Académique Documents
Professionnel Documents
Culture Documents
http://journals.cambridge.org/EUR
* Edgar MORIN, L'homme et la mort devant Vhistoire (Paris, Correa, 1951), version
revisee et augmentee sous le titre L'homme et la mort (Paris, Seuil, 1976); B. G.
GLASER & A. L. STRAUSS, Awareness of Dying (Chicago, Aldine, 1965); ID. Time for
Dying (Chicago, Aldine, 1968); G. GORER, Death, Grief and Mourning in Contem-
porary Britain (London, The Cresset Press, 1965); Philippe ARIES, La mort inversee.
Le changement des attitudes devant la mort dans les societes occidentales, Archives
europeennes de sociologie, VIII (1967), 169-195 ; ID. The Hour of our Death (London,
Allen Lane, 1981) [transl. from French]; Anne-Marie GUILLEMARD, La retraite, une
mort sociale : sociologie des conduites en situation de retraite (Paris/La Haye, Mouton,
1973); Louis-Vincent THOMAS, Anthropologie de la mort (Paris, Payot, 1975);
L. PRIOR, The Social Organization of Death (Basingstoke, Macmillan, 1989).
172
Arch, europ. sociol., XXXII (1991), 172-196. — 0003-9756/91/0000-591 $02.50 ^) 1991 A.E.S.
SOCIAL DEATH
During the 1960s, two major ethnographies of death and dying were
carried out in the U.S.; namely, that by Sudnow (1967) and that by
Glaser and Strauss (1965; 1968). These studies examined the social
definition of death and the process of dying in American hospitals where
'most people in the United States now die' (Glaser and Strauss 1965,
viii). By the 1980s, the proportion dying in American hospitals had risen
to 80 % (Turner 1984, 127) whilst in Britain 70 % had come to die by
this time in hospitals or in similar settings (OPCS 1989, Series DHI). In
both ethnographies, social death is distinguished from clinical or bio-
logical death and special attention is given to the former. In the words of
Glaser and Strauss: 'Our intent was, above all, to ask whether people can
die socially before they die biologically, and what this means for human
relationships' (1965, x). Sudnow confirms Glaser and Strauss's sugges-
tion that hospital patients can, and indeed frequently do, die socially
before they have been defined as biologically dead (1967, 74).
Like Glaser and Strauss, Sudnow cites Goffman's concept of the
'non-person' as a source for the notion of 'social death'. Goffman defines
'non-persons' as 'standard categories of persons who are sometimes
treated in their presence as if they were not there' (1959, 152). He gives as
173
MICHAEL MULKAY & JOHN ERNST
examples, servants, slaves, the very old, the very young and the sick.
Goffman does not explore this concept in depth. Nevertheless, he seems
to imply that the application of the category 'non-person' by participants
is implicit rather than explicit and that it normally involves no more than
a passing and contextually specific denial of another's social existence.
These kinds of social actors, he maintains in the quotation above, are only
sometimes treated as if they did not exist. In other words, the young, the
old and the sick are not uniformly ignored in all contexts; and on some
occasions their individuality, that is, their existence as persons, may be
recognized by others. Even slaves, whose exclusion from the social world
of their masters is most enduring and most complete, are able to create a
fuller social life among themselves (Patterson 1982, 6).
Goffman's essential insight is that our social existence is not neces-
sarily continuous. As we come to occupy certain social roles, we may
temporarily drop out of the social world that is being created around us.
Furthermore, Goffman's examples indicate that this process of social
deconstruction may often be related in some way to people's biological
condition; for instance, to their age or to their state of health. In certain
circumstances, human beings can come to be regarded as living biological
organisms without social attributes. Glaser and Strauss, and Sudnow,
develop this idea in the context of the modern American hospital. The
concept of 'social death' is designed to express their observation that a
hospital patient can become implicitly a permanent non-person whilst
still alive, as other people foresee the patient's biological death and
change their conduct with the patient's biological termination in mind.
