Family Configurations

A Structural Approach of Family Diversity in Late Modernity

Eric D. Widmer

1
INTRODUCTION – FROM THE FAMILY INSTITUTION TO FAMILY
CONFIGURATIONS......................................................................................................4

CHAPTER 1 – A CONFIGURATIONAL PERSPECTIVE ON FAMILIES.....................8

Configurations......................................................................................................................................................8
Interdependencies...............................................................................................................................................11
Family Issues Revisited......................................................................................................................................13
The Family Network Method.............................................................................................................................17
Data.....................................................................................................................................................................20
Conclusion..........................................................................................................................................................21

CHAPTER 2 - WHO ARE MY FAMILY MEMBERS ?.................................................23

Family Members Beyond the Nuclear Family...................................................................................................24
Family Boundaries..............................................................................................................................................27
A Universe of Family Members .......................................................................................................................31
Structured Diversity............................................................................................................................................33
Family configurations of Parents........................................................................................................................36
Family Structures and Configurations................................................................................................................38
Conclusion..........................................................................................................................................................40

CHAPTER 3 – FAMILY SOCIAL CAPITAL ...............................................................42

Interdependencies Beyond the Nuclear Family..................................................................................................43
Transitivity..........................................................................................................................................................45
Dimensioning Social Capital..............................................................................................................................50
Blood and Friendship in Families.......................................................................................................................53
Trusting Family Members.................................................................................................................................. 55
Emotional Interdependencies in an International Perspective............................................................................58
Does Social Capital Matter?...............................................................................................................................62
Conclusion..........................................................................................................................................................65

CHAPTER 4 – FAMILY CONFLICTS..........................................................................67

Being Interdependent in Conflict.......................................................................................................................67
Support and Interference.................................................................................................................................... 69
Ambivalences.....................................................................................................................................................70
Closeness and Conflicts ....................................................................................................................................71
Intergenerational Conflicts.................................................................................................................................75
Conclusion..........................................................................................................................................................77

CHAPTER 5 - POST-DIVORCE FAMILIES ...............................................................79

Social Support, Divorce and Remarriage...........................................................................................................80
Social Capital after Divorce...............................................................................................................................81
Conflict and Ambivalence in Post-divorce Families..........................................................................................84
Ambivalence in Older Adulthood.......................................................................................................................86
Divorce, Family Structures and Configurations.................................................................................................91
The Cultural Meaning of Stepfamily Statuses....................................................................................................95
Conclusion..........................................................................................................................................................99

2
CHAPTER 6 – FAMILIES AND PSYCHIATRIC PROBLEMS..................................101
Family Systems and the Configurational Perspective......................................................................................102
Family Composition and Social Capital .........................................................................................................103
Intellectual Impairment ....................................................................................................................................108
Variability Again...............................................................................................................................................111
Conclusion........................................................................................................................................................116

CHAPTER 7 – SHORT TERM CHANGES IN FAMILIES.........................................118

Focus Points......................................................................................................................................................119
Adaptation Within Stability..............................................................................................................................121
Structural Instability.........................................................................................................................................123
Conclusion........................................................................................................................................................128

CHAPTER 8 – FAMILY TRAJECTORIES................................................................131

The Family Life Cycle Revisited......................................................................................................................133
Empirical Family Trajectories..........................................................................................................................135
Life Trajectories and Family Configurations....................................................................................................137
Family Trajectories of Individuals with Psychiatric Problems.........................................................................140
Conclusion........................................................................................................................................................143

CONCLUSION – INDIVIDUALIZED FAMILIES........................................................146

Individualization...............................................................................................................................................147
The Fabric of Diversity.....................................................................................................................................149
Social Integration .............................................................................................................................................151
The Disadvantaged ..........................................................................................................................................154
The Family in Late Modernity..........................................................................................................................156
References........................................................................................................................................................ 158

3
 INTRODUCTION – From the Family Institution to Family Configurations

In the nineteen twenties, Ernest Watson Burgess, a leading representative of the Chicago School

and one early family sociologist, characterized the great changes that families experienced during his

life time by a new relational model (Burgess, 1926; Burgess, Locke & Thomes, 1960). Rather than

building a moralistic discourse regarding the decline of the Family, as so many scholars of his time

did, Burgess earnestly tried, by the use of empirical research, to understand the patterns that made the

new kind of family experiences that he discovered so different from those of the close past.

Ultimately, his research stressed the emergence of the Companionship Family, which replaced the

Family Institution: love against parental supervision of mate selection, privatization of family life

against community interference, equality between husband and wife against patriarchy. In his view,

the Companionship model, based on a long term commitment of spouses linked by democratic

arrangements and a functional specialization within the nuclear family, was about to fully erase from

Western societies the last remnants of rural forms of family life, doomed by modernity. the Family

was constituted by a small and lasting group of interacting personalities linked by meaningful social

roles, a feeling of shared belonging and a legitimate although unequal division of labour. In Burgess's

view, actual interactions rather than legal contracts defined families (Burgess, 1926). Based on the

empirical investigation of such interactions, he stressed the reorganization of modern families along

new structural principles, rather than interpreting the already increasing divorce rates as evidence for

the disorganization of the Family.

In the eves of the nineteen eighties, after divorce, non-marital cohabitation and changing gender

roles made it clear that things had changed in the family realm, some prominent scholars cast doubt on

the survival of a functional stable model of family. The diversity of families had become so great, they

said, that any attempt to find general principles underlying their organization was doomed. Indeed,

childless families, dual earner families, living apart couples, single parent families, stepfamilies, same

sex partnerships, families constituted by friends and adopting families are various alternatives that

4
currently compete with the model of the nuclear family, constituted by the main earner male, his wife

and their co-resident biological children 1. Each of these families can be further decomposed in

subtypes with their own structures and roles. Stepfamilies, for instance, cover a variety of situations:

couples with resident stepchildren from the female partner, couples with resident stepchildren from

the male partner; couples with resident stepchildren on both sides; couples with a mix of shared and

non-shared children; couples with resident stepchildren and non resident stepchildren, etc. The

number of possible family structures following divorce and remarriage is so important that one may

wonder if the stepfamily as an homogeneous concept is truly valid.

In relation with this diversity of family structures, a debate there is an ongoing debate among

sociologists, especially in North America, regarding the use of the plural or the singular concerning the

family realm. Should we talk about “the Family” or about “families”? The majority of family

researchers have opted for the second solution by arguing that the diversity of family structures and

experiences has become so great that refering to “the Family” is at best confusing, at worst a gross

scientific misconception, laden by normative biases stemming from a nostalgia for the nuclear family.

Frightened by the perspective of imposing a single model of family life as a norm or an ideal to which

all family structures should be referred, some scholars have celebrated the diversity and fluidity of

family experiences (Stacey, 1993 & 1996).

Others, destabilized by the diversity of families, have developed a pessimistic perspective on their

future. As a matter of fact, sociology since its origin has often developed negative views regarding

family change. The lack of a normative model of family commitment, stemming from

individualization trends, will supposedly make the Family collapse. The books of David Popenoe in

the United States are often given as particularly eloquent examples of that perspective (Bengston,

2001): the decreasing number of children in families, the increasing divorce rates and economic

independence of women are indicators of the declining role of families in modernity (Popenoe, 1993;

1
The nuclear family was defined by anthropologist Peter Murdock (1949) as a residence unit constituted by a
married heterosexual couple and its biological and socially dependent offspring . The functionalists added to this
definition a gendered dimension, with the organization of the family into two subset of roles, one instrumental,
associated with the male breadwinner, the other, expressive, with his home-focused wife (Parsons & Bales,
1956).

5
Popenoe, 2005). In Europe, Ulrich Beck and colleagues took over the same line of thought by

emphasizing the deleterious influences of individualization forces on the Family. A mass society of

eremites unable to maintain any long commitment is about to come (Beck, 1986; Beck & Beck-

Gernsheim; 2002). Family is a “zombie category” shred apart by the strong cultural and economic

enforcement on individuals to live a life of their own, which has left little in terms of private

commitments and family care. In a more positive light, Anthony Giddens emphasized the self-

referencing internal process of intimate partnership formation and development. His perspective,

although less pessimistic, reduced the significance of the social embeddedness of dyadic processes

and hence made it difficult to understand for whom and how family relationships still continue to

matter. Giddens’ concept of “pure relationship,” which is an archetype of late modernity according to

him, focuses on self-exploration, negotiation and symmetry in power relations, and decidedly weaker

external constraints on intimate relationships (Giddens, 1991 and 1995). In all cases, the possibility for

scholars to find models of long-term commitments in such individualized families has been considered

highly improbable.

While acknowledging the impact of individualization on families, this book asserts that the

diversity of family relationships still leaves much space to long term commitments. The functions of

support and social integration filled by families according to Burgess (1925) and Parsons and Bales

(1956), are still central in contemporary societies. But in order to make sense of the diversity of

current family commitments, another perspective on families is necessary. Sadly enough, quantitative

research continues for the most part to focus on the nuclear family, and qualitative research frequently

falls short of looking for rules and patterns underlying the apparent fluidity of family life in late

modernity.

Following the work of Burgess, prominent scholars from the United States and Europe sought to

develop an understanding of the relational models characterizing contemporary families and their

embeddedness in social structures (Kantor & Lehr, 1975; Kellerhals & Montadon, 1991; Kellerhals,

Widmer & Levy, 2004; Olson et al., 1989; Reiss, 1971; Roussel, 1992). To achieve this end, they built

6
a series of typologies of conjugal or family interactions. Their emphasis on relational logics

characterized by systemic properties lead to account for couples and families by a small set of distinct

types with specific functional properties. This has made great advances in the knowledge of families

possible. There is much indeed to command in those systemic approaches of the Family, which

stressed the necessity of going beyond individual cases and single dimensions of family life.

Following their lead, the configurational approach emphasizes the interdependencies existing among

large sets of family members and uses social network methods and concepts to uncover how family

relationships work beyond the nuclear family. Based on a series of empirical quantitative studies, it

reveals the importance of these interdependencies for a variety of issues. Divorce and remarriage, the

role of families for social integration, family conflict and ambivalence, families of individuals with

psychological problems, and changes in family relationships throughout the life course will be

addressed in such a perspective. The data come from the United States and Switzerland, two societies

known for their high rates of divorce and family turnover.

This book searches patterns shaping individualized families. It intends to uncover some of the

principles that account for their diversity, while going beyond cases in their singularity, in order to

show how families in late modernity relate to “the Family” as a set of informal rules that organize

positive and negative interdependencies existing among a large number of family members.

7
 CHAPTER 1 – A configurational perspective on families

What do we need another perspective on families for? The decline of the nuclear family

consecutive to the pluralization of life trajectories has made the work of sociologists more difficult.

Indeed, family members that matter cannot be defined a priori, using the household as a natural limit

to the Family. The number of relationships to be taken into account is much greater than that between

partners, or between parents and their resident children. In sum, the Family cannot be theorized as a

small group with obvious boundaries, a clearly defined and consensual division of labour, and a

collective identification. As a matter of fact, approaches that focus on the nuclear family disregard

significant dimensions of social integration belonging to families in late modernity.

An alternative is therefore needed. The configurational perspective on families traces complex

patterns of emotional, cognitive and practical interdependencies among large numbers of family

members beyond the nuclear family. It focuses on interdependencies between partners, children and

other individuals such as relatives and friends (Widmer, 2006; Widmer & Jallinoja, 2008). Its main

assumption is that key family dyads, such as the conjugal and the parent-child dyads, are embedded in

a larger set of family interdependencies which explain how they develop. Relatives can be a source of

strength for couples and parents by acting as a resource to help alleviate their conflicts. Support from

relatives and friends may however also foster tensions and conflicts for structural reasons that we will

uncover. Simultaneously, key family dyads influence the way in which family configurations are

shaped. This chapter stresses some of the main features of the configurational perspective on families.

The various issues that it raises will be tackled in later chapters.

Configurations

The concept of configuration was first proposed in the nineteen thirties by Jacob Moreno (1934),

a once most praised scientist as the founder of sociometry and of the sociodrama. Moreno defined a

8
configuration as an collection of individuals of any size, from the smallest personal networks, which

he termed “social atoms”, to humanity as a whole, in which meaningful ties link individuals with each

other. He first stressed that configurations concerned actual relationships rather than relationships as

defined by organizational charts, administrations or census offices.

He also put much emphasis on the fact that configurations are patterned: any dyad belonging to a

configuration is influenced by the shape of the configuration as a whole. Interdependencies within a

configuration are not randomly organized but follow informal rules, such as reciprocity. Indeed,

individuals tend to balance what is given with what is received in most relationships. An imbalance in

reciprocal exchanges triggers either frustration and conflict, or an inequality of power between the

exchange partners. Moreno also stressed that “chaining”, or what was later called “transitivity” or

“structural balance”, occurs in social relationships. One significant result of research on networks is

that they develop in bundles. Typically, friends of friends also become friends, and friends of enemies

become enemies. Based on Fritz Heider’s theory of balance (Heider, 1946; Heider, 1958), social

network scholars later found that people feel uncomfortable when their friends do not like each other

and they avoid these situations in their personal networks (Cartwright & Harary, 1956; Festinger,

1957, Newcomb, 1961). Transitive hierarchies and power structures have higher legitimacy in the eyes

of individuals than non-transitive ones. Transitivity has significant social consequences, as it

strengthens role differentiation, cohesion, and clustered subgroups. To summarize, transitivity fulfills

the individual’s need for consistency and creates differentiated and cohesive social structures.

The concept of configuration as it relates to networks, was taken over from the nineteen thirties to

the nineteen nineties by German sociologist Norbert Elias who contributed greatly to various fields of

sociology (Dunning & Mennell, 2003). Elias (1994) defined configurations as “structures of mutually

oriented and dependent people” (p. 214). Individuals, Elias proposed, are interdependent in a

configuration because each one fulfills some of the others' needs for social recognition, power,

emotional proximity, financial and practical resources, sexuality, or other socially defined needs

9
(Quintaneiro, 2005). Interdependencies are not dyadic in nature, Elias stated, but rather organized in

large networks. Individuals develop a variety of ties with family members, friends and colleagues,

who branch out to other persons. As such, configurations have to deal with power issues: resources

are scarce and individuals, while cooperating, also compete for them within groups. This competition

creates tensions and conflicts that are beyond an individual's control. The patterns of

interdependencies that characterize configurations, therefore, are commonly unintended. They, in turn,

shape the cooperation strategies and the conflicts that occur in each dyad belonging to them.

Based on this theoretical stance, the configurational perspective posits that family dyads are

interdependent (Widmer & Jallinoja, 2008). It stresses on the one hand that parent-child or couple

relationships are shaped by the larger networks of interdependencies with relatives, friends and others

in which they are embedded. On the other hand, patterns of interdependencies depend to a significant

extent on partnerships and parent-child relationships (Widmer, 2004). Various research results pointed

at the fact that what happens in couples influences networks. For instance, the courtship process

interferes with other strong ties and tends to lower their importance. Therefore, individuals are

concerned about mate selection of their network members. When couples split, conjugal networks are

again profoundly changed. When individuals remarry, this has implications for a large number of

persons beyond their couples or their households. Courtship, conjugal roles, conjugal quality, divorce

and remarriage show that the fragility of couples in Western societies is intertwined with larger

relational contexts that they shape while being shaped by them (Widmer, 2004).

Although the configurational perspective cannot be considered a theory at this stage of its

development, it makes a number of assumptions that facilitate the study of complex patterns of

relationships, such as those characterizing families in late modernity. First, families that matter are not

defined by institutional criteria such as belonging to the same household or being married together.

Family interdependencies, what we need others for, and the tensions and conflicts that they set up, are

given prime importance. Second, the configurational perspective rejects the assumption that family

10
dyads can be analyzed as independent and separate entities each with their own logic. Instead, it

focuses on the influence of the larger configuration of relationships in which each dyad is embedded.

Finally, a configurational perspective on the Family emphasizes its temporal and spatial nature. It

measures change and stability of family relationships in individual life courses and across historical

periods. Both Elias and Moreno emphasized the developmental dimensions of configurations. Elias

was the most vivid about it, as he claimed that the historical dimension of configurations should never

be overlooked. Because he stressed the changing balance of tensions as a main feature of large social

configurations such as the royal court of Versaille, historical sociology is the only disciplin that fitted

his expectations. Similarly, Moreno emphasized the need of following configurations over time. As his

empirical work was based on smaller groupings and involved sociometric measurements, he focused

on short term changes. Both authors however stress the importance of time, either historical or

personal, as a main feature of configurations.

