European Journal of Sociology

http://journals.cambridge.org/EUR

Additional services for European Journal

of Sociology:

Email alerts: Click here

Subscriptions: Click here
Commercial reprints: Click here
Terms of use : Click here

The sociological analysis of social

meanings of suicide

Jack D. Douglas

European Journal of Sociology / Volume 7 / Issue 02 / November 1966, pp 249 -

275
DOI: 10.1017/S0003975600001430, Published online: 28 July 2009

Link to this article: http://journals.cambridge.org/

abstract_S0003975600001430

How to cite this article:

Jack D. Douglas (1966). The sociological analysis of social meanings of
suicide. European Journal of Sociology, 7, pp 249-275 doi:10.1017/
S0003975600001430

Request Permissions : Click here

Downloaded from http://journals.cambridge.org/EUR, IP address: 130.126.162.126 on 16 Mar 2015

ON SUICIDE
 JACK D. DOUGLAS

The sociological analysis

of social meanings of suicide

[There are] two kinds of knowledge, which we may call for convenience
dynamical and statistical. The statistical method of investigating social questions
has Laplace for its most scientific and Buckle for its most popular expounder.
Persons are grouped according to some characteristic, and the number of persons
forming the group is set down under that characteristic. This is the raw
material from which the statist endeavours to deduce general theorems in
sociology. Other students of human nature proceed on a different plan. They
observe individual men, ascertain their history, analyse their motives, and compare
their expectation of what they will do with their actual conduct. This may be
called the dynamical method of study as applied to man. However imperfect
the dynamical method of study of man may be in practice, it evidently is
the only perfect method in principle, and its short-comings arise from the limit-
ation of our powers rather than from a faulty method of procedure. If we betake
ourselves to the statistical method, we do so confessing that we are unable to
follow the details of each individual case, and expecting that the effects of wide-
spread causes, though very different in each individual, will produce an average
result on the whole nation, from a study of which we may estimate the character
and propensities of an imaginary being called the Mean Man.
Clerk MAXWELL, Science and Freezvill.

The history of ideas demonstrates conclusively that certain ideas

can become so pervasive and central to the thought of a culture that
over many centuries the members of that culture unquestioningly apply
these ideas in many different ways to new fields of experience (i).
Such ideas are what we shall call metaphysical ideas. Such metaphys-
ical ideas normally form the ground for common-sense discourse. The
history of ideas has shown that they also form the ground for, and
frequently constitute much of the substance of, serious intellectual
works. Though science in the western world was born and developed
partly as an explicit revolution against all such unexamined, "un-
empirical" ideas, recent work in the history of science has led to the
conclusion that scientific thought is largely the result of and partly
constitued by just such metaphysical ideas (2). Moreover, more recent

(1) Lovejoy's classic study of the (2) E.A. BURTH'S classical work was
Great Chain of Being idea in Western one of the first to clearly apply this gener-
culture is one such conclusive demon- al idea to science. See his The Metaphys-
stration. ical Foundations of Modern Science,
(New York, Doubleday, 1954).

249
ArcHiv. europ. sociol., VII (1966), 249-275.
 JACK D. DOUGLAS

work in the history of science has led to the conclusion that once
scientific ideas have been accepted by the members of a scientific
discipline, these ideas in turn come to form the unexamined ground
and substance of the normal scientific works within that discipline.
Though these ideas thus have far more in common with common-sense
and humanistic discourse than most scientists would ever care to admit,
there are some important differences which are taken into consideration
by giving the established, unexamined ideas of sciences a different
name—that of paradigmatic ideas (3).
Though sociological thought is still to some extent in a non para-
digmatic stage of thought construction, it does seem reasonably clear
that since about the middle of the 19th century sociological thought
has been constructed both of unexamined or metaphysical ideas and
of paradigmatic ideas. Insofar as sociological thought has been para-
digmatic, it has been multiply paradigmatic: that is, sociological
thought has since that time normally been intentionally constructed in
such a way that it will be seen by members of the (roughly defined)
discipline to fit certain (multiple) ideas held by those members to
constitute "sociological thought" (4). These paradigmatic ideas were
the foundations upon which whole works were constructed. Some-
times one or two paradigmatic ideas formed the basis for a sociologist's
lifetime work (5). In general, however, those works were constructed
around a number of paradigmatic ideas, even though these ideas often
conflicted with each other when they were examined in relation to
each other (6).
Most of the twentieth century sociologists who have done works
(3) See Thomas S. KUHN, The Struc- vacillation on the part of some. Why did
ture of Scientific Revolutions (Chicago, "sociology" win out over the others, at
The University of Chicago Press, 1962). least in France? Did its generality allow
(4) Until the latter part of the 19th its proponents to gain the support of more
century the terms used instead of "sociol- traditions of thought in their political
ogy" were "moral statistics", "public struggles to found an independent profes-
hygiene", "medical statistics", "social sion in the academic setting? Was it the
statistics", etc. Since these various disci- failure to achieve such a general name
plines of thought actually constituted the that led to the death of moral statistics in
paradigmatic works for those works (such Germany? This whole critical question
as DURKHEIM'S Suicide) still treated as of the naming of disciplines awaits serious
sociological paradigms by sociologists, it investigation.
seems apparent that one should use the (5) Halbwachs has previously argued
term "sociological" in referring to them, that this is the key to understanding
even though the earlier authors did not Quetelet's many works. See La The'orie de
meet that often used rule-of-thumb test of Vhomme moyen : Essai sur Quetelet et la
the self-imputation of the name "sociol- statistique morale (Paris 1913).
ogy". This whole problem of naming (6) Rather than any synthesis of the
brings into sharp focus the whole ques- paradigmatic ideas, the early sociologists
tion of the processes by which this many were normally content to provide a unify-
faceted set of disciplines came to share ing theme to their work, which usually
the one name of "sociology". It is clear consisted of some general goal of social
that there was much conflict over the welfare, such as national power, alle-
whole question of naming that there was viating suffering, or eradicating evil.

2SO
 THE SOCIOLOGICAL ANALYSIS OF SOCIAL MEANINGS OF SUICIDE

on suicide have taken Durkheim's Suicide as their major paradigm:

that is, they have constructed their works in such a way that members
of the discipline will see (7) their works as being further expositions
of the fundamental (paradigmatic) ideas in Durkheim's Suicide. In
doing so, they have implicitly assumed that the fundamental questions
concerning the proper nature of a sociological investigation of suicide
have already been answered in Suicide. This means that the funda-
mental ideas of these twentieth century works are not at all apparent.
One must certainly see them in the context of Durkheim's Suicide if
he is to adequately understand them. Once again, however, the funda-
mental or paradigmatic ideas of Durkheim's Suicide cannot be very
adequately understood until one puts it in the context of the many
nineteenth century works on suicide which Durkheim used to con-
struct his own work.
For these reasons, our investigation of the sociological methods of
analyzing social meanings of suicide must begin with an investigation
of the metaphysical and paradigmatic ideas of the nineteenth century
works which formed the context of Durkheim's Suicide. We shall
then be ready to critically examine these ideas as they are found in
their most developed form in Durkheim's Suicide. From this critical
examination we shall be able to proceed to our exposition of what
now seems to be the best sociological method of analyzing social mean-
ings of suicide.

Nineteenth Century Sociological Thought Concerning Suicide.

The metaphysical ideas of the nineteenth century sociologists were

very close to the metaphysical ideas of common-sense, especially as
they found these common-sense ideas expressed by literary and philo-
sophical authors involved in the practical activities of administering
public health programs and government bureaus of vital statistics.
Throughout the entire nineteenth century, there was a great profu-
sion of literary and philosophical works on suicide. This great interest
of the intellectuals in suicide had begun in the eighteenth century when
the ethics of suicide became a fundamental issue in the great debate
over the real and proper nature(s) of society (8). This concern with
(7) It must be specifically noted here (8) See. L.G. CROCKER, Discussion
that sociologists have often used many of Suicide in the Eighteenth Century,
devices for presenting their works as Journal of the History of Ideas, XIII
more in agreement with the professionally (1952), pp. 47-72.
accepted paradigm, Durkeim's Suicide, The importance of suicide in French
than they in fact were. Halbwachs', for thought of this period was in turn prob-
example, tried to show that his work, Les ably partly the result of the importance
causes du suicide, was in agreeement "in of the subject in English thought in the
principle" with Durkheim's Suicide even seventeenth century. See S.E. SPROTT,
though he in fact rejected much of Dur- The English Debate on Suicide (La Salle,
kheim's basic argument. Illinois, Open Court Press, 1961).

