Suicide
Suicide
I. Social AspectsJack D. Douglas
II. Psychological Aspects (1)Edwin S. Shneidman
III. Psychological Aspects (2)Norman L. Farberow
I. SOCIAL ASPECTS
Suicide has been an object of fundamental concern to Western men of all cultural periods, though the importance given to suicide itself, as well as the degree to which it has entered into other concerns, has varied greatly from one period to another. “Suicide,” said Goethe, “is an incident in human life which, however much disputed and discussed, demands the sympathy of every man, and in every age must be dealt with anew” ([1811-1833] 1908, vol. 2, p. 125).
This is not to say that suicide has failed to interest people in other cultures but, rather, that the reasons have usually been different. In feudal Japan, for instance, suicide was an ultimate act of honor, redemption, or union. In the Western world, however, suicide has always been regarded as fundamentally problematic. Possibly because suicide in the West has nearly always been interpreted as voluntary self-destruction, with no element of social constraint or obligation, philosophers from Plato to Camus have used it as a starting point for reflection on the eternal problems of human existence. Indeed, ambivalence toward the fact of his own existence is one of Western man’s most persistent characteristics; life and death, the relation of man to man, and the relation of man to himself have been subject to continual questioning, and suicide has been seen as relevant to all of these. In such a climate of thought, it is not surprising that neither settled opinion nor emotional consensus has ever been reached concerning the nature of suicide.
Early writings . Early Western writers on suicide were primarily concerned either with its desirability (including its honorability) or with its morality. But all of them, whether Greek, Roman, Jew, or Christian, were concerned to some degree with the particular facts of actual cases of suicide. Most importantly, from our point of view, they implicitly assumed that before one could properly judge the actor in the case, one had also to know his intentions, the situation in which he found himself, and the nature and outcome of his actions.
After the normative arguments against suicide by Augustine, the Council of Aries, and the Council of Bourges, there is little evidence that the normative evaluation of suicide was considered to be very problematic until the seventeenth century. But there remained the cognitive problem of deciding which set of facts concerning intentions, situations, and actions should be imputed to the category of suicide. Most important to Christian thinkers was the problem of categorizing “martyrs”: were they suicides or not? This cognitive problem was to be just as difficult for sociologists in the twentieth century as for theologians in the fifth century (see Halbwachs 1930).
The nineteenth century . The fundamental forms of thought about suicide, including the implicit assumptions that suicide must be studied as a form of normative action and that suicide is connected with the problem of man’s relation to God and man’s relation to man, were all part of the literary and common-sense background of sociological thought about suicide in the nineteenth century. But there were many other, more specific influences on these nineteenth-century works.
Almost all of the many works on suicide in the eighteenth and nineteenth centuries included extensive and detailed considerations of the historical cases, the literary examples, and the philosophical treatments of suicide. The works of Appiano Buona-fède, Louis F. Bourquelot, Carl F. Stäudlin, Albert des Étangs, Pierre J. C. Debreyne, Alexandre Brière de Boismont, and Alfred Legoyt, to mention but a few, were greatly concerned with such sources. Among the literary, philosophical, and historical influences on the developing sociology of suicide, three can be singled out as most important.
(1) Knowledge of the high frequency of suicide among the upper classes in ancient Rome, and of the honorableness of suicide among the Romans, led to the idea that suicide is fundamentally a matter of social custom. According to Voltaire and other thinkers of the Enlightenment, suicide in a society where suicide is neither the custom nor positively valued or expected was due to a failing, weakness, or disease. This background assumption helped to produce both the psychological theory that suicide is caused by insanity (or character weakness, or disequilibrium), a theory of great influence even today (Dahlgren 1945; Achille-Delmas 1932; Deshaies 1947), and the sociological theory that suicide is caused by a failure of the normative control of individuals by society (Brière de Boismont 1856; Morselli 1879; Durkheim 1897; Cavan 1928).
(2) In the eighteenth century, suicide became a focal point in the ethical argument between the philosophes and the church supporters. Largely as an outgrowth of this struggle, the relation of religion to suicide became a leading concern of the nineteenth-century sociologists. Specifically, this tradition of thought was largely responsible for the explanations of the great statistical increase in suicide in nineteenth-century Europe in terms of meaninglessness of life, lack of discipline, self-centeredness, moral disorganization, and materialism. These developments are dealt with more fully below.
(3) Death, especially death by suicide, was a favorite theme in the literature of the romantic movement. Indeed, the most important contribution of this literary tradition to the developing sociology of nineteenth-century Europe was probably that it helped to focus attention on suicide. Largely because of its treatment in literature, suicide was seen by the educated public as a fundamental social problem. Nineteenth-century Europeans, especially the French, were as frightened of la manie du suicide (as Tissot [1840] called it) as twentieth-century Americans are of juvenile delinquency and the Mafia. But the literary concern with suicide also provided the core of certain very important, specific “theoretical” explanations. Of fundamental importance was the prime romantic symbol of an isolated, lonely hero of a poetic (or intellectual) bent who wanders far from human society in search of the impossible and, failing of the impossible, becomes increasingly melancholy and enamored of eternity. The idea that self-imposed isolation produces “melancholy,” and thence suicide, was hardly novel; indeed, Robert Burton’s seventeenth-century treatise, The Anatomy of Melancholy, had carefully documented its classical sources. But this prime symbol of romanticism contained the most unclassical ideas of both egoism and anomie. Indeed, the romantic image of suicide seems to have been so much the mythical model which the sociologists of suicide had in mind that both Morselli ([1879] 1882, p. 297) and Durkheim (1897) had a strong tendency to treat egoism and anomie as almost identical. It is even possible to specify the work of Chateaubriand, especially as treated by Briére de Boismont (1856, pp. 39-40), as the major source of Durkheim’s idea of anomie suicide (Douglas 1965, pp. 20-26).
Some of the fundamental ideas about suicide which the sociologists inherited from history, philosophy, literature, and common sense became the cores of the various sociological theories of suicide. There were other influences on these theories which will be described later. But the sociologists of the period were dependent upon these nonscientific and, most importantly, nonempirical sources for their ideas about suicide as a specific phenomenon. Though this is somewhat less true today, the sociological study of suicide is in substantially the same condition.
The definition of suicide
A great number of different definitions of suicide have been proposed by the students of suicide during the first two hundred years of the term’s general use. Almost all of these definitions have included, to varying degrees and in various combinations, one or more of the following conceptual dimensions: (a) the initiation of an act that leads to the death of the initiator; (b) the willing of an act that leads to the death of the wilier; (c) the willing of self-destruction; (d) the loss of will; (e) the motivation to be dead (or to die, or to be killed) which leads to the initiation of an act that leads in turn to the death of the initiator; (f) the knowledge of an actor concerning the relations between his acts and the objective state of death; (g) the degree of central integration of the decisions of an actor who decides to initiate an action that leads to the death of the actor; (h) the degree of firmness or persistence of the decision to initiate an act that leads to the death of the initiator; (i) the degree of effectiveness of the act in producing death. (For a detailed presentation of the various definitions, see Schneider 1954, pp. 9-59; for an analysis of the definitions of suicide, see Douglas 1965.)
The great profusion and confusion of definitions of suicide have been very largely the result of combining an abstract, a priori approach to defining the concept with an assumption of “verbal realism” —i.e., the assumption that if the same name is in fact used to refer to a set of phenomena, then that set of phenomena must in fact have some shared property that is designated by the shared word. Durkheim and almost all other sociologists and psychologists who have attempted to carefully define suicide, have assumed that one should arbitrarily define the concept to fit his scientific methods and his theoretical purposes but that the definition should not differ too much from common usage (Durkheim [1897] 1951, pp. 41-42). Durkheim further assumed that the most common definition of suicide was that of death caused by an action initiated by the actor with the intention of causing his own death. Since, however, teleology is anathema to positivists, Durkheim decided that intentions are too “intimate” to yield valid information and that, consequently, “knowledge” by the actor of the deadly consequences of his actions should be the fundamental factor in deciding that a death is a suicide, even if the consequences are relatively uncertain (ibid., pp. 43-46).
Unfortunately, Durkheim’s definition of suicide was quite irrelevant to his work on suicide. As Maurice Halbwachs most astutely argued, Durkheim was almost totally dependent on the official suicide statistics for his data, so the only definition that really mattered was that (rather, those) which the officials had in mind when they categorized the causes of deaths. Since Halbwachs believed that one could not know what definitions these officials used, he concluded that sociologists must be satisfied with no definition at all (Halbwachs 1930). Actually, the coroners and doctors who do the categorizing seem generally to be firm believers in the common-sense theory of teleology. Moreover, the laws governing the certification of cause of death very generally specify “intention” as necessary for a classification of suicide. Benoit-Smullyan (1948) has also pointed out that since Europeans do not generally consider self-sacrifices to be suicides, Durkheim’s “altruistic suicides” would not be included in the official statistics on suicide. Consequently, it seems clear that Durkheim’s definition of suicide was not so much irrelevant to his data as it was a complete distortion of the meaning of the data.
With the exception of Halbwachs’ work, the published sociological works on suicide since Durkheim’s Suicide have generally assumed that Durkheim’s definition is best; they have even failed to note Halbwachs’ valid criticism of it. They have, however, also either implicitly assumed “intention” to be a necessary factor, or else they have totally avoided the problem. The result has been both confusion and a failure to recognize the problems involved in defining suicide. The sociology of suicide is in obvious need of a system of categories which will describe the observed phenomena that are significant in terms of a theory of suicide. Such descriptive categories must be worked out in partial independence of the theories in order to aid their development, but they must also be worked out in partial dependence on developing theories. However, the first critically important thing for sociologists to do is to make direct observations of the empirical phenomena which are called suicide in any society, so that definitions will be relevant and so that the meanings of the data will be known to the theorists.
Official statistics on suicide
Largely because of the rapid acceleration in statistics-keeping in eighteenth-century Europe, the systematic recording and tabulating of suicides had become standard practice in many of the governmental regions of Europe by the first quarter of the nineteenth century. The basic ideas of “moral statistics,” especially the fundamental idea that a given incidence of a phenomenon such as suicide is regularly associated with a given population, had been clearly formulated in the eighteenth century, certainly by 1741, when Johann Peter Süssmilch published his famous treatise on the “divine order” underlying demographic phenomena. The traditions of “political arithmetic,” of studying racial and national character, and of comparative studies of suicide rates were synthesized by Adolphe Quetelet (1835) in order to produce his “average man” theory of suicide rates. Quetelet’s fundamental idea was that each stable “social system” produces a stable, average personality type about which the individual personalities tend to cluster. The average man, therefore, has a given, stable probability of committing suicide in a given period of time.
