Class, Crisis and the State

1

Methodological

Introduction

The essays in this book have been heavily shaped by the academic context in which they were written. As a graduate student in sociology I constantly confronted the hegemony of an empiricist, positivist epistemology in the social sciences. In virtually every debate over Marxist ideas, at some point I would be asked, prove it! To the extent that Marxist categories could be crystallized into testable hypotheses, non-Marxists were willing (sometimes) to take those ideas seriously; to the extent that debate raged simply at the level of theory, non-Marxists found it relatively easy to dismiss our challenges.

Marxists in the social sciences reacted to these pressures in several distinct ways. Perhaps the dominant response was to dismiss the attacks of non-Marxist social scientists as reflecting bourgeois ideology and/or a positivist methodology. It was common in Marxist student circles to argue that the very enterprise of formulating testable hypotheses was inimical to a Marxist methodology. Historical and dialectical explanation was counterposed to predictive, linear explanations. Particular hostility was reserved for the battery of quantitative techniques used in American sociology: even to use regression equations in a research project was to abandon the essence of Marxism. The demand that we prove theoretical claims through empirically testable propositions, therefore, was treated as purely ideological. To accept the demand would be to give up the battle by accepting the methodological principles of positivist social science.

A second response was to try to generate empirical studies which would prove our arguments to even the most stubborn opponent. Of particular importance in this vein was the large number of power structure studies produced in the 1960s and early 1970s criticizing pluralist interest-group theory. Such studies contributed greatly to legitimating the use of certain Marxist categories in social research and to demonstrating the ideological character of much pluralist theory. But as many Marxist critics of such research have stressed, much of the dialectical character of Marxist theory was lost in the process. In a sense, a large part of such Marxist empirical work can be seen as using Marxist categories without using Marxist theory.

Naturally, there is a third alternative: the attempt to develop empirical research agendas firmly rooted within not only the categories, but the logic, of Marxist theory. Such an approach would reject the positivist premise that theory construction is simply a process of empirical generalization of law-like regularities, but would also insist that Marxist theory should generate propositions about the real world which can be empirically studied.

This third strategy is only beginning in the United States. In effect it is an attempt simultaneously to engage in debate with mainstream social theory and to develop a style of empirical research which advances Marxist theory. Potentially, the research generated by this orientation may become an important contribution by North American Marxists to Marxist social science.¹

The essays in this book should be seen, in part, as contributing to the formation of this third response to positivist social science. While none of the essays constitutes an empirical investigation of a specific historical or structural problem, they are all intended to help establish the theoretical preconditions for such investigations.

The development of a stronger tradition of theoretically-structured empirical investigation within Marxism has three important preconditions: first, it is necessary that Marxists develop a broad range of research competences so that they can in fact conduct empirical investigations in a sophisticated and sensitive way. Second, it is essential to have a deep grasp of Marxist theory, so that the propositions developed do not merely tap the surface level of Marxist categories but are in fact systematically linked to the inner logic of the theory itself. Finally, it is important to know how to link that theory to concrete research agendas. The essays in this book are primarily relevant to the second and third of these issues. In order to understand how they attempt to accomplish this, it will be helpful to examine briefly the methodology of theory-construction which underlies them.

Linking Theory to Data in Social Research

One of the central epistemological premises of Marxist theory is the distinction between the level of appearances and the underlying social reality which produces those appearances.² This is not to say that appearances are purely ephemeral, inconsequential mystifications. On the contrary, the immediately encountered social experience of everyday life is extremely important. People starve at the level of appearances, even if that starvation is produced through a social dynamic which is not immediately observable. The point of the distinction between appearances and underlying reality is not to dismiss appearances, but rather to provide a basis for their explanation. The central claim is that the vast array of empirical phenomena immediately observable in social life can only be explained if we analyse the social reality hidden behind those appearances. If we remain entirely at the level of appearances we might be able to describe social phenomena, and even predict those phenomena, but we cannot explain them.³

Marxists, then, have generally stressed the importance of elaborating a theory of the underlying structures of social relations, of the contradictions embedded in those structures, of the ways in which those underlying structures generate the appearances which people encounter in everyday life. The classic example of such an analysis is, of course, Marx’s discussion of surplus value in Capital: the equality of exchange relations (commodity relations) in the capitalist market hides the real relations of exploitation within production. One can very easily predict exchange relations by simply investigating characteristics operating at the level of the market (indeed, this is one of the essential projects of neoclassical economics) but in order to explain them it is necessary to explore the dynamics embedded in production relations themselves.

