The Rules of Sociological Method by Emile Dürkheim Edited with an Introduction by Steven Lukes Translated by W. D. Halls Iffll THE FREE PRESS New York London Toronto Sydney The Rules of Sociological Method Chapter V Rules for the Explanation of Social Facts The constitution of species is above all a means of grouping the facts so as to facilitate their interpretation, but social morphology is only one step towards the truly explanatory part of the science. What is the method appropriate for explanation? utility. I Most sociologists believe they have accounted for phenomena once they have demonstrated the purpose they serve and the role they play. They reason as if phenomena existed solely for this role and had no determining cause save a clear or vague sense of the services they are called upon to render. This is why it is thought that all that is needful has been said to make them intelligible when it has been established that these services are real and that the social need they satisfy has been demonstrated. Thus Comte relates all the drive for progress of the human species to this basic tendency, 'which directly impels man continually to improve his condition in all respects',1 whereas Spencer relates it to the need for greater happiness. It is by virtue of this principle that Spencer explains the formation of society as a function of the advantages which flow from co-operation, the institution of government by the utility which springs from regulating military co-operation,2 and the transformations which the family has undergone from the need for a more perfect reconciliation of the interests of parents, children and society. But this method confuses two very different questions. To demonstrate the utility of a fact does not explain its origins, nor 119 -w- 120 The Rules of Sociological Method how it is what it is. The uses which it serves presume the specific properties characteristic of it, but do not create it. Our need for things cannot cause them to be of a particular nature; consequently, that need cannot produce them out of nothing, conferring in this way existence upon them. They spring from causes of another kind. The feeling we have regarding their utility can stimulate us to set these causes in motion and draw upon the effects they bring in their train, but it cannot conjure up these results out of nothing. This proposition is self-evident so long as only material or even psychological phenomena are being considered. It would also not be disputed in sociology if the social facts, because of their total lack of material substance, did not appear - wrongly, moreover -bereft of intrinsic reality. Since we view them as purely mental configurations, provided they are found to be useful, as soon as the idea of them occurs to us they seem to be self-generating. But since each fact is a force which prevails over the force of the individual and possesses its own nature, to bring a fact into existence it cannot suffice to have merely the desire or the will to engender it. Prior forces must exist, capable of producing this firmly established force, as well as natures capable of producing this special nature. Only under these conditions can facts be created. To revive the family spirit where it has grown weak, it is not enough for everybody to realise its advantages; we must set directly in operation those causes which alone can engender it. To endow a government with the authority it requires, it is not enough to feel the need for this. We must address ourselves to the sole sources from which all authority is derived: the establishment of traditions, a common spirit, etc. For this we must retrace our steps farther back along the chain of cause and effect until we find a point at which human action can effectively intervene. What clearly demonstrates the duality of these two avenues of research is that a fact can exist without serving any purpose, either because it has never been used to further any vital goal or because, having once been of use, it has lost all utility but continues to exist merely through force of custom. There are even more instances of such survivals in society than in the human organism. There are even cases where a practice or a social institution changes its functions without for this reason changing its nature. The rule of is pater est quern justae nuptiae declarant has remained substantially the same in our legal code as it was in ancient Roman law. But Rules for the Explanation of Social Facts 121 while its purpose was to safeguard the property rights of the father over children born of his legitimate wife, it is much more the rights of the children that it protects today. The swearing of an oath began by being a kind of judicial ordeal before it became simply a solemn and impressive form of attestation. The religious dogmas of Christianity have not changed for centuries, but the role they play in our modern societies is no longer the same as in the Middle Ages. Thus words serve to express new ideas without their contexture changing. Moreover, it is a proposition true in sociology as in biology, that the organ is independent of its function, i.e. while staying the same it can serve different ends. Thus the causes which give rise to its existence are independent of the ends it serves. Yet we do not mean that the tendencies, needs and desires of men never actively intervene in social evolution. On the contrary, it is certain that, according to the way they make an impact upon the conditions on which a fact depends, they can hasten or retard development. Yet, apart from the fact that they can never create something out of nothing, their intervention itself, regardless of its effects, can only occur by virtue of efficient causes. Indeed, a tendency cannot, even to this limited extent, contribute to the production of a new phenomenon unless it is itself new, whether constituted absolutely or arising from some transformation of a previous tendency. For unless we postulate a truly providential harmony established beforehand, we could not admit that from his origins man carried within him in potential all the tendencies whose opportuneness would be felt as evolution progressed, each one ready to be awakened when the circumstances called for it. Furthermore, a tendency is also a thing; thus it cannot arise or be modified for the sole reason that we deem it useful. It is a force possessing its own nature. For that nature to come into existence or be changed, it is not enough for us to find advantage in this occurring. To effect such changes causes must come into play which require them physically. For example, we have explained the constant development of the social division of labour by showing that it is necessary in order for man to sustain himself in the new conditions of existence in which he is placed as he advances in history. We have therefore attributed to the tendency which is somewhat improperly termed the instinct of self-preservation an important role in our explana- 122 The Rules of Sociological Method tion. But in the first place the tendency alone could not account for even the most rudimentary form of specialisation. It can accomplish nothing if the conditions on which this phenomenon depends are not already realised, that is, if individual differences have not sufficiently increased through the progressive state of indetermina-tion of the common consciousness and hereditary influences.3 The division of labour must even have begun already to occur for its utility to be perceived and its need to be felt. The mere development of individual differences, implying a greater diversity of tastes and abilities, had necessarily to bring about this first consequence. Moreover, the instinct of self-preservation did not come by itself and without cause to fertilise this first germ of specialisation. If it directed first itself and then us into this new path, it is because the course it followed and caused us to follow beforehand was as if blocked. This was because the greater intensity of the struggle for existence brought about by the greater concentration of societies rendered increasingly difficult the survival of those individuals who continued to devote themselves to more unspecialised tasks. Thus a change of direction was necessary. On the other hand if it turned itself, and for preference turned our activity, towards an ever increasing division of labour, it was also because it was the path of least resistance. The other possible solutions were emigration, suicide or crime. Now, on average, the ties that bind us to our country, to life