Quetelet’s probabilistic theory was specifically intended to explain the observed regularity (or order) in the official suicide statistics. This regularity, which was first observed for the different nations of Europe, seemed to the moral statisticians to be a remarkable demonstration of the deterministic, lawlike nature of the actions of individuals. Since the seventeenth century, Europeans had believed suicide to be more frequent in some nations than in others. The French considered suicide to be as English as gambling (they seem to have believed that more Englishmen preferred death to life because the fog made living so unworthwhile), though Voltaire thought this impression was due to more newspaper coverage of suicide in England. But before the nineteenth century, the explanations of these (assumed) regular differences in suicide frequencies always involved the assumption that the individual will was the ultimate cause of the action. In the early nineteenth century both the psychiatric theory that all suicides are insane (and, therefore, do not cause their own actions) and the direct comparison of national suicide rates (without any real consideration of the individuals who committed the actions) combined to eliminate the individuals (or wills, or personalities) from consideration as possible causes of suicide. Quetelet’s early work on the comparisons of social rates of suicide (work that gained much support from the researches of Gustave F. Étoc-Demazy and Andre M. Guerry) was quickly given a metaphysical basis by the English positivists, especially by Henry Thomas Buckle. This largely eliminated Quetelet’s variable of average personality; the direct comparison of other external, objective data (such as geographical-ecological distributions and marriage rates) with suicide rates was now an established method with a metaphysical justification, that of positivistic determinism.
Largely because of the vast influence of Brière de Boismont’s great work on suicide (1856), based primarily on case studies of 4,595 suicides, including 1,328 suicide notes (a source of data first used by Guerry as early as 1833), and 265 attempted or planned suicides, the more important works on suicide by Wagner (1864) and Morselli (1879) did not deny causal significance to various individual factors, such as motives. They did, however, consider the comparative analysis of officially computed suicide rates to be necessary for establishing causes of suicide “. . . such as the most positive mode of psychological study would fail to discover in the individual” (Morselli [1879] 1882, p. 10). Durkheim’s sociologistic theory of suicide proposed that suicide rates are a sine qua non of the sociological study of suicide, and the great influence of his Suicide (the first edition of which appeared in 1897, although it was not translated into English until 1951) quickly made the statistical approach almost identical with the sociological approach, especially in the United States. Indeed, Kruijt (1960) and others have recently assumed the two to be totally identical.
Validity of official data . Since sociologists have relied almost exclusively on the official statistics and coroners’ reports on suicide for the data from which to develop and test theories of suicide, the question of the validity of the official data is absolutely critical in any evaluation of these sociological works. Throughout the nineteenth century most students of suicide were highly critical of the official statistics on suicide. Esquirol (1838), Brière de Boismont (1856), Legoyt (1881), Strahan (1893), and many others believed the official statistics grossly underestimated the actual number of suicides. Even Morselli (1879) most emphatically warned against the dangerous misuse of suicide statistics. But the sociologists, especially Durkheim (who included very little consideration of the value of his data in Suicide), consistently assumed that the evidence was good enough for the kinds of positivistic (i.e., objective, by definition) analyses they wanted to make. In general, they argued that (a) the stability of the statistics shows that there are no significant errors in them and (b) there are no good reasons to conclude that there are consistent biases in the data, i.e., the errors cancel each other out.
The first argument is based on the mistaken assumption that errors cannot be patterned or stable. Buckle (1857-1861, p. 18) even noted that the frequency of incorrectly addressed letters is extremely stable from year to year without noting the obvious implication for stable suicide statistics. Though contemporary sociologists (especially Americans) who have used the official records on suicide rarely have explicitly considered their value as evidence, they seem generally to have assumed that the second argument given above is correct. The most fundamental reasons for denying the validity of the argument, and therefore the adequate validity of the statistics, are as follows.
(1) It seems quite likely that the official categorizers of suicide in different nations, states, and cities have used different abstract and operational definitions of suicide. We have noted above that the formal definitions of suicide by the students of suicide have varied greatly and in a most complex fashion. But there have been some consistent differences in definitions between groups, such as those between the psychiatrists, the psychologists, and the sociologists. Why would one not expect to find similar consistent differences in the abstract and operational definitions of suicide used by different groups of official categorizers of causes of death? A coroner of a large city who will categorize a cause of death as suicide only if a suicide note is found clearly will produce results consistently different from those of a coroner in another city who uses a different rule-of-thumb definition, such as that there must be an eyewitness report, a clearly perceptible motive, etc. Only “verbal realism” can lead one to assure a priori that the coroner in Lapahaw, Georgia, applies the word “suicide” in a fashion substantially similar to the way the word is applied by the coroner in Los Angeles, California.
(2) It also seems most plausible to assume that there are consistent differences in the abstract and operational definitions of the terms of the different definitions, such as “cause,” “responsible,” “sane,” “death,” and “intention.” For example, some official agencies will not categorize the cause of death as suicide if death occurs three days after the injurious act; and it is worth observing here that all bureaucracies create ad hoc rules of thumb in order to effectively apply general policy to their own particular problems.
(3 ) A great deal of evidence has been presented by Herman Krose, Georg von Mayr, Halbwachs (1930, pp. 19-39), and Achille-Delmas (1932) which has shown both that (a) there are consistent differences in the administrative practices of dealing with suicide (and related categories) and that (b) changes (hence, differences) in administrative practices regarding suicide produce immediate and highly significant changes in suicide rates. Two brief examples are instructive. Halbwachs (1930, pp. 33-34) found that, though a ruling in 1866 made certification of cause of death by a doctor or public health official obligatory throughout France, in fact almost all rural deaths escaped certification; certification of cause of deaths in public places was made by police officials, and certification by family doctor was universally acceptable. He also found that a reform in the Prussian Bureau of Statistics in 1883 produced a 20 per cent stable increase in the suicide rate, beginning in 1884 (Douglas 1965, pp. 259-377).
These and other facts strongly suggest that the relationship between statistical organizations and the suicide rates they produce is subject to the following principle: other things being equal, suicide rates vary directly with the degree of professional medical training of the categorizers, the average rate of man-hours devoted to “cause of death” categorization, and the independence of the categorizers from “interested parties” (for which see below). This general principle leads us to expect, among other things, that (a) urbanization and suicide rates will vary directly; (b) both industrialization and the wealth of populations will vary directly with suicide rates (a possible explanation of the great increase in suicide rates of Europe in the nineteenth century and of developing nations today); and (c) periods of general disorganization or reorganization, such as wars, will produce decreases in suicide rates (because the officials are fewer and have more important things to do). And, in general, these predictions are strongly supported.
(4) The most important set of “other things” that are not equal, and which we must expect to cause great biases in official statistics on suicide, are the various social meanings of suicide, especially as they relate to approval or disapproval. Because of his positivist philosophy, his peculiar theory of “social pathology,” and his theory of “juridical norms,” Durkheim assumed that all groups in Europe were equally against suicide. With the partial exception of the works of Halbwachs (1930) and Cavan (1928), almost all sociological theories of suicide since Durkheim have implicitly assumed the normative definitions of suicide to be both invariant and highly negative. This is a totally fallacious and most unfortunate assumption. In his brilliant study of French attitudes toward suicide, Albert Bayet (1922) conclusively demonstrated that over a long period of time there was a consistently far more negative (even horrified) attitude toward suicide on the part of the “simple,” uneducated population (largely rural) than on the part of the more educated, upper-class population. Such patterned differences in the normative meanings of suicidal actions might produce different frequencies in actual (or real) suicidal actions, but they would seem far more likely to produce differences in the frequency and strength of attempts to conceal suicide both by the suicidal individual and by his significant others. Moreover, one would expect that the more a primary group has to lose from having one of its members socially categorized as a suicide, the greater will be the frequency and strength of attempts to conceal suicide. These two factors make up what can be called a group coefficient of attempted concealment of suicidal actions. We must also expect, however, that different groups will have different coefficients of success in such attempts, the major determinants of which would seem to be degree of social influence and degree of social integration of the group (or individual) making the attempt at concealment. Obviously, these two coefficients could produce many biasing effects on official statistics, the most likely of which seems to be that, given a certain degree of negative normative definition of suicidal action, the degree of social integration of the suicide’s primary group into the general community will vary directly with the coefficient of attempted concealment and with the coefficient of success in attempted concealment. This means, of course, that official statistics on suicide will tend to be biased in such a direction that they will support an integration theory of suicide, such as Durkheim’s.
(5) Over half a century before the publication of Durkheim’s Suicide, Jean-Étienne Esquirol (1838) had established that some methods of committing suicide make valid categorization of the cause of death more difficult. Since, moreover, there seem to be patterned variations between social categories in suicide methods, we must expect biases in official statistics from this source. For example, women use barbiturates and gas far more frequently than do men. Suicides by barbiturates and gas are very difficult to distinguish from accidents. Since such doubtful cases are almost always categorized as “accidental” or “accidental suicide,” we must expect bias in the direction of lower official suicide rates for women.
Two conclusions regarding the official statistics on suicide seem justified: (a) we have little specific, systematic knowledge about the means employed by different statistical bureaus to arrive at these figures; (b) what knowledge we do have about these figures and the means of arriving at them strongly supports our arguments that they are highly biased in certain directions. In general, at the present time there seems to be no adequate justification for using official statistics on suicide to build or test a scientific theory of suicide. There seems to be every reason for not using them (Douglas 1965, pp. 259-377).
Sociology and the official data . Almost every published theory of suicide that has been called sociological has depended on the official statistics on suicide for its testing. This means, of course, that the testing of these theories is subject to all of the biases of the official statistics and therefore cannot generally be accepted.
But this does not necessarily mean that the theories of suicide are themselves wrong. All the so-called sociological theories have been general abstract theories which the authors believed to be applicable to a great variety of social behavior—for instance, to all deviant behavior. These theorists have never derived their general theories from the official statistics, still less from any actual observations of cases of suicide. On the other hand, they have always known from the beginning most of the patterns of suicide shown by the official statistics, so that they have been able to select (no doubt unconsciously) just those patterns for consideration in the testing stage which their theories can “predict.” In spite of rhetorical disclaimers that any such bias exists, the fact remains that all published sociological theories of suicide have been supported at a high level of statistical significance by “tests” using the official statistics on suicide. Moreover, since all of these works after Durkheim’s Suicide have included only the most perfunctory consideration of alternative theories for explaining the same data, there is at present no justification for considering any one theory better than any other, unless one were to consider sophistication of argument an acceptable criterion, in which case Durkheim would hold the field unopposed (Douglas 1965, pp. 112-259).