251
 JACK D. DOUGLAS

suicide as a social problem grew rapidly in the nineteenth century,

eventually reaching a point at which a large segment of the population
seems to have believed that there was a veritable 'mania' for suicide
sweeping Europe (9).
This great concern with suicide led to the publication of a huge
number of works on suicide which expressed the metaphysical ideas
concerning suicide of common sense and which directly influenced the
works of the sociologists. The works of Buonafede, Bourquelot, des
Etangs, Debreyne, Lisle, de Boismont, and Legoyt were all quite
important in this respect (10).
From these many works, and probably from their own direct
common-sense experience as well, the sociologists derived (or "absorb-
ed") their most important metaphysical ideas for their works on
suicide. Though by no means the only important metaphysical ideas
in their works (11), the three most important of these for our purpose
are the following:
(1) Social actions are in some way caused (or motivated) by mean-
(9) Tissot wrote a very influential sophical works on suicide in this essay, it
work expressing both this idea and the is important here to note that this concern
idea that an increasing "spirit of revolt" with suicide as a problem is still very
was responsible for such social actions, great in the Western world today. In fact,
an idea which became more important in though it may have been a temporary
the sociological works as the century result of the Second World War, there is
progressed. (See De la manie du suicide, every indication that the last few decades
Paris 1841.) The same idea was express- has been a period in which suicide was
ed in the far more balanced work of Jan considered to be a more serious philoso-
MASARYK: "Die Selbstmordneigung tritt phical and moral problem than at any
gegenwartig in alien civilisirten Landern time in the last century. Albert CAMUS
mit erschreckender Intensitat auf..." (See great work, The Myth of Sisyphus, begins
Der Selbstmord als Sotiale Massener- with the brutal assertion that "there h
scheinung der Modernen Civilisation (Wien but one truly serious philosophical prob-
1881). lem, and that is suicide". For other
Much of this belief in the spreading important examples see P.L. LANDSBERG,
"mania" of suicide was probably the The Experience of Death: the Moral Prob-
result of the great importance of suicide lem of Suicide (New York, The Philo-
in romantic literature. (See, for example, sophical Library, 1953) ; Leon MEYNARD,
Maigron's attack on the romantics, Le ro- Le suicide, itude morale et metaphysique
mantisme et les mceurs, Paris 1910.) But (Paris, Presses Universitaires de France,
the feeling of certainty that the quantity 1954), and Georg SIEGMUND, Sein oder
of suicide was steadily increasing was Nichtsein: Die Frage des Selbstmordes
probably the result of the steady increase (Trier, Paulinus-Verlag, 1961).
in the official registrations of deaths (10) Many of the most important
caused by suicide. This steady rise in the works forming the link between philo-
official counts was most likely the result sophical, literary, and common-sense works
of a steady growth in the registration and the sociological works were quoted
activities of the officials, but there was extensively (and referred to in the biblio-
little critical attitude toward the statistics graphy) in Legoyt's very influential work,
except on the part of doctors who were Le suicide ancien et moderne (Paris 1881).
involved in the problems of trying to (11) One of the most important meta-
categorize the causes of death. physical ideas in these sociological works
(Though we shall not be directly con- will not be directly dealt with here at all.
cerned with recent literary and philo- This is their universally shared idea that

252
 THE SOCIOLOGICAL ANALYSIS OF SOCIAL MEANINGS OF SUICIDE

ings held by the individual and shared by other members of the

society (12). This idea was and is one of the most fundamental
common-sense ideas concerning persons. It was accepted and used by
the sociologists even when certain of their paradigmatic ideas conflicted
with it; and it remained very largely unexamined among the French
and Italian sociologists who were most important in constructing
sociological thought on suicide, even when the German sociologists
began to consider the problem of determining social meanings to be
the most fundamental problem of sociological thought. Though I do
not think such a fundamental proposition should go unexamined, space
limitations will prevent our treating it as problematic here; it will
remain a metaphysical idea for our purpose in this essay.
(2) Individuals know the meanings of their own actions and they
also know the meanings of other individuals' actions. The first half
of this metaphysical idea was denied by one of the paradigmatic ideas
of sociological works in the later half of the 19th century (see below),
but 'the second half, the ability of the observer (the sociologist) to
understand the meanings of the actions of others, was retained.
(3) Meaningful social actions, especially actions which are moral or
immoral, are just as subject to counting and quantitative analyses as
are physical objects and properties. This idea, which was first clearly
stated by certains philosophers at the end of the 17th century and first
used in the analysis of social statistics by Sussmilch in the 18th cen-
tury (13), remained very largely unexamined throughout the 19th
century and the first half of the 20th century. (This particular idea
is to some degree paradigmatic, in that it was quite explicitly stated
by the moral statisticians and others and in that they sometimes
examined the ways in which one should go about quantifying. How-
ever, it is treated here as being primarily a metaphysical idea because
the sociologists never really questioned it and, therefore, never really
attempted to establish or demonstrate its validity. From the seven-
teenth century on, men of practical affairs, such as government officials
and physicians, had come to assume that the social world, like all the
rest of the world, was subject to quantitative precision (14). By the

suicidal actions are "immoral" actions, the paradigmatic thought of the "social
both in an absolute sense (as in Durkeim's mechanists" and some of the more
conception of "social pathology") and in extreme positivists who simply sought to
the sense that this is a universally shared use the theories of the natural sciences as
social meaning of suicide in Western paradigms for analyzing social actions,
societies. This idea was behind the expli- For an excellent treatment of the "social
cit theoretical idea that a breakdown or mechanists", see PITIRIM A. SOROKIN,
decrease in "social constraint" (or "social Contemporary Sociological Theories (New
organization", etc.) was in some way the York, Harper, 1928), pp. 3-63.
cause of the increasing suicide rates in (13) Johann Peter SUSSMILCH, Die Gott-
Western societies. liche Ordnung in den Veranderungen des
(12) This metaphysical idea sharply Menschlichen Geschlechts (Berlin 1761).
distinguished sociological thought from (14) For an excellent treatment of the

253
 JACK D. DOUGLAS

nineteenth century the only questions concerned matters of how one

could most effectively quantify the social world.)
These metaphysical ideas pervade the sociological works of the nine-
teenth century—and most of those of the twentieth century. They
were, however, most influential in determining the specific forms of
thought in the sociological works through both their influence on the
paradigmatic ideas of the sociological work and their combinations
with the paradigmatic ideas in these works. The most important
paradigmatic ideas of the 19th century sociological works on suicide
(and on most other social phenomena) are the following:
(1) The stability of suicide rates proves that the official suicide
statistics (from which these rates were taken) are reliable and valid.
As de Guerry had said, such regularity could not be the result of
chance:

La statistique criminelle devient aussi positive, aussi certaine que les autres
sciences d'observation lorsqu'on sait s'arreter aux faits bien constates, et les
grouper de maniere a les degager de ce qu'ils offrent d'accidentel. Ses resultats
generaux se presentent alors avec une si grande regularity qu'il est impossible
de les attribuer au hasard. Chaque annee voit se reproduire le meme nombre de
crimes dans le meme ordre, dans les memes regions; chaque classe de crimes a
sa distribution particuliere et invariable, par age, par saison; tous sont accom-
pagnes, dans des proportions pareilles, de faits accessoires, indifferents en ap-
parence, et dont rien encore n'explique le retour (15).

(2) The stability of suicide rates indicates that these actions are
caused by some lawful factors external to or not controlled by the
individuals commiting them. Again, de Guerry had stated the case
rather well:

Si nous considerons maintenant le nombre infini de circonstances qui peuvent

faire commettre un crime, les influences exterieures ou purement personnelles qui
en determinent le caractere, nous ne saurons comment concevoir, qu'en dernier
resultat, leur concours amene des effets si constants, que les actes d'une volonte
libre viennent ainsi se developper dans un ordre fixe, se resserrer dans des limites
si etroites. Nous serons forces de reconnaitre que les faits de l'ordre moral sont
soumis, comme ceux de l'ordre physique, a des lois invariables, et qu'a plusieurs
egards, la statistique judiciaire presente une certitude complete (16).

(3) The most important external factors causing these actions are
social factors (17).
general cultural development in the West- (17) Both de Guerry and Quetelet
em World, of the ideas and values of expressed this position well. Consider,
quantitative precision see G.N. CLARK, for example, the following statement by
Science and Social Welfare in the Age of Quetelet: "Society includes withing it-
Newton (Oxford, Clarendon Press, 1949). self the germs of all the crimes commit-
(15) A.M. GUERRY, Essai sur la statis- ted, and at the same time the necessary
tique morale de la France (Paris 1833), facilities for their development. It is the
p. 9. social state, in some measure, which
(16) Ibid. p. 11. prepares these crimes, and the criminal

254
 THE SOCIOLOGICAL ANALYSIS OF SOCIAL MEANINGS OF SUICIDE

(4) The most important social factors in the causation of social rates
of suicide are the social meanings, especially the "moral customs" of
the social groups. This paradigmatic idea was clearly an explicit
formulation of one of the most important metaphysical ideas (18).
(5) There exists a reasonably small set of highly abstract social
meanings (called the "social system" or the "social structure") which
are the causes of specific patterns of social actions such as suicide
rates; and the only valid sociological theory of such patterns (or sui-
cide rates) will be one in terms of states of this set of abstract mean-
ings. Unlike the other paradigmatic ideas, this idea was not fully
developed until Durkheim's Suicide (19). As we shall show further
below, however, this idea, in the specific form used by Durkheim,
had slowly developed throughout the century.

The Conflict 0} Methods.

These metaphysical and paradigmatic ideas, especially when combin-
ed with a few lesser paradigmatic ideas (see below), formed the basis
for two fundamentally different methods for analyzing the social mean-
ings of suicidal actions. At first, these two methods were pretty well
combined, at least syncretistically, in the works of statistical medi-
cine (20). As the 19th century progressed, however, these two tra-
ditions of methodological analysis, the case study method and the (moral)
statistical method, came into increasing conflict. Out of this conflict
the moral statisticians created certain paradigmatic ideas to justify
their own method and to "invalidate" the case study method. These
paradigmatic ideas, combined with the more general metaphysical and
paradigmic ideas already discussed, came to constitute the distinctive
ideas of Durkheim's sociologistic position (21).

The Case Study Method.

