Essays in Sociological Theory

I

The Role of Ideas

in Social Action

THE SUBJECT of this paper has given rise to much controversy which has, on the whole, turned out to be strikingly inconclusive. It may be suggested that, in part at least, this is a result of two features of the discussion. On the one hand, sides have tended to be taken on the problem in too general terms. Ideas in general have been held either to have or not to have an important role in the determination of action. As opposed to this tendency, I shall attempt here to break the problem down into different parts, each of which fits differently into the analytical theory of action.

On the other hand, the discussion has, for the taste of the present writer, been altogether too closely linked to philosophical problems and has seldom been brought fairly into the forum of factual observation and theoretical analysis on the empirical level. This paper is to be regarded as a theoretical introduction to attempts of the latter sort.

I am far from believing that social or any other science can live in a kind of philosophical vacuum, completely ignoring all philosophical problems, but even though, as I have stated elsewhere,1 scientific and philosophical problems are closely interdependent, they are nevertheless at the same time independent and can be treated in relative abstraction from each other. Above all, from the fact that this paper will maintain that ideas do play an important part in the determination of action, it is not to be inferred that its author is committed to some kind of idealistic metaphysics of the sort from which it has so often been inferred that ideas must arise through some process of immaculate conception unsullied by social and economic forces or that they influence action by some automatic and mysterious process of self-realization or emanation without relation to the other elements of the social system.

The paper, then, will be devoted to the statement of a theoretical framework for the analysis of the role of ideas on an empirical, scientific basis. Without apologies, I shall start with an explicit definition of my subject matter. Ideas, for the purposes of this discussion, are concepts and propositions, capable of intelligible interpretation in relation to human interests, values and experience. So far as, qua ideas, they constitute systems, the relations between these concepts and propositions are capable of being tested in terms of a certain type of norm, that of logic.

The definition just given is so stated that it can serve as the definition of a variable in a system of interdependent variables. That is, it is a combination of logical universals to which many different particulars, the values of the variable, may be fitted. Since the present concern is wholly scientific, the sole important questions to be asked are three: 1. Do differences which are accurately ascertainabie obtain between the specific content of the ideas held by different individuals or groups in social systems at different times? 2. Is it possible to establish important relations between these differences and other observable aspects of, or events within, the same social systems? 3. Are these relations such that the ideas cannot be treated as a dependent variable, that is, their specific content deduced from knowledge of the values of one or more other observable variables in the same system? If all three of these questions can be answered in the affirmative, it may be claimed that ideas play an important role in the determination of social action in the only sense in which such a claim has meaning in science. Ideas would be an essential variable in a system of theory which can be demonstrated to work, to make intelligible a complex body of phenomena. Whether in an ultimate, ontological sense these ideas are real, or only manifestations of some deeper metaphysical reality is a question outside the scope of this paper.

Ideas obviously could not be treated as a variable in systems of social action unless their specific content varied from case to case. But besides the variations of specific content from case to case, it may be possible, as has been suggested, to divide them into certain broad classes which differ appreciably from one another in their relations to action. How these classes shall be defined, and how many there are, are pragmatic questions in the scientific sense; the justification of making a distinction between any two classes is that their members behave differently in their relations to action. Whether this is the case or not is a question of fact. I shall outline such a classification and then present an analysis of the role of each so as to demonstrate the importance of making the distinctions.

The first class may be termed existential ideas. The concepts which comprise such ideas are the framework for describing or analyzing entities, or aspects or properties of them, which pertain to the external world of the person who entertains the ideas, the actor. These entities either are or are thought to be existent at the time, to have existed, or to be likely to exist. The reference is to an external reality in some sense. The ideas involve existential propositions relative to some phase or phases of this reality, real or alleged. The most general type of norm governing existential ideas is that of truth.

Of existential, as of other ideas, it is convenient to distinguish two subclasses, the distinction between which is of cardinal importance. The one are empirical ideas, the concepts and propositions of which are, or are held to be, capable of verification by the methods of empirical science. All other existential ideas, on the other hand, I shall class together as nonempirical, regardless of the reasons why they are not scientifically verifiable.2

The second main class are what may be called normative ideas. These refer to states of affairs which may or may not actually exist, but in either case the reference is not in the indicative but in the imperative mood. If the state of affairs exists, insofar as the idea is normative the actor assumes an obligation to attempt to keep it in existence; if not, he assumes an obligation to attempt its realization at some future time. An idea is normative insofar as the maintenance or attainment of the state of affairs it describes may be regarded as an end of the actor. The states of affairs referred to may also be classified as empirical and nonempirical according to the above criteria.3

The first set of problems to be discussed concerns the role of empirical existential ideas. I think it fair to say that no branch of social science has been subjected to more thorough and rigorous analysis than this, so it forms an excellent starting point.4 The context in which this analysis has taken place is the range of problems surrounding the concept of the rationality of action in the ordinary sense of the maximization of efficiency or utility by the adaptation of means to ends. It is the sense of rationality which underlies most current analysis of technological processes in science, industry, medicine, military strategy and many other fields, which lies at the basis of economic theory, and much analysis of political processes regarded as processes of maintaining, exercising, and achieving power.

