//img.uscri.be/pth/93b267b5bccec66c3a9bb58fff29921baa5aeb0b
Cet ouvrage fait partie de la bibliothèque YouScribe
Obtenez un accès à la bibliothèque pour le lire en ligne
En savoir plus

The sibling tie in adulthood : an analysis based on meeting frequency - article ; n°1 ; vol.43, pg 41-65

De
26 pages
Revue française de sociologie - Année 2002 - Volume 43 - Numéro 1 - Pages 41-65
What can be said of the sibling tie in adulthood ? Analysis of a survey conducted by Insee of meetings between or among brothers and sisters (survey sample : 6,000 French households) points to three main conclusions. First, there is little that is normative about the sibling tie: number of meetings varies greatly by individual. In contrast to direct filiation, relations between siblings seem determined by choice and interest more than status-conditioned obligations. Second, inclination is nonetheless more likely to regulate sibling ties for men; for women, regulation is more likely to be status-conditioned. Third, the sibling tie is structurally secondary to direct filiation (mother and father-adult children) : how often siblings see each other is a function of father's and/or mother's presence, and declines when individuals settle into couples and have their own children. The fact that the sibling tie functions as a substitute when direct filial ties are impoverished or lost further illustrates that, of the two direct blood ties, the sibling tie is structurally secondary. These properties of siblingship are probably particular to the cognatic or non-unilineal, kindred system which is that of modern Western kinship.
25 pages
Source : Persée ; Ministère de la jeunesse, de l’éducation nationale et de la recherche, Direction de l’enseignement supérieur, Sous-direction des bibliothèques et de la documentation.
Voir plus Voir moins

Emmanuelle Crenner
Jean-Hugues Déchaux
Nicolas Herpin
Amy Jacobs
The sibling tie in adulthood : an analysis based on meeting
frequency
In: Revue française de sociologie. 2002, 43-1. pp. 41-65.
Abstract
What can be said of the sibling tie in adulthood ? Analysis of a survey conducted by Insee of meetings between or among
brothers and sisters (survey sample : 6,000 French households) points to three main conclusions. First, there is little that is
normative about the sibling tie: number of meetings varies greatly by individual. In contrast to direct filiation, relations between
siblings seem determined by choice and interest more than status-conditioned obligations. Second, inclination is nonetheless
more likely to regulate sibling ties for men; for women, regulation is more likely to be status-conditioned. Third, the sibling tie is
structurally secondary to direct filiation (mother and father-adult children) : how often siblings see each other is a function of
father's and/or mother's presence, and declines when individuals settle into couples and have their own children. The fact that the
sibling tie functions as a substitute when direct filial ties are impoverished or lost further illustrates that, of the two direct blood
ties, the sibling tie is structurally secondary. These properties of siblingship are probably particular to the cognatic or non-
unilineal, "kindred" system which is that of modern Western kinship.
Citer ce document / Cite this document :
Crenner Emmanuelle, Déchaux Jean-Hugues, Herpin Nicolas, Jacobs Amy. The sibling tie in adulthood : an analysis based on
meeting frequency. In: Revue française de sociologie. 2002, 43-1. pp. 41-65.
http://www.persee.fr/web/revues/home/prescript/article/rfsoc_0035-2969_2002_sup_43_1_5565R. franc, sociol, 43, Supplement, 2002, 41-65
Emmanuelle CRENNER
Jean-Hugues DÉCHAUX
Nicolas HERPIN
The Sibling Tie in Adulthood: an Analysis
Based on Meeting Frequency
Abstract
What can be said of the sibling tie in adulthood? Analysis of a survey conducted by
Insee of meetings between or among brothers and sisters (survey sample: 6,000 French house
holds) points to three main conclusions. First, there is little that is normative about the s
ibling tie: number of meetings varies greatly by individual. In contrast to direct filiation, re
lations between siblings seem determined by choice and interest more than status-
conditioned obligations. Second, inclination is nonetheless more likely to regulate sibling
ties for men; for women, regulation is more likely to be status-conditioned. Third, the s
ibling tie is structurally secondary to direct filiation (mother and father-adult children): how
often siblings see each other is a function of father's and/or mother's presence, and declines
when individuals settle into couples and have their own children. The fact that the sibling tie
functions as a substitute when direct filial ties are impoverished or lost further illustrates
that, of the two direct blood ties, the sibling tie is structurally secondary. These properties of
siblingship are probably particular to the cognatic or non-unilineal, "kindred" system which
is that of modern Western kinship.
