Read an Excerpt
ON SOCIAL ORGANIZATION AND SOCIAL CONTROL By MORRIS JANOWITZ THE UNIVERSITY OF CHICAGO PRESS Copyright © 1991 The University of Chicago
All right reserved. ISBN: 978-0-226-39303-2 Chapter One Styles of Theory Construction
Sociologists entertain differing notions about the nature of sociological explanations. As sociology became an academic discipline, it became philosophically concerned with the applicability and limits of positivism, and in time, the potentials of logical positivism as a basis of sociological inquiry. This involvement with positivism has not resulted in the partition of sociology. While there are extreme phenomenologists in sociology who reject the relevance of positivism, most "working" sociologists are concerned with the appropriate balance of emphasis in the discipline. It is as if the participants required the close proximity of their opponents for stimulation.
The basic issues become clear if one considers the inquiries that sociologists actually undertake rather than the formulae they claim to adhere to. However, because sociology is incremental, each sociologist does not start afresh. The investigator must make a number of decisions about the logic of his approach which are influenced by past accomplishments. He has a set of alternative ways of constructing his categories and defining the properties of social organization which can be thought of as his style of theory construction. (The possibility of a revolutionary breakthrough is never denied, but there is no basis for anticipating its immediate or even remote eventuality.) Systemic analysis as one style of theory construction supplies a set of decisions about five categories of analysis: general versus middle-range theory; equilibrium versus developmental analysis; interdependent versus autonomous structures; generalized versus differentiated units of analysis; and social versus socioenvironmental boundaries.
General Theory versus Middle-Range Theory
Immediately after 1945, sociologists debated the alternative merits of general theory versus theory of the middle range. Like many such sociological issues, it has subsided without clear-cut resolution. However, on this issue, systemic analysis is explicit: it rejects the relevance of these alternative formulations because they are both part of a common approach. The terms of reference in the distinction were offered with specific intellectual objectives in mind by Robert K. Merton. He was concerned that sociological theory not be limited to the history of ideas, but that theory construction have an influence on empirical research and that empirical research, in turn, on the reformulation of sociological theory. He was writing in the context of his own work as a member of a university-affiliated center during a period of expansion of survey research methodology. He wanted to relate the findings of this methodology to the broader, more enduring interests of academic sociologists. It may well be that the language of his argument created a debate which he did not intend. His formulations gave rise to an antithesis between his approach and that of the comprehensive-system builders, particularly Talcott Parsons and his colleagues. The difference between these approaches is one in preferred degree of abstraction and generality of concepts and propositions.
Merton argues, "from all of this it would seem reasonable to suppose that sociology will advance in the degree that its major concern is with developing theories of the middle range and will be frustrated if attention centers on theory in the large. I believe that our major task today is to develop special theories applicable to limited ranges of data, theories, for example, of class dynamics, of conflicting group pressures, of the flow of power and the exercise of interpersonal influence, rather than to seek at once the 'integrated' conceptual structure adequate to derive all of these and others." There is a close affinity between this style of theory construction and that of systemic analysis. The emphasis is on a range of testable hypotheses for which no single postulated schema will suffice....
But any set of hypotheses requires a group of givens on which they can rest. Systemic analysis refers to testable and delimited hypotheses to be found in theories of the middle range. But there can be no testable hypotheses without some elements-explicit, at that-of a general theory. The argument appears overdrawn at best, except to the extent that particular general theories do not generate testable hypotheses. Naturally, there is considerable latitude in the logical structure, specificity, and their substantive content of the givens. The result of the enormous discussion of this so-called controversy is to point out that some theories, because of their global character or their internal definitions and categories, do not produce testable hypotheses. In the language of systemic analysis, such testable hypotheses do not have a life of their own but are derived in part from, or at least dependent on, elements which are to be found in what Robert K. Merton calls grand theory.
An essential aspect in the debate between the advocates of grand theory and advocates of middle-range analysis centers on the appropriate theoretical terms of reference in a sociological analysis. Again, at this point, the observer might be struck by the lack of agreement in the theoretical terms of reference which sociologists employ in their empirical research efforts. Every abstraction is not a theory, nor can the term "concept" be applied to every abstraction. Much of the discussion in sociology about appropriate theoretical terms of reference is both commonplace and arbitrary. But to dismiss these problems as being a form of oversophistication and to engage in excessive elaboration is self-defeating. It is more important to note the degree of operative convergence in sociological analysis, since systemic analysis does require terms which will expedite the translation of findings from one investigation to another and from one subject to another.