When applied on families, this set of assumptions emphasizes the embbededness of partnerships

and parenting in large and complex sets of relationships with steprelatives, in-laws, grandparents,

uncles and aunts, cousins, friends or neighbors, and even care professionals considered as family

members. A structural approach of relationality and embeddedness of families and personal lives is

currently lacking. This book reconsiders some main issues of family research by defining families as

large, open and personal configurations rather than as small, closed and collectively organized groups.

It illustrates the fruitfulness of this theoretical shift by presenting a series of empirical results.

Interdependencies
Interdependency is a central concept of the configurational perspective. It stems from the fact that

individuals depend on a variety of others, without necessarily being aware of that dependency and

willing to admit it in standard surveys or comprehensive qualitative interviews. If practical services

and money transfers clearly constitute a set of interdependencies within families, they are certainly

not the only or even the most significant ones. There is a long tradition in sociology that emphasizes
11
the cognitive and emotional importance of group members. Concepts such as the “reference person”

(Hyman, 1942) or the “reference individual” (Merton, 1957) or the “orientational others” (Kuhn,

1964), stress the importance of specific persons as cognitive and emotional benchmarks throughout

one's life. Kuhn (1964) defines orientational others as people to whom individuals are committed

emotionally and psychologically, who provide individuals with a concept of self, and who influence an

individual's self-definition through communication. Note that, once established, the persistence of

someone as an orientational other does not require frequent, recent, long or even positive interactions

(Milardo, 1989; Surra & Milardo, 1991).

It is readily apparent that many family members beyond the nuclear family meet Kuhn’s

definition of the orientational other (Kuhn, 1964). Frequency of interactions, practical help or

financial help are not the sole indicators of family interdependencies. Various contingencies from late

modernity may decrease such interdependencies while cognitive and emotional interdependencies

remain strong. People may not receive money from their parents or they may not see them regularly

because they do not need the provision of money, or because they live far away from them. Despite

that, they may still be very much emotionally and cognitively interdependent with them. Emotional

support and communication are prime features of relationships of individuals with their parents and

siblings in adulthood. It comes as no surprise that partners also play such a role, some after divorce

also. Although sociologists often prefer the hard facts of money, face to face interactions and domestic

support, the strength of feelings and cognitions should not be underestimated as indicators of

interdependencies, especially in the family realm. Indeed, some anthropologists have cleverly argued

that financial or material support gain a special value in families as they are interpreted as proofs of

love (Schneider, 1980). From the work of Marcel Mauss (1992), we know that material exchanges

have meaning beyond their monetary value. In all cultures, including Western cultures, exchanges of

material goods in families are signs of interpersonal acknowledgement (Caplow, 1982). Individuals

care about family gifts and support because they give meaningful information on whether or not they

12
matter for other family members. A large number of studies show that perceived support, rather than

received or provided support, influences individual development 2. The strong feeling that family

members are concerned by us is a key dimension of interdependencies.

Interdependencies are also unintended (Newton, 1999). Cooperation in configurations creates

unexpected tensions and conflicts because it is associated with various constraints that individuals

enforce on each other by their interdependencies (Elias, 1984; Widmer, Giudici, Le Goff, Pollien,

2009). Resources in time, love, money, support and social recognition are scarce and their unequal

distribution within family configurations is subject to power and control attempts that make them shift

from one state of balance to another state over time. Those shifts go beyond individual control because

of the complex patterns of interdependencies shaping configurations (Letonturier, 2006). The balance

of tensions and cooperation in the configuration in turn shapes processes of cooperation and conflict

occurring in each dyad (Elias, 1983). By linking power issues and conflicts with positive

interdependencies, the configurational perspective enables researchers to understand them as social

processes rather than as outcomes of a group's failure to function properly. This is especially

significant, as we will see, for the understanding of families in late modernity.

Family Issues Revisited

Various scholars have felt the need of crossing the borders of the nuclear family and of

addressing family relationships as sets of interdependencies, both negative and positive, with an

emphasis on cognitive and emotional dimensions. Research on conjugal interactions and parenting

has stressed the importance of taking the relational context of nuclear families into account.

Demographers have acknowledged the impact of close family members and friends on fertility issues

(Widmer & Jallinoja, 2008). Gerontologists have underlined the importance of personal networks for

2
Perceived support is the perception that social support would be available should an individual wish to
access it (Sarason et al., 1991). It is strongly correlated with a variety of developmental outcomes (Sarason,
Sarason & Shearin, 1986; Sandler & Barrera, 1984; Sarason et al., 1991; Wethington & Kessler, 1986).
Findings on the impact of received support on adjustment are far more ambiguous (Sarason et al., 1991;
Wethington & Kessler, 1986).
13
the understanding of health (including psychological health) in old age (Guilley & Lalive d'Epinay,

2008). A series of studies on conjugal interactions and satisfaction have also shown the embbededness

of couples in larger social networks (Widmer, 2004; Widmer, Kellerhals & Levy, 2004 and 2006a).

How shall we proceed from there? A first central issue, from a configurational perspective, relates

to the very definition of “families”. What is a family? How do we define families that matter, that truly

influence one's life? What are the boundaries of such families, especially those stemming from divorce

and remarriage? The usual assumption of family research is that significant family units are obvious.

They are defined either as including members of a single household or individuals linked by marriage

or biological parenthood. On the contrary, the next chapter stresses that no institutional criteria such

as those is comprehensive enough to define families that matter. It underlines the emergence of

significant family interdependencies that cannot be circumscribed in reference with a household or

with a limited set of institutionalized statuses. It first recalls studies on the ways in which individuals

define the boundaries of their family. Then, the chapter focuses on empirical studies that were

conducted in the United States and Switzerland. It presents several types of family configurations,

each with a distinct emphasis on children, stepparents, in-laws, grand-parents, siblings, partners or

previous partners, and friends defined as family members. Using interdependencies rather than

institutionalized roles to define families is a prerequisite to the understanding of their overall relational

logic from a configurational perspective.

A second major issue that this book addresses is the contribution of families to social integration.

The large increase of non-marital cohabitation, divorce and serial remarriages in recent decades lead

several prominent sociologists to question the ability of families to provide care and meaning to their

members in late modernity. Families are no more in charge of social integration said Beck and Beck-

Gersheinheim (2002). The chapter, “Family Social Capital” first reviews the evidence pointing to the

integrative role of family members beyond the nuclear family. Then it describes two alternate

conceptualizations of relational resources provided by family members in terms of bonding and

14
bridging social capital. As we shall see, the composition of family configurations makes a great

difference for social capital.

Because of the emphasis on the integrative function of families in many scholarly works, only

little interest for conflict developed until the rise of divorce in the nineteen sixties, which pressured

sociologists to address family conflict as an issue. Since then, rather broad explanations of family

conflict were given, stemming from the contradiction between individualization trends of women and

men (for instance, Beck and Beck-Gernsheim, 2002). The configurational perspective proposes to

analyse conflict as intertwined with positive family interdependencies. Individuals develop family

conflicts because they depend on each other for various resources that cannot be found in other

relationships: sexual intimacy, emotional support, communication and social recognition are central

dimensions that relate families to the construction of personal identity. Dependencies are never easy to

tackle as they create fairness and power issues. Therefore, conflict and support are not opposite

processes in families but rather the two faces of the same coin. In order to understand how family

conflicts develop in late modernity, it is necessary to take into account the complex patterns of

contradicting interdependencies linking a large number of individuals.

In that regard, post-divorce families are a telling case. Individuals belonging to them are

hypothesized to be part of divorce chains or remarriage chains, in which interdependencies exist

among a large number of persons living in different households and linked indirectly (e.g. a child and

her father’s new partner’s children’s father). When divorce and partnering occur more than once, on

several sides (father, mother, parents’ partners) or in multiple generations (one’s parents but also one’s

grandparents), interdependencies become complex and create a variety of power issues. Based on

several empirical studies, the book carefully considers flows of emotional support and conflict in post-

divorce families. Rather than assuming that a single pattern of interdependencies capture the

complexity of post-divorce family configurations, it emphasizes the diversity of such contexts in terms

15
of composition, social capital and conflict. It is indeed plainly wrong, as we will see, to equate all

post-divorce family configurations to the same patterns of interactions.

Another crucial issue of family research concerns the interplay between family relationships and

psychological outcomes. Are individuals with psychological problems embedded in similar family

configurations than other individuals? The interrelation existing between family relationships and

psychological health has been a classical focus of system theory (Kerr & Bowen, 1988; Broderick,

1993; Minuchin, 1974). The results of various empirical studies done on clinical samples will be

reviewed, that show that the model of the nuclear family should indeed be revisited if one wishes to

better understand the relational contexts of individuals experiencing psychiatric problems. Again,

much variability of family configurations exists in clinical samples. It is a shortcoming, as we shall

see, to consider that the majority of individuals in psychotherapy or in daycare psychiatric facilities

have similar family configurations.

Family configurations are not static entities but constantly adapt to life events and life transitions.

The issue of change over time will be addressed, as it is a central assumption of the configurational

perspective that configurations are evolving in ways that are never fully intended by individuals

belonging to them. Life trajectories give many opportunities for new interdependencies to shape

family configurations. Marriages, births, sicknesses, residential moves, divorces, or deaths of family

members at the same time destroy and create family interdependencies. Measuring changes over time

requires either retrospective or longitudinal data with a variety of limitations due to the cost of

collecting such information. Despite these limitations, some results will be given about the ways in

which family configurations change over time pending on specific life events and on the overall

pattern of one’s life trajectory.

16
The Family Network Method

Since the mid nineteen nineties, we have been developing a methodology for assessing family

relationships in a configurational perspective based on social network methods. The Family Network

Method (FNM) (Widmer, 1999b; Widmer & Lafarga, 2000) provides an alternative way for studying

family relationships as sets of emotional and cognitive interdependencies. Its theoretical bases and

empirical measurements significantly differ from standard surveys dealing with family issues in a

quantitative perspective. It is based on the cognitive network or “socio-cognitive” approach of social

networks, which was first proposed in the nineteen eighties and nineteen nineties by sociologists and

anthropologists interested in biases of perceptions concerning relationships in social networks

(Krackhardt, 1987). In this approach, individuals interviewed do not only report on their own ties, but

on interdependencies which exist among all individuals included in the network that they belong to.

In the Family Network Method (FNM), respondents first give a list of persons that they consider

as significant family members. They are instructed that the term “significant” refers to those people in

their family who have played a role, either positive or negative, in their life during the past year. A

statement is read to respondents that further emphasizes that they should not only refer to the people of

their family who are significant to them because they love them or respect them, but also to those who

have upset them or have made them angry during the last year. The term “family” is left undefined and

respondents are asked to use their own definition of what they intend by « family ».

Based on this list of family members, various questions are asked to respondents about emotional

support, conflict and influence, three fundamental dimensions of all interpersonal relationships

(Widmer, 1999a). Respondents not only have to estimate their own relationships with their family

members, but also the relationships existing among all family members. The question regarding

emotional support reads: « From time to time most people discuss significant personal matters with

other people. Who would give emotional support to X during routine or minor troubles? » Answers are

then transformed into a square socio-matrix such as the one of Table 1. Each relationship is coded by a

17
one. When a relationship is inactive, the corresponding cell is set to zero. The instrument takes from

10 minutes to 60 minutes to fill in clinical samples, and from 10 to 30 minutes in non-clinical samples.

Respondents only estimate a series of dyadic relationships with straightforward “exist/does not exist”

decisions to make. They proceed systematically throughout the list of their family members and a

great amount of relational information about their family configuration is gathered in a reasonable

amount of time. Note that the relationships are not necessarily reciprocal. The information make it

possible to analyse relationships among family members with ease; it also permits to visualize

interdependencies of various kinds in family configurations. The graph Figure 1 of captures in a

simple and straightforward way respondents’ perceptions of how their family configurations are

structurally organized. In Figure 1, arrows linking individuals represent flows of emotional support in

one family configuration. Note that arrows initiate from support seekers and point to support

providers.

FIGURE 1 ABOUT HERE

Figure 1. Perceived emotional support in a family configuration

Note: Arrows point to support providers

The graph of Figure 1 provides information at several levels. Firstly, one achieves an overview of

family interdependencies. This family configuration is fully connected, as no individual is or sub-

group is an isolate. Secondly, one also sees that some individuals are more central than others:

interestingly, the respondent is not the most central individual in his own configuration but her mother

is. Thirdly, the interdependencies are strongly influenced by the status of family members. The

mother's, father's and partner's closest relatives are spatially differenciated, although interdependencies

exist among them as well.

18
 INSERT TABLE 1.1 HERE

Table 1. Matrix of emotional support in a family configuration

Sociometric matrices and their corresponding graphs offer a straightforward way to measure and

illustrate structural properties of interest to family researchers such as group boundaries, cohesion,

subgroups, balance and reciprocity within one’s family configuration (Wasserman & Faust, 1995) .

Other concepts dealing with family interactions such as relational roles and power distribution have

ready-to-use equivalents in sociometric research. For instance, family balance can be approached

using measures on transitivity; relational role structures are measured in reference with structural

equivalence models, and subgroups by clicks. In considering families as configurations, researchers

benefit from a range of measurements widely used in social network analysis 3.

In many respects, the Family Network Method (FNM) is distinct from questionnaires measuring

social support or support networks. First, it does not intend to enlist support providers or protective

persons, but all family members, be they supportive, disruptive or both at the same time. Second,

rather than focusing on the respondents' relationships, it asks them to estimate relationships among all

family members. In this respect, it is complementary to instruments measuring key dimensions of

perceived support such as how many support needs are met, support availability and satisfaction

toward support, or support appraisal (Procidano & Heller, 1983 ;Sarason; Levine, Basham, 1983;

Vaux et al., 1986). The Family Network Method is also less sensitive to social desirability than

standardized questionnaires about family issues. Individuals do not have a clear view of the picture

that they provide while responding to “yes/no” questions about the large number of the relationships

that they have to consider. For instance, they do not feel ashamed to report a persisting conflict

between themselves and their partners, because the issue is included in a much larger list of conflicts
3
This book focuses on three sets of measures related to density, connectivity and centrality, which are described
in chapter three.
19
that exist among their family members. Contrary to assessment tools measuring support or interactive

networks, the Family Network Method is not designed to measure actual interactions, such as

frequencies of contacts or amount of help actually provided to respondents by their family members. It

focuses on significant members of the family rather than on family members regularly met or who

provide support.

Data
This book is based on research done in the United States and Switzerland between 1998 and 2010.

It is not a comparative piece but it nevertheless intends to consider emergent family realities from the

two countries, in the light of the configurational perspective. Various datasets from both contexts are

used. An extended case study, that of Betty, a twice divorce woman in her fifties, living on the West

coast of the United States, and her 15 family members living across the country, is considered in

several chapters4. The family configuration of Betty provides several insights about how family

boundaries are defined, as well as about the links between family support and family conflict. Family

members of Betty were interviewed twice with a nine month lag, in order to measure the permanence

of family configurations over time. A sample of undergraduate students of a university in the U.S were

also interviewed. They were recruited through direct contacts with the author and several of the

author’s students. These analyses were complemented by several datasets collected in Switzerland by

the author and his colleagues. A first sample includes 1534 couples in which both partners were

interviewed separately, two times within five years, about their kinship and friendship networks and

their conjugal interactions and their networks of relatives and friends. Another sample of 500

undergraduate students from various Swiss universities were asked about their family configurations;

A sample of 100 women in middle adulthood, mothers of at least one child, were asked about their

family configurations; one third of them experienced divorce and remarriage. In order to approach the

complex issue of family configurations and psychological frailty, three samples of individuals with a
4
Case studies, we believe, do a lot for the advancement of knowledge, as they help researchers to
highlight the main issues that need to be addressed using large samples and statistical analyses.
20
clinical background were collected, including a sample of 80 adults in psychotherapy, with five waves

of interviews over a year and half; a sample of 60 individuals with psychiatric disorders

institutionalized in a daycare facility, and a sample of 25 individuals with psychiatric disorders and

mental impairment. One family member was also interviewed for 17 of them. In order to estimate how

distinct or similar family configurations are in the United States, Switzerland and other Western

countries we used the module social networks of the International Social Survey Program, which

complemented the samples that we personally collected by twenty representative national samples 5.

