T HIS BOOK is a big, ambitious undertaking, organized in 31 long chapters covering subjects which range from religion and civilization in economic life to money, banking, wages, and incentives. But, as the title suggests, it is essentially a research tool. It is meant not so much to be and reviewed as to be The test is whether people end up coming back to it more and more or less and less as time goes on. Judging from those areas of economics and sociology in which I have worked, I would predict people will return to The Handbook of Economic Sociology more and more. The chapter by Chris Tilly and Charles Tilly on labor market structures, the area I know best, is the most comprehensive review of the subject; it handles with rare sophistication, material drawn from across the social sciences. The essay by Alejandro Portes on the informal sector and that by Ivan Light and Stavros Karageorgis on the ethnic economy also consolidate areas of study dispersed over the literature of a variety of different disciplines. But I doubt that any single person is in a well-informed position to pass judgment on all of the essays in the volume. Nonetheless, the publication of a book like this provides an occasion to reflect upon the field of scholarly endeavor, to consider what it represents as a complement to conventional economics and, possibly, as an alternative. For this, it seems reasonable to the text, or at least peruse it, chapter by chapter. The first thing to be said about approaching The Handbook of Economic Sociology in that way is that it is a true handbook: The editors, Neil J. Smelser and Richard Swedberg, provide very little guidance on how it might be read as opposed to referred to. It has no real introduction. It invites readers to pick out chapters at random, following their own inclinations. This, moreover, turns out to be a very frustrating experience. It leaves one wondering what economic sociology is, or even, what economics is that economic sociology is not. Absent some other guide, one seems forced back to basic definitions. In introductory economics-at least when I teach it-we offer two of these. One defines economics broadly as the study of how people employ scarce resources and distribute them over time and among competing demands (Paul Samuelson 1961). The other is much narrower and more focused:

JN THE LAST DECADE, the levels of unemployment in industrialized countries have risen dramatically. Rates now are typically two, and in some cases, three and four times those prevailing in the 1950s and 1960s. The fact that these new higher levels are widely tolerated suggests that our thinking about the meaning and significance of unemployment has changed enormously over the period. Such a change is certainly present in the thinking among professional economists. In the 1960s the standard view was that the unemployed represented unutilized resources; their existence in an economy where the vast majority of people had unsatisifed wants was seen as a major social paradox and the most important unsolved intellectual puzzle of the capitalist economic system. This fed the rationale for Keynesian countercyclical fiscal policy and government deficit spending: The government could reasonably print money in order to hire the unemployed because the resources absorbed in the process were essentially free Today, a good number of professional economists, certainly in the United States but to a lesser extent throughout the world, have come to view measured unemployment in industrial economies as an artifact in at least three senses. It is a statistical artifact of a measurement process that classifies as unemployed people who are not really available for work. It is an institutional artifact of a system of social insurance and public welfare that encourages an extension of the process of job search. And it is an artifact of the language that uses a term which in everyday parlance means forced idleness for activities that have important productive functions akin to the functions of inventories, information processing, and investment associated with the utilization of capital goods. These interpretations to be sure hardly constitute a consensus about the meaning of unemployment. But they are no more diverse than the range of views that underlay the older orthodoxy. And for policy makers and economic researchers, they carry a single message: There are many more serious problems toward which to direct attention. These new views about unemployment were developed out of a set of ideas originally associated with the Chicago School of economics, where the emphasis-at once positive and normative-was placed on the competitive market as the gover* A Review of Alexander Keyssar, Out of Work: The First Century of Unemployment in Massachusetts. Cambridge: Cambridge University Press, 1986 and Robert Salais, Nicolas Baverez, and B6nedicte Reynaud, L'invention du chomage: Histoire et transformations d'une categorie en France des annees 1890 aux annees 1980. Paris: Presses Universitaires de France, 1986.

Introduction, 602. — I. The internal innovative process, 604; the search, 604; the design and construction of equipment, 608; job design given capital equipment, 612. — II. The role of outside institutions, 615. — III. Conclusions, 619.

The dictionary defines paradigm as a model or pattern. In grammar, an example of conjugation or declension, showing word in all its inflectional forms. A paradigm is thus practical example which perfectly illustrates an abstract principle. Its hallmark, in this dictionary definition, is the complete correspondence between the abstract and the applied, between theory and praxis. The term was adapted to the discussion of theory by Thomas Kuhn, and its current vogue in economics is attributable directly to his book. As Kuhn used the term, paradigm retains its double meaning of theory and of praxis, but the relationship between the two becomes ambiguous. Scientists believe, Kuhn seems to argue, that their inquiry is based upon theory of how the world operates and how the investigation of its operation ought to proceed. But science is in fact largely set of practices in which members of given community customarily engage. Students seeking to follow their professors in careers in which the latter will judge them and ultimately determine their fate are not encouraged to andertake projects which depart radically From those which their professors conceive For them. When suggestions for such projects trise, either from students or competing nembers of the community, they are as often reated by sarcasm and ridicule as the subect of seasoned discourse and debate. In this )rocess, what students acquire is less an abtract understanding of what they are doing han set of habits, or instincts, about what constitutes legitimate mode of inquiry or plausible explanation. The disjuncture between theory and practice which arises in this way creates the potential for scientific revolution. To ask in this context about the paradigm to which notions of labor market segmentation belong is thus to ask question about the relationship between that mode of research and the theory and practice of economics. I never considered myself revolutionary; indeed quite the contrary. But as an exponent of labor market segmentation in the community of normal economics, I can assure you that labor market segmentation does not fit the paradigm. The sentiments and reactions which Kuhn tells us greet the abnormal and aparadigmatic in discipline, that is, fury, disdain, resentment, sarcasm, and condescension have definitely greeted labor market segmentation. This is matter of fact; an observation about praxis. The question is then what aspect of the conceptual structure of the discipline accounts for this practical result. In most discussions, labor market segmentation is contrasted with human capital theory, but this does not account for the hostile reception it has received. It fails, first, because the treatment accorded market segmentation by the profession is not that different from the treatment accorded human capital when it first appeared. Gary Becker describes his personal experience on this score in the introduction to his book of essays, The Economic Approach to Human Behavior, and the disdain and antagonism which he attributes to his colleagues are at least as great as anything experienced by the proponents of segmentation. But, perhaps more fundamentally, the two views are not, as we shall see shortly, necessarily in conflict. *Professor of economics, Mitsui Professor for Probems of Contemporary Technology, Massachusetts Intitute of Technology.