and to feeling for our fellows are stronger and more resistant sentiments than the habits which can deter us from narrower specialisation. Thus these habits had inevitably to give ground as every advance occurred. Thus, since we are ready to allow for human needs in sociological explanations, we need not revert, even partially, to teleology. For these needs can have no influence over social evolution unless they themselves evolve, and the changes through which they pass can only be explained by causes which are in no way final. What is even more convincing that the foregoing argument is the study of how social facts work out in practice. Where teleology rules, there rules also a fair margin of contingency, for there are no ends - and even fewer means - which necessarily influence all men, even supposing they are placed in the same circumstances. Given the same environment, each individual, according to his temperament, adapts himself to it in the way he pleases and which he prefers to all others. The one will seek to change it so that it Rules for the Explanation of Social Facts 123 better suits his needs; the other will prefer to change himself and to moderate his desires. Thus to arrive at the same goal, many different routes can be, and in reality are, followed. If then it were true that historical development occurred because of ends felt either clearly or obscurely, social facts would have to present an infinite diversity and all comparison would almost be impossible. But the opposite is true. Undoubtedly external events, the links between which constitute the superficial part of social life, vary from one people to another. Yet in this way each individual has his own history, although the bases of physical and social organisation remain the same for all. If, in fact, one comes even a little into contact with social phenomena, one is on the contrary surprised at the outstanding regularity with which they recur in similar circumstances. Even the most trivial and apparently most puerile practices are repeated with the most astonishing uniformity. A marriage ceremony, seemingly purely symbolic, such as the abduction of the bride-to-be, is found to be identical everywhere that a certain type of family exists, which itself is lined to a whole political organisation. The most bizarre customs, such as the 'couvade', the levirate, exogamy, etc. are to be observed in the most diverse peoples and are symptomatic of a certain social state. The right to make a will appears at a specific phase of history and, according to the severity of the restrictions which limit it, we can tell at what stage of social evolution we have arrived. It would be easy to multiply such examples. But the widespread character of collective forms would be inexplicable if final causes held in sociology the preponderance attributed to them. Therefore when one undertakes to explain a social phenomenon the efficient cause which produces it and the function it fulfils must be investigated separately. We use the word 'function' in preference to 'end' or 'goal' precisely because social phenomena generally do not exist for the usefulness of the results they produce. We must determine whether there is a correspondence between the fact being considered and the general needs of the social organism, and in what this correspondence consists, without seeking to know whether it was intentional or not. All such questions of intention are, moreover, too subjective to be dealt with scientifically. Not only must these two kinds of problems be dissociated from each other, but it is generally appropriate to deal with the first kind before the second. This order of precedence corresponds to 124 The Rules of Sociological Method that of the facts. It is natural to seek the cause of a phenomenon before attempting to determine its effects. This method is all the more logical because the first question, once resolved, will often help to answer the second. Indeed, the solid link which joins cause to effect is of a reciprocal character which has not been sufficiently recognised. Undoubtedly the effect cannot exist without its cause, but the latter, in turn, requires its effect. It is from the cause that the effect derives its energy, but on occasion it also restores energy to the cause and consequently cannot disappear without the cause being affected.4 For example, the social reaction which constitutes punishment is due to the intensity of the collective sentiments that crime offends. On the other hand it serves the useful function of maintaining those sentiments at the same level of intensity, for they could not fail to weaken if the offences committed against them remained unpunished. 5 Likewise, as the social environment becomes more complex and unstable, traditions and accepted beliefs are shaken and take on a more indeterminate and flexible character, whilst faculties of reflection develop. These same faculties are indispensable for societies and individuals to adapt themselves to a more mobile and complex environment.6 As men are obliged to work more intensively, the products of their labour become more numerous and better in quality; but this increase in abundance and quality of the products is necessary to compensate for the effort that this more considerable labour entails.7 Thus, far from the cause of social phenomena consisting of a mental anticipation of the function they are called upon to fulfil, this function consists on the contrary, in a number of cases at least, in maintaining the pre-existent cause from which the phenomena derive. We will therefore discover more easily the function if the cause is already known. If we must proceed only at a second stage to the determination of the function, it is none the less necessary for the complete explanation of the phenomenon. Indeed, if the utility of a fact is not what causes its existence, it must generally be useful to continue to survive. If it lacks utility, that very reason suffices to make it harmful, since in that case it requires effort but brings in no return. Thus if the general run of social phenomena had this parasitic character, the economy of the organism would be in deficit, and social life would be impossible. Consequently, to provide a satisfactory explanation of social life we need to show Rules for the Explanation of Social Facts 125 how the phenomena which are its substance come together to place society in harmony with itself and with the outside world. Undoubtedly the present formula which defines life as a correspondence between the internal and the external environments is only approximate. Yet in general it remains true; thus to explain a fact which is vital, it is not enough to show the cause on which it depends. We must also - at least in most cases - discover the part that it plays in the establishment of that general harmony. II Having distinguished between these two questions, we must determine the method whereby they must be resolved. At the same time as being teleological, the method of explanation generally followed by sociologists is essentially psychological. The two tendencies are closely linked. Indeed, if society is only a system of means set up by men to achieve certain ends, these ends can only be individual, for before society existed there could only exist individuals. It is therefore from the individual that emanate the ideas and needs which have determined the formation of societies. If it is from him that everything comes, it is necessarily through him that everything must be explained. Moreover, in society there is nothing save individual consciousnesses, and it is consequently in these that is to be found the source of all social evolution. Thus sociological laws can only be a corollary of the more general laws of psychology. The ultimate explanation of collective life will consist in demonstrating how it derives from human nature in general, either by direct deduction from it without any preliminary observation, or by establishing links after having observed human nature. These expressions are almost word for word those used by Auguste Comte to characterise his method. 