The survey includes a large number of variables concerning family members beyond the nuclear

family. This dataset, although rough in the information that it provides, is a valuable international

source of data on relationships beyond the nuclear family.

Conclusion

The configurational perspective on families developed in this book underscores the cognitive and

emotional interdependencies that exist among large numbers of family members. The aim of the

configurational approach is to discover how family interdependencies relate to support and caring, but

also to conflict, control and interference, in central family dyads such as partnerships and parent-child

relationships. The configurational perspective emphasizes a bidirectional causality, where individuals

and dyads contribute to shaping the family configurations that they belong to and are shaped by them

(Elias, 1991). Therefore, a reciprocal causation where individuals and their actions and motivations

influence the interdependencies with configuration members is in line with the configurational

perspective. Instead of conceptualizing families as composed by autonomous individuals lead by

abstract values or personal lifestyles, the configurational perspective focuses on the larger network of

interdependencies in which they are embedded. In that sense, the approach is structural. It stresses the

existence of informal rules that account for the alternatives available to individuals in their family life.

These rules, we believe, are the bases on which much of the current diversity of families develop.
5
http://www.gesis.org/en/data_service/issp/data/2001_Social_Networks_II.htm
21
Therefore, uncovering them contributes to the understanding of the individualized families of late

modernity.

22
 CHAPTER 2 - Who Are my Family Members ?

Individualization theorists who forecast the decline of the Family have a specific type of families

in mind: the nuclear family, constituted by an heterosexual married couple with its co-resident

biological children (Stacey, 1993). This definition of the Family arbitrarily stipulates who counts as a

family member and who can or even should be disregarded. Fortunately, there are alternatives for

defining family members that matter. From the perspective of Jacob Moreno (1934), configurations

build up on shared interests and concerns. People belong to a configuration because they are linked to

a space, an activity or a person that make them interdependent. In other words, they share a focus

point, as American sociologist Feld called a social, psychological, legal or physical entity around

which joint activities are organized (Feld, 1981). Configurations come into being and evolve because

individuals modify their activities, visit other spaces or change their concerns for persons. It is notable

that loving the same person, having a similar interest for the same activity or working in the same

place create patterns of interdependencies that go beyond positive and intended interactions. Indeed,

focus points do not necessarily need to be positive or emerging from autonomous choices 6.

Moreno's and Feld's insights about the importance of joint activities or concerns for similar

persons have direct implications for the study of families. The configurational perspective states that

families should not be defined using institutional criteria. Considering households as the natural

settings of families, or marriage as the sole object of family sociology, we argued in the introduction,

overlooks the complexities of families in late modernity. How then should the boundaries of families

be defined? Who is a family member? What should be taken in consideration when referring to an

6
From there, Moreno (1934) studied a number of configurations, from orphanages and school classes to larger
groupings, using basic social network tools. Interestingly, he also considered what he called“social families”,
that is educational facilities in which adolescents or children were placed under the care of adults within a family
like kind of settings.
23
individual's family? Blood, marriage, sexuality, intimacy, long term commitments? Or something that

encompasses all those meaningful dimensions of the Family?

The definition of family boundaries is indeed a serious matter, as setting who is in and who is out

of the family has consequences for the understanding that researchers are likely to propose regarding

the functions and structures of families in late modernity. As the Family concept is currently

overloaded with various normative expectations concerning the proper way to build a partnership, to

rise children, to work as a woman or a man, to give or to receive various forms of support, some

authors abandoned it for other concepts, such as close relationships (Scanzoni & Marsiglio, 1991;

Scanzoni et al., 1989) or personal life (Smart, 2007). From a configurational perspective, family

relationships carry a series of functions, emotional, cognitive and practical, that would not be well

understood by dropping the concept of family. More precisely, there is no other concept as the Family

that carries the very idea of strong intimate interdependencies, negative and positive, in some respects

chosen, in other respects enforced, with significant consequences on the long run. The concept is

therefore still fruitful while its definition should certainly be reconsidered. This chapter starts by

recalling some issues related with family contexts beyond the nuclear family. It proceeds by focusing

on empirical research that we conducted in the United States and Switzerland on the composition of

family configurations.

Family Members Beyond the Nuclear Family

The emergence of family configurations that are not circumscribed to the household or to a

limited set of family roles has been underlined throughout the last two decades (Cherlin &

Furstenberg, 1994; Scanzoni & Marsiglio, 1993; Weston, 1997, Widmer, 1999b; Widmer & Jallinoja,

2008), and an interest in family relationships beyond the domestic unit has developed. Research on

post-divorce families draws attention to the fact that households and families are not one and the same.

In many cases, the emotional interdependencies between children and their non-resident parents

24
remain strong and non-resident parents still have much psychological import for the child. On the

other hand, divorced parents are likely to remarry and become emotionally dependent on a new

partner. Therefore, many stepfamilies are part of divorce chains or remarriage chains (Bohannan,

1970), in which emotional and cognitive linkages exist among persons living in different households.

Consequently, when members of a stephousehold are asked who are part of their family, a different

answer is received from every one of them (Cherlin & Furstenberg, 1994). Interestingly, two partners

in a first marriage also have quite different family members. As a matter of fact, in one empirical

study that addressed the “shared versus specific” dimensionality of partners’ kinship networks,

partners reported on average fewer than half of their family members as shared (Stein, Bush, Ross &

Ward, 1992). Such evidence suggests that family configurations are indeed individualized. This is not

only due to the changes associated with divorce and remarriage, but also to the bi-laterality of Western

kinship systems (Lévi Strauss, 1949; Murdock, 1949), which has for centuries emphasized rights and

obligations of individuals both on their mother's side and their father's side.

The conceptualization of family ties as chains of interdependencies indeed does not only concern

families facing divorce and its aftermaths. Since the 1970, a large number of United States and United

Kingdom studies have identified the contribution of larger family contexts to nuclear families. As

reported by Lee and Adams (Adams, 1970; Lee, 1980), the primary focus of research on relational

contexts of couples in the sixties and seventies was undoubtedly the issue of the isolation of the

nuclear family from its kinship network, a central proposition of the structural-functionalist

perspective on families (Parsons & Bales, 1956). Since then, the hypothesis that nuclear families are

isolated from their kinship network was rejected on empirical grounds (Widmer, 2004). Adults

frequently keep strong emotional interdependencies with their parents, who represent a large

proportion of one’s support network. Interdependencies are played out in emotional and cognitive

terms as much as in practical and financial support. Despite the fact that they live in different

households and do not belong to the nuclear family of individuals any more, parents and siblings of

25
adults continue to have a great emotional significance (Fehr & Perlman, 1985). This is also the case in

old age. The majority of elderly people have a close relative alive and a child, with many having

grandchildren. In Switzerland, as four over five individuals above 80 have at least one contact per

week with their family members (Guilley & Lalive d'Epinay, 2008). These social interactions are

linked with emotional closeness: a great majority of individuals acknowledge that they have at least

one person in their family to which they feel very close. Old age is therefore on average characterized

by the maintenance of family resources rather than a decisive decrease of those resources (Bengston,

Rosenthal & Burton, 1990; Bengston, Harootyan, 1994). Overall, at all stages of adult life, family ties

remain meaningful in late modernity (Bengston, 2001).

The nuclear family has been for years a model rather than a reality. The development of single

living since the nineteen sixties has been stressed. The causes of the large increase of such living

arrangement are related to changes at both ends of adult life: the transition to adulthood has become

less homogeneous with many individuals spending an extended period of time by themselves before

possibly entering a stable intimate relationship. The increase of life expectancy has made many

widowers live an extended period of time without a partner. It is a misunderstanding, however, to

draw from the fact that 50 percent of women in later years live by themselves that they do not have a

family of their own. Links with grown up children and grandchildren, as well as with siblings,

constitute a large share of the elderly' s social networks, with much import for one's health and life

expectancy (Bengtson, Harootyan, 1994). Intergenerational family relationships remain vivid, if not

necessarily positive, in old age.

Are families only structured by blood and marriage? The distinction between friendship and

family ties may not be as clear-cut as the two concepts suggest. For instance, gays and lesbians create

non-kin families that include friends and lovers (Scanzoni & Marsiglio, 1993; Weston, 1997). Friends

frequently become godfathers and godmothers by being given a family status that crosses the blood

divide. Another reason for casting doubt on the definition of families as nuclear is provided by the

26
globalization of contemporary societies. Various ethnic groups in late modern societies do not fit

within the nuclear family scheme. For instance, strong kinship interdependencies exist in African-

American and Hispanic families in the United States (Aschenbrenner, 1973; Madsen, 1964, Stack,

1983). In the African-American community, grandmothers often play a key role in rising children,

especially when resources of parents are scarce or when the mother misses a partner. In the Mexican-

American community, couples are frequently embedded in large networks of relatives of various

kinds.

Considered one by one, non-nuclear family forms concern only a minority of individuals. When

added up, however, they represent a majority of living arrangements. As a matter of fact, the nuclear

family in its standard form has become a quantitatively marginalized way of life in societies such as

the United States or Switzerland (Bengston, 2001; Stacey, 1993). Defining family members that

matter using the household membership is therefore inadequate. Individuals depend on various

relationships, in which adult children, ageing parents, siblings, ex-partners, lovers or friends

considered as family members play a significant role.

Family Boundaries

An alternative to the approach of families as nuclear is therefore needed. The Family Network

Method (Widmer, 1999b; Widmer, Chevalier & Dumas, 2005; Widmer & La Farga, 2000) was drawn

from social network research (Wasserman & Faust, 1994). Let us exemplify its usefulness for the

definition of family boundaries by a case study. Betty is a 54-year-old female working as a social

worker in a state run facility for abused and neglected children in a middle-sized American town

located on the West Coast of the United States. She has been married twice, with both marriages

ending in divorce. Betty has two children from her first marriage, as well as a fairly large number of

blood relatives consisting of three sisters, two brothers, and her mother. She was chosen as a case

study primarily because she experienced both divorce and remarriage. In order to determine whom

27
Betty defined as her significant family members, she was asked to give the first name of the persons in

her family who were significant for her at the time of the interview. Betty included nine significant

family members who then had to report who their own significant family members were. Then, we

interviewed all persons who were included by at least two persons included by Betty as significant

family members, under the assumption that they were actors in her family configuration. Betty

included an extended range of individuals as significant family members: her two children, her two ex-

partners, two close friends, her mother, and two sisters. After interviewing them, five individuals were

added because they were included by at least two persons cited by Betty. There were three other

siblings, a sister-in-law, and Betty’s former mother-in-law from her first marriage. Note that she did

not herself include her two brothers and one sister. Figure 2 depicts the family configuration up to a

distance of three from Betty (her direct cites, the joint cites of at least two of her cites, and some of

their own cites). The larger nodes represent the fifteen core members of Betty’s family configuration.

Nodes in black indicate the persons that have been included directly by Betty as significant family

members. The graph shows a total of 79 persons included with 145 arcs linking them together. A

description of the status of each family member is provided in Table 2.

28
 INSERT FIGURE 2 HERE

Figure 2. Relationships in the family configuration of Betty

Insert table 1.2 here
Table 2. Family members of Betty (relation to Betty, age, gender, education, occupation)

From Figure 2, one sees that Betty’s family configuration is not a bounded small group corresponding to

a nuclear family, but rather corresponds to a widespread open network featuring a configuration of directly

or indirectly interdependent individuals. Among the persons included, two thirds were cited by only one

individual. To further assess the boundaries of this configuration, a matrix of citation matches among the 15

core family members was computed7. The overall match is .14, meaning that, on average, family members of

Betty share 14 percent of their own family members with others. When only the nine persons included

directly by Betty, plus Betty, are considered, this percentage remains about the same (15 percent). Twenty

seven of the 105 pairs composed by the 15 individuals included in the family configuration of Betty do not

share any significant family members at all, and 25 share only one person, Betty, in 21 cases. Two thirds of

the dyads share fewer than four persons as significant family members, and only 10 percent report sharing

more than six members. On average, each pair of individuals shares only 2.6 members. Those various

estimates show that the extent to which each person’s family configurations overlaps with others is small,

thus lending support to the conclusion that family configurations are individualized structures. Every

individual has to a large extent a family configuration of her own. Parents and children do not have the same

family configuration, as well as partners. Even siblings, especially in adulthood, have non overlaping family

configurations. This does not mean that family configurations are fully personal either. Some family

members are obviously shared by several individuals, as shown in Figure 2.

Family boundaries cannot be taken for granted in late modernity. Family configurations, which include

individuals with whom significant interdependencies have developed, expand into several directions, from

current and past sexual relationships, members of the kin network, relatives by marriage, and friends or co-

workers considered as significant family members. As significant family units are networks of

interdependencies rather than cohesive subgroups defined by a common residence and institutionalized roles,

7
This index computes the proportion of family members that any two individuals share, compared with the total number
of family members that they cite.
30
the type of social integration one can expect from them is different from what is provided by smaller and

more connected groups such as nuclear families. Emotional support and conflicts in family configurations,

as we shall see, are much more varied and individualized than what is often assumed.

A Universe of Family Members

Who are my significant family members? In a series of additional empirical studies, we instructed

respondents that by “significant family members” we were referring to “those people in their family who

played a role, either positive or negative, in their life during the past year”. The interviewer had to mention

further that they were not only interested in the people that were significant to the respondents because they

loved them or respected them, but also in those who had upset them or had made them angry during that

year. Note that this question does not assume that a high level of face to face interactions currently exists

with family members. It does not suggest either that the set of family members constitutes a bounded

functional unit8. The only assumption made is that individuals included are cognitively and emotionally

significant and considered as family members. We also asked respondents to recall family members who

were significant in a negative way. This part of the definition deviates from the assumption that only positive

memberships define groups, a strategy which leads to biases (Milardo, 1983; Milardo, 1988). As a matter of

fact, intimate relationships such as family relationships are not free of violence and conflicts, and the sense

of family identity has complex components, negative as well as positive (Berscheid, 1983; Firth, Hubert &

Forge, 1970 ; Peterson, 1983; Sprey, 1969). The assumption that close relationships are by nature positive is

rejected by configurational studies, which stress that conflict relationships shape interdependencies just as

much as positive relationships.

In a first study, based on 25 college students from a north American campus, the average number of

significant family members is 8.6, living in 5.7 different households. The variety of family members is

noticeable with a total of 19 distinct types of family members included, such as fathers, mothers, sisters,

brothers, but also grandparents, steprelatives and in-laws. Of the 25 respondents, 24 include their mother in
8
This assumption was made by functionalism, which makes the understanding of power and conflict relationships in
families difficult, as the family is conceived as an homogeneous entity. From this perspective, any internal division of
interests is considered dysfunctional as it contradicts the postulate of homogeneity. Although the functionalist
perspective on families has faced many criticisms, notably in feminist work, the lack of empirical research on family
conflicts in sociology is one of its long lasting consequence.
31
their family configuration; 19, one or all of their siblings; 18, their father; ten, one or more grandparents; ten,

aunts and uncles and seven include cousins. Seven respondents also include stepparents, and eight include a

partner. Hence, significant family contexts spread well beyond the nuclear family in several directions. In

another study based on a sample of 229 college students, there is a dominance of the mother, the father,

brothers and sisters, who are included by a large majority of respondents. Mothers are clearly prominent, as

almost all respondents include them as family members. Again, the rank of inclusion of mothers is the lowest

of all with the least variance. In other words, mothers are included first by a large number of respondents. As

such, they have the highest cognitive and emotional salience as family members. Fathers and siblings come

next, although they are less often included. Other blood relatives are also well represented, with grandparents

scoring highest. On average, relatives on the mother’s side are more prominent than relatives on the father’s

side. For instance, the mothers' sisters are included more than twice as often as the fathers' sister, and cousins

on mothers' side are much more frequently included in family configurations than on fathers' side. Thus,

individuals develop, on average, more interdependencies with their mother’s relatives than their father’s

relatives (Coenen-Huther et al., 1994; Firth, Hubert, & Forge, 1970) . In addition to blood connections,

friends are commonly include as family members; approximately 30 percent of respondents include a female

friend as part of their family, and 20 percent included a male friend. Partnership is also well represented, as

39 percent of respondents include their partners as significant family members. Seven percent extend their

significant family configuration to their in-laws by citing their partner’s mother. A stepfather and a

stepmother are included by about one respondent in ten.