The common feature of all these modes of analysis of action is its conception as a process of attaining specific and definite ends by the selection of the most efficient means available in the situation of the actor. This, in turn, implies a standard according to which the selection among the many possible alternative means is made. There is almost universal agreement that the relevant basis of selection in this kind of case involves the actor’s knowledge of his situation, which includes knowledge of the probable effects of various possible alternative ways of altering it which are open to him. One of the necessary conditions of rationality of his action is that the knowledge should be scientifically valid.5

Valid empirical knowledge in this sense is certainly a system of ideas. It consists of concepts and propositions and their logical interrelations. Moreover, in all the above analyses of action, this knowledge is treated as a variable in the system of action; according to variations in its specific content, the action will be different. In explaining, above all, failure for the actual course of action to conform with a rational norm describing the best course, we continually refer to features of the store of knowledge of the actor. We say He did not know … with the implication that if he had, he would have acted differently, and He supposed erroneously that … , with the corresponding implication that if he had not been in error on the level of knowledge, he would also have acted differently. Thus, two of the coordinates of variation of knowledge which are relevant to its role in action are that in the direction of ignorance and of error. There is, for the attainment of any given end in any given situation, a certain minimum of valid knowledge which is adequate. If the knowledge actually falls short of this, if the actor is ignorant of any important features of the situation, or if his ideas are invalid, are in error, this is an adequate explanation of the failure of his action to be rational.

The analytical scheme in which the role of valid empirical knowledge in this sense has been most highly elaborated and conceptually refined is economic theory. Knowledge is a basically important variable in the system of economic theory, and he who would radically deny a role in action to ideas must find a satisfactory alternative explanation of all the uniformities of human action which have been established by two centuries of economic analysis, or demonstrate that the supposed uniformities do not exist.

But exactly the same thing is true of what we ordinarily call technology. The very processes of technological change to which many of our materialists assign so fundamental a role are in part a function of knowledge, i.e., of ideas, in exactly the same sense in which economic processes are. And there, far more than in the narrowly economic realm, knowledge has become a variable which we think of as to a high degree autonomous. For it takes, to a large extent, the form of theoretically systematized scientific knowledge rather than common sense. Surely the development of modern aniline dyes, the radio, or alloy steels, cannot be understood without reference to the essentially autonomous developments of science on which they depend.

Marxian theory has, however, classed technology among the material factors in social change, while ideas form part of the superstructure. Whence does this peculiar procedure derive? Two important sources of it may be noted. In the first place, Marxian theory has neither a rigorous concept of ideas, nor a classification of different kinds of ideas. Hence, when those ideas which Marxians habitually term ideologies behave differently from the scientific basis of technology, they tend to ignore the fact that the latter is also made up of ideas, and generalize the behavior of the former into that of ideas in general. Secondly, Marxian theory rests on an analytical basis essentially different from that which is the starting point of the present discussion. For it, the total concrete structure of the industrial enterprise is a factor, technology, social organization and all. The present attempt is to break down entities like this into simpler elements, the classification of which cuts across the Marxian dichotomy of ideal and material factors. There is no inherent reason why the Marxian choice of variables should be ultimate. The only scientific test as between it and another, such as that under discussion here, is the pragmatic one: which is the more illuminating in the understanding of certain empirical problems.

Every human society possesses a considerable stock of empirically valid knowledge, both of the nonhuman environment in which its members act, and of themselves, and of each other. That this knowledge is empirical and not theoretically systematized in the sense of modern science does not alter the fact. Moreover, a very large part of the action of the members of all societies is to be understood in terms of this knowledge. Lévy-Bruhl’s theory that primitive men do not think logically has, so far as it bears upon this point, been definitely discredited.6

But in addition to ideas which will stand the test of scientific validity, there are current in every society many ideas which in one respect or another diverge from this standard. So far as their reference is existential rather than normative or imaginative, the question arises as to what is the basis of this divergence. In answer to this question, a certain positivistic bias is very widely prevalent, and must be guarded against. It is the view, implicit or explicit, that divergence from the standard of empirical verifiability is always and necessarily a matter of empirical shortcomings in the sense that the ideas in question are not only, negatively, not verifiable, but that they can be shown to be positively wrong, that is that the basis of their unverifiability is ignorance or error, or both. This judgment clearly implies that there is available an adequate positive scientific standard by which to judge them.