'Since the social phenomenon', he asserts, 'conceived of in its totality, is only basically a simple development of humanity without any creation of faculties at all, as I have established above, the whole framework of effects that sociological observation can successively uncover will therefore necessarily be found, at least in embryo, in that primordial type which biology has constructed beforehand for sociology'.8 This is because, in his view, the dominant fact of social 126 The Rules of Sociological Method life is progress, and because progress furthermore depends on a factor exclusively psychical in kind: the tendency that impels man to develop his nature more and more. Social facts may even derive so immediately from human nature that, during the initial stages of history, they could be directly deduced from it without having recourse to observation.9 It is true, as Comte concedes, that it is impossible to apply this deductive method to the more advanced phases of evolution. This impossibility is purely of a practical kind. It arises because the distance from the points of departure and arrival becomes too considerable for the human mind, which, if it undertook to traverse it without a guide, would run the risk of going astray.10 But the relationship between the basic laws of human nature and the ultimate results of progress is none the less capable of analysis. The most complex forms of civilisation are only a developed kind of psychical life. Thus, even if psychological theories cannot suffice as premises for sociological reasoning, they are the touchstone which alone permits us to test the validity of propositions inductively established. 'No law of social succession', declares Comte, 'which has been elaborated with all the authority possible by means of the historical method, should be finally accepted before it has been rationally linked, directly or indirectly, but always irrefutably, to the positivist theory of human nature'. 11 Psychology will therefore always have the last word. This is likewise the method followed by Spencer. In fact, according to him, the two primary factors of social phenomena are the external environment and the physical and moral constitution of the individual.12 Now the first factor can only influence society through the second one, which is thus the essential motivating power for social evolution. Society arises to allow the individual to realise his own nature, and all the transformations through which it has passed have no other purpose than to make this act of self-realisation easier and more complete. It is by virtue of this principle that, before proceeding to any research into social organisation, Spencer thought it necessary to devote almost all the first volume of his Principles of Sociology to the study of primitive man from the physical, emotional and intellectual viewpoint. 'The science of sociology', he states, 'sets out with social units, conditioned as we have seen, constituted physically, emotionally and intellectually and possessed of certain early acquired notions and correlative feelings'.13 And it is in two of these sentiments, fear of Rules for the Explanation of Social Facts 127 the living and fear of the dead, that he finds the origin of political and religious government.14 It is true that he admits that once it has been constituted, society reacts upon individuals.15 But it does not follow that society has the power to engender directly the smallest social fact; from this viewpoint it has causal effectiveness only through the mediation of the changes that it brings about in the individual. Thus it is always from human nature, whether primitive or derived, that everything arises. Moreover, the influence which the body social exerts upon its members can have nothing specific about it, since political ends are nothing in themselves, but merely the summary expression of individual ends.16 Social influence can therefore only be a kind of consequent effect of private activity upon itself. Above all, it is not possible to see of what it may consist in industrial societies whose purpose is precisely to deliver the individual over to his natural impulses by ridding him of all social constraint. This principle is not only at the basis of these great doctrines of general sociology, but also inspires a very great number of particular theories. Thus domestic organisation is commonly explained by the feelings that parents have for their children and vice versa; the institution of marriage by the advantages that it offers husband and wife and their descendants; punishment by the anger engendered in the individual through any serious encroachment upon his interests. The whole of economic life, as conceived of and explained by economists, particularly those of the orthodox school, hangs in the end upon a purely individual factor, the desire for wealth. If we take morality, the basis of ethics is the duties of the individual towards himself. And in religion one can see a product of the impressions that the great forces of nature or certain outstanding personalities awaken in man, etc., etc. But such a method is not applicable to sociological phenomena unless one distorts their nature. For proof of this we need only refer to the definition we have given. Since their essential characteristic is the power they possess to exert outside pressure on individual consciousnesses, this shows that they do not derive from these consciousnesses and that consequently sociology is not a corollary of psychology. This constraining power attests to the fact that they express a nature different from our own, since they only penetrate into us by force or at the very least by bearing down more or less heavily upon us. If social life were no more than an 128 The Rules of Sociological Method Rules for the Explanation of Social Facts 129 extension of the individual, we would not see it return to its origin and invade the individual consciousness so precipitately. The authority to which the individual bows when he acts, thinks or feels socially dominates him to such a degree because it is a product of forces which transcend him and for which he consequently cannot account. It is not from within himself that can come the external pressure which he undergoes; it is therefore not what is happening within himself which can explain it. It is true that we are not incapable of placing constraints upon ourselves; we can restrain our tendencies, our habits, even our instincts, and halt their development by an act of inhibition. But inhibitive movements must not be confused with those which make up social constraint. The process of inhibitive movements is centrifugal; but the latter are centripetal. The former are worked out in the individual consciousness and then tend to manifest themselves externally; the latter are at first external to the individual, whom they tend afterwards to shape from the outside in their own image. Inhibition is, if one likes, the means by which social constraint produces its psychical effects, but is not itself that constraint. Now, once the individual is ruled out. only society remains. It is therefore in the nature of society itself that we must seek the explanation of social life. We can conceive that, since it transcends infinitely the individual both in time and space, it is capable of imposing upon him the ways of acting and thinking that it has consecrated by its authority. This pressure, which is the distinctive sign of social facts, is that which all exert upon each individual. But it will be argued that since the sole elements of which society is composed are individuals, the primary origin of sociological phenomena cannot be other than psychological. Reasoning in this way, we can just as easily establish that biological phenomena are explained analytically by inorganic phenomena. It is indeed certain that in the living cell there are only molecules of crude matter. But they are in association, and it is this association which is the cause of the new phenomena which characterise life, even the germ of which it is impossible to find in a single one of these associated elements. This is because the whole does not equal the sum of its parts; it is something different, whose properties differ from those displayed by the parts from which it is formed. Association is not, as has sometimes been believed, a phenomenon infertile in itself, which consists merely in juxtaposing externally facts already given and properties already constituted. On the contrary, is it not the source of all the successive innovations that have occurred in the course of the general evolution of things? What differences exist between the lower organisms and others, between the organised living creature and the mere protoplasm, between the latter and the inorganic molecules of which it is composed, if it is not differences in association? All these beings, in the last analysis, split up into elements of the same nature; but these elements are in one place juxtaposed, in another associated. Here they are