By all these counts, significant family members extend well beyond the household and the nuclear

family. Non-resident parents and siblings, partners' relatives, stepparents as well as stepchildren and

stepsiblings, and friends considered as family members, constitute a significant share of family

configurations. Members of the nuclear family, that is the partner and the co-resident and dependent

children, are most of the time included first in the list of family members. They nevertheless are absent in

many family configurations just because, as we shall see, individuals are childless or single. When they are

present, other relationships support or interfere with the interdependencies existing with them.

32
Structured Diversity

Family configurations are diverse. Table 3 provides several telling examples of this diversity, drawn

from the sample of young adults. The respondent in the first family configuration includes her parents and

her siblings. She goes on with her mother's siblings, some of their children and partners, and with her

maternal grandmother. When she is done with this side of her family, she continues with her father's

relatives, although with fewer inclusions. Interestingly, a partner is not included in that case. As a result, her

family configuration is mostly constituted by blood relatives, with a balanced mix of uncles, aunts and

cousins. It does not focus on a specific family generation, as members of generations +2 (grandparents), +1

(parents, uncles and aunts) and 0 (cousins and siblings) are present. A rather similar composition

characterizes the second configuration but with the father's relatives exclusively. In that case, no mother's

relative is included in the family configuration; this creates a strong unbalance between the two sides of the

kinship network. This is not the case of the third configuration, in which both sides of the kinship network

are balanced, except for grandparents who come from the mother's side. Contrary to the other cases, the

fourth case is horizontal rather than vertical, as the siblings and their partners represent a large number of

persons and a significant share of the family configuration. The inclusion of the sibling's partners gives a

distinct significance to this configuration, as not only blood but also partnership constitutes an underlying

rule of the family configuration.

INSERT TABLE 2.2 HERE

Table 3 Composition of eight sampled family configurations

The fifth family configuration is also kinship oriented with the inclusion of grandparents. However,

uncles, aunts and cousins are under-represented. The configuration is more diverse as it includes a partner

and a number of in-laws, along with some friends. The emphasis on blood ties is weaker than in the two

previous configurations. Interdependencies, this example confirms, can be developed through partnerships

and friendships and lead to family configurations that are not inherited from parental kinship networks. The

diversification of family ties beyond blood ties is also present in the sixth family configuration, which is the

smallest of all. In addition to parents and siblings, there are three friends considered as family members.

33
Families by choice, in which kinship ties are replaced by chosen individuals from the pool of friends,

neighbours or co-workers, have become a reality. Interestingly, the presence of friends in that case precludes

the inclusion of blood relatives. We will further see if this is a general rule.

Another deviation from the world of blood ties is found in family configurations seven and eight. Both

of them include family members stemming from divorce and remarriage. In the seventh family

configuration, half-siblings as well as parents' partners are included. Interestingly, the father comes further

down in the list, suggesting that his cognitive and emotional importance is lower than in other family

configurations, in which fathers are included first. The family configuration seven is very small, as no blood

tie from the kinship network is added to the circle of parents and their new or former partners and children.

Overall, there is a great diversity among those family configurations. The number of family members

included varies, from seven to fifteen. The family members included are distinct as well, from mother and

father to friends or kinship ties such as grandparents, cousins, uncles or aunts. In some cases, in-laws

predominate; in other cases, one side of the kinship network (either the mother's side or the father's side)

have a much larger number of members. The realm of kinship is extended in various ways, related to

marriage, either of the respondent (in laws) or of her parents (stepparents, half-siblings). In all cases, a

logical order of inclusion is followed. Usually, partners, parents, children and siblings are included first in

the list of family members. As such, these listings share a common feature with other free listing tasks, in

which the most representative members of a category come always first (D'Andrade, 1995). They are

followed by other family members linked to each other. Family members come in bundle; respondents

include one side of their kinship networks first, and then the other side. There is also chaining across family

members: the mother's sister is always present if the mother's sister's children are included. In other words,

one does not include a remotely linked blood family member, without including the intermediaries that

connect the respondent to her. In that sense, family configurations build up by chaining and transitivity. Only

a few missing links can be pointed at in the listing presented in Table 3.

In order to uncover what underlying principles account for the composition of family configurations,

various exploratory analyses were run 9. Seven types of family configurations were revealed by cluster
9
A detailed presentation of this study is was published in the Journal of Social and Personal relationships (Widmer,
2006).
34
analysis, a method that helps researchers uncover subgroups in samples 10. A first type of family

configurations focuses on blood relatives, with the inclusion of family members of different generations.

Such family configurations have a balanced set of grandparents, from the mother’s side as well as from the

father’s side. They are vertically rather than horizontally oriented as they do not include aunts, uncles or

cousins. They also exclude friends, in-laws and step-relatives. Therefore, they have the smallest size of all,

with only 7.5 members on average. They were referred to as the “Beanpole” family type (Bengston,

Rosenthal, & Burton, 1990; Widmer, 2006) , that is, a family whose living members come from several

generations, but with few members in each generation. This type of family configuration, demographers

propose, is becoming increasingly prominent due to decreasing fertility rates (Ruggles, 1996). In fact, it is

the modal category in this sample of young adults, with one third of respondents included in it.

The second type includes respondents who add friends to the set of relatives of the beanpole family

configuration, while putting less emphasis on grandparents. In that type, friends represent as many as 4.5

persons, whereas other blood relatives, in-laws and steps are less present. Respondents in post-divorce

family configurations, by contrast, include a large number of step-relatives in their family configuration and

under-represent biological fathers 11. Stepfathers are on average more cited than stepmothers and come first

in the list of family members. In conjugal configurations, respondents show a strong orientation toward their

partner and in-laws. The partner and the partner’s mother are over-represented, as well as other in-law

relationships, such as the partner’s father and the partner’s siblings. Not only are partners included more

frequently than in other types of configurations, but they also come first in the list, with an average rank of 3

(compared with a rank of 5 on average). There is a focus on ties created by the conjugal bond for this small

minority of cases (5 percent). Other respondents develop a mother-oriented family configuration: the

mothers’ sister, the mother’s sister’s children, as well as the mother’s sister’s partner, are over-represented.

This is also the case, although to a lesser extent, of the mother’s mother and the mother’s father. This family

10
Cluster analysis makes it possible to go beyond the effects of specific relational dimensions and find configurations. It
has been used to construct typologies of networks in various research.. In all of the cases described in this book, we
examined a sequence of hierarchical cluster analyses based on Ward’s method of clustering. Instead of partitioning the
observations into some predetermined number of clusters in a single step, this hierarchical procedure produces step by
step splits (Everitt, 1993). Ward’s method minimizes within-cluster variance and thus produces good estimates of
cluster groupings.
11
Such configurations are considered in Chapter 5.
35
configuration is the second most common in frequency after the beanpole family configuration, as it

concerns about one respondent in five in the sample of young adults. The father-oriented family

configuration presents the same pattern as the mother-oriented configuration but on the father’s side:

relatives of the father are over-represented in it. Sibling family configurations are centered on siblings of

respondents and their partners and offspring.

Overall, all seven types of family configurations extend well beyond the nuclear family in this life stage.

Even the beanpole family configuration does not end with co-resident parents, children, partners and

siblings, as the large number of other significant relatives, especially grandparents, shows. The various types

of family configuration together demonstrate the existence of a set of informal principles accounting for the

inclusion of individuals in family configurations. Blood, especially on the mother’s side, is a major criterion

used to identify significant family members. Partnership is a second major criterion, working both at the

level of respondents (in-law family configurations) and at their parents’ level (post-divorce family

configurations). The development of “families by choice” (Weston, 1997) is a third criterion for defining

family members, with friends considered as family members constituting a significant share of family

configurations in young adulthood. A fourth criterion is constituted by the community of siblings. Overall,

there is much diversity in configurations of young adults, but also a large impact of structures. The diversity

of families is channelled into distinct types, responding to well-known principles of Western kinship systems

(Levi-Strauss, 1968; . Murdock, 1949).

Family configurations of Parents

There is a diversity of family configurations in early adulthood. This life stage is however a moratorium

(Erikson, 1968) in which various life styles can be experimented, thus probably leading to a greater number

of family alternatives than in other life stages. Individuals who went through steady partnership and

parenthood may have developed more standardized sets of interdependencies with their family members.

Therefore, results for other life stages are necessary before the variety of family configurations is confirmed.

In a study of 101 women with a partner and at least a co-resident child aged between 6 and 15, a variety of

family configurations was found once more.

36
In this parental stage, the beanpole family again constitutes a significant share of family configurations, with

the inclusion of respondents' parents and of their partners' parents. While all respondents in this configuration

include their own parents, eighty six percent of them have included a mother in law and 67 percent a father

in law. The conjugal configuration is again present, with its emphasis on the partner's relatives. For

instance, the partner's mother is more often included (55 percent) than the respondent's father (45 percent)

and the siblings in-laws are more often included than the siblings (1.54 versus .73). Therefore, respondents

shift their interdependencies towards their partner's side and for some reasons disregard the

interdependencies in their family of origin. The sibling configuration that was found for young adults is in

that case split into two more specialized configurations, once on the brothers' side, the other on the sisters'

side. In both cases, contrary to what happens for young adults, not only siblings are included in family

configurations, but also their partners and children. Therefore, the sibling group, that was already present in

young adulthood, is expanded and becomes a configuration of persons indirectly related by a sibling tie. In

both cases, a large number of friends are added to the siblings, so that these configurations become mixed 12.

Compared with the sample of young adults, a nuclear family configuration is present at that stage. In nuclear

family configurations, there is a smaller number of significant family members, with a focus on the partner

and the children. Other family members are also present, but without a clear emphasis on any category of

relatives, unlike in the other family configurations. The post-divorce type is also present in this parental

stage, although with a distinct composition. Instead of divorced parents, step-siblings and half siblings, this

type includes previous partners (most often the children's other parents) or the children of the current partner

(such as stepchildren). Interestingly, the father and siblings of the respondent are much less often included in

the post-divorce family configurations than in other configurations. The process of divorce and remarriage in

some cases involves estranging one self from the family of origin 13. Quite distinctly, other individuals have a

kinship configuration, in which grandparents, uncles, aunts, nieces and nephews are included, creating the

largest family configurations of all, with an average number of 15.5 family members.

Overall, there is much stability of the major types of family configurations from young adulthood to the

12
Therefore, the friendship configuration that is found for young adults is not present any more. It is replaced by this
horizontal family configuration.
13
See chapter 5.
37
parental stage. In other words, the main types of family configurations are present in both life stages, with

the exception of the nuclear family configuration. This does not mean that individuals remain in the same

configuration throughout their life, as these results are not longitudinal but compare distinct individuals in

different life stages. There are also differences among configurations at the two life stages: conjugal and

nuclear family configurations are much more present in the parental stage; beanpole family configurations

obviously do not include as many grandparents in older adulthood than in younger adulthood, and siblings

and friends are fused in a single family configuration in the parental stage. Although the same underlying

principles shape family configurations in both life stages, a significant adaptation to demographic changes

occurs in the process of growing older. In other words, family configurations evolve through the life course

as they adjust to births, divorces, deaths, household memberships, migrations and other life events. These

transformations, however, often represent structural shifts rather than revolutions. Strong interdependencies

with a sister, for instance, translate into an active role as an aunt when the siblings reach the parental stage;

the connection with the nephew later leads to the construction of a tie between cousins.

Family Structures and Configurations

The similarity of configurations in distinct life stages suggests that they may evolve on the basis of

constrains encapsulated in family structures, which refer to the number of children one has had, the relatives

still alive, the number of co-resident persons or the basic facts of divorce and marriage. Are family

configurations a full replication of family members actually being available? Individuals at earlier and later

stages of their life course are embedded in distinct demographic structures. The elderly, for instance, are

obviously embedded in a truncated family structure as they do not have relatives from their parents'

generation any more and because their own family generation has shrunk due to earlier deaths. Therefore, the

beanpole family configuration takes a distinct shape when individuals grow old. While it is true that the

underlying structures of family configurations vary according to life stages, they also vary according to

cohorts, ethnicity and social class, due to differential access to partnership, parenthood and social inequalities

of life expectancy (Bengston, 2001).

38
 The link between family configurations and family structures can be tackled by refering to the number

and status of relatives alive. In the study on the parental stage, respondents reported on how many children,

sisters and brothers, uncles and aunts they had, as well as whether their parents and grandparents were alive.

In the case in which they had divorced and remarried, they also reported how many children their current

partner had. Based on this set of information 14, four types of demographic structures were identified.

Avuncular family structures concerns cases in which there are many uncles and aunts. Grandparents have

had many children and the respondent has therefore a large kinship network. This does not mean that

interdependencies with uncles, aunts and cousins are active, but only that many relatives are available for

interactions and interdependencies. Vertical family structures focus on the respondent's parents and siblings

with a very limited number of uncles and aunts, as well as few cousins available. The extended family

structure is similar to the previous one but it expands towards the partner's relatives, who are also alive in

great numbers. In other words, the partner has a large set of living parents and siblings. In stepfamily

structure, the partner has children from a previous relationship and the respondent's children have half-

siblings on their father's side.

INSERT TABLE 2.2 HERE

Table 4. Distribution of family configurations by family structures (in percent)

Table 4 shows the links between family configurations and family structures. Individuals embedded in

an avuncular structure are overrepresented in nuclear family configurations. When parents pass away, and

only uncles and aunts remain, there is indeed a shrinkage of family interdependencies. By contrast,

individuals who are either in a vertical or an extended demographic structure are overrepresented in beanpole

family configurations. Still having parents around in adulthood is a condition promoting significant

intergenerational contacts that go beyond them and include a larger set of relatives. Obviously demographic

14
The indicators of demographic structures are rather limited in that case. Additional information should be included in
future research concerning a larger sets of relatives by blood or alliance (cousins, in-laws, etc.), as well as their spatial
and social locations.
39
structures which include steprelatives lead to an over-representation of post-divorce configurations. As we

shall see, however, post-divorce structures lead to other configurations than post-divorce configurations 15.

Overall, the statistical association between demographic structures and family configurations is neither

weak nor strong. This indeed proves that family configurations do not build up from a social vacuum: they

respond to the structural constraints imposed by marriages, divorces, births and deaths. But they are not fully

determined by the demography of the family either. Other dimensions of life such as the development of

trust and intimacy with specific individuals throughout a history of interdependencies add another layer of

complexity that makes each individual's family configuration unique in many respects.

Conclusion

Negative assessments of the Family by scholars or the mass media were triggered by the decreasing

importance of nuclear families in late modernity. The nuclear family is often considered as the natural form

that all families should have in order to be considered proper families. The pluralization of the life course,

since the nineteen sixties, have made the complexity of family interdependencies increase and resources are

sought beyond the boundaries of the nuclear family 16. There is nothing natural in the nuclear family, and its

decreasing functional importance is not synonymous to a decline of the Family altogether. By asking

individuals to define their family configurations, an attempt was made to set the boundaries of families by

using alternative criteria to the household, marriage or biological parenthood. Those criteria are indeed too

narrow to capture the set of interdependencies that link individuals with family members in late modernity.

The configurational perspective focuses on individuals and their family ties, rather than on families as

homogeneous and well defined groups. This theoretical shift has several empirical consequences. When

asked about their significant family members, individuals refer to a variety of family configurations that go

well beyond the household. Only a few of them acknowledge a lack of family members, which makes it clear

that the family realm is not on the verge of collapse. Family members still matter, even though the duration

of partnerships has decreased. The number of alternatives to manage for defining family configurations is

15
The link between the stepfamily in its demographic basis is considered in chapter 5.
16
The nineteen sixties may represent an historical exception for the long time span that individuals remain in
the same partnerships.
40
indeed significant. A variety of ways of building one own family ties exist Indeed, individuals build up

individualized family configurations17.

The individualization of family ties meets soon, however, some limits. The variety of family

configurations refers to a small number of models, with known underlying principles. Individuals focus on

their paternal or maternal ties, their partner's parents and siblings, their steprelatives or their friends. The

number of informal principles that structure family configurations is limited and the available solutions

logically organized. The family configurations that they refer to are not the result of free choice. As a matter

of fact, family configurations are embedded in demographic constraints, related with partnerships, fertility

options, divorces, migrations, social mobility and deaths. In that sense, family configurations are not the

highly personalized expressions of life styles or individual preferences, but, as we shall later see, the results

of series of events and transitions that happened in life courses of family members.