The method of determining the meanings causing suicidal actions
by studying the individual cases of suicide was firmly grounded in the
is merely the instrument to execute them, tween the external categories and the
Every social state supposes, then, a cer- internal, meaningful states seems to have
tain number and a certain order of crimes, dissuaded him from developing the idea,
these being merely the necessary conse- See, for example, his excellent statement
quences of its organization." (A Treatise of this position on page 114 of Ibid,
on Man, (Edinburgh 1842), p. 6, originally (20) The most important work in this
published as Sur Vhomme in Paris in school of thought is ESQUIROL'S Maladies
1835.) mentales (Paris 1838).
(18) This idea was clearly developed (21) Though our discussion of sociolo-
by the time of MORSELLI'S, Suicide: An gism is somewhat different, distinc-
Essay on Comparative Moral Statistics tive aspects of Durkheim's sociologism
(New York 1882). have previously been well discussed by
(19) Morselli seems to have shared Edward TIRYAKIAN in Sociologism and
much of this idea, but his greater caution Existentialism (Englewood Cliffs, N.J.,
in constructing statistical arguments be- Prentice-Hall, 1962).

255
 JACK D. DOUGLAS

metaphysical ideas that actions are caused by meanings shared and

understood by all members of the society. Given these two ideas, it
was apparent that to explain suicide, one had to know its meanings
and that one could best know these meanings by self-reports of mean-
ings. The increasing use, primarily by doctors, of the ideas of scientific
induction in explaining any actions led to the development of this
common-sen-sical method into an attempt to provide more abstract
explanations by observing a number of cases of suicide. This informal
statistical method (or method of analytic induction) (22) was increas-
ingly formalized. It developed from a simple, unstructured review
of the classical literary cases of suicide in such works as those of
Montaigne (23) and Voltaire (24) into the careful counting by Brierre
de Boismont (25) and others of the number of suicide notes giving
certain motives for the ensuing suicidal actions. The methods used
by the members of this methodological tradition for determining the
meanings of suicide, however, did not become formalized along the
lines of the statistical method. These researchers and theorists did
not attempt to impose any sets of predetermined categories upon these
meanings, rather, their analyses stayed close to the linguistic categories
used by the suicides themselves. Nor did they make much attempt to
analyze the meanings by any more formal methods.
Though, as we shall shortly argue, this case study method was in

(22) One of the finest discussions of ly on literary cases taken from the
this method in relation to other methods Romantics to construct his types of sui-
used by sociologists is by L.L. BERNARD cide. E. DURKHEIM, Suicide (Glencoe,
in The Development of Methods in So- The Free Press, 1951), pp. 227-294.
ciology, The Monisi, XXXVIII (1928), 292-
320. Montaigne discussed suicide at length (24) Voltaire's essay on suicide in the
in "Custom of the Island of Cea". It Philosophical Dictionary made use not
should be stressed that the use of literary only of the traditional historical cases but
and historical cases of suicide remained also of contemporary newspaper cases.
very important throughout the nineteenth From such sources he attempted to arrive
century. Such works as that of Legoyt at certain scientific conclusions about
were largely based on historical material suicide : 1) suicide is more frequent in
and in certain fundamental respects cities than in rural regions; 2) the expla-
this form of data was far more reliable nation of the greater urban frequency of
and valid than that of twentieth century suicide is that cities produce more melan-
sociologists. (See the extensive criticisms cholia (or depression) in individuals
of official statistics below.) Historical because they have more free time from
and literary case material still remains, physical labor to think; 3) suicide can
of course, the best source for studies of be physically inherited because moral
changes in the moral meanings of suicide. character is inherited (an idea which
Such material was used very well for this Morselli and many others most emphat-
purpose by W.E.H. LECKY in History of ically accepted); and 4) some suicides,
European Morals from Augustus to Char- such as Euripide's Phaedra, commit sui-
lemagne, 2 vols. (New York 1869). It was cide in order to get revenge against some-
used excellently in the finest study done one (an idea which became of great
thus far on the moral meanings of suicide, importance only in the twentieth century
A. BAYET'S he suicide et la morale (Paris, works on suicide).
Alcan, 1922).
(25) BRIERRE DE BOISMONT, DU suicide
Durkheim himself relied almost entire- et de la folie suicide (Paris 1856).

256
 THE SOCIOLOGICAL ANALYSIS OF SOCIAL MEANINGS OF SUICIDE

certain respects more valid than the statistical method used by the
sociologists of the latter part of the 19th century, still the method
involved far too simplistic a view of social meanings. For our pur-
poses here, it is sufficient to note that the method failed to consider the
complexity and difficulty of understanding one's self. It assumed far
too high a degree of self-awareness, rationality, and honesty on the
part of social actors. It was in part because of their realization of this
that the moral statisticians and, subsequently, the Durkheimian sociol-
ogists rejected the case study method of determining and analyzing
the social meanings of suicide. This was shown in two specific para-
digmatic ideas they both adopted and created in their analyses of the
works based on the case study method:
a) suicide itself is an irrational action, so that individuals who
commit suicide cannot be presumed to understand the meanings of
their own actions (26) ;
b) men are in general infinitely complex and/or irrational, so that
the whole case study method of analyzing meanings by taking the
statement and actions of individuals as the fundamental data is invalid.
There were, however, other reasons, probably more important
reasons, for the rejection of the case study method by these groups
of sociologists. For one thing, some of them were anxious to create
their own professional and academic discipline, and this seemed to
demand a method that was distinct from the common-sense methods
still used by the (philosophical and psychological) case study method-
ologists. As Durkheim said, " [...] if there is to be a social science,
we shall expect it not merely to paraphrase the traditional prejudices
of the common man but to give us a new and different view of
them" (27). Even more importantly, however, there was the distinc-
tive mathematical method, the statistical method, to which the moral
statisticians had become increasingly committed. The fundamental
idea of this whole mathematical method was that individual differences
must, for certain purposes, be eliminated or overlooked. Quetelet had,
to some extent, attempt to salvage consideration of individuals by his

(26) This idea was largely adopted placed the proposal to establish a regu-
from the psychiatric theorists, especially lar course in sociology in the Faculty of
from Esquirol's work (op. dt.) in which Letters at Bordeaux, enabled us to devote
suicide was considered to be a symptom ourselves early to the study of social
of insanity. science and, indeed, to make it our voca-
(27) The Rules of Sociological Method tion. Therefore, we have been able to
(Glencoe, The Free Press, 1962), xxxvu. abandon these general questions (of philo-
The importance of the new method for sophical sociology) and to attack a
the new academic discipline (or the other certain number of definite problems. The
way around) was indicated by DURKHEIM very force of events has thus led us to
himself in his Introduction to The Rules construct a method that is, we believe,
of Sociological Method : "A happy combi- more precise and more exactly adapted to
nation of circumstances, among the the distinctive characteristics of social
more important of which may rightly be phenomena" {Ibid. p. ix.).

257
 JACK D. DOUGLAS

average man theory; but data on individuals was increasingly rejected

until the fundamental paradigm of the sociologistic position almost (28)
completely eliminated such data:
c) society must be considered to be a different level of reality from
individuals, so that analyzing data from individuals is invalid (29).
Unfortunately, as we shall see, this paradigmatic idea had as one
of its consequences the elimination of all scientific means of determin-
ing and analyzing the social meanings of suicide.

The Statistical Method of Analysing the Social Meanings of Suicide.

Though it extends back to the works of the political arithmeticians

and Sussmilch by direct lines of influence, the statistical method of
the sociologists in the 19th century developed primarily in three major,
interrelated schools of thought, each making somewhat different use
of the fundamental metaphysical and paradigmatic ideas: the medical
statisticians and medical hygienists, deriving largely from the French
"ideologues" (30), and best represented by Villerne and Parent
Duchatelet; the probability theorists, deriving largely from the direct
influence of Laplace on Quetelet, who in turn influenced the works
of all the traditions; and the demographic works, deriving from many
different sources, including Sussmilch through his influence on Malthus
and others. These traditions of thought were increasingly synthesized
and carried forward during the century by the moral statisticians,
especially by the works of Guerry, Etoc-Demazy)(3i), de Boismont,
Lisle (32), Morselli, CEttingen (33), Wappaus, Masaryk, Wagner (34),
and Bertillon (35). At the end of the century stands the great attempt
at a synthesis of the fundamental metaphysical and paradigmatic ideas

(28) Durkheim dit not actually com- other intellectual forces leading in the
pletely eliminate data on individuals, even same direction : political theory had come
when consistency called for this. Through to treat the nation state as independent
his Aristotelian ideas concerning causality, of the individual members; race psychol-
Durkheim retained a form of negative ogy had developed ideas of extra-indi-
(nay-saying) causality for individuals. This vidual, meaningful forces causing actions
was a form of material causality whereas by individuals: and the organic analogy
the efficient causality, which is what con- was a powerful one throughout the nine-
cerns science, was society. For an anal- teenth century.
ysis of such aspects of Durkheim's argu- (30) See F. PICAVET, Les ideologues
ment see Appendix II, "The Individual (Paris 1891).
and Society in Durkheim's Suicide", of (31) Recherches statistiques sur le suicide
Jack D. DOUGLAS, The Social Meanings (Paris 1844).
of Suicide, forthcoming by Princeton Uni- (32) Du suicide (Paris T856).
versity Press, rg6y. (33) Die Moralstatistik (Erlangen 1882).
(29) The statistical method of argument (34) Die Gesetzmossigkeitin den scheinbar
would surely not be sufficient in itself to willhurlichen menschlichen Handlungen
explain this elimination of data on indi- (Hamberu 1864).
viduals. Besides the factors of founding (35) Cours tlementaires de statistique
an independent profession, there were (Paris 1895).