Social death, then, in the hospital setting, grows out of the recognition
that a patient is dying clinically. Sudnow, in particular, emphasises that
the term 'dying' is an 'essentially predictive term [...] Seeing " dying " is
seeing the likelihood of [clinical] death within some temporal perspective
[...]' (1967, 64). Similarly, Glaser and Strauss show that 'death expec-
tations are a key determinant in how everyone acts during the dying
sequence' (1968, 13). In other words, when subjects are seen to be dying,
those with whom they come into contact respond by seeking to impose a
special frame of reference in terms of which further interaction is to be
organized (Sudnow 1967, 68-9). These interpretative frames can vary
considerably. In some instances, as death approaches, the subject's social
life may become more intense and long dormant personal relationships
may be briefly revitalized. But the two ethnographies show clearly that
the more usual situation in the modern hospital is that the patient's
involvement in social life is increasingly restricted by the progressive
withdrawal of other parties. The sequence of physical decline which we
call 'dying' is accompanied by a sequence of social decline which occurs
as other people begin to grieve, to reduce the frequency of personal
contact, to avoid 'difficult' topics, and so on, in anticipation of the
patient's expected demise (Glaser and Strauss 1968, 25). This process is
174
SOCIAL DEATH
175
MICHAEL MULKAY & JOHN ERNST
centrate on those who are still alive (Glaser and Strauss 1965, 249-53).
All of this is well documented in both studies. Nevertheless, we must not,
as analysts, adopt unquestioningly the limited perspective of the medical
staff. We must remember that other parties are normally involved in the
death sequence and that they may well see the process differently.
Furthermore, death sequences are seldom wholly confined to the medical
setting, even in cases where the person dies in hospital. It seems advi-
sable, therefore, to consider the possibility that, for some participants,
social existence may be taken to persist after biological death, and that, in
other settings, the biologically dead may continue to contribute actively
to social life.
Glaser and Strauss themselves describe instances where patients'
social existence seems to continue, at least fleetingly, after death, even in
the hospital setting. They report, for example, that relatives often speak
to and/or caress a patient who has just died. However, the analysts
dismiss this behaviour as abnormal and as requiring a special type of
explanation. They comment that the 'deceased person's awareness is thus
taken into account, in a strange way, by the living, even though, rea-
listically, that problem should disappear after the patient has died' (1965,
114). In other words, Glaser and Strauss treat the living participants as
having made a mistake when they continue to include a dead person
within their social world. Glaser and Strauss suggest that people find it
difficult to adjust to the transition from life to death and that they make
this mistake because they forget momentarily that they are now dealing
with a corpse. Yet, Glaser and Strauss provide no independent evidence
to support this interpretation; and such evidence is particularly necessary
because this is the only point in their analysis where they override and
explain away what appears to be participants' definition of the situation.
In contrast, Sudnow is willing to accept without reservation that social
death my occur 'well after " biological " or " clinical " death' (1967, 75).
Furthermore, he attempts to furnish empirical support for this claim.
Sudnow's evidence, however, is rather weak and unconvincing. In the
first place, the examples he gives are all very short term, involving
minutes, or at most hours, of 'social life' after clinical death. More
critically, all of Sudnow's illustrations contain an element of pretence.
Thus, in some instances, staff reassure anxious relatives that the patient is
still alive when they know that the patient has actually died—for example,
on the operating table. In situations of this kind, the patient clearly
continues to live for those relatives who accept the staff's assurances. Yet
this continued existence is based on an illusion created by others and is,
in this respect, similar to what Glaser and Strauss regard as the 'mistake'
of talking to a corpse. In addition, this form of 'social life' is necessarily
brief and can only last while the pretence of the patient's clinical survival
can~be maintained.
176
SOCIAL DEATH
177
MICHAEL MULKAY & JOHN ERNST
178
SOCIAL DEATH
dead is the physical separation of men from women. Sudnow points out
that the 'sexual segregation of body care found in the handling of dead
bodies is thus similar to that which governs some aspects of the care of
live ones' (1967, 78). However, these socially denned and socially
differentiated forms of handling do not constitute social life in the sense
with which we are employing this term. For the hospital rules do not
define the corpses as active agents. On the contrary, the rules for
handling corpses are general procedures which deal with a broad dis-
tinction between male and female bodies and which involve no necessary
recognition of the individuality of any specific corpse. They are intended,
in the hospital context, to regulate the conduct of the living, not to
articulate a social relationship in which the dead themselves participate.
Similar considerations apply in relation to the 'right to proper burial'
or the 'right to rest in peace'. Assertions of such rights may grow out of a
continuing relationship between living persons and dead individuals.
