Suicide (1897)

[Excerpt from Robert Alun Jones. Emile Durkheim: An Introduction to Four Major Works. Beverly Hills, CA: Sage Publications, Inc., 1986. Pp. 82-114.]

Outline of Topics

  1. What is Suicide?
  2. Extra-social Causes
  3. Social Causes and Social Types
    1. Egoistic Suicide
    2. Altruistic Suicide
    3. Anomic Suicide
  4. Suicide as a Social Phenomenon
  5. Critical Remarks

What is Suicide?

Explanation requires comparison; comparison requires classification; classification requires the definition of those facts to be classified, compared, and ultimately explained. Consistent with The Rules of Sociological Method, therefore, Durkheim began his 1897 work with a warning against notiones vulgares, together with an insistence that

our first task... must be to determine the order of facts to be studied under the name of suicide... we must inquire whether, among the different varieties of death, some have common qualities objective enough to be recognized by all honest observers, specific enough not to be found elsewhere and also sufficiently kin to those commonly called suicides for us to retain the same term without breaking with common usage.1

Durkheim's initial effort at such a definition indeed followed common usage, according to which a "suicide" is any death which is the immediate or eventual result of a positive (e.g., shooting oneself) or negative (e.g., refusing to eat) act accomplished by the victim himself.2 But here Durkheim immediately ran into difficulties, for this definition failed to distinguish between two very different sorts of death: the victim of hallucination who leaps from an upper story window while thinking it on a level with the ground; and the sane individual who does the same thing knowing that it will lead to his death. The obvious solution -- i.e., to restrict the definition of suicide to actions intended to have this result -- was unacceptable to Durkheim for at least two reasons. First, as we have seen (p. 64 above), Durkheim consistently tried to define social facts by easily ascertainable characteristics, and the intentions of agents were ill-fitted to this purpose. Second, the definition of suicide by the end sought by the agent would exclude actions -- e.g., the mother sacrificing herself for her child -- in which death is clearly not "sought" but is nonetheless an inevitable consequence of the act in question, and is thus a "suicide" by any other name.

The distinctive characteristic of suicides, therefore, is not that the act is performed intentionally, but rather that it is performed advisedly -- the agent knows that death will be the result of his act, regardless of whether or not death is his goal. This criterion is sufficient to distinguish suicide, properly so-called, from other deaths which are either inflicted on oneself unconsciously or not self-inflicted at all; moreover. Durkheim insisted that such a characteristic was easily ascertainable, and that such acts thus formed a definite, homogeneous group. Hence Durkheim's definition: Suicide is applied to all cases of death resulting directly or indirectly from a positive or negative act of the victim himself, which he knows will produce this result.

This definition, however, was subject to two immediate objections. The first was that such foreknowledge is a matter of degree, varying considerably from one person or situation to another. At what point, for example, does the death of a professional dare-devil or that of a man neglectful of his health cease to be an "accident" and start to become "suicide"? But for Durkheim to ask this question was less to raise an objection to his definition than to correctly identify its greatest advantage -- that it indicates the place of suicide within moral life as a whole. For suicides, according to Durkheim, do not constitute a wholly distinctive group of "monstrous phenomena" unrelated to other forms of behavior; on the contrary. they are related to other acts, both courageous and imprudent, by an unbroken series of intermediate cases. Suicides, in short, are simply an exaggerated form of common practices.

The second objection was that such practices, however common, are individual practices, with individual causes and consequences, which are thus the proper subject matter of psychology rather than sociology. In fact, Durkheim never denied that suicide could be studied by the methods of psychology, but he did insist that suicide could also be studied independent of its individual manifestations, as a social fact sui generis. Indeed, each society has a "definite aptitude" for suicide, the relative intensity of which can be measured by the proportion of suicides per total population, or what Durkheim called "the rate of mortality through suicide, characteristic of the society under consideration."3 This rate, Durkheim insisted, was both permanent (the rate for any individual society was less variable than that of most other leading demographic data, including the general mortality rate) and variable (the rate for each society was sufficiently peculiar to that society as to be more characteristic of it than its general mortality rate); and, just as the first would be inexplicable were it not "the result of a group of distinct characteristics, solidary with one another, and simultaneously effective in spite of different attendant circumstances," so the second proved "the concrete and individual quality of these same characteristics, since they vary with the individual character of society itself."4 Each society, Durkheim thus concluded, is predisposed to contribute a definite quota of suicides; and it was this predisposition5 which Durkheim proposed to study sociologically.

Thus defined, Durkheim's project again fell naturally into three parts: first, an examination of those extra-social causes sufficiently general to have a possible effect on the social suicide rate (but which in fact influence it little, if at all); second, the determination of the nature of the social causes, the way in which they produce their effects and their relations to those individual conditions normally associated with the different kinds of suicide; and third, the more precise account of the "suicide aptitude" described above, of its relation to other social facts, and of the means by which this collective tendency might be counteracted.

Extra-Social Causes

Durkheim suggested that, a priori, there are two kinds of extra-social causes sufficiently general to have an influence on the suicide rate. First, within the individual psychological constitution there might exist an inclination, normal or pathological, varying from country to country, which directly leads people to commit suicide. Second, the nature of the external physical environment (climate, temperature, etc.) might indirectly have the same effect. Durkheim took up each in turn.

The annual rate of certain diseases, like the suicide rate, is both relatively stable for a given society and perceptibly variable from one society to another; and since insanity is such a disease, the demonstration that suicide is the consequence of insanity (a psychological fact) would successfully account for those features of permanence and variability which had led Durkheim to suggest that suicide was a social fact sui generis. Durkheim was thus particularly concerned to eliminate insanity as a probable cause of suicide, and he did so by attacking that hypothesis in its two most common forms: the view that suicide itself is a special form of insanity, and the view that suicide is simply an effect of various types of insanity. The first Durkheim dismissed by classifying suicidal insanity as a "monomania" -- a form of mental illness limited to a single act or object -- and then arguing that not a single incontestable example of such monomania had yet been shown to exist. The second he rejected on the ground that all suicides committed by the insane are either devoid of deliberation and motive altogether or based on motives that are purely hallucinatory, while many suicides are "doubly identifiable as being deliberate and springing from representations involved in this deliberation which are not purely hallucinatory."6 There are many suicides, therefore, not connected with insanity.

But what about psychopathic conditions which fall short of insanity -- neurasthenia and alcoholism -- but which nonetheless are frequently associated with suicide? Durkheim responded by showing that the social suicide rate bears no definite relation to that of neurasthenia, and that the latter thus has no necessary effect on the former; and alcoholism was discarded as a putative cause on evidence that the geographical distributions of both alcohol consumption and prosecutions for alcoholism bear no relation to that of suicides. A psychopathic state, Durkheim concluded, may predispose individuals to commit suicide, but it is never in itself a sufficient cause of the permanence and variability of suicide rates.

Having dismissed pathological states as a class of causes, Durkheim turned his attention to those normal psychological conditions (race and heredity), which, again, are sufficiently general to account for the phenomena question. The view that suicide is the consequence of tendencies inherent in each major social type, for example, was undermined by the enormous variations in social suicide rates observed within the same type, suggesting that different levels of civilization are much more decisive. But the argument that suicide is hereditary had first to be distinguished from the more moderate view that one inherits a predisposition to commit suicide; for the latter, as in the case of neurasthenia, is not an "explanation" of suicide at all. The stronger argument -- that one inherits a semi-autonomous psychological mechanism which gives rise to suicide automatically -- was then rejected on the grounds that its most dramatic manifestation (the regularity with which suicide sometimes appears in the same family) can be explained by other causes (contagion), and that as within racial types, there are patterned variations within the same family (between husbands and wives) which. on this hypothesis, would be rendered inexplicable.

But if normal or abnormal psychological predispositions are not, by themselves, sufficient causes of suicide, might not such predispositions acting in concert with cosmic factors (climate, seasonal temperature etc.) have such a determinative effect? The conjunction of such predispositions with climate, Durkheim answered, has no such influence; for while the geographical distribution of suicides in Europe varies according to latitude and thus roughly according to climate as well, these variations are better explained by social causes. Montesquieu's suggestion that cold, foggy countries are most favorable to suicide was equally discredited by the fact that, in every country for which statistics were available, the suicide rate is higher in spring and summer than in fall and winter.