258
 THE SOCIOLOGICAL ANALYSIS OF SOCIAL MEANINGS OF SUICIDE

of these traditions of thought, Durkheim's Suicide (36). Since Durk-

heims's work is the best of these moralstatistical works (37) and since it
is this work which still serves as the sociological paradigm for research
and theory on suicide, we must briefly examine Durkheim's methods
of determining and analyzing the social meanings of suicide.
On the most general level, the task which Durkheim set himself in
Suicide was that of systematically relating the already established sta-
tistical relations between certain official statistics on official categoriza-
tions of individuals (such as married, divorced, etc.) and certain official
statistics on suicide rates to the egoist-anomie theory of immoral actions
developed by certain Romantic authors (38), and previously applied
only partially to the statistics on suicide by Boismont and Morselli (39).
The official statistics on suicide were treated as being almost com-
pletely non-problematic (except in the instance of official categorizations
of motives, which conflicted with Durkheim's whole theory of suicide).
In writing Suicide Durkheim accepted as valid Morselli's idea that
"A corpse is a corpse", by which Morselli meant that an official cate-
gorizer of the causes of death could very reliably and validly tell if an
individual had committed suicide.
There was actually no adequate justification for this assumption.
The medical examiners and theorists of suicide, including Esquirol,
had long been aware of the great difficulty of determining whether or
not a corpse was the result of a suicide (40). Most importantly of all,
these problems were the direct result of the social meanings of suicide
shared by various groups of vital statisticians. Unfortunately for
Durkheim's own arguments, the official categorizations of a death as
caused by "suicide" were generally most dependent on their impu-
tations of an intention to die by one's own action: since one of the

(36) This interpretation of Durkheim's (38) Chateaubriand was especially im-

Suicide has previously been established portant in this respect. For some of the
in the author's The Social Meanings of details on this relationship see Jack D.
Suicide, Part I, chap. 1 and 2, forthcom- DOUGLAS, The Sociological Study of Sat-
ing by Princeton University Press. The cide: Suicidal Actions as Socially Mean-
whole history of the social research of ingful Actions (unpublished Ph. D. thesis,
these many schools has been dealt with Princeton University, Princeton), pp. 20-
in The Rise of Social Research, ed. by 26.
Anthony Oberschall and Jack D. Douglas, , . _,. . .. ... .. ^ . , . ,
, ,, . . „ j -n = > (,g) This formulation of Durkheim s
forthcoming by Harper and Row. . , , , . ., .,
, . _ . , , , ., task leaves out of consideration the very
1371 Since there are some who would . . . , ., . , , „
i. i_. J- i ii- T-. 1 1. • 1 c- • important intellectual warfare aspect of
probably dispute calling Durkheim's Sui- c K ., ., . . ., . . , . p,.. .
.. , , , ° .. .. , ,. Suicide, that is, the task of lustifying
cide a work of moralstatistics we should ,, . , , . . . , ', °
. ., . „ . .. . , . the independent existence of a profes-
pomt out that Suicide was reviewed in * •1 -r^ • •
f, . . , . T, , , . , . sion of sociology. This is very lmpor-
the first volume of L Annie sociologique . j. • j 1 j- ._, , , ,
, „ , „ . , J ., . tant for understanding6 the whole work
(1896-1897), 397-406, under the general
v y v/i, • > * ' * ' ° but not for understanding s that Fpart of
category of "Statistique morale". We . ,. , ,
. . . . , , . 1 , . , , it which concerns us here.
might also point out that the unnamed
author of this review was very critical of (40) Especially valuable in this respect
Durkheim's "denying" all causal signifi- is H.W. RDMSEY, Essays and Papers on
cance to individual factors. Some Fallacies of Statistics (London 1875).

259
 JACK D. DOUGLAS

critical dimensions of meanings involved in the statutory definitions of

"suicide" as a cause of death and in the general common-sense mean-
ings of "suicide" in the Western World is precisely that of "intention
to die" (41), the official categorizations of "suicide" can in general be
only as valid and reliable as official categorizations of "intention".
Since Durkheim thought official categories of intentions or motives to
be completely invalid and unreliable, he should have concluded the
same thing about official statistics on suicide. (The essentially
problematic nature of the social meanings of 'suicide' is a far more
important reason for considering the official statistics on suicide to
have unknown—and possibly unknowable—meanings, but we shall deal
with this below.)
Durkheim probably believed to some extent that the stability of the
official statistics on suicide meant that they could not to any significant
degree be the result of errors. This view had been expressed much
earlier by Guerry, was used intermittently throughout the nineteenth
century, and was ultimately given its most brilliant exposition by Halb-
wachs in Les causes du suicide (42). The argument is invalid, even
in its most brilliant form, for two reasons which these theories failed
to consider: even "errors" are normally found to fit certain stable
patterns (43); and the complex nature of the social meanings of suicide
which lead to a greater number of different ways of categorizing sui-
cide, actually lead one to expect highly stable probability distributions
in the official categorizations of suicide (44).
Besides this fundamental problem with the official statistics on
suicide, there are a large number of biases in the official statistics that
parallel the official categories of social structure (class, occupation, age,
etc.) in such a way as to greatly bias any theory constructed from or
tested by them. Most importantly, it seems most reasonable to expect
that in societies, such as the Western World, in which some groups
morally stigmatize suicides and their families, there will be both differ-
ential tendencies on the parts of members of different (official) cate-
gories to have any "suspicious" deaths within their families categorized
as something other than "suicide" and differential degrees of success
in these attempts. Moreover, it seems most reasonable to expect that

(41) For a consideration of the many currently preparing an English edition

dimensions of meaning involved in at- of The Causes of Suicide, to be published
tempts to formally define "suicide" and by the University of California Press.)
of the implications of this for determin- (43) This had in fact been noted by
ing the common-sense meanings of sui- Buckle in a footnote concerning errors
cide, see Jack DOUGLAS, "The Social in the addresses of London mail, but he
Aspects of Suicide", forthcoming in failed to draw the general implication,
volume IX of the new International (44) The whole argument and mass of
Encyclopedia of the Social Sciences; and data against the validity and reliability
Appendix I of The Social Meanings of of official statistics on suicide has been
Suicide, op. cit. dealt with in great detail in Part III
(42) M. HALBWACHS, Les causes du of Jack D. DOUGLAS, The Sociological
suicide (Paris 1930). (The author is Study of Suicide, op. cit. pp. 259-406.

260
 THE SOCIOLOGICAL ANALYSIS OF SOCIAL MEANINGS OF SUICIDE

the attempts and the frequencies of success will be greatest precisely

in those groups which are most "integrated" in the society. The
general implication of this very involved argument is that the greater
the degree of social "integration" of a group, the lower will be its
official suicide rate as a result of the nature of the official categorization
process itself. Consequently, one can explain the fundamental cor-
relations in Durkheim's argument in terms of the nature of the official
categorization processes (45).
In general, the use of official statistics on suicide to derive and test
sociological theories of suicide was based on fallacious assumptions and
arguments. Moreover, the failure of the sociologists using the statis-
tical method of analyzing social meanings to see this was the direct
result of their failure to see the importance of social meanings in
determining the official statistics and of their whole statistical method
of analyzing social meanings.
On the one side of Durkheim's general argument were the official
statistics on suicide. On the other side were the official statistics on
other official categories such as marriage, divorce, education, etc. Durk-
heim by no means assumed that correlation constituted causality or
explanation. The explanation of these statistical relations in terms of
social meanings was provided by Durkheim in different ways: 1) his
causal (or backward) definitions of suicide types provided him with
some social meanings; 2) he derived the meanings of the other official
categories from his common-sense observations and reasonings; 3) he
inferred social meanings from the juristic norms or laws of the society;
4) he inferred post hoc some meanings of these other categories from
their relative associations with the statistics on types of suicide (which
had been given meanings by definition, though these definitions really
were partly descriptive definitions derived from Durkheim's knowledge
of the works using the case study method of analyzing social mean-
ings (46); and (5) he inferred the meanings of the relations from many
different petitio principii created for the purpose.
Durkheim's petitio principii method of argument has been very
well criticized by both Alpert and Needham and will not concern us
here (47). Durkheim's post hoc or causal definitions of the suicide
types actually added little social meaning to the rest of his analyses.
(45) Ibid. official categorization processes in specific
ways.
Most of the previous arguments against (46) De Boismont was especially impor-
the use of official statistics have been tant in this manner and probably provided
by psychiatrists. (See, especially, the fine most of Durkheim'sknowledge of individual
work by F. ACHILLE-DELMAS, Psychologic cases of suicide. See Jack D. DOUGLAS,
pathologique du suicide (Paris, Alcan, 1933). ibid. pp. 20-112.
Though these works have often been (47) See, especially, Rodney NEEDHAM'S,
right in particulars, they have never tried "Introduction" to E. Durkheim and M.
to show that there exist (systematic) biases Mauss, Primitive Classification (Chicago,
in the statistics and that these biases are The University of Chicago Press, 1963),
related to the social structure and the pp. xiii-xv.