At least in the field of empirically known systems of existential ideas, it can be stated with confidence that this class, which may be called un-scientific ideas, does not exhaust the departures from empirical verifiability, but that, in addition, there is a class of concepts and propositions which are unverifiable, not because they are erroneous, but because, as Pareto put it, they surpass experience. Such ideas as that the universe is divided between a good and an evil principle, that souls go through an unending series of reincarnations, that the only escape from sin is by divine grace, are in this category. They are nonscientific rather than unscientific.7

What, then, can be said about the role of such nonscientific ideas? So far as they are existential rather than normative or imaginative in character, there are certain formal similarities with empirical, scientifically valid ideas. The latter may, in one aspect, be considered as mechanisms of orientation of the actor to his situation. Insofar as man is treated as a purposive being, attempting rationally to attain ends, he cannot be considered as fully oriented to his situation until, among other things, he has adequate knowledge of the situation in the respects which are relevant to the attainment of the ends in question, or other functionally equivalent mechanisms.

But the role of existential ideas has so far been considered only in one context, that of the basis of choice of means to given ends. There is in addition the necessity of cognitive orientation of another sort, an answer to the problem of justification of the ends which are in fact pursued.8 If the justifications men give of why they should pursue their ultimate ends are systematically and inductively studied, one fact about them stands out. One very prominent component of all known comprehensive social systems of such justifications must be classed as nonempirical. The more the attempt is made to state the explicit or implicit major premises of such arguments clearly and sharply, the more evident it becomes that they are metaphysical rather than scientific propositions. This, I maintain, is true of all known social systems; whether it is ultimately possible to eliminate these nonempirical elements is not a relevant question in the present context.

But the mere demonstration that a certain class of phenomena exists does not prove that their description involves, for the purposes in hand, important variables. The question is not whether nonempirical existential ideas are always to be found in social systems, but whether important features of these social systems can be shown to be functions of variations in the content of these ideas. How is this problem to be attacked?

Most attempts in this field have been couched in terms of the historical or genetic method alone. Of course the only possible causal factors9 in the genesis of any particular state of affairs are components of particular antecedent states of affairs in the same sequence. But even then causal relationship can be demonstrated only by the use of general concepts and generalized knowledge of uniformities. The question here at issue does not touch the explanation of particular facts, but the establishment of uniformities. The only possible procedure by which this can be done in our field is comparative method which permits the isolation of variables. It is the strict logical counterpart of experiment. One important reason for the unsatisfactory character of the discussion of these problems revolving about Marxism is the fact that it has been almost uniformly couched in genetic, historical terms, as the Marxian theory itself is, and analytical generalizations as to the role of ideas cannot in principle be either proved or disproved by such a method. Hence the indeterminate issue of the controversy.

By far the most significant empirical studies available in this particular field are those of Max Weber in the sociology of religion.10 Weber was interested in a particular problem of historical imputation, that of the relative role of material factors and of the religious ideas of certain branches of Protestantism in the genesis of what he called rational bourgeois capitalism. But Weber’s methodological insight showed him that, in the absence of well-established general informities touching the role of ideas, it was hopeless to attack the problem by more and more elaborate genetic studies of the immediate historical background of modern capitalism. So he turned to the comparative method, the study of the influence of variations in the content of religious ideas.

A variable cannot, of course, be isolated unless other possibly important variables can, within a relevant range of variation, either be held constant or their independence demonstrated. Weber attempted to deal with this problem by showing that, in the different societies he treated, before the development of religious ideas in which he is interested, the state of the material factors and their prospective autonomous trends of development was, in the relevant respects, essentially similar. That is, for instance, in his three best worked out cases, those of China, India, and Western Europe, he attempted to estimate the relative favorableness or unfavorableness of the economic situations, the conditions of production, to a capitalistic development. The outcome of his studies in this respect was the judgment that there is a high degree of similarity in all three societies in this respect, with, if anything, a balance of favorableness in favor of India and China.11

But the fact remains that only in Europe did the development of capitalism actually take place. What accounts for the radically different outcomes in the three civilizations? It is a fact that the development of religious ideas in the three cases took quite different courses. In relation to this variable, an adequate range of variation to account for the differentiation is demonstrable, whereas in the case of the material factors it is not. This places the burden of proof on him who would advance a materialistic explanation. He must show that differentiating elements on this level were present of which Weber did not take account.

However, Weber did not leave his account of the role of religious ideas at this point. In terms of a more generalized conceptual scheme, the theory of action, or his "verstehende Soziologie," he analyzed certain mechanisms by which ideas can and do exert an influence on action. On the basis of this analysis, he worked out what is the probable effect on certain aspects of secular social life of adherence to each of the dominant systems of religious ideas, Confucianism, Hinduism and Protestantism, and found these deductions verified in that the actual facts corresponded, as seen in comparative perspective, with expectations in terms of reasoning from this hypothesis.

He further strengthened his case by working out, in an elaborate analysis of evidence from various sources in terms of his conceptual scheme, an understanding of many of the specific mechanisms of the process by which this influence has probably been exerted and verified this analysis in considerable detail.