Is suicide, then, as the Italian statisticians Ferri and Morselli believed, an effect of the mechanical influence of heat on the cerebral functions? Durkheim here objected on both conceptual and empirical grounds -- that this theory presumes that the constant psychological antecedent of suicide is a state of extreme excitation, where in fact it is frequently preceded by depression; and, in any case, that the suicide rate is in decline in July and August, and thus does not vary regularly with temperature. The "revised" Italian argument -- that it is the contrast between the departing cold and the beginning of the warm season that stimulates the psychological predispositions -- was equally rejected by Durkheim as inconsistent with the perfect continuity (steady increase from January to June, steady decrease from July to December) of the curve representing the monthly variations of the suicide rate.

Consistent with the argument of The Rules (Chapter VI) Durkheim insisted that such a perfectly continuous variation could be explained only by causes themselves varying with the same continuity; and, as a first clue to the nature of these causes, he pointed out that the proportional share of each month in the total number of annual suicides is perfectly parallel with the average length of the day at the same time of the year. Other clues follow: suicide is more common by day than by nights in morning and afternoon than at midday, and on weekdays than on weekends (except for an increase of female suicides on Sundays). In every case, Durkheim observed, suicide increases in those months, days of the week and hours of the day when social life is most active, and decreases when collective activity declines. Anticipating the argument of Book Two, Durkheim thus suggested that suicide is the consequence of the intensity of social life; but before he could proceed to explain how such a cause might produce such an effect, Durkheim had to deal with one other "psychological" theory -- Tarde's argument that social facts in general, and suicide in particular, can be explained as the consequence of imitation.

The term "imitation," Durkheim began, is used indiscriminately to explain three very different groups of facts: (1) that complex process whereby individual states of consciousness act and react upon one another in such a way as to produce a new, collective state sui generis (2) that impulse which leads us to conform to the manners, customs and moral practices of our societies; and (3) that largely unpremeditated, automatic reproduction of actions just because they have occurred in our presence or we have heard of them. The first, Durkheim insisted, can hardly be called "imitation," for it involves no act of genuine reproduction whatsoever;7 the second involves an act of reproduction, but one inspired both by the specific nature of the manners, customs, and practices in question, and by the specific feelings of respect or sympathy they inspire, and thus one ill-described by the term "imitation"; only in the third case, where the act is a mere echo of the original, and subject to no cause outside of itself, is the term warranted. Hence Durkheim's definition: " Imitation exists when the immediate antecedent of an act is the representation of like act, previously performed by someone else; with no explicit or implicit mental operation which bears upon the intrinsic nature of the act reproduced intervening between representation and execution."8

Thus defined, of course. imitation is reduced to a purely psychological phenomenon; for while the synthesis of individual consciousnesses into a collective state sui generis and conformity to obligatory beliefs and practices are both highly social, "imitation properly so-called" is mere repetition, creating no intellectual or moral bond between its agent and his antecedent. We imitate other human beings in the same way that we reproduce the sounds of nature, physical objects, or the movements of non-human animals; and since no clearly social element is involved in the latter, neither is there such an element in the former. To suggest that the suicide rate might be explained by imitation, therefore, was to suggest that a social fact might be explained by a psychological fact -- a possibility Durkheim had already denied in The Rules.

Durkheim's definition9 clearly reduced the number of suicides attributable to imitation. But it did not eliminate them; on the contrary. Durkheim insisted that there was no other phenomenon so "contagious" as suicide. But it did not follow that this contagious quality necessarily had social consequences -- affected the social suicide rate -- for its consequences might instead be merely individual and sporadic: and if imitation did not affect the suicide rate, it was doubtful ( pace Tarde) that it had any social consequences whatsoever, for no phenomenon was more affected by imitation than suicide.10

If, on the other hand, imitation does influence suicide rates, Durkheim suggested, this should be reflected in the geographical distribution of suicides -- the rate typical of one country should be transmitted to its neighbors; and, indeed, contiguous geographical areas do reveal similar suicide rates. But such a geographical "diffusion" of suicides might equally well be explained by the parallel "diffusion" of distinctively social influences throughout the same region. In addition to similarity of rates in geographically contiguous areas, therefore, the "imitation hypothesis" further requires that there be a "model" of particularly intense suicidal activity, and that this activity be "visible" enough to fulfill its function as a model to be imitated. These conditions are in fact fulfilled by the major urban centers in western European countries; thus, we ought to expect the geographical distribution of suicides to reveal a pattern of concentration around major cities, with concentric circles of gradually less intensive suicidal activity radiating out into the countryside. Instead, we find suicide occurring in roughly homogeneous masses over broad regions with no central nuclei, an observation which suggests not only the complete absence of any local influence of imitation, but the presence of the much more general causes of the social environment. Most decisive, however, is the fact that an abrupt change in that social environment is accompanied by an equally abrupt change in the suicide rate. one which is not reflected beyond the bounds of the social environment in question, and thus one which could hardly be explained as the consequence of imitation.

But Durkheim's argument in fact went much further than this denial that, its individual effects notwithstanding, imitation is an insufficient cause for variations in the suicide rate; for, in addition, he insisted that imitation alone has no effect on suicide whatsoever. This extension of his argument was the consequence of Durkheim's more general theoretical commitment to the view that the thought of an act is never sufficient to produce the act itself unless the person thinking is already so disposed; and the dispositions in question, of course, are the result of social causes. Imitation, therefore, is not a real cause, even of individual suicides: "It only exposes a state which is the true generating cause of the act," Durkheim concluded, "and which probably would have produced its natural effect even had imitation not intervened, for the predisposition must be very strong to enable so slight a matter to translate it into action."11

Social Causes and Social Types

Durkheim's argument so far is a perfect example of his characteristic "argument by elimination" -- the systematic rejection of alternative explanations of a given phenomenon in order to lend authority to the "sole remaining" candidate.12 He thus claimed to have shown that, for each social group, there is a specific tendency to suicide that can be explained neither by the "organic-psychic constitution" of individuals nor by the nature of the physical environment; and as his discussion of geographic and seasonal variations of suicide has already hinted, the tendency in question must thus be, in itself, a collective phenomenon, and must depend upon social causes.

But is there, in fact, one "single, indestructible" suicidal tendency? Or are there rather several, which should be distinguished from one another and then studied separately? Durkheim had already pondered this difficulty in Book One, in his discussion of suicide by insanity, and his solution there was repeated here. Briefly, the suicidal tendency, single or not, is observable only in its individual manifestations (individual suicides); thus, Durkheim proposed to classify suicides into distinct "types" or "species" according to their similarities and differences, on the assumption that there would be as many types as there were suicides having the same essential characteristics, and as many "tendencies" as there were types.

This solution, however, immediately raised another problem. In his treatment of suicides by insanity, Durkheim had at his disposal many good descriptions of individual cases -- of the agent's psychological state prior to the act, of his preparations to commit the act, of the manner in which the act was performed, etc. But such data were almost completely unavailable for suicides committed by sane people, a fact which rendered classification by external manifestations impossible. Durkheim was thus forced to alter his strategy -- indeed, to "reverse the order of study" altogether, adopting an "etiological" rather than "morphological" system of classification. Assuming, as always, that any given effect has one, and only one corresponding cause, Durkheim argued that there must be as many special types of suicide as there are special causes producing them: "Without asking why [these types of suicide] differ from one another," Durkheim proposed, "we will first seek the social conditions responsible for them; [we will] then group these conditions in a number of separate classes by their resemblances and differences, and we shall be sure that a specific type of suicide will correspond to each of these classes."13

How, then, do we determine the causes of suicide? One answer was simply to rely on statistical records of the "presumptive motive of suicide" (apparently construed as a cause) kept by officials in most modern societies; but, despite its obvious convenience and plausibility, Durkheim rejected this resource for at least two reasons. First such "statistics of the motives of suicides" were actually statistics of officials opinions of such motives, which thus embodied not only difficult assessments of material fact, but still more difficult explanations and evaluations of actions performed at will. Second, regardless of the credibility of such reports, Durkheim simply denied that motives were true causes, a characteristic position he supported by pointing to the contrast between relatively constant proportions of different classes of "motive explanations" (both over time and across occupational groups) and extremely variable suicide rates themselves (over the same time period and across the same occupational groups). These "reasons" to which suicides are ascribed, Durkheim thus insisted, are only apparent causes, individual repercussions of more general states which they only imperfectly express: "They may be said to indicate the individuals weak points, where the outside current bearing the impulse to self-destruction most easily finds introduction. But they are no part of this current itself, and consequently cannot help us to understand it."14 Disregarding such individual repercussions, therefore, Durkheim turned directly to the "states of the various social environments" (religious confessions, familial and political society, occupational groups) across which the variations in suicide rates occur, and within which their causes might be found.