Major theories of suicide
Durkheim’s sociologistic theory. In Suicide Durkheim was, of course, primarily concerned with proving by demonstration that sociology is a scientific discipline sui generis. The work was a polemic in a great ideological war; and as Halbwachs has said (1930, p. 3), its argumentative power has made it convincing (though, I would add, only to those already convinced of its general position). To achieve his goal of demonstration, Durkheim made use of the mass of published material on suicide statistics, the many statistical relations already established between suicide rates and social relations by the flourishing school of “moral statisticians,” and the methods of analysis developed by Bertillon and many others. Earlier sociologists, especially Morselli (1879), had considered “society,” and in particular the morals of a society, to be the most important cause of a given suicide rate. It was from these previous works that Durkheim took his specific ideas, such as “egoism” and “lack of moral restraint,” about what caused suicide rates to vary. But previous theorists, with the major exception of Quetelet, had proposed multifactored explanations; Durkheim proposed a general synthesis. In general, he took his notion of statistical relations and methods from the moral statisticians and, like many of the moral statisticians before him, turned to the romantics, and the psychologists whom they had influenced, for his fundamental causal variables. Durkheim’s own contribution was to translate “egoism,” “need for an external, moral authority,” and similar psychological variables into “social” (or cultural) variables which he finally reduced to the two opposing moral dimensions of egoism-altruism and anomie-fatalism.
Unfortunately for Durkheim’s ambitious program, Suicide (1897) is an extremely uneven work. The theory presented in Books 1 and 2 is for the most part extremely positivistic, relating external (or “objective”) variables to other external variables. In Book 3, however, the theory is radically different: internal variables are related both to each other and to external variables. The positivistic version is the one usually given primacy by American sociologists (see, for example, Gibbs & Martin 1964); but the later version is the more considered, developed, and tenable theory. The core of the latter is very simple, though its presentation is confused by Durkheim’s penchant for multiple terms (e.g., “disintegration,” “disorder,” “disequilibrium,” “lack of unity or cohesion”), petitio principii, and lack of conceptual clarity. He assumes that for all societies there is some optimal equilibrium or integration of egoism-altruism, on the one hand, and anomie-fatalism, on the other. Any change in the relative strengths of these ideas (or forces) will produce an increase in “social disintegration,” which will lead in turn to an increase in the suicide rate and in the particular type of social relation associated with the “force” that is on the increase. Hence, there is a statistical relation between suicide rates and the incidence of different types of social relations.
The critical flaw in the work is the lack of any means of measuring these forces of egoism, anomie, etc., either independently of each other or independently of a change in suicide rates. Yet Durkheim assumed that these were the fundamentally important variables: any change in suicide rates or social relations meant to him that one or more of the forces was increasing relative to the others. But how did he determine which force was increasing? We must conclude that he inferred it from the associated changes in social relations, and that, for the most part, he assumed he knew from common sense what these changes meant. Unfortunately, the concrete examples Durkheim gives us will not support the notion that his common sense was superior, as a scientific instrument, to anyone else’s (Douglas 1965, pp. 7-112).
The major contribution of Durkheim’s Suicide was its grand conception of the general nature of sociological theory. Durkheim’s intentions in this work went much too far beyond the possibilities of the theory and data of his day—and of ours.
Halbwachs’ theory of suicide. The work of Maurice Halbwachs (1930) began as a supplement to Durkheim’s Suicide but became a radical re-orientation of sociological theory. Halbwachs concluded that Durkheim, in spite of his extensive use of replicative analyses, had overlooked the high degree of overlap of the variables he found to be related to suicide rates. The only fundamentally significant independent relation, Halbwachs argued, is between suicide rates and the degree of complexity of a society: thus, the rural style of life (or sociocultural system) is simpler than the urban and therefore has a lower suicide rate. But Halbwachs also rejected Durkheim’s sociologism. He maintained that there is a complementary relation between the motives imputed to suicides and social situations of isolation. To support his theory, Halbwachs presented the most extensive and thorough analysis of official statistics in all of the sociological literature. As we have already said, however, it seems clear that there is a fundamental bias in the official statistics in the direction of underestimating rural rates (Douglas 1965, pp. 195-209).
Social disorganization theories . Unlike almost all other recent sociological works on suicide, the various “ecological” works on suicide have been done quite outside the tradition initiated by Durk-heim. Ruth Cavan’s Suicide (1928), the first of these works, was predominantly influenced by Morselli. Most ecological studies of suicide have relied heavily on some form of social disorganization theory. Various population variables, such as high mobility and complexity, are hypothesized as the causes of a relative lack of effect of social values on behavior (this being the meaning that “social disorganization” usually has for these writers), which in turn is hypothesized as the cause of suicide and similar deviant acts.
The social disorganization approach to suicide involved certain extreme assumptions which have more recently been shown to be clearly unacceptable. For one thing, it assumed cultural homogeneity of values and behavior patterns, that is, it was taken for granted that all of the many ethnic groups of cities like Chicago shared the same values and behavior practices with regard to any form of “deviance,” such as suicide was thought to be. An extensive literature has attacked this assumption when applied to delinquency, and Bayet (1922) has amassed enough evidence to prove that attitudes toward suicide may vary considerably even in a nation with a supposedly homogeneous culture. The social disorganization approach also assumed that cultural values are the only social meanings that determine rates of suicidal actions. But there is no evidence whatsoever in favor of this assumption—and, indeed, it was contradicted by Cavan’s own arguments concerning the importance of spite and other purely personal motives in causing suicide (Cavan 1928).
Cavan’s work, the earliest of the significant ecological works on suicide in America, did not involve the so-called ecological fallacy (for which see Robinson 1950). In fact, Cavan clearly recognized the fundamental principle involved in this fallacy, even to the extent of criticizing the statistical arguments of Morselli and Durkheim in terms of it (1928, p. 289). Realizing what was needed in order to avoid this fallacy, she did attempt to provide evidence regarding individual cases of suicide. In general, she argued that social disorganization causes individual disorganization, and that the latter manifests itself in patterns of increasing inability to cope with crisis situations. In an attempt to demonstrate the validity of this theory, she presented lengthy selections from the personal documents of two individuals who committed suicide. Though it seems clear now that the attempted demonstration involves an imposition of the general category of “individual disorganization” upon the statements of the individuals, it is also clear that Cavan’s work, which was followed in most important details by that of Schmid (1928), involved more consideration of real cases of suicide than any other sociological work to appear before the later 1960s. Moreover, her material on the case of Marion is still among the best available on real cases of suicide (see the discussion in Douglas 1965, pp. 140-167, 471-483).
Recent sociological theories . Most of the recent sociological theories of suicide have been formulated ostensibly within a Durkheimian content, yet most of them are psychosocial theories that are actually quite different from Durkheim’s.
The one theory that actually seems to be socio-logistic in the Durkheimian manner is that proposed by Gibbs and Martin (1964). This theory, although derived from far too simple an interpretation of Durkheim’s Suicide, is true to its positiv-istic aspect in rejecting any consideration of real cases of suicide and in attempting to relate official statistics on official categories (such as marriage) to official statistics on suicide. The basic idea of the theory is that the more socially integrated (or less “conflictful”) a set of statuses (such as age, race, marital status), the more frequently that set of statuses will be occupied by members of the society and the less frequently it will be associated with suicide. This whole approach suffers from all of the weaknesses already detected in Durkheim’s and from some important additional ones. Thus there is almost no consideration of the real-world (as opposed to official) social meanings of statuses, whereas Durkheim did at least make use of well-informed common sense to provide him with social meanings for the status categories he was interested in analyzing. In addition, the “testing” of the theory involved the analysis of only four of the truly immense number of statuses in our society, and all four of these are the standard categories used in analyses of official data on suicide, so that any sociologist making an analysis of the relations between these categories and suicide rates knows what the relations are before he begins (Douglas 1965, pp. 121-135).
Most of the other recent theories of suicide that purport to be sociological—the most significant is that of Henry and Short (1954)—include psychological variables. They almost all agree that socio-economic status change (or “reverse of fortune,” as Brière de Boismont called it) is the basic sociological cause of suicide, but each proposes a different personality theory (frustration-aggression theory, self theory, loss-of-meaning theory, etc.) to explain why some few individuals commit suicide when confronted with a given status change and almost all others do not. The basic problem is that status change is “significantly” related to an immense number of things. These statistical works never involve any demonstration that other factors are not causing both the supposed cause and the suicide. They rarely include consideration of the many alternative, conflicting theories that purport to explain the same official statistics. They generally involve careful arrangement and choice of data so that statistically “significant” results can be obtained. For example, in the work by Henry and Short (1954), social classes in America are dichotomized into high and low, an arrangement which completely obscures the U-functional relationship almost always found to exist between socioeconomic status and official suicide rates in the Western world. Moreover, theoretical argument in these works is frequently quite divorced from statistical analysis; in some cases, the two can be related only through the addition of many ad hoc assumptions (see, for instance, the analysis of Henry & Short 1954 in Douglas 1965, pp. 209-229). In some of these works, changes in the population bases for the official suicide statistics were actually not taken into consideration. For example, some of the most important analyses made by Henry and Short were for periods during which new states were being added almost each year to the U.S. Death Registration Area. This fact had led Dublin and Bunzel (1933) to use only the original Death Registration Area, but Henry and Short completely overlooked the whole problem.
However, a few of these psychosocial works, as this mingling of psychological and sociological approaches can be called, have introduced valuable new orientations into the study of suicide. Thus Breed’s interviews of the surviving families of individuals who committed suicide (1963) have finally introduced into sociology a research method long established in psychiatry. This method has moved sociologists one step further toward direct observation of real-world cases of suicidal events. Indeed, it has shown them that henceforth they must consider the effects of a suicide on others as part of the suicide phenomenon. The case studies reported by psychologists (Deshaies 1947; Schneider 1954) and the sociocultural studies of suicide made by a few anthropologists and psychiatrists (Devereux 1931; Hendin 1964) make clear what should have been obvious to sociologists long ago: suicidal actions are socially meaningful actions, and individuals commit them in order to communicate something to themselves and others about themselves and about others. Even an individual whose primary goal is a state of nonbeing will not commit suicide unless he can do so in such a way as to communicate, to himself and possibly to others as well, just the right meanings. The empirical study of suicidal actions as socially meaningful— cognitively, affectively, and normatively—opens new directions of highly fruitful research and theory (Douglas 1965, especially pp. 377-543).
Only the most comprehensive observation and description of the everyday actions and statements relevant to suicide are likely to result in scientifically useful empirical and theoretical generalizations concerning suicidal actions as socially meaningful actions. Systematic analyses of existing case studies have, however, already resulted in some very important generalizations concerning the social meanings of suicidal actions in the Western world. The best attempt to generalize about suicide in this way for a non-Western society is by Devereux (1931); there are some important contributions to such an approach in other works (see, for instance, Bohannan 1960).
Perhaps the most important high-level generalization about suicide that can be made at this time is that the situated meanings of suicidal acts are often very different from their abstract meanings. In other words, the meaning of a suicidal act for those who are directly involved with it will very rarely be the same as the meaning it has for those who are not so involved, and will certainly not be the same as for the individual who, by committing suicide, is trying to communicate something (Douglas 1965, pp. 406-440). This finding has two fundamental implications for all investigation of the social meanings of suicide and, perhaps, for all of sociology. First, it is not possible to predict or explain specific types of social events, such as suicide, in terms of abstract social meanings, such as values favoring suicide. Second, it is not possible to study the situated social meanings of suicide, which are most important in its causation, without reference to actual instances of suicidal acts with which the individuals to be questioned have been directly involved. This generalization leads one to question the value of any method for investigating any realm of social action if it attempts to abstract members of society from the involvements of their everyday lives.