associated in one way, there in another. We are even justified in wondering whether this law does not even extend to the mineral world, and whether the differences which separate inorganic bodies do not have the same origin. By virtue of this principle, society is not the mere sum of individuals, but the system formed by their association represents a specific reality which has its own characteristics. Undoubtedly no collective entity can be produced if there are no individual consciousnesses: this is a necessary but not a sufficient condition. In addition, these consciousnesses must be associated and combined, but combined in a certain way. It is from this combination that social life arises and consequently it is this combination which explains it. By aggregating together, by interpenetrating, by fusing together, individuals give birth to a being, psychical if you will, but one which constitutes a psychical individuality of a new kind.17 Thus it is in the nature of that individuality and not in that of its component elements that we must search for the proximate and determining causes of the facts produced in it. The group thinks, feels and acts entirely differently from the way its members would if they were isolated. If therefore we begin by studying these members separately, we will understand nothing about what is taking place in the group. In a word, there is between psychology and sociology the same break in continuity as there is between biology and the physical and chemical sciences. Consequently every time a social phenomenon is directly explained by a psychological phenomenon, we may rest assured that the explanation is false. Some will perhaps argue that, although society, once formed, is the proximate cause of social phenomena, the causes which have determined its formation are of a psychological nature. They may concede that, when individuals are associated together, their 130 The Rules of Sociological Method Rules for the Explanation of Social Facts 131 association may give rise to a new life, but claim that this can only take place for individual reasons. But in reality, as far as one can go back in history, the fact of association is the most obligatory of all, because it is the origin of all other obligations. By reason of my birth, I am obligatorily attached to a given people. It may be said that later, once I am an adult, I acquiesce in this obligation by the mere fact that I continue to live in my own country. But what does that matter? Such acquiescence does not remove its imperative character. Pressure accepted and undergone with good grace does not cease to be pressure. Moreover, how far does such acceptance go? Firstly, it is forced, for in the immense majority of cases it is materially and morally impossible for us to shed our nationality; such a rejection is even generally declared to be apostasy. Next, the acceptance cannot relate to the past, when I was in no position to accept, but which nevertheless determines the present. I did not seek the education I received; yet this above all else roots me to my native soil. Lastly, the acceptance can have no moral value for the future, in so far as this is unknown. I do not even know all the duties which one day may be incumbent upon me in my capacity as a citizen. How then could I acquiesce in them in advance? Now, as we have shown, all that is obligatory has its origins outside the individual. Thus, provided one does not place oneself outside history, the fact of association is of the same character as the others and is consequently explicable in the same way. Furthermore, as all societies are born of other societies, with no break in continuity, we may be assured that in the whole course of social evolution there has not been a single time when individuals have really had to consult together to decide whether they would enter into collective life together, and into one sort of collective life rather than another. Such a question is only possible when we go back to the first origins of any society. But the solutions, always dubious, which can be brought to such problems could not in any case affect the method whereby the facts given in history must be treated. We have therefore no need to discuss them. Yet our thought would be singularly misinterpreted if the conclusion was drawn from the previous remarks that sociology, in our view, should not even take into account man and his faculties. On the contrary, it is clear that the general characteristics of human nature play their part in the work of elaboration from which social life results. But it is not these which produce it or give it its special form: they only make it possible. Collective representations, emotions and tendencies have not as their causes certain states of consciousness in individuals, but the conditions under which the body social as a whole exists. Doubtless these can be realised only if individual natures are not opposed to them. But these are simply the indeterminate matter which the social factor fashions and transforms. Their contribution is made up exclusively of very general states, vague and thus malleable predispositions which of themselves could not assume the definite and complex forms which characterise social phenomena, if other agents did not intervene. What a gulf, for example, between the feelings that man experiences when confronted with forces superior to his own and the institution of religion with its beliefs and practices, so multifarious and complicated, and its material and moral organisation! What an abyss between the psychical conditions of sympathy which two people of the same blood feel for each other,18 and that hotchpotch of legal and moral rules which determine the structure of the family, personal relationships, and the relationship of things to persons, etc.! We have seen that even when society is reduced to an unorganised crowd, the collective sentiments which arise within it can not only be totally unlike, but even opposed to, the average sentiments of the individuals in it. How much greater still must be the gap when the pressure exerted upon the individual comes from a normal society, where, to the influence exerted by his contemporaries, is added that of previous generations and of tradition! A purely psychological explanation of social facts cannot therefore fail to miss completely all that is specific, i.e. social, about them. What has blinkered the vision of many sociologists to the insufficiency of this method is the fact that, taking the effect for the cause, they have very often highlighted as causal conditions for social phenomena certain psychical states, relatively well defined and specific, but which in reality are the consequence of the phenomena. Thus it has been held that a certain religiosity is innate in man, as is a certain minimum of sexual jealousy, filial piety or fatherly affection, etc.. and it is in these that explanations have been sought for religion, marriage and the family. But history shows that these inclinations, far from being inherent in human nature, are either completely absent under certain social condi- 132 The Rules of Sociological Method tions or vary so much from one society to another that the residue left after eliminating all these differences, and which alone can be considered of psychological origin, is reduced to something vague and schematic, infinitely removed from the facts which have to be explained. Thus these sentiments result from the collective organisation and are far from being at the basis of it. It has not even been proved at all that the tendency to sociability was originally a congenital instinct of the human race. It is much more natural to see in it a product of social life which has slowly become organised in us, because it is an observable fact that animals are sociable or otherwise, depending on whether their environmental conditions force them to live in common or cause them to shun such a life. And even then we must add that a considerable gap remains between these well determined tendencies and social reality. Furthermore, there is a means of isolating almost entirely the psychological factor, so as to be able to measure precisely the scope of its influence: this is by seeking to determine how race affects social evolution. Ethnic characteristics are of an organic and psychical order. Social life must therefore vary as they vary, if psychological phenomena have on society the causal effectiveness attributed to them. Now we know of no social phenomenon which is unquestionably dependent on race, although we certainly cannot ascribe