The Division of Labor in Society (1893)

[Excerpt from Robert Alun Jones. Emile Durkheim: An Introduction to Four Major Works. Beverly Hills, CA: Sage Publications, Inc., 1986. Pp. 24-59.]

Outline of Topics

  1. Durkheim's Problem
  2. The Function of the Division of Labor
  3. The Causes of the Division of Labor
  4. Abnormal Forms of the Division of Labor
  5. Critical Remarks

Durkheim's Problem

In 1776, Adam Smith opened The Wealth of Nations with the observation that "the greatest improvements in the productive powers of labour, and the greatest part of the skill, dexterity, and judgement with which it is anywhere directed, or applied, seem to have been the effects of the division of labour."1 Despite the numerous economic advantages thus derived, however, Smith insisted that the division of labor was not itself the effect of any human wisdom or foresight; rather, it was the necessary, albeit very slow and gradual, consequence of a certain propensity in human nature -- "the propensity to truck, barter, and exchange one thing for another."2 Common to all men, this propensity could be found in no other animals; and, subsequently encouraged by the recognition of individual self-interest, it gave rise to differences among men more extensive, more important, and ultimately more useful than those implied by their natural endowments.

More than a century later, Durkheim could observe, apparently without exaggeration, that economists upheld the division of labor not only as necessary, but as "the supreme law of human societies and the condition of their progress.3 Greater concentrations of productive forces and capital investment seemed to lead modern industry, business, and agriculture toward greater separation and specialization of occupations, and even a greater interdependence among the products themselves. And like Smith, Durkheim recognized that this extended beyond the economic world, embracing not only political, administrative, and judicial activities, but aesthetic and scientific activities as well. Even philosophy had been broken into a multitude of special disciplines, each of which had its own object, method, and ideas.

Unlike Smith, however, Durkheim viewed this "law" of the division of labor as applying not only to human societies, but to biological organisms generally. Citing recent speculation in the "philosophy of biology" (see the works of C.F. Wolff, K.E. von Baer, and H. Milne-Edwards), Durkheim noted the apparent correlation between the functional specialization of the parts of an organism and the extent of that organism's evolutionary development, suggesting that this extended the scope of the division of labor so as to make its origins contemporaneous with the origins of life itself. This, of course, eliminated any "propensity in human nature" as its possible cause, and implied that its conditions must be found in the essential properties of all organized matter. The division of labor in society was thus no more than a particular form of a process of extreme generality.

But if the division of labor was thus a natural law, then (like all natural laws) it raised certain moral questions. Are we to yield to it or resist it? Is it our duty to become thorough, complete, self-sufficient human beings? Or are we to be but parts of a whole, organs of an organism? In other words, is this natural law also a moral rule ? If so, why, and in what degree? In Durkheim's opinion, the answers of modern societies to these and similar questions had been deeply ambivalent -- i.e., on the one hand, the division of labor seemed to be increasingly viewed as a moral rule, so that, in at least one of its aspects, the categorical imperative" of the modern conscience had become: Make yourself usefully fulfill a determinate function"4; on the other hand, quite aside from such maxims endorsing specialization, there were other maxims, no less prevalent which called attention to the dangers of over -specialization, and encouraged all men to realize similar ideals. The situation was thus one of moral conflict or antagonism, and it was this which Durkheim sought first to explain and then to resolve.

This in turn calls for two final observations. First, the method of this explanation and resolution was to be that of the so-called "science of ethics"; for Durkheim was convinced that moral facts like the division of labor were themselves natural phenomena -- they consisted of certain rules of action imperatively imposed upon conduct, which could be recognized, observed, described, classified, and explained. Second, this explanation itself was but a preliminary step to the solution of practical social problems; for Durkheim always conceived of societies as subject to conditions of moral "health" or "illness," and the sociologist as a kind of "physician" who scientifically determined the particular condition of a particular society at a particular time, and then prescribed the social "medicine" necessary to the maintenance or recovery of well-being.

Durkheim's problem thus defined, his solution fell quite naturally into three principal parts:

  1. the determination of the function of the division of labor;
  2. the determination of the causes on which it depended; and
  3. the determination of those forms of "illness" which it exhibited.

The Function of the Division of Labor

The word "function," Durkheim observed, can be used in two, quite different, senses:

  1. to refer to a system of vital movements (e.g., digestion, respiration, etc.) without reference to the consequences of these movements; or
  2. to refer to the relationship between these movements and the corresponding needs of the organism (e.g., digestion incorporates food essential to replenish nutritional resources of the body, while respiration introduces the necessary gases into the body's tissues; etc.).

Durkheim insisted on the second usage; thus, to ask "what is the 'function' of the division of labor?" was simply to ask for the organic need which the division of labor supplied.

But at first sight, the answer to this question seemed all too clear; for, as Smith had already observed, the division of labor improves both the skill of the worker and the productive power of society, and thus its "function" would simply be to produce and secure those economic, artistic. and scientific advantages subsumed under the word "civilization." Against this, Durkheim presented two arguments. The first, which reveals Durkheim's deep, if ambivalent, debt to Rousseau, was that, if the division of labor has no other role than to render "civilization" possible, then there would be no reason to grant it the status of a "moral" fact -- of rules of action imperatively imposed upon conduct. On the contrary, if the average number of crimes and suicides is employed as the "standard of morality," Durkheim argued, we must conclude that immorality increases as the economy arts, and sciences progress. At its very best, therefore, civilization would be morally indifferent; and if its productions were the sole function of the division of labor, then it, too, would participate in this moral neutrality.

Durkheim's second argument was that, if the division of labor has no other role than to make civilization possible, then it would have no reason for existence whatsoever; for civilization, by itself, has no intrinsic value; rather, its value is derived entirely from its correspondence to certain needs. But these needs, Durkheim argued, are themselves the product of the division of labor. If the division of labor existed only to satisfy them, its only function would be to diminish needs which it itself had created. And this made little sense to Durkheim, for, while it might explain why we have to endure the division of labor, it would hardly be consistent with the fact that we desire occupational specialization and push it forward relentlessly. For the last to be intelligible, we must assume that the division of labor satisfies needs which the division of labor has not itself produced.

What, then, are these "needs" satisfied by the division of labor? As a first step toward an answer, Durkheim posed a paradox as old as Aristotle -- that, while we like those who resemble us, we are also drawn toward those who are different, precisely because they are different. In other words, difference can be as much a source of mutual attraction as likeness. The key to resolving the paradox, Durkheim suggested, lies in recognizing that only certain kinds of differences attract -- specifically, those which, instead of excluding one another, complement one another: "If one of two people has what the other has not, but desires, in that fact lies the point of departure for a positive attraction."5 In other words, we seek in others what we lack in ourselves, and associations are formed wherever there is such a true exchange of services -- in short, wherever there is a division of labor.

But if this is the case, we are led to see the division of labor in a new light6 -- the economic services it renders are trivial by comparison with the moral effect it produces. Its true function, the real need to which it corresponds, is that feeling of solidarity in two or more persons which it creates. Thus, the role of the division of labor is not simply to embellish already existing societies, but to render possible societies which, without it, would not even exist; and the societies thus created, Durkheim added, cannot resemble those determined by the attraction of like for like. Rather, they must bear the mark of their special origin.

The last point laid the immediate foundations for the next step in Durkheim's argument. Thus far, he had shown only that, in advanced societies, there is a social solidarity derived from the division of labor, something already obvious from two facts: that the division of labor does produce a kind of solidarity, and that the division of labor is highly developed in advanced societies. The question which remained was both more important and more difficult to answer: To what degree does the solidarity produced by the division of labor contribute to the general integration of society? This question was important because only by answering it could Durkheim determine whether this form of solidarity was essential to the stability of advanced societies, or was merely an accessory and secondary condition of that stability; and it was difficult because an answer required the systematic comparison of this form of solidarity with others, in order to determine how much credit, in the total effect, was due to each. Such a comparison in turn required a classification of the various types of solidarity to be compared, and here Durkheim faced one of the most formidable obstacles to his science of ethics: the fact that, as a "completely moral phenomenon," social solidarity did not lend itself to exact observation or measurement.

Durkheim's way of surmounting this obstacle was to substitute for this internal, moral fact an "external index" which symbolized it, and then to study the fact in light of the symbol. This external symbol was law -- i.e., where social life exists, it tends to assume a definite, organized form, and law is simply the most stable and precise expression of this organization. Law reproduces the principal forms of solidarity; and thus we have only to classify the different types of law in order to discover the different types of solidarity corresponding to them.