This does not imply that there are no patterns of meanings common to all the events that members of a society call “suicide” or “suicidal.” When one looks at the meanings imputed to particular suicidal actions, one does find that certain general features tend to recur. Most importantly, any suicidal action is usually believed to mean something fundamental about the self of the individual committing it, or about the situation (especially the persons involved) in which he committed it, or about both of these. Whether the specific meanings realized will be directed to the actor’s self or to his situation will depend on the imputations of causality made by the various persons in the situation: will they see the individual as the cause of his own actions—that is, as “responsible” for them —or will they see him as having been caused (”driven”) to do them by circumstances (loss of job, family trouble, etc.)? The individual committing the suicidal action often attempts to place one of these two general constructions upon the action by pointing out in some way the external cause that is to be “blamed” for his suicide.
There is no one meaning or set of meanings that can be imputed to all (or even most) suicidal actions. Just what meanings are imputed in each case will depend on (1) the intentions of the various actors; (2) the socially perceived ways in which the actions are committed; (3) the specific patterns of suicidal meanings that are realized; (4) the argument processes before, during, and after the suicidal actions. It should be clearly noted that whether or not actions are socially categorized as “suicidal” depends on precisely the same sort of process. The obvious example is that of individuals who with various ends in mind (avoiding embarrassment to their families, loss of insurance money, or disgrace to themselves), arrange suicides that are designed to be taken for accidents.
It is probably not possible for individuals to construct any meanings they please for their actions, though individual creativity does extend the limits immensely, and all cases include imponderable idiosyncrasies. However, the limits remain. There are, first of all, various criteria of plausibility of motives, or rationality, though it is very likely that in some instances individuals intend their actions to be considered implausible or irrational. Second, a relatively few patterns of situated meaning play important parts in most interpretations of particular suicidal actions, and it therefore seems likely that individuals take these patterns into consideration when attempting to construct the meanings of their actions for others. The most common patterns of this sort in the Western world are those involving such motives as the search for help (Sacks 1966), sympathy, escape, repentance, expiation of guilt, self-punishment, and “seriousness” (Douglas 1965, pp. 440-511).
Much careful description and analysis in this area remain to be done (ibid, pp. 511-540). However, the basic problems and the most appropriate methods for solving them now seem clear, and there is great promise of rapid development in both the empirical and the theoretical study of suicide.
Jack D. Douglas
[See also Deviant behavior; Integration, article onsocial integration; Sociology, article onthe early history of social research; and the biographies ofDurkheim; Halbwachs; Quetelet.]
BIBLIOGRAPHY
Achille-Delmas, FranÇois 1932 Psychologie pathologique du suicide. Paris: Alean.
Bayet, Albert 1922 Le suicide et la morale. Paris: Alean.
BenoÎt-Smullyan, Émile 1948 The Sociologism of Émile Durkheim and His School. Pages 499-537 in Harry Elmer Barnes (editor), Introduction to the History of Sociology. Univ. of Chicago Press.
Bohannan, Paul (editor) 1960 African Homicide and Suicide. Princeton Univ. Press.
Breed, Warren 1963 Occupational Mobility and Suicide Among White Males. American Sociological Review 28:179-188.
BriÈre de Boismont, Alexandre 1856 Du suicide et de la folie suicide. Paris: Baillière.
Buckle, Henry Thomas (1857-1861) 1913 The History of Civilization in England. 2d ed. 2 vols. New York: Hearst.
Cavan, Ruth S. 1928 Suicide. Univ. of Chicago Press.
Dahlgren, Karl G. 1945 On Suicide and Attempted Suicide. Lund (Sweden): Universitets Bokhandel.
Deshaies, Gabriel 1947 La psychologie du suicide. Paris: Presses Universitaires de France.
Devereux, George 1931 Mohave Ethnopsychiatry and Suicide. Washington: Government Printing Office.
Douglas, Jack D. 1965 The Sociological Study of Suicide: Suicidal Actions as Socially Meaningful Actions. Ph.D dissertation, Princeton Univ.
Douglas, Jack D. 1966 The Sociological Analysis of Social Meanings of Suicide. Archives européennes de sociologie 7:249-275.
Douglas, Jack D. 1967 The Social Meanings of Suicide. Princeton Univ. Press.
Dublin, Leonard I.; and Bunzel, Bessie 1933 To Be or Not to Be: A Study of Suicide. New York: Smith & Haas.
Durkheim Émile (1897) 1951 Suicide: A Study in Sociology. Glencoe, 111.: Free Press. ? First published in French.
Esquirol, Jean-ÉTIENNE D. (1838) 1845 Mental Maladies. Philadelphia: Lea & Blanchard. ? First published as Des maladies mentales considérées sous les rapports médical-hygiéniques et médico-legal.
Gibbs, Jack P.; and Martin, Walter T. 1964 Status Integration and Suicide. Eugene: Univ. of Oregon.
Goethe, Johann Wolfgang Von (1811-1833) 1908 Poetry and Truth From My Life. 2 vols. London: Bell. ? First published in German.
Guerry, AndrÉ M. 1833 Essai sur la statistique morale de la France. Paris: Crochard.
Halbwachs, Maurice 1930 Les causes du suicide. Paris: Alean.
Hendin, Herbert M. 1964 Suicide and Scandinavia: A Psychoanalytic Study of Culture and Character. New York: Grune.
Henry, Andrew F.; and Short, James F. JR. 1954 Suicide and Homicide: Some Economic, Sociological and Psychological Aspects of Aggression. Glencoe, 111.: Free Press.
Kruijt, Cornelius S. 1960 Zelfmoord: Statistich-sociologische verkenningen. Suicide: Sociological and Statistical Investigations. Assen (Netherlands): Van Gorcum. ? Contains a summary in English.
Legoyt, Alfred 1881 Le suicide ancien et moderne. Paris: Droiun.
Morselli, Enrico (1879) 1882 Suicide. New York: Appleton. ? First published in Italian.
Quetelet, Adolphe (1835) 1842 A Treatise on Man and the Development of His Faculties. 2 vols. Edinburgh: Chambers. ? First published in French.
Robinson, W. S. 1950 Ecological Correlations and the Behavior of Individuals. American Sociological Review 15:351-357.
Sacks, Harvey 1966 The Search for Help: No One to Turn to. Unpublished manuscript.
Schmid, Calvin F. 1928 Suicides in Seattle, 1914 to 1925: An Ecological and Behavioristic Study. University of Washington, Publications in the Social Sciences, Vol. 5, no. 1. Seattle: Univ. of Washington Press.
Schneider, Pierre-B. 1954 La tentative de suicide. Neuchâtel (Switzerland): Delachaux & Niestlé.
Sthahan, Samuel 1893 Suicide and Insanity. London: Sonnenschein.
SÜssmilch, Johann P. (1741) 1788 Die göttliche Ordnung in den Veränderungen des menschlichen Geschlechts, aus der Geburt, dem Tode und der Fort-pflanzung. 3 vols. Berlin: Verlag der Buchhandlung der Realschule.
Tissot, Joseph 1840 De la manie du suicide et de l’esprit de révolte: De leur causes et de leur remèdes. Paris: Lagrange.
Wagner, Adolf H. G. 1864 Die Gesetzmässigkeit in den scheinbar willkuhrlichen menschlichen Handlungen vom Standpunkte der Statistik. Part 2. Statistik willkuhrlichen Handlungen. Hamburg (Germany): Boyes & Geisler.
II. PSYCHOLOGICAL ASPECTS (1)
The clear definition and meaningful classification of various suicidal phenomena are fundamental to advancements in the treatment, prevention, and investigation of self-destruction in man. The purpose of this article is to present some new conceptions of suicidal phenomena, first by identifying current confusions and then by proposing a taxonomic scheme that attempts to embrace many of the diverse aspects of self-destructive behavior.
Definition
As a beginning, a straightforward definition of suicide might read: “Suicide is the human act of self-inflicted, self-intentioned cessation.” At least five points are to be noted in this brief definition: (1) it states that suicide is a human act; (2) it combines both the decedent’s conscious wish to be dead and his actions to carry out that wish; (3) it implies that the motivations of the deceased may have to be inferred and his behaviors interpreted by others, using such evidence as a suicide note, spoken testimony, or retrospective reconstruction of the victim’s intention; (4) it states that the goal of the action relates to death, rather than to self-injury, self-mutilation, or inimical or self-reducing behaviors; and (5) it focuses on the concept of the cessation of the individual’s conscious introspective life. (An explication of the concept of “cessation” may be found in Shneidman 1964.)
Difficulties of definition. Assuming the validity of the definition of suicide cited above—that a human being, with the intention of stopping his life, inflicts upon himself the equivalent of a mortal wound—the meaning would seem clear enough if it were stated that a certain individual had “committed suicide.” On the other hand, confusions of meaning arise immediately if a specific individual is labeled as “suicidal.” Although suicide seems to be not too difficult to define, suicidal phenomena are, in fact, very complicated. Some of the current confusions relating to the term “suicidal” may be listed as follows.
(1) The word “suicidal” is used indiscriminately to cover different categories of behavior. For example, one cannot be sure whether it is being used to convey the idea that an individual has (a) committed suicide, (b) attempted suicide, (c) threatened suicide, (d) exhibited depressive behavior with or without suicidal ideation, or (e) manifested generally self-destructive or inimical patterns. Of recent writers, Stengel and Cook (1958) especially have emphasized the importance of differentiating specifically between data on attempted suicide and data regarding individuals who have committed suicide. But the fact remains that the classification of suicidal behavior currently most common in everyday clinical and research use is a rather homely, supposedly common-sense division: in its barest form it implies that all humanity can be divided into two groupings, suicidal and nonsuicidal; and then, with seeming meaningfulness, it divides the suicidal category into subgroups of committed, attempted, and threatened. Although this elaboration is more sophisticated than the suicidal-versus-nonsuicidal view of life, it remains neither theoretically nor practically adequate for understanding or treatment.
(2) There is constant confusion in respect to the temporal dimension of suicidal acts. One sees the word “suicidal” used to convey the information that an individual was self-destructive (or manifested behavior in any of the other categories listed immediately above), is currently self-destructive, or will be so. This obviously contains confusions among statements that are postdictive (relating to the past), “paridictive” (relating to the more-or-less present), and predictive (relating to the future). Most diagnoses of individual “suicidality” are post hoc definitions in that they refer to those cases in which an individual is labeled as “suicidal” only after he has committed suicide. Statistics on suicide are, of course, based primarily on post hoc definitions of suicide. The primary difficulty in all such cases lies in determining whether or not the individual actually intended to kill himself. The case of the individual who writes a suicide note and then shoots himself is fairly clear, but many cases of death are unclear or equivocal as to mode of death. For example, in the case of an individual who “jumps or falls” from a high place or who is found dead of barbiturate poisoning, the question is often raised whether the case was suicidal or accidental. The coroner’s traditional concern has been with assessing whether God (natural and accidental deaths) or man (suicidal or homicidal deaths) is responsible. The term “suicide” as used by coroners in the certification of death is a medicolegal term and includes, as a sine qua non, the concept that the person played a major role in bringing about his own demise, that is, that it was his intention to die. It should be obvious that statistics on suicide can be greatly influenced (in any city or country) by the manner in which these equivocal deaths are labeled.