Sociologists and anthropologists of kinship in modern societies have only
rarely studied relations between/among siblings in adulthood. This is surpris
ing, especially since the tie between brother(s) and/or sister(s) is one of the
two that constitute the elementary or nuclear family, the basic unit of the
Western kinship system (the other being of course that of direct filiation, fa
ther and mother-children). In modern Western societies, as among the Eski
mos, families" the kinship network can be described as "interlocking elementary
(Ghasarian, 1996, p. 48). Each individual belongs to two such el
ementary or nuclear families: the one into which he or she was born -the fami
ly of orientation- and the one he or she forms with his or her spouse (or
opposite-sex partner) W and children: the procreation family.
(l)The number of unmarried couples in known as the PaCS {Pacte civil de solidarité),
France is high. Living together in what is known which permits all unmarried couples, hetero-
as a union libre is widely accepted and practiced: sexual and homosexual, to officialize their
four out of ten children are born to unmarried situation and thereby acquire some of the legal
couples. In 2000 the French parliament officially rights of married couples,
recognized this arrangement by passing a law
41 Revue française de sociologie
The ties that make up the elementary family are direct consanguinity ties;
there are no intermediate ones. Brother and sister, like father, mother, son,
and daughter, are "primary" relatives. This is why these ties are the most
likely to be maintained throughout one's life. In our society, relations with the
nuclear family of orientation constitute the core of the kinship network opera
tive in adulthood. We know how central the role of relations between father
and mother and adult children is, but we know almost nothing of the relations
that may exist between siblings, even though, of all primary kinship ties, this
one lasts the longest because of closeness in sibling ages. The "structural" po
sition of the sibling tie within the kinship network is thus a very strong one,
and may be expected to endow it with a key role. Analyzing the organization
of the Western kinship system as a whole requires studying the sibling tie.
The richest contributions on the question date from the 1950s and 60s,
when anthropologists of kinship, until then specialized exclusively in the
study of unilineal systems, became interested in bilateral or undifferentiated
systems (those that do not take sex into account in defining kinship ties), of
which Western kinship is one. The most stimulating theoretical text is without
a doubt Cumming and Schneider's 1961 article, "Sibling solidarity: a property
of American kinship", published in American Anthropologist. Taking up an
idea launched by Pherson (1954) and on the basis of a 220-respondent survey
conducted in Kansas City, Cumming and Schneider claimed that the sibling
tie is the archetypal kinship tie in undifferentiated systems characterized by
egalitarian kinship relations -the American system, for example. The tie this
study found to exist between adult siblings is strong and by soli
darity, namely at the level of affect; but it is also marked by a high degree of
autonomy, choice, and egalitarianism. It, rather than the more normative and
hierarchical direct filial tie, seemed to the authors particularly in step with the
cultural orientations of modern society, infused as they are with individual
ism. In sum, the "modernity" of the sibling tie could be opposed to the
"passe" quality of filiation.
This thesis was not really tested. Virtually no further surveys bearing on
the theme were conducted, even in the United States, where analysis of the
sibling tie had in fact long been neglected (Irish, 1964). In the last few years,
North American sociologists have shown renewed interest in the study of sib
ling groups. As we shall see, a number of American statistical surveys have
essentially confirmed Cumming and Schneider's 1961 interpretation. At the
present time, therefore, there is consensus among North American specialists
around the idea that the sibling tie is both strong and determined by individual
initiative rather than status-determined norms.