17
This diversity was already present in the past. For a long time, anthropologists have stressed that one
feature of kinship in the Western world is its bilaterality (Levi-Strauss, 1968; . Murdock, 1949): because
individuals get their kinship connections both on their mother's and father's sides, with little specialization of
exchanges on each side, individuals have never been parts of homogeneous kinship groups in the West as it
is the case in other cultures. They were rather pushed to develop personal connections within the large
number of relatives provided by blood or marriage. In other words, family ties have been personal for a long
time in European societies.
41
 CHAPTER 3 – Family Social Capital

For years, the Family has been considered a crucial group for maintaining social integration. In the

sociological research of the nineteen fifties and sixties, under the influence of Talcott Parsons and Robert

Bales and their well-known book “Family: Socialization and Interaction Process” (Parsons & Bales, 1956),

it was assumed that only nuclear families could take care of the functional imperatives of creating and

maintaining individuals fit into modern life by socializing them and providing them with various forms of

support. Since then, the quantitative importance of the nuclear family has diminished. Does that mean that

the contribution of the Family to social integration has decreased as well ?

Overall, the problem with scholarship emphasizing the integrative function of the nuclear family is that

it assumes that the family, as a cohesive and homogeneous group, has this function for all its members. From

this perspective, family life benefits to husbands and wives, parents and children alike. Strong inequalities of

access to emotional and instrumental support however exist between partners, or parents and children, in the

highly heterogeneous and individualized family configurations revealed in the previous chapter. Therefore, a

recent trend of sociological research has conceptualized the integrative function of families as social capital

(Furstenberg & Kaplan, 2004), a concept that encompasses individual resources stemming from the

possession of a durable network of acquaintance or recognition (Bourdieu, 1985). The more

interdependencies people have with other individuals, the more likely they are to mobilize resources, be they

instrumental (money, domestic support), informational or emotional. Social capital has various positive

consequences for individuals, such as promoting their physical and psychological health and increasing their

adaptability to disruptive events or demanding life transitions (Cohen & Wills, 1985). Overall, the concept of

social capital conceives social integration as an individualized resource rather than as a group attribute. As

such, it fits the approach of families as configurations. Rather than considering that a family as a whole has

to fulfil an integrative function in order to play its role in society, the social capital concept points at the fact

that individuals have unequal access to family relational resources that are meaningful for mastering key

dimensions of life, such as experiencing a fulfilling conjugal or parent-child relationship, or avoiding

42
depression and other psychological problems (Cohen & Wills, 1985; Duck, 1997; Widmer, Giudici, Le Goff

& Pollien, 2009).

Interdependencies within family configurations provide various resources to individuals, such as new

information, material help, emotional support and companionship. Family configurations are central in these

respects, as support of various kinds is more frequent among family members than in other configurations,

such as the work place. This chapter asserts that, although in many respects unique, family-based social

capital takes various forms, each with its own distinct logic. It begins by reviewing the evidence pointing to

the functional importance of family members beyond the nuclear family. Then, it proposes two alternate

conceptualizations of relational resources provided by families in terms of either bonding or bridging social

capital. It proceeds by showing that these two kinds of social capital depend on the composition of family

configurations.

Interdependencies Beyond the Nuclear Family

The theme that relatives are not of prime importance to the functioning of the nuclear family was clearly

enunciated in the functional perspective on families in the nineteen fifties and nineteen sixties (Parsons &

Bales, 1956). While family sociology, for the most part, has focused on parent-child and conjugal

relationships as autonomous entities, some research has contradicted this statement, identifying the

contribution of family configurations to the functioning of key dyads of the nuclear family (Widmer, 2004).

At first, research mostly focused on material exchanges linking nuclear families and their kinship networks.

Contrary to the thesis of the nuclear family’s isolation, individuals have regular contacts with relatives. If

alternatives to face-to-face interactions, such as phone calls or emails, are also considered, a large majority of

individuals maintain meaningful interactions with relatives. There are, however, great variations in the

amount and significance of interactions that individuals develop with them. Some individuals interact daily

with relatives while the interactions of other individuals are more casual. Of course, family interactions

depend on the composition of one’s family configuration. The number of relatives in the area influences the

amount of regular family interactions. Men and women differ in their interactions with family members.

Typically, women have a higher proportion of contacts with family members than men; they are more

43
frequently in charge of family sociability, organizing special gatherings or regular forms of contact with

relatives on their side and their partners’ side. Individuals are not only embedded in interactions with

relatives; they are also functionally embedded in them. In the United States, as in Switzerland, a large

percentage of individuals benefit from financial, domestic and emotional help provided by family

configurations. In other words, interdependencies, not isolation, is the rule in family configurations, with

only a minority of individuals being isolated from their family members.

If the thesis of functional isolation of the nuclear family proved to be inadequate, there are, however,

obvious limits to the exchanges existing within family configurations 18. In many cases, significant exchanges

only concern the immediate kinship circle, drawn from the individuals’ parents and siblings and their own

children. Most financial and domestic help is provided by relatives in a direct ascending line. In other words,

instrumental support mostly circulates from parents to their adult children. Parents also provide tangible help

with raising their grandchildren. Only a limited proportion of instrumental exchanges within family

configurations concerns relatives beyond those genealogical limits.

The instrumental assistance in horizontal relationships, for instance from siblings, is much less intense.

As compared with parents, adult siblings do not support each other as much in terms of money or domestic

tasks. In addition, normative imperatives to instrumental help are fairly low in sibling relationships, contrary

to what is expected in parent-child relationships. Siblings are strong resources of emotional support however,

frequently surpassing parents in that regard. Much of the support beyond parent-child ties is emotional and

based on affinity rather than on status. Instrumental support in family configurations is limited as it only

concerns a small number of persons. Specific events such as giving birth, changing homes or divorcing

trigger much of the instrumental help provided in family configurations, although long lasting instrumental

exchanges and support exist in some cases, especially for couples with small children. Overall, however,

exchanges should not be considered as indicators of permanent material interdependencies within family

configurations.

18
This review and its main results are further developped in a previous publication (Widmer, 2004).
44
 That said, many family interdependencies are not instrumental. Indeed, emotional and cognitive

interdependencies are main functions of families, as Parsons and Bales (1956) rightly pointed out 19. In

that sense, family members are part of what family researchers Surra and Milardo (1991) called

psychological networks, which are composed of people to whom individuals are committed emotionally,

who provide individuals with a concept of self, and who can sustain or alter one’s identity through

communication20. As underscored by the anthropologist Firth and his colleagues, already in the nineteen

sixties, the concept of family is of strong affective significance because “it expresses a sense of identity with

specific persons who are members of one’s kin universe” (Firth, Hubert, & Forge, 1970). Despite the fact

that they live in different households and no longer belong to the nuclear family of respondent, parents and

siblings of adults continue to have great emotional and cognitive significance. For instance, most siblings

wish to keep in touch with each other and are each aware, at a minimum, of the whereabouts and general

circumstances of the other. Social comparison between siblings, even in adulthood, is frequent. The

cognitive function of family members does not end when they do not share a common residence any more.

The importance of family members as orientational others 21 point to emotional support and perceived support

as main sources of social capital in families.

Transitivity

Research on family resources focuses on quantity of support, either instrumental or emotional. This has

proven to be a useful approach to the subject, as quantity of support does indeed matter for conjugal and

parental relationships: individuals with more support do better on a variety of outcomes (Cohen & Wills,

1985; Sandler & Barrera, 1984; Sarason et al., 1991; Widmer & Weiss, 2000). But quantity is not the only

option; quality or, in other words, “structure”, also matters. So far, the structures of social capital provided

by family configurations is not well known. The way in which support is organized within family

configurations has implications for what happens in key family dyads. One important result of research on
19
This point may seem obvious. However, most large datasets available on intergenerational transfers in the United
States or Europe put a strong emphasis on material and financial support and relegate the issue of emotional transfers to
a marginal status. Therefore, we know little about the emotional and cognitive dimensions of family configurations.
20
Note that, once established, the persistence of someone as a member of one's psychological network does not require
frequent, recent, or long interactions.
21
See introduction.
45
social networks is that relationships tend to promote a state of transitivity or so called “balance”. In the case

of families, this state is achieved when emotionally close family members of close family members are also

close, or when close family members of disliked or remote family members are kept at distance or are

disliked as well. On the other hand, a state of intransitivity or imbalance exists when close family members

of close family members are disliked or disregarded, or when disliked family members of close family

members are emotionally close. Figure 3 presents a transitivity case and an intransitivity case in two

hypothetical family triads. In the transitive case, the respondent emotionally depends on her mother, who

depends on her partner, and the respondent depends on her stepfather as well. In the intransitive case,

although the respondent is supported by her mother who is supported by her partner, the respondent does not

acknowledge her mother's new partner as an emotional resource. This situation is likely to create

complications.

INSERT FIGURE 3.1 HERE

Figure 3. Transitivity in triads

As a matter of fact, following the theory of Fritz Heider (1946; 1958), a renowned psychologist, people

feel uncomfortable when they believe the members of their personal networks do not like each other.

Therefore, they try to avoid such states in their actual relationships. They also underestimate the occurrence

of this situation and overestimate the occurrence of relational balance. In other words, they tend to organize

and perceive their relationships with others transitively. As a matter of fact, individuals are more at ease

reporting transitive relationships existing in their configurations. They remember transitive structures better

than non-transitive ones, and they systematically bias the perception of their own relationships toward

transitivity22.

These statements are supported by a number of results from research on social networks. It is impossible to
22

make reference to all researchers who contributed to this line of research (for instance, Davis, 1979; Davis &
Leinhardt, 1972; De Sotto & Albrecht, 1968; Freeman, 1992; Heider, 1958; Krackhardt, 1987; Killworth &
Bernard, 1976; Kumbassar, Romney & Batchelder, 1994; Moreno, 1934; Newcomb, 1961).

46
 This tendency of social relationships towards transitivity has various implications for groups. First, it is

a prerequisite for the ordering of status, prestige, power and influence in groups: when triads are transitive,

situations of “double bind” or relational inconsistencies, in which two individuals give contradictory orders

to a third individual, seldom occur, as individuals are placed in an ordered line of command: Individual A

obeys to individual B who obeys to individual C, and A also obeys to C. Authority is much strengthened in

transitive triads. Support also has a specific quality when it is organized transitively: The fact that an

individual receives support from a second and a third individual, while the second individual also receives

support from the third individual, creates a high level of consistency in support compared with a

configuration in which the first individual receives support from a second individual, who receives support

from a third individual but the first individual does not receive support from the third individual. In a

transitive triad, the two helpers can coordinate themselves while supporting the first person, which is not the

case in an intransitive triad. Transitivity also increases role differentiation and clustered subgroups. When

any two persons indirectly connected by a third one are directly connected, the likelihood of having sparse

configurations is low and the likelihood of finding clearly defined group boundaries in a network is high.

Transitivity leads in time to dense configurations or, at least, to dense subsets of relationships in a

configuration, that is subgroups of individuals linked to each other by third parties. In a word, transitivity

creates social integration within configurations.

How does that apply to family configurations? Family ties frequently develop following a transitive

logic. The British anthropologist Harris made this point: “Persons related by familial kinship matter because

of familial involvement rather than because of kinship recognition.” A sister’s son matters because of the

emotional links between the two sisters, and the mother with the son, though the aunt lacks a direct

emotional involvement with her nephew in the first place. The emotional significance of the nephew is the

result of the juxtaposition of two familial relationships (Harris, 1990). This process is possible because

intimate ties are likely to combine and create additional relationships. This is especially true, as suggested by

Harris, in the case of blood ties, for which there is a tendency to spread across generational lines. Because of

parents, grandparents and grandchildren connect together. Because of siblings, uncles and nephews, aunts

and nieces, become interdependent.

47
 In the exploratory research done on young adults in the United States 23, 17 out of 25 individuals

perceived their family configurations to be transitively organized. Therefore, in the majority of cases,

individuals had structurally balanced family configurations. This result indeed confirms the intuition of

James Coleman (1988), for whom family ties are synonymous to bonding social capital, e.g., a set of

strongly interconnected individuals linked by long term, multifunctional, status-based and transitive ties.

Much of the work emphasizing family relationships as social capital is indeed based on its definition in terms

of bonding social capital. From this perspective, social capital is hypothesized to stem from family

configurations characterized by an orientation towards blood relatives, who have usually known each other

for a long time and have statutory obligations to remain in contact. The links between siblings, their own

offspring and their parents is secured by a series of reinforcing interdependencies associated with the family

history across generations. Cousins learn to know each other and respect their aunts and uncles under the

pressure of their parents, who are subordinated to their own parents. The transitive and well connected

configurations stemming from blood relatedness enhances expectations, claims, obligations and trust among

individuals because of an increase of normative control. It also facilitates communication flows by

multiplying the number of information channels and reducing the number of intermediaries within a chain. In

closed family configurations, support has a collective nature, as several individuals coordinate their efforts

when helping another.

Of the 25 American college students, however, eight presented significant deviation from transitivity.

The assumption that transitivity and bonding social capital are overly dominant in contemporary families is

indeed challenged by demographic and social changes, sometimes confusingly interpreted as family decline.

Family configurations of late modernity create room for imbalance and intransitivity. As a matter of fact,

scholars have emphasized that relationships among stepparents and stepchildren are different than

relationships between parents and children. For instance, stepfathers develop a strong relationship with their

partners, who keep a strong tie with their biological children from their previous marriage. However, because

relationships among stepparents and stepchildren are less intimate, less supportive, and are associated with

23
See chapter two.
48
more conflicts than relationships between parents and children, remarriage is likely to create intransitive

triads such as the one presented in Figure 3, in which two strong ties coexist with a negative or weak tie.

Relationships with in-laws are associated with structural imbalance as well, as they are more distant or

more tense than relations with parents. Although the partner to partner relationship and the mother-son

relationship are usually strong, the mother-in-law - daughter-in-law relationships are often associated with

tensions, which are even greater when the daughter-in-law has a child (Fischer, 1983). Those tendencies

create imbalance in triads that include in-laws. Thus, people are likely to report a significant amount of

intransitivity in their set of family members. For instance, both a parent and a partner are close to a person,

but at the same time, this person acknowledges that they are only weakly, and even negatively, connected to

each other24.

Consider again the case of Betty, this 54-year-old female, with two children married twice, with both

marriages having ended in divorce. Interdependencies in Betty‘s family configuration follow a complex

pattern. In order to measure social capital in this family, we asked each of them to report the degree of

closeness they feel toward other members 25. Results for the family configuration as a whole, as well as for

only those members included directly by Betty, are reported in Table 5.

INSERT TABLE 3.1 HERE

Table 5. Closeness in the family configuration of Betty (percent)

Only a minority of relationships are among ‘very close’ or ‘close’ persons. Two thirds of the

relationships among family members of Betty are considered ‘somewhat close’ or less. This result

holds true even when one considers only individuals included by Betty. It is especially noteworthy

that 28 percent of the configuration members do not know each other at all. This example points at

the fact that bonding social capital and transitivity cannot be taken for granted in contemporary families. The

24
One solution to this dilemma is for the individual to develop strong interdependencies with her in-laws. This solution
has however consequences for the partnership and for interdependencies with blood relatives.
25
The question was: “ For each of the following persons, please tell me if s/he is:” a very close person to you; close;
somewhat close; acquaintance; barely know; not known at all.
49
lack of transitivity of relations with step-parents, in-laws and former partners suggests that many of them do

not provide a bonding type of social capital.

Does that mean that Betty lacks social capital? When asked, she reported that she could indeed count on

her family members, even though she had to do it on an individual basis rather than to expect a collective

type of support, except for support provided by her siblings. The situation created by her two divorces is

indeed a peculiar one, revealing another type of resources than bonding social capital. A second

conceptualization of social capital applies on the experience of Betty by the bridging potential of social ties.

Contrary to Coleman’s perspective (1988), Ronald Burt (2001), a sociologist and network researcher,

describes social capital as a function of brokerage opportunities: the weaker interdependencies between

subgroups of a network create holes in the structure that provide some persons – brokers or bridgers - with

opportunities to mediate the flow of information between network members and hence to control the projects

that bring them together. Burt, along with another sociologist, Mark Granovetter (1973), argues that full

transitivity may be detrimental to individuals, as it is associated with social homogeneity, a lack of personal

control over oneself and an absence of non-redundant information. According to research on firms and trade

markets, brokerage positions in networks provide greater opportunities to innovate, a larger autonomy in

transactions, and access to new information. Bridging social capital, that is, the ability of playing a

gatekeeper position between two unrelated sides of a network, is a resource provided to individuals by some

family configurations more than by others. Betty has indeed an intermediary position between her previous

partner, her children, her friends considered as family members, her siblings and her mother. This situation

has some advantages for Betty, as she has achieved a level of autonomy that is unknown in dense family

configurations. This enables her, as we will see in chapter 7, to deal with the relationships with her mother in

a responsive way.