(3) There are confusions relating to the characteristics of suicidal behaviors. In many past and current investigations of self-destructive behaviors, several different dimensions of behavior have unfortunately been thrown together. (If one were, for example, studying homicidal phenomena, it is unlikely that he would fail to differentiate among homicides committed on the highway, in the bedroom, during armed robberies, on the field of battle, etc.—and yet a comparable lack of discrimination has been characteristic of many studies of suicidal phenomena.) These issues also relate to the phenomenologic and semantic confusions in the use of the word “suicidal.” Some individuals who “die by their own hand” do not necessarily “commit suicide.” That is to say, they do not, in their own minds, kill themselves or seek death. Instead of fleeing into a vaguely conceptualized “death,” they behave so as to escape from aspects of life. For some the concept of death, or final cessation of being, does not enter into their thinking. They rather indulge in either a planned or a momentary impulsive act—and termination of life is the result. Then we say that they have “committed suicide,” whereas such an individual might conceivably have left a note which states that “. . . all I did was to swallow those pills ... I just wanted some relief at that moment. . . . The tension was so great I had to do something. ... I did not know what I was doing or what would happen. ... It was a gamble. ... I was desperate.. . .” Operationally, it would seem that the key characteristics of suicidal behaviors do not lie in the differences in method (shooting, sedation, cutting, hanging) but rather in the differences in the individual’s life phase, in the lethality of his acts, and in his intention vis-à-vis death, as indicated below in (4), (5), and (6), respectively. That is, it may very well be that the confusions listed above in (1), (2), and (3) might, in large measure, be avoided by making the very distinctions suggested in the subsequent paragraphs.
(4) Many confusions arise because each individual’s attitudes toward his own death are biphasic: that is, any adolescent or adult, at any given moment, has (a) more or less long-range, pervasive, relatively habitual orientations toward his own death. These characterological orientations are an integral part of his total psychological make-up and reflect his philosophy of life, need systems, aspirations, identifications, conscious beliefs, etc. And he is also capable of having (b) relatively short-lived, acute, fairly sudden shifts of his orientations toward his own death. Indeed, this is what is usually meant when one says that an individual has “become suicidal.” It is therefore crucial in any complete assessment of an individual’s orientations toward his own death to know both his habitual and his current (today’s) orientations. Failure to make this distinction is one reason why many current efforts to relate “suicidal state” to psychological test data or to case-history data have been barren and confusing. For individuals in their “normal” (usual for them) state, their habitual and their current orientations toward their own demise will be the same; for individuals who are acutely disturbed, their current orientations toward their demise will often reflect this perturbation. “Being suicidal” involves being disturbed, although not necessarily psychotic.
(5) Popular accounts of “suicidal” behaviors often focus on the method used (for example, wrist cutting) or on the precipitating cause (ill health, losing money) without regard for one of the key dimensions in the assessment of suicidal behavior, namely lethality. The primary clinical goal of any suicide prevention agency is to keep people out of the coroner’s department. Thus, an individual’s unhappiness, perturbation, loneliness, alcoholism, schizophrenia, homosexuality, depression, etc., are relevant in suicide prevention primarily as they bear on the assessment of his lethality potential. To say that an individual has “threatened” suicide or has “attempted” suicide is relatively uninformative without some indications of the potentialities of a lethal outcome with which the threat or attempt was made. Recently, at the Suicide Prevention Center in Los Angeles specific procedures have been evolved for the rapid assessment of the lethality of an individual’s suicidal potential (see Litman & Farberow 1961; Tabach-nick & Farberow 1961).
(6) The most obfuscating confusions relating to suicidal phenomena may occur if the individual’s intentions in relation to his own cessation are not considered. “Suicide” also has an administrative definition. In the United States (and most of the countries reporting to the World Health Organization), “suicide” defined from the point of view of the coroner or vital statistician is simply one of the four modes of death, the others being natural, accidental, and homicide (N-A-S-H). This traditional fourfold classification of all deaths leaves much to be desired. Its major deficiency is that it emphasizes relatively adventitious details in the death (that is, whether the individual was invaded by a lethal virus or a lethal bullet, which may make little difference to the deceased) and, more important, it erroneously treats the human being as a Cartesian biological machine rather than appropriately treating him as a psychosocial organism, thus obscuring the individual’s intentions in relation to his own cessation. Further, the traditional N-A-S-H classification of deaths completely neglects the concepts of contemporary psychology regarding intention and purpose, the multiple determination of behavior, unconscious motivation, etc.
Much of the problem arises because of a confusion between methods and purpose. Although it is true that the act of putting a shotgun in one’s mouth and pulling the trigger with one’s toe is almost always related to lethal self-intention, this particular isomorphic relationship between method and intent does not hold for most other means, such as ingesting barbiturates or cutting oneself with a razor. Individuals can attempt to attempt suicide, attempt to commit suicide, or attempt to be nonsuicidal. Cessation intentions may range all the way from deadly ones, through the wide variety of ambivalences, rescue fantasies, cries for help, and psychic indecisions, all the way to clearly formulated nonlethal intention in which a semantic usurpation of a “suicidal” mode has been consciously employed.
Deaths classified by intention. As a way out of this impasse it is suggested that all deaths—in addition to their being labeled as natural, accidental, suicide, or homicide—also be designated as intentioned, subintentioned, or unintentioned (Shneidman 1963).
Intentioned deaths. An intentioned cessation is one in which the deceased played a direct, conscious role in his own demise. The death was due primarily to the decedent’s conscious wish to stop his conscious life and to his actions in carrying out that wish. In this category would be individuals who seek their own demise, initiate or participate in their own demise, or take calculated risks where the odds are critically unfavorable to their survival, as in Russian roulette, where the individual bets his life on the objective probability of as few as five out of six chances that he will survive.
Subintentioned deaths. Subintentioned deaths are those in which the deceased played an important indirect, covert, partial, or unconscious role in his own demise. The death is suspected to be due in some part to the actions of the decedent which seemed to reflect his unconscious wishes to hasten his death, as evidenced by his own carelessness, foolhardiness, neglect of self, imprudence, resignation to death, mismanagement of alcohol or drugs, disregard of life-saving medical regimen, brink-of-death patterns, etc. This concept of subintentioned death is similar in some ways to Karl Menninger’s concepts (1938) of chronic suicide, focal suicide, and organic suicide, except that these relate to self-defeating ways of continuing to live, whereas the notion of subintentioned death relates to ways of stopping the process of living.
In terms of the traditional N-A-S-H classification of modes of death, it is important to note that some instances of all four types can be found in the category of subintentioned deaths, depending upon the particular details of each case. There is a growing literature on the role of the individual in his own natural, accidental, suicidal, or homicidal death. Subintentioned deaths involve what might be called the psychosomatics of death, that is, cases in which essentially psychological processes—like fear (including fear of voodoo), anxiety, derring-do, hostility, withdrawal, etc.—seem to play some role in exacerbating the catabolic or physiological processes which bring on termination, as well as those cases in which the individual seems to play an indirect, largely unconscious role in inviting or hastening his own demise [see Psychosomatic Illness; see also Gengerelli & Kirkner 1954; Macdonald 1961; Weisman & Hackett 1961; Weiss 1957; Wolfgang 1959].
It can be noted that the subintentioned death implies more than ambivalence toward wanting to be dead and wishing to be rescued (the to-be-or-not-to-be inner dialogue ) found in practically every suicide. It rather reflects the active indirect (largely unconscious) participation of the individual in hastening his own demise.
Unintentioned deaths. Unintentioned deaths are those in which the deceased played no significant psychological role in his own cessation. The death was due entirely to failures within the body or to assault from without (by a bullet, a blow, a steering wheel, etc.) in a decedent who, rather unambivalently, wished to continue to live. Such a person at the time of his death is, as it were, going about his business (even though he may be lying in the hospital) with no conscious intention or strong drive in the direction of effecting or hastening his demise. What happens is that something from the outside—outside of his mind— happens to him. This lethal “something” may be a cerebral vascular accident, a myocardial infarction, some malfunction, some catastrophic catabolism, some invasion—whether by bullet or by virus—which he did not, in any part, himself generate. “It” happens to “him.”
Practical applications. A practical application of the above schematization of death in terms of the individual’s intention has been made since 1960 in Los Angeles. This procedure, labeled the “psychological autopsy” (Curphey 1961; Litman et al. 1963), has been used in those cases that are equivocal as to mode of death. The procedure consists of the use of especially adapted interview techniques to generate psychological information about the deceased. Survivors, friends, and professional acquaintances of the deceased are interviewed. Clues are sought, especially in relation to the prodromal aspects of suicide. A judgment is made along traditional lines (for example, probable suicide, probable accident, probable natural cause, suicide-accident undetermined, suicide-natural undetermined, etc.). An additional judgment is also made in terms of the intentioned, subintentioned, and unintentioned categories. Although one would, a priori, expect all suicidal deaths to be intentioned and all natural, accidental, and homicidal deaths to be unintentioned, the findings in the Los Angeles procedures indicate that the relationship between traditional modes of death and types of intention is a complicated one, with the crucial role occupied by subintentioned deaths.
It might be protested, inasmuch as the assessments of these intentioned states involve the appraisal of unconscious factors, that some workers (especially lay coroners) cannot legitimately be expected to make the psychological judgments required for this type of classification. To this, one answer would be that medical examiners and lay coroners make judgments of this nature every day of the week. The fact is that in the situation of evaluating a possible suicide, the coroner often acts (sometimes without realizing it) as psychiatrist and psychologist and as both judge and jury in a quasi-judicial way. This is because certification of death as suicide does, in itself, imply some judgments of reconstruction of the victim’s motivation or intention. Making these judgments is an inexorable part of a coroner’s function. It might be much better for these psychological dimensions to be made explicit and for an attempt, albeit crude, to be made to use them, rather than for these psychological dimensions to operate on an unverbalized level.
Theoretical and taxonomic positions
We now turn from our consideration of the definitional problems of suicide to a brief historical resume of the major theoretical and taxonomic positions in relation to suicidal phenomena.