It is one thing to make the epistemological claim that explanation requires the decoding of hidden contradictions; it is another to develop a strategy for studying the social world which allows one to link systematically such underlying structural processes to empirically observable phenomena. General maxims about moving from the concrete to the abstract and back to the concrete are not very helpful. The problem is how to move from the concrete to the abstract, and how to move back.

In the absence of a coherent strategy for linking systematically the abstractions of Marxist theory to concrete research, two problems are likely to arise. On the one hand, Marxist theory often tends to become very ideological and immutable to transformation from empirical study. The frequent impression in Marxist research that all of the answers are pre-given, are known prior to the investigation, is at least partially the result of the methodological distance between the general theory and the facts of history. On the other hand, Marxist research often becomes purely descriptive, contributing only marginally to the development of Marxist theory. Historical movements are richly described using Marxist categories, but those descriptions are difficult to translate into transformations of theory. While one should not exaggerate these two tendencies, nevertheless the advancement of Marxist theory is at least in part retarded by the lack of clear strategies for linking theory and research.⁴

In order to facilitate the development of such strategies within Marxism, two general tasks are important. First, it is essential that Marxist theory be formulated in a comprehensible way. This may seem trivial, but the opacity of much Marxist theoretical work is a tremendous obstacle to using such work as a basis for systematic empirical investigation. In particular, it is critical to distinguish within Marxism between assumptions or premises which are not subject to transformation by historical investigation, and propositions which are;⁵ and it is important to distinguish between definitions of concepts and propositions about those concepts. To be sure, theoretical debates over the definitions of concepts and theoretical debates about the actual dynamics of the social world are related. Definitions should not be arbitrary, and a theory of social structures influences the very definitions of those structures. Nevertheless, the two types of theoretical discussion should not be confused, at least if the goal is to develop a conceptual apparatus that can be used in empirical research.

Clarity, however, is not enough. It is also important to develop a more systematic way of understanding the causal relations between the structural categories of Marxist theory and the level of appearances tapped in empirical investigation.⁶ That is, historical investigation gathers data at the level of appearances (by definition): events, personal ties, manifest economic variables, institutional arrangements, demographic distributions, and so on. In some sense these phenomena constitute effects of structural relations. The problem is to define more systematically what effects means. If empirical investigation is to be directly linked to the logic of the theory itself, then much greater rigour in understanding the logic of causality implicit in the theory is necessary.

Some steps in this direction have been made by Louis Althusser and other so-called structuralist Marxists. The concepts of over-determination and, more broadly, structural causality, have provided at least a preliminary formulation of the relationship between structures and their manifest effects.⁷ This concept of causality, however, has been very difficult to use explicitly in empirical studies. While this may be due partially to the high level of abstraction at which Althusser and others have discussed these concepts, it is also due to certain problems in the conceptualization of structural causality itself. In particular, the global notion of structural causality contains within itself several distinct forms of causality. In order to make the concept of structural causality accessible for empirical research, therefore, it needs to be broken down into this plurality of types of causation.

Modes of Determination and Models of Determination

What follows is a provisional attempt at elaborating a more differentiated schema of structural causality compatible with Marxist theory. The discussion will revolve around what I shall label modes of determination, that is, a series of distinct relationships of determination among the structural categories of Marxist theory and between those categories and the appearances of empirical investigation. These diverse modes of determination will then be organized into what can be called models of determination, that is, schematic representations of the complex interconnections of the various modes of determination involved in a given structural process. Such models of determination can be considered symbolic maps of what Althusserians have generally referred to as structured totalities.

Before discussing these diverse modes of determination, it must be emphasized that the schematic diagrams representing the models of determination are largely heuristic devices. They are designed to make explicit those linkages among categories which are either vague or implicit in theoretical statements. The diagrams themselves may appear to be highly mechanistic and rigid, not allowing for the dynamic movements which lie at the heart of a dialectical view of history. The intention, however, is to develop a way of representing the structural constraints and contradictions present in a given society which make that dynamic movement a non-random process.