26l
 JACK D. DOUGLAS

The critical aspect of his whole method of determining and analyzing

social meanings of suicide was the method he used to determine and
analyze the meanings of the official categories other than suicide.
Bayet, in his great work on Le suicide et la morale, was the first
sociologist to see clearly that Durkheim's whole method of determin-
ing and analyzing the meanings of these categories (and, one should
add, of the types of suicide as well) was common-sensical:
En tout cas, ce qui est grave, c'est qu'il faut croire l'auteur sur parole. Oil
sont les usages prouvant que les protestants "punissent le suicide" ? Par quoi
s'exprime "l'eloignement" pour ceux qui touchent au suicide ? Quels faits permet-
tent de dire que la morale commune reprouve le suicide ? Durkheim ne le dit pas.
Sans doute est-il d'avis que la morale de son temps est la sienne et qu'il la
connait. Mais on peut supposer aussi, sans aucun paradoxe, que notre propre
morale nous est en un sens fort etrangere. Le temoignage du plus grand philo-
sophe ne peut remplacer, au point de vue scientifique, des observations soumises
au controle a la critique (48).

Durkheim did not clearly see the need for any scientific method of
determining and analyzing meanings: probably because he started
Suicide with a positivistic method which emphasized external (morpho-
logical) factors as the causes of social actions such as suicide (49). In
the course of the work he came increasingly to see the social meanings
of such categories or factors as being the critical causes or explanations
of the suicide rates, but he had no objective method of determining
social meanings other than the use of common-sense and the existing
juristic norms. In one part of Suicide he argued, as he had earlier,
that only the juristic norms were an adequate indication of the moral
meanings, since they were not subject to individual variations (50).
His whole argument, however, rested on the absolutely fundamental
assumption that the moral meanings of suicide do not vary within
Western cultures, since, otherwise, variations in suicide rates could
be explained by his own theory in terms of variations in these moral
meanings rather than in terms of variations in the social meanings of
anomie-fatalism and egoism-altruism. (Bayet saw how fundamental
Durkheim's assumption of the invariance in moral meanings was to
Durkheim's whole argument and proceeded to demonstrate in immense
detail just how unjustified Durkheim's assumption was for his own
society, France.) This meant that he was left with nothing but com-
(43) A. BAYET, Le suicide et la morale not, however, see the reasons for this in the
{Paris, Alcan, 1922), p. 3. very nature of Durkheim's positivistic
(49) Roger LACOMBE, in his excellent gen- method nor did he have a specific solution
eral critique of La me'thode sociologique de to offer to the problem.
Durkheim (Paris, Alcan, 1926) has argued (50) As Bayet {op. cit.) argued so excel-
that the fundamental weakness of Durk- lently, Durkheim's whole " realist" concep-
heim's whole method was his failure to tion of law and its relations to morality and
recognize the need for any scientific means actions is completely untenable. Bayet,
of determining the inner (psychological) however, did not see that this method of
meanings which Durkheim believed to be analyzing moral meanings was a necessary
the causes of social actions. Lacombe did outcome of Durkheim's whole method.

262
 THE SOCIOLOGICAL ANALYSIS OF SOCIAL MEANINGS OF SUICIDE

mon sense to determine and analyze the variations in the social

meanings of the categories which he was using to explain variations in
the suicide rates.
To adequately demonstrate that Durkheim did rely almost exclu-
sively upon his common-sense understandings of the official categories
of social relations to derive or impute the critical meanings between
these official categories of social relations and his general, meaningful
causes of anomie-fatalism and egoism-altruism would demand an
extremely long argument. Since, moreover, this argument has already
been presented elsewhere, we shall content ourselves with one instance
of this critical flaw in Durkheim's argument. When Durkheim invoked
the supposed differences between the masculine mind and the feminine
mind to explain the less "constraining" effect on women of marriage,
he was merely following the established practice among students of
suicide. Morselli had argued, for example, that men are, by their
nature, much more subject to "egoistical motives" than women:

As to her causes (physical and moral) the greatest excess of men is found
in the group of vices, in that of financial embarrassments, and in weariness of
life, that is to say, amongst the egoistical motives, whilst among women, after
mental diseases, there predominate passions, domestic troubles, shame and remorse
(especially in cases of illegitimate pregnancy). Among the causes which urge
them to leave this life woman always exhibits that spirit of self-denial, that
delicacy of feeling and of love, which, inspire all her acts (51).

Now, Durkheim does not at all dispute the greater egoistic suicide rate
of men. Largely working with the same statistics on family associations
and suicide that Morselli used, Durkheim found more or less the same
statistical relations and, working in the same theoretical tradition, he
agreed, in his own way, that the difference was man's greater egoism.
But he then supplies a completely contrary (and very complex) com-
mon-sense interpretation of this supposedly greater masculine egoism,
an interpretation that happens to fit all of the details of his own theory:

This is also why woman can endure life in isolation more easily than man.
When a widow is seen to endure her condition much better than a widower and
desires marriage less passionately, one is led to consider this ease in dispensing
with the family a mark of superiority; it is said that woman's affective faculties,
being very intense, are easily employed outside the domestic circle, while her
devotion is indispensable to man to help him endure life. Actually, if this is
her privilege it is because her sensibility is rudimentary rather than highly
developed. As she lives outside of community existence more than man, she is
less penetrated by it; society is less necessary to her because she is less impregn-
ated with sociability. She has few needs in this direction and satisfies them
easily. With a few devotional practices and some animals to care for, the old
unmarried woman's life is full. If she remains faithfully attached to religious
traditions and thus finds ready protection against suicide, it is because these very
simple social forms satisfy all her needs. Man, on the contrary, is hard beset
in this respect. As his thought and activity develop, they increasingly overflow
(51) H. MORSELLI, Suicide, op. cit. p. 305.

263
 JACK D. DOUGLAS

these antiquated forms. But then he needs others. Because he is a more complex
social being, he can maintain his equilibrium only by finding more points of sup-
port outside himself, and it is because his moral balance depends on a larger
number of conditions that it is more easily disturbed (52).

Even when he was aware of common-sense interpretations that were

completely contrary to his own, he continued to use his common-sense
interpretations with complete confidence.

The Method of Using Systematic Comparisons of Situated Communi-

cations to Determine and Analyze the Social Meanings of Suicide.

The logical outcome of the paradigmatic ideas of those works using

the statistical method of determining and analyzing social meanings is
that found in Durkheim's sociologistic theory: they have no scientific
means of determining and analysing social meanings in terms of real-
world events that can be objectively observed and replicated because
they have denied all epistemological value to such forms of data. They
must, however, have social meanings to explain the suicide rates.
Consequently, they must use their common sense ideas (their deus ex
machina data) about the meanings of certain categories. Being un-
constrained by any methods of observations for arriving at such mean-
ings, they use their common-sense ideas to impute those meanings to
the category which their abstract theories have predicted. The argu-
ment has now come full circle and appears very convincing until one
considers the nature of the critical evidence, the social meanings (54).
Though the details are very difficult, the general nature of the
remedy is clear: in order to determine and analyze the social meanings
of suicide, and, thence, to be able to determine what causal relations
exist between these meanings and the various types of suicidal actions,
sociologists must develop scientific methods of observing, describing
and analyzing communicative actions concerning real-world cases of
suicide. Since we have already seen the fundamental weaknesses of

(52) E. DURKHEIM, Suicide, op. dt. et qu'on connaissait desormais la solution,

pp. 215-216. Est-ce la dialectique, sont-ce les statis-
(53) The whole argument concerning this tiques qui emportaient la conviction ? L'ua
interpretation of Suicide can be found in et l'autre sans qu'on sut bien toujours
Part I of Jack D. DOUGLAS, The Social distinguer ce qui etait l'autre. Quelquefois
Meanings of Suicide, forthcoming by Prince- la dialectique plus que les faits, non par
ton University Press. la faute de Durkheim d'ailleurs. Mais
(54) Though being very cautious in cela presentait plus d'un inconvenient. On
criticizing the master of the French school, ne s'apercevait pas que l'edifice reposait
Halbwachs clearly saw the reliance of sur des fondements qui n'etaient point
Suicide upon the "dialectique": partout aussi solides. Comment en eut-il
« En fermant cet ouvrage, plus d'un ete autrement ? II n'y a pas d'oeuvre scien-
lecteur, surtout plus d'un lecteur philo- tifique que de nouvelles experiences n'obli-
sophe, a sans doute eu le sentiment que gent a reviser et completer. » M. HALB-
le probleme du suicide ne se posait plus, WACHS. Les causes du suicide, op. cit., p. 3.