The result of this very comprehensive comparative study in all these phases was not only to build up a strong case for his original historical thesis, that the ideas of ascetic Protestantism actually did play an important causal role in the genesis of modern capitalism. It also resulted in the formulation of a generalized theory of the role of nonempirical existential ideas in relation to action. It is this which is of primary interest here.

It was not Weber’s view that religious ideas constitute the principal driving force in the determination of the relevant kinds of action. This role is rather played by what he called religious interests. A typical example is the interest in salvation, an interest which has in turn a complex derivation from, among other things, certain stresses and strains to which individuals are sometimes subjected in social situations where frustration of the worldly ends seems inevitable and founded in the nature of things. But the mere interest in salvation alone is not enough. The question arises as to what kinds of specific action it will motivate. This, Weber’s comparative analysis shows, will be very different according to the structure of the existential religious ideas according to which the individual achieves cognitive orientation to the principal nonempirical problems he faces in his situation.

For example, on the basis of the generally imminent, pantheistic conception of divinity of Indian philosophy, and more specifically of the doctrines of Karma and Transmigration, to seek salvation in a radical sense through concrete achievement in worldly spheres would be meaningless. If such action contravened the traditional order, it would be reprehensible for that reason and set the actor back on his quest for salvation; if not, it could only generate more Karma and lead to endless rebirths. The only meaning of salvation is escape from the wheel of Karma in completely otherworldly mystical and ascetic exercises. For the Calvinist, on the other hand, mystical union with the divine is entirely excluded by the absolute transcendentality of God. He has been placed in this world to do God’s will in the building of the Kingdom. His eternal fate is settled by Predestination, but he can become certain of salvation through proving his faith by active labor in the vineyard, by doing God’s will.

The function of religious ideas is, in relation to the interest in salvation, to define the situation, to use W. I. Thomas’ term. Only by reference to these ideas is it possible to understand, concretely, what specific forms of action are relevant to attainment of salvation, or certainty of it. Weber succeeded in showing that rational, systematic, workmanlike labor in a worldly calling has had this significance to ardent believers in Calvinism and related religious movements, whereas it would be totally meaningless to a believer in Karma and Transmigration on a pantheistic background no matter how strong his interest in salvation. In this sense, the content of the religious ideas is a significant variable in the determination of the concrete course of action.

So far discussion has been confined to the role of existential ideas. These have been dealt with in two quite different contexts. Empirical ideas have been analyzed in their relation to the problem of selection of means according to the norm of rationality. Nonempirical ideas, on the other hand, have been treated in relation to the teleological problem of orientation of the actor, the justification of selection of ends to pursue. There is a gap between these two treatments which must now be filled. Selection of means has no significance except in relation to ends, while what has been called teleological orientation is equally meaningless unless there is, facing actors, a problem of choice between alternative ends.

Indeed the whole analytical procedure which has here been followed implies that a fundamental role in action is played by normative elements.12 In the first place, analysis of the underlying assumptions involved in treatment of empirical knowledge as an independent variable in the choice of means has shown that both a positive role of ends, and the existence of determinate relations of ends in a more or less well-integrated system are essential to the attribution of causal importance to knowledge. Rational action, in the sense of action guided by valid knowledge, is at the same time action which is normatively oriented. Similarly, the definition of the situation with reference to religious interests could have no meaning apart from the contention that it made a difference to the course of action what ends, among the various alternatives, were chosen.

Not only is action normatively oriented in the sense of pursuing ends, it is also subject to certain normative conditions, to rules which guide it. For instance, in pursuing the end of closing a profitable deal, a businessman may consider himself subject to the condition that it shall be done honestly. From some points of view, such rules may be considered themselves as ends, but they are not the immediate ends of the course of action under analysis. They appear rather as considerations limiting the acceptable range of alternative means, choice among which is to be guided by considerations of rational efficiency.

Now both ends and guiding norms involve a cognitive element, an element of ideas, however little the normative pattern may be exhausted in these terms. That such an element is involved may be brought out by considering the implications of the questions which are inevitably asked when we try to understand action in terms of such normative elements. What is the end … of a given course of action;—for instance, what is meant by making a profitable deal; or what do you mean … by the norm to which a course of action is subject,—for instance, by honesty in making a deal? It is obvious that the answers to all questions must be in the form of propositions, that is, of ideas. But in this case, ideas are in some sense imputed, not only to the sociological observer of action, but to the actor himself. It is a question not of what honesty means to the observer, but to the actor. It means, for instance, among other things, that he should not attempt to get the other party’s consent to the deal by making statements about his product as true which he knows to be false.

The essential point for present purposes is that, in so far as analysis of action in terms of orientation to ends and norms is scientifically useful at all, it implies two things: 1. That it is possible to impute to the actor with adequate precision for the purposes in hand, not only a will to attain certain ends or conform with certain norms, but a content of those ends and norms which is capable of formulation as a set of ideas. 2. That variations in this content stand in functional relations to the facts of the system of action other than the system of ideas of the actor.