to this proposition the value of a law. But we can at least assert that it is a constant fact in our practical experience. Yet the most diverse forms of organisation are to be found in societies of the same race, while striking similarities are to be observed among societies of different races. The city state existed among the Phoenicians, as it did among the Romans and the Greeks; we also find it emerging among the Kabyles. The patriarchal family was almost as strongly developed among the Jews as among the Hindus, but it is not to be found among the Slavs, who are nevertheless of Aryan race. By contrast, the family type to be found among the Slavs exists also among the Arabs. The maternal family and the clan are observed everywhere. The precise nature of judicial proofs and nuptial ceremonies is no different among peoples most unlike from the ethnic viewpoint. If this is so, it is because the psychical element is too general to predetermine the course of social phenomena. Since it does not imply one social form rather than another, it cannot explain any such forms. It is true that there are a certain number of facts which it is customary Rules for the Explanation of Social Facts 133 to ascribe to the influence of race. Thus this, in particular, is how we explain why the development of literature and the arts was so rapid and intense in Athens, so slow and mediocre in Rome. But this interpretation of the facts, despite being the classic one, has never been systematically demonstrated. It seems to draw almost all its authority from tradition alone. We have not even reflected upon whether a sociological explanation of the same phenomena was not possible, yet we are convinced that this might be successfully attempted. In short, when we hastily attribute to aesthetic and inherited faculties the artistic nature of Athenian civilisation, we are almost proceeding as did men in the Middle Ages, when fire was explained by phlogiston and the effects of opium by its soporific powers. Finally, if social evolution really had its origin in the psychological make-up of man, one fails to see how this could have come about. For then we would have to admit that its driving force is some internal motivation within human nature. But what might such a motivation be? Would it be that kind of instinct of which Comte speaks, which impels man to realise increasingly his own nature? But this is to reply to one question by another, explaining progress by an innate tendency to progress, a truly metaphysical entity whose existence, moreover, has in no way been demonstrated. For the animal species, even those of the highest order, are not moved in any way by a need to progress, and even among human societies there are many which are content to remain stationary indefinitely. Might it be, as Spencer seems to believe, that there is a need for greater happiness, which forms of civilisation of every increasing complexity might be destined to realise more and more completely? It would then be necessary to establish that happiness grows with civilisation, and we have explained elsewhere all the difficulties to which such a hypothesis gives rise. 19 Moreover, there is something else: even if one or other of these postulates were conceded, historical development would not thereby become more intelligible; for the explanation which might emerge from it would be purely teleological. We have shown earlier that social facts, like all natural phenomena, are not explained when we have demonstrated that they serve a purpose. After proving conclusively that a succession of social organisations in history which have become increasingly more knowledgeable have resulted in the greater satisfaction of one or other of our fun- 134 The Rules of Sociological Method damental desires, we would not thereby have made the source of these organisations more comprehensible. The fact that they were useful does not reveal to us what brought them into existence. We might even explain how we came to conceive them, by drawing up a blueprint of them beforehand, so as to envisage the services we might expect them to render - and this is already a difficult problem. But our aspirations, which would thereby become the purpose of such organisations, would have no power to conjure them up out of nothing. In short, if we admit that they are the necessary means to attain the object we have in mind, the question remains in its entirety: How, that is to say, from what, and in what manner, have these means been constituted? Hence we arrive at the following rule: The determining cause of a social fact must be sought among antecedent social facts and not among the states of the individual consciousness. Moreover, we can easily conceive that all that has been stated above applies to the determination of the function as well as the cause of a social fact. Its function can only be social, which means that it consists in the production of socially useful effects. Undoubtedly it can and indeed does happen that it has repercussions which also serve the individual. But this happy result is not the immediate rationale for its existence. Thus we can complement the preceding proposition by stating: The function of a social fact must always be sought in the relationship that it bears to some social end. It is because sociologists have often failed to acknowledge this rule and have considered sociological phenomena from too psychological a viewpoint that their theories appear to many minds too vague, too ethereal and too remote from the distinctive nature of the things which sociologists believe they are explaining. The historian, in particular, who has a close contact with social reality cannot fail to feel strongly how these too general interpretations are incapable of being linked to the facts. In part, this has undoubtedly produced the mistrust that history has often manifested towards sociology. Assuredly this does not mean that the study of psychological facts is not indispensable to the sociologist. If collective life does not derive from individual life, the two are none the less closely related. If the latter cannot explain the former, it can at least render its explanation easier. Firstly, as we have shown, it is undeniably true that social facts are produced by an elaboration sui generis of psychological facts. But in addition Rules for the Explanation of Social Facts 135 I i ; | this action is itself not dissimilar to that which occurs in each individual consciousness and which progressively transforms the primary elements (sensations, reflexes, instincts) of which the consciousness was originally made up. Not unreasonably has the claim been made that the ego is itself a society, just as is the organism, although in a different way. For a long time psychologists have demonstrated the absolute importance of the factor of association in the explanation of mental activity. Thus a psychological education, even more than a biological one, constitutes a necessary preparation for the sociologist. But it can only be of service to him if, once he has acquired it, he frees himself from it, going beyond it by adding a specifically sociological education. He must give up making psychology in some way the focal point of his operations, the point of departure to which he must always return after his adventurous incursions into the social world. He must establish himself at the very heart of social facts in order to observe and confront them totally, without any mediating factor, while calling upon the science of the individual only for a general preparation and, if needs be, for useful suggestions.20 Ill Since the facts of social morphology are of the same nature as physiological phenomena, they must be explained according to the rule we have just enunciated. However, the whole of the preceding discussion shows that in collective life and, consequently, in sociological explanations, they play a preponderant role. If the determining condition for social phenomena consists, as we have demonstrated, in the very fact of association, the phenomena must vary with the forms of that association, i.e. according to how the constituent elements in a society are grouped. Furthermore, since the distinct entity formed by the union of elements of all kinds which enter into the composition of a society constitutes its inner environment, in the same way as the totality of anatomical elements, together with the manner in which they are arranged in space, constitutes the inner environment of organisms, we may state: The primary origin of social processes of any importance must