264
 THE SOCIOLOGICAL ANALYSIS OF SOCIAL MEANINGS OF SUICIDE

the early case study method, this earlier method will not help much.
The 20th century case study methods of psychiatry and psychology are
also of little use, since they are based primarily on certain abstract,
predetermined, genetic theories of action (55) which leave out of con-
sideration almost all aspects of social meanings and, thereby, falsify
the nature of human action.
What is called for is a whole new sociological method for determin-
ing and analyzing the communicative actions which can be observed
and replicated in real-world cases of suicide. This method must retain
the emphasis on observation and description of the earlier case-study
methods, but it must also retain the emphasis on comparative studies
of patterns of meanings of the statistical method.
The ideal method would involve, at the first step of observation, the
exact recording of all verbal and non-verbal communicative acts involv-
ed in a case of suicide. The next step would be the preliminary
analysis of the patterns of invariant linguistic items (i.e. words, phrases,
sentences, facial expressions, etc.). Following this, one would next
analyze such communications to determine the varying structures in
which these linguistic items appear: that is, one would determine the
usages (or constructions—see below) made with these linguistic items.
One would then attempt to determine the relations between the general
situations or contexts of the social actors, as denned by the actors,
and the constructions appearing. Lastly, one would be ready to attempt
a more general theory relating meanings to each other and to actions.
We need hardly point out that this ideal is quite Utopian at this
time. At the present time the best approximation to the ideal method
which is practicable involves the careful use of the best cases of suicid-
al communications and actions recorded by psychiatrists and other
persons involved, such as the persons commiting the actions, analyzed
by one's common-sense understandings of the meanings of such com-
municative acts in our society (56). Such analyses of cases must be
comparative and must not be predetermined by some set of personality
or society categories assumed to be the only right categories for under-
standing the social world.
Such an analysis is clearly a very lengthy one that we cannot pos-
sibly fully undertake here, even though the details of the method will
not be clear until one can actually see a number of cases being compar-
atively analyzed. However, my previous analyses of just this sort have
led to conclusions which have fundamental implications for all socio-
logical investigations and theory and which, consequently, are especial-
(55) The inadequacies and the limits the social meanings of suicide in other
of usefulness of psychological and psychi- cultures poses very difficult problems,
atric case studies of suicide have been The best attempt to do so thus far is
analyzed at length in Jack D. DOUGLAS, G. DEVEREUX'S Mohave Efhnopsychiatry
The Sociological Study of Suicide, op. cit. and Suicide (Washington, U.S. Govt. Print-
PP- 377-406. ing Office, 1961).
(56) The determination and analysis of

265
 JACK D. DOUGLAS

ly worth summarizing here (57). Following the brief statement of

these general conclusions, we shall briefly analyze a few cases of
"revenge suicide" to demonstrate some of the fundamental details
of this method of determining and analyzing the social meanings
of suicide.
The sociologist who begins to consider the actual ways in which
officials go about deciding that a given instance of death is a "suicide"
soon discovers a startling fact about "suicide" in the Western World.
First of all, one discovers that the term "suicide" does not have any
clear and distinct meaning in the Western World, not even in terms
of the formal definitions proposed by theorists or those decreed by law
for the officials charged with categorizing the "causes of death." The
following are the most common dimensions of meanings found in dif-
ferent combinations in the formal and legal definitions of "suicide":
(1) the initiation of an act that leads to the death of the initiator;
(2) the willing of an act that leads to the death of the wilier;
(3) the willing of self-destruction;
(4) the loss of will;
(5) the motivation to be dead (or to die, or to be killed) which leads
to the initiation of an act that leads to the death of the initiator;
(6) the knowledge of an actor concerning the relations between his
acts and the objective state of death;
(7) 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;
(8) the degree of firmness or persistence of the decision (or willing)
to initiate an act that leads to the death of the initiator;
(9) the degree of effectiveness of the actions in producing death (58).
The profusion and complexity of different definitions of suicide certain-
ly makes it apparent that different official statistics on suicide cannot
be comparable. But, far more importantly, it indicates that "suicide"
is an essentially problematic conception: that is, "suicide" is not a
socially meaningful linguistic category that is either applied rightly or
wrongly to any given instance of social action; rather, it is a category
with different (abstract) meaning that can rightfully either be applied
or not applied to certain types of actions. Whether or not the category
is actually imputed in a given situation is dependent on factors that
are partially independent of the meanings of the term itself. Most
especially, whether or not the category is imputed in a given situation
is dependent upon the argument processes that take place between the
participants in the situation. (For example, it is most common for

(57) See Jack D. DOUGLAS, The Social live de suicide (Paris, Delachaux, 1954),
Meanings of Suicide, op. cit., Part. IV. pp. 9-59. For an analysis of these dimen-
(58) For a detailed presentation of the sions of meanings in the formal definitions,
various formal, theoretical definitions of see Jack D. DOUGLAS, Appendix II, The
"suicide" see P.B. SCHNEIDER, La tenta- Social Meanings of Suicide, op. cit.

266
 THE SOCIOLOGICAL ANALYSIS OF SOCIAL MEANINGS OF SUICIDE

members of the family and friends to argue with the official categorizers
to the effect that this instance of death should not be categorized as
a "suicide".)
This essentially problematic nature of the social meanings of "sui-
cide" means, for one thing, that there is no such thing as one definite,
necessarily valid (socially meaningful) " Suicide rate". The whole idea
of a given, necessarily valid (or invalid) "Suicide rate" for a Western
society is a complete misconception of the meaningful nature of "sui-
cide" in Western societies, even if one is only concerned with the
categorizations made by officials. (One would expect, however, that the
meanings are less essentially problematic among one group with a high
degree of shared experience than between different groups of this
nature). Whatever the state of reality, the "facts", one can construct
many different, equally "valid", socially meaningful counts of "sui-
cide", and this is true even if, for some totally arbitrary reason, one
attends only to the categorizations of officials. The problem with
statistics on suicide, therefore, is not at all one of simply devising
more accurate measurements. The fundamental problem is that of
determining and analyzing the social meanings of suicide must be solved
before one can attempt any quantitative analyses.
However, it is even more important to discover that, aside from
being essentially problematic, "suicide" is also situationally (or prag-
matically) problematic. That is, regardless of the abstract social mean-
ings given to the category of "suicide", the actual or realized imputa-
tions of the category in any given instance of death are dependent upon
many other factors, most especially upon the intentions of the partici-
pants. (An obvious instance of this which illustrates the difference
between the two types of problems of meanings is the individual who
specifically intends to commit what he imputes the category of "sui-
cide" but does it in such a way that the local officials will categorize
it as an "accident".)
Since there are no specific meanings imputed to all (or even most)
suicidal actions, the meanings of such actions must be constructed by
the individuals committing them and by the others involved through
their interactions with each other. Just what specific meanings are
realized or actually imputed will depend on the intentions of the various
actors, the socially perceived ways in which the actions are committed,
the specific patterns of suicidal meanings (see below) which are realiz-
ed, and the whole 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. That is, individuals construct arguments involv-
ing behavior and statements in order to communicate to other arguers
the meanings which best fit their intentions. The obvious examples
are those individuals who construct situations which they believe will
be categorized as "accidents", with the intentions of avoiding "embar-

267
 JACK D. DOUGLAS

rassment" for their families, stigmatization of their character, loss of

insurance money, etc.)
The study of suicide in terms of specific situations or real-world
instances of "suicide" leads one to see that the meanings imputed to
"suicide" by individuals not involved in a concrete, real-world situation
of suicidal actions are very different from the meanings imputed to
"suicide" by individuals who are involved in a concrete situation of
suicidal action at the time of the communication. This means that
the situated meanings are very different from the abstract meanings.
This finding has two fundamental implications for all investigations of
the social meanings of suicide and, most likely, for all of sociology.
First of all, it is not possible to predict or explain types of social events,
such as suicide, in terms of abstract social meanings such as egoism
or general social values against suicide. This is a denial of both the
fundamental paradigmatic idea of Durkheim's sociologists theory and
of the (derivative) fundamental paradigmatic idea of most general
theories in sociology today. Secondly, it is not possible to study the
situated social meanings of suicide, which are most important in ihi
causation of suicide, by any means (such as questionnaires or laboratory
experiments) which involve abstracting the communicators from con-
crete instances of suicide in which they are involved. This generali-
zation leads one to question the value of any method of investigation
of any realm of social action which attempts to abstract the members
of the society from the involvements of their everyday lives.
This does not mean that there are not patterns of meanings common
to those events imputed the category of "suicide" or "suicidal" by
members of the society. When one looks at the meanings imputed in
concrete situations one does find certain general dimensions of mean-
ings occurring in almost all cases. Most importantly, any suicidal action
is believed to mean something fundamental (or substantial) about the
self of the individual committing it, or about the situation (especially
the significant others) in which he committed the action, or about some
combination of the self and the situation. Whether the specific mean-
ing realized (see below) will be directed to the self or to the situation
of the actor will depend on the imputations of causality made by the
various interactors: that is, will they see the individual as the cause
(or "responsible") for his own actions or will they see him as having
been caused ("driven") to do it by the situation (loss of job, family
trouble, etc.). The individual committing the suicidal action very
generally attempts to place one of those two general constructions of
meanings upon the action by making use of such devices as pointing
out the external cause which is to be imputed causality (or blame)
for his suicide.
It is not possible for individuals to construct just any meanings
whatsoever for their actions, though individual creativity does extend
the limits immensely and all cases include imponderable idiosyncrasies.

268
 THE SOCIOLOGICAL ANALYSIS OF SOCIAL MEANINGS OF SUICIDE

There are, first of all, the various criteria determining the plausibility
of arguments, which limit very greatly the meanings which can be
realized, though it is even most likely that in some instances individuals
even intend to have their arguments considered to be implausible or
irrational ("insane", "senseless", etc.). There are, secondly, rela-
tively few patterns of situated meanings which one finds playing impor-
tant parts in most of the realized meanings concerning "suicidal
actions" and which, consequently, are very generally used by indiv-
iduals to intentionally construct certain overall meanings. (Because
of their frequency it seems most likely that individuals must take these
patterns into consideration in some way in their attempts to construct
the meanings of their actions for others.) The most common patterns
of meanings of this sort in the Western World are those involving
"revenge" (59), "the search for help" (60), "sympathy", "escape",
"repentance", "expiation of guilt", "self-punishment", and "serious-
ness" (61).
We can best illustrate a part of the method of analyzing such mean-
ings of suicide (62) by briefly presenting a few comparative case
analyses. For this purpose we shall analyze the "revenge" meanings
of suicide.
As we have noted above, suicidal actions have, as one of their
potential dimensions of meaning, the meaning that something is funda-
mentally wrong with this situation. This means that suicidal actions
can be used reflexively to say that something is fundamentally to blame
for their situation. Because of the generally shared values against
hurting others, especially against pushing another person too far ("to
the brink"), individuals can use such reflexive meanings to achieve
goals while still alive through attempting suicide but "failing" to die
(or succeeding in not dying, depending on just what their goal is).
We can see this use of the reflexive meaning of suicide rather clearly
in the following case:
Mr. F. B., born 1902, had for several years before his admission to hospital
shown increasing irritability, suspiciousness and lack of inhibition. In 1945 he
became openly paranoid and depressed, and at the same time made excessive

(59) The author is currently analysing concentrated on "suicide" rather than

the cross-cultural "revenge" meanings of on the very broad spectrum of different
suicide in a work entitled Revenge Suicide, social categories of "suicidal" phenom-
to be published by Prentice-Hall. ena in the Western world. The reason
(60) Harvey Sacks has previously ana- for this is simply one of simplification,
lyzed these patterns of meanings in "No We could not discuss other categories
One to Turn to", in E. SCHNEIDMAN, ed., in any detail in such a brief work. The
Essays in Self-Destruction, to be published general properties of "suicide" probably
by Holt, Rinehart, and Winston. hold for these other categories, but there
(61) See Jack D. DOUGLAS, The Socio- are many important, specific differences.
logical Study of Suicide, op. cit. pp. 440-511. (Some of these have been discussed briefly
(62) We should point out that through- in Jack D. DOUGLAS, The Social Meanings
out this paper we have discussed and of Suicide, op. cit.)