Without prejudging the validity of Cumming and Schneider's thesis for the
French case, we can use it to examine the reality of French adult sibling ties in
contrast to two other types of social tie: direct filiation, since siblingship is
the relation established with one's other primary blood relatives, and
friendship ties, since the weakness of status-determined norms and strength
of personal inclination and affinity that characterizes siblingship should logi
cally, following Cumming and Schneider, make it similar to friendship. Our
42 Emmanuelle Crenner, Jean-Hugues Déchaux, Nicolas Herpin
approach may be described as structural in that by seeking to identify the fac
tors that determine how often siblings see each other, we are trying to grasp
the place of the sibling tie in the overall organization of kinship, and the
properites deriving from that place. (2) This approach in no way requires us to
concur with structural-functionalist anthropology, to see kinship as a sort of
collective individual, a single, indivisible entity. Instead it enables us to see it
as a network of dyadic relations, each of which can only be defined in relation
to the others. For this reason, we shall consider the expressions "kinship" and
network" synonymous. "kinship
entraide'" We shall be using iNSEE's "Réseau de parenté et survey [Kinship
networks and mutual support], conducted in October 1997 on a sample of
6,000 French households (see Appendix for presentation of survey). The pre
sent article is only the first of several that aim to analyze the sibling tie in
contemporary French society; here we will only be discussing results relative
to how often and under what general circumstances siblings see each other;
questions of mutual aid and support will be considered in a later study.
Siblingship is not a strongly normed tie
Three general observations highlight that the sibling tie is only weakly nor
mative. First, number of meetings varies greatly by individual and individ
ual's family and social situation. Second, there is no difference between men
and women as to overall number of meetings, whereas traditionally, for kin
ship relations, this difference is marked and indicates that for women there are
heavy status-determined obligations. Lastly, the longer siblings have lived to
gether the more likely it is that they will meet (personal affinity ties are
forged in a situation of cohabitation), with the exception of dyadic sibling
groups, where sibling relations seem more subject to norms.
(2) Clearly we are not using the term "struc- of analysis that is careful to account for the
tural" in the same way as Lévi-Strauss, for organization and functioning of kinship as a
whom kinship structures involved analysis of structure, the way components work together,
marriage ties, specifically the rules governing and the internal coherence of the whole. Struc-
matrimonial exchange but rather in the sense tural analysis of siblingship thus aims to
given it by ethnologists of kinship, following identify what properties characterize the sibling
Radcliffe-Brown. Authors as different as tie, with the idea that these depend on its
Davenport, Freeman, Leach, Murdock, and position in the overall organization of kinship.
Parsons have used the same term, though they Using the term this way in no way requires us
did not necessarily have the same theoretical to adopt what are today largely discredited
"structural" does not structural-functionalist theories of kinship, approach. In this sense,
designate a method or theory, but simply a type
43 Revue française de sociologie
Variability of relations and weakness of status-determined obligations
Despite the collateral shrinking of kinship networks due to the continual
decrease in French fertility rates over the last 30 years, it remains a common
thing today to have an adult sibling living outside the household. Sev
enty-four percent of individuals aged 15 or over have at least one brother or
sister who doesn't live with them, and that figure is over 84 % for persons
aged 25 to 59 (Crenner, 1998). On average, one has two siblings outside the
household, but this masks great dispersion: one in four of those surveyed have
none, and 20 % have more than three. (3)
In general, one is less likely to see one's sibling(s) than one's other pr
imary blood relations: 35 meetings a year, as opposed to 86 with mother, 85
with parents, and 69 with father. But given how widely the figure varies by in
dividual, this average does not have much meaning. Statistically, meeting fr
equency figures are highly dispersed: 1 5 % of respondents had not met their
brother and/or sister in the course of the year while 10% had seen
him/her/them once a week. This seems at first glance to confirm the notion
that this tie is not highly normed. In fact, absence of all contact with siblings
(15 %) is much higher than for mother (6 %) or father (8 %). This indicates
that individuals feel less compelled to maintain contact with their closest
collaterals than with their parents (direct filiation).
Sibling relations, then, would very much seem to be what individuals want
them to be (Cumming and Schneider, 1961; Muxel, 1998). This high variabil
ity, amplifying a common tendency in modern Western kinship relations, was
underlined several times in earlier surveys (Adams, 1968; Firth, Hubert, and
Forge, 1969; Allan, 1977). Compared with relations with father and mother or
adult children, the status-determined obligations (of contact and support) are
much less strong for siblings, and relations among them, which are more vari
able, are also more up to the individual's discretion; that is, more determined
by choice and interest.
Geographical distance strongly influences how often siblings see each
other, more so than it does seeing father and mother (see Table I). This obser
vation confirms the results of earlier studies conducted in France (Gokalp,
1978) and the United States (Lee, Mancini, and Maxwell, 1990; White and
Riedmannn, 1992) and corroborates the hypothesis that the sibling tie is
weakly normed. Individuals don't bother to travel as far when the norm for
meeting frequency is not very strong and when not meeting carries no real
sanction.