At least six basic modes of determination can be distinguished within the global concept of structural causality: structural limitation, selection, reproduction/nonreproduction, limits of functional compatibility, transformation and mediation. While these modes of determination are highly interdependent, and thus a full understanding of any one of them presupposes an understanding of all, nevertheless it will be helpful to define each of them.

1. Structural Limitation: This constitutes a pattern of determination in which some social structure establishes limits within which some other structure or process can vary, and establishes probabilities for the specific structures or processes that are possible within those limits. That is, structural limitation implies that certain forms of the determined structure have been excluded entirely and some possible forms are more likely than others. This pattern of determination is especially important for understanding the sense in which economic structures ultimately determine political and ideological structures: economic structures set limits on the possible forms of political and ideological structures, and make some of those possible forms more likely than others, but they do not rigidly determine in a mechanistic manner any given form of political and ideological relations.

A good example of such structural limitation determination is the relationship between the economic structure and the forms of the state in feudal society. Given the nature of economic relations in classical feudalism—the control of the immediate means of production by the peasantry, the appropriation of the surplus product through coercion, the limited amount of surplus available, etc.—a representative democracy with universal suffrage was structurally impossible as a form of the state, i.e. it fell outside the structural limits established by economic structures. Within those limits, however, a fairly wide variety of state forms could occur, ranging from highly decentralized manorial systems of political rule, to relatively centralized Absolutist states. While the given structure of feudal economic relations may have shaped the likelihood of different specific forms of the feudal state, it did not determine uniquely which form occurred.

Structural limitation does not imply that every structurally possible form of the state (or other structure determined by a relation of structural limitation) is necessarily functional for the reproduction of the determining structure. We shall deal with this question in some detail below in the discussion of limits of functional compatibility as a mode of determination. For the moment it is simply important to note that the range of structurally limited possibilities and the range of functional possibilities do not necessarily coincide. In fact, part of our understanding of the concept of contradiction will hinge on the various ways in which a non-correspondence between structural and functional limitation is generated. More on this later.

2. Selection: Selection constitutes those social mechanisms that concretely determine ranges of outcomes, or in the extreme case specific outcomes, within a structurally limited range of possibilities. In a sense, selection can be seen as a form of second-order limitation: the setting of limits within limits. Much of the analysis of specific historical conjunctures can be thought of as investigations of the concrete patterns of selection that occur within broadly defined structural limits.

There are two complementary forms of selection, which can be termed positive and negative selections.⁸ Negative selection involves those mechanisms which exclude certain possibilities. Positive selection, on the other hand, involves mechanisms which determine specific outcomes among those that are possible. What is typically referred to as decision-making processes revolve around processes of positive selection. Taken together, positive and negative selection determine the concrete structural outcomes within limits determined by structural limitation.

A good illustration of selection can be seen in the interrelationship between the economic structure (forces and relations of production), the state structure and class struggle: the economic structure establishes limits of variation on both class struggle and the structure of the state: the state in turn acts as a selection mechanism on forms of class struggle, shaping those struggles within limits established by the underlying economic structure. These patterns of determination are illustrated in the simple model of determination in Figure 1.1

Figure 1.1 Illustration of Selection and Structural Limitation

In the case of feudal economic structures, this model of determination would read as follows: Given the basic structure of feudal economic relations, only certain forms of class struggle are possible, and different specific forms have greater probability than others. For example, the possibility of revolutionary socialist struggles organized through political parties is beyond the limits of variation imposed by feudal economic relations. Within the broad range of possible class struggles that could occur, the structure of the state was an important selection mechanism which determined whether struggles would take the form of land invasions, grain riots, millenarian movements, peasant flight from landlords, etc.