Durkheim and other sociological positions . In this century (or at least since 1897, the date of the publication of Durkheim’s Le suicide) there have been two major approaches to the definition and understanding of suicide: the sociological and the psychological. The former historically is identified primarily with Émile Durkheim, French sociologist (1897). Durkheim was interested not so much in suicidal phenomena per se as he was in the explication of his own sociological method. He used the analysis of suicidal phenomena as the occasion to work out four types of factors in suicide—altruistic, egoistic, anomic, and fatalistic—but discussed only the first three.
There are studies generally classified as “sociological” (in the sense that they present statistical or ecological data, but not in the sense that they follow Durkheim’s interest in explaining the variety of man’s moral commitments to his society) which are worthy of note, among them Sainsbury (1955) and Stengel and Cook (1958) in England, and Dublin (1963) in the United States. Dublin’s book, especially, furnishes the student with an encyclopedia of statistical information about suicide phenomena.
Freudian and other psychological positions . The psychological approach is identified primarily with Freud, who is generally acknowledged to have first stated comprehensive psychological insights into suicide. Freud’s conceptualization of suicide (1917a; 1917b) was that of a primarily intra-psychic phenomenon, stemming from within the mind, primarily the unconscious mind, of the individual. In one of his formulations Freud envisaged suicide as being the result of a process wherein feelings of love and affection which had originally been directed toward an internalized love object had become, as a result of rejection and frustration, angry, hostile feelings; however, because the object had become internalized and part of the self, the hostile feelings were then directed toward the self. Thus suicide, from a psychoanalytic point of view, might be described as murder in the 180th degree.
The classical Freudian approach not only tended rather systematically to ignore social factors but also tended to focus on a single complex or psychodynamic constellation. But we now know that individuals kill themselves for a number and variety of psychologically felt motives: not only hate and revenge, but also dependency, shame, guilt, fear, hopelessness, loyalty, fealty to self-image, pain, and even ennui. Just as no single formula or pattern can be found to explain all human homosexuality or prostitution, so no single psychological pattern is sufficient to contain all human self-destruction. (For a comprehensive review of this topic, see Litman 1967.)
Synthesis—the concept of self. The synthesis of the psychological position, with its clinical emphasis on the individual internal drama within the single mind, and the sociological position, discussed in the previous article, remains to be accomplished. A recent study bearing on this point (Shneidman & Farberow 1960) emphasized the interplay between both the social and psychological factors as mutually enhancing roles in each individual’s suicide. This finding is consistent with that of Halbwachs (1930), whose position—unlike that of his mentor, Durkheim—was that the “social” and “psychopathological” explanations of suicide are complementary rather than antithetical. A synthesis of these two lies in the area of the “self,” especially in the ways in which social forces are incorporated within the totality of the individual. In understanding suicide, one needs to know the thoughts and feelings and ego functionings and unconscious conflicts of an individual, as well as how he integrates with his fellow man and participates morally as a member of the groups within which he lives.
Taxonomic positions . Not many classifications of suicidal phenomena have been proposed. Durk-heim’s classification distinguished between anomic, egoistic, and altruistic suicide (1897). To this can be added Menninger’s classification of the sources of suicidal impulses, namely the wish to kill, the wish to be killed, and the wish to die (1938). Menninger also classified subsuicidal phenomena into chronic suicide (asceticism, martyrdom, addiction, invalidism, psychosis); focal suicide (self-mutilation, malingering, multiple accidents, impotence and frigidity); and organic suicide (involving the psychological factors in organic disease). A composite listing of other rubrics would include the following: suicide as communication; suicide as revenge; suicide as fantasy crime; suicide as unconscious flight; suicide as magical revival or reunion; suicide as rebirth and restitution. A classification of suicidal types in terms of cognitive or logical styles has been proposed (Shneidman 1961). This classification, which stemmed from the logical analysis of suicide notes, divides the thinking styles of suicides into three types: logical, catalogical, and paleological.
Thus, in general, we see that the term “suicidal” is a broad one, starting from a base line of the individual who consciously and advertently takes his own life, through those individuals who by virtue of their unconscious mechanisms hasten their demise, through those individuals who indulge in partial or focal or chronic suicide, and perhaps even to those many individuals who by their own daily inimical acts truncate and diminish the full scope of their potential self-actualizations.
A concluding note concerning the current professional status of suicide prevention: Since 1955 there has been a marked spurt in interest in suicide prevention activities. In the United States, in 1958, there were 3 suicide prevention centers; in 1960, there were 5; in 1964, 9; in 1965, 15; and in 1967, 40. A Center for Studies of Suicide Prevention was established within the National Institute of Mental Health (NIMH) in 1966. A new multidisciplinary profession, suicidology—the scientific and humane study of human self-destruction—has come into being; special training courses in suicidology are offered at Johns Hopkins University.
Edwin S. Shneidman
[See alsoDeath. Other relevant material may be found in Psychoanalysis; Self concept; Medical care; and in the biography of Durkheim.]
BIBLIOGRAPHY
The bibliography for this article is combined with the bibliography of the article that follows.
III. PSYCHOLOGICAL ASPECTS (2)
The continuum of suicidal activity comprises total self-destruction (death); self-injury (but non-lethal), including crippling, maiming, painful and nonpainful activities; threats and other verbalized indications of intention toward self-destruction or self-injury; feelings of despair, depression, and unhappiness (which may not include thoughts of self-injury, but frequently do); and thoughts of separation, departure, absence, and relief and release. Somewhere within this continuum the thought or impulse becomes translated into action. The clinical impression is that once the psychological defenses against suicidal activity have been breached and the action has occurred, the possibility for further future acting out when emotional tension and strain recur is facilitated. It seems likely also that increasingly serious behavior results, possibly because of feelings of guilt or of feelings that such behavior is necessary in order to communicate with equivalent impact. The factor that determines the behavior remains puzzling. Perhaps the controls developed through coping mechanisms and defense patterns, brought to bear at different levels of dysfunctioning (Menninger 1963), play a crucial role, and the intensive study of these might provide the understanding necessary for suicide prevention for different individuals in various suicidal situations.
Characteristics of the suicidal person . The identification of the suicidal person is not especially difficult once the process has begun. Most persons considering self-destruction will identify themselves by communicating this tendency either behaviorally or verbally long before any specific act occurs. The typical suicidal person will generally reveal all or most of the following characteristics: (1) ambivalence—the desire, either conscious or unconscious or both, to live and to die, present at the same time; (2) feelings of hopelessness and helplessness, futility, and inadequacy to handle problems; (3) feelings of either physical or psychological exhaustion, or both; (4) marked feelings of unrelieved anxiety or tension, depression, anger, and/or guilt; (5) feelings of chaos and disorganization with inability to restore order; (6) mood swings, for example, from agitation to apathy or withdrawal; (7) cognitive constriction, inability to see alternatives, limitation of potentialities; (8) loss of interest in usual activities, such as sex, hobbies, and work; (9) physical distress, such as insomnia, anorexia, psychasthenia, and psychosomatic symptoms.
Prediction of lethal behavior . The more difficult problem is the evaluation of the suicidal person in terms of the relatively immediate potentiality of lethal acting out. The suicidal crisis presents the professional person with the need for quick appraisal. A schedule, evolved from experiences in the Suicide Prevention Center in Los Angeles, lists the following criteria used in evaluation of the emergency situation (see Litman & Farberow 1961). (1) Age and sex: older white males generally have the highest suicidal potentiality; a young female, on the other hand, is usually less lethally suicidal. (2) Suicidal plans: specificity about time, place, and method, plus the means for carrying out a plan, indicate high suicidal danger; vague, diffuse talk by a nonpsychotic about dying indicates that the situation is less serious. (3) Resources: external sources of support and interest, such as family, relatives, friends, physician, or hospital, are helpful; when the patient’s resources are exhausted the suicidal potentiality rises. (4) Prior suicidal behavior: a past history of suicidal behavior indicates greater present danger, and the seriousness of the prior suicide attempt adds an additional important consideration. (5) Onset of suicidal behavior: the acute suicidal crisis may be more immediately serious but also more amenable to intervention. The chronically suicidal person, especially the alcoholic or borderline schizophrenic, who presents repeated feelings of depression, is the more serious long-term therapeutic problem. (6) The medical situation: many patients visit physicians with minor physical complaints which are in reality indications of severe depression. Studies (Motto & Greene 1958; Robins et al. 1959; Dorpat & Ripley I960; and Shneidman & Farberow 1961) have indicated that more than 50 per cent of the patients who had committed suicide had seen their physicians within three months prior to their deaths. (7) Loss of a loved one: where death, separation, quarrel, or divorce from someone close has recently occurred, the suicidal potentiality is increased. (8) Communication: if communication still exists between the patient and others the suicidal risk is lowered; when the communication breaks and the person withdraws the danger increases. (9) Reaction of the referring person: if interest and concern about the patient continue, this is helpful. If, however, the referring person is angry, rejecting, and attempting to rid himself of responsibility, the potentiality is increased.
Psychological theories
The psychological theories have been summarized by Jackson (1957), who divided them roughly into nonpsychoanalytic and psychoanalytic.
Nonpsychoanalytic formulations . The nonpsychoanalytic formulations refer to “exhaustion” causing restriction of the field of consciousness so that an “organic depression” occurs; failure of adaptation; a disturbance of balance of will to live from a dynamic fixation of infantile attachment; infantile protest and hostility against harsh, restraining figures; narcissism in a rigid personality; compensations for homicidal impulses against members of the immediate family; and spite in children.
Psychoanalytic formulations . Most of the psychoanalytic theories stern from two of Freud’s theoretical contributions: his elaboration of the dynamics of depression in Mourning and Melancholia (1917b) and his postulations of the death instinct in Beyond the Pleasure Principle (1920). Depression, and consequently suicide, occurs as a result of strong, aggressive urges directed against an introjected object formerly loved but now hated. Menninger (1938) adopts the concept of a death instinct and elaborates it by postulating three elements in suicide: a wish to kill, a wish to be killed, and a wish to die. Zilboorg (1936; 1937) considers suicide a way of thwarting outside forces and one method of gaining immortality. O’Connor (1948) adds the feeling that the person achieves omnipotence by a return to early power narcissism. Palmer (1941) suggests that arrested psychosexual development as a result of the unavailability of important figures at crucial stages is the basic mechanism. Garma (1943) stresses the loss of an important love object, suicide being used to recover it. Bergler (1946) describes the introjec-tion suicide, aggression against guilt feelings; hysteric suicide, unconscious dramatization of how one does not want to be treated; and the miscellaneous type, such as paranoid schizophrenics reacting to voices. Farberow (1961), in the collection The Cry for Help, summarizes his contributors’ several theoretical approaches to suicide. To the formulations already presented above, some of his contributors add frustrated dependency; longing for spiritual rebirth and seeking to re-establish contact with the self by destruction of the ego; strong inferiority feelings and veiled aggression in dependent individuals with “pampered life style”; a depressive, hateful type of personality structure developed from interpersonal experiences; alienation and feelings of disparity between idealized self and real self; and the person’s attempts to validate his self according to his own “constructions.”