269
 JACK D. DOUGLAS

sexual demands. He was afraid of committing suicide (three members of his

family had killed themselves), and threatened his wife and child. When his wife
started separation proceedings he became extremely depressed and self-accusatory.
One evening he drank acid with suicidal intent and told his wife that he had
done so. He was immediately admitted to a general hospital and thence transfer-
red to the observation ward and finally to a mental hospital. His paranoid and
depressive symptoms remained stationary and he settled down to a dull retarded
state. His wife did not continue with the separation proceedings, but visited
him regularly, and said that she would not divorce him as he might try again
to take his life. At the time of the follow-up interview, seven years after the
suicidal attempt, he was still in the hospital [...] (63).

In this case we can see quite clearly that the wife interpreted her
husband's suicidal actions as being a direct result of her separation
proceedings, that is, it was clear to her that this particular, immediately
preceding situation was a "cause" of his suicidal actions. In line with
her interpretation of his suicidal actions and her desire that he not
commit suicide, she changed the situation back to what it had been
before the "causing" situation.
However, there is also an "aggressive" possibility in this general
reflexive meaning of suicidal actions (64). By using various methods
of pointing out who or what is the cause (i.e., "to blame") for one's
suicidal actions one can produce negative social sanctions and self-
blame against specific entities and persons (65). In analyzing such
meanings it seems essential to begin comparing the more culturally
obvious instances and only after understanding the many details of
these to proceed to the more culturally involved, subtle cases. One
of the more culturally obvious cases of revenging oneself upon some
specific person(s) is the following:
A young clerk twenty-two years old killed himself because his bride of four
months was not in love with him but with his elder brother and wanted a divorce
so that she could marry the brother. The letters he left showed plainly the
suicide's desire to bring unpleasant notoriety upon his brother and his wife, and
to attract attention to himself. In them he described his shattered romance and
advised reporters to see a friend to whom he had forwarded diaries for further
details. The first sentence in a special message to his wife read: " I used to
love you; but I die hating you and my brother, too." This was written in a

(63) Quoted from STENGEL and COOK, missed the whole social nature of such
Attempted Suicide (London, Oxford Uni- an action.
versity Press, 1958), p. 56. (65) One can construct the meanings
(64) Psychologists and sociologists have of his suicidal actions in such a way as
very generally come to consider "suicide" to "blame the world"or "blame Hollywood"
to be one form of "aggression". See, for (as seems to have been partially true in
example, A.F. HENRY and J.F. SHORT, the death of Marilyn Monroe) and so on.
Suicide and Homicide: Some Economic, However, because our commonsense theo-
Sociological and Psychological Aspects of ries of persons tend to strongly emphasize
Aggression (Glencoe, The Free Press, 1954). persons as the causal factors in explaining
They have not seen, however, that the our actions, it is especially easy to see
success of such an aggressive intent is other individuals as the something that
due to the reflexive nature of the social is to blame for a suicidal action,
meanings of suicide. They have really

27O
 THE SOCIOLOGICAL ANALYSIS OF SOCIAL MEANINGS OF SUICIDE

firm hand; but as his suicide diary progressed, the handwriting became erratic
and then almost unintelligible as he lapsed into unconsciousness. Some time after
turning on the gas he wrote: "Took my 'panacea' for all human ills. It won't
be long now." An hour later he continues: "Still the same, hope I pass out
by 2 a.m. Gee, I love you so much, Florence. I feel very tired and a bit dizzy.
My brain is very clear. I can see that my hand is shaking—it is hard to die
when one is young. Now I wish oblivion would hurry"—the note ended there.
Another note regretted the inconvenience to the landlady for using her prem-
ises as a death-house. Still another read: " 'To whom it may interest: The
cause of it all: I loved and trusted my wife and trusted my brother. Now I
hate my wife, despise my brother and sentence myself to die for having been
fool enough to have ever loved any one as contemptible as my wife has proven
to be. Both she and her lover (my brother) knew this afternoon that I intended
to die tonight. They were quite pleased at the prospect and did not trouble to
conceal their elation. They had good reason to know that I was not jesting.' "
The brother who is twenty-three years old spoke frankly to the police about
his friendship with his brother's wife. Though separated in childhood when the
parents had drifted apart, the two brothers had later on become inseparable
companions until shortly before the tragedy, when both fell in love with the
same girl. The younger man attempted suicide when his love was not returned
and upon his recovery, the girl agreed to marry him out of pity—but later on
she found she could not live up to her bargain. After a few weeks of married
life, the husband discovered the relationship existing between his wife and his
brother. He became much depressed and threatened suicide. The day before his
death, there was a scene and when assured that the two were really deeply in
love with each other, the clerk retorted: "All right, I can do you more harm
dead than alive" (66).

This case shows very clearly the general structure of "revenge"

meanings that suicidal actions are very generally intended to construct
in the minds of others and in the individual committing the suicidal
action. First of all, and most strikingly clear in this case, is the
importance of pointing-out the person one thinks is to blame, the person
whom one intends to be held responsible for the suicidal action by
others and by the person blamed. The general problem of analyzing
such meanings is that of determining how it is that individuals in our
culture go about imputing causality of social actions and what it means
(normatively, especially) to be the cause of this type of action (67).
In the present case the pointing-out of whom the actor takes to be to
blame and whom he thinks should be blamed by others (and by them-
selves) is very clear: he left many notes and made many statements
so that those whom he blamed would be perfectly clear. But the point-
fee) This case is quoted from Louis I. sible for the actions of others far more
DUBLIN and Bessie BUNZEL, TO Be or than is the case in most of the United
Not To Be (New York, Smith and Haas, States. It would thus seem that the
1933) PP' 294-295. fundamental difference, a difference which
(67) Hendin has tried to show that in makes it far more possible to use suicidal
Denmark persons are believed to be not actions as a threat (thus to achieve whatever
only a fundamental cause of the actions one wants to achieve) and as revenge,
of others, as they are in the whole Western is due to a difference in the social den-
World, but that they should be held respon- nition of "responsibility" for the actions

271
 JACK D. DOUGLAS

ing-out of those he wants to be taken as the culprits is not enough:

to have them defined as the causes by others (or by themselves) he
has to show that there exists some typical situation which is typically
believed to cause a typical motive, which in turn is believed to cause
certain typical actions (such as suicidal actions). In this case he did
these things by trying to show that he had been betrayed by the im-
moral action of his brother and his wife.
Now, since the construction of such a meaning is clearly dependent
upon much more evidence than simply his own statements, the mean-
ing that is created in other minds is partly dependent upon the mean-
ingful responses to this blaming by those pointed-out by him as to
blame. In this case they tried to appear sincere (freely talking with
the police) and show that she never really had loved him, was simply
doing him a kind deed, and could not help herself from betraying
him because of the force of love (i.e., it was not "really" a betrayal at
all, though it might be admitted that it looked that way from his stand-
point). The problem which such strategies face is that the committing
of a suicidal action also means that one is highly committed to what
he says (that he is serious and sincere, as this action of ultimate
commitment shows) and that he is deserving of "sympathy" because
of what the external situation has forced him to do (68). In such a
situation those who are blamed have a very difficult time of it trying
to define things in a way more acceptable to themselves (especially if
they "know" that what he says is "true"). And they can hardly
argue that they are more right or that they are more sympathetic.
They would seem to have only two courses of action that promise any
success: that of redefining what happened (it wasn't "really" the way
he said it was), which is the path chosen by the brother and wife in
this case, and that of redefining the person who commited the suicidal
action—he was "crazy" or he was just trying to harm us, so that he
is not so sympathetic after all. This might well help to explain why
it is that so many individuals who attempt suicidal actions are treated
as "crazy" (sent to psychiatrists, etc.) by their significant others and
why individuals who intend their suicidal actions to blame others use
(welfare, etc.) of others, rather than in they seem to be variations on the generally
any difference in definitions of the causal- shared patterns.
ity of such events. See HENDIN, Suicide (68) Here we have had to bring in
and Scandinavia (New York Grune and some of the other dimensions and pat-
Stratton, 1964), pp. 28-29. terns of the social meanings of suicide.
This shows the fundamental need for a
This example brings up the whole prob- general comparative approach in which
lem of sub-cultural patternings in the one sees each pattern or dimension in the
meanings of suicide. Clearly there are context of all the others. Any other
such patternings, some of which are na- approach completely distorts the mean-
tional. ("Romantic suicide" in Germany ings. (This has been done in Part IV of
would be another instance.) However, Jack D. DOUGLAS, The Social Meanings
though little is known about them as yet, of Suicide, op. cit.)

