Key Problems of Sociological Theory

John Rex

• Publisher: Routledge
• Year: 2006
• eBook ISBN: 9781134685899
• Topic: Social Sciences
• Subtopic: Sociology
• Index: Social Sciences

This is Volume of VII twenty-two on a series on Social Theory and Methodology. Originally published in 1961, this book was written because of the author's sense of the inadequacies of a sociological tradition dominated by empiricism and positivism. The tradition of empiricism leads to attempts to settle public issues by reference to crude ad hoc generalisations. So "right-wing" facts are refuted by "left-wing" facts and vice versa, and in the argument which ensues nothing becomes clear except the value-biasses which the authors seek desperately to conceal. The tradition of positivism on the other hand fails in refusing to interpret observed correlations of fact except in terms of the natural sciences. So the sociologist often appears to have derived little more insight through his precise methods than the untutored layman is able to do through trusting to intuition and common-sense.

I
SOCIOLOGY AS A SCIENCE

THE claim of sociology is that the disciplines of scientific argument can be beneficially applied to the study of the relations between men in society. Thus at the outset Comte saw sociology as introducing to the study of social affairs the ‘positive’ methods which had made great advances in the natural sciences possible. In a different tradition Marx called his socialist analysis of nineteenth century society ‘scientific’ and contrasted it with other socialist theories which he called ‘utopian’. As Durkheim developed his polemic against Comte and Spencer, he argued that their approach was not scientific enough and urged that social facts like natural facts should be treated as ‘things’. Pareto argued for the application of ‘logicoexperimental’ methods in sociology. And even Max Weber, who was much the most sensitive of the masters of sociology to the case against a science of society, found it necessary to demand of the proponents of ‘understanding’ as the appropriate method for history and the social studies, that their arguments should measure up to the canons of scientific proof.

Nor has there been any substantial retreat from this position amongst recent sociologists. Indeed they have been reinforced by such doctrines as logical positivism with its emphasis upon the verification principle, and behaviourism, which remains as a substantial influence in the social sciences, even though some of its more extreme implications may have been abandoned. Thus for example we have the radical neo-positivism of G.A.Lundberg, who contrasts the ‘vast amount of trial-and-error blundering and emotional squirmings’ of our social adjustments with the relatively systematic adjustments which we make to the physical world, where ‘events are immediately referred to their proper place in the framework of science’. ¹So one could go on. A great deal of sociology is written in this crusading spirit, crusading for the application of scientific methods in academic environments, which are often hostile to a science of society for the worst of reasons. One certainly does not want to oppose the spirit which lies behind this crusade. It is true that what Lundberg calls the mental ‘hygiene’ of scientific method is greatly needed in our thinking, and, not least, in our academic thinking, about society and politics. What needs to be asserted, however, is that the crusade could be more effectively carried on if we devoted a little more time to thinking about what actually is meant by the scientific method. All too often our crusaders assume that there is a single and agreed set of principles which we have only to take over from the natural scientists and apply to society. This is a naïve view and it is time that sociologists became better acquainted with the present situation in the philosophy of science, and with the answers of philosophers of science to those who ask them for an account of the scientific method.

There was a time when the problem of the philosophy of of science appeared to be that of setting down the principles of inductive logic in such a way that they would bear comparison with the logic of deductive argument. But those days are surely past. What characterizes the philosophy of empirical science today is its increased humility. It is widely recognized that the arguments of empirical science can never have the same sort of certainty as the rational demonstrations of deductive logic do. Science is not thought of as the search for a set of final and absolute truths. Rather it is seen as an always relatively imperfect and incomplete attempt to explain and to predict the events which we experience. And, because this is recognized, the philosophers have abandoned any attempt to legislate to the scientist. Instead they turn to the man whose work has led to relatively successful prediction and adjustment, and ask him what his methods and arguments were, in order to discover which arguments and methods really work. And it is always to be expected that new sorts of argument and method will be found, which may have uses elsewhere.

Thus methodology or the philosophy of empirical science has ceased to be a mainly normative discipline. It has, itself, become empirically oriented. The philosophy of science merges inevitably with the history of empirical investigation. And the philosopher of science must, of necessity be very familiar with the theory and the research methods of the science he is studying.

The situation confronting the methodologist of the social sciences, however, is, in some respects, peculiar. He cannot simply content himself with asking what methods of proved success ‘come naturally’ to the sociologist, as a philosopher might ask, for example, of the chemist or the biologist. Partly this is because there are so few methods of proved success in sociology. Partly it is because the forms of argument which come most naturally in social discourse are the most misleading. But most of all it is because the great sociologists have all in one way or another disciplined their work by modelling their investigations on those of one or other of the branches of natural science. There is, of course, a great deal to be learned from an analysis of the arguments which have arisen among historians about the interpretation of history, and it is in this field that one can best see the arguments which ‘come naturally’ to students of society. But even here the most profitable task for the methodologist is the comparative one, that is of comparing the methods used by historians to arrive at their conclusions with those of the scientist. We shall do well to begin, therefore, by considering some of the models of scientific argument which sociologists have drawn from the natural sciences. Our first aim should be to set them out and this will be done in the present chapter. In later chapters we shall be concerned with the actual problems presented to the sociologist by the nature of his subject matter, and by the nature of the questions which he feels called upon to ask. That is to say we shall be concerned with problems of conceptual analysis and sociological theory.