They may, therefore, be used by the living to defend or, in some other
way, give expression to that relationship. But neither recognition of these
rights, nor acceptance of the obligations they entail, necessarily implies
involvement in such a relationship. In other words, it is possible to fulfil
these obligations and to implement these rights without attributing to the
dead any vestige of social existence. Social life, then, depends, not on the
existence of procedural rules or general obligations concerning treatment
of the dead by the living, but on the continuation of individual actors in
the experiences and activities of the living. Social death consists of the
cessation of these experiences and activities.
This completes our preparatory examination of the concept of 'social
death'. Sudnow, in the course of his analysis, warns against the temp-
tation to define this phenomenon in too loose a fashion, so that it becomes
equivalent to 'any instance of radically asocial treatment of a person'
(1967, 75). He prefers to avoid such a usage because it is bound to be
'analytically ambiguous'. We have attempted, to some degree, to follow
Sudnow's advice. We have tried to be careful and reasonably precise in
our claims and to restrict the scope of our definition of 'social death'. We
do not imagine, however, that we have managed completely to avoid
'analytical ambiguity'. It is comforting to note, therefore, that Sudnow
himself, within the same paragraph as his warning, admits that his own
analysis is not entirely free from this defect (1967, 75). In our view,
fruitful investigation often requires a degree of conceptual looseness.
Thus Patterson's (1982) imaginative extension of the concept 'social
death' to the topic of slavery has shown that a more flexible use of this
idea can help to elucidate a much wider range of social activities than
Sudnow's narrowly ethnographic approach was able to envisage.
Nevertheless, we have, like Sudnow, linked our notion of 'social death'
firmly to participants' recognition of 'biological and/or clinical death'. We
179
MICHAEL MULKAY & JOHN ERNST
will now try to show that our revised conception of 'social death' can be
used to throw new light on social responses to the changing pattern of
biological death in modern society.
180
SOCIAL DEATH
1980, 88 % of men in this age group were actually retired from work
(Turner 1984). One consequence of this situation has been the formation
of an elderly population whose members are socially and economically
disadvantaged (Turner 1984, 117). Not only are the elderly largely
excluded from the workforce, they are also denied full participation in the
family life of their own descendants and in the wider community (Turner
1984). As an indication of the social isolation of the elderly, we can
consider the extent to which they are visited by friends and relatives. In
Greater London, for example, in the 1970s, only a quarter of elderly
people received a visit more than once a week; and one in six were never
visited at all. Even in less urban areas such as Yorkshire, only four out of
ten elderly people were visited more than once a week (Hunt 1979).
The general situation, then, is that as people become older and the
likelihood of their entering a medically denned dying sequence increases,
they are formally retired from paid employment and the social contacts of
their mature years are severed. Retirement from work means reduction of
income, often to very low levels, which in turn further restricts their
social activities. Family and communal involvement is reduced and
elderly people are likely to find themselves increasingly subject to a
general sense of physical aversion which is akin, in a milder form, to the
revulsion caused by dead bodies. From the time of retirement, elderly
people in Britain and the U.S. are channelled collectively away from the
main areas of social activity and their social ties with the wider society are
progressively weakened in anticipation of their biological end.
As we grow old we are gradually marginalized within the community and begin to
lose a number of personal contacts, which takes the form of a personal or psychologi-
cal contraction (dying) from social relations. Through the process of dying our social
contacts are gradually diminished and we find ourselves socially isolated. Finally, this
process is terminated by a biological death which brings to a conclusion the long
history of our personal disengagement. We can therefore regard dying like aging as a
gradual attrition of social relations combined with increasing dependency often on
institutions which are somewhat anonymous and bureaucratic (Turner 1984, 125-6).
181
MICHAEL MULKAY & JOHN ERNST
on, or take their meaning from, any understanding of a long term future'
(1967, 70-1). In other words, the death sequences that come to envelop
elderly people within the hospital setting are accentuated versions of a
broader social death sequence within which such people are already
located.
In England today, over 60 % of elderly people die in hospitals (OPCS
Series DHI no. 2) where the bureaucratic 'neglect of the patient as a
person' (Field 1989, 147) converges, to varying degrees, with the phy-
sical, emotional and communicative withdrawal of the living from those
on the death sequence. A smaller, but significant, proportion of the
elderly end their days in residential and nursing homes. In 1987, 8 % of
all deaths among the elderly in England and Wales occurred in such
homes; in the case of women, the proportion was 11 % (OPCS Series DHI
no. 2). These 'homes' provide an institutionalized form for the final
separation of people on the death sequence from 'ordinary' social actors.