be sought in the constitution of the inner social environment. i 136 The Rules of Sociological Method We may be even more precise. In fact, the elements which make up this environment are of two kinds: things and persons. Apart from the material objects incorporated in the society, among things must be included the products of previous social activity -the law and the customs that have been established, and literary and artistic monuments, etc. But it is plain that neither material nor non-material objects produce the impulsion that determines social transformations, because they both lack motivating power. Undoubtedly there is need to take them into account in the explanations which we attempt. To some extent they exert an influence upon social evolution whose rapidity and direction vary according to their nature. But they possess no elements essential to set that evolution in motion. They are the matter to which the vital forces of society are applied, but they do not themselves release any vital forces. Thus the specifically human environment remains as the active factor. The principal effort of the sociologist must therefore be directed towards discovering the different properties of that environment capable of exerting some influence upon the course of social phenomena. Up to now we have found two sets of characteristics which satisfy that condition admirably. These are: firstly, the number of social units or, as we have also termed it, the 'volume' of the society; and secondly, the degree of concentration of the mass of people, or what we have called the 'dynamic density'. The latter must be understood not only as the purely physical concentration of the aggregate population, which can have no effect if individuals - or rather groups of individuals - remain isolated by moral gaps, but the moral concentration of which physical concentration is only the auxiliary element, and almost invariably the consequence. Dynamic density can be defined, if the volume remains constant, as a function of the number of individuals who are effectively engaged not only in commercial but also moral relationships with each other, i.e. who not only exchange services or compete with one another, but live their life together in common. For, since purely economic relationships leave men separated from each other, these relationships can be very active without people necessarily participating in the same collective existence. Business ties which span the boundaries which separate peoples do not make those boundaries non-existent. The common life can be affected only by the number of people who effectively Rules for the Explanation of Social Facts 137 co-operate in it. This is why what best expresses the dynamic density of a people is the degree to which the social segments coalesce. For if each partial aggregate forms an entity, a distinct individuality separated from the others by a barrier, it is because in general the activity of its members remains localised within it. If, on the other hand, these partial entities are entirely fused together, or tend to do so, within the total society, it is because the ambit of social life to this extent has been enlarged. As for the physical density - if this is understood as not only the number of inhabitants per unit of area, but also the development of the means of communication and transmission - this is normally in proportion to the dynamic density and, in general, can serve to measure it. For if the different elements in the population tend to draw more closely together, it is inevitable that they will establish channels to allow this to occur. Furthermore, relationships can be set up between remote points of the social mass only if distance does not represent an obstacle, which means, in fact, that it must be eliminated. However, there are exceptions,21 and one would expose oneself to serious error if the moral concentration of a community were always judged according to the degree of physical concentration that it represented. Road, railways, etc. can serve commercial exchanges better than they can serve the fusion of populations, of which they can give only a very imperfect indication. This is the case in England, where the physical density is greater than in France but where the coalescence of social segments is much less advanced, as is shown by the persistence of parochialism and regional life. We have shown elsewhere how every increase in the volume and dynamic density of societies, by making social life more intense and widening the horizons of thought and action of each individual, profoundly modifies the basic conditions of collective life. Thus we need not refer again to the application we have already made of this principle. It suffices to add that the principle was useful to us in dealing not only with the still very general question which was the object of that study, but many other more specialised problems, and that we have therefore been able to verify its accuracy already by a fair number of experiments. However, we are far from believing that we have uncovered all the special features of the social environment which can play some part in the explanation of social facts. All we can say is that these are the sole 138 The Rules of Sociological Method features we have identified and that we have not been led to seek out others. But the kind of preponderance we ascribe to the social environment, and more especially to the human environment, does not imply that this should be seen as a kind of ultimate, absolute fact beyond which there is no need to explore further. On the contrary, it is plain that its state at any moment in history itself depends on social causes, some of which are inherent in society itself, while others depend on the interaction occurring between that society and its neighbours. Moreover, science knows no first causes, in the absolute sense of the term. For science a fact is primary simply when it is general enough to explain a great number of other facts. Now the social environment is certainly a factor of this kind, for the changes which arise within it, whatever the causes, have repercussions on every part of the social organism and cannot fail to affect all its functions to some degree. What has just been said about the general social environment can be repeated for the particular environments of the special groups which society includes. For example, depending on whether the family is large or small, or more or less turned in upon itself, domestic life will differ considerably. Likewise, if professional corporations reconstitute themselves so as to spread over a whole area, instead of remaining enclosed within the confines of a city, as they formerly were, their effect will be very different from what it was previously. More generally, professional life will differ widely according to whether the environment peculiar to each occupation is strongly developed or whether its bonds are loose, as is the case today. However, the effect of these special environments cannot have the same importance as the general environment, for they are subject to the latter's influence. Thus we must always return to the general environment. It is the pressure that it exerts upon these partial groups which causes their constitution to vary. This conception of the social environment as the determining factor in collective evolution is of the greatest importance. For if it is discarded, sociology is powerless to establish any causal relationship. Indeed, if this order of causes is set aside, there are no concomitant conditions on which social phenomena can depend. For if the external social environment - that which is formed by Rules for the Explanation of Social Facts 139 neighbouring societies - is capable of exercising some influence, it is only upon the functions of attack and defence; moreover, it can only make its influence felt through the mediation of the internal social environment. The principal causes of historical development would not therefore be found among the circumfusa (external influences). They would all be found in the past. They would themselves form part of that development, constituting simply more remote phases of it. The contemporary events of social life would not derive from the present state of society, but from prior events and historical precedents, and sociological explanations would consist exclusively in linking the present to the past. It is true that this may seem sufficient. Is it not commonly said that the purpose of history is precisely to link up events in their sequence? But it is impossible