Dimensioning Social Capital

Social capital is a multidimensional construct. Bridging and bonding social capitals should not be

regarded as opposites, as they may, in some cases, appear together. How do we know what types of social

capital is present in a family? First, the sheer number of individuals present in the family configuration is an

50
indicator of social capital: the more family members there are, the larger the family-based social capital is.

This is even more true when focusing on supportive family members. Knowing how many others are

resources for respondents permits to measure the extent to which they are embedded in a set of active

relationships. Overall, this indicator refers to a bonding type of social capital, as interdependencies have a

local character26.

Density, which is computed as the number of existing connections divided by the number of pairs of

family members included by respondent, e.g., potential connections, is another indicator of bonding social

capital. It can be computed for respondents’ supporting family members, for respondents' supported family

members, or for the full family configuration. In his founding article on social capital, James Coleman

(1988) stressed the beneficial impact of dense set of interdependencies on various kinds of communities,

including schools and businesses. Dense networks indeed feature many transitive triads. When dyadic

relationships are under the scrutiny of third parties that are interconnected, the two individuals can trust each

other more, as they know that deviations from the rules will be known throughout the network. Parents in

contact with each other and with the school authorities are more able than disconnected parents to frame their

children's behaviours and support them in their school work. The higher the density, the more likely the

individual is to be embedded in collective support and control. For instance, the density of the family

configuration presented in Figure 4 is .47, meaning that about half of the supportive ties possible in this case

exist. This is significantly more than in Figure 5, where only one fifth of possible connections are perceived

by respondent as existing. Density is the most commonly used indicator of bonding social capital.

A third measure concerns the number of components present in family configurations. A component is a

subset of individuals that is disconnected from the configuration, and in which all individuals can reach each

other either directly, or indirectly throughout their interdependencies with intermediaries. Two persons who

are connected neither directly nor indirectly do not belong to the same component. The more components

there are in a family configuration, the less it is able to work as a group. The change of number of

components when the respondents are removed by the researcher from the graph capturing their family

26
They are local as they concern the individuals to whom the respondent is directly connected by strong ties. Recall
that all family members are not necessarily supportive, as they may be included in family configurations because of the
conflicts and tensions that they have generated (see chapter one).
51
configurations is an informative indicator of bridging social capital. If the number of components greatly

increases when respondents are taken out of the graph, their position of intermediaries is confirmed: their

bridging social capital is high, as their removal makes the graph significantly less connected. For instance,

the family configuration presented in Figure 5 splits into three components when the respondent is removed,

while the family configuration in Figure 4 remains fully connected after her removal.

Finally, the betweenness centrality captures the proportion of connections for which the respondent is in

a position of intermediary. In Figure 4, for instance, the respondent has a high betweenness centrality (both

in terms of supportive and supported family members) as many members of her family configuration do not

have direct connections with each other, but must use the respondent as an intermediary in order to be

connected. This is not the case of the family configuration in Figure 5, in which a large majority of family

members have direct connections with each other, and therefore do not need the respondent as an

intermediary. Betweenness centralization is a generalization of the respondent’s betweenness centrality. It

measures the proportion of interdependencies in the family configuration captured by any individual. A

family configuration is said to be centralized if a small number of individuals lie between all other members’

chains of interdependencies27.

The combination of size, density, components and centrality provides a good approximation of the

structures of social capital in family configurations. They should be used conjointly, as some individuals

may benefit from both bridging social capital and bonding social capital. Four types of relational resources

hence characterize family configurations (Table 6).

INSERT TABLE 3.2 HERE

Table 6. Types of social capital in family configurations

Individuals with no family-based social capital are those with the least interdependencies. They are

disadvantaged in terms of emotional and instrumental resources. Individuals with bridging social capital and

no bonding social capital are likely to have experienced a series of family discontinuities of various natures,

such as regular changes of residence or serial divorces and remarriages, both of their parents and on their

27
Usually, respondents are the most central nodes in the graph, as the Family Network Method measures personal
networks, also called “ego-network”.
52
own. Those experiences might not necessarily be disruptive in an ontological sense pertaining to their

identity, but they indeed create many focus points 28 in a life time and, as a consequence, distinct subsets of

family members, for which they are bridgers or brokers 29. Individuals with bonding social capital and no

bridging social capital in their family configurations are likely to have experienced a stable life trajectory

both in their childhood and their adulthood, with a swift transition from the household of their parents, to

their own status as partners and parents. Those individuals have supportive but also quite controlling family

configurations, which may help them to progress in their local environment, but which do not link them with

a diversity of family members and which decrease their structural autonomy. Individuals who have a

comprehensive type of social capital are the most adaptative to the constraints of modern life, as they benefit

at the same time from local embeddedness and from diverse and open family environments. As experiences

learned in the family contexts during childhood and adolescence have enduring consequences 30, being

confronted to any of this four types of social capital is likely to shape distinct life trajectories and life

projects, in which localism and openness are unequally developed.

Blood and Friendship in Families

Does the composition of family configurations31 matter for social capital? All family configurations do

not provide the same amount and type of social capital. In the previous section, we stressed that blood ties

may create more transitivity than non-blood ties such as marriage. Thus, two partners frequently have quite

different family members, because they do not have the same blood connections (Stein, Bush, Ross & Ward,

1992). Indeed, the types of family configurations presented in Chapter 2 have consequences for bonding and

bridging social capital32. As exemplified in Figure 4, respondents in beanpole family configurations, where

28
See chapters 7 and 8 for a detailed presentation on the interrelations between life events, focus points and family
configurations.
29
This indeed represents a specific kind of family integration, in which individuals are intermediaries among otherwise
disconnected subgroups. The model of Simmel (1999) on intersecting circles is a classical reference accounting for this
type of social integration.
30
A large literature exists on the impact of non-normative events in childhood on the development of individuals in
adulthood life. For instance, individuals who have experienced a high level of family conflict in childhood or
adolescence are more likely to develop poor conjugal satisfaction, a phenomenon described as belonging to the divorce
cycle (Wolfinger, 2005).
31
See chapter 2.
32
These results are presented in more details in a series of publications (Widmer, 1999b; Widmer, 2004; Widmer, 2006;
Widmer, 2007).
53
blood relationships are dominant, have a low centrality, as their family members are commonly

interdependent. Although the mother’s side and the father’s side do not fully overlap, they are connected by

the mother’s and the father’s interdependencies with their own in-laws. The respondent is part of a large

group of siblings, which is itself included in a large, mostly blood-oriented, family configuration. The central

position of the grandparents is typical of beanpole family configurations.

INSERT FIGURE 3.2 AND 3.3 HERE

Figure 4. A beanpole family configuration

Figure 5. A friendship family configuration

In family configurations characterized by blood relationships, individuals are embedded in dense sets of

interdependencies. Respondents have a low centrality in their own families, e.g., many interdependencies

among their family members do not depend on them, and their family configurations are resistant to their

own removal. As these configurations include the grandparents, the parents, some uncles, some aunts, and in

many cases, some cousins, they represent a typical case of bonding social capital, with respondents being

under the scrutiny of a large number of interconnected blood relatives. Recall that this type of social capital

has advantages and disadvantages of its own. Individuals in these family configurations have less autonomy

in their everyday life and decisions. But they also benefit from a collective rather than dyadic form of

support and normative framing. When family configurations are dense, emotional support flow through

multiple channels, with the probability of having them all at one point suddenly disrupted being small, and

the likelihood of having several persons collaborating in giving support being high. Dense family

configurations also exert more normative pressure towards conformity, as many family members interact

while shaping each individual's behaviours and ideas.

Figure 5 features a friendship family configuration: although friends develop interdependencies with

each other, they are disconnected from other family members, except for the respondent. Thus, because of

this separation of the family configuration into two or more components, respondents have a high centrality

in such family configurations and are integrated in a large number of otherwise disconnected subgroups. In
54
friendship family configurations, bridging social capital is hence dominant. These “star-like” family

configurations split into several separate subgroups when respondents are removed. Therefore, individuals

control the flow of communication much more than in other types of family configurations. In fact, family

members cannot contact each other without their help. Therefore, they have a great amount of autonomy

from their family members in their everyday lives and decisions. However, they do not benefit from the

collective support and the normative framing associated with beanpole family configurations. On the other

hand, these “star-like” family configurations also exert less normative pressure towards conformity, as

family members are much less able to collaborate for influencing respondents; they may even exert

contradictory normative influences on respondents.

What happens when older adults are considered? Overall, identical results were found in the study of

101 women in older adulthood33 regarding the impact of blood and intergenerational ties versus non-blood

horizontal ties on bridging and bonding social capital. Family configurations in which a high number of

blood ties exist and that include parents of the respondents feature a higher density, a lower centralization

and a lower number of components than family configurations constituted by siblings and friends, which

develop a bridging type of social capital 34. Families by blood or by choice (Weston, 1997) have indeed

distinct consequences for social integration also in the parental stage.

These results shed light on the effects of family configurations on social capital. Family configurations

vary in the extent to which they include fathers, stepparents, parents in laws, friends, siblings' partners and

children, and former partners 35. These variations have a strong impact on the social capital that they make

available to individuals. In terms of bonding social capital, family configurations oriented towards

grandparents, uncles, aunts and cousins are optimal. On the other hand, they lack the autonomy provided to

individuals by bridging social capital in friendship family configurations.

Trusting Family Members

The importance of trust for the Family and for society has been underlined in the last decades. Trust

allows individuals to be active in a variety of situations while having only a limited knowledge and little

33
See Chapter two.
34
The case of post-divorce family configurations will be considered in chapter five.
35
See Chapter one.
55
control over them. It makes actions and relations possible in situations where face-to-face interactions are not

possible (Giddens, 1990). Therefore, trust is a resource for individual actions as well as for collective

purposes in families. It makes it possible for individuals to take risks while accepting the possible state of

disappointment when a decision is made that is contrary to their expectations. Trust helps people get over the

complexities of family interactions in late modernity.

In a functionalist perspective, the nuclear family is normatively trustworthy. Unless one's nuclear

family is on the verge of collapse, one should trust its members. The mother-wife is normatively defined as

the main provider of love, and the father-husband as the main breadwinner; children also have specialized

roles according to their sex and their age. In this perspective, all members are trusted, as they fill clear

functional roles and social statuses in the family unit. The normative dimension of family statuses may

however have less importance in late modernity societies, because they are associated with less clearly

defined expectations. The role of a stepfather or a stepmother, for instance, was acknowledged as

normatively undefined (Cherlin & Furstenberg, 1994); expectations toward stepparents are not provided by

the societal context. In a time of desinstitutionalization of marriage itself (Cherlin, 2004), it may well be that

trust in family roles beyond the nuclear family has gained some importance. In order to see how trust

develops in families, we again referred to the family configurations of women with a least one co-resident

child in their household36. Trust was measured by asking how much respondents trust each of their family

members37.

Results show that trust in family members goes well beyond the nuclear family, as it concerns not

only spouses or co-resident children but also non-co-resident parents and siblings. Trust is therefore not

limited to the household but crosses its boundaries to link individuals who have a history of shared intimacy.

Family statuses, however, cover a variety of processes that, to a large extent, explain their effect on trust.

Indeed, trust is accounted for by dyadic processes, such as the development of reciprocal supportive

interactions between family members. Father, mother and siblings are especially trusted in adulthood. These
36
Analyses on trust and family configurations are presented in more details in a forthcoming publication (de Carlo &
Widmer, forthcoming).
37
. The question reads: “For all persons included in your list, can you tell me how much you trust her?” The possible
responses were: absolute trust, a large trust, some trust, low trust, no trust at all. The answers were initially coded on a
five point scale. The scale was then dichotomized to perform logistic regressions. The active modality in further
analysis corresponds to higher scores on the scale.
56
family statuses are activated by a variety of relational interdependencies between specific persons. In other

words, individuals trust in their relatives in adulthood because positive and reciprocal interactions have

developed throughout their lives. Doing the family work (Schneider, 1980) in daily interactions provides the

basis for trust to develop. The highest levels of trust are found in cases in which both respondents and family

members provide and receive support. Another main factor of trust holds in the extent to which family

members achieve a high level of priority in one's life, a situation we referred to as loyalty. Various events

such as weddings, birthdays, and other collective gathering create new focus points in which the

interdependencies with family members can be expressed publicly by gifts and other material means (Finch,

1989). The level of interdependency toward family members is also expressed when non-normative events

such as divorce, health problems or death of a family member occur. The extent to which individuals care for

each other in those hardships provides clear indications to all about the level of priority given to each

relationship.

Overall, the development of trust relates to family members with whom a long history of positive

interdependencies exists. Those individuals belong to the larger family configuration and respond to its

organization of interdependencies. Respondents embedded in dense and supportive family configurations

have a much greater likelihood of benefiting from a high level of trust. In other words, the level of individual

trust in family members not only depends on one's relationships with others, but also on the patterns of

interdependencies that exist overall in family configurations. Families in which there is a great number of

supportive interdependencies trigger a high level of trust in each dyad. In other words, bonding social capital

is associated with greater trust. Therefore, trust does not only depend on the personal ability of respondents

to develop active relationships with family members, but also on the overall density of interdependencies in

their family configurations. As we saw in the previous section, the extent to which family configurations

develop dense sets of ties depends on their composition. Indeed, beanpole family configurations develop a

higher density of supportive interactions. The inclusion of parents, siblings, grandparents, uncles, and aunts

in one's family configuration makes the likelihood of benefiting from trust higher because many family

members are interconnected. In other types of family configurations, such as friendship families, with a

lower density of interdependencies, trust is linked with the development of dyadic relationships, without
57
additional positive impact of the family configuration. In all cases, trust depends to a large extent on the set

of reciprocal supportive relationships that have been built, or not, in childhood, adolescence, and early

adulthood by individuals. These sets of dyadic interdependencies are not fully personal, however, as they are

embedded in larger configurations of family interdependencies.

Emotional Interdependencies in an International Perspective

How do the United States and Switzerland currently stand compared with other countries concerning

emotional support provided by interpersonal relationships? To what extent family members beyond the

nuclear family are defined as important sources of emotional support? Comparing families across cultures

has been a concern of anthropological research for decades. Scholars have pointed out that many societies

around the world share common normative attitudes toward family relationships, including the incest taboo,

the regulation of sexuality and fertility by marriage, and a concern for intergenerational relationships (Lévi-

Strauss, 1949; Murdock, 1949). The same authors however also underline that family relationships present a

great variety of patterns from culture to culture. Even in the Western world, countries differ in their

emphasis on the nuclear family, paternal or maternal kin.

To address the importance of family members for social capital in various countries, we capitalize on data

from the International Social Survey Program (ISSP), made available by the Zentralarchiv fuer Empirishe

Sozialforschung, Koeln. An established program of cross-national collaboration, the ISSP has facilitated social

science surveys since 1985 (Smith, 1992). In 2004, independent research institutions replicated survey questions

on social networks. Data are available for 37 mainly Western and industrial countries. Because the focus of this

book is on family configurations in the United States and Switzerland, two democratic countries with a market

economy, we focus on countries that share these features. These include formerly socialist states of The Czech

Republic, East Germany, Hungary, Poland and Slovenia; the Southern European states of Italy and Spain; the

Scandinavian social democracies of Norway Denmark and Finland; the conservative welfare states of France,

Austria, and West Germany; and the liberal welfare states of Australia, Canada, Great Britain, New Zealand,

Northern Ireland, Switzerland and the U.S. (Esping-Andersen, 1999; Stier, Lewin-Epstein, and Braun, 1999).

Sample sizes range from 428 in East-Germany or 912 in Great-Britain to 1560 in Norway. ISSP data improve

58
on data presented in this book, both because of the large number of countries included and because of their

representativeness38.

Respondents were asked the following question: “Now suppose you felt just a bit down or depressed,

and you wanted to talk about it. Who would you turn to first for help?” An additional question was then

asked: “And who would you turn to second if you felt a bit down or depressed and wanted to talk about?”