Egoistic Suicide

Durkheim first asked the different religious confessions affect suicide. If we look at a map of Western Europe, for example, we see that where Protestants are most numerous the suicide rate is highest, that where Catholics predominate it is much lower, and that the aptitude of Jews for suicide is lower still, though to a lesser degree, than that of Catholics. How are these data to be explained?

Again, Durkheim escorted the render through an argument by elimination. In many of the societies under observation, for example, Jews and Catholics are less numerous than Protestants; thus it is tempting to explain their lower suicide rates as the consequence of that rigorous moral discipline which religious minorities sometimes impose upon themselves in the face of the hostility of surrounding populations. But such an explanation, Durkheim observed, ignores at least three facts: first, suicide is too little an object of public condemnation for religious hostility to have this effect: second, religious hostility frequently produces not the moral conformity of those against whom it is directed, but rather their rebellion against it; and third, the reduced suicide rate of Catholics relative to Protestants is independent of their minority status -- even in Spain. Catholics commit suicide less frequently.

The last point in particular suggested an alternative explanation -- that the cause for lower rates of suicide is to be found within the nature of the religious confession itself. But such an explanation, Durkheim insisted, cannot refer to the religious percepts of the confession. for there Catholics and Protestants prohibit suicide with equal emphasis; rather, the explanation must proceed from one of the more general characteristics differentiating them, and that characteristic -- indeed, "the only essential difference between Catholicism and Protestantism" -- is that the latter permits free inquiry to a greater degree than the former.15

But if the proclivity of Protestantism for suicide must thus be related to its spirit of free inquiry, this "free inquiry" itself requires explanation, for it brings as much sorrow as happiness, and thus is not "intrinsically desirable." Why, then, do men seek and even demand such freedom? Durkheim's answer: "Reflection develops only if its development becomes imperative, that is, if certain ideas and instinctive sentiments which have hitherto adequately guided conduct are found to have lost their efficacy. Then reflection intervenes to fill the gap that has appeared, but which it has not created."16 In other words, Protestantism concedes greater freedom of thought to the individual because it has fewer commonly accepted beliefs and practices. Indeed, it was this possession of a common, collective credo that, for Durkheim, was the essence of religious society itself, and that distinguished it from those merely temporal bonds which unite men through the exchange and reciprocity of services, yet permit and even presuppose differences; and, precisely to the extent that Protestantism lacked such a credo, it was a less strongly integrated church than its Roman Catholic counterpart.

Durkheim then suggested that this explanation is consistent with at least three other observations. First, it would account for the still lower suicide rates of Jews who, in response to the hostility17 directed against them, established strong community ties of thought and action, virtually eliminated individual divergences, and thus achieved a high degree of unity, solidarity, and integration. Second, of all the great Protestant countries England has the lowest suicide rate; and it also has the most "integrated" of Protestant churches. And third since knowledge is the natural consequence of free inquiry, we should expect that suicide increases with its acquisition, and Durkheim had little trouble demonstrating that this was the case.

But this last demonstration did raise an anomaly: the Jews, who are both highly educated and have low suicide rates. But for Durkheim, this was the proverbial exception that proves the rule. For the Jew seeks to learn, not in order to replace traditional beliefs with individual reflection, but rather to protect himself from others' hostility through his superior knowledge. "So the exception," Durkheim observed "is only apparent";

it even confirms the law. Indeed, it proves that if the suicidal tendency is great in educated circles, this is due, as we have said, to the weakening of traditional beliefs and to the state of moral individualism resulting from this; for it disappears when education has another cause and responds to other needs.18

Finally, it should be noted that the combined effect of these observations on religious confessions and suicide was an implicit celebration of the Third Republic in general and its program of secular education in particular. For, as Durkheim was pleased to make clear, the long-acknowledged correlation between the growth of knowledge and suicide could not be taken to mean that the former "causes" the latter; on the contrary, knowledge and suicide are independent effects of a more general cause -- the decline of traditional beliefs. Moreover, once these beliefs have declined, they cannot be artificially reestablished,19 and thus free inquiry and the knowledge that results become our only resources in the effort to replace them. Finally, Durkheim had shown that the prophylactic effect of religion on suicide owed little to its condemnation of suicide, its idea of God, or its promise of a future life; rather, religion protects man from suicide "because it is a society. What constitutes this society is the existence of a certain number of beliefs and practices common to all the faithful, traditional and thus obligatory. The more numerous and strong these collective states of mind are," Durkheim concluded, "the stronger the integration of the religious community, and also the greater its preservative value."20

But if religion thus preserves men from suicide because it is a society, other "societies" (e.g., the family and political society) ought to have the same effect. After developing a statistical measure of the immunity to suicide enjoyed by various groups,21 for example, Durkheim was able to show that, while marriage alone has a preservative effect against suicide, this is limited and benefits only men; the larger family unit, on the other hand, provides an immunity which husband and wife share. Similarly, when one marital partner dies, the survivor loses a degree of suicidal immunity; but this loss is less a consequence of the severing of the conjugal bond alone than of the more general shock to the family that the survivor must endure. Finally, the immunity to suicide increases with the size of the family,22 a fact Durkheim attributed to the greater number and intensity of collective sentiments produced and repeatedly reinforced by the larger group.

Similarly, the examination of political societies showed that suicide, quite rare in a society's early stages, increases as that society matures and disintegrates. During social disturbances or great popular wars, by contrast, the suicide rate declines, a fact that Durkheim claimed is susceptible of only one interpretation -- that these disturbances "rouse collective sentiments, stimulate partisan spirit and patriotism, political and national faith alike, and concentrating activity toward a single end, at least temporarily cause a stronger integration of society."23

Suicide thus varies inversely with the degree of integration of the religious, domestic, and political groups of which the individual forms a part; in short, as a society weakens or "disintegrates," the individual depends less on the group, depends more upon himself, and recognizes no rules of conduct beyond those based upon private interests. Durkheim called this state of "excessive individualism" egoism, and the special type of self-inflicted death it produces egoistic suicide.

