THE PROBLEM OF CLASSES

l. We here mean by classes those social phenomena with which we are all familiar—social entities which we observe but which are not of our making. In this sense every social class is a special social organism, living, acting, and suffering as such and in need of being understood as such.1 Yet the concept of class occurs in the social sciences in still another meaning—a meaning shared with many other sciences. In this sense it still corresponds to a set of facts, but not to any specific phenomenon of reality. Here it becomes a matter of classifying different things according to certain chosen characteristics. Viewed in this sense, class is a creation of the researcher, owes its existence to his organizing touch. These two meanings are often annoyingly mixed up in our social-science thinking, and we therefore emphasize what should be self-evident, namely, that there is not the slightest connection between them as a matter of necessity. Whenever there is any actual coincidence of their contents, this is either a matter of chance, or—if it is really more than that—must be demonstrated, generally or specifically, by means of pertinent rules of evidence. It can never be assumed as a matter of course. This word of caution applies especially to the field in which theoretical economics operates. In theoretical economics, a landlord—the very term implies the confusion we oppose—is anyone who is in possession of the services of land. But not only do such people not form a social class. They are divided by one of the most conspicuous class cleavages of all. And the working class, in the sense of economic theory, includes the prosperous lawyer as well as the ditch-digger. These classes are classes only in the sense that they result from the scholar’s classification of economic subjects. Yet they are often thought and spoken of as though they were classes in the sense of the social phenomenon we here seek to investigate. The two reasons that explain this situation actually make it more troublesome than it would otherwise be. There is, first, the fact that the characteristic by which the economist classifies does have some connection with the real phenomenon. Then there is the fact that the economic theorist finds it exceedingly difficult to confine himself strictly to his problems, to resist the temptation to enliven his presentation with something that fascinates most of his readers—in other words, to stoke his sputtering engine with the potent fuel of the class struggle. Hence the amusing circumstance that some people view any distinction between economic theory and the facts of social class as evidence of the most abysmal failure to grasp the point at issue; while others see any fusion of the two as the most abysmal analytical blundering. Hence, too, the fact that the very term class struggle, let alone the idea behind it, has fallen into discredit among the best minds in science and politics alike—in much the same way that the overpowering impression of the Palazzo Strozzi loses so much by its inescapable juxtaposition with the frightful pseudo-architecture of modern apartment houses.

2. Of the many sociological problems which beset the field of class theory—the scientific rather than the philosophical theory, the sociological rather than the immediately economic—four emerge distinctly. First, there is the problem of the nature of class (which is perhaps, and even probably, different for each individual scientific discipline, and for each purpose pursued within such a discipline)—and, as part of this problem, the function of class in the vital processes of the social whole. Fundamentally different, at least theoretically, is the problem of class cohesion—the factors that make of every social class, as we put it, a special living social organism, that prevent the group from scattering like a heap of billiard balls. Again fundamentally distinct is the problem of class formation—the question of why the social whole, as far as our eye can reach, has never been homogeneous, always revealing this particular, obviously organic stratification. Finally, we must realize—and we shall presently revert to this point—that this problem is again wholly different from the series of problems that are concerned with the concrete causes and conditions of an individually determined, historically given class structure—a distinction that is analogous to that between the problem of the theory of prices in general and problems such as the explanation of the level of milk prices in the year 1919.

We are not, at this point, seeking a definition that would anticipate the solution of our problem. What we need, rather, is a characteristic that will enable us, in each case, to recognize a social class and to distinguish it from other social classes—a characteristic that will show on the surface and, if possible, on the surface alone; that will be as clear or as fuzzy as the situation itself is at first glance. Class is something more than an aggregation of class members. It is something else, and this something cannot be recognized in the behavior of the individual class member. A class is aware of its identity as a whole, sublimates itself as such, has its own peculiar life and characteristic “spirit.” Yet one essential peculiarity—possibly a consequence, possibly an intermediate cause—of the class phenomenon lies in the fact that class members behave toward one another in a fashion characteristically different from their conduct toward members of other classes. They are in closer association with one another; they understand one another better; they work more readily in concert; they close ranks and erect barriers against the outside; they look out into the same segment of the world, with the same eyes, from the same viewpoint, in the same direction. These are familiar observations, and among explanations which are traditionally adduced are the similarity of the class situation and the basic class type.