Whether normative ideas constitute a variable independent of others in the system of action, is to be tested by essentially the same kind of procedure which was outlined in the case of Weber’s treatment of religious ideas. Weber himself showed that it is a variable in part dependent on nonempirical ideas. This would make it, insofar, relatively independent of material factors. But at the same time, there is no essential reason why an important range of variability independent in turn of metaphysical and religious ideas does not exist.

The foregoing analysis of the role of ideas in action has been presented in general terms, with appeal to generally known facts, and to two bodies of technically specific evidence, that employed in economic and technological analyses of rational action, and in Max Weber’s studies of the role of religious ideas. It is impossible within the limits of such a paper to detail any significant sample of the enormous mass of empirical evidence, from these and other sources, which supports the main lines of the analysis. I should not, however, like to close without mentioning one other set of considerations which seem to be greatly to strengthen the case for my thesis.

It has already been remarked that demonstration of causal relationship in any particular historical sequence cannot be derived from observation of the facts of that particular sequence alone; it is necessary to be able to apply to these facts generalized theoretical knowledge derived from comparative analysis of a series of different particular situations. Only by this procedure can variables be isolated and the functional relationships of their values be worked out and verified.

Hence the problem of the role of ideas cannot be treated adequately in terms of ad hoc recitation of the facts of certain examples. It involves systematic theoretical analysis of action, of the relation of the same variables to many different concrete situations. In both the two cases which have been most fully analyzed above, the theorems relative to the role of ideas are not isolated, but are an integral part of more comprehensive bodies of theory. Thus the analysis of the role of empirical ideas in rational action may be regarded as an application to this particular problem of one of the most highly developed bodies of generalized theoretical knowledge in the social field, economic theory. This has the effect of greatly strengthening the evidence for the particular theorem, for it is verified not only directly with reference to the kind of facts here discussed, but indirectly in that it is logically interdependent with all the other theorems of economic science. So far as they are mutually interdependent, the facts which support any one serve also to verify the others.

In the case of religious ideas, there is no such generally recognized and used body of theory into which the results of Weber’s empirical studies can be fitted. But it has already been remarked that Weber himself did in fact develop a body of such theory to a high degree of systematization in the course of his studies. The theoretical structure he developed is, in his own work, applicable to, and verified in terms of, many other problems than that of the role of ideas. But more than this. My own recently published analysis of certain phases of the development of social theory in the last generation13 has shown that in these theoretical results Weber converged with remarkable exactitude and detail on a structure in all essentials like that developed by other theorists with quite different starting points and empirical interests. In particular Durkheim, whose interest was not specifically in the problem of the role of ideas at all, but in the basis of social solidarity, arrived at a set of categories in the field of religion which corresponds point for point with that of Weber. Weber’s theoretical analysis of the role of nonempirical ideas is in fact part of a much broader system of analytical social theory, the emergence of which can be traced in a number of sources quite independent of Weber.

Moreover not only did Weber, Durkheim, and others converge on this particular part of a theoretical system, dealing mainly with religion, but as, among other things, very important parts of the work of both men show, this common scheme of the sociology of religion is in turn part of a still broader theoretical system which includes the economic and technological analysis of the role of empirical knowledge in relation to rationality of action. Both sets of problems belong together, and are part of the same more generalized analysis of human action.14

To conclude. The actual controversy over the role of ideas has been much more a battle of the implications of rival philosophical and other extrascientific points of view than it has been the result of careful, empirical analysis of the facts. I suggest that leaving these philosophical considerations aside and embarking on such careful study will very probably result in much reduction of the difference of opinion. The thesis put forward in this paper seems to me not only to fit very important bodies of well established and carefully analyzed facts. It also fits in with a body of generalized theoretical knowledge of human social action, which has already accumulated a heavy weight of scientific authority behind it in a. large number of different factual fields. This seems to me to justify taking the positive role of ideas as a working hypothesis for further empirical research. The result of such research will, as always, be to modify the formulations of the problem, and of theorems which appear to be verified, from forms which seemed acceptable when the research process began. But such modification is not refutation of a theory; it is the normal course of scientific progress to which the superseded theory itself makes an essential contribution.

1 The Structure of Social Action, 20 ff, New York, 1937.

2 This residual category is formulated for the immediate purposes in hand and its use is not to be held to imply that no distinctions between subclasses of nonempirical ideas are important for any other purposes.

3 There is a third class of ideas which may be called imaginative. The content of these refers to entities which, are neither thought to be existent nor does the actor feel any obligation to realize them. Examples would be a Utopia which is not meant as defining a program of action, or the creation of an entirely fictitious series of situations in a novel. At least the most obvious significance of such ideas in relation to action is as indices of the sentiments and attitudes of the actors rather than as themselves playing a positive role. To inquire whether indirectly they do play a role would raise questions beyond the scope of this paper and they will be ignored in the subsequent discussion. They are mentioned here only to complete the classification.