Contributing factors
From the variety of theories and multitude of factors in suicide, it is apparent that suicidal phenomena are both widespread and complex. They reflect common sociological roots and influences and, at the same time, express singular personal experiences and impulses. A psychology of suicide must take as many such factors as possible into account if understanding is to be attained. A single schema encompassing the major factors which enter into the understanding of any suicidal event can, as yet, be only a desirable goal. The following factors are necessary considerations in any comprehensive overview of suicidal behavior.
Sociological background . It is obvious that any event needs to be viewed in the setting within which it occurs, but suicide especially has varying significance when it takes place in such widely different countries as, for example, Denmark, Italy, Japan, or the southern part of the United States. National differences are further compounded by racial, religious, and economic factors. Similarities and specific differences will be found. Studies of suicide in some of the tribes of east Africa (Bohannan 1960) and in aboriginal tribes of central India (Elwin 1943) have shown motives remarkably similar to those of Western cultures, for example, domestic strife or loss of social status, as well as features specific to these tribes, such as intervention by the gods or bewitchment by ghosts.
Fluctuations of economic status or the changing political scene influence suicide. For example, suicide has varied with economic depressions, as when the rate soared in the United States during the depression of the mid-1930s and dropped markedly during the war years of the 1940s. An investigation by Arkun (1963) of the suicide rates in Turkey during the periods from 1927 to 1946 (after sweeping social reforms) and 1950 to 1960 (after World War n) showed startling changes in rates which, during the earlier period, could be attributed to the upheavals in the culture of Turkey and especially to the change in the role of women because of the reforms of Ataturk. Japan’s suicide rate has always been high; but whereas in ancient times the traditions of hara-kiri and seppuku were prominent, today much of the suicide rate is contributed to by a younger age group, such as students in universities who are faced with failure or fear of it (Iga 1961).
Cultural background . Cultures often surround death with taboos and rituals which illustrate feelings about death and dying and which include attitudes toward suicide. Myths and folklore illustrate some of the attitudes, as in the history of the Vikings and the tales of Valhalla, and contribute to the condemnation or condoning of suicide.
The culture may also determine interpersonal relationships that influence the occurrence of suicide. Hendin (1964) examines the “Scandinavian suicide phenomenon” and arrives at an explanation for the high suicide rates in Denmark and Sweden in contrast to those in Norway by a determination of the “psychosocial characteristics.” Using psychoanalytic methods, he finds differences in the psychodynamic constellations of the three countries, such as in dependency aspects, attitudes toward performance and accomplishment, handling of aggression and guilt feelings, relationships between the sexes, methods of discipline, and other dynamic features.
Individual demographic characteristics . Epidemiological aspects of suicide, such as age, sex, nationality, race, religion, marital status, education, financial status, have all been studied exhaustively for various countries (Dublin 1963). Where the individual falls in respect to each of these provides immediately invaluable information about the suicidal person. Important to include here also is the physical and mental status of the individual. A chronic, debilitating, or possibly fatal illness such as emphysema, cardiac disease, or cancer, or a recurrent mental illness which hospitalizes the individual for several months every two or three years will obviously influence the individual suicide (Farberow et al. 1963; Shneidman et al. 1962).
Psychological factors . It is within the psychological factors that the core of the problem of understanding suicide is met. These factors include not only the current personality status and psycho-dynamic constellations of the individual in question but also the motivations for his suicidal behavior, the reasons why his actions, thoughts, or feelings lead him to suicide. The categorization of motivations into interpersonal and intrapersonal factors seems to offer a meaningful classification of many of the various phenomena. The distinction is arbitrary, of course, for it is practically impossible for a suicide to occur without both types of relationships being involved. Nevertheless, one or the other aspect will often predominate.
Interpersonal motivations in suicide occur when the suicidal person attempts by his behavior either to bring about an action on the part of another person or persons or to effect a change in attitude or feeling within another person or persons, or both. The suicidal behavior can thus be seen as a means to influence, persuade, force, manipulate, stimulate, change, dominate, reinstate, etc., feelings or behavior in someone else. The other person is most often someone who has been in a close relationship, such as spouse, fiancee, or member of the family. Infrequently, the object of the behavior is more generalized, and it may be society itself.
Interpersonal motivations can be found, of course, in all ages but are usually predominant in the younger and middle-aged groups. A typical example is that of the girl between the ages of 20 and 25 who is reacting with strong feelings of rejection to a quarrel with a loved one or to divorce or separation. Her emotional state is one of agitation, dependency, immaturity, poor judgment, and impulsivity. Her suicidal behavior is used to express anger or feelings of rejection and to force a change in the rebuffing person or to arouse guilt feelings in him. Much of the behavior is still verbal, some is impulsive acting out, and most of it contains an “appeal” element (Stengel & Cook 1958).
Less often the aim is the expiation of or the need to express the guilt the person feels for having done something either imaginary or real in the relationship with another person. Ambivalence about dying is relatively low, inasmuch as the person, despite the fact that he is engaging in suicidal behavior, does not usually wish to die.
Intrapersonal motivations appear most often in older persons and thus in situations in which ties with others have dissipated. The individual’s action seems aimed primarily at expressing the pressures and stresses from within and at fulfilling important needs in himself. The typical person is a male aged 60 or over who has recently suffered the death of a loved one, whose physical condition has deteriorated so that there is illness or pain, or whose children are married and so live their own separate lives. There are intense feelings of loneliness, feelings of not being needed any longer, of no longer being able to work effectively, perhaps because of physical condition, or of feeling that life has been lived and holds no more. The mood is often depressed, withdrawn, and physically and emotionally exhausted. There may be strong need for expiation and for atonement stemming from excessive feelings of guilt. An important dynamic is the need to maintain “psychological integrity” (Appelbaum 1963) or self-esteem or self-concept, even by the paradoxical act of self-destruction. Ambivalence is again low inasmuch as the person, if he embarks on a suicidal course, usually does so with full intent to die. [See Aging, article onpsychological aspects.]
Some cases of suicide occur in which it is difficult to distinguish whether interpersonal or intra-personal motivations are predominant. Rather it seems that each is equally present, although perhaps not always in the same strength at the same time. A typical example is a middle-aged person, the precipitating suicidal stimulus is the death of a loved one, separation, divorce, loss of job, loss of status, or sometimes a crippling, debilitating illness. Such a person is generally depressed, anxious, frustrated, and sometimes agitated, showing poor judgment and disorganization which will sometimes extend to psychotic or near-psychotic proportions. Marked symptoms of frustrated dependency, hostility, and aggression, perhaps because of rejection or masochism, and the two elements “appeal” and “ordeal” are readily seen (Stengel 1956). The ordeal element is especially apparent in the greater ambivalence about dying and the marked tendency to leave survival up to fate, destiny, or chance. Suicidal attempts are usually more lethal, but there are also more provisions for rescue, both conscious and unconscious.
Importance of work. Work takes on special significance for this group. Often with premature dissolution of relationships in the middle-aged group, work becomes the principal source of self-significance and self-esteem. The nonpersonal aspects of the work itself, rather than the people involved with whom interaction on the job occurs, become important. So long as the person is able to function in his job and to lose himself in its details, there is sufficient defense against suicidal impulses. For the very severely and chronically depressed person, work may provide a cover for the feeling of emptiness and void from which he is continually trying to escape. The routine of work keeps him busy and prevents him from thinking about himself. Not to work provides him with time during which he is free to think about himself and to feel useless and empty. Once the work is lost, perhaps through some personal difficulty, physical crisis, or enforced retirement, a crucial defense seems to be breached and suicidal impulses will burst through.
Can any suicide be entirely intrapersonal? One suggestion has been that this may occur only in the psychotic. However, the problem may well be only one of understanding on the part of the observer. The psychotic may be reacting entirely on an interpersonal basis but in a bizarre or devious process which simply is not comprehended by others.
Feelings about death and the afterlife. Most often the individual simply reflects the prevailing attitudes of his culture about suicide, death, or life after death. However, the individual may arrive at his own conceptualizations, which may vary markedly. Convictions of eternal peace after death, of the possibility of reunion with a deceased loved one, visions of hell-fire and brimstone, or of pain and unmitigated suffering, belief in the supernatural, or faith in magic may be key factors in an individual’s suicidal behavior.
Communication . Seen as a communication process, suicidal behavior often achieves clearer perspective. In most instances, the suicidal activity occurs at the end of a long train of events that have finally led the person to the decision that life is no longer worth living. Accompanied by many communications along this course, the suicidal act itself then becomes a communication which may have many meanings and much significance. The communication in suicidal behavior can be grouped under five headings. (1) Form: the communication may be verbal, including written, or nonverbal and behavioral. (2) Directness or indirectness: the communication may be straightforward and clear, or disguised and indirect. Withdrawal, giving away prized possessions, remarks about not needing articles, fantasies of death, burial, or rescue from dangerous situations may occur. (3) Substance or content: the communication may contain expressions of affect, either fixing or expiating guilt or blame, explanations of the suicidal act, or instructions and directions to survivors, as in wills and suicide notes (Shneidman & Farberow 1957a; Tuckman et al. 1959). (4) Object of the communication: in most interpersonal situations, the communication is directed to a specific person or persons; in intrapersonal motivations it is more often directed to society in general. (5) Purpose: the communication may be overt or indirect in aim. Sometimes it is a cry for help, a plea to be stopped or to be rescued, a means for expression of hostility and hate, a final fixing of blame, a way to cause shame or arouse guilt, or a way to assume blame, absolve others, and expiate one’s own guilt.
Countersuicidal controls
The factors that mitigate against suicide are as important as those that influence toward it. In many instances, the fact that self-destruction is not chosen as the way out of seemingly unbearable situations—as in the concentration camps of Germany, or when the individual is subject to continuous pain and discomfort in the terminal stages of cancer—impresses the observer. The controls may be external or internal.
External controls . External controls refer to all the controls that society may bring to bear on an individual to keep him conforming and alive, such as taboos, religion, myths, mores, group and subgroup identifications, marriage, family, children. Also significant are the actions of others toward the individual. Indications of support, understanding, interest, and concern, especially on the part of the “significant other” but also by hospital and professional personnel, may be the essential preventive factor.
Internal controls . Internal controls may stem from the ideals, standards, morality, conscience, or feelings of responsibility of the individual. In addition, the ego structure of the individual may provide him with flexibility, adaptability, independence, and feelings of self-esteem which will permit him to endure severe emotional stress. Or he may be more vulnerable because of rigidity, overdependence, and poor self-concept.