But why does individualism thus cause suicide? The traditional view -- that man, by his psychological nature, cannot live without some transcendent, eternal reason beyond this life -- was rejected on the ground that, were our desire for immortality so great, nothing in this life could satisfy us; whereas, in fact, we do take pleasure in our temporal lives, and the pleasures we take are not merely physical and individual, but also moral and social, bath in their origin and in their purpose. Durkheim thus returned to the conception of the duality of human nature first found in The Division of Labor:

... social man superimposes himself upon physical man. Social man necessarily presupposes a society which he expresses or serves. If this dissolves, if we can no longer feel it in existence and action about and above us. whatever is social in us is deprived of all objective foundation... Thus we are bereft of reasons for existence: for the only life to which we could cling no longer corresponds to anything actual; the only existence still based upon reality no longer meets our needs... So there is nothing more for our efforts to lay hold of, and we feel them lose themselves in emptiness.24

It is in this social (rather than the earlier. psychological) sense therefore that our activity needs an object transcending it; for such an object is implicit within our moral constitution itself, and cannot be lost without this constitution losing its raison d'être to the same degree. In this state of moral confusion, the least cause of discouragement gives birth to desperate self-destructive resolutions a suicidal tendency that infects not only individuals but whole societies; and, precisely because these intellectual currents are collective they impose their authority on the individual and drive him even further in the direction he is already by internal disposition inclined to go. Ironically therefore, the individual submits to the influence of society at the very moment that he frees himself from it: "However individualized a man may be, there is always something collective remaining -- the very depression and melancholy resulting from this same exaggerated individualism."25

Altruistic Suicide

But if excessive individuation thus leads to suicide, so does insufficient individuation: thus, among primitive peoples, we find several categories of suicide -- men on the threshold of old age, women upon the deaths of their husbands, followers and servants upon the deaths of their chiefs -- in which the person kills himself because it is his duty. Such a sacrifice, Durkheim argued, is imposed by society for social purposes; and for society to be able to do this, the individual personality must have little value, a state Durkheim called altruism, and whose corresponding mode of self-inflicted death was called obligatory altruistic suicide.26

Like all suicides, the altruist kills himself because he is unhappy;27 but this unhappiness is distinctive both in its causes and in its effects. While the egoist is unhappy because he sees nothing "real" in the world besides the individual, for example, the altruist is sad because the individual seems so "unreal"; the egoist sees no goal to which he might commit himself, and thus feels useless and without purpose while the altruist commits himself to a goal beyond this world, and henceforth this world is an obstacle and burden to him. The melancholy of the egoist is one of incurable weariness and sad depression, and is expressed in a complete relaxation of all activity the unhappiness of the altruist, by contrast, springs from hope, faith even enthusiasm, and affirms itself in acts of extraordinary energy.

Altruistic suicide thus reflects that crude morality which disregards the individual, while its egoistic counterpart elevates the human personality beyond collective constraints; and their differences thus correspond to those between primitive and advanced societies. But altruistic suicides do occur among more civilized peoples -- among the early Christian martyrs and the French revolutionaries -- and in contemporary French society, Durkheim insisted, there even exists a "special environment" in which altruistic suicide is chronic: the army Military suicide thus represents an evolutionary survival of the morality of primitive peoples: "Influenced by this predisposition," Durkheim observed, "the soldier kills himself at the least disappointment, for the most futile reasons, for a refusal of leave, a reprimand an unjust punishment, a delay in promotion, a question of honor, a flush of momentary jealousy, or even simply because other suicides have occurred before his eyes or to his knowledge."28 The "contagious" suicides ascribed by Tarde to psychological causes Durkheim thus insisted, are rather explained by the moral constitution which predisposes men to imitate one another's actions.

Finally, Durkheim's discussion of altruistic suicide aptly illustrates some of the most characteristic arguments of the work as a whole -- his rejection of any definition of suicide which appeals to subjective mental states (motives purposes, etc.), his suggestion that self-inflicted deaths reflect the most general moral sentiments of the societies within which they occur, and the view that such suicides are thus merely exaggerated expressions of behavior which, in more moderated form, would be labeled "virtuous." However pure the motives which led to the "heroic" suicide of Cato, for example, it was not different in kind from that of one of Frazer's primitive Polynesian chiefs; and where altruistic suicides reflect a courageous indifference to the loss of one's life (albeit to the loss of others' lives as well), its egoistic counterpart exhibits a praiseworthy respect and sympathy for the suffering of others (albeit a concern to avoid one's own suffering and sacrifices as well).

Anomic Suicide

Egoistic and altruistic suicide, as we have seen, are the respective consequences of the individual's insufficient or excessive integration within the society to which he belongs. But quite aside from integrating its members, a society must control and regulate their beliefs and behavior as well; and Durkheim insisted that there is a relation between a society's suicide rate and the way it performs this important regulative function. Industrial and financial crises, for example, increase the suicide rate, a fact commonly attributed to the decline of economic well-being these crises produce. But the same increase in the suicide rate, Durkheim observed, is produced by crisis resulting in economic prosperity; "Every disturbance of equilibrium," he insisted, "even though it achieved greater comfort and a heightening of general vitality, is an impulse to voluntary death."29 But how can this be the case? How can something generally understood to improve a man's life serve to detach him from it?

No living being, Durkheim began, can be happy unless its needs are sufficiently proportioned to its means; for if its needs surpass its capacity to satisfy them, the result can only be friction, pain, lack of productivity, and a general weakening of the impulse to live. In an animal, of course, the desired equilibrium between needs and means is established and maintained by physical nature -- the animal cannot imagine ends other than those implicit within its own physiology, and these are ordinarily satisfied by its purely material environment. Human needs, however, are not limited to the body alone; indeed, "beyond the indispensable minimum which satisfies nature when instinctive, a more awakened reflection suggests better conditions, seemingly desirable ends craving fulfillment."30 But the aspirations suggested by such reflections are inherently unlimited there is nothing in man's individual psychology or physiology which would require them to cease at one point rather than another. Unlimited desires are, by definition, insatiable, and insatiability is a sure source of human misery: "To pursue a goal which is by definition unattainable," Durkheim concluded, "is to condemn oneself to a state of perpetual unhappiness."31

For human beings to be happy, therefore, their individual needs and aspirations must be constrained; and since these needs and aspirations are the products of a reflective social consciousness, the purely internal, physiological constraints enjoyed by animals are insufficient to this purpose. This regulatory function must thus be performed by an external, moral agency superior to the individual -- in other words, by society. And since the constraints thus applied are borne unequally by a society's members, the result is a "functional" theory of stratification resembling that of Kingsley Davis and Wilbert Moore32 -- society determines the respective value of different social services, the relative reward allocated to each, and the consequent degree of comfort appropriate to the average worker in each occupation.

This classically conservative doctrine is tempered by two qualifications. First, the scale of services and rewards is not immutable, but rather varies with the amount of collective revenue and the changing moral ideas of the society itself; and second, the system must secure some degree of legitimacy -- both the hierarchy of functions and the distribution of these functions among the population must be considered "just" by those subject to it. These caveats entered, however, Durkheim insisted that human happiness can be achieved only through the acceptance of moral (that is, social) constraints.

But what has this to do with suicide? Briefly, when society is disturbed by some crisis, its "scale" is altered and its members are "reclassified"33 accordingly; in the ensuing period of dis-equilibrium, society is temporarily incapable of exercising its regulative function, and the lack of constraints imposed on human aspirations makes happiness impossible. This explains why periods of economic disaster, like those of sudden prosperity, are accompanied by an increase the number of suicides, and also why countries long immersed in poverty have enjoyed a relative immunity to self-inflicted death.

Durkheim used the term anomie to describe this temporary condition of social deregulation, and anomic suicide to describe the resulting type of self-inflicted death; but in one sphere of life, he added, anomie is not a temporary disruption but rather a chrome state. This is the sphere of trade and industry, where the traditional sources of societal regulation -- religion, government, and occupational groups -- have all failed to exercise moral constraints on an increasingly unregulated capitalist economy. Religion, which once consoled the poor and at least partially restricted the material ambitions of the rich, has simply lost most of its power. Government, which once restrained and subordinated economic functions, is now their servant, thus, the orthodox economist would reduce government to a guarantor of individual contracts, while the extreme socialist would make it the "collective bookkeeper" -- and neither would grant it the power to subordinate other social agencies and unite them toward one common aim. Even occupational groups, which once regulated salaries, fixed the price of products and production, and indirectly fixed the average level of income on which needs were based, has been made impotent by the growth of industry and the indefinite expansion of the market. In trade and industry, therefore, "the state of crisis and anomy is constant and, so to speak, normal. From top to bottom of the ladder greed is aroused without knowing where to find ultimate foothold. Nothing can calm it," Durkheim concludes, "since its goal is far beyond all it can attain."34 And thus the industrial and commercial occupations are among those which furnish the greatest numbers of suicides.

Quite aside from such economic anomie, however, is that domestic anomie which afflicts widows and widowers as well as those who have experienced separation and divorce.35 The association of the latter with an increased tendency to suicide had already been observed,36 but had been attributed to marital selection -- divorced couples are more apt to have been recruited from individuals with psychological flaws, who are also more apt to commit suicide. Characteristically, Durkheim rejected such individual, psychological "explanations" for both suicide and divorce arguing instead that we should focus on the intrinsic nature of marriage and divorce themselves.