The implicit and explicit issues in the debate between grand theory and middle-range theory resolve themselves when there is a degree of logical clarity which makes it possible to separate the hypotheses that are being tested in a given research from the assertions (assumptions) on which they rest and which are therefore taken for granted in the course of the investigation. This distinction is essential if any real progress is to be made in codification of research findings.
It is important to distinguish between types of givens. There are those givens which are no more than assumptions. They are taken for granted during the course of a particular investigation, and can be weakened by specific empirical findings. Such assumptions are different from the "overarching" givens, those postulates found in grand theory, which represent basic outlooks and are not likely to be rejected by particular empirical research but can be abandoned or modified only by the longterm superiority of alternative sets of hypotheses drawn from an alternative theoretical perspective. In the absence of revolutionary science, this is indeed slow, difficult, and painful....
Equilibrium versus Development Analysis
Sociology reacts strongly to developments in other research disciplines. In particular, it continues to reflect the imprint of biological and physiological constructs.... The intellectual record spans attempts to incorporate Charles Darwin's evolution and natural selection to the contemporary development of cybernetics and the biology of the brain. The most persistent influence has been by the use of the analogy of equilibrium. Distinct from such analogous reasoning has been the effort to explore the linkages of biological factors and social organization or social behavior as a research task.
Equilibrium, as a property of social organization heavily emphasized by L. J. Henderson and his disciples, has to be juxtaposed to the alternative which might best be called developmental analysis. Systemic analysis acknowledges the priority of the developmental posture, at least as a point of entry; at this decisive point it finds equilibrium analysis too confining.
In the extensive literature on equilibrium analysis, the underlying postulate is the inherent tendency of the component subsystems and the overall system of social organization to remain in a steady state. Modifications and deviations tend to be eliminated, reduced, or narrowly restricted. The equilibrium approach, in effect, perpetuates the older distinction between static and dynamic analysis. It asserts the necessity of stipulating in advance all the variables required. In this view, understanding social structure is a prerequisite for explicating social change.
Theoretical orientations which make use of the property of equilibrium handle the question of change by the "principle of emergence," a philosophical construct of long tradition. However, as pointed out by various philosophers, including Abraham Edel, this "principle" is not without problems and limitations for empirical investigators. As applied to social organization analysis, the "principle of emergence" has been described: "... at various levels of organizational complexity systems emerge which have properties which cannot be inferred from or explained in terms of the operation of their component parts or elements and that these emergent properties must be treated as causally relevant variables in the theory. By implication, at each emergent level, certain new degrees of freedom are created." This formulation asserts that the new properties cannot be inferred from the prior state, but it is still causally relevant. The property of emergence violates the development toward theoretical closure which the equilibrium format is designed to facilitate. Thus equilibrium analysis is as much an aspiration as a reality.
Developmental analysis is both more time-bound and more interested in making explicit so-called emergent properties. It does not make the sharp distinction between static and dynamic analysis or between structural analysis and the analysis of social change. There are many unsolved and unsolvable problems of such a format, but it supplies a more realistic and comprehensive orientation, beginning with the aspiration that the postulates of change be explicit and substantive, rather than implicit and purely formal-as is often the case in the equilibrium format. To use such an orientation may well "mess up" the analysis. For example, secondary, intervening variables are stated in advance; but the possibility that additional elements will be introduced to the analysis cannot be ruled out in advance.
If equilibrium analysts, or some of them, postulate that social change occurs gradually, developmental analysis postulates that the rates of social change vary and are a subject for empirical investigation. Likewise, if equilibrium analysis implies that change has its locus outside the system under analysis, in developmental analysis change can originate within and outside the social organization under investigation.
It is a primary responsibility of the sociologist to collect and chart empirical indicators of social change. At this point, sociology and "history" converge. All empirical sociology is history, but history is not all empirical sociology. The collection of social indicator data cannot be limited to particular hypotheses but must reflect the collective judgment of social scientists about the range of descriptive data required to write contemporary social history. In the 1920s such a research notion was presented in the writings of William F. Ogburn. Almost a half-century later, the institutionalized collection of such indicators in sociology has begun to occur."