They are the culmination of the process whereby the elderly are pro-
gressively excluded from the ongoing social world.
Residential homes have become an increasingly important location for
the enactment of the death sequence among the elderly. During the last
decade or so, recourse to residential care has become common practice for
people over 85 years, and amongst those who are defined as having 'high
dependency' needs. Between 1977-87, the number of elderly persons
living in non-hospital institutional care increased by 41 %. This growth
seems likely to continue for the foreseeable future (Sinclair 1988; Health
and Personal Social Services Statistics for England 1989). In short, the
bureaucratic management of the death sequence is still expanding.
Despite the changing structure of residential care in England—with
greater numbers of recipients, higher levels of government finance, and a
marked increase in private, commercial forms of provision—there is little
indication that institutional regimes are appreciably different from the
'impersonal, dehumanising and controlling model' of elderly residential
care first identified by Goffman and Townsend in the 1960s (Hughes and
Wilkin 1987, 420). Although physical conditions have generally improved
since then, treatment in many contemporary homes for the elderly is still
designed to suppress residents' individuality, initiative and social enga-
gement (Booth 1985; Wilkin, Hughes and Jolley 1985; Hughes and
Wilkin 1987; Sinclair 1988). Whereas in some pre-literate societies the
unwanted old are buried alive (Goody 1962, 44), in our society they are
immured in residential homes where, out of sight and largely out of
mind, they can be left whilst the process of biological decline takes its
inevitable course.
When elderly people enter a residential or nursing home, all those
involved are aware that the entrants are most unlikely to come out alive
(Hockey 1985). As Glaser and Strauss report, when geriatric patients are
sent to an age-segregated institution, they know that 'it is all over but the
182
SOCIAL DEATH
dying' (1968, 52). For those left behind, the symbolic significance of this
transition is so evident that in many cases they engage in what Fulton and
Fulton (1971) call 'anticipatory grief. In such circumstances, persons are
mourned as if they were biologically dead, even though they may
continue to live on in the home for years. Such pre-emptive mourning
helps those outside the home to deal with the crisis of separation. It
enables them to begin to adjust in advance so that, by the time death
comes, they are already 'grieved out' (Glaser and Strauss 1968, 151-4).
As a result of this process, however, they become more distanced from
the person in the home, who is pushed further along the social death
sequence as his or her past relationships are disrupted and lose their
meaning. Outsiders will often, of course, try to maintain contact with
elderly residents. But the symbolic divide between the full social actors
on the outside and the physically and socially segregated insiders makes
such relationships little more than pretence. For, from the outside, the
residents are, owing to their inability to perform meaningful social action,
already dead in social terms.
This does not mean that, internally, residential homes for the elderly
are completely devoid of social life. What may often happen in such
settings is that the limited range of activities available to residents
becomes socially structured in relation to a basic distinction between
those who are deemed to be close to death and those for whom death is
not thought to be imminent. In situations of this kind, the more active
residents strive to re-establish the meaning of their own existence by
separating themselves physically and symbolically from other residents.
This is well documented in Hockey's (1985) study of a residential home
in North Eastern England.
In this home, the 'fit' residents were distinguished by staff and by
other residents from the 'frail'. These two categories of residents were
kept apart spatially and were treated in quite different ways. In parti-
cular, in their dealings with the 'fit', staff stressed their liveliness and
employed a linguistic repertoire which denied the immediate relevance of
death. In contrast, in their treatment of the 'frail', staff presumed the
imminence of biological death and used language forms which referred to
death directly as well as by implication. These categories were main-
tained despite the fact that the 'fit' often died unexpectedly and the 'frail'
sometimes lived on well beyond their allotted span.
For the staff, the distinction between the 'frail' and the 'fit' enabled
them to focus their attention on the former and, thereby, to sustain a
sense that the home was more than a 'house of death'. Although, from the
outside, such homes are seen as locales for the socially dead, on the
inside, residents may be defined in ways which emphasise the conti-
nuation of social life. This is particularly important for the newer and
'fitter' residents who are even more committed than staff to the spatial
and social segregation of the 'frail'. The consequence for the 'fit' is that
183
MICHAEL MULKAY & JOHN ERNST
they can, in this way, retain some vestigial sense of personal worth and
social significance. The consequence for the 'frail' of this collusion
between the 'fit' and the staff is, of course, that they cease to exist as
social actors. The ultimate outcome for the 'frail' is that they are sys-
tematically prevented from participating in this last, culturally impov-
erished setting in much the same way that the elderly in general are
excluded from the wider society. For other participants, the 'frail' tend to
become invisible. 'As a result, the final transition of a " frail " resident to
sick bay and to the mortuary often goes by unremarked. A much earlier
separation ensures that their name and identity is unknown and of little
significance' (Hockey 1985, 42).