3. Reproduction/non-reproduction: Reproduction/non-reproduction is a more complex mode of determination than structural limitation and selection. To say that one structure functions to reproduce another implies that the reproducing structure prevents the reproduced structure from changing in certain fundamental ways. To say that the capitalist state, for example, reproduces capitalist economic relations means that it prevents those economic relations from changing into non-capitalist economic relations, and furthermore, that in the absence of such a reproduction process the economic structure potentially (but not inevitably) would change in such ways. Reproduction thus is also a kind of limiting process: it maintains the reproduced structure within certain limits of variation. The essential difference from structural limitation is that in the latter case there is no presumption that the determined structure would necessarily change in the absence of the specific structural limitation process, whereas in the case of reproduction such changes would normally occur. Reproduction/non-reproduction is symbolized in Figure 1.2.

Figure 1.2 Illustration of Reproduction/non-reproduction and Limitation as Modes of Determination

To say that the capitalist state is necessary for the reproduction of capitalist economic relations is not to say that the capitalist state always functions in a perfectly optimal way for the reproduction of those economic relations. It is quite possible for the effects of the state to be far less than optimal, and even under certain circumstances, for it to become non-reproductive. Reproduction/non-reproduction must therefore be understood as a variable relation of determination, not an absolute one.

4. Limits of functional compatibility: If the state is not always optimally functional for the reproduction of economic relations—indeed, if it is possible for the state to become non-reproductive—then we need some way of expressing the processes which determine which forms of the state will be functional and in what ways. This is what is meant by limits of functional compatibility: the mode of determination which determines which forms of the state will be reproductive and which non-reproductive. Stated in somewhat different terms, limits of functional compatibility determine what the effects of a given structure of the state will be on economic structures. This relation is illustrated in Figure 1.3 over the page.

Figure 1.3 Illustration of Limits of Functional Compatibility

As expressed in this model of determination, the economic structure both sets limits of variation on the structure of the state, and determines the extent to which it will itself be reproduced by the actual structure of the state which emerges. The crucial issue is that these two modes of determination do not necessarily coincide. The limits of functional compatibility are not intrinsically coordinated with the limits of structural variation. This is precisely what makes it possible for a form of the state to emerge which is non-reproductive of economic structures, and thus for a structural contradiction to exist between economic and political structures. When such a situation arises, either there will occur a fairly rapid transformation of the economic structures or the structures of the state will be altered in ways which make it once more reproductive. To a large extent, class struggle determines which of these outcomes will in fact occur.

A good example of limits of functional compatibility as a mode of determination is the relationship of the state to economic structures in the transition from feudalism to capitalism: In the early periods of the Absolutist state in Western Europe, these state structures could be considered reproductive of a limited development of the capitalist mode of production within a still largely feudal social structure. Perry Anderson describes this relation as follows: "The apparent paradox of Absolutism in Western Europe was that it fundamentally represented an apparatus for the protection of aristocratic property and privileges, yet at the same time the means whereby this protection was promoted could simultaneously ensure the basic interests of the nascent mercantile and manufacturing classes. … There was always a potential field of compatibility at this stage between the nature and programme of the Absolutist State and the operations of mercantile and manufacturing capital."⁹ As capitalism expanded, however, the Absolutist state increasingly became an obstacle to capital accumulation. Its feudal character constantly ended by frustrating and falsifying its promises for capital. In our terms, this structure of the state gradually became non-reproductive of the emerging economic relations even though it still fell within the structural limits of variation. The eventual result was the bourgeois revolutions: the resolution of the functional incompatibility of the Absolutist state through its violent transformation.

5. Transformation: Transformation refers to a mode of determination by which class struggle (practices) directly affect the processes of structural limitation, selection and reproduction/non-reproduction. Transformation is thus fundamental to the dialectical character of patterns of determination as understood in Marxist theory: class struggle, which is itself structurally limited and selected by various social structures, simultaneously reshapes those structures. The word simultaneously is important in this formulation: social structures do not first structurally limit and select class struggle, after which class struggle transforms those structures. Class struggle is intrinsically a process of transformation of structures, and thus the very process which sets limits on class struggle is at the same time transformed by the struggles so limited. This dialectical relationship between transformation and limitation is represented in Figure 1.4.