4 Much of this analysis is discussed in The Structure of Social Action. See esp. chap. 4, 161 ff.; chap. 5, 180 ff.; chap. 9, 344 ff.

5 Efficiency involves choice among two or more alternative ways of attaining an end. The validity of knowledge alone is not a sufficient criterion to determine the relative efficiency of the different alternatives. Statement of the other necessary criteria would involve difficult questions far beyond the scope of this paper.

6 See especially B. Malinowski, "Magic, Science and Religion," ed. by Robert Redfield (Free Press, Glencoe, III., 1948).

7 I do not wish to maintain that this distinction possesses ontological significance. To do so would be to alter the plane of the discussion of this paper, which has set out to adhere to the scientific level. Inevitably, the basis of the distinction must be found in current standards of scientific methodology. From this point of view, a nonempirical proposition is one, not only which cannot, because of practical difficulties, be verified with present techniques, but which involves, in the strict operational sense, meaningless questions, questions which cannot, in the present state of our scientific and methodological knowledge, be answered by a conceivable operation or combination of them. Whether, at some future time, a completely positivistic philosophy will be capable of demonstration is another question. But I should like to point out that objection to this distinction usually involves the positivistic philosophical position; it is arbitrarily laid down that all departures from the standard of empirical verifiability must be in terms of ignorance and error. The position taken here is such that the burden of proof is on him who would object to the distinction. It is his task to show empirically that what have here been called unscientific and nonscientific ideas in fact do not stand in different relations to action. This shifts the argument from the methodological to the factual plane.

8 On this problem, see The Structure of Social Action, chap. 5, 205ff.

9 Factors in the sense of concrete events or states of affairs, or parts or aspects of them, not of generalized, analytical elements like mass or ideas. The two are often confused. See The Structure of Social Action, chap. 16, 610ff.

10 Gesammelte Aufsatze zur Religionssoziologie. 3 vols. The most comprehensive secondary accounts in English are in L. L. Bennion, Max Weber’s Methodology, and The Structure of Social Action, chaps. 14 and 15.

11 This part of Weber’s work was not methodologically completely rigorous, but allowance for this does not affect his general conclusions.

12 The problem of the significance of normative elements in action is extensively treated throughout The Structure of Social Action.

13 The Structure of Social Action. See esp. chaps. 17 and 18:

14 The case of Pareto is particularly interesting in this respect. Pareto has been very widely heralded as one of the major prophets of anti-intellectualism, as one of the principal social theorists who radically denied an important role to ideas. Did he not lay particular emphasis on nonlogical action?

To those who have followed the above argument closely, two facts should make one suspicious of this interpretation. First, Pareto was well trained in economic theory, and insofar as he attributes importance to the elements it analyzes, to the interests, he must, ipso facto attribute importance to ideas. But not only this; he makes the conception of rationality in precisely the technological-economic sense the starting point of his own broader analysis of action. Nonlogical action is precisely action insofar as it cannot be understood in terms of this standard of rationality.

It turns out on analysis that his main theoretical scheme as such involves no theorem at all as to the role of ideas, except empirical existential ideas. His actual thesis is, not that other ideas have no role, but that beyond the range of applicability of this kind of conception of rationality or logical action, the ideas which do have a role cannot claim empirical scientific validity. But in his actual treatment there is much evidence that he attributes a very important role to nonempirical existential and normative ideas. This conclusion is strongly confirmed by the circumstances that Pareto’s general conceptual scheme converges in all essential respects with the broader more general theoretical structure of which I have spoken, which may also be found in the works of Max Weber and Durkheim. It would indeed be strange, in the light of this fact, if there were a radical disagreement between them on so basic a theorem as that of the role of ideas.

The interpretation of Pareto as a radical anti-intellectualist appears to arise mainly from two sources. On the one hand, there is, in the formulation of his approach to the analysis of action a source of anti-intellectualistic bias (The Structure of Social Action, 272, Note 1), which does not however play any substantive part in the main theoretical structure. This is indicative of the fact that his own theory was imperfectly integrated; and there are, underlying this, currents of thought which tend in this direction. But more important than this basis in Pareto’s own work is the fact that the general majority of Pareto’s interpreters have approached his work with an interpretive bias which enormously exaggerates the importance of these tendencies. The source of this bias is the fact that interpretation has been predominantly in terms of a positivistic system of general social theory. See The Structure of Social Action, chaps. 5-7.

II

The Professions

and Social Structure

COMPARATIVE STUDY of the social structures of the most important civilizations shows that the professions occupy a position of importance in our society which is, in any comparable degree of development, unique in history. Perhaps the closest parallel is the society of the Roman Empire where, notably, the Law was very highly developed indeed as a profession. But even there the professions covered a far narrower scope than in the modern Western world. There is probably in Rome no case of a particular profession more highly developed than in our own society, and there was scarcely a close analogy to modern engineering, medicine or education in quantitative importance, though all of them were developed to a considerable degree.