How far Hockey's conclusions apply in detail to other residential
homes is not yet clear; although the findings of other, more general
studies suggest that the regime observed by Hockey is unexceptional
(Hughes and Wilkin 1987). In so far as her study is representative, it
shows that, within our residential homes for the elderly, the death
sequence of society at large is repeated in microcosm. It is ironic that,
even inside these institutions which exist to cater for the socially dead,
participants find it necessary to deny death's presence (Aries 1981) and,
in so doing, to condemn others, and in due course themselves, to yet
another manifestation of social death.
Between the early 1900s and the early 1980s, the life expectancy of
females born in England increased by 25.3 years, compared to an increase
of 23.3 years for males (McPherson and Coleman 1988). In the U.S.,
over the same period, the difference between the life expectancy of white
males and white females increased from 2.1 years to 6.9 years (Kearl
1989, 132). Thus women have benefited rather more than men from the
great expansion in the length of human life that has occurred during this
century. Furthermore, not only do women now experience a greatly
extended lifespan, but they also spend, on average, much less of their
lives than in the past bearing and giving birth to children (Turner 1984).
In modern Britain, the vast majority of these children will survive to
adulthood and will, indeed, outlive their parents. Throughout the 1800s,
in contrast, a large proportion of children were destined to die during
infancy. Thus in Liverpool, in 1899, the number of children dying before
the age of five was 136 per 1,000 in the more prosperous wards of the
city. 'In the poorer areas it stood at 274 and in the meanest of streets, it
was a catastrophic 509 per 1,000' (Walvin 1983, 23). Since then, infant
mortality has been reduced to a remarkably low level. In England and
Wales, for example, the number of children dying in their first year
has fallen dramatically from 154 per 1,000 live births in 1900 to 9 per
184
SOCIAL DEATH
i,ooo live births in 1985 (Mitchell 1982; Halsey 1988). The consequence
for British women has been that, unlike their Victorian counterparts, they
are very unlikely to spend a significant part of their child-bearing years
dealing with the deaths of their own children.
In addition, the general increase in life expectancy means that the
deaths of parents and of other close relatives are now likely to occur much
later in a woman's life cycle than in the past (Jalland and Hooper 1986).
As a result, during the main years of their maturity, women are now
much less involved with death as a normal part of family life. This is, of
course, also the case for men. But the social impact of these demographic
changes has been much greater for women owing to their central role in
domestic activities, including those ritual activities associated with death.
In the last century, both children and adults almost always died at home
in the family setting (Walvin 1986). For example, in 1851 only 15.5 % of
deaths occurred in public institutions (Registrar General 1855). Within
the domestic setting, the rituals of bereavement were overwhelmingly the
responsibility of women and it was on women 'that the external signs of
grief were hung' (Morley 1971, 63).
Men passed almost the whole burden of mourning to their womenfolk,
who were required to mourn relations by marriage in the same manner as
relations of blood. In what we may loosely call the 'Victorian middle
class', the wife would be expected to mourn an aunt or uncle for three
months, a grandparent for nine months, and a son or daughter for one
year. The most lengthy and elaborate of all mourning sequences was that
of a wife for a husband. Mourning dress of an exaggerated severity was
worn for the first two years as a sign of 'inner desolation'. Black,
light-absorbing crape was used to express the widow's grief and her
withdrawal from the world. For the first year of bereavement, the widow
could undertake no social activity outside the home. After this she could,
if she wished, gradually resume her place in society. It was not unknown,
however, for the bereaved woman to remain, like Queen Victoria herself,
in more or less permanent seclusion (Gorer 1965, 79; Morley 1971, 68).