(3) In this connection it should be observed that in Western societies fertility rates have
that there is nothing universal about the sibling been decreasing continuously. If only from a
tie. Of all primary blood relations {Ego with demographic perspective, then, the structural
father, mother, son, daughter, brother, sister), position of the sibling tie is secondary. As will
the sibling tie is the least common. Within the be shown further on, this conclusion is
kinship system, siblingship thus represents a supported by other, specifically relational
mere possible relation; it is not part of all arguments,
individuals' lives. This is especially true given
44 Emmanuelle Crenner, Jean-Hugues Déchaux, Nicolas Herpin
Table I. - Influence of distance between homes on meeting frequency between siblings and
between mother-child and father-child
Mother*** Sibling* Father**
Less than 10 km 99 165 179
10 - 49 km 31 57 65
50 - 499 km 19 11 19
Morethan 500 km 2 5 8
All distances 35 69 86
Data source: "Réseau de parenté et entraide", [Kinship networks and mutual support] of IssEE 's
Enquête Permanente sur les Conditions de Vie des Ménages survey [Ongoing survey of household
living conditions] for October 1999. [Same data source for all tables.]
Sample: French residents over 14 *having at least one brother or sister who does not live with
them, **father not living with them, ***mother not living with them.
Reading: Persons over 14 with at least one sibling not them saw this/those sibling(s) 99
times over the previous 12 months when their home and sibling's home were less than 10 km away
from each other.
Large families are readily considered more united than others. This is not
at all the case for sibling relations within them. Meeting frequency decreases
regularly by size of group (see Table IV below): 46 meetings in sib
ling groups of two, 41 in groups of three, 35 in groups of four, and only 28 in
groups of five or more. This means that meeting frequency per sibling de
clines even more. We see that an alternative hypothesis from the one we be
gan with -namely, that individuals devote a fixed amount of time to seeing
siblings and that meeting frequency is therefore constant by sibling group
size- is not confirmed either. (4) This result, already obtained by Lee,
Mancini, and Maxwell (1990), proves that large family size does not reinforce
status-determined obligations among siblings; in fact, relations within this
type of sibling group appear less normed than in others, to the point that here
there seem to be virtually no norms at all. (5)
Moreover, we would have had to go beyond the (4) It should be noted that the equivalency
consequences of divorce or separation to we have established between meeting
frequency (counted) and time devoted to consider other serious events in family history,
meetings (not counted, and which may vary such as grave illness, a parent's death, children
being removed from parents' home and placed greatly) is only approximative (see survey
presentation in Appendix). in foster homes, parents taking in cousins for
(5) It would have been interesting to assess long periods, etc. In this connection we can
only note that "broken" the impact of divorce and family recomposition sibling groups are a
when these intervened during childhood, as fairly rare phenomenon, as may be observed
well as of other major events in family history, indirectly in the proportion of siblings who live
on meetings between adult siblings. Ivsee's less than 10 years in the same household: 14 %.
"Réseau de parenté et entraide"1 provides only Among these we do of course find fragmented
families, but also wide age gaps between fragmentary information on family history,
however. We do not know, for example, if children of the same parents as well as
father and mother are or are not divorced, and of different ones. Impact of family
therefore can know nothing of what may history on adult sibling meeting frequency
happen as a result of divorce or separation. would itself be a worthy object of study.
45 Revue française de sociologie
Variations by occupation-social category (Insee's "Professions et
catégories socioprofessionnelles'''' [Pcs] nomenclature) and educational level
correspond to those for kinship relations in general: seeing siblings increases
in inverse proportion to place in the socio-occupational hierarchy, with farm
ers having the same number as industrial manual workers (39 meetings, as op
posed to 2 1 for senior executives and intellectual professions), and educational
level, with individuals who have obtained the baccalauréat [French national
high school degree] and anything below it at the same level for meeting
frequency (37, as opposed to 16 for individuals with at least the baccalauréat
plus two years of higher education). Variations by income present a similar
distribution, but with less spread: the wealthiest individuals (4th quartile) see
their siblings less than those with low or middle incomes (1st and 2nd
quartiles): 29 meetings to 38 (see Table XIII below).