The three major models of scientific investigation which sociologists have taken over from the natural sciences are, firstly, that which sees science as primarily classificatory, secondly, that which sees it as a search for laws, and finally that which sees it as concerned with the establishment of causal relations and sequences. We may now consider each of these in turn, and then go on to consider whether in fact there are not more fundamental aspects of scientific method which underly all of them.

SCIENCE AS CLASSIFICATION

The great tradition of thought in the philosophy of science in England since the Reformation has been an empiricist one. Its aims were clearly stated by Francis Bacon in the Novum Organum, when he contrasted the methods of empirical science with those of the scholastic thinkers as follows:—

There are only two ways of searching into and discovering truth. The one flies from senses and particulars to the most general axioms, and from these principles, the truth of which it takes for settled and immoveable, proceeds to judgment and the discovery of middle axioms. And this way is now in fashion. The other derives from the senses and particulars, rising by gradual and unbroken ascent, so that it arrives at the most general axioms last of all. This is the true way, but as yet untried. ¹

This new and ‘true’ way for empirical science had two possible applications. One was to search for laws by a process of induction by simple enumeration, i.e. by noting the characteristics or behaviour of as large a number of instances of a phenomenon as possible and proceeding to generalize about all instances of that phenomenon. To the question of the validity and the usefulness of this procedure we shall have to return when considering the nature of scientific laws. The second application concerns us more immediately. This was the attempt to describe and to classify the objects in the world which was made by many biologists and natural historians shortly after Bacon’s time. In its early stages, biology, with its meticulous observation and description, followed by classification in terms of observable characteristics, seemed to be the empirical science par excellence, the one sphere in which the ‘true way’ had come gloriously into its own.

Inevitably the drive towards an empiricist method for sociology has led to proposals for the application in sociology of the proved methods of the biological sciences, and noone has rested his account of sociological method more firmly on the experience of biology than Durkheim. It will be convenient therefore to follow Durkheim in order to understand more fully the implications of the adoption of such a model of scientific method by the sociologist.

In The Rules of Sociological Method, ¹ Durkheim outlines five stages of scientific investigation. These are: (1) Definition of the subject matter in terms of some observable characteristic; (2) Description of normal types after a study of many cases; (3) Classification into species, genera etc.; (4) Comparative and causal investigation of the reasons for variation; (5) The attempt to discover any general law that might emerge in the course of these various stages. Thus in a passage which is striking for its doctrinaire insistence on the one ‘true way’ for the scientist, Durkheim discusses the question of the economist’s study of ‘value’ as follows: ‘If value had been studied as any fact of reality ought to be studied, the economist would indicate first of all, by what characteristics one might recognize the thing so designated, then classify its varieties, investigate by methodical inductions what the causes of its variations are, and finally compare these general results in order to abstract a general formula’. ² Let us now consider what would be involved in the application of this method in sociology.

Turning to the first stage we find that, in point of fact, it involves two separate tasks. For the indication of the characteristics by which a thing is designated will involve, first of all, the definition of a total field of study, and secondly the definition of the sort of ‘thing’ which will be found in this field. In the case of biology the field is first of all defined as covering plant and animal life, and then the key characteristics of each species have to be indicated. Unfortunately Durkheim does not distinguish between the two problems involved here and his account of the problems of sociological method is consequently confused.

In any case it will be apparent that both tasks raise difficulties when they are undertaken with regard to social data. Clearly, it is not at all possible to indicate the characteristics by which the sphere of the social is demarcated, with anything like the ease with which the biologist is able to demarcate his field. Nor is it always practical to seek for ‘things’ in the social world to be classified into species in the way that plants and animals are. For these reasons it may well be impossible even to begin applying biological methods of investigation in all fields of sociology without serious distortion.

It is in fact very difficult to find an empirical characteristic which might serve to demarcate the social, and Durkheim appears to recognize this in practice, for his own definition of the social involves a complex theoretical concept. Social facts are said to be distinguished by the fact that they are external to the individual and exercise restraint over him. Clearly this definition does not tell us by what empirical characteristics social facts may be recognized, though by confusing the perspective of the scientist with that of the observed participant (who can distinguish the social from the non-social in this way) Durkheim gives us the impression that it does. But even he recognizes that it is unsatisfactory, for he goes on to give a second definition of a social fact as, ‘every way of acting, which is general throughout a given society, while at the same time existing independently of its individual manifestations’. ¹ But this definition is no more satisfactory than the first, for, though in its first part it seems to offer an empirical means of differentiating the social from the non-social, Durkheim would insist that this alone would be an insufficient characterization of the social. Everything therefore turns on the second part of the definition, but here we are faced with a metaphysical conception of little use in actual research.