Developmental analysis at times postulates some hypothetical end state, for example, Harold D. Lasswell's "garrison state." The goal is to examine social change in terms of movement toward or away from this construct. This emphasis in developmental analysis includes the charting of "master trends," to use C. W. Mills's term, which supply the societal context for the investigation of particular institutions. Within such a historical and analytic framework, it becomes feasible to explore the limits of equilibrium analysis.
To argue that developmental models are more appropriate for the study of society hardly implies that systemic analysis of social control has no aspirations for general formulations. It does mean that sociologists must pay attention to "historical" issues in research. To aspire to narrow the distinction between history and sociology is commendable only if the sociologists remain aware of the looseness of notions such as the "principle of emergence" and other escape hatches that historians have allowed themselves.
Likewise, there is a question of the appropriate time span involved in developmental analysis, a subject on which sociologists have not displayed great expertise. There is no general answer; but sociologists have erred in limiting the periods in their developmental analysis, whether the object of study is a particular institutional innovation or a pattern of societal change. Explanations of the limits vary; the dramatic character of particular contemporary events, inherent limitations of sociological inquiry and available data, or even the desire to participate in policy making. However, intellectual criticism of sociological research has had an increasing influence on broadening and extending sociological time perspectives. In summary, systemic analysis encompasses a concern with the equilibrium model and an emphasis on developmental analysis; the actual admixture is related to both the specific research issue and the style of the investigator.
Interdependent versus Autonomous Structures
One aspect of sociological theory which has fairly general acceptance is that embodied in the holistic approach. The elements of social organization have meaning in terms of all other elements. Most of the founding figures of sociology made some equivalent reference to the idea of holistic analysis. The incremental development in sociology has reaffirmed this notion; the influence of social anthropology has reinforced it. However, sociologists require frequent reminders of this aspect of sociological theory, because it is very difficult to implement in research.
The emphasis of holistic analysis makes the meaning and relevance of any factor or dimension of analysis rest on an understanding of all the other factors and dimensions; this is an essential part of the pursuit of macrosociology. For better or worse, sociology is the discipline concerned with "the context." However, sociological theories based on a holistic framework, as Pierre L. van den Berghe points out, can use fundamentally different conceptions of explanation and causality.
In fact, to speak of a holistic orientation is to refer to no more than the view that, in social organization, causation is to some degree reciprocal among the component elements. It does not answer the question of the degree of interdependence of the components in any particular social organization. On this essential issue-the degree of postulated interdependence among component structural elements-sociological writers differ consistently and to a considerable degree; the direction of systemic analysis is to take for granted at least a considerable degree of actual and potential autonomy between elements, or to assert that there are conditions under which such autonomy is operative.
The distinction of equilibrium versus developmental analysis is relevant at this point. Equilibrium versus developmental analysis postulates different degrees of interdependence. Equilibrium orientations assert a great degree of effective interdependence of component elements of any given social organization. In this type of theory construction, the structure of a society is characterized by a continuous network of persons and groups, so that a change in one component has a direct and discernible effect throughout the entire social organization.
However, if one proceeds by means of systemic analysis, the degree of interdependence is problematic; more specifically, developmental analysis asserts that some autonomous groups are required to achieve effective integration in a given social organization. Thereby systemic analysis, which uses developmental constructs, rejects the idea of complete and comprehensive interdependence.
In particular, the very notion of social control in an advanced industrial society underlines the crucial role of autonomous or relatively autonomous groups. This asserts no less than that some groups in an advanced industrial society will be autonomous or relatively so if the process of social control is to be effective. There is no reason to deny the possibility that in a primary group, a community, a large bureaucracy, or [a] nation, effective social control is compatible with and requires some autonomy in particular constituent groups. In fact, examination of social control in an advanced industrial society indicates that, because of the increased complexity of the division of labor, the importance of such autonomous structures is enhanced if social rather than coercive control is to be achieved.
(Continues...)
Excerpted from ON SOCIAL ORGANIZATION AND SOCIAL CONTROL by MORRIS JANOWITZ Copyright © 1991 by The University of Chicago. Excerpted by permission.
All rights reserved. No part of this excerpt may be reproduced or reprinted without permission in writing from the publisher.
Excerpts are provided by Dial-A-Book Inc. solely for the personal use of visitors to this web site.