When close relatives and particularly when her husband died, the
middle class Victorian woman was required to enter a formal social death
sequence; that is, the range of her conduct was drastically curtailed, she
was excluded from many of her normal social activities, and indeed she
continued to exist only in the most distant fashion for many of the people
in her customary social world. The conventional response of the Vic-
torian woman to death in her immediate family was a kind of surrogate
death of her own. It is interesting to note that servants, those prototypical
non-persons mentioned by Goffman, were also put into mourning when a
member of the family died (Morley 1971, 72). But the adult women of the
family suffered a much more complete restriction of their social being.
Given the high adult and infant mortality rates of the time, the great
majority of middle class women would have spent many years, in total,
185
MICHAEL MULKAY & JOHN ERNST
enclosed within the ritual death sequences through which expression was
given to the collective bereavement of their families.
Among poorer people, for obvious economic reasons, this extended
ritual confinement and symbolic adornment of women was impossible
(Walvin 1986). For many Victorian women, the labours required to
maintain a daily subsistence for the living had to take precedence over
ritual labours for the dead. Nevertheless, the middle class pattern of
activity in relation to death 'percolated to the lower classes' (Morley
1971, 17) and furnished a model of proper conduct to which many less
wealthy people aspired. Thus, members of the labouring classes put aside
more money for 'decent burial' than for any other purpose (Morley 1971;
Walvin 1986). The essential feature of such burial, as opposed to disposal
in a pauper's grave, was that it provided an identifiable site for the dead
person which could be the focus for communal and family mourning. As
in the case of the middle class, the responsibility for such mourning lay
heavily upon the women of the working class.
The central involvement of women in mourning and the partial social
death of bereaved women required by the rules of Victorian etiquette was
not only a symbolic expression of loss. It also served as a mechanism
whereby the social existence of deceased family members was extended.
For one of the primary tasks of bereaved women was to keep alive the
memory of the family dead by means of regular prayers, by continuing to
celebrate their birthdays and other anniversaries, by visiting their graves,
and so on (Morley 1971; Walvin 1986). These activities were given
significance for the great majority of women by belief in an afterlife and
by hope of reunion in a world beyond the grave. 'The hopes of all,
whether unsophisticated, or sophisticated, were fixed on Heaven' (Morley
1971, 102). Women were required to spend so much of their lives in the
borderland between life and death because they were responsible, within
the family setting, for maintaining strong social bonds between the living
and the constantly increasing number of the dead.
Most of this now, of course, has gone. Belief in the afterlife has greatly
declined and with it the rituals of mourning and the other forms of
associated conduct (Gorer 1965; Morley 1971; Turner 1983). In these
respects, modern women may appear to live both longer and freer lives
that are less oppressed by the deaths of others. Yet social death is still a
more significant feature of women's lives than of the lives of men;
although its form now is very different. Because women outlive men, on
average, by a considerable margin, most married women alive today are
destined to become widows (Kearl 1989). Rees's study of a rural
community, in which the ratio of widows to widowers was more than
three to one, is not unrepresentative. Thus men are much less likely than
women to be left alone as biological death draws near. This suggests that
the most extreme form of social death for the majority of women in
present day society, as in Victorian times, is likely to occur after the death
186
SOCIAL DEATH
of their husband. Rees's study showed that many of the widows in his
population during the 1960s did maintain some kind of social life with
their dead partners. But these 'hallucinations of widowhood' experienced
in private by elderly, isolated women are no more than pale shadows of
the socially supported relationships with the dead that gave meaning to
widowhood in the Victorian era.
We observed in the previous section that retirement from paid
employment is a major transition in men's lives and that this enforced
social disengagement marks for men the beginning of the death sequence.
This change in men's status is of considerable importance for their wives,
who in many instances will also retire from the official workforce as the
family cycle runs its course. Thus both women and men experience more
or less together the marked restriction of social activities that is charac-
teristic of all social death sequences. However, the change is less dramatic
and less consequential in the short run for married women owing to the
fact that, with few exceptions, their unpaid labour in the home continues
undiminished after their husband's or their own retirement. Accordingly,
women's entry to the death sequence will normally involve less dis-
ruption to the established pattern of their lives and a less significant
reduction in their social activities.
Given that caring for their menfolk provides a central focus for the
daily activities of elderly women, the character of the death sequence
changes most significantly for them when their partner dies. Referring to
that generation of American women who are now very old, Kearl writes
that: 'So totally dependent are their identities on those of their spouses
that when their husbands die these females lose a significant portion of
their very self-concept' (1989, 485). With the advent of widowhood
comes a radical alteration in women's social existence and, in many cases,
a great acceleration in their move toward social death. It may well be that,
for many women, this is where the social death sequence really begins.