This proposal encountered two immediate difficulties. The first was that some social relations are regulated not by law, but by custom; moreover, custom is frequently at odds with law, and thus may express an altogether different form of social solidarity. Here Durkheim resorted to one of his favorite (and least convincing) defenses -- i.e., the distinction between the normal and the pathological. The conflict between law and custom arises where the former no longer corresponds to existing social relations, but maintains itself by habit, while the latter corresponds to these new relations, but is denied juridical expression. But such conflict, Durkheim insisted, is both rare and pathological; the normal condition is one in which custom is the very basis of law, in which custom alone can manifest only secondary forms of social solidarity, and thus in which law alone tells us which forms of social solidarity are essential. This purely arbitrary distinction, incidentally, reveals not only a profound discomfort with the ethnographic study of primitive societies, but a concerted effort to rationalize this discomfort as well.

The second objection was that social solidarity does not completely manifest itself in any perceptible form whatsoever, for law (and even custom) are but the partial, imperfect manifestations of internal psychological states which are thus the more appropriate focus for our investigations. Durkheim's response contained three interrelated arguments; first, that we can determine the nature of social solidarity scientifically only by studying its most objective and easily measurable effects (such as law); second, that, while solidarity "depends on" such internal states, these are not equivalent to social solidarity itself; and, finally, that these states themselves depend on social conditions for their explanation, a fact which explains why at least some sociological propositions find their way into the purest analyses of psychological facts.7

How, then, do we classify the different types of law? If the classification is to be scientific, Durkheim argued, we must do so according to some characteristic which both is essential to laws and varies as they vary. This characteristic is the sanction -- i.e., "Every precept of law can be defined as a rule of sanctioned conduct. Moreover, it is evident that sanctions change with the gravity attributed to precepts, the place they hold in the public conscience the role they play in society."8 These sanctions, Durkheim then observed, fall into two classes: repressive sanctions (characteristic of penal laws), which consist in some loss or suffering inflicted on the agent, making "demands on his fortune, or on his honor, or on his life, or on his liberty, and deprive him of something he enjoys."9; and restitutive sanctions (characteristic of civil, commercial, procedural, administrative, and constitutional laws), which consist "only of the return of things as they were, in the re-establishment of troubled relations to their normal state."10

The two types of law thus classified according to their characteristic sanctions, Durkheim was now in a position to determine the types of solidarity corresponding to each. The first of these Durkheim called mechanical solidarity -- that type of solidarity characterized by repressive sanctions. And since acts calling forth such sanctions are (by definition) "crimes," then the inquiry into the nature of mechanical solidarity became an inquiry into the nature of crime.

What, then, is "crime"? While acknowledging that there are many kinds of crime, Durkheim was convinced that they all contained a common element; for otherwise the universally identical reaction to crimes (repressive sanctions) would itself be unintelligible. Nonetheless, the enormous variety of crimes suggested that this common element could not be found among the intrinsic properties of criminal acts themselves; rather, it had to be found in the relations which these acts sustain with certain external conditions. But which relations? After some characteristic annihilations of competing proposals, Durkheim concluded that the only common element in all crimes is that they shock sentiments which, "for a given social system, are found in all healthy consciences."11 And this also explains why penal (as opposed to civil) law is "diffused" throughout the whole society rather than centralized in a special magistrate -- the sentiments to which penal law corresponds are immanent in all consciousnesses.12

But what about acts like incest -- acts which provoke widespread aversion, but are merely "immoral" rather than "criminal"? Durkheim replied that "crimes" properly so-called have an additional distinctive property not shared by simply "immoral" acts: the sentiments they offend must have a certain average intensity. And again, this greater intensity of sentiments responsive to crime as opposed to immoral acts is reflected in the fixity of penal law over time, by contrast with the great plasticity of moral rules. Finally, the sentiments responsive to criminal acts are also more well-defined than those nebulous sentiments evoked by immorality.

Durkheim's definition of crime thus led directly to his notion of the conscience collective -- "the totality of beliefs and sentiments common to the average citizens of the same society"13 -- which Durkheim then endowed with quite distinctive characteristics: it forms a determinate system with its own life; it is "diffuse" in each society and lacks a "specific organ"; it is independent of the particular conditions in which individuals find themselves; it is the same in different locations, classes, and occupations; it connects successive generations rather than changing from one to another; and it is different from individual consciences, despite the fact that it can be realized only through them. A "crime," therefore, is simply an act which offends intense and well-defined states of this conscience collective, a proposition which describes not simply the "consequences" of crime, but its essential property: "We do not reprove it because it is a crime, but it is a crime because we reprove it."14

But aren't there acts which do not offend the conscience collective, but which are nonetheless severely sanctioned by the state? And are there then two distinct types of crime? Durkheim insisted there are not, for the effects called forth by criminal acts are the same in either case, and the same effect must have the same cause. Durkheim was thus led to argue that the state derives its authority from the conscience collective, and becomes its directive organ and its symbol but, while the state never completely frees itself from this source of its authority, it does become an autonomous, spontaneous power in social life. The extent of the state's power over the number and nature of criminal acts depends on the authority it receives from the conscience collective; and this authority can be measured either by the power the state exerts over its citizens, or by the gravity attached to crime against the state. As Durkheim would show, this power was greatest and this gravity most pronounced in the lowest, most primitive societies; and it was in these societies that the conscience collective enjoyed the greatest authority.

In effect, therefore, Durkheim argued that crime is characterized its capacity to provoke punishment. But if this was the case, crime ought to explain the various characteristics of punishment, and any, demonstration that it did so would augment the plausibility of Durkheim's initial argument. What, then, are the characteristics of punishment? Disregarding the conscious intentions of those applying it, Durkheim insisted that the characteristics of punishment are what they have always been -- its mood is passionate; its function is vengeance, even expiation; its intensity is variable or "graduated"; its source is society rather than the individual; its cause is the violation of a moral rule; and its form is "organized" (unlike the "diffuse" repression of merely immoral acts, its implementation is the act of a definitely constituted body or tribunal). In short, punishment is "a passionate reaction of graduated intensity that society exercises through the medium of a body acting upon those of its members who have violated certain rules of conduct."15

How are these characteristics to be explained? Durkheim first observed that every state of conscience is an essential source of life, and everything that weakens such a state "wastes and corrupts" us; thus we react energetically against those ideas and sentiments which contradict our own. But the ideas and sentiments offended by crime, Durkheim argued, have particular features which in turn explain the special characteristics of punishment: i.e., because these sentiments are held with particular strength, the reaction is passionate; because these sentiments transcend individual mental states, mere restitution is unacceptable, and revenge and even expiation are called for; because the vivacity of such sentiments will nonetheless vary, the intensity of the reaction will also be variable; because such sentiments are held collectively, the source of the reaction will be society rather than the individual; and because these sentiments are well-defined, the reaction to their violation will be organized.

Having begun by establishing inductively that "crime" is an act contrary to strong and well-defined states of the conscience collective, therefore, Durkheim confirmed this definition by showing that crime thus defined accounts for all the characteristic features of punishment; and since the whole point of Durkheim's inquiry into the nature of crime was its promise to reveal the nature of mechanical solidarity, we might reasonably ask what has been thus revealed. Durkheim's answer was that the cause of mechanical solidarity lies in the conformity of all individual consciences to a common type, not only because individuals are attracted to one another through resemblance but because each is joined to the society that they form by their union; inversely, the society is bound to those ideas and sentiments whereby its members resemble one another because that is a condition of its cohesion.

Durkheim thus introduced an idea which would assume increasing importance in his later work: the duality of human nature. Briefly, in each of us there are two consciences -- one containing states personal to each of us, representing and constituting our individual personality; the other containing states common to all, representing society, and without which society would not exist. When our conduct is determined by the first, we act out of self-interest; but when it is determined by the second, we act morally, in the interest of society. Thus the individual, by virtue of his resemblance to.other individuals, is linked to the social order. This is mechanical solidarity, which, as we have seen, is manifested through repressive law; and the greater the number of repressive laws, the greater the number of social relations regulated by this type of solidarity.

The very nature of restitutive sanctions, however, indicates that there is a totally different type of social solidarity which corresponds to civil law; for the restitutive sanction is not punitive, vengeful, or expiatory at all, but consists only in a return of things to their previous, normal state. Neither do violations of civil laws evoke the milder, more diffuse disapproval of merely moral transgressions, in fact, we can imagine that the laws themselves might be quite different than they are without any feeling of moral repugnance being aroused. Durkheim thus concluded that such laws, manifested in restitutive sanctions, could not derive from any strong state of the conscience collective, but must have some other source.