Marriage, Durkheim explained, ought to be understood as the social regulation not only of physical instinct, but also of those aesthetic and moral feelings which have become complicated with sexual desire over the course of evolution. Precisely because these new aesthetic and moral inclinations have become increasingly independent of organic necessities, the moral regulation of monogamic marriage has become necessary: "For by forcing a man to attach himself forever to the same woman," Durkheim observed, "it assigns a strictly definite object to the need for love, and closes the horizon."37 Divorce would then be understood as a weakening of this matrimonial regulation, and wherever law and custom permit its "excessive" practices the relative immunity to self-inflicted death thus guaranteed is undermined, and suicides increase.

As we have already seen, however, the immunity guaranteed by marriage alone is enjoyed only by the husband, both partners participating only in the immunity provided by the larger domestic society; similarly, it is husbands rather than wives who are afflicted with increased suicide rates where divorces are "excessive." Why don't divorce rates affect the wife? Durkheim's quintessentially Victorian answer was that the mental life of women -- and thus the "mental character" of their sexual needs -- is less developed than that of men; and since their sexual needs are thus more closely related to those of their organism, these needs find an efficient restraint in physiology alone, without the additional, external regulation of that monogamic matrimony required by males. This was an observation however, from which Durkheim derived an un-Victorian inference: since monogamic matrimony provides no suicidal immunity to the wife, it is a gratuitous form of social discipline which she suffers without the slightest compensatory advantage. The traditional view of marriage -- that its purpose is to protect the woman from masculine caprice, and to impose a sacrifice of polygamous instincts upon the man -- is thus clearly false; on the contrary, it is the woman who makes the sacrifices, receiving little or nothing in return.38

To this "etiological" classification of suicides by their causes, Durkheim added a "morphological" classification according to their characteristic effects or manifestations. Suicides like that of Lamartine's Raphael, for example, committed out of a morbid mood of melancholia -- were considered the consequence and expression of egoistic suicide, as were the more cheerfully indifferent "Epicurean" suicides of those who, no longer able to experience the pleasures of life, see no reason to prolong it. Altruistic suicide, as we have already seen, is characterized by the serene conviction that one is performing one's duty, or a passionate outburst of faith and enthusiasm; while anomic suicide, though equally passionate, expresses a mood of anger and disappointment at aspirations unfulfilled.

Just as there are different types of suicide distinguishable by their causes, therefore, there are different species of moods or dispositions through which these types are expressed. In actual experience, however, these types and species are not found in their pure, isolated state; on the contrary, different causes may simultaneously afflict the same individuals, giving rise to composite modes of suicidal expression. Egoism and anomie, for example, have a special "affinity" for one another -- the socially detached egoist is often unregulated as well (though usually introverted, dispassionate, and lacking in those aspirations which lead to frustration), while the unregulated victim of anomie is frequently a poorly integrated egoist (though his boundless aspirations typically prevent any excessive introversion). Similarly, anomie may be con joined with altruism -- the exasperated infatuation produced by anomie may coincide with the courageous, dutiful resolution of the altruist. Even egoism and altruism, contraries though they are, may combine in certain situations -- within a society undergoing disintegration, groups of individuals may construct some ideal out of whole cloth, devoting themselves to it to precisely the extent that they become detached from all else.

Finally, Durkheim found no relation whatsoever between the type of suicide and the nature of the suicidal acts by which death is achieved. Admittedly, there is a correlation between particular societies and the popularity of certain suicidal acts within them, indicating that the choice of suicidal means is determined by social causes. But the causes which lend one to commit suicide in a particular ways Durkheim insisted are quite different from those which lead one to commit suicide in the first place; the customs and traditions of a particular society place some instruments of death rather than others at one's disposal, and attach differing degrees of dignity even to the various means thus made available. While both are dependent on social causes, therefore, the mode of suicidal act and the nature of suicide itself are unrelated.

Suicide as a Social Phenomenon

At any given moment, therefore, the moral constitution of a society -- its insufficient or excessive degree of integration or regulation -- establishes its contingent rate of voluntary deaths, its "natural aptitude" for suicide; and individual suicidal acts are thus mere extensions and expressions of these underlying currents of egoism, altruism, and anomie. Moreover, the terms that Durkheim employed in making this argument -- "collective tendencies," "collective passions," etc. -- were not mere metaphors for average individual states; on the contrary, they are "things," sui generis forces which dominate the consciousnesses of individuals. In fact, the stability of the suicide rate for any particular society could have no other explanation:

... the numerical equality of annual contingents... can only be due to the permanent action of some impersonal cause which transcends all individual cases.... The proof that the reality of collective tendencies is no less than that of cosmic forces is that this reality is demonstrated in the same way, by the uniformity of effects.39

Such an argument, Durkheim admitted suggests that collective thoughts are of a different nature from individual thoughts, that the former have characteristics which the latter lack. But how can this be if there are only individuals in society? Durkheim's response was an argument by analogy alluded to in The Division of Labor40 and developed more fully in "Individual and Collective Representations" (1898). The biological cell, Durkheim observed, is made up exclusively of inanimate atoms; but surely this doesn't mean that there is "nothing more" in animate nature. Similarly individual human beings, by associating with one another, form a psychical existence of a new species, which has its own manner of thinking and feeling: "When the consciousness of individuals, instead of remaining isolated, becomes grouped and combined," Durkheim observed, "something in the world has been altered. Naturally this change produces others, this novelty engenders other novelties, phenomena appear whose characteristic qualities are not found in the elements composing them."41 Social life, Durkheim thus admitted, is essentially made up of representations; but collective representations are quite different from their individual counterparts. Indeed, Durkheim had no objection to calling sociology a kind of psychology, so long as we recall that social psychology has its own laws which are not those of individual psychology.

Moreover, it is simply not true that there are "only individuals" in society. First, a society contains a variety of material things (e.g., written laws, moral precepts and maxims, etc.) which "crystallize" social facts, and act upon the individual from without; and second, beneath these immobilized, sacrosanct forms are the diffused, mingling subjacent sentiments of which these material formulae are the mere signs, and which are equally external to the individual conscience. The result was a critique of Quetelet reminiscent of Kant's rejection of any empiricist ethics. Struck by the statistical regularity of certain social phenomena over time, Quetelet had postulated "the average man" -- a definite type representing the most generalized characteristics of people in any given society. Such an approach, Durkheim insisted, makes the origin of morality an insoluble mystery; for it conflates the collective type of a society with the average type of its individual members, and since the morality of such individuals reaches only a moderate intensity, the imperative, transcendent character of moral commands is left without an explanation. Beyond the vacuous conception of "God's will," Durkheim insisted, "no alternative exists but to leave morality hanging unexplained in the air or make it a system of collective states of conscience. Morality either springs from nothing given in the world of experience, or it springs from society."42

In fact, these three currents of opinion -- that the individual has a certain personality (egoism), that this personality should be sacrificed if the community required it (altruism), and that the individual is sensitive to ideas of social progress (anomie) -- coexist in all societies, turning individual inclinations in three different and opposed directions. Where these currents offset one another, the individual enjoys a state of equilibrium which protects him from suicide; but where one current exceeds a certain strength relative to others, it becomes a cause of self-inflicted death. Moreover, this strength itself depends on three causes: the nature of the individuals composing the society, the manner of their association, and transitory occurrences which disrupt collective life. The first, of course, is virtually immutable, changing only gradually over a period of centuries; the only variable conditions, therefore, are social conditions, a fact which explains the stability observed by Quetelet so long as society remains unchanged.