Modern widows, of course, are not necessarily old. Nonetheless, the
concentration of biological death among the elderly means that widows
today will be significantly older collectively than in Victorian times. Yet,
owing to their greater longevity, they will often have many more years to
live. Writing about the American situation, Kearl estimates that the
average duration of widowhood may now be close to twenty years.
These long years of widowhood are not subject to the strict control
which drastically limited women's social activities in the last century, yet
which, at the same time, gave them social significance. They are, rather,
likely to be years of increasing personal emptiness and declining social
involvement. It may be that, for a proportion of the women now
approaching this major change in social status, the consequent reduction
in social participation will be less complete and less personally damaging
than in the past. 'For more recent cohorts of women [...] individuals may
have more anticipatory socializations for widowhood because of earlier
187
MICHAEL MULKAY & JOHN ERNST
188
SOCIAL DEATH
189
MICHAEL MULKAY & JOHN ERNST
Victorian times are now experienced in a society which has yet to develop
a social death sequence that is appropriate to the new demographic
situation.
In Britain today, excluding stillbirths and neonatal deaths, accidents
are the most common cause of death among children (Foster and Smith
1987). Such deaths are unexpected, not only in the sense that each
particular death occurs without warning, but also in the sense that
modern parents do not see childhood as a time when death is likely. Thus
parents often react initially with disbelief as well as with shock, anger and
dismay (Foster and Smith 1987). The emotional distress of the bereaved
is so great, partly because the child's firmly anticipated future in this
world is removed instantaneously, but also because expectations of life
after death leading to eventual reunion are either entirely absent or are
held in most cases with less conviction than in the past. In the great
majority of accidental deaths in childhood, the bereaved are deprived of a
graduated social death sequence involving the lost child, either before or
after the child's death, in the course of which they might adjust to their
grief and to the destruction of the personal world in which the child had
played such an important part (Foster and Smith 1987).
We cannot know whether the experiences of parents today upon the
death of a child are any more or less severe than in the last century.
Nevertheless, it is clear that present day parents have no well-established
patterns of social conduct and associated belief to help them through the
traumatic events, centred around and processed by the medical
bureaucracy, which follow the accidental death of a daughter or son.
Acquaintances, friends, even other members of the family network are
likely to respond by pointedly avoiding the topic and also by reducing the
frequency of their contact with the bereaved (Gorer 1965; Foster and
Smith 1987). As a result of the death of child, the social activities of the
parents will tend to become restricted in a way that is analogous to,
though less formalized than, the restrictions imposed on the grieving
Victorian mother. In other words, on the sudden death of a child, parents
today are likely to undergo a form of social death sequence by association.
In the modern world, however, this sequence has no wider symbolic
or social significance. Parents are no longer ritually separated from the
living in order to keep alive the memory of, or to maintain contact with,
the newly dead. They are, rather, set apart to forget, and to prepare to
replace, the dead (Foster and Smith 1987). Their social exclusion is, like
that experienced by the very old, but another expression of the customary
flight from death in our society by those who are not directly affected
(Kearl 1989). The consequence is that the parents, unprepared and
lacking social support, typically respond to their bereavement and to the
unexpected social isolation that follows with bewilderment accompanied
by a devastating sense of personal loss.
I thought there was no way we would get over the dreadful pain and awful hurt
190
SOCIAL DEATH
inside, which you can't explain to people. And it was total darkness even in the middle
of the day when there was bright sunshine. In my mind it was just total darkness
(A bereaved parent quoted in Foster and Smith 1987, 18).
One basic feature or such accidents is that death takes the survivors by
surprise. One source of their pain is that they have had no time in which
to readjust their relationship with the child whilst the child was still alive
(Foster and Smith 1987). In the case of deaths from cancer, however,
which is the second most frequent cause of death among the young in
contemporary Britain, time for readjustment is available. Childhood
death from cancer is a comparatively slow process. In this respect, it
resembles the typical death sequence of the elderly. Indeed, as the disease
progresses, children dying from cancer come increasingly to look and act
like the aged.
Dying children are more like the elderly than their own healthy peers—without
futures, worried, often passive, unhappy, burdened with responsibilities for others
and their feelings. They even resemble old people—either bald or dishevelled,
emaciated or bloated, incapacitated, generally sickly, and most of all losing, failing
with time (Bluebond-Langner 1978, 213).