An indication of this source was afforded by an examination of the conditions under which such rules are established. Briefly, there are some relationships (typically, those involving contractual obligations) which the consent of the interested parties is not sufficient to create or to change; on the contrary, it is necessary to establish or modify such relationships juridically, by means of law. While contracts are entered and abrogated through the efforts of individuals, therefore, they have a binding, obligatory power only because they are supported and enforced by society. Most important, the contractual relations thus regulated are not "diffused" throughout the society; they do not bind the individual to society, but rather bind special parties in the society to one another.

The cooperative relations thus formed create what Durkheim called organic solidarity, which is derived not from the conscience collective, but from the division of labor. For, where mechanical solidarity presumes that individuals resemble one another, organic solidarity presumes their difference; and again, where mechanical solidarity is possible only in so far as the individual personality is submerged in the collectivity, organic solidarity becomes possible only in so far as each individual has a sphere of action peculiar to him. For organic solidarity to emerge, therefore, the conscience collective must leave untouched a part of the individual conscience so that special functions, which the conscience collective itself cannot tolerate, may be established there; and the more this region of the individual conscience is extended, the stronger is the cohesion which results from this particular kind of solidarity.

Durkheim had thus postulated two distinct types of social solidarity (mechanical and organic), each with its distinctive form of juridical rules (repressive and restitutive). In order to determine their relative importance in any given societal type, therefore, it seemed reasonable to compare the respective extent of the two kinds of rules which express or symbolize them. The preponderance of repressive rules over their restitutive counterparts, for example, ought to be just as great as the preponderance of the conscience collective over the division of labor; inversely, in so far as the individual personality and the specialization of tasks is developed, then the relative proportion of the two types of law ought to be reversed.

In fact, Durkheim argued, this is precisely the case. Despite the flimsy ethnographic evidence supporting such generalizations, Durkheim argued that the more primitive societies are, the more resemblances (particularly as reflected in primitive religion) there are among the individuals who compose them16; inversely, the more civilized a people, the more easily distinguishable its individual members.17 Durkheim's discomfort with the ethnographic literature was still more evident when he turned to the nature of primitive law. Relying on Sir John Lubbock's Origin of Civilization (1870) and Herbert Spencer's Principles of Sociology (1876-85), he suggested that such law "appears to be entirely repressive"18, but insisting that such observations necessarily lack precision, Durkheim instead pointed to the evidence of written law. Moving from the Pentateuch to the "Twelve Tables" (451-450 B.C.) of the Romans to the laws of early Christian Europe, therefore, Durkheim argued that the relative proportions of repressive to restitutive laws are precisely those which his theory would lead us to expect.19

When we reach the present, therefore, we find that the number of relationships which come under repressive laws represents only a small fraction of social life; thus, we may assume that the social bonds derived from the conscience collective are now much less numerous than those derived from the division of labor. But one might still argue that, regardless of their number, the bonds which tie us directly to our societies through shared beliefs and sentiments have greater strength than those resulting from cooperation; and to this hypothetical objection, Durkheim had two independent answers.

First, he felt that, regardless of their undeniable rigidity the bonds created by mechanical solidarity, even in lower societies, were inferior to those created by organic solidarity in their more advanced counterparts. Here, again, Durkheim's ethnographic resources were limited to a few passages cited from Spencer, Fustel de Coulanges, and Theodor Waitz's Anthropologie der Naturvölker (1859); but the source of this conviction, in any case, was less empirical than theoretical. Where, as in lower societies, the conscience collective is virtually coextensive with the individual conscience, each individual "contains within himself all that social life consists of," and thus can carry "society" wherever he wishes to go; inversely, the society, given its rudimentary division of labor, can lose any number of its members without its internal economy being disturbed. Thus, from both standpoints the bonds connecting the individual to society based upon the conscience collective are less resistant to disseveration than those based upon the division of labor.

Durkheim's second answer was that, as society evolves from a lower to a higher type, the bonds created by mechanical solidarity become still weaker. The strength of mechanical solidarity, Durkheim argued, depends on three conditions:

  1. the relation between the volume of the conscience collective relative to the individual conscience;
  2. the average intensity of the states of the conscience collective; and
  3. the degree of definition of the states of the conscience collective.

As we have seen, intense and well-defined states of the conscience collective are the basis of repressive laws; and, since we have also seen that the proportion of such laws has declined, it seems reasonable to assume that the average intensity and degree of definition of the conscience collective have also declined. The same, Durkheim admitted, cannot be said about the relative volume of the conscience collective; for, while that "region" of the conscience collective manifested by repressive laws has no doubt contracted, that region of the same conscience expressed through less intense and more vague sentiments of custom and public opinion may in fact have expanded. But meanwhile, Durkheim argued, the volume of the individual conscience has grown in at least equal proportions; for, "if there are more things common to all, there are many more that are personal to each."20 The most we can say of the relative volume of the conscience collective, therefore, is that it has remained the same; for it certainly has not gained, and it may have lost. And if we could prove what we already have good reason to assume -- that the conscience collective has become both less intense and more vague over time -- then we could be sure that mechanical solidarity has become weaker over the same period.

How could such proof be provided? Not by comparing the number of repressive rules in different societal types, Durkheim emphasized, for this number alone does not vary exactly with the sentiments thus represented. Instead, Durkheim simply grouped the rules into classes corresponding to the types of sentiments aroused by their violation. The result was a list of "criminological types," whose number would necessarily correspond to the number of intense, well-defined states of the conscience collective: "The more numerous the latter are, the more criminal types there ought to be, and consequently, the variations of one would exactly reflect the variations of the other."21

The conclusion of Durkheim's investigation, of course, was that a large number of criminological types -- those expressed by repressive laws governing sexual relations, domestic, and, most dramatically, religious life -- had progressively disappeared over the centuries; and this in turn suggested that the states of the conscience collective had indeed become less intense and more vague, and that mechanical solidarity was commensurately weakened. The notable exception here, as Durkheim was careful to point out, were those states of the conscience collective which have the individual as their object, as in the protection of the individual's person and rights. And this Durkheim (in effect) suggested, is indeed an exception which proves the rule; for it could become possible only if the individual personality had become far more important in the society, and thus only if the personal conscience of each individual had grown considerably more than the conscience collective itself. To this other proofs were added: the decline of religion (which, at this time Durkheim literally defined as strong, commonly held beliefs) and the disappearance of those proverbs and adages whereby "collective thought condenses itself."22 All conspired to make the same point: that the conscience collective had progressed less than the individual conscience, becoming less intense and distinct, and more abstract and indecisive.

Will the conscience collective then disappear? Durkheim thought not, at least in part because of the "notable exception" mentioned above -- it not only survives, but becomes more intense and well-defined, in so far as its object is the individual: "As all the other beliefs and all the other practices take on a character less and less religious, the individual becomes the object of a sort of religion. We erect a cult on behalf of personal dignity which, as every strong cult already has its superstitions."23 But while it is from society that this cult gathers its force, it is not to society, but to ourselves, that it attaches us; thus Durkheim denied that it was a true social link, and repeated his argument that all such links derived from likeness have progressively weakened. If society itself is to survive, therefore, there must be some other "true social link" which replaces it, and this, of course, is organic solidarity, the product of the division of labor.

But if the way in which men are linked together has thus evolved from mechanical to organic solidarity, there should be parallel changes in the structural features of the societies themselves. What kind of social structure, therefore, might we expect to find in a society whose cohesiveness is based primarily on resemblances? Briefly, we would expect what Durkheim called the horde -- an absolutely homogeneous mass of indistinguishable parts, devoid of all form. arrangement, or organization. Durkheim admitted that no societies fitting this description had ever been observed; but among both the Iroquois24 and Australian25 tribes, he found societies made up of a number of groups of this kind. Durkheim thus gave the name clan to the horde which had become an element of a more extensive group, and used the term segmental societies with a clan base to refer to peoples thus constituted through an association of clans.

Durkheim chose the term "clan" because these groups are both familial (i.e., all members are regarded as "kin," most are consanguineous, and they practice collective punishment, collective responsibility, and, once private property appears, mutual inheritance) and political (i.e., not all members are consanguineous, some merely bear the same name; it attains dimensions much larger than any "family," and the heads of clans are the sole political authorities). Most important, however, the clan is internally homogeneous, and its solidarity is thus based on resemblances.26 Even the clans themselves must bear certain resemblances if segmental organization is to be possible, although their differences must also be sufficient to prevent them from "losing themselves" in one another. This, then, is the social structure of mechanically solidary societies.