Both questions had the same answer categories that enlisted family members and friends: husbands-wife-

partner, mother, father, daughter, son, sister, brother, other blood relatives, in-law relatives, close friends and

a series of alternative social roles such as neighbours, members of the clergy or psychologists, which we

recoded into a single category. A response category was also included for individuals who said that they had

nobody to turn to. The percentage distributions for the 20'000 respondents were aggregated into a matrix of

20 rows (the countries) by 24 columns (the 12 response categories for each of the two questions), which is

presented in Table 7 and Table 8.

Overall, partners and friends are defined as the most significant providers of psychological support.

More than half respondents cite a partner as a significant support provider. This does not mean, however,

that only partners are important. Indeed, on average, 18 percent individuals cite their mother as significant,

16 percent a daughter, 13 percent a sister. Male family statuses are less frequently cited, as fathers are only

cited by four percent, son by eight percent and brothers by five percent of respondents. Other blood

relatives only account for three percent of citations, as are in-laws. Only a very small minority do not cite

someone as a significant help provider. Overall, there is a diversity of family statuses cited beyond the

conjugal tie. Adults indeed keep on being connected with their parents and siblings, as the results show.

Older adults use their children as support providers. This is especially the case for the second individual to

whom one would ask emotional support in case of need. Table 8 indeed shows distributions with much larger

variances than Table 7. Therefore, family ties are used as backup and support for partnerships.

INSERT TABLE 3.3 HERE

38
Note, however, that ISSP data do not enable researchers to address the structural dimension of support as they do not
include information about ties linking family members of respondents with each other. Only the ties between
respondents and their family members are considered. They also focus on positive ties and exclude conflicts and
ambivalences.
59
Table 7. First individual to be called for emotional support ( percent)

INSERT TABLE 3.4 HERE

Table 8. Second individual to be called for emotional support ( percent)

Beyond commonalities shared by all countries, do countries present distinct patterns concerning

support providers? A data reduction strategy is called for in order to identify such configurations of

relationships in such large datasets. Cluster analysis was used to group countries into “attitude regimes”

that share common views about support provided by family members and friends. From hierarchical cluster

analysis, three clusters of attitudes towards friends and family members emerge, which are summarized in

Table 9.

INSERT TABLE 5 ABOUT HERE

Table 9. Cluster of attitudes towards family members and friends as emotional support providers
(percent)

The United States, Switzerland, Italy, Northern Ireland and France belong to a first cluster of countries

characterized by a lower centrality of partners in comparison with other countries. The United States and

Switzerland, despite their distinct linguistic contexts and the spatial distance separating them, are part of a

group of countries with a more diverse set of significant family members than other countries. In the two

contexts in which the analyses presented in the previous chapters took place, conjugal relationships have less

clear-cut primacy than in other countries. The United States and Switzerland are known for their high rate of

divorce and a minimalistic intervention of State policies in family life, as well as a liberal type of social

welfare. The explanation is however not straightforward, as some of the other countries included in this

cluster share a fully different historical path, with lower rate of divorce. Australia, Germany (East and West),

Great-Britain, Norway, New-Zealand, Canada, Denmark, Finland belong to a second cluster in which there is

more emphasis on partners whereas the other family terms are slightly less frequently included as help

providers. Indeed, other blood relatives and in-laws are less often cited than in the first group of countries.

This focus on partners is however not overwhelming, as only half of respondents in those countries include

60
partners as their first help providers and 59 percent include a partner overall. Austria, Hungary, Czech

Republic, Slovenia, Poland and Spain constitute a third cluster of countries in which individuals are slightly

more centred on children. Daughters are included by 20 percent of individuals belonging to this cluster as

providing support, against 11 percent in the kinship oriented cluster and 15 percent in the partner oriented

cluster. The emphasis on children is therefore somewhat stronger than in other contexts.

Overall, the profiles of the three clusters are very similar and the differences among them rather slight.

Are they family models or variants around a common cultural understanding of family relationships? To

assess the relative significance of cluster effects and country effects on family social capital, we generalize a

one-way, nested ANOVA design, following a methodology that we have developed to investigate other issues

(Widmer, Treas & Newcomb, 1998; Treas & Widmer, 2000). This yields a decomposition of the sums of

square of Tables 14 and 15 into the components attributable to shared attitudes, attitudes common to multi-

country clusters, and attitudes specific to individual countries. The three component sums of squares add to the

total sum of squares, that is the total variations of attitudes towards helpful others in the set of twenty countries

considered. The larger the sum of squares between clusters is , the more distinct the cluster from the average

attitude profile. The smaller the within sum squares is, the more homogeneous the cluster and the less

important national cultures 39. The total variance in tables 14 and 15 is 4.87 and the consensus model is 4.48,

that is, 92 percent of the total variance. For the three clusters, the between-cluster sum of squares is 20 (four

percent) and the within-cluster sum of squares is .19 (four percent). The consensus across the 20 western

countries considered here is therefore very high, as only a very small portion of the variance is attributable to

clusters or countries or to profiles specific to countries. Overall, there is a large homogeneity in the way in

which country profiles consider emotional support provided by family members and others. The consensus

is significantly greater in this realm than in the realm of attitudes towards non marital sex or women work

(Widmer, Treas & Newcomb, 1999; Treas & Widmer, 2000). The cluster profiles only account for a very

39
We compute the sum of squares across the 20 rows of table 14, using the "average" response as the basic
model. The average response is 8 percent, the result of dividing the 100 percent total for each of the two
variables (first supportive individual and second supportive individual) by 12, its number of response
categories. Thus, we use .08 to compute the sum of squares for the overall data set. To find out how much
is shared by countries, we compute the sum of squares between the mean proportion of each variable and
the ..08 mean of the overall data set. We subtract from this number the within and between clusters’ sum of
squares.
61
small part of the total variance. There are no distinct family regimes within Western countries, but rather

distinct cultural sensitivities built on a common understanding of how family and friendship work for

emotional support.

Cluster analysis identifies three cultural variants expressing the common understanding of the

resources provided by family members and friends in slightly distinct ways: the multiplex orientation, the

conjugal orientation and the children orientation. Since the cluster analysis shows the United States and

Switzerland belong to the multiplex orientation, this study confirms that the two countries from which the

various samples and case studies presented in this book come, share a similar approach to the family realm

within a large consensus among countries regarding the importance of family ties for emotional support.

In any case, the nuclear family as a cultural construct stressing the monopoly of the spouse for the

provision of emotional support, is associated with social conditions that have dramatically altered during the

last 30 years: permanent employment with generous benefits, stability of social and cultural norms,

economic growth, a social welfare state, and a conjugal tie meant to last. Even though this model has

changed, there is a great similarity in the ways in which the contribution of family members to social capital

is defined across national contexts, as the analyses performed indicate. In all countries, partners play a major

role in psychological support. Although the other family statuses taken separately only represent a modest

percentage of the total, they provide a major part of the help when considered together. In other words,

interpersonal ties beyond the nuclear family matter. Adults maintain or develop strong emotional

interdependencies towards their parents, in-laws, siblings and other blood or non-blood family members.

The isolation of the nuclear family is no more granted, its primacy is no more insured. There are many

alternatives to ties between husbands and wives in the realm of the family. Family based social capital

beyond the nuclear family in Western countries is therefore confirmed by a large and representative survey

such as the ISSP.

Does Social Capital Matter?

The configurational perspective states that any family dyad is to a large extent explained by the family

configuration in which it is embedded. Does this hypothesis hold true? In other words, does social capital

matter for key family relationships such as partnerships and parent-child relationships? System theory, from
62
Watzlawick to Minuchin (Broderick, 1993), has emphasized that subsystems constituting the nuclear family,

such as the conjugal, the parent-child and the sibling subsystems, are interdependent. The configurational

perspective goes a step further and asserts that they are parts of larger family contexts, from which they

draw resources while trying to protect their boundaries for the sake of their own functioning.

Family configurations indeed influence conjugal satisfaction in a variety of ways. In a series of scientific

articles, we addressed the impact of configurations of family members and friends on conjugal quality and

parenting, using a representative sample of 1,534 couples in which both partners were interviewed

(Kellerhals, Widmer & Levy, 2004; Widmer, Kellerhals & Levy ,2004; Widmer, Legoff, Hammer,

Kellerhals, Levy, 2006). Various questions about interdependencies with relatives and friends were asked to

each partner independently. Using a cluster analysis, six distinct types of configurations were identified,

mixing friends and relatives. Sparse networks refer to couples disconnected from relatives and friends, thus

representing a case of isolation. About one fifth of couples are friendship oriented rather than kinship

oriented. Laterality is another feature that has proven to be structurally and functionally significant. In about

four cases over 10, partners do not have equally present and supportive networks. In one couple over 10,

strong support is associated with attempts to control and interference by family members. Couples with

sparse networks are those that show the lowest conjugal quality; a lack of connections with larger

configurations of ties has indeed a negative impact on conjugal quality for both men and women. Only one

configuration is systematically associated with significantly improved conjugal quality. As a matter of fact,

configurations have a positive impact on conjugal quality only when they are strong for both partners, with a

large share of supportive blood ties, which do not interfere with the conjugal dyad.

This is also the case for the parent-child dyad40. As is true for conjugal dyads, most empirical research

on parent-child dyads has disregarded their integration in larger configurations of interdependencies. In

contrast, some scholars have underlined that support from relatives enhances parenting (Belsky, 1984;

Belsky, 1990; Chen & Kaplan, 2001; Cowan, Powell & Cowan, 1998). Various mechanisms explain the

impact of family configurations on parenting. Parents with more social support and fewer negative

interactions with family members provide more sensitive care to their children and have fewer conflicts with
40
The influence of configurations on parenting and child development is extensively covered in Widmer, Legoff,
Hammer, Kellerhals, Levy (2006).
63
them. Support from family members enhances the psychological well-being and the self-efficacy of parents.

For instance, a mother’s self confidence as a parent is bolstered by praise from a supportive network

member. The psychological resources of parents are of great importance to the parenting process. More

mature parents, with more robust psychological well-being, are more able to provide adequate stimulation to

their children. Family configurations and friends have an indirect effect on parent-child relationships and

parenting by their influence on the conjugal subsystem and the psychological well-being of parents. In

decreasing the likelihood for parents of experiencing psychological distress or conjugal problems and

conflicts, and in increasing the quality of their coping strategies, configurations constituted by active family

members and friends on both parents’ sides positively influence parenting and parent-child relationships by

strengthening the conjugal subsystem, which is the principal support subsystem for parents.

Both conjugal and parent-child relationships depend to a large extent on social capital provided by

family configurations to which they belong. This has implications for a large array of issues. For instance,

the concept of dyadic coping, which comes from psychological research on conjugal quality, makes

researchers focus on the internal interactions of couples as predictors of separation and divorce. The

importance of external factors potentially influencing conjugal coping has been acknowledged by family

psychologists. For instance, they point to the increasing risk of divorce stemming from stress originating

outside the conjugal relationship (Bodenmann et al., 2004). One can hence conceptualize coping strategies of

couples as embedded coping, e.g.,strategies connected to the family configuration (but also to friends and

possibly to other acquaintances) in which couples are embedded in their everyday lives, which provides them

with social capital. Some studies in clinical psychology underline that resolution skills are not the central

element on which therapists must intervene when dealing with couples in conflict. Couples with poor

problem-solving skills are as satisfied as couples with good problem-solving skills when positive affects

between partners are still strong, a result consistent with the hypothesis that the resolution of problems is less

critical than how partners define and understand the context in which their differences of opinions are

discussed41. Taking family configurations into account is helpful to understand why working on a couple’s

problem resolution skills and interpersonal communication abilities is not always effective at solving

41
See Widmer et al. (2009) for further developments and references.
64
problems over a long period of time, or why some couples with troubles have more difficulty fixing their

problems than others with apparently the same psychological and communication abilities but with quite

divergent family configurations.

Conclusion

The pluralization of family configurations created by the increase of divorce and alternative family

forms since the nineteen sixties has often been regarded as detrimental to the functions played by the Family

in the realm of social integration. Overall, the assessment of family decline made by individualization

theorists is strongly associated with their conceptualization of the Family as a small group with obvious

boundaries and much stability on the long run (Bengston, 2001). As the nuclear family has decreased in

number, the general public and some sociologists have eagerly assumed that the integrative function of

families was endangered, with various negative consequences for society at large, such as increasing

delinquency and school dropouts of children and adolescents, and raising life dissatisfaction and depression

of adults.

In opposition to this perspective, the configurational approach of families points at the individualized

resources that individuals develop in unbounded and heterogeneous family contexts. Even though the dyads

of the nuclear family play a key role in one’s life, they are embedded in a large set of family

interdependencies that provide them with social capital. Relationships with parents, siblings and their

partners, grandparents, uncles, aunts, cousins, but also friends or neighbours considered as family members,

constitute family configurations that have decisive integrative functions over one’s life trajectory. They have

a strong influence on the more central family dyads in their daily functioning. They may also play key roles

in delimited time periods of life, especially in life transitions such as the transition to partnership,

parenthood, divorce and widowhood.

the Family is a central configuration of late modernity. It provides resources and meaning to individuals,

as various national surveys taken in the United States and Switzerland show 42. There are, however, great

42
The International Social Surveys and the World Values Survey both reveal that the family satisfaction is considered to
be the most significant factor in achieving overall life satisfaction in most Western countries, including the US and
Switzerland.
65
differences in the structures of social capital provided by family configurations according to their

composition. Family configurations based on blood ties and intergenerational relationships provide a

bonding type of social capital, with much collective support and trust, but little individual autonomy. Quite

distinctly, family configurations based on friends considered as family members provide bridging social

capital by which family life opens up to a variety of ties, but with a more active individualized relational

work needed. Such family configurations make individuals integrate into a large set of realities while

keeping the emotional interdependencies to family members. This alternate form of social capital has

profound and not necessarily negative consequences. Individuals who develop bridging social capital have a

high level of structural autonomy and therefore more capabilities for action. Of course, this comes with a

cost, as being a bridger means keeping possibly contradicting interdependencies on both sides of the bridge.

This requires a great personal investment in time, energy and sociability. Therefore, individuals in that

situation might not be as involved in maintaining lasting strong ties with specific family members, as they

have more alternatives to consider, and less time and energy to invest in each of them individually. Overall,

the normative expectations associated with family life in late modernity has changed with the development

of bridging social capital in families; support imposed by dense family configurations is being replaced by

relational structures in which individualized relational resources come to play a prominent role. Social

capital in families is less and less defined by rights of birth or marriage obligations. In a social context where

conjugal dyads have become more fragile, family configurations have resources to propose, but at some cost.

The next chapter develops this point further by focusing on family conflicts and ambivalences.

66
 CHAPTER 4 – Family conflicts

Family conflicts were for a long time of little interest to sociologists, who considered that their main

responsibility was to address the contribution of the nuclear family to social integration. Family conflicts

were left to psychologists or social workers, as personality disorders, poverty or social exclusion were

supposed to explain their occurrence. That changed with the large increase of divorce in the nineteen sixties

and nineteen seventies, which provided serious evidence that indeed sociologists had to address family

conflict as an issue. One understanding of it stems from individualization theory, which emphasizes the

importance of massive cultural shifts towards the privatization of partnerships and the prominence of

individual self-realization over the maintenance of family ties on the long run 43. Family conflict is indeed

often interpreted as a sign of family decline. The configurational perspective takes a different stand and

analyses family conflicts as intertwined in commitments towards family members. Family interdependences

are costly. The focus on the contributions of families to social integration has somehow obstructed the

acknowledgement that interdependencies within family configurations create a large number of tensions and

ambivalences44. This chapter first goes back on the interconnections linking conflict and interdependencies

within families, stressing how easily family support becomes interference, a situation that triggers

ambivalences of various kinds. Using several datasets, collected in the United States and Switzerland, it

proceeds by showing that family conflicts frequently happen in strongly interdependent dyads or triads,

within specific family configurations.

Being Interdependent in Conflict

The previous chapter revealed that various emotional and material interdependencies link individuals

with a large number of family members in several distinct types of configurations. This, however, should be

somehow paid for. German sociologist Georg Simmel emphasized in the early years of the twentieth century

that competition and fights are social interactions in their own right (Simmel, 1955). Conflicts are not the
43
These trends certainly have some reality, but overemphasizing their explanatory power makes individuals believe that
families are doomed by late modernity because of its emphasis on self-development and individual autonomy, while
providing only few insights on how conflicts develop in empirical cases.
44
This emphasis on positive interactions is most obvious in work on intergenerational solidarity done in the nineteen
nineties (Bengtson & Harootyan, 1994).
67
opposite of solidarity, as they often develop because of positive interdependencies. They are rooted in many

family relationships, as they connect configuration members and make them interact intensely. During

conflicts, communications are often at a high level; in many cases, family members do not develop conflicts

because they want their relationships to end. They rather want them to be maintained but changed (Widmer,

1999a). If one has a conflict with a parent or a sibling, it is because the set of interdependencies that link the

two individuals is valued: something needs to be changed in order for the interdependency to remain active.