191
MICHAEL MULKAY & JOHN ERNST
192
SOCIAL DEATH
We have seen, then, that the death of young persons in our society
causes special problems, even in situations where the process of dying is
slow and where the eventual outcome is recognized in advance by all
concerned. The underlying reason for this is that the socially organized
death sequences focussing upon the young which were appropriate in a
society characterized by a high mortality rate among children, and which
provided support and compensation for those involved, have disappeared
almost completely as the frequency of such deaths has fallen to remark-
ably low levels. In our society, as the concentration of death among the
elderly has become firmly established and has come to be taken for
granted, so the death of the young has come to be almost unthinkable as
part of the ordinary course of events.
Yet, in a small percentage of young lives, biological death still comes
desperately early. Our customary response when such a death approaches
is akin to constructing a social death sequence in reverse; that is, instead
of initiating an anticipatory process of social decline, we try hard to create
a sense of continuing social involvement for the stricken child. We try to
pretend, for the child but also for ourselves, that the unthinkable is not
happening and to draw the child with us into a world of make-believe
where children never die. In this way, we seek to delay the child's social
death until pretence and denial are no longer possible.
Concluding remarks
We have tried to show that the concept 'social death' can be used to
elucidate major historical developments in society at large. In order to
explore the complex historical changes that have occurred in people's
conduct associated with death and dying, we have employed two kinds of
systematic comparison. First, we have compared demographic profiles of
biological death with profiles of social death. Second, we have compared
the profiles of biological and social death in modern Britain with those of
Victorian times. Although we have furnished much detailed documen-
193
MICHAEL MULKAY & JOHN ERNST
194
SOCIAL DEATH
REFERENCES
ANDERSON, M., 1985, The emergence of the in Everyday Life (Garden City, New York,
modern life cycle in Britain, Social History, X, Doubleday).
69-87. GOODY, J., 1962, Death, Property and the
ARIES, P., 1981, The Hour of Our Death Ancestors (London, Tavistock).
(London, Allen Lane) [transl. from French]. GORER, G., 1965, Death, Grief and Mourning
BLUEBOND-LANCNER, M., 1978, The Private in Contemporary Britain (London, The Cresset
Worlds of Dying Children (Princeton, Prince- Press).
ton University Press). GUILLEMARD, A.-M., La retraite, une mart
BOOTH, T., 1985, Home Truths: old people's sociale : sociologie des conduites en situation de
homes and the outcome of care (Aldershot, retraite (Paris/La Haye, Mouton, 1973).
Gower). —, La vieillesse et I'Etat (Paris, P.U.F.,
CLEGG, F., 1988, Decisions at a Time of Grief 1980).
(University of Hull, unpublished). — et Remi LENOIR, Retraite et echange
cso (Central Statistical Office) 1989. Annual social : tentative d'explication des systemes de
Abstract of Statistics (London, HMSO). relations sociales en situation de retraite (Paris,
FALKINGHAM, J., 1989, Dependency and C.E.M.S., 1974).
ageing in Britain: a re-examination of the HALSEY, A. H. (ed.), 1988, British Social
evidence, Journal of Social Policy, XVIII (2), Trends since igoo (Basingstoke, Macmillan).
2H-33- Health and Personal Social Services Statistics
FIELD, D., 1989, Nursing the Dying (Lon- for England igSg (London, HMSO).
don, Tavistock/Routledge). HOCKEY, J., 1985, Cultural and social
FOSTER, S. & P. SMITH, 1987, Brief Lives interpretations of 'dying' and 'death' in a
(London, Arlington Books). residential home for elderly people in the
FULTON, R. & J. FULTON, 1971, A psy- North East of England, Curare, VIII (1),
chological aspect of terminal care: anticipatory 35-43-
grief, Omega, II, 91-100. HUGHES, B. & D. WILKIN, 1987, Physical
GLASER, B. G. & A. L. STRAUSS, 1965, care and quality of life in residential homes,
Awareness of Dying (Chicago, Aldine). Ageing and Society, VII, 399-425.
—, 1968, Time for Dying (Chicago, Aldine). HUNT, A., 1978, The Elderly at Home: a
GOFFMAN, E., 1959, The Presentation of Self study of people aged 65 and over living in the
195
MICHAEL MULKAY & JOHN ERNST
196