But there is also a social structure to which organic solidarity corresponds. Typically, such societies are constituted not by homogeneous segments, but by a system of different organs, each of which has a special role, and which themselves are formed of differentiated parts. These parts are also arranged differently: rather than being merely juxtaposed or mingled, they are coordinated and subordinated to one another around a central organ, which exercises a regulative action on the entire organism. Finally, the place of each individual in such societies is determined not by his name or kin-group, but by the particular occupation or social function to which he is committed.

This is what Durkheim called the organized societal type which, because of its sharp differences from the segmental type, can advance only in so far as the latter is gradually effaced. But Durkheim was also aware of the considerable complexity of the transition from one to the other, and provided a particularly subtle account of the almost parasitical manner in which the new occupational "organs" at first utilize the old familial system (as when Levites became priests), the subsequent process whereby consanguineous ties give way to less resistant bonds based upon territorial allegiances, and, finally, the complete triumph of the fully "organized" societal type over the structural constraints of its earlier, "segmental," counterpart. As with the primitive horde, Durkheim admitted that this organized type was nowhere presently observable in its purest form; but he added that "a day will come when our whole social and political organization will have a base exclusively, or almost exclusively, occupational."27

Thus far, Durkheim's argument would have appeared relatively familiar to his contemporaries, for it bore an unmistakable similarity to that found in Spencer's Principles of Sociology (1876-1885), particularly in its emphasis on the growth of individuality with the advance of civilization. This similarity was sufficiently upsetting to Durkheim to provoke a more detailed account of his differences with Spencer. For the latter, for example, the submersion of the individual in lower societies was the result of force, an artificial suppression required by the essentially despotic, "military" type of organization appropriate to an early stage of social evolution. For Durkheim, by contrast, the effacement of the individual was the product of a societal type characterized by the complete absence of all centralized authority; military personality in lower societies was a consequence not of suppression, but of the fact that, in those societies, the "individual," as such, did not exist. Reversing Spencer's argument, therefore, Durkheim saw the emergence of despotic authority not as a step toward the effacement of the individual, but as the first step toward individualism itself, the chief being the first personality to emerge from the previously homogeneous social mass.

But there was more to this than a typical Durkheimian annihilation of an intellectually inferior opponent; for Durkheim sought to establish two important propositions. The first of these was hinted at in our earlier discussion of Durkheim's view of the state -- that when we find a governmental system of great authority. we must seek its cause not in the particular situation of the governing, but in the nature of the societies governed. The second was that altruism, far from being a recent advance over man's selfish, egoistic tendencies, is found in the earliest societies; for, as we have seen, Durkheim had a dualistic conception of human nature, and thus both egoism and altruism were natural expressions of the human conscience at all stages of social evolution.

What, then, is the essential difference between lower societies and our own? Durkheim's answer was again worked out in opposition to Spencer, whose own answer again appeared quite similar. Spencer had observed, like Durkheim, that in industrial societies a cooperative form of solidarity is produced automatically as a consequence of the division of labor. But if Spencer thus recognized the true cause of social solidarity in advanced societies, Durkheim argued, he had not understood the way in which it produced its effect; and, misunderstanding this, Spencer had misunderstood the nature of the effect (i.e., social solidarity) itself.

Consider only two features of Spencer's conception of social solidarity: because industrial solidarity is produced automatically, it does not require the regulation or intervention of the state in order to produce or maintain it; and because the sphere of societal action is thus drastically reduced, the only surviving link between men is the relationship of contracts, freely entered and freely abrogated, according to the self-interest of the parties involved. Durkheim's initial response was that, if this is truly the character of societies whose solidarity is produced by the division of labor, we might with justice doubt their stability; for "self-interest" creates only the most ephemeral, superficial sort of social bond, and in fact disguises a more fundamental, albeit latent and deferred, conflict. The large and increasing volume of restitutive law, moreover, hardly suggested to Durkheim that the regulative intervention of the state in contractual relations was decreasing; on the contrary, it suggested that unregulated contracts alone were insufficient to secure equal justice for their contending parties -- particularly the worker in contractual relations between labor and management. While Spencer was right to point to the increase in the number of social relationships governed by contract, he ignored the parallel increase in the number of non -contractual relations; but most important, he ignored the fact that, even within the contract, "everything is not contractual" -- i.e., a contract assumes the predetermination of the rights and obligations of the contracting parties, a function performed not only by state-regulated contract law, but also through the less formal but nonetheless imperative structures of custom.

In short, Spencer did not understand the nature of social solidarity nor did he understand the function of the division of labor. Whatever its economic advantages, the function of the division of labor was pre-eminently moral. In fact, contrasting the solidarity created by occupational specialization with the "inferior" bonds forged by its mechanical counterpart, Durkheim insisted that the moral character of society is more pronounced in the "organized" type. Precisely because the modern individual is not sufficient unto himself, for example, it is from society that he receives all that is necessary to life; thus is created his strong sentiment of personal dependence which inspires those mundane sacrifices we call "moral acts" and, in occasional, extreme cases, those acts of complete self-renunciation which Durkheim would take up in Suicide (1897). On its side, society learns to regard its members not as indistinguishable units that could be lost without serious disruption to its internal economy, but as irreplaceable organic parts which it cannot neglect, and towards which it has important obligations. It was the perfection of this moral function toward which all social evolution tended.

The Causes of the Division of Labor

Durkheim was always concerned to distinguish the causes of a social fact from its functions, and the division of labor was no exception. Indeed, he insisted, the causes of the division of labor could not possibly consist in some anticipation of its moral effects; for, as we have seen, those effects became evident only after a lengthy process of social evolution, and could hardly be foreseen. In a different sense, however, Durkheim's inquiry into causes rehearsed his earlier analysis of functions; for, just as the earlier discussion began with Durkheim's rejection of Adam Smith's argument that the function of the division of labor was the advancement of civilization, so the later discussion began with a negative assessment of that "classic" explanation, attributed to political economy in general, whereby the cause of the division of labor would be "man's unceasing desire to increase his happiness."28

Against this explanation, which would reduce the division of labor to purely individual and psychological causes, Durkheim launched a three-pronged attack. First, he challenged the axiom on which the explanation rests -- namely, the assumption that man's desire to increase his happiness is indeed unceasing. Here Durkheim's early experience in Wundt's psychological laboratory served him well, for he was able to cite the famous law of the German experimental psychologist E.H. Weber (later quantified by Gustav Fechner) to the effect that the smallest increment in a stimulus required to produce a, difference in the sensation experienced is not an absolute amount, but is rather relative to the magnitude of the stimulus in question. As a corollary to this law, Durkheim insisted that the intensity of any agreeable stimulus can increase usefully (i.e., contribute to increased pleasure) only between two extremes. An increase in monetary wealth, for example, must be of a certain size if pleasure is to be its result; inversely, a person thoroughly accustomed to large increases in wealth estimates the value of such increases accordingly, and is equally denied pleasure proportionate to the stimulus received. The increase in income experienced by the man of average wealth is thus the one most apt to produce a degree of pleasure proportionate to its cause. If the cause of the division of labor were the desire for happiness, therefore, social evolution would surely have come to a stop long ago; for the maximum happiness of which men are capable would have been achieved through a relatively moderate development of social differentiation and its resulting stimuli.29 This insistence that the human capacity for happiness is very limited, a kind of Aristotelian ethics augmented by Wundt's Grunzuge der physiologische Psychologie (1874), remained one of Durkheim's most constant and characteristic ideas.

Second, Durkheim regarded it as very doubtful that the advance of civilization increases human happiness in any case. Here Durkheim initially sounds like Rousseau: while he admitted that we enjoy pleasures unknown to earlier societies, he observed also that we experience forms of suffering that they were spared, and added that it is not at all certain that the balance is in our favor. But it soon becomes clear that, again. Durkheim's more fundamental source was Aristotle. Even if social progress did produce more pleasure than pain, Durkheim thus insisted, this would not necessarily bring more happiness; for "pleasure" describes the local, limited, momentary state of a particular function, while "happiness" describes the health of the physical and moral species in its entirety, the extent to which that species has realized its true nature. Thus, the normal savage is just as happy as the normal civilized man, an argument supported not only by Waitz's Anthropologie der Naturvölker (1859), but also by the rapid rise in the suicide rate commensurate with the advance in civilization, a phenomenon in which Durkheim already had a powerful interest.