It seems evident that many of the most important features of our society are to a considerable extent dependent on the smooth functioning of the professions. Both the pursuit and the application of science and liberal learning are predominantly carried out in a professional context. Their results have become so closely interwoven in the fabric of modern society that it is difficult to imagine how it could get along without basic structural changes if they were seriously impaired.

There is a tendency to think of the development and application of science and learning as a socially unproblematical process. A vague sort of curiosity and beyond that mere possession of the requisite knowledge are held to be enough. This is evidenced by the air of indignant wonder with which technologically minded people sometimes cite the fact that actual technical performance is well below the theoretical potentialities of 100 percent efficiency. Only by extensive comparative study does it become evident that for even a moderate degree either of the development or the application of science there is requisite a complex set of social conditions which the technologically minded seldom think of, but incline to take for granted as in the nature of things. Study of the institutional framework within which professional activities are carried on should help considerably to understand the nature and functions of some of these social constants.

The professions do not, however, stand alone as typical or distinctive features of modern Western civilization. Indeed, if asked what were the most distinctive features, relatively few social scientists or historians would mention the professions at all. Probably the majority would unhesitatingly refer to the modern economic order, to capitalism, free enterprise, the business economy, or however else it is denominated, as far more significant. Probably the only major exception to this would be the relatively prominent attention given to science and technology, but even these would not be thought of mainly in relation to the professional framework, but rather as handmaidens of economic interests.

Not only is there a tendency to empirical concentration on the business world in characterizing this society, but this is done in terms which tend to minimize the significance of the professions. For the dominant keynote of the modern economic system is almost universally held to be the high degree of free play it gives to the pursuit of self-interest. It is the acquisitive society, or the profit system as two of the most common formulas run. But by contrast with business in this interpretation the professions are marked by disinterestedness. The professional man is not thought of as engaged in the pursuit of his personal profit, but in performing services to his patients or clients, or to impersonal values like the advancement of science. Hence the professions in this context appear to be atypical, to some even a mere survival of the mediaeval guilds. Some think that these spheres are becoming progressively commercialized, so that as distinctive structures they will probably disappear.

There are various reasons for believing that this way of looking at the essence of modern society is a source of serious bias in the sociological interpretation of the situation. The fact that the professions have reached a uniquely high level of development in the same society which is also characterized by a business economy suggests that the contrast between business and the professions which has been mainly started in terms of the problem of self-interest, is not the whole story. Possibly there are elements common to both areas, indeed to our whole occupational system, which are at least as important to their functioning as is self-interest to business, disinterestedness to the professions. The concrete interpenetration of the two, as exemplified in the role of engineers and lawyers in the conduct of business enterprises would suggest that. The study of the professions, by eliminating the element of selfinterest in the ordinary sense, would seem to offer a favorable approach to the analysis of some of these common elements. This paper will deal with three of them which seem to be of particular importance to the modern occupational structure as a whole, including business, the professions, and government.

But before entering on their discussion a further point may be noted. In much of traditional thought about human action the most basic of all differences in types of human motivation has been held to be that between egoistic and altruistic motives. Correlative with this there has been the tendency to identify this classification with the concrete motives of different spheres of activity: the business man has been thought of as egoistically pursuing his own self-interest regardless of the interests of others, while the professional man was altruistically serving the interests of others regardless of his own. Seen in this context the professions appear not only as empirically somewhat different from business, but the two fields would seem to exemplify the most radical cleavage conceivable in the field of human behavior.

If it can be shown that the difference with respect to self-interest does not preclude very important institutional similarities in other respects, a further possibility suggests itself. Perhaps even in this respect the difference is not so great as our predominantly economic and utilitarian orientation of thought would lead us to believe. Perhaps even it is not mainly a difference of typical motive at all, but one of the different situations in which much the same commonly human motives operate. Perhaps the acquisitiveness of modern business is institutional rather than motivational.

Let us, however, turn first to the elements of the common institutional pattern of the occupational sphere generally, ignoring for the moment the problem of self-interest. The empirical prominence of industrial technology calls attention immediately to one of them. Industrial technology in the modern world has become to a large extent applied science. One of the dominant characteristics of science is its rationality in the sense which is opposed to traditionalism. Scientific investigation, like any other human activity when viewed in terms of the frame of reference of action, is oriented to certain normative standards. One of the principal of these in the case of science is that of objective truth. Whatever else may be said of this methodologically difficult conception, it is quite clear that the mere fact that a proposition has been held to be true in the past is not an argument either for or against it before a scientific forum. The norms of scientific investigation, the standards by which it is judged whether work is of high scientific quality, are essentially independent of traditional judgments.