To this extent the behavior of people toward one another is a very dependable and useful symptom of the presence or absence of class cohesion among them—although it does not, of course, go very deeply, let alone constitute a cause. Even more on the surface—a symptom of a symptom, so to speak, though it hints at a far-reaching basic orientation—is the specific way in which people engage in social intercourse. These ways are decisively influenced by the degree of “shared social a priori,” as we might say with Simmel. Social intercourse within class barriers is promoted by the similarity of manners and habits of life, of things that are evaluated in a positive or negative sense, that arouse interest. In intercourse across class borders, differences on all these points repel and inhibit sympathy. There are always a number of delicate matters that must be avoided, things that seem strange and even absurd to the other class. The participants in social intercourse between different classes are always on their best behavior, so to speak, making their conduct forced and unnatural. The difference between intercourse within the class and outside the class is the same as the difference between swimming with and against the tide. The most important symptom of this situation is the ease or difficulty with which members of different classes contract legally and socially recognized marriages. Hence we find a suitable definition of the class—one that makes it outwardly recognizable and involves no class theory—in the fact that intermarriage prevails among its members, socially rather than legally.2 This criterion is especially useful for our purposes, because we limit our study to the class phenomenon in a racially homogeneous environment, thus eliminating the most important additional impediment to intermarriage.3

3. Our study applies to the third of the four questions we have distinguished—to the others only to the extent that it is unavoidable. Let us begin by briefly discussing three difficulties in our way—a consideration of each of them already constituting an objective step toward our goal.

First: We seek to interpret the class phenomenon in the same sense in which we understand social phenomena generally, that is, as adaptations to existing needs, grasped by the observer—ourselves—as such. We shall pass over the logical difficulties inherent in even this simple statement, such as whether it is admissible to apply our own conceptual modes to cultures remote from us. There is also the question of the extent to which the condition of culturally primitive peoples in our own time may be taken as a clue to the past state of modern civilized peoples, and the even more important question of the extent to which historical data are at all valid for theoretical purposes. One difficulty, however, we must face. Unless specifically proven, it is an erroneous assumption that social phenomena to which the same name has been applied over thousands of years are always the same things, merely in different form. This is best seen in the history of social institutions. Anyone will realize that common ownership of land in the ancient Germanic village community—supposing, for the moment, that its existence had been proven—is something altogether different from common land ownership in present-day Germany. Yet the term ownership is used as though it always implied the same basic concept. Obviously this can be true only in a very special sense, to be carefully delimited in each case. When taken for granted, it becomes a source of one-sided and invalid constructions. The fact that there may occur in the language of law and life of a given period expressions that we regard as equivalent to our chosen concept, proves nothing, even when those expressions were actually used in an equivalent sense. Similarly, the actuality of the institution we call marriage has changed so greatly in the course of time that it is quite inadmissible to regard that institution always as the same phenomenon, from a general sociological viewpoint and without reference to a specific research purpose. This does not mean that we renounce the habit, indispensable in analysis, of seeking, wherever possible, the same essential character in the most diverse forms. But the existence of that character must be a fact, its establishment the result of study, not a mere postulate. This applies to our problem as well. When we speak of “the” class phenomenon and take it to mean that group differences in social values, found everywhere, though under varying conditions, are everywhere explained by the same theory, that is not even a working hypothesis, but merely a method of presentation in which the result is anticipated—a result that has meaning only from the viewpoint of the particular theory in question. “Master classes,” for example, do not exist everywhere—if, indeed, the concept of “master” has a precise content at all.