Durkheim's third argument dealt with a revised version of the "happiness hypothesis" which might have met the objections of his first two -- that pleasure (which is at least an element in happiness) loses its intensity with repetition, and can be recaptured only through new stimuli, meaning more productive work (and hence, through the division of labor). Progress would thus be, quite literally, an effect of boredom. But to this Durkheim had several objections. First, such a "law" would apply to all societies, and thus it could provide no account of why the division of labor advances in some societies and not in others. Second, Durkheim denied the assumption on which the argument is based: namely, that repetition alone reduces the intensity of pleasure. So long as our pleasures have a certain variety, he argued, they can be repeated endlessly; only if the pleasure is continuous and uninterrupted does its intensity wane. But even if continuity thus does what repetition cannot, Durkheim continued, it could not inspire us with a need for new stimuli; for if continuity eliminates our consciousness of the agreeable state, we could hardly perceive that the pleasure attached to it has also vanished. Even novelty itself is but a secondary, accessory quality of pleasure, without which our ordinary pleasures, if sufficiently varied, can survive very well. In short, boredom is an insufficient cause to so painful and laborious an effect as the development of the division of labor.

Having thus dismissed individualistic, psychologistic causes, Durkheim argued that we must seek the explanation of the division of labor in some variation within the social context, and added that his earlier discussion of its function already pointed in the direction of an answer. Durkheim had shown how the organized structure (and thus the division of labor) had developed as the segmental structure had disappeared; thus, either the disappearance of the segmental structure is the cause of the division of labor, or vice versa. Since, as we have seen, the segmental structure is an insurmountable obstacle to the division of labor, the latter hypothesis is clearly false; the division of labor can thus appear only in proportion as the segmental structure has already begun to disappear.

How does this occur? Briefly, Durkheim suggested that, instead of social life being concentrated in a number of small, identical individual segments, these parts begin to extend beyond their limits, exchange movements, and act and react upon one another. Durkheim called this dynamic or moral density, and suggested that it increases in direct ratio to the progress of the division of labor. But what produces this "moral density"? Durkheim pointed to two causes. First, the real, material distance between members of a society must be reduced both spatially (e.g., the growth of cities) and technologically (e.g., advances in communications and transportation), for such "material density" multiplies the number of intra-societal relations. Second, this effect is reinforced by the sheer "social volume" of a society (the total number of its members). Thus, Durkheim argued that the division of labor varies in direct ratio to the dynamic or moral density of society, which is itself an effect of both material density and social volume.30

But how does this double cause (material density and social volume) produce its ultimate effect (the division of labor)? Here again, Durkheim had to confront the competing explanation of Herbert Spencer. In First Principles (1862), Spencer had argued that all homogeneous masses are inherently unstable and thus tend toward differentiation, and that they differentiate more rapidly and completely as their extension is greater. But in Spencer's theory, such extension produces differentiation, not by itself, but only in so far as it exposes parts of the social mass to diverse physical environments, thus encouraging diverse aptitudes and institutional specialization. Durkheim in fact agreed that a diversity of external circumstances has this differentiating effect; but he denied that this diversity was sufficient to cause (rather than merely accelerate ) an effect so dramatic as the division of labor.

For his own explanation, Durkheim turned to Darwin's Origin of Species (1859), arguing that an increased material density and social volume cause the division of labor, not because they increase exposure to diverse external circumstances, but because they render the struggle for existence more acute. According to Darwin, so long as resources are plentiful and population size is limited, similar organisms can live side by side in relative peace; but where population increases and resources become scarce, conflict and competition ensue, and this conflict is just as active as the organisms are similar and pursue similar needs. Where organisms are different and pursue different needs, on the other hand, what is useful to one organism will be of no value to another, and conflict will diminish.

Human populations, Durkheim argued, adhere to the same law. In so far as a social structure is "segmental" in character, each segment has its own organs, kept apart from like organs by the divisions between segments. With the growth in the "material density" and "social volume" of the society, these divisions disappear, the similar organs are put into contact with one another, and competition between them ensues. Those groups which triumph then have a larger task, which can be discharged only through a greater internal division of labor; those organs which are vanquished can henceforth maintain themselves only by specializing on a fraction of the social function they previously performed; but in either case, the division of labor is advanced.

Thus, the conflict and competition resulting from an increase in social volume and density produces advances in the division of labor just as the latter mitigates against the negative consequences of the former. In the modern city, for example, large and highly condensed populations can coexist peacefully as a consequence of occupational differentiation: "The soldier seeks military glory, the priest moral authority, the statesman power, the businessman riches, the scholar scientific renown. Each of them can attain his end without preventing the others from obtaining theirs."31 Nothing in this process, Durkheim added, implies an increase in happiness, or that the pursuit of happiness might be its goal: on the contrary, "everything takes place mechanically" as the result of an inexorable law of social progress.

Finally, Durkheim argued, it is a corollary of this law that the division of labor can be established only among the members of an already constituted society. For the effect of these same forces (e.g., opposition, conflict, competition, etc.) upon a number of independent individuals could only be further diversification without the development of compensatory social bonds32, while Durkheim had already shown that the division of labor creates moral linkages even as it differentiates. Durkheim thus argued that the individuals among whom the struggle for existence is waged must already belong to the same, mechanically solidary society. In opposition to Spencer's view that a society is the product of cooperation, therefore, Durkheim supported Comte's argument that cooperation already presupposes the spontaneous existence of society.33 This, in turn, became the basis for Durkheim's reply to Brunetière at the height of the Dreyfus Affair. Far from being destructive of the social order, individualism is itself the product of society, and expresses a particular stage in its ongoing, structural evolution.

Durkheim had thus argued forcefully that the division of labor is caused by changes in the volume and density of societies. But this was not yet a complete explanation, for Durkheim recognized that such specialization was not the only possible solution to the struggle for existence which then ensued. Others included emigration, colonization, resignation to a precarious existence, and even suicide. The division of labor was thus a contingent rather than a necessary consequence of changes in the social environment, and for it rather than its alternatives to result, it was essential that the influence of at least two secondary factors -- the conscience collective and heredity -- be significantly reduced.

Durkheim's argument concerning the "progressive indetermination" of the conscience collective has already been described; but now Durkheim attempted to explain it, focusing equally on the growth of rationality and the decline of tradition. In early societies, Durkheim began, everyone is related to specific objects of their environment (e.g., animals, trees, plants, etc.) in roughly the same way, and the states of conscience representing this environment take on a parallel similarity; the fusion of these individual consciences thus results in a conscience collective which is sharp, decisive, and well-defined. As these societies become more voluminous and their populations more diversely situated, however, common objects can no longer create common experiences and representations; in so far as it is to remain "common," therefore, the conscience collective must necessarily become less concrete and well-defined, and more general and abstract. The "animal" becomes the "species," the "tree" becomes "trees in general and in abstracto," the "Greek" and the "Roman" become the concept of "man"; and a similar process of progressive abstraction up to the level of universalizable concepts persists in law, religion. and morality. This explains the difficulty we have in understanding primitive societies. Our own minds, dominated by the logic and rationality this evolutionary process has produced, see in earlier societies only bizarre, fortuitous combinations of heterogeneous elements; but in fact, these are simply societies dominated by concrete sensations and representations rather than abstract concepts.34

But in so far as the conscience collective thus becomes less concrete and decisive, it necessarily has less of an impact on individual thought and behavior. Precise states of conscience act in a manner analogous to instinctive reflexes; more general principles affect behavior only through the intervening reflections of intelligence. Thus, "deliberated movements have not the spontaneity of involuntary movements. Because it becomes more rational, the [ conscience collective ] becomes less imperative, and for this very reason, it wields less restraint over the free development of individuals."35 But the cause of this growth of rationality, again, is the increase in the volume of the society's population and the environmental diversity thus implied.

Still more important than the "progressive indetermination" of the conscience collective, however, is the decline of tradition; for the strength of the conscience is due to the fact not only that its states are shared, but also that they are the legacy of previous generations. This authority of tradition is well supported in societies of the segmental type, which, as we have seen, have a familial as well as a political base; but as the segmental organization is undermined, individuals no longer feel bound to their kin-group or even their place of origin; migration ensues, and the authority of tradition weakens commensurately. But here, again, the decline of tradition is the consequence of those factors -- social volume and density -- which gradually dissipate the segmental form of social organization. In other words, just as it is purely mechanical causes which lead to the individual's submersion in the conscience collective, it is similarly mechanical causes ( not the "utility" of emancipation) which subvert that conscience and lead to individual freedom.