What is true of science as such is in turn true of its practical applications. Insofar as a judgment of what is the best thing to do rests on scientific considerations, whether it be in technology or in medicine, the merely traditional way of doing it as the fathers have done it, fails to carry normative authority. The relevant questions are, rather, objective,—what are the facts of the situation and what will be the consequences of various alternative procedures? Furthermore rationality in this sense extends far beyond the boundaries of either pure or applied science in a technical sense. The business man, the foreman of labor, and not least the non-scientific professional man such as the lawyer, is enjoined to seek the best, the most efficient way of carrying on his function, not to accept the time-honored mode. Even though the range of such rational considerations be limited by ends which are institutionally kept outside discussion, as the financial well-being of the enterprise or, as in the law, certain accepted principles of the Common Law, still, within the limits, traditionalism is not authoritative.

It should be noted that rationality in this sense is institutional, a part of a normative pattern: it is not a mode of orientation which is simply natural to men. On the contrary comparative study indicates that the present degree of valuation of rationality as opposed to traditionalism is rather unnatural in the sense that it is a highly exceptional state. The fact is that we are under continual and subtle social pressures to be rationally critical, particularly of ways and means. The crushing force to us of such epithets as stupid and gullible is almost sufficient indication of this. The importance of rationality in the modern professions generally, but particularly in those important ones concerned with the development and application of science serves to emphasize its role in the society at large. But this is even more impressively the case since here it is divorced from the institutionalized expectation of selfinterest typical of the contractual pattern of business conduct.

In quite a different way the role of the professions serves to bring out a second widely pervasive aspect of our general occupational pattern. There is a very important sense in which the professional practitioner in our society exercises authority. We speak of the doctor as issuing orders even though we know that the only penalty for not obeying them is possible injury to the patient’s own health. A lawyer generally gives advice but if the client knew just as well what to do it would be unnecessary for him to consult a lawyer. This professional authority has a peculiar sociological structure. It is not as such based on a generally superior status, as is the authority a Southern white man tends to assume over any Negro, nor is it a manifestation of superior wisdom in general or of higher moral character. It is rather based on the superior technical competence of the professional man. He often exercises his authority over people who are, or are reputed to be, his superiors in social status, in intellectual attainments or in moral character. This is possible because the area of professional authority is limited to a particular technically defined sphere. It is only in matters touching health that the doctor is by definition more competent than his lay patient, only in matters touching his academic specialty that the professor is superior, by virtue of his status, to his student. Professional authority, like other elements of the professional pattern, is characterized by specificity of function. The technical competence which is one of the principal defining characteristics of the professional status and role is always limited to a particular field of knowledge and skill. This specificity is essential to the professional pattern no matter how difficult it may be, in a given case, to draw the exact boundaries of such a field. As in all similar cases of continuous variation, it is legitimate to compare widely separated points. In such terms it is obvious that one does not call on the services of an engineer to deal with persistent epigastric pain, nor on a professor of Semitic languages to clarify a question about the kinship system of a tribe of Australian natives. A professional man is held to be an authority only in his own field.

Functionally specific technical competence is only one type of case in which functional specificity is an essential element of modern institutional patterns. Two others of great importance may be mentioned to give a better idea of the scope of this institutional element. In the first place, in the classic type of contractual relationship, rights and obligations are specifically limited to what are implicitly or explicitly the terms of the contract. The burden of proof that it is really owed, is on him who would exact an obligation, while in many other types of relationship the opposite is true, the burden of proof that it is not due is on the one who would evade an obligation. Thus in an ordinary case of commercial indebtedness, a request for money on the part of one party will be met by the question, do I owe it? Whether the requester needs the money is irrelevant, as is whether the other can well afford to pay it. If, on the Other hand, the two are brothers, any contractual agreements are at least of secondary importance; the important questions are, on the one hand, whether and how urgently the one needs the money, on the other whether the second can afford it. In the latter connection it comes down to a question of the possible conflict of this with what are recognized as higher obligations. In the commercial case it is not necessary even to cite what other possible uses for the money may be involved, the question is only why it should be paid. In the kinship case the question is immediately why the request should not be met, and the only satisfactory answer is the citing of higher obligations with which it conflicts. Commercial relations in our society are predominantly functionally specific, kinship relations, functionally diffuse.

Similarly if a doctor asks a patient a question the relevant reaction is to ask why he should answer it, and the legitimizing reply is that the answer is necessary for the specific function the doctor has been called upon to perform, diagnosing an illness for instance. Questions which cannot be legitimized in this way would normally be resented by the patient as prying into his private affairs. The patient’s wife, on the other hand, would, according to our predominant sentiments, be entitled to an explanation as to why a question should not be answered. The area of the marriage relationship is not functionally specific, but diffuse.

Functional specificity is also essential to another crucial pattern of our society, that of administrative office. In an administrative or bureaucratic hierarchy, authority is distributed and institutionalized in terms of office. By virtue of his office a man can do things, particularly in the sense of giving orders to others, which in his private capacity he would not be allowed to do at all. Thus the treasurer of a