Psychology of the survivors
The suicide has great impact on the survivors, regardless of whether it had predominantly interpersonal or intrapersonal motives. These reactions generally vary directly in intensity with distance of relationship with the suicidal person. Among the group in a close relationship, the spouse, children, family, relatives, close friends, or therapist, a variety of feelings and reactions may be aroused. These may include (1) strong feelings of loss, accompanied by sorrow and mourning; (2) strong feelings of anger for (a) being made to feel responsible, or (b) being rejected in that what was offered was refused; (3) guilt, shame, or embarrassment with feelings of responsibility for the death; (4) feelings of failure or inadequacy that what was needed could not be supplied; (5) feelings of relief that the nagging, insistent demands have ceased; (6) feelings of having been deserted, especially true for children; (7) ambivalence, with a mixture of all the above; (8) reactions of doubt and self-questioning whether enough was attempted; (9) denial that a suicide has occurred, with a possible conspiracy of silence among all concerned; and (10) arousal of one’s own impulses toward suicide.
Among those in a more distant relationship, such as neighbors, employer or fellow employees, the hospital, or society, the reactions may also be those of (1) anger because of (a) a feeling that the suicidal person has rejected his social and moral responsibilities, or (b) being made to feel responsible, or (c) an implied accusation of not enough concern, interest, or caring about its members and fellow man; (2) guilt that not enough was offered to make the person want to live; (3) rejection, resulting from the suicidal person’s obvious choice to do without them; and (4) uneasiness, manifested by a vague need for self-examination to determine what was wrong or to rationalize the discomfort away.
Norman L. Farberow
[See alsoDeath. Other relevant material may be found in Depressive disorders; Medical care; and in the biography of Durkheim.]
BIBLIOGRAPHY
Appelbaum, Stephen A. 1963 The Problem-solving Aspect of Suicide. Journal of Protective Techniques and Personality Assessment 27:259-268.
Arkun, Nezahat 1963 Intiharin psikodinamikleri. Istanbul (Turkey): Baha Matbaasi.
Bergler, Edmund 1946 Problems of Suicide. Psychiatric Quarterly (Supplement) 20:261-275.
Bohannan, Paul (editor) 1960 African Homicide and Suicide. Princeton Univ. Press.
Curphey, Theodore J. 1961 The Role of the Social Scientist in the Medicolegal Certification of Death from Suicide. Pages 110-117 in Norman L. Farberow and Edwin S. Shneidman (editors), The Cry for Help. New York: McGraw-Hill.
Dorpat, Theodobe L.; and Ripley, Hebbert S. 1960 A Study of Suicide in the Seattle Area. Comprehensive Psychiatry 1:349-359.
Dublin, Louis I. 1963 Suicide: A Sociological and Statistical Study. New York: Ronald.
Durkheim, ÉMILE (1897) 1951 Suicide: A Study in Sociology. Glencoe, 111.: Free Press. ? First published in French.
Elwin, Verrier 1943 Maria Murder and Suicide. Oxford Univ. Press.
Farberow, Norman L. 1961 Summary. Pages 290-321 in Norman L. Farberow and Edwin S. Shneidman (editors), The Cry for Help. New York: McGraw-Hill.
Farberow, Norman L.; and Shneidman, Edwin S. (editors) 1961 The Cry for Help. New York: McGraw-Hill. ? Contains a bibliography on suicide, 1897-1957, on pages 325-388.
Farberow, Norman L.; SHNEIDMAN, E. S.; and Leonard, C. V. 1963 Suicide Among General Medical and Surgical Hospital Patients with Malignant Neoplasms. U.S. Veterans Administration, Medical Bulletin No. 9. Washington: Veterans Administration.
Farberow, Norman L. et al. 1966 Suicide Among Patients With Cardiorespiratory Illnesses. Journal of the American Medical Association 195:422-428.
Freud, Sigmund (1917a) 1959 The Psychogenesis of a Case of Homosexuality in a Woman. Volume 2, pages 202-231 in Sigmund Freud, Collected Papers. International Psycho-analytic Library, No. 10. London: Hogarth; New York: Basic Books.
Freud, Sigmund (1917b) 1959 Mourning and Melancholia. Volume 4, pages 152-170 in Sigmund Freud, Collected Papers. International Psycho-analytic Library, No. 10. London: Hogarth; New York: Basic Books. ? Authorized translation from the German under the supervision of Joan Riviere. The first English edition was published in 1925.
Freud, Sigmund (1920) 1950 Beyond the Pleasure Principle. Authorized translation from the 2d ed., by C. J. M. Hubback. International Psycho-analytic Library, No. 4. New York: Liveright. ? First published as Jenseits des Lustprinzips. A paperback edition, translated by James Strachey, was published in 1959 by Bantam Books.
Gahma, Angel (1943) 1944 Sadism and Masochism in Human Conduct: Part 2. Journal of Clinical Psycho-pathology and Psychotherapy 6:355-390. ? First published in Spanish.
Gengerelli, Joseph A.; and Kirkner, Frank J. (editors) 1954 The Psychological Variables in Human Cancer, A symposium presented at the Veterans Administration Hospital, Long Beach, California, October 23, 1953. Berkeley: Univ. of California Press.
Halbwachs, Maurice 1930 Les causes du suicide. Paris: Alean.
Hendin, Herbert M. 1964 Suicide and Scandinavia: A Psychoanalytic Study of Culture and Character. New York: Gruñe.
Iga, Mamoru 1961 Cultural Factors in Suicide of Japanese Youth With Focus on Personality. Sociology and Social Research 46:75-90.
Jackson, Don D. 1957 Theories of Suicide. Pages 17-21 in Edwin S. Shneidman and Norman L. Farberow (editors), Clues to Suicide. New York: McGraw-Hill.
Kubie, Lawrence S. 1964 Multiple Determinants of Suicidal Efforts. Journal of Nervous and Mental Disease 138:3-8.
Litman, Robert E. 1967 Sigmund Freud on Suicide. Unpublished manuscript.
Litman, Robert E.; and Farberow, Norman L. 1961 Emergency Evaluation of Self-destructive Potentiality. Pages 48-59 in Norman L. Farberow and Edwin S. Shneidman (editors), The Cry for Help. New York: McGraw-Hill.
Litman, Robert E. et al. 1963 Investigations of Equivocal Suicides. Journal of the American Medical Association 184:924-929.
Macdonald, John M. 1961 The Murderer and His Victim. Springfield, 111.: Thomas.
Menninger, Karl A. 1938 Man Against Himself. New York: Harcourt.
Menninger, Karl A. 1963 The Vital Balance: The Life Process in Mental Health and Illness. New York: Viking.
Motto, Jerome A.; and Greene, Clara 1958 Suicide and the Medical Community. Archives of Neurology and Psychiatry 80:776-781.
National Institute OF Mental HealthBulletin of Suicidology. ? Published since 1967.
O’Connor, William A. 1948 Some Notes on Suicide. British Journal of Medical Psychology 21:222-228.
Palmer, D. M. 1941 Factors in Suicide Attempts: Review of 25 Consecutive Cases. Journal of Nervous and Mental Disease 93:421-442.
Ridout, Aileen B. 1962 Suicide as a Factor in Public Health. Journal of the Royal Institute of Public Health and Hygiene 25:115-128.
Robins, Eli et al. 1959 The Communication of Suicidal Intent: A Study of 134 Consecutive Cases of Successful (Completed) Suicide. American Journal of Psychiatry 115:724-773.
Sainsbury, Peter 1955 Suicide in London: An Ecological Study. London: Chapman. ? A paperback edition was published in 1956 by Basic Books.
Shneidman, Edwin S. 1961 Psycho-logic: A Personality Approach to Patterns of Thinking. Pages 153-190 in Conference on Contemporary Issues in Thematic Apperceptive Methods, Fels Research Institute, 1959, Contemporary Issues in Thematic Apperceptive Methods. Edited by Jerome Kagan and Gerald S. Lesser. Springfield, 111.: Thomas.
Shneidman, Edwin S. 1963 Orientations Toward Death: A Vital Aspect of the Study of Lives. Pages 200-227 in Robert W. White (editor), The Study of Lives: Essays on Personality in Honor of Henry A. Murray. New York: Atherton.
Shneidman, Edwin S. 1964 Suicide, Sleep, and Death: Some Possible Interrelations Among Cessation, Interruption, and Continuation Phenomena. Journal of Consulting Psychology 28:95-106.
Shneidman, Edwin S. (editor) 1967 Essays in Self Destruction. New York: International Science Press.
Shneidman, Edwin S.; and Farberow, Norman L. (editors) 1957a Clues to Suicide. New York: McGraw-Hill.
Shneidman, Edwin S.; and Farberow, Norman L. 1957b Some Comparisons Between Genuine and Simulated Suicide Notes in Terms of Mowrer’s Concepts of Discomfort and Relief. Journal of General Psychology 56:251-256.
Shneidman, Edwin S.; and Farberow, Norman L. 1960 A Socio-psychological Investigation of Suicide. Pages 270-293 in Henry P. David and J. C. Brengelmann (editors), Perspectives in Personality Research. New York: Springer.
Shneidman, Edwin S.; and Farberow, Norman L. 1961 Statistical Comparisons Between Attempted and Committed Suicides. Pages 19-47 in Norman L. Farberow and Edwin S. Shneidman (editors), The Cry for Help. New York: McGraw-Hill.
Shneidman, Edwin S.; Farberow, Norman L.; and Leonard, C. V. 1962 Suicide: Evaluation and Treatment of Suicidal Risk Among Schizophrenic Patients in Psychiatric Hospitals. U.S. Veterans Administration, Medical Bulletin No. 8. Washington: Veterans Administration.
Stengel, Erwin 1956 The Social Effects of Attempted Suicide. Journal of the Canadian Medical Association 74:116-120.
Stengel, Erwin; and Cook, Nancy G. 1958 Attempted Suicide: Its Social Significance and Effects. London: Chapman.
Tabachnick, Norman D.; and Farberow, Norman L. 1961 The Assessment of Self-destructive Potentiality. Pages 60-77 in Norman L. Farberow and Edwin S. Shneidman (editors), The Cry for Help. New York: McGraw-Hill.
Tuckman, Jacob; Kleiner, R. J.; and Lavell, M. 1959 Emotional Content of Suicide Notes. American Journal of Psychiatry 116:59-63.
Weisman, Avery D.; and Hackett, Thomas P. 1961 Predilection to Death: Death and Dying as a Psychiatric Problem. Psychosomatic Medicine 23:232-256.
Weiss, James M. A. 1957 The Gamble With Death in Attempted Suicide. Psychiatry 20:17-25.
Westermabck, Edward A. (1906-1908) 1924-1926 The Origin and Development of the Moral Ideas. 2d ed. 2 vols. London: Macmillan.
Wolfgang, Marvin E. 1959 Suicide by Means of Victim-precipitated Homicide. Journal of Clinical and Experimental Psychopathology and Quarterly Review of Psychiatry and Neurology 20:335-349.
Zilboorg, Gregory 1936 Suicide Among Civilized and Primitive Races. American Journal of Psychiatry 92: 1347-1369.
Zilboorg, Gregory 1937 Considerations in Suicide, With Particular Reference to That of the Young. American Journal of Orthopsychiatry 7:15-31.