But don't the occupational specialities of more organized societies simply reproduce the conscience of the primitive segment, and exercise the same regulative function. For at least three reasons, Durkheim's answer was an emphatic no: first, the occupational conscience affects only the occupational life, beyond which the individual enjoys much greater freedom; second, the occupational conscience is shared by fewer individual minds, has commensurately less authority, and thus offers less resistance to individual transgressions than its collective counterpart; and third, the same causes (i.e., increased volume and density) which progressively undermine the conscience collective have a similar, if less dramatic, effect within the occupational group. Thus, "not only does occupational regulation, because of its very nature, hinder less than any other the play of individual variation, but it also tends to do so less and less."36

The other "secondary factor" whose influence had to be reduced in order for the division of labor to emerge was the role of heredity. Durkheim was particularly concerned with this because, according to John Stuart Mill's Principles of Political Economy (1848), the first condition of the division of labor was that "diversity of natures" whose principal function was to classify individuals according to their capacities. If this were the case, Durkheim argued, heredity would, constitute an even more insurmountable obstacle to individual variability than the conscience collective; for, where the latter chained us only to the moral authority of our familial group, the former would bind us to our race, and thus to an utterly impersonal, congenital past, totally oblivious to our individual interests and aspirations. Thus, the greater the role of heredity in a society's distribution of tasks (as, for example, in the caste system, or in rigidly stratified societies), the more invariable that distribution, and the more difficult it is for the division of labor to make headway. It was Durkheim's goal, however, to show that, for at least two reasons, the role played by heredity in the distribution of tasks has declined in the course of social evolution.

First, Durkheim observed, aptitudes appear to be less transmissible by heredity precisely to the degree that they are more specialized; in so far as a society has a more complex division of labor, therefore, the relative role played by heredity in determining individual capacities will have been reduced. In short, social evolution produces new modes of activity requiring capacities that heredity simply cannot transmit. Second, Durkheim insisted, even those capacities that heredity can transmit (e.g., instincts) decline both in number and strength with social evolution.37 Whether conceived relatively or absolutely, therefore. the contribution of heredity to the determination of individual tasks has been progressively reduced, and has thus presented few obstacles to the continuing growth of the division of labor.

This led Durkheim to some general conclusions about the distinction between the division of physiological labor and its social counterpart. Precisely because it is imposed by birth, Durkheim argued, the function of the biological cell is immutably fixed; but in society, hereditary dispositions are not predestinary, and the individual's specialized function is largely self-determined. Durkheim thus denied the view of Comte and Spencer that "substitution" (i.e., one part of an aggregate exchanging function with another) was a characteristic of lower rather than higher evolutionary forms38; on the contrary, in social evolution, function becomes independent of structure in direct proportion to the increasing complexity of society. This in turn explains the origin and development of "civilization"; for as social volume and density increase, men can maintain themselves only through harder work and the intensification of their faculties, which inevitably produces a higher state of culture.

But Durkheim's theory of social evolution was not quite so mechanistic as the account above implies; for, while he urged that civilization was thus the effect of necessary causes, and denied that it was the result of the desire for happiness, he nonetheless argued that it was also "an end, an object of desire, in short, an ideal."39 This paradoxical quality of civilization was based, once again, on Durkheim's distinction between the normal and the pathological. At each stage in the history of a given society, he suggested, there is a "certain intensity" of the collective life which is "normal"; and if everything in the society happens "normally," this state is realized automatically. But, in fact, everything does not happen normally; societies, like individual organisms. are subject to disease, and this prevents them from realizing their natural, ideal condition. Under these circumstances, Durkheim argued, it is not only legitimate but also essential that the sociologist intervene, ascertain the degree of collective activity appropriate to existing conditions, and attempt to realize this ideal state of health (or "golden mean") by the proper means.40 And precisely because the "conditions" here referred to would constantly change, the social ideal would always be definite without ever becoming definitive: "Thus, not only does a mechanistic theory of progress not deprive us of an ideal, but it permits us to believe that we shall never lack for one."41

Finally, these observations led Durkheim to a sociological reformulation of the mind-body problem posed in Descartes' Meditations (1641). The progress of the individual conscience, as we have seen, is in inverse ratio to that of instinct, not because that conscience "breaks up" instinct, but because it "invades" the territory that instinct has ceased to occupy. Instinct, of course, has regressed because of the increasing importance of sociability; thus, the rational superiority of human beings over lower animals is a consequence of their superior sociability. Durkheim thus agreed with the observation of the "spiritualist" philosophers42 that modern "psycho-physiology" would never be able to explain more than a small fraction of psychic phenomena through reference to organic causes; for psychic life, in its highest manifestations, is simply much too free and complex to be understood as a mere extension of physical life. But this is not to say that psychic life cannot be explained by natural causes; for society, no less than organic processes, is a part of nature. There is thus a vast region of the individual conscience which is both unintelligible to "psycho-physiology" and yet perfectly amenable to scientific investigation. Durkheim thus called for a "socio-psychology" which would investigate those psychic facts which have social causes. Far from deriving social facts from the essential features of human nature, such a positive science, pace Spencer, would derive human nature from society.

Abnormal Forms of the Division of Labor

The normal function of the division of labor, as we have seen, is to produce a form of social solidarity; but, like all social (as well as biological) facts, the division of labor may present "pathological" forms which produce different and even contrary results. Durkheim was especially concerned to study these forms for two reasons: first, if it could not be proved that they were deviant and exceptional, the division of labor might be accused of "logically implying" them; and second, the study of such deviant forms might help us better to understand those conditions supportive of the normal state. Eventually Durkheim focused on three types of such pathological forms, not because they exhausted the range of deviant cases, but because they seemed the most general and most serious.

The first type, already identified by Comte43, is found where individuals, increasingly isolated by their more specialized tasks, lose any sense of being integral parts of some larger whole. This reflects a lack of mutual adjustment among the parts of the social organism which Durkheim called the anomic division of labor, citing certain commercial and industrial crises, the conflict between capital and labor, and the "scholastic" specialization of scientific investigation among its examples. And what was particularly alarming, again, was that this form of social disintegration increased with the growth of the division of labor, and thus appeared to be its natural rather than pathological consequence.

How was such a consequence to be avoided? Comte's answer, based on his acceptance of the view that social integration is not a spontaneous product of the division of labor, was that an independent, governmental organ (i.e., the state, as informed by the positive philosophy) was necessary to realize and maintain social unity. Durkheim, by contrast, was extremely skeptical of the efficacy of government regulation of the economy; for the problems afflicting economic institutions arose from a multiplicity of particular circumstances of which only those closest to those problems have any knowledge. And, in any case, he rejected Comte's premise as well; as with all organisms the unity of society was to be obtained by the "spontaneous consensus of parts."44

To overcome the anomic division of labor. therefore, we must first determine the conditions essential to the normal state of organic solidarity. These conditions include not only a system of organs necessary to one another, but also the predetermination of the way in which these mutually necessary organs and their functions are to be related. This predetermination is the critical role of rules of conduct, which are themselves the product of habit and tradition. Very briefly, certain groups of people (organs) engage in definite forms of action (functions) which are repeated because they cling to the constant conditions of social life; when the division of labor brings these different organs and their functions together, the relations thus formed partake of the same degree of fixity and regularity; and these relations, being repeated, become habitual, and, when collective force is added, are transformed into rules of conduct.

The difficulty with the anomic division of labor, of course, is that such rules either do not exist or are not in accord with the degree of development of the division of labor. How can such a situation arise? Typically, something is interposed between otherwise contiguous organs so that the mutual stimulation created by their functions becomes less frequent, less intense, and less determined; the organs lose the sense of mutual dependence that mutual stimulation would normally create, and, as a consequence, the rules reflecting those relations remain vague, ill-defined, and fail to perform their proper integrative function. In commercial and industrial crises, for example, the growth and separation of producers and their markets has proceeded to the extent that the former cannot rationally predict the behavior of the latter; in the conflict between labor and capital, the development of large-scale industry and the factory system has separated the worker both from his family and from his employer; and in the specialization of scientific investigation, the moral and social sciences in particular have not yet understood their relationship to one another and to the older sciences, and have thus ignored the collaborative nature of the work in which they are engaged. But in each case, anomie is the consequence not of the division of labor itself, but of those exceptional and abnormal circumstances under which otherwise contiguous organs become separated, thus preventing the adequate development of rules of conduct.