For variables that identify social position in the broad sense, then (socio-
occupational, economic, and cultural-level), the sibling tie is no different
from other kinship ties, namely direct filiation. We might have expected that
as a tie more determined by personal inclination than status, closer to the a
rchetype of the chosen relation especially valued in mid-level and high social
categories (Bidart, 1997, pp. 244-246), the sibling tie would be more favored
in these milieux. Our results invalidate that hypothesis. As Coenen-Huther,
Kellerhals, and Allmen (1994, pp. 157-160) have shown regarding mutual as
sistance, the weak normativity of the sibling tie is a general given of the
Western kinship system; it transcends differences in social milieu. On this
point, Cumming and Schneider's thesis is confirmed.
Women do not see their siblings more than men
The second general observation is absence of any difference between men
and women for overall number of meetings: 35 for men, 34 for women. Varia
tions by age confirm this balance (see Table II below). The differences are
very small, increase moderately after age 40, first for men (men aged 40 to 59
see their siblings more often), then, after 60, for women. Telephone contact is
more frequent among women (27 calls versus 20 for men) -not surprising
given what we know of women's central role in telephone sociability. This
does not invalidate the previous observation, since the number remains low:
on average, 24 calls per year.
That there should be no difference between men and women is at first sur
prising, since it goes against the thesis that in kinship relations women have
specific responsibilities. More exactly, it invalidates the idea often put fo
rward that sibling relations are women's affair. Several British monographs
(Young and Willmott, 1983 [1957]; Firth, Hubert, and Forge, 1969) under
lined women's greater implication in relations with brothers and sisters. It is
true that these monographs are hardly recent and, with the exception of Firth
et al. (1969), concerned working-class neighborhoods. In the Us, women's
greater involvement in sibling relations has recently been confirmed for both
46 Emmanuelle Crenner, Jean-Hugues Déchaux, Nicolas Herpin
contact and affective or support relations, regardless of whether assistance
given is expected or effective (White and Riedmann, 1992).
Table II. - Influence of sex and age on number of meetings with sibling(s)
> 60 <30 30-39 40-59 All ages
69 35 28 25 Men 35
34 68 33 23 28 Women
68 34 26 27 Both sexes 35
Sample: French residents over 14 having at least one sibling who does not live with them (Same
sample for most tables; indicated when different.)
Reading: Men aged 30 to 39 with at least one sibling not living with them saw this/those sibling(s)
35 times over the previous 12 months.
In France and neighboring countries, more recent surveys have discovered
greater implication of women in terms not of meeting frequency but declared
affective closeness. A survey by the Institut National des Études
Démographiques (JNED) entitled "Proches et parents" [Close friends, and rel
atives] shows that the affective tie within the sibling group -measured by a
personal inclination indicator (ratio between number of relatives declared
"close" and number of surviving relatives)- is stronger for women than men
(Bonvalet, Gotman, and Grafmeyer, 1997, p. 40). A survey conducted in
Switzerland by Kellerhals' team reached a similar conclusion: women are
more likely than men to say they feel close to their siblings (Coenen-Huther,
Kellerhals, and Allmen, 1994, p. 66).
Our result does not contradict this last point since it measures only fr
equency of meetings among siblings, not affective intensity of the relation.
And as Walker and Thompson (1983) have shown in the particular case of
mother-daughter relations, meeting frequency between relatives is not necess
arily a good indicator of relational intimacy. There is no reason not to say
that women are more tied to their siblings affectively, even though they don't
see them any more than men do. If there is a difference, which is what earlier
studies suggest, it would then have more to do with the content and quality of
relations rather than their periodicity. As we shall see further on, this differ
ence can be seen in individual's choice of which sibling(s) to see. (6)
The importance of length of shared life
How often adult siblings see each other is naturally conditioned by how
long they lived together (see Table III below). The frequency of meetings in
creases parallel to length of shared life.
(6) If the affective quality of sibling be less where women are concerned and their
relations is stronger for women than men, then treatment of siblings not significantly different
the real impact of personal preference should iated.