It must surely be admitted that Durkheim’s attempt to provide a purely empirical criterion of the social is a failure. And this is not surprising. For the fact of the matter is that the actual data with which sociology is concerned, and which it seeks to explain, consists of human behaviour and the products of human behaviour, the same data with which psychologists, economists and historians have to deal. The difference lies not in the data, but in the different theoretical frame of reference, in terms of which the data is interpreted. It was Durkheim’s great merit as a sociological theorist that he saw and insisted upon the distinctiveness of sociological, as contrasted with psychological explanations. Unfortunately, however, his empiricist bias as a methodologist prevented him from clarifying the true nature of the difference.

There still remains to be considered the possibility of defining in terms of empirical characteristics the various sorts of ‘thing’ which are to be classified as social species. The difficulty here is that we do not readily experience the social world as being made up of a number of discrete things, each with its own definite spatial boundary, in the way that the biologist does in studying plant and animal life. It seems likely that the real reason why Durkheim insists upon the otherwise rather odd rule that we should consider social facts as things is that he expects sociology to follow a classificatory method, and this method presupposes that the world is experienced in terms of ‘things’. If it is not experienced in this way, however, and it proves difficult to conceptualize our experience in these terms, the classificatory method can only be used at the cost of a considerable distortion of experience.

Durkheim’s own discussion of the corollaries of saying that social facts are things does not clarify his meaning. He seems primarily concerned to oppose ideological analysis, or, as he says, the tendency ‘to focus our consciousness upon, to analyse and combine our ideas’ and to insist on the necessity of studying the social world outside our own heads. But this world consists of social relations (whose meaning might well be further analyzed in terms of the goals, aspirations, expectations, understandings and ideas of the related persons) and the real problem is whether a world of this kind can be thought of as consisting of discrete things

It would seem that it is sometimes possible to think of systems of social relations in this way, but that on other occasions it is more convenient and more illuminating to think of a continuous social process, which could be analysed in terms of its elements, but not in terms of kinds of ‘thing’. The ‘functionalist’ school in social anthropology, in criticizing the misuse of the comparative method, are sometimes insisting on this point (e.g. when they argue that it is not possible to make a comparative study of something like sacrificial customs, because any particular custom loses its meaning when it is torn out of context). The type of work, in which it would seem to be both permissible and useful to regard social facts as things, is in the comparative study of associations and institutions like trade unions, forms of local government, of political parties, or of social processes like political revolutions. This was the sort of work which the Webbs called sociological in Britain, and it is the same approach which characterizes work such as Duverger’s recent sociological study of political parties. ¹

It would seem, in fact, that in these fields one is dealing with sets of facts which common-sense is prepared to regard as things, quite independently of whether they belong to the realm of the social or not. For this reason the usefulness of the biological type of enquiry in sociology does not depend entirely on our ability to give an empirical criterion for demarcating the social. Common-sense recognizes the existence of certain classes of things and the sociologist can without distortion of the commonsense picture of the world, go on to describe and classify the species.

It should be noted that the ‘things’ which Durkheim has in mind are ‘societies’ ¹ and that he insists as against Comte that it is possible to study ‘societies’ rather than simply ‘humanity’. He specifically rejects the view of Pascal, of which Comte approves that, ‘the entire succession of men, through the whole course of the ages must be regarded as one man, always living and incessantly learning’. ² But even ‘societies’ are not the clearly demarcated ‘things’ which the comparative and classificatory method demands, and only a superficial analysis can make them appear so. To this question it will be necessary to return in later chapters.

For the moment, then, we may accept that it is possible within certain limited spheres to employ the specimen collecting and classificatory methods of the biologist. Granting that this is so, we may now go on to the next stage which Durkheim discusses, namely that of distinguishing the normal specimen from the pathological.

In principle one can imagine this problem as arising in any science, whatever its subject matter, because there is no subject in which instances of what we regard as the same phenomenon do not differ from one another in their detailed appearance; but the problem seems to loom larger in the biological and social sciences for two reasons. Firstly, there probably is more variability among biological and social specimens than is found amongst the phenomena dealt with by the physical sciences, and, secondly, the physicist is sometimes able to avoid the problem by formulating idealized laws, which explain, without claiming to describe, empirical events. The main problem here is now, in the absence of some pre-existing explanatory theory, to write in general terms about a particular class of things or objects. Of course, to some extent the very fact that we are able to name things is an indication that we have some general concept which we attach to the name, but science demands that this general concept should be made explicit and its exact implications defined. This is what the definition of the normal specimen is supposed to do. Durkheim’s contention is that the true scientific way of arriving at the concept of the normal...

Table of contents

• Cover Page
• Title Page
• Copyright Page
• PREFACE
• I: SOCIOLOGY AS A SCIENCE
• II: EMPIRICIST SOCIOLOGY
• III: THE SUBJECT MATTER OF SOCIOLOGY
• IV: THE PROBLEM OF FUNCTIONALISM
• V: THE ACTION FRAME OF REFERENCE
• VI: VALUES IN SOCIOLOGICAL THEORY
• VII: THE THEORY OF SOCIAL CONFLICT AND CHANGE
• VIII: CONFLICT AND THE ANALYSIS OF CLASS
• IX: OBJECTIVITY AND PROOF IN SOCIOLOGY
• X: CONCLUSIONS: THE SCOPE OF SOCIOLOGY