But it is not sufficient simply that there be rules, for sometimes the rules themselves are the source of the problems. Where the lower classes become dissatisfied with the position granted them by custom or law, for example, we find a strictly regulated form of organization which Durkheim called the forced division of labor, which is nonetheless a potential source of dissension and civil war. The causes of this pathological form are clear. In society, as we have seen, there is a great distance between the hereditary dispositions of the individual and the social function he will fill; and the "space" thus left open to striving and deliberation is also vulnerable to influences which deflect the individual from the role most consistent with his tastes, aptitudes, and capacities. But for the division of labor to produce solidarity, it is not sufficient that each individual have his specialized task; it is still necessary that this task be appropriate to him. The "forced division of labor" is thus the consequence of that structural condition in which the distribution of social functions does not correspond to the distribution of natural talents.

Again, Durkheim insisted that this condition was not a necessary consequence of the division of labor, but rather the product of particular circumstances. "Normally" the division of labor arises spontaneously, and the harmony between individual natures and social functions is the inevitable consequence of each individual's unimpeded pursuit of those tasks for which he is best suited. But here the difficulty arises. For social inequalities thus to express no more than natural inequalities requires a social context in which the latter can be neither increased nor decreased by any external cause; in other words, it requires absolute equality of external conditions, and Durkheim was well aware that no such society had ever existed.

Durkheim was thus in the seemingly awkward position of defining as "normal" a feature which the division of labor had never presented in its pure state. Nonetheless, as always, he was optimistic. Pointing to the progressive decline of the caste system, the increasing accessibility of public office to the average citizen, and the growth of social assistance whereby the disadvantages of birth could b overcome, Durkheim argued not simply that progress toward social justice had been made or that it was a good to be pursued, but that the elimination of external inequalities and realization of the ideal of structural spontaneity was essential -- indeed, indispensable -- to that form of solidarity upon which "organized" societies themselves depend. Social justice would emerge, quite literally, because it had to if advanced societies were to exist at all.

Equality of external conditions was thus necessary if each individual was to find his proper function in society; but it was also necessary if these functions were to be linked to one another. This was particularly evident in contractual relations, which are the juridical expression of those exchanges necessary to the division of labor. Precisely because such exchanges between functions in advanced societies are necessary, contracts must be kept; but unless contractual relations were to remain precarious, they must be kept not just through fear of force, but spontaneously. And it is to fulfill this condition of spontaneity that we say contracts must involve "free consent."

But what does "free consent" mean? In order to answer this question, Durkheim first had to define his notion of the "social value" of an object of exchange. Such a value, Durkheim insisted, is equivalent not ( pace Ricardo) to the labor the object might have cost, but to the amount of energy capable of producing "useful social effects" which the object contains; this, in turn, varies according to the sum of efforts necessary to produce the object, the intensity of the needs which it satisfies, and the extent of the satisfaction it brings. The price of an object deviates from this value, Durkheim argued, only under "abnormal" conditions; thus, the public finds "unjust" every exchange where the price of the object bears no relation to the trouble it cost and the social service it renders. According to Durkheim, therefore, a contract is "freely consented to" only if the services exchanged have an equivalent social value, expressing an equilibrium of wills which is consecrated by a contract; and because this equilibrium is produced and maintained by itself, and expresses the nature of things, it is truly spontaneous.

For the obligatory force of a contract to be complete, therefore, expressed consent alone is not sufficient; the contract must also be just. Social value, however, cannot be determined a priori, but only in the process of exchange itself; thus, for justice to be the rule of contracts, it is necessary, once again, for the entreating parties (labor and management) to be placed in conditions that are externally equal. And here again, Durkheim revealed his evolutionary optimism: the emphasis on "consent" (and especially "free" consent) appears as a very recent development, and contractual law increasingly detracts value from those contracts entered under unequal conditions. If a strong conscience collective was the preemptive need of all lower societies, the requirement and ideal goal of modern societies is social justice.

Durkheim's third pathological form of the division of labor arose from his observation that the functions of an organism can become more active only on the condition that they also become more continuous one organ can do more only if the other organs do more, and vice versa. Where this continuity is lacking, the functional activity of the specialized parts decreases, resulting in wasted effort and loss of productive capacity; but, as always, Durkheim was less concerned with the economic than with the moral consequences of such an abnormal condition. Where the functional activity of the parts languishes, Durkheim thus warned, the solidarity of the whole is undermined.

For precisely this reason, the first concern of intelligent, scientific management will be to suppress useless tasks, to distribute work so that each worker is sufficiently occupied, and thus to maximize the functional activity of each social organ. Increased activity in turn produces greater continuity, an augmented sense of the mutual dependence of the parts on one another, and a stronger bond of solidarity. But where mismanagement prevails, the activity of each worker is reduced, functions become discontinuous, and solidarity is undermined.

But again, Durkheim insisted that such mismanagement and inactivity is the exception rather than the rule, a judgment for which he gave at least four reasons. First, the same factors that cause us to specialize (the increase in social volume and density) also cause us to work harder, for the competition within each speciality increases as the specialities themselves become more numerous and divided. Second, the division of labor itself, by saving time otherwise wasted in passing from one function to another, increases the efficiency of the individual worker. Third, functional activity grows with the talent and competence of the individual worker, and both are naturally increased by the repetition of similar tasks. And finally, as labor becomes divided, work becomes a permanent occupation, then a habit, and ultimately a need -- a progression which increases the functional activity of all workers subject to it.

What, then, is the "first principle" of ethics? And what is the relation of ethics to society? Among the most incontestable of moral rules, Durkheim observed, is that which orders us to internalize the conscience collective of the groups to which we belong; and the "moral" quality of this rule is derived from the essential function it serves in preventing social disintegration. But the contrary rule, which orders us to specialize, is no less imperative; and it too is "moral" because obedience to it, after a certain stage in social evolution, is essential to social cohesion. An initial answer to both questions above, therefore, is that moral rules render "society" possible: "Everything which is a source of solidarity is moral, everything which forces man to take account of other men is moral, everything which forces him to regulate his conduct through something other than the striving of his ego is moral, and morality is as solid as these ties are numerous and strong."45

Durkheim thus opposed the more Kantian tradition which removed moral consciousness from its societal context and defined it through freedom of the will. On the contrary, morality consists in a state of social dependence, and thus deprives the individual of some freedom of movement; and society, far from consisting of external threats to the autonomy of the will, provides the sole foundation upon which that will can act: "Let all social life disappear," Durkheim argued, "and moral life will disappear with it, since it would no longer have any objective."46 Even Kant's "duties of the individual towards himself" are properly understood as duties toward society, for they are the product of collective sentiments which the individual must not offend. The "categorical imperative" of modern society, therefore, is to concentrate and specialize our activities, contract our horizons, choose a definite task, and immerse ourselves in it completely.

The predictable objection to this injunction, of course, was that such specialization implies a narrowing of the individual personality, rendering each of us an "incomplete" human being. But why, Durkheim asked, is it more natural to develop superficially rather than profoundly? Why is there more dignity in being "complete" and mediocre rather than in living a more specialized, but intense, existence? Durkheim, in other words, was re-invoking the Aristotelian principle that man ought to realize his nature as man, though with the added caveat that this nature is not historically constant, but rather varies according to the needs of the societal type in question. Moreover, to be a "person" means to be an autonomous source of action, to possess something empirical and concrete which is ours and ours alone; and this condition, by sharp contrast with the "apparent" liberty and "borrowed" personality of individuals in lower societies, is the product of the division of labor.

