives: as such, they implied that the welfare and self-fulfillment of every member should be sought after (pp ). In a central essay in the volume, Comte, Popper and Positivism, Giddens gives an historical treatment of the development of the philosophy of science in relation to sociology. The general framework of the chapter is a careful and developed analysis of positivism as a conception of the methodology of the social sciences. Tracing the concept from Comte through the Vienna school philosophers (Schlick, Neurath, Carnap, and Hahn, for example), to the logical empiricism of Hempel, Oppenheimer, and Nagel, Giddens carefully works through the relationships between philosophical debates about positivism and the development of empirical sociology. This discussion sheds much light on the profound impact which logical positivism had on the formation of sociological thought and method. This chapter by itself would serve as an admirable introduction to the philosophy of science for graduate students in the social sciences. 15

16 The final few chapters offer discussion of several important contributors to social theory in the twentieth century Habermas, Marcuse, Garfinkel, and Foucault. These essays have the feeling of occasion pieces rather than fully developed analysis, but each essay provides some insight into the subject at hand. In sum, each of these books must be counted at an important contribution to contemporary discussions of social theory. Giddens book shows a remarkable erudition in the full range of social theory. Giddens offers insight into the logic of the views he discusses, as well as the complicated relations which obtain among them. The book will appeal as strongly to readers who know a good deal about these traditions as it will to the novice. The Barnes book has a distinctive point of view, and makes a substantive case for a particular approach to social theory. Some readers will find this approach highly congenial to their own work, and others will not. But all will be stimulated by Barnes s capable argumentation and exposition of the issues. Barnes and Giddens each give evidence to what is surely true: that the traditions of classical social theory continue to provide insight into modern society. Hands, D. Wade. Testing, Rationality, and Progress: Essays on the Popperian Tradition in Economic Methodology. Lanham, Maryland, Pp $40.00 (cloth). Ethics 1994 This book is a thoughtful, sustained reflection on the suitability of the methodological ideas of Karl Popper and Imre Lakatos in application to the history of economics. As Hands notes, Popper and Lakatos have been the philosophers of science who have had the greatest influence on methodologically minded economists. Wade Hands is himself a methodologically sophisticated economist, and he does a fine job of bridging between current work in philosophy of science and the methodological debates that have occurred within the economics profession in the past fifteen years. The central themes of the book are these: Is it useful to consider whether economic theories are falsifiable? And can we understand the development of economic thought within the framework of Lakatos s theory of progressive research programs? Hands is in general rather skeptical about the utility of either Popper s falsifiability criterion (pps. 67 ff.) or Lakatos s more complex criteria of novelty and progressivity (pp ) as a basis for evaluating economic thought and theory. Particularly useful are the author s efforts to consider various methodological maxims in light of specific episodes in recent economic thought. The book is a valuable contribution to the growing literature in the philosophy of economics. 16

17 Gordon, Scott. The History and Philosophy of Social Science. Routledge, London, Pp. x, 668, index, $ Philosophy of the Social Sciences 1994 It is difficult to characterize the genre of The History and Philosophy of Social Science. The book contains much discussion of many of the great texts and authors in the history of the social sciences Marx, Spencer, Comte, Durkheim, Weber, Bentham and yet it is not really a developed history of the social sciences. It is interspersed with discussions of problems of methodology in the social science the role of laws in social science, the role of modeling, the concepts of progress and order in social theory, the logical character of social and historical explanation and yet it is something less than a coherent exposition of a theory of the philosophy of social science. It is too long and intricate to be a popular exposition of social theory; and yet it is underdocumented for a scholarly contribution. And finally, it is a dense scholarly book that contains neither footnotes nor bibliography. (Books and articles are referred to only by title and date, internal to the text.) The reader is thus left with a sense of puzzlement about the audience the author has in mind. I can think of only one group: beginning graduate students in social philosophy or political science who want to get a preliminary synopsis of the historical development of the social sciences. This is a book of almost 700 pages, treating a great variety of topics. So it is perhaps useful to begin by offering a recital of some of the themes Gordon raises: problematic social concepts society, science, progress, order (chapters 1, 2, 8, 10) philosophy of social science themes (chapters 3, 14, 18) history of economics and political economy (chapters 5, 6, 9, 17) history of social theory (chapters 11, 12, 13, 15) history of political theory (chapters 4, 7) biology and social science (chapter 16) Given the length of the book, it is impossible to summarize or evaluate all the main points. So a reviewer may perhaps be permitted to turn to a venerable social science technique: sampling. History of the social sciences. Most of the book falls generally within the history of the social sciences. The author provides an exposition of the central ideas of an imposing list of figures in the history of the social sciences broadly construed. The breadth of knowledge represented is impressive; Gordon writes with confidence on subjects as diverse as Montesquieu s interpretation of the 17

18 English Constitution (pp. 82 ff.), Plato s account of the division of labor (pp. 59 ff.), and Comte s stages of human evolution (pp. 292 ff.). Most subjects are treated with a broad brush, however: a quick review of the philosophical precursors, a brief account of the intellectual environment, and an account of the theoretical ideas. Most satisfactory and substantive is Gordon s exposition of the history of political economy. About 300 pages are devoted to this topic (dispersed over non-consecutive chapters). Gordon provides fairly detailed exposition of the main stages of development of political economy since the seventeenth century: the evolution of classical political economy (Quesnay, Smith, Malthus, Ricardo); Marxist economic theory; the marginalist revolution (Marshall, Pareto); and the Keynsian revolution. It is disappointing, however, that few substantive connections are made between developments in the history of political economy and the other historical topics discussed. Thus Gordon s exposition of the history of economics as a discipline is essentially free-standing within the book, and one is tempted to think that this material would have been better published separately in a volume devoted to the topic. At the same time, it should be noted that there are already reasonably authoritative books available on this topic: Joseph Schumpeter s magisterial History of Economic Analysis (Oxford, 1954) and Mark Blaug s Economic Theory in Retrospect (Irwin, 1968). Gordon offers little on the history of political economy that could not be found in these two books. The next largest topic treated by Gordon is the history of sociology and social theory: Spencer, Marx, Durkheim, and Weber, as well as a number of lesser figures. Spencer, Durkheim, and Weber each receive exposition of about 25 pages, while Marx s sociology and economics are treated together in a separate chapter of greater length. The exposition is brief but adequate; Gordon identifies and explains the central ideas of each author fairly clearly and accurately. As is the case in the history of political economy, there already exist a number of standard books on the history and development of sociology and social theory. A History of Sociological Analysis, edited by Tom Bottomore and Robert Nisbet (Basic, 1978), covers the history of sociology in perhaps more insightful terms. And Raymond Aron s Main Currents in Sociological Thought (Basic, 1965) is a coherent and insightful treatment of the full range of its subject. It is not obvious that Gordon s book has added much to the existing literature in this area. Least satisfactory is Gordon s treatment of the history of political theory. Much of his discussion falls in the domain of political philosophy. In a chapter on political theory and political philosophy (chapter 4) Gordon offers a discussion of the political theories of Plato, Hobbes, Locke, and Montesquieu; later chapters present the thought of Hume and Smith (chapter 7), and Mill (chapter 11). Much of this material is fairly distant from Gordon s own 18