The decisive influence of these currents, however, is rarely exerted throughout an entire society; on the contrary, its effect is typically felt within those particular environments whose conditions are especially favorable to the development of one current or another. But the conditions of each individual environment are themselves dependent on the more general conditions of the society as a whole -- the force of altruism in the army depends on the role of the military in the larger civilian population; egoistic suicide increases among Protestants to the extent that intellectual individualism is a feature of the entire society; and so on. No collective sentiment can affect individuals, of course, when they are absolutely indisposed to it, but the same social causes that produce these currents also affect the way individuals are socialized, so that a society quite literally produces citizens with the appropriate dispositions at the same time that it molds the currents to which they will thus respond. Durkheim did not deny, therefore, that individual motives have a share in determining who commits suicide but he did insist that the nature and intensity of the "suicidogenic" current were factors independent of such psychological conditions. Indeed, this was why Durkheim could claim that his theory, however "deterministic," was more consistent with the philosophical doctrine of free will than any psychologistic theory which makes the individual the source of social phenomena; for the intensity of his currents, like the virulence of an infectious disease, determines only the rate at which the population will be affected, not the identity of those to be struck down.

The last remark hinted at what we have seen to be one of Durkheim's preoccupations -- his repeated efforts to resolve philosophical quandaries by sociological means; and he soon turned to another: Should suicide be proscribed by morality?43 This question, Durkheim observed, is typically dealt with by formulating certain general moral principles and then asking whether suicide logically contradicts these or not. But Durkheim insisted instead on an empirical sociological approach, examining the way in which real societies have actually treated suicide in the course of history, and then inquiring into the reasons for this treatment. This examination indicated that suicide has been long, widely, and severely condemned, but that such condemnations fell into two categories, indicating two historical stages. In antiquity, suicide was a civil offense, and though the individual was forbidden to end his own life, the state might permit him to do so on certain occasions. But in modern societies, suicide is viewed as a religious crime, and the condemnation is thus both absolute and universal. The distinctive element in the second stage, Durkheim insists, is the Christian conception of the human personality as a "sacred" thing; henceforth, in so far as he retains his identity as a man, the individual shares that quality sui generis which religions ascribe to their gods: "He has become tinged with religious value; man has become a god for men. Therefore any attempt against his life suggests sacrilege."44

To Montesquieu or to Hume, such an argument, based on a religious premise, was less than compelling. But to Durkheim, agnostic though he was, the religious vestments of the argument were purely symbolic and did little to discredit it; on the contrary, for Durkheim, every symbol (however mystical) must correspond to something real, and the reality to which the "sacred individual" corresponds is that body of collective sentiments which, with the growth of social volume and density, the division of labor, and individual differences, has elevated the individual personality above that primitive, homogeneous community within which it was literally non-existent.45 This view that the human person is in some sense sacred, Durkheim insisted, is virtually the only common bond joining a modern society's members; far from injuring only himself, therefore the man who commits suicide violates the most fundamental maxim of the social orders a transgression which is reflected in and in turn justifies, its severe moral prohibition.

Such an appeal to the sacredness of individual life necessarily raised the question of the relation between suicide and homicide; and this in turn led Durkheim to another attack on the "Italian school" of Ferri and Morselli, for whom such acts were the result of the same psychological cause (moral degeneracy) under different social conditions (suicide is simply a homicide which, repressed by a pacific social environment, is directed back toward the self). Durkheim denied of course that the causes of suicide and homicide are either "psychological" or "the same," and also that the social conditions under which they occur are so consistently different; for, as we have seen there are different kinds of suicide with different, non-psychological causes, and while some of these are identical to those of homicide others are quite opposed to them. Egoistic suicide, for example results from conditions of disintegration and social indifference which, by reducing the intensity of the passions and increasing the respect for the individual, decreases the tendency to homicide. Altruistic suicide, by contrast, springs from a reduced respect for the individual life, as does homicide; but these are the social conditions of primitive rather than civilized societies. Anomie suicide, however is produced by that more modern mood of exasperation and world-weariness which is equally conducive to homicide; and which kind of death will result is largely determined by the moral constitution of the individual in question. If suicide and homicide vary inversely, Durkheim thus concluded it is not ( pace Ferri and Morselli) because they are differing social expressions of the same psychological phenomenon; on the contrary, it is because most modern suicides result from conditions of egoism which are hostile to homicides. And if the relationship between suicide and homicide is not perfectly inverse, it is because the special social conditions which favor either anomie or altruistic suicide are also favorable to homicide.46

Here was another sociological answer to a venerable philosophical question -- i.e., whether our feelings for others are mere extensions of our feelings for ourselves or, by contrast, are independent of such selfish sentiments altogether. Durkheim's answer was that both alternatives are misconceived. Feelings for others and feelings for ourselves are not unrelated, but neither does one spring from the other; on the contrary, both are derived from a third source: that estimate of the moral value of the individual rendered by the conscience collective at any point in time. Where that estimate is low, as in primitive societies, our indifference to the pain and sadness of others, for example, is matched by our indifference to our own; but where that estimate is high, as in advanced societies, our concern for our own comfort is balanced by a concern for that of others. Our egoistic instincts, of course, will weaken feelings when applied to the first, and strengthen them in application to the second; but the same moral condition exists and is active in both cases.

Like The Division of Labor in Society, Suicide concludes with some thoroughly practical questions: What attitude should modern societies take towards suicide? Should reforms be undertaken to restrain it? Or must we accept it as it is? Again as in The Division of Labor, Durkheim's answers to these questions depended on whether the current state of suicide is to be considered "normal" or "abnormal," and, as he had already shown through the example of crime in The Rules of Sociological Method, the "immorality" of suicide did not necessarily point to the latter. On the contrary, the statistical data going back to the eighteenth century, as well as legislation surviving from still earlier periods, suggested to Durkheim that suicide was a normal element in the constitution of all societies. In primitive societies and the modern military, for example, the strict subordination of the individual to the group renders altruistic suicide an indispensable part of collective discipline. Again, in societies where the dignity of the person is the supreme end of conduct, egoistic suicide flourishes. And again, in societies where economic progress is rapid and social restraints become slack, anomie suicides are inevitable.

But don't such currents of altruism, egoism, and anomie cause suicide only if excessive? And might such currents not be everywhere maintained at the same level of moderate intensity? Durkheim's initial response echoed his discussion of crime in The Rules -- there are special environments within each society which can be reached by such currents only if the latter are strengthened or weakened far above or below the more general societal norm. But again, as with crime, these special modifications of the current are not merely necessary; they are also useful, for the most general collective state is simply that best adapted to the most general circumstances, not to those exceptional circumstances to which a society must also be adapted. A society in which intelligent individualism could not be exaggerated, for example, would be incapable of radical innovation, even if such innovation were necessary; inversely, a society in which such individualism could not be significantly reduced would be unable to adapt to the conditions of war, in which conformity and passive obedience are elevated into virtues. It is essential, therefore that such "special environments" be preserved as a part of the more general existence, so that a society might both respond to particular conditions and evolve gradually over time.47

Thus the spirit of renunciation, the taste for individuation, and the love of progress each have their place in every society, and cannot exist without generating suicide. But this does not mean that every suicido-genic current is "normal"; on the contrary, these currents must produce suicides only in a certain measure which varies from one society to another as well as over time. Here Durkheim was particularly concerned to dismiss the view that suicide, the rate of which had increased exponentially in western Europe since the eighteenth century, was the "ransom money" of civilization, the inevitable companion of social progress. The rash of suicides which accompanied the growth of the Roman Empire, Durkheim admitted, might support such a view; but from the height of Rome to the Enlightenment, suicide rates increased only slightly, while Roman culture was assimilated and then surpassed by Christianity, the Renaissance, and the Reformation. Social progress, therefore, does not logically imply suicide, and the undeniably rapid growth of suicide in the late nineteenth century should be attributed not to the intrinsic nature of progress, but rather to these special conditions under which this particular phase of progress has occurred; and even without knowing the nature of these conditions, Durkheim insisted that the very rapidity of this growth indicated that they are morbid and pathological rather than normal.48

How, then, was this "pathological phenomenon" to be overcome? Durkheim clearly considered the present indulgence toward suicide excessive, but felt that increased penalties for self-inflicted deaths would be inefficacious. The proposed imposition of severe penalties, for example, ignored the fact that suicide is but an exaggeration of acts regarded as virtuous, which a society could hardly be expected resolutely to condemn; and the milder moral penalties (e.g., refusal of burial, denial of civil, political, or familial rights), like education, fail to touch suicide at its source. Indeed, both the legal and the educational systems are themselves products of the same currents that cause suicide itself.