to conceive how the state which civilisation has attained at any given time could be the determining cause of the state which follows. The stages through which humanity successively passes do not engender each other. We can well understand how the progress realised in a given era in the fields of law, economics and politics, etc., makes fresh progress possible, but how does the one predetermine the other? The progress realised is a point of departure which allows us to proceed further, but what stimulates us to further progress? We would have to concede that there was a certain inner tendency which impels humanity constantly to go beyond the results already achieved, either to realise itself more fully or to increase its happiness, and the purpose of sociology would be to rediscover the order in which this tendency has developed. But without alluding afresh to the difficulties which such a hypothesis implies, in any case a law to express this development could not be in any sense causal. A relationship of causality can in fact only be established between two given facts. But this tendency, presumed to be the cause of development, is not something that is given. It is only postulated as a mental construct according to the effects attributed to it. It is a kind of motivating faculty which we imagine as underlying the movement which occurs, in order to account for it. But the efficient cause of a movement can only be another movement, not a potentiality of this kind. Thus all that we can arrive at experimentally is in point of fact a series of changes between which there exists no causal link. The antecedent state does not produce the subsequent one, but the relationship be- 140 The Rules of Sociological Method tween them is exclusively chronological. In these conditions any scientific prediction is thus impossible. We can certainly say how things have succeeded each other up the present, but not in what order they will follow subsequently, because the cause on which they supposedly depend is not scientifically determined, nor can it be so determined. It is true that normally it is accepted that evolution will proceed in the same direction as in the past, but this is a mere supposition. We have no assurance that the facts as they have hitherto manifested themselves are a sufficiently complete expression of this tendency. Thus we are unable to forecast the goal towards which they are moving in the light of the stages through which they have already successively passed. There is no reason to suppose that the direction this tendency follows even traces out a straight line. This is why the number of causal relationships established by sociologists is so limited. Apart from a few exceptions, among whom Montesquieu is the most illustrious example, the former philosophy of history concentrated solely on discovering the general direction in which humanity was proceeding, without seeking to link the phases of that evolution to any concomitant condition. Despite the great services Comte has rendered to social philosophy, the terms in which he poses the sociological problem do not differ from those of his predecessors. Thus his celebrated law of the three stages has not the slightest causal relationship about it. Even if it were true, it is, and can only be, empirical. It is a summary review of the past history of the human race. It is purely arbitrary for Comtc to consider the third stage to be the definitive stage of humanity. Who can say whether another will not arise in the future? Similarly, the law which dominates the sociology of Spencer appears to be no different in nature. Even if it were true that we at present seek our happiness in an industrial civilisation, there is no assurance that, at a later era, we shall not seek it elsewhere. The generality and persistence of this method is due to the fact that very often the social environment has been perceived as a means whereby progress has been realised, and not the cause which determines it. Furthermore, it is also in relationship to this same environment that must be measured the utilitarian value, or as we have stated it, the function of social phenomena. Among the changes caused by the environment, those are useful which are in harmony with the Rules for the Explanation of Social Facts 141 existing state of society, since the environment is the essential condition for collective existence. Again, from this viewpoint the conception we have just expounded is, we believe, fundamental, for it alone allows an explanation of how the useful character of social phenomena can vary without depending on arbitrary factors. If historical evolution is envisaged as being moved by a kind of vis a tergo (vital urge) which impels men forward, since a dynamic tendency can have only a single goal, there can exist only one reference point from which to calculate the utility or harmful-ness of social phenomena. It follows that there exists, and can only exist, a single type of social organisation which fits humanity perfectly, and the different societies of history are only successive approximations to that single model. It is unnecessary to show how such a simplistic view is today irreconcilable with the acknowledged variety and complexity of social forms. If on the other hand the suitability or unsuitability of institutions can only be established in relation to a given environment, since these environments are diverse, a diversity of reference points thus exists, and consequently a diversity of types which, whilst each being qualitatively distinct, are all equally grounded in the nature of the social environment. The question just dealt with is therefore closely connected to the constitution of social types. If there are social species, it is because collective life depends above all on concomitant conditions which present a certain diversity. If, on the contrary, the main causes of social events were all in the past, every people would be no more than the extension of the one preceding it, and different societies would lose their individuality, becoming no more than various moments in time of one and the same development. On the other hand, since the constitution of the social environment results from the mode in which the social aggregates come together - and the two phrases are in the end synonymous - we have now the proof that there are no characteristics more essential than those we have assigned as the basis for sociological classification. Finally, we should now realise better than before how unjust it would be to rely on the terms 'external conditions' and 'environment' to serve as an indictment of our method, and seek the sources of life outside what is already alive. On the contrary, the considerations just mentioned lead us back to the idea that the causes of social phenomena are internal to the society. It is much 142 The Rules of Sociological Method rather the theory which seeks to derive society from the individual that could be justly reproached with seeking to deduce the internal from the external (since it explains the social being by something other than itself) and the greater from the lesser (since it undertakes to deduce the whole from the part). Our own preceding principles in no way fail to acknowledge the spontaneous character of every living creature: thus, if they are applied to biology and psychology, it will have to be admitted that individual life as well develops wholly within the individual. IV From the set of rules which has just been established, there arises a certain conception of society and collective life. Two opposing theories divide men on this question. For some, such as Hobbes and Rousseau, there is a break in continuity between the individual and society. Man is therefore obdurate to the common life and can only resign himself to it if forced to do so. Social ends are not simply the meeting point for individual ends; they are more likely to run counter to then. Thus, to induce the individual to pursue social ends, constraint must be exercised upon him, and it in the institution and organisation of this constraint that lies the supreme task of society. Yet because the individual is regarded as the sole and unique reality of the human kingdom, this organisation, which is designed to constrain and contain him, can only be conceived of as artificial. The organisation is not grounded in nature, since it is intended to inflict violence upon him by preventing him from producing anti-social consequences. It is an artifact, a machine wholly constructed by the hands of men and which, like all products of this