Second: The class membership of an individual is a primary fact, originally quite independent of his will. But he does not always confirm that allegiance by his conduct. As is well known, it is common for nonmembers of a class to work with and on behalf of that class, especially in a political sense, while members of a class may actually work against it. Such cases are familiar from everyday life—they are called fellow travelers, renegades, and the like. This phenomenon must be distinguished, on the one hand, from a situation in which an entire class, or at least its leadership, behaves differently from what might be expected from its class orientation; and, on the other hand, from a situation in which the individual, by virtue of his own functional position, comes into conflict with his class. There is room for differences of opinion on these points. For example, one may see in them aberrations from the normal pattern that hold no particular interest, that have no special significance to an understanding of society, that are often exceptions to the rule more apparent than real. Those who view the class struggle as the core of all historical explanation will generally incline to such opinions and seek to explain away conflicting evidence. From another viewpoint, however, these phenomena become the key to an understanding of political history—one without which its actual course and in particular its class evolution become altogether incomprehensible. To whatever class theory one may adhere, there is always the necessity of choosing between these viewpoints. The phenomena alluded to, of course, complicate not only the realities of social life but also its intellectual perception. We think that our line of reasoning will fully answer this question, and we shall not revert to it.

Third: Every social situation is the heritage of preceding situations and takes over from them not only their cultures, their dispositions, and their “spirit,” but also elements of their social structure and concentrations of power. This fact is of itself interesting. The social pyramid is never made of a single substance, is never seamless. There is no single Zeitgeist, except in the sense of a construct. This means that in explaining any historical course or situation, account must be taken of the fact that much in it can be explained only by the survival of elements that are actually alien to its own trends. This is, of course self-evident, but it does become a source of practical difficulties and diagnostic problems. Another implication is that the coexistence of essentially different mentalities and objective sets of facts must form part of any general theory. Thus the economic interpretation of history, for example, would at once become untenable and unrealistic—indeed, some easily demolished objections to it are explained from this fact—if its formulation failed to consider that the manner in which production methods shape social life is essentially influenced by the fact that the human protagonists have always been shaped by past situations. When applied to our problem, this means, first, that any theory of class structure, in dealing with a given historical period, must include prior class structures among its data; and then, that any general theory of classes and class formation must explain the fact that classes coexisting at any given time bear the marks of different centuries on their brow, so to speak—that they stem from varying conditions. This is in the essential nature of the matter, an aspect of the nature of the class phenomenon. Classes, once they have come into being, harden in their mold and perpetuate themselves, even when the social conditions that created them have disappeared.

In this connection it becomes apparent that in the field of our own problem this difficulty bears an aspect lacking in many other problems. When one seeks to render modern banking comprehensible, for example, one can trace its historical origins, since doubtless there were economic situations in which there was no banking, and others in which the beginnings of banking can be observed. But this is impossible in the case of class, for there are no amorphous socities in this sense—societies, that is, in which the absence of our phenomenon can be demonstrated beyond doubt. Its presence may be more or less strongly marked, a distinction of great importance for our solution of the class problem. But neither historically nor ethnologically has its utter absence been demonstrated in even a single case, although there has been no dearth either of attempts in that direction (in eighteenth-century theories of culture) or of an inclination to assume the existence of classless situations.4 We must therefore forego any aid from this side, whatever it may be worth,5 though the ethnological material nevertheless retains fundamental significance for us. If we wanted to start from a classless society, the only cases we could draw upon would be those in which societies are formed accidentally, in which whatever class orientations the participants may have either count for nothing or lack the time to assert themselves—cases, in other words, like that of a ship in danger, a burning theater, and so on. We do not completely discount the value of such cases, but quite apparently we cannot do very much with them. Any study of classes and class situations therefore leads, in unending regression, to other classes and class situations, just as any explanation of the circular flow of the economic process always leads back, without any logical stopping point, to the preceding circular flow that furnishes the data for the one to follow. Similarly—though less closely so—analysis of the economic value of goods always leads back from a use value to a cost value and back again to a use value, so that it seems to turn in a circle. Yet this very analogy points to the logical way out. The general and mutual interdependence of values and prices in an economic situation does not prevent us from finding an all-encompassing explanatory principle; and the fact of regression in our own case does not mean the non-existence of a principle that will explain the formation, nature, and basic laws of classes—though this fact naturally does not necessarily furnish us with such a principle. If we cannot derive the sought-for principle from the genesis of classes in a classless state, it may yet emerge from a study of how classes function and what happens to them, especially from actual observation of the changes in the relationship of existing classes to one another and of individuals within the class structure—provided it can be shown that the elements explaining such changes also include the reason why classes exist at all.