The recent, pathological growth of suicide must thus be attacked at its egoistic and anomie.49 The rapid increase of egoistic suicides, for example, could be attributed to the increasing failure of society to integrate its individual members; and it could be counteracted only by re-establishing the bonds between the individual and the social group: "He must feel himself more solidary with a collective existence which precedes him in time, which survives him, and which encompasses him at all points."50

Which social groups were best prepared to exercise this reintegrative function? Certainly not the state, Durkheim insisted, for political society is too distant from the individual to affect his life forcefully and continuously. Neither is religion a binding force; for while the Roman Catholic Church once exercised an integrative influence, it did so at the cost of a freedom of thought it no longer has the authority to command. Even the family, traditionally the central cohesive force in the life of the individual, has proved susceptible to the same disintegrative currents responsible for the rapid increase of suicide. In fact, the state, religion, and the family were able to prevent suicides only because they were cohesive, integrated societies in themselves; and, having lost that character, they no longer have that effect.

But there is a group -- the "occupational group" or "corporation" -- that has enormous integrative and thus preventative potential. "Its influence on individuals is not intermittent," Durkheim emphasized for "it is always in contact with them by the constant exercise of the function of which it is the organ and in which they collaborate. It follows the workers wherever they go.... Wherever they are, they find it enveloping them, recalling them to their duties, supporting them at need. Finally," he concluded, "corporate action makes itself felt in every detail of our occupations, which are thus given a collective orientation."51

To fulfill this potential, however, the occupational groups must become a recognized organ of public life, outside of (though subject to)52 the state, and be granted definite social functions -- the supervision of insurance, welfare, and pensions; the settling of contractual disputes; the regulation of working conditions; etc. But above all, the occupational group must exercise a moral function: "Besides the rights and duties common to all men," Durkheim explained, "there are others depending on qualities peculiar to each occupation, the number of which increases in importance as occupational activity increasingly develops and diversifies. For each of these special disciplines," he concluded, "an equally special organ is needed, to apply and maintain it."53

But if this is the best way to combat "corrosive individual egoism," it is also the best means to combat anomie;54 for the same groups that re-integrate the individual into social life can also serve to regulate his aspirations: "Whenever excited appetites tended to exceed all limits," Durkheim explained, the corporation would have to decide the share that should equitably revert to each of the cooperative parts. Standing above its own members, it would have all necessary authority to demand indispensable sacrifices and concessions and impose order upon them.... Thus, a new sort of moral discipline would be established, without which all the scientific discoveries and economic progress in the world could produce only malcontents.55

The pathological increase in suicides is thus a result of the "moral poverty" of our age, Durkheim insisted, and a new moral discipline is required to cure it; but as always, he insisted that this moral poverty itself had structural causes, and thus a reform of social structure (i.e., decentralized occupational groups)56 was required to relieve its most morbid symptoms.

Critical Remarks

As the first systematic application of the methodological principles set out in his "manifesto" of 1895, Suicide reveals their limitations as well as their advantages, and thus provides an occasion for considering a number of difficulties -- argument by elimination, petitio principii, an inappropriate and distortive language, etc. -- which, though typical of Durkheim's work as a whole, are perhaps most clearly seen here. Durkheim's characteristic "argument by elimination," for example, pervades both The Division of Labor and The Elementary Forms, but there is no better example of its power to both persuade and mislead than Durkheim's discussion of "extrasocial causes" in Book One of Suicide. Briefly, the argument consists of the systematic rejection of alternative definitions or explanations of a social fact, in a manner clearly intended to lend credibility to the sole remaining candidate -- which is Durkheim's own. Durkheim's use of this technique, of course, does not imply that his candidate does not deserve to be elected; but as a rhetorical device, argument by elimination runs at least two serious risks: first, that the alternative definitions and/or explanations might not be jointly exhaustive (other alternatives may exist); and, more seriously, that the alternative definitions and/or explanations might not be mutually exclusive (the conditions and causes they postulate separately might be conjoined to form perfectly adequate definitions and/or explanations other than Durkheim`s' "sole remaining" candidates). Durkheim's persistent use of this strategy can be attributed to his ineradicable belief, clearly stated in The Rules, that a given effect must always have a single cause, and that this cause must be of the same nature as the effect.57

Petitio principii -- the logical fallacy in which the premise of an argument presumes the very conclusion yet to be argued -- is, again, a feature of Durkheim's work as a whole. In The Elementary Forms, for example, Durkheim first defined religion as a body of beliefs and practices uniting followers in a single community, and later he concluded that this is one of religion's major functions. But there is no clearer instance of this style of argument than Durkheim's "etiological" classification of the types of suicide, which of course presupposes the validity of the causal explanations eventually proposed for them. The point, again is not that this automatically destroys Durkheim's argument; but it does make it impossible to entertain alternative causes and typologies, and thus to evaluate Durkheim's frequently ambitious claims.58

Durkheim's repeated insistence that sociology is a science with its own, irreducible "reality" to study also led him to adopt a language that was both highly metaphorical and systematically misleading. This is first evident in The Division of Labor, where abundant biological metaphors continuously suggest that society is "like" an organism in a variety of unspecified and unqualified ways; and it is still more pernicious in The Elementary Forms, where the real themes of the work -- the social origin of religious beliefs and rituals, their symbolic meanings, etc. -- are frequently disguised beneath the obfuscatory language of "electrical currents" and "physical forces."59 Suicide combines the worst elements of both; and in particular, this language made it difficult if not impossible for Durkheim to speak intelligibly about the way in which individual human beings perceive, interpret, and respond to "suicido-genic" social conditions.60

Finally, it might be argued that Durkheim's central explanatory hypothesis -- that, when social conditions fail to provide people with the necessary social goals and/or rules at the appropriate levels of intensity their socio-psychological health is impaired, and the most vulnerable among them commit suicide -- raises far more questions than it answers. Aren't there different kinds of "social goals and rules," for example, and aren't some of these dis -harmonious? What is socio-psychological "health"? Isn't it socially determined, and thus relative to the particular society or historical period in question? Why are disintegrative, egoistic appetites always described as individual, psychological, and even organic in origin? Aren't some of our most disruptive drives socially generated? And if they are, aren't they also culturally relative? Why are some individuals rather than others "impaired"? And what is the relationship (if, indeed, there is one) between such impairment and suicide? The fact that these questions and others are continuously begged simply reiterates an earlier point -- that Durkheim's macro-sociological explanations all presuppose some social-psychological theory, whose precise nature is never made explicit.61