Figure 1.4 Illustration of Transformation as a Mode of Determination

It is especially important to understand the relationship between the concept of contradiction and the notion of transformation. In our discussion of limits of functional compatibility, I argued that the potential non-correspondence between structural limitation and limits of functional compatibility as modes of determination made possible the contradictions between structures. For that possibility to become actualized, however, class struggle must affect social structures through relations of transformation. Class struggles are, above all, struggles over social structures. This means that even if at a given point in time the structure of the state falls within the limits of functional compatibility determined by economic structures, there is no reason for that compatibility to be automatically reproduced over time. Class struggles transform economic relations, thus changing the reproduction requirements themselves; and class struggle transforms the state, thus making it potentially less reproductive over time. There is thus a systematic tendency for the contradictions between classes (class struggle) to generate contradictions between social structures (non-reproductive relations of determination).

To describe a mode of determination as transformation does not imply that a particular structure will in fact necessarily be transformed. In the case illustrated in Figure 1.4, the transformation of state structures is a consequence of class struggle, and it may well happen that the forces for the maintenance of existing structures of the state may be stronger than the forces for transformation. Defining a mode of determination as a relationship of transformation means that what is at issue is the transformation of structures, not that such transformation always takes place.

6. Mediation: Mediation is in some ways the most complex mode of determination. It defines a mode of determination in which a given social process shapes the consequences of other social processes. A mediating process must be distinguished from what is commonly called an intervening process or variable in sociology. This distinction is illustrated in Figure 1.5. An intervening variable is simply a variable which is causally situated between two other variables. X causes Y which in turn causes Z. A mediating variable, on the other hand, is one which shapes the very relationship between two other variables: Y causes the way in which X affects Z. In a sense a mediating process can be viewed as a contextual variable: processes of mediation determine the terrain on which other modes of determination operate.

Figure 1.5 The Difference between an Intervening Variable and a Mediating Variable

Mediation is especially important in analysing the relationship between class struggle and relations of structural limitation, selection and reproduction. For example, it is often argued that the bureaucratic structure of the capitalist state acts as an important selection mechanism in determining the actual activity of the state (policies, interventions, etc.). Class struggle decisively mediates this selection relation: the identical structures of the state will have very different consequences for state activity depending upon the relationship of class struggle to the state. When class struggles remain completely external to the institutions of the state, bureaucratic structures may effectively select state policies which optimally serve the interests of capital. When class struggles occur within the state apparatus itself—when civil service workers and teachers become unionized, state employees go on strike, welfare workers support their clients, etc.—the same formal state structure can select very different sorts of state interventions. This pattern is illustrated in Figure 1.6.

Figure 1.6 Illustration of Mediation as a Mode of Determination

Class struggles similarly mediate reproduction determinations. The extent to which a given state structure is reproductive of economic relations may be conditioned by the kinds of class struggles in the society. Where class struggle is very intense and very politicized, bourgeois democratic structures may prove quite unreproductive; where class struggle is very economistic and apolitical, the identical structures may function very reproductively.

Finally, class struggle also mediates relations of structural limitation. Structural limitation does not simply define those forms of the determined structures which are impossible; it also determines the relative likelihood of various possible forms of that structure. Class struggle can mediate this relationship and alter the pattern of probabilities. This kind of mediation is particularly important in periods of revolutionary transformation of structures. For example, after a socialist revolution, a variety of new state forms are structurally possible. To the extent that the working class has a history of active participation in bourgeois democratic struggles, the likelihood that a genuinely democratic form of the socialist state will emerge is increased.

If we take all six of these modes of determination together, we can create a model of determination of the relationships among economic structures, state structures, state interventions and class struggle. This model is presented in Figure 1.7. This model could of course be made more complex. Other elements could be added, such as the role of ideology. Or, more complex interconnections among the elements could be posed. For example, it could be argued that the structures of the state themselves mediate the transformation relationship between class struggle and the state (i.e. the structures of the state shape the extent to which they can be transformed by class struggle).¹⁰ In the present context, the issue is not so much the completeness of this specific model of determination, but the demonstration that this kind of model is a useful way of clarifying the relationships among elements in a theory.

Figure 1.7 Combined Model of Determination of Economic Structure, State Structure, State Policies and Class Struggle

Models of determination such as the one illustrated in Figure 1.7 should not be thought of as the end product of a serious historical investigation. Rather they are a prelude to such research. They are designed to lay