272
 THE SOCIOLOGICAL ANALYSIS OF SOCIAL MEANINGS OF SUICIDE

less direct means of blaming (i.e., so that they won't be clearly just
"aggressive") (69).
This analysis has also led us to see that social causality or "blame",
like "suicide" itself, is the result of suicidal argument processes be-
tween the many participants, real and imagined (70). (It also leads
us to expect that society as a whole must eventually be analyzed in
terms of such meaningful conflicts rather than in terms of assumed
patterns of shared meanings as is generally done today.)
As an example of the far more involved, subtle "revenge" meanings
that can be constructed for suicidal actions, we shall present and briefly
analyze the case of Marguerite as reported in Deshaies excellent work:

Thirty-eight years of age, divorced for several years, Marguerite S. had lost
her parents and a child. She lived alone and worked as a saleswoman in a
Parisian department store. Very pretty, refined, well-balanced, not at all emotion-
al, she had never had the least psychopathic trouble. Of normal intelligence, she
was very affectionate, obliging, gentle, devoid of all aggressiveness, on the whole
passive and somewhat listless. She had a coquetterie, self-confidence, and the
simpering, slightly childish manner common to pretty women. She scarcely
exploited her charms, and lived rather turned in upon herself, waiting for events
instead of preparing for them.
In 1938 there came an event. Chance linked her with an engineer, Guy, two
years younger than she. She became his mistress. Their liaison was not
interrupted by the war, on the contrary—it reinforced it, at least on Marguerite's
part. And then, these were the sad years of the occupation, with their anxieties
and common hopes. Even though they didn't live together, Marguerite strained
her ingenuity to better Guy's existence; their sexual understanding was perfect.
Marguerite was happy and loved passionately, without reserve, without after-
thought, with no eye to the future—even though their liaison was without a
formal engagement. (Guy, a methodical and prudent fellow, had taken care to
make this clear from the beginning.)
For Marguerite, Guy represented the universe. He was at the same time child
and lover, family and master, the reason for living and the aim of living. He
had bound up all her capacity for affection, filled the emptiness of an incomplete
existence, and made the ideal teacher around whom her deeper personality could
develop. It was not a matter of a thunderbolt, but of a slow, steady, building
up of layers which united her indissolubly to her object. The hold of the object
manifested itself in everything though without being anything of an obsession,
for the self had no place to struggle, its adhesion was complete. To give one-
self body and soul is not a vain image, the oblative form of love, the purest
perhaps, the proof, as in the present case, a fortunate passion, a normal passion.
Why must it be so rarely given out ?
A November evening in 1943 produced the catastrophe. With consideration,
but with decision, Guy announced to her the end of their liaison. Marguerite
experienced an intense emotion, with cardiac pain, facial congestion, tears, astenia.
"It's impossible, oh! no, it isn't true! Tell me it's nightmare! My Guy, my
Guy, you are everything for me, you are my God, you are my soul! Can one
live without his soul? It's not possible. You're all my life, everything [...] To
whom will I tell my troubles, my thoughts ? It's horrible! You are my sole
reason for living, without you I no longer have anything. I feel as if my head
will burst. Must I pay with all the tortures of the heart and mind for the few
(69) This whole analysis has been taken suicidal process has been given in Chap. V
from The Social Meanings of Suicide. of Part IV in Ibid.
(70) An extensive analysis of such a

273
 JACK D. DOUGLAS

hours of joy I've known? My darling, if you go away it's either madness or
death for me [...] I don't wish it, but I wouldn't have the strength to bear [...]"
Destroyed, poor Marguerite stopped working, stayed inside, and lost interest
in everything that wasn't the object of her passion. Her disarray was total, a
veritable cataclysm in which, at one blow, all the affective organization of her
life foundered, all her attachments were broken, all her interests vanished, her
whole future dissolved. The duration stopped at the threshold of the present,
burdened with a past which could no longer lead to anything. The weeks rolled
by, their alternating hope and despair determined by the attitude of Guy, who
was sometimes softened and charitable, sometimes hardened and pitiless accord-
ing to the predominance of his tender feelings or his rational will to carry out
his plan.
[...] Meanwhile Marguerite had gone back to work and was striving to reconquer
her lover. She was calm, with a normal deportment, no longer emotional; never-
theless, her passion remained unchanged. After a sad Christmastime, cruelly
nourished by the tender memories of the past, she appeared for the first time
animated by an aggressive tendancy which gave to the idea of suicide the charact-
er of vengeance. "I am suffering, I can't stand it any longer, I wish to die.
I wish I were able to hate you! I have given you everything and you have
given me a hard heart. Why haven't I met another man who had a heart less
hard? Why do I love you so much? I would like to hate, I would like to hate
you. I would like to kill you and then myself [...] Have pity!" "But I know
what I will do: it isn't you who will leave me, it is I. I shall die before you,
under your eyes, I want you to see me die. I want you to see me dead and
I want the image of me always between that woman and you." Jealousy explodes
and works against the unknown rival: "I will scratch that woman's eyes out.
I will kill her. I will kill her. She has no right to marry you. You are my
whole life, without you it's the end of the world, without you I can't go on
living." This aggressiveness was transitory and suicide continued to be seen as
a liberation from suffering and also a way to free Guy from the problems and
boredom she had created for him.
The idea of suicide was active and accepted, but Marguerite still hesitated,
perhaps from a lack of courage, but especially because a vague hope persisted,
since Guy continued to see her. Her sleep was troubled by expressive nightmares:
train accidents, falls down precipices; she is going to hide herself under the water,
deeper and deeper and she suffocates; she throws herself out the window of
Guy's building; Guy suddenly appears at her home and she tells him: "I shall
remain with you always" ; sometimes some sexual dreams, exceptionally dreams
of war (bombing at which she is impassively present).
Then all hope disappeared. Guy definitively maintained his decision and told
her to "remake her life" without him. And then since she could no longer live
either with him or without him, she decided to kill herself. She still loved Guy
as much, and in a letter which she addressed to him on April 29, 1914, she told
him again all her love, and very tenderly, with neither irony nor complaints, she
wished him happiness before telling him adieu. The next day they took her
body from the Seine (71).

The thing that seems most striking about Marguerite's suicide is the
great difference in the general "tone" or over-all meaning (that is, the
context of meaning determined by the dominant meanings) of the last
communication with her lover, her "farewell" letter, from the earlier
communications about her intention to commit suicide. In the earlier
stages of the struggle it was clear that she was fighting and that suicide
(71) G. DESHAIES, Psychologic du suicide 1947). (This case has been translated
(Paris, Presses Universitaires de France, by the author of this essay.)

274
 THE SOCIOLOGICAL ANALYSIS OF SOCIAL MEANINGS OF SUICIDE

was being used as a "weapon", a threat of making the other responsible

for the grave and immoral injury to herself. In the last communica-
tion, which, most unfortunately, Deshaies did not reproduce, she was,
providing Deshaies' interpretation is correct, not at all "aggressive",
quite to the contrary, she was kind and loving toward him. It is, of
course, quite possible on the surface of it to believe that she had a
"change of heart" and decided not to blame her lover for her death.
But, even if this were what she had intended, this is probably not the
meaning of such a communication, or not the sole communication, to
someone, such as the lover himself, who looks at it. The meaning to
an observer, especially one involved with her in such a way as to be
somewhat sympathetic, is that she has at the very end expressed her
deepest love for him and has been so forgiving as to wish him only
the greatest happiness. This not only makes her more dependent upon
him, because of the great love, and, therefore, makes him more the
cause of her suicidal action; but it also makes her far more "sympa-
thetic". The over-all result of her change in tone in her communica-
tions is to make her lover even more "guilty" or to "blame" for her
terrible action, probably even in his own eyes. (It is, however, entirely
possible that her lover, being subjected to blame, might still see it in
terms of the first tone of her communications, especially since the first
parts of a sequence of communications tends to strongly determine the
context of the latter parts of the communications (72). If so, then he
would quite possibly interpret her "farewell" as being just such an
indirect form of "aggression.") The expression of "sympathy" can
itself be an action intended, or in fact even if not intended, to produce
the greatest injury; and it seems especially possible to do this when one
commits a suicidal action, so that the only "aggression" expressed is
that turned upon oneself in such a way as to make the other, whom
one expresses only love for, appear to be the one "really" responsible
for the injury (73).
This last analysis is clearly dependent primarily upon the common-
sense understandings of the sociological observers. As such, it is clearly
less replicative or scientific. Such analyses as these can only be verified
after the more objective, replicative, comparative studies of the actual,
direct statements of the social actors have led to highly verifiable
results. The immediate goals of sociological investigations of suicide
must be to provide the necessary descriptive material and the analyses
of such material. A very great deal of further careful descriptions and
analyses of the whole suicidal processes remains to be done. However,
the basic problems and the methods for solving them now seem clear.

(72) This suggested "primacy effect" is meanings of suicidal phenomena,

one of the very many properties of communi- (73) This analysis has been taken from
cations which one would have to under- Part IV of Jack D. DOUGLAS, The Social
stand quite well before he could give any Meanings of Suicide, op. cit.
nearly definitive interpretations of the

275