  1. 1897b: 42.
  2. Durkheim recognized that this distinction was already not without ambiguity: "In one sense, few cases of death exist which are not immediately or distantly due to some act of the subject. The causes of death are outside rather than within us, and are effective only if we venture into their sphere of activity" (1897b: 43).
  3. 1897b: 48.
  4. 1897b: 51.
  5. Durkheim did not deny, therefore, that individual conditions may cause individual suicides, nor did he deny that this was a profitable area of study for psychologists; but most of these conditions are insufficiently general to affect the suicide rate of the society as a whole, and thus they were of no interest to the sociologist (cf. 1897b: 51-52).
  6. 1897b: 67.
  7. It is interesting to see how little Durkheim claimed to understand this process in 1897 (cf. 1897b: 125-127, 130 n10), particularly in light of its significance in Durkheim's later explanation of religious belief and ritual (cf. 1912: 240 245).
  8. 1897b: 129.
  9. Durkheim also distinguished between suicides caused by moral contagion (originating in one or two individual cases and then repeated by others) and those caused by moral epidemic (originating in the whole group under the influence of a common pressure): the first involved imitation and were thus attributable to psychological causes while the second was a social fact subject to social causes (cf. 1897b: 131-130).
  10. This scorns to have been the main point of this chapter altogether: cf. Durkheim s conclusion: No fact is more readily transmissible by contagion than suicide. yet we have lust seen that this contagiousness has no social effects. If imitation is so much without social influence in this case, it cannot have more in others: the virtues ascribed to it are therefore imaginary (1897b: 141, 142).
  11. 1897b: 141.
  12. Cf. Lukes, 1972: 31-32.
  13. 1897b: 147.
  14. 1897b: 151.
  15. 1897b: 157. Here Durkheim was not denying the idealistic nature of Roman Catholicism by contrast, for example, with Greco-Latin polytheism and Hebrew monotheism; rather, he was arguing that Protestantism stresses religious individualism and freedom of personal interpretation, while "all variation is abhorrent to Catholic thought" (1897b: 158).
  16. 1897b: 158.
  17. Durkheim thus acknowledged the role of minority status, not because religious hostility Imposes some "higher morality," but because it forces the minority to achieve greater unity and integration (cf. 1897b: 159-160).
  18. 1897b: 168.
  19. Cf. the similar point in The Division of Labor (1893: 409).
  20. 1897b: 170.
  21. This was Durkheim's famous "coefficient of preservation"--the number showing how many times less frequent suicide is in one group than in another at the same age. Thus. when Durkheim said that the coefficient of preservation of husbands of the age of 25 in relation to unmarried men is 3, he meant that, if the tendency to suicide of married persons at this time of life is represented by 1, that of unnamed persons must be represented by 3. When the coefficient of preservation sinks below 1, Durkheim described it as a "coefficient of aggravation" (1897b: 177).
  22. Durkheim thus rejected the Malthusian connection between limitation of family size and general well-being: "Actually [this restriction] is so much a source of the reverse condition that it diminishes the human desire to live. Far from dense families being a sort of unnecessary luxury appropriate only to the rich, they are actually an indispensable staff of daily life" (1897b: 201).
  23. 1897b: 208.
  24. 1897b: 213.
  25. 1897b: 214.
  26. "Obligatory" altruistic suicide is the essential type, from which two others derive -- i.e., "optional" altruistic suicide, in which a concurrence of circumstances makes self-inflicted death praiseworthy, thus encouraging it without requiring it; and "acute" altruistic suicide (of which "mystical" suicide is the "perfect pattern"), in which the individual kills himself for the pure joy of sacrifice and self-renunciation (cf. 1897b:222-225).
  27. This is not to say that a melancholy view of life automatically increases the inclination to suicide. Christians, for example, have a gloomy conception of this life combined with an aversion to suicide, a conjunction Durkheim attributed to their "moderate individualism" (cf. 1897b: 226).
  28. 1897b: 238-9.
  29. 1897b: 246.
  30. 1897b: 247. This argument -- that desires are simple and few in the "state of nature," but multiply with advancing civilization -- is one that we (and presumably Durkheim) owe to Rousseau's Discourse on the Origin of Inequality (1755).
  31. 1897b: 247-8.
  32. Cf. Davis and Moore. "Some Principles of Stratification", American Sociological Review, Vol. X (April, 1945). pp. 242-249.
  33. Durkheim used the term "repressive anomy" to describe the condition produced by a reclassification downward in the social hierarchy, and "progressive anomy" to describe its upward counterpart (cf. 1897b: 285).
  34. 1897b: 256.
  35. Cf. Book Two, Chapter 3.
  36. Cf. the study of Bertillon (September. 1882), summarized by Durkheim on p. 260.
  37. 1897b: 270.
  38. 1897b: 275-6.
  39. 1897b: 309. Durkheim thus presented an alternative explanation for a phenomenon -- the statistical regularity of certain social phenomena over time -- first analyzed scientifically in Adolph Quetelet's Sur l'homme et le développment de ses facultés ou Essai de physique sociale (1835) and Du système social et des lois qui le régissent (1848). See Durkheim's discussion of these works on pp. 300-304.
  40. Cf. 1893: 96-100.
  41. 1897b: 310-11.
  42. 1897b: 318. Durkheim insists that the alternative view -- that a whole is qualitatively identical with the sum of its parts, and an effect qualitatively reducible to the sum of its causes -- would render all change inexplicable; and he again attacks what he takes to be Tarde's defense of this position (cf. 1897b: 311-312).
  43. See, for example, the famous arguments against such proscription presented in Montesquieu's Persian Letters (1721) and Hume s "On Suicide" (1783).
  44. 1897b: 334.
  45. This modern "cult of man" should not be confused with the "egoistic individualism" discussed earlier. The latter represents an insufficient state of integration which detaches the individual from society with dangerous consequences, the former unites the members of a society in a single thought, the disinterested impersonal conception of an "ideal humanity" which transcends and subordinates private, selfish goals (cf. 1897b: 330-337).
  46. Durkheim recognized that the homicides produced by conditions of anomie and those produced by conditions of altruism could not be "of the same nature"; like suicide, therefore. homicide "is not a single, individual criminological entity, but must include a variety of species very different from one another" (1897b: 358).
  47. Cf. the discussion of crime presented in 1897b: 98-104.
  48. Cf. 1897b: 85-107. To these symptoms of pathology, Durkheim adds the rise of philosophical pessimism. Comparing his own intellectual milieu with that of Zeno and Epicurus, he points to the systems of Schopenhauer and Hartmann, and the more broadly based intellectual movements of anarchism, aestheticism, mysticism, and revolutionary socialism as evidence of a "collective melancholy" which "would not have penetrated consciousness so far if it had not undergone a morbid development" (1897b: 370).
  49. The causes of altruistic suicide, as we have seen, played no role in the "morbid effervescence" of nineteenth-century suicides, and appeared instead to be declining. Fatalistic suicide was already a subject of merely historical interest.
  50. 1897b: 373-374.
  51. 1897b: 379. Cf. the important Preface to the second edition of The Division of Labor (1902) which extends this argument.
  52. The failure of the state to perform this regulative function in the past, Durkheim suggests, led to the eventual suppression of the medieval guilds altogether: "... if similar corporations of different localities had been connected with one another instead of remaining isolated. so as to forth a single system, if all these systems had been subject to the general influence of the State and thus kept in constant awareness of their solidarity, bureaucratic despotism and occupational egoism would have been kept within proper limits" (1897b: 381-382).
  53. 1897b: 380.
  54. Though not all kinds of anomie. The only way to reduce suicides arising from conjugal anomie (i.e.. divorce), Durkheim suggests, is to make marriage more indissoluble; but by thus diminishing the suicides of husbands, we would also increase those of wives, for whom the matrimonial bond is of considerably less benefit. This dilemma can he overcome, however. if we recall that the differential advantage enjoyed by the husband is due to the fact that his aspirations are of societal origin while the wife is more influenced by physiology. While Durkheim was sure that a woman could never fulfill the same social functions as a man, therefore, he still felt that granting women a more active and important role in society would eventually secure them the same advantages from matrimony hitherto enjoyed only by men (cf. 1897b: 384 386).
  55. 1897b: 383. This parallel efficacy is due to the fact that, at least in part, anomie results from the same cause -- the "disaggregation" of social forces -- as does egoism; but in each case "the effect is different, depending on the "point of incidence" of this cause, and whether it "influences active and practical functions, or functions that are representative. The former it agitates and exasperates; the latter it disorients and disconcerts" (1897b: 382).
  56. "Decentralization" was the watchword of a number of Durkheim's contemporaries who, recognizing the impotence of the monolithic state in the face of egoism and anomie, sought to restore to local groups some of their old autonomy. Durkheim's proposed reform was thus a specific variation on this more general theme. The particular advantage of occupational decentralization, he urged, is that, because each of these new centers of moral life would be the focus of only specialized activity, the individual could become attached to them, and they could become attached to one another, without the solidarity of the country as a whole being undermined (cf. 1897b: 390-391).
  57. Lukes, 1972: 31-33.
  58. Cf. Lukes, 1972: 31.
  59. In his defense of The Elementary Forms before the Société francaise de philosophie (1913), Durkheim insisted that his primary concern had been to point out the "dynamo-genic" quality of religious belief.
  60. Cf. Lukes, 1972: 34-36.
  61. Cf. Lukes, 1972: 213-222.

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