While Durkheim thus shared the sense of some contemporaries that theirs was an age of profound crisis, he denied that the crisis was intellectual or spiritual" in its causes. On the contrary, it was the consequence of far-reaching structural changes undergone by society in a very short time; thus, while the morality corresponding to the segmental societal type had regressed, the "new" morality of the organized type had not advanced rapidly enough to fill the void thereby left in our consciences. The corrective for this crisis, therefore, was not to resuscitate the outworn dogmas of the past, but to reduce external inequality and increase justice, and thus to render the new, still discordant organs and functions harmonious. This was an enterprise, Durkheim concluded, in which social structure set the terms, while social theory set the goals:

In short, our first duty is to make a moral code for ourselves. Such a work cannot be improvised in the silence of the study; it can arise only through itself, little by little, under the pressure of internal causes which make it necessary. But the service that thought can and must render is in fixing the goal that we must attain.47

Critical Remarks

The Division of Labor in Society was a seminal contribution to the sociology of law and morality, and remains a sociological "classic" by any standards. By the same standards, however, it also contains undeniable shortcomings which have limited its appeal to modern sociologists. An immediate difficulty, for example, is Durkheim's insistence that social solidarity is an exclusively "moral" phenomenon, of which law is the "externally visible symbol," an insistence which ignores the frequent conflict of some moral principles with others, some laws with other laws, and morality with legality generally. Durkheim, of course, did not deny the existence of such conflict; but he did suggest that it was "pathological," not a part of the "normal" functioning of society, and thus placed it beyond the central focus of his sociological vision. Similarly, Durkheim implied that the state is merely an instrument whose authority reflects the disposition of the conscience collective, an implication which excludes most of the concerns explored so brilliantly by Max Weber -- the means by which one group in a society achieves asymmetrical control over another; the personal, subjective standards by which the first judges the behavior of the second and renders it consequential; and so on. The point here is not simply that Durkheim did not choose to discuss these issues; rather, the point is that he could not, given the reasons why he chose to study law in the first place -- as an external index" of the more fundamental moral conditions of the social order.48

Second, Durkheim clearly overstated the role of repressive law relative to the institutions of interdependence and reciprocity (e.g., kinship, religious ritual, economic and political alliance, etc.) in primitive societies. Malinowski's Argonauts of the Western Pacific (1922), for example, has provided ample evidence of the significance and complexity of relations of exchange among the Trobriand Islanders. In part, this may be attributed to Durkheim's ignorance (or rather dismissal) of the ethnographic literature on primitive peoples, for his pronouncements on "primitive" legal systems in The Division of Labor are largely based on inferences drawn from the Hebrew Torah, the Twelve Tables of the ancient Romans, and the laws of early Christian Europe; but he seems to have got these wrong as well. The religious and moral exhortations of the Torah, for example, are largely devoid of "penal" sanctions, and coexisted with a predominantly secular legal system maintained by their "restitutive" counterparts; the sanctions attached to the Twelve Tables were almost equally restitutive; and the gradual emergence of the state as the preemptive legal institution of early modern Europe witnessed an increase in the relative proportion of repressive laws. Indeed, Durkheim understated the role of repressive law even in advanced industrial societies, in part because he ignored the fact that the nineteenth-century system of penal incarceration replaced the custom of compensating the victims of some crimes financially, and in part because he disregarded the punitive, stigmatizing aspect of many civil laws.49

Finally, it is difficult to share Durkheim's confidence in the self-regulating quality of organic solidarity. Durkheim's account of the "anomic" division of labor alone, for example, exposed all the evils of unregulated capitalism--commercial and industrial crises, class conflict, meaningless, alienated labor, etc.50 But his analysis of these evils was notoriously uncritical; because organic solidarity has evolved more slowly than its mechanical counterpart has passed away, the Third Republic endures a "pathological, disintegrative void" -- an analysis which simultaneously implies that these evils are not endemic to modern societies (and thus eviscerates any criticism of them), and conveniently locates the conditions for the successful functioning of "organized" societies in some unspecified, Utopian future. As his work developed, however, Durkheim gradually relinquished the evolutionary optimism which underlay this mechanical, self-regulating" conception of the division of labor, became increasingly attracted to socialism and the potentially regulatory function of occupational groups51, and granted greater emphasis to the independent role of collective beliefs in social life.


  1. Smith, 1776: 3
  2. Smith, 1776: 13.
  3. 1893: 39.
  4. 1893: 43.
  5. 1893: 68-9.
  6. Durkheim acknowledged that Comte was "the first to have recognized in the division of labor something other than a purely economic phenomenon" (1893: 62).
  7. Durkheim's particular reference here was to Alexander Bain's The Emotions and the Will (1859) and Herbert Spencer's Principles of Psychology (1855).
  8. 1893: 68-9.
  9. 1893: 69.
  10. 1893: 69.
  11. 1893: 73.
  12. Though Durkheim admits that penal law is often "administered" through particular magistrates.
  13. 1893: 79. The French word conscience embraces both the English words "conscience" and "consciousness"; thus it embraces moral and religious beliefs and sentiments on the one hand, and cognitive beliefs and sentiments, on the other. Since translation into either English usage might create confusion, I have henceforth left this term in the original French.
  14. 1893: 81. Durkheim acknowledged the reversion of psychology to Spinoza here, as in "things are good because we like them, as against our liking them because they are good."
  15. 1893: 96.
  16. This introduced one of several disagreements between Durkheim and Gabriel Tarde. In Les Lois de l'imitation (1890). Tarde had suggested that civilization produces social similarities. Durkheim acknowledged the growth of similarities between societies and even between occupational types. but insisted that the individuals within such societies and occupations had. in fact. become progressively differentiated (cf. 1893: 137-8).
  17. Here Durkheim was not above citing phrenological data provided by Gustave LeBon.
  18. 1893: 138.
  19. Durkheim acknowledged that this predominance of repressive law in ancient societies might have alternative explanations, including that presented in Henry Sumner Maine's Ancient Law (1861) -- that reduction of law to the written word took place during a period of violence and barbarism, so that the laws were a reaction against customary behavior. Again, however, Durkheim's view of law, like his view of the state, emphasized that it expresses and symbolizes customs, reacting against them only with a force it has borrowed from them; cf. Durkheim (1893:146).
  20. 1893: 153.
  21. 1893: 154.
  22. 1893: 171.
  23. 1893: 172.
  24. Cf. Lewis Henry Morgan, Ancient Society (1877).
  25. Cf. Lorimer Fison and A.W. Howitt, Kamilaroi and Kurnai (1880)
  26. This led Durkheim into an extended criticism of Spencer, who agreed that social evolution begins from a state of homogeneity, but who also argued that such homogeneity was "inherently unstable." Durkheim, by contrast, emphasized the strong, coherent social life of such groups, based upon an abundance of common beliefs and practices (cf. 1893: 179).
  27. 1893: 190.
  28. 1893: 233.
  29. Durkheim also argued that even these limited pleasures follow their causes by considerable periods of time; thus, those generations inaugurating such advances experience not pleasure but pain, and thus the expectation of pleasure could hardly have been their motive.
  30. Durkheim acknowledged that Comte had already come to much the same conclusion, citing Cours de philosophie positive, IV, 455.
  31. 1893: 267.
  32. Here Durkheim cited the support of Darwin's "law of the divergence of characters" (cf. 1893: 276).
  33. Durkheim's citations are to Spencer, Principles of Sociology, 111, 331; and Comte, Cours de philosophie positive, IV, 421.
  34. This was the source of Durkheim's later, extended disagreement with Lucien Levy-Bruhl.
  35. 1893: 290-1.
  36. 1893: 303.
  37. Durkheim corroborated this argument with the observation that even that degree of physiological conformity required to speak meaningfully of a "race" at all seems rapidly to be disappearing, and added an extended commentary on Sir Francis Galton's Natural inheritance (1889) both to confirm and explain it (cf. 1893: 323-327).
  38. Durkheim here cites Comte, Cours de philosophie positive, Vl, 505; and Spencer, Principles of Sociology, 11, 57.
  39. 1893: 339.
  40. For Durkheim of course, these means included, above all else, moral education, and excluded, without qualification, revolutionary political activity (cf. 1893: 340-341).
  41. 1893: 344. This was Durkheim's answer to Spencer's argument, presented first in Social Statics (1850) and extended in First Principles (1862), that social evolution has a limit, beyond which it cannot pass, in the perfect adaptation of the individual to the natural environment.
  42. In French philosophy, the term "spiritualism" is typically reserved for Maine de Biran 11766 1824), Felix Ravaisson-Mollien (1813-1900), and Henri Bergson (1859-1941).
  43. Durkheim cites Cours de philosophie positive, IV, 429; and Espinas, Les Sociétés animaux (1877).
  44. 1893: 360.
  45. 1893: 398.
  46. 1893: 399.
  47. 1893: 409.
  48. Cf. Lukes and Scull, 1983: 10-15.
  49. Cf. Lukes and Scull, 1983: 115.
  50. Lukes, 1972: 174.
  51. Cf the preface to the second edition of The Division of Labor in Society (1902a).

To return to the Durkheim Home Page, click here.
For additional information, send a message to