kind, is only what it is because men have willed it so; an act of volition created it, another one can transform it. Neither Hobbes nor Rousseau appear to have noticed the complete contradiction that exists in admitting that the individual is himself the creator of a machine whose essential role is to exercise domination and constraint over him. Alternatively, it may have seemed to them that, in order to get rid of this contradiction, it was sufficient to conceal it from the eyes of its victims by the skilful device of the social contract. It is from the opposing idea that the theoreticians of natural law Rules for the Explanation of Social Facts 143 and the economists, and more recently Spencer, 22 have drawn their inspiration. For them social life is essentially spontaneous and society is a natural thing. But, if they bestow this characteristic upon it, it is not because they acknowledge it has any specific nature, but because they find a basis for it in the nature of the individual. No more than the two thinkers already mentioned do they see in it a system of things which exists in itself, by virtue of causes peculiar to itself. But while Hobbes and Rousseau only conceived it as a conventional arrangement, with no link at all in reality, which, so to speak, is suspended in air, they in turn state its foundations to be the fundamental instincts of the human heart. Man is naturally inclined to political, domestic and religious life, and to commercial exchanges, etc., and it is from these natural inclinations that social organisation is derived. Consequently, wherever it is normal, there is no need to impose it by force. Whenever it resorts to constraint it is because it is not what it ought to be, or because the circumstances are abnormal. In principle, if individual forces are left to develop untrammelled they will organise themselves socially. Neither of these doctrines is one we share. Doubtless we make constraint the characteristic trait of every social fact. Yet this constraint does not arise from some sort of artful machination destined to conceal from men the snares into which they have stumbled. It is simply due to the fact that the individual finds himself in the presence of a force which dominates him and to which he must bow. But this force is a natural one. It is not derived from some conventional arrangement which the human will has contrived, adding it on to what is real; it springs from the heart of reality itself; it is the necessary product of given causes. Thus to induce the individual to submit to it absolutely of his own free will, there is no need to resort to deception. It is sufficient to make him aware of his natural state of dependence and inferiority. Through religion he represents this state to himself by the senses or symbolically; through science he arrives at an adequate and precise notion of it. Because the superiority that society has over him is not merely physical, but intellectual and moral, it need fear no critical examination, provided this is fairly undertaken. Reflection which causes man to understand how much richer or more complex and permanent the social being is than the individual being, can only reveal to him reasons to make 144 The Rules of Sociological Method Rules for the Explanation of Social Facts 145 comprehensible the subordination which is required of him and for the feelings of attachment and respect which habit has implanted within him.23 Thus only singularly superficial criticism could lay us open to the reproach that our conception of social constraint propagates anew the theories of Hobbes and Machiavelli. But if, contrary to these philosophers, we say that social life is natural, it is not because we find its origin in the nature of the individual; it is because it derives directly from the collective being which is, of itself, a nature sui generis; it is because it arises from that special process of elaboration which individual consciousnesses undergo through their association with each other and whence evolves a new form of existence.24 If therefore we recognise with some authorities that social life presents itself to the individual under the form of constraint, we admit with others that it is a spontaneous product of reality. What logically joins these two elements, in appearance contradictory, is that the reality from which social life emanates goes beyond the individual. Thus these words, 'constraint' and 'spontaneity', have not in our terminology the respective meanings that Hobbes gives to the former and Spencer to the second. To summarise: to most of the attempts that have been made to explain social facts rationally, the possible objection was either that they did away with any idea of social discipline, or that they only succeeded in maintaining it with the assistance of deceptive subterfuges. The rules we have set out would, on the other hand, allow a sociology to be constructed which would see in the spirit of discipline the essential condition for all common life, while at the same time founding it on reason and truth. Notes 1. Comte, Cours dephilosophiepositive, IV, p. 262. 2. Spencer, Principles of Sociology, vol. II, part V, ch. II, p. 247. 3. Division du travail social, II, chs 3 and 4. 4. We would not wish to raise questions of general philosophy which would be inappropriate here. However, we note that, if more closely studied, this reciprocity of cause and effect could provide a means of reconciling scientific mechanism with the teleology implied by the existence and, above all, the persistence of life. 5. Division du travail social, II, ch. 2, and especially pp.l05ff. 6. Ibid., pp. 52-3. 7. Ibid., p. 301ff. 8. Comte, Cours de philosophie positive, IV, pp. 333-4. 9. Ibid., IV, p. 345. 10. Ibid., IV, p. 346. 11. Ibid., IV, p. 334. 12. Spencer, Principles of Sociology,, vol. I, part I, ch. 2. 13. Ibid., vol. I, part I, ch. XXVII, p. 456. [Durkheim paraphrases. The exact quotation reads: 'Setting out with social units as thus conditioned physically, emotionally and intellectually, and as thus possessed of certain early-acquired ideas and correlative feelings, the science of sociology has to give an account of all the phenomena that result from their combined actions.'] 14. Ibid., p. 456. 15. Ibid., p. 15. 16. 'Society exists for the benefit of its members; not its members for the benefit of society ... the claims of the body politic are nothing in themselves, and become something only in so far as they embody the claims of its component individuals' (vol.1, pt II, ch. II, pp. 479-80). 17. In this sense and for these reasons we can and must speak of a collective consciousness distinct from individual consciousnesses. To justify this distinction there is no need to hypostatise the collective consciousness; it is something special and must be designated by a special term, simply because the states which constitute it differ specifically from those which make up individual consciousnesses. This specificity arises because they are not formed from the same elements. Individual consciousnesses result from the nature of the organic and psychical being taken in isolation, collective consciousnesses from a plurality of beings of this kind. The results cannot therefore fail to be different, since the component parts differ to this extent. Our definition of the social fact, moreover, did no more than highlight, in a different way, this demarcation line. 18. Inasmuch as it may exist before all animal life. Cf. on this point, A. Espinas, Des societes animates, (Paris, 1877) p. 474. 19. Division du travail social, II, ch. I. 20. Psychical phenomena can only have social consequences when they are so closely linked to social phenomena that the actions of both are necessarily intermingled. This is the case for certain socio-psychical phenomena. Thus a public official is a social force, but at the same time he is an individual. The result is that he can employ the social force he commands in a way determined by his individual nature and thereby exert an influence on the constitution of society. This is what occurs with statesmen and, more generally, with men of genius. The latter, although they do not fulfil a social role, draw from the collective sentiments of which they are the object an authority which is itself a social force, one which they can to a certain extent place at the service of their personal ideas. But it can be seen that such cases The Rules of Sociological Method by Emile Dürkheim Edited with an Introduction by Steven Lukes Translated by W. D. Halls THE FREE PRESS New York London Toronto Sydney