47 Revue française de sociologie
Table III. - Influence of length of shared life on number of meetings with sibling(s)
Both sexes Men Women
Less than 10 years 24 26 20
10-14 years 32 35 32
15-19 33 37 31
20-25 years 42 42 46
More than 25 years 47 40 58
All durations 35 34 35
Reading: Men with at least one sibling not living with them saw this/those sibling(s) 26 times in
the previous 12 months when they had lived less than 10 years together in the family of origin and
40 times when they had lived more than 25 years together in the family of origin.
This influence, which confirms the observed tendency to non-differentia
tion of sibling relations with age, also weakens regularly with age. The great
est differences are to be found among the young (under 30), while among
older people (over 60) the differences are smallest. If the influence of this fac
tor decreases with age, this is probably because the time since siblings ceased
living together has itself increased, effacing the preferences that life together
gave rise to. However, siblings who lived together more than 20 years always
see each other more frequently than others -at all ages, even over 60. This
seems to suggest that 20 years of life together for siblings marks a "qualitative
leap": that length of time is enough to affect feelings and behavior in a way
that is not effaced by lapse of time since living together. Childhood involves
undergoing various trials. The fact of having lived through them together
makes brothers and sisters "closer", more sensitive to each other's different
points of view, more able to anticipate each other's reactions and to accept
each other's suggestions and criticism. We can see why having lived together
for a long time when young brings about increased frequency of meetings.
Still, the break that occurs at 20 years of shared life only applies to large
sibling groups -three or more. Sibling groups of two are less affected by
length of shared life, except when it is short (less than 15 years) or very short
(less than 10 years). Here the break falls at 15 years of shared life. After that
length of time, two siblings see each other much more often than three or
more (Table IV below). The 20-year threshold thus differentiates large sibling
groups from the others, but is inoperative in groups of two, where one sees
one's brother or sister even if shared life was not as long, on condition that it
not have been too short either.
What explains the stronger propensity of dyadic groups to generate rela
tively strong ongoing relations once individuals have left the family house
hold, even when length of shared life is not very long? One possible factor is
strength of status-determined norms. Though in general these are vague and
non-constraining, it is nonetheless more difficult to circumvent them when
there are only two siblings. On the contrary, when there are at least three, per
sonal inclination logic takes over more easily, and the play of preferences
48 Emmanuelle Crenner, Jean-Hugues Déchaux, Nicolas Herpin
(age, sex, or other) may lead a sibling to neglect another, as is the case for the
minority sex in sexually mixed triads (see below). Firth, Hubert, and Forge
(1969) showed that the fact of having only one sister or brother led to keeping
up good relations with him or her, whereas in sibling groups of at least three,
the proportion of good relations declines. Only long shared life -at least 20
years- can counterbalance this centrifugal effect and unite the group. In other
words, the regulation of relations in large sibling groups has more to do with
the interpersonal aspect of the relation and with what conditions it, such as
length of shared life, than with any normative framework.
Table IV. - Influence of length of shared life and sibling group size on sibling meeting frequency
4+ brothers 3 brothers All sizes 1 brother or 2 brothers
and/or sisters and/or sisters and/or sisters 1 sister
< 10 years 24 34 22 25 23
41 36 27 10-14 years 32 38
40 32 26 15-19 years 33 49
> 20 44 37 43 49 48
41 35 28 All durations 35 46
Reading: Persons over 14 with at least one sibling saw that/those sibling(s) 34 times in the
previous 12 months if they had lived together less than 10 years during childhood and 49 times if
they had lived together more than 20 years.
The secondary structural position of the sibling tie
While intense in youth, relations between siblings decline substantially in
adulthood, particularly at the moment of settling into a couple, and even more
markedly when children are born. This inflection illustrates the secondary
position of sibling relations compared to parent-children relations (direct
filiation), the two together constituting the primary relations structure. Once a
procreation family exists, filiation relations take over. Furthermore, if young
adult siblings see each other frequently, this is also in part thanks to their fa
ther and mother. Only the presence of niece(s) or nephew(s) can partially
compensate for the secondary structural position of the sibling tie.
Youth is the golden age of sibling relations
Variations by age are strong, but not continuous (see Table V below).
Brothers and sisters still see each other frequently when aged 15 to 19, only
half as often between 20 and 29, and half that often between 30 and 39. Fre
quency of relations stabilizes at a low level after 40. This progression is com
pensated for in part by telephone use, which increases over 60 but remains
below what it was under 30.
49