New interdependencies are often sought by family members through conflict. For instance, the higher

dissatisfaction of women in marriage is in part due to their higher investment in family life. A major

modality of conflict is one in which women are overly concerned with family issues such as money, leisure

time or division of household tasks, and want things to improve in order to maintain the conjugal bond

(Widmer et al., 2003). Partners frequently go through conflict to achieve a new organization of

interdependencies between them.

In some other cases, the interdependency, while displeasing, cannot be disregarded: two persons living

in the same household fight because they cannot avoid being in contact with each other 45. Third parties such

as parents and partners, because of the social pressure towards transitivity, create interdependencies between

other family members who do do not like each other in the first place. The other person has to be dealt with,

for instance because of a shared residence, a financial or emotional dependency. Many couples, parents and

children develop regular conflicts without willing to put an end to their relationships. This is especially true

in relationships that cannot be easily ended, such as those between parents and children, siblings in

childhood and adolescence, or couples with young children. Conflict means in those cases that the

interdependency between the two individuals cannot be avoided. Feelings of hate, jealousy and bother

towards a person who is an intimate relate to the feelings of closeness and dependency that one feels

towards this person. Violence between partners is an example of such interdependency. Men who become

violent with women relate to various forms of emotional dependencies. Violent relationships keep on

because individuals depend on each other on various grounds, emotional, financial, etc. In still other cases,

positive interdependencies trigger negative feelings. Providing regular support to elderly parents is linked

with ambivalence (Lüscher, 2002, 2004, 2009). Supportive relationships create tensions and conflicts that
45
Indeed, as Moreno (1934) and Feld (1981) stress, focus points make people interdependent. See chapter 7 for further
developments.
68
could be avoided by decreasing the positive exchanges linking family members, and the negative value of the

relationship is directly connected with its positive value. It is impossible to understand why in that case

children develop frustrations concerning their ageing parents without taking their supportive behaviours into

account. In all these cases, there is a unity between the negative and the positive faces of relationships.

Overall, family conflicts are not the opposite of positive ties as they are embedded in the set of the

interdependencies linking individuals. For those reasons, conflict is not, in many cases, a consequence of

poor family interdependencies but a part of the mix of interdependencies that define each family

configuration in a unique manner.

Support and Interference

The larger family context in which conjugal or parent-child conflicts develop plays a central role in

conflicts. Emotional and instrumental resources are scarce and their distributions within family

configurations are related with power issues. Consider conjugal relationships once more. The previous

chapter ended with the conclusion that bonding social capital constrains individuals in a structure of

interdependencies that leaves little space for autonomous projects. The logic of support itself may indeed be

at times disruptive. As a matter of fact, bonding social capital does not only have positive consequences.

Some scholars have stressed the burdens associated with transitive configurations. Amoral familism, family

interference and family overcare stress that a bonding social capital within families have detrimental effects

that a bridging social capital escape from .

As a matter of facts, empirical research points at a curvilinear effect by support of family members and

friends on partnerships (See for example, Johnson & Milardo, 1984; Julien et al., 1994; Holman, 1981). One

major problem of being embedded in a dense family configuration is interference, that is, some configuration

members intrude into conjugal relationships, for instance by giving advices on matters that concern the

couple first. When family members are too involved in providing support, they become interfering, and as

such, they may endanger conjugal or parent-child dyads. Intervention of third parties in an existing conjugal

conflict reinforces partners’ self-legitimacy, thus making a consensual solution less likely. Partners report

greater marital conflict and ambivalence in conjugal relationships when wives frequently interact with

friends rather than with relatives. Therefore, the amount of support received by couples is not linearly

associated with their satisfaction. Support provided in configurations has quite different consequences for
69
conjugal dyads according to its perception or not as an interference. The effect of bonding social capital on

parent-child relationships is also curvilinear, as it does not facilitate parent-child relationships and parenting

when the expectations of parents and other configuration members about the child are inconsistent, or when

family members are perceived by parents as competitors rather than as supporters in the parenting process

(Widmer, Legoff, Hammer, Kellerhals, Levy, 2006).

The presence of strong interdependencies with configuration members, characterized at the same time

by interference attempts and various types of support, is associated with an increased likelihood of couples

to end with a divorce (Widmer, Kellerhals & Levy, 2004). As such, conjugal dissatisfaction and divorce are

no more considered as the failure of a self-sufficient couple, but as one outcomes from a series of processes

that include not only the partners, but also their parents, their siblings, their friends and their colleagues.

Interference of family members and friends modifies the balance of interdependencies within couples.

Family support often leads to conflict as it makes individual autonomy and the primacy of couples over other

family relationships decrease.

Ambivalences
In the late nineteen nineties, in order to investigate such widespread cases, ambivalence was proposed as

a promising concept to address the interplay between solidarity and conflict in intergenerational

relationships, between adult children and their parents 46. Ambivalence is defined as a situation in which a

lasting co-occurrence of positive and negative dimensions coexist within a relationship (Lüscher, 2002 and

2004). Adults and their ageing parents are torn in many cases between the opposite norms of providing or

receiving support, and maintaining their autonomy. In some cases, adult children wish to support their

parents more but they cannot because of other commitments. Lüscher and Pillemer (1998) report several

cases in which support given for a long period of time to an older adult endangers the conjugal relationship,

and the relationship that the care giver develops with her own children, because of the difficulty of investing

in all relationships at the same time. Likewise, according to Lüscher and Pillemer (1998), ambivalence in

mother-daughter relationships is increased in adulthood when daughters have children. Such ambivalences

are explained by the fact that investments done in the care of their mother by individuals lowers their

46
Kurt Lüscher and his collaborators are among the first to propose, by the end of the nineteen nineties to apply the
concept of ambivalence to intergenerational family relationships (Lüscher, 2002, 2004).
70
investment in their partnership or their children. Individuals intertwined in such contradictions, develop

ambivalent feelings toward family members. Indeed, there is a great number of cases in which care for an

older parent by a child, or, to the contrary, financial support from a parent to an adult child, are loaded with

ambivalence. Rather than being « either love or hate» types of relationships, family relationships are in many

cases at the same time loaded with negative and positive feelings. Consider sibling relationships: because

siblings in childhood and adolescence are compelled to live together, they develop a large number of

conflicts (Widmer, 1999a). Indeed, relationships embedded in a strong structural or normative pressure to be

persistent, trigger negative feelings while maintaining a high level of positive interdependencies. Not only

intergenerational and sibling relationships but also conjugal relationships, among other family relationships,

are loaded with ambivalence, as a variety of results on conflict management by couples shows (Widmer et

al., 2003). Ambivalences, the configuration perspective sustains, are embedded in family interdependencies.

Indeed, family resources in love, money, time and emotional support are scarce and claimed by large

numbers of individuals who compete for their allocation.

Closeness and Conflicts

Consider again the case of Betty, this 54-year-old female married twice, with both marriages having

ended in divorce. Chapter 3 showed that interdependencies in Betty‘s family configuration follow a complex

pattern with many supportive relationships. Table 10 reports the percentage of support, conflict and

ambivalent relationships in the 210 dyads that constitute the family configuration, according to the closeness

that individuals feel towards others47.

INSERT TABLE 10 ABOUT HERE

Table 10. Support, conflict and ambivalence according to the level of closeness

Conflicts are much more frequent among family members of Betty who consider each other very

close or close than between family members who are only somewhat close or acquaintances. Seldom does

conflict appear between individuals who are not closely related. Obviously, individuals who barely know or

do not know each other at all are not in conflict. This case study points at the fact that family configurations

include a mix of conflicts and supportive relationships, especially when family members are close. Just
47
Ambivalence is measured as the simultaneous presence of conflict and emotional support in any dyad.
71
compare the supportive ties in Betty's family, as reported by each family member in Figure 6, with conflict

ties in Figure 7.

INSERT PICTURE 4.1 HERE

Figure 6. Supportive ties in Betty's family configuration

Family members of Betty are densely connected both in terms of conflict and support. One notices a

symmetry in patterns of conflict and support. Betty, her mother and siblings, constitute a subgroup of

individuals highly interconnected by support and conflict. Indeed, there are persistent fights among siblings

concerning the care to be provided to their mother. The mother of Betty continues to play an active role in

the configuration, as she wishes to be in charge of the life of her grown up children. The contradictions in the

parent-child dyads time after time spillover conjugal relationships, in which misunderstandings occur about

what should be invested in the partnership and in intergenerational relationships.

INSERT PICTURE 4.2 HERE

Figure 7. Conflict ties in Betty's family configuration

This is especially the case of the dyads linking some of Betty's siblings and their partners. A high

level of both conflict and support characterizes those dyads 48. In other words, a large number of ties are

simultaneously tense and supportive. Interestingly, these ties are located in the centre of the configuration,

between the siblings of Betty and their mother, rather than on its periphery.

There is, however, another type of ambivalence in this family configuration, stemming from a

contradiction between several relationships. Interestingly, Betty does not provide emotional support to

neither her son, her son's father, nor to her former mother-in-law, but she does so with her daughter. This

daughter has developed such supportive ties with these three family members, and has only developed

serious conflict with her brother. Thus, there is an imbalance in many triads in which Betty is embeded, as

individuals connected positively to her, are linked directly by tense relationships. This ambivalence is

structural rather than dyadic, as it stems from contradictions within a structure of interdependencies among at

least three actors. In that case, feelings towards each family member are unidimensional as they are either

positive or negative. The balance of interdependencies is however unset in triads that do not comply with the

48
This explains why there is only a moderate but still positive correlation between conflict and support. Indeed, the
Pearson correlation between conflict ties and supportive ties is .18 in that case. It is rather weak but still statistically
significant below the .05 level.
72
transitivity principle. This situation is therefore distinct from dyadic ambivalence, in which tensions coexist

within the same relationship, rather than between relationships. Because dyadic and structural ambivalences

neither overlap nor are disjunct, conflict and support are only moderately correlated. Following Lüscher

(2004, 2009), four cases can be theoretically derived from the conjunction of conflict and support, as

illustrated by Table 11.

INSERT TABLE 4.2 ABOUT HERE

Table 11. Typology of support and conflict in family configurations

Each combination of conflict and support draws a distinct configuration of ties. Configurations

characterized by neither conflict nor solidarity are not active. One may refer to this situation as atomization

(Lüscher, 2004). No strong long lasting ties, and no orientational others are available; their absence is not

compensated by weak ties and a larger autonomy. This situation rises serious concerns regarding the family's

survival chances and the social capital that it provides to its members. In the case of atomization, individuals

are not strongly interdependent, neither in positive nor in negative terms, neither strongly, nor weakly.

Family configurations characterized by a high density of conflict and no or only few supportive ties are

featured by cases of captivation in which individuals are compelled to stay in regular contact, without any

interest and active taste for being together. Why staying together if only conflict arises and emotional

closeness is gone? The lack of resources or the scarcity of alternatives explain captivation, such as when

partners cannot afford a divorce because of housing or financial problems, or when emotional

interdependencies with children make such an issue unthinkable. In families with problems of integration to

the larger society, unemployment, health problems or drug use, there is a spillover effect from the external

problems on the internal interactions that often leads to captivation (Widmer et al., 2003). Another origin of

captivation is generated by the normative impediments constraining family members to regularly interact.

Children are impelled to live with their parents during childhood and adolescence. This involves a series of

obligations that come into contradiction with norms of personal autonomy and self-development, which are

so strongly valued in other areas of social life in modernity. Adolescent sibling are especially marked by

such contradictions, as they have not chosen each other (Widmer, 1999a). It is noticeable that when

adolescence ends, the frequency of interactions among them significantly decreases, making family conflicts

73
also less likely. Family configurations of older adults undermined by age problems also correspond to such a

case, as well as individuals with psychiatric problems 49. In all those cases, individuals have no other

alternative than relating either to their family members or to institutions. When care becomes over-care and

is forced by circumstances, conflicts and ambivalence are likely to develop.

Individuals with supportive ties and no conflict belong to family configurations in which there is an

emphasis on the group solidarity rather than on individuals. There is little space for a diversity of interests or

contradictions in such families. In the large study on couples (Widmer et al., 2003), two types of conjugal

interdependencies focus on fusional values, with an emphasis on the group rather than on the individual.

Interestingly, the low level of conflict in such couples was associated with a low level of conflict in parent-

child dyads and in sibling dyads. In that case, the high density of emotional support stemmed from

interdependencies among family members that were perceived as beneficial by those involved.

Ambivalence is characterized by a high density of both conflict and support in the same family

configuration. Individuals develop conflicts with family members on whom they are emotionally dependent.

Family support is loaded with ambivalence as it implies a decrease of individual autonomy and a higher

functional dependency of partners and parents towards their relatives. Ambivalence within couples fires back

on relationships with members of the family configurations, as conjugal dyads play a central role in them and

are, as such, a focus point for many family members. Children, parents, siblings and friends receive

contradicting messages from ambivalent couples split between various interdependencies. This is associated

with the systemic nature of conflict in family configurations: a high density of supportive ties frequently

leads to the development of interference and conflict of interests.

Conflict and support should not be conceptualized as opposite dimensions in family configurations.

They can appear together or each by itself. Conflicts are at time the consequences of dense and transitive

networks of interdependencies. In other cases, however, they are the only surviving ties after positive

interactions have been shattered. The meaning of conflicts to a large extent depends on the overall patterns of

supportive interdependencies characterizing each family configuration.

49
See chapter 6.
74
Intergenerational Conflicts

In what family configurations do ambivalence and conflict develop most? Recall from Chapter 3 that

family configurations provide distinct types of social capital according to their composition. Beanpole family

configurations, centred on several generations of blood relatives, provide a bonding type of social capital, a

situation in which all individuals in the configuration are interconnected. This situation creates a high level

of collective solidarity as well as a high level of control, as individuals join together in order to help or

impose a normative framing to any family member. As a matter of fact, this type of social capital creates

tensions and conflicts, as it makes individuals and couples face the interference of family members. Table 12

shows that beanpole configurations are associated with a higher level of ambivalence than other family

configurations50. In other words, the density of support in beanpole family configurations triggers a high

density of conflict. A great number of relationships in such configurations are both negative and positive,

mixing conflict and support. Therefore, one may call such configurations ambivalent as many of their dyads

are torn apart by conflicting forces. Figure 8 considers a typical case of a beanpole family configuration.

Ambivalent ties are represented by dashed lines and supportive ties are represented by light solid lines. There

is no tie characterized by conflict only in this family configuration.

INSERT PICTURE 4.3 HERE

Figure 8. Conflict, support and ambivalence in a beanpole family configuration

Full arcs: only support provided. Dashed arcs: support and conflict

In this beanpole family configuration, all conflict ties are also support ties. In other words, conflicts are

interrelated with solidarity practices. Members of the configuration at the same time upset each other and

support each other. Intergenerational relationships are commonly represented in those ambivalent ties: the

mother, the father and the grandmother of the respondent all have developed ambivalent relationships with

one or several of their children. Other ambivalent relationships however also exist among conjugal partners:

the respondent's sister and her partner; the couple of grandparents, the maternal uncle and his partner, the

50
The average density of ambivalence is computed from the index proposed by Willson et al.
(2003) for dyadic relationships: the absolute value of the difference between density of support and
density of conflict, added to the sum of the density of support and of conflict.
75
paternal uncle and his partner, the mother and the father. In this highly interconnected configuration,

intergenerational ambivalence comes with conjugal ambivalence. As we saw previously, these two

phenomena are interdependent. A high level of interference in one relationship is associated with an

increased level of conflict in the other relationship. Ambivalent relationships are graphically at the centre of

the family configuration, as they constitute its backbone.

Far from being marginal phenomena, ambivalences play an active role in this family configuration. One

explanation is that relationships in such family contexts involve a large share of obligations. Recall that, due

to the density of interdependencies, a bonding type of social capital is present in beanpole family

configurations. Individuals are, in other words, bound to others in transitive triads: they cannot escape from

dyadic relationships, as family members are linked with each other by third parties. This large number of

third parties means that individuals cannot change or end a dyadic relationship without it to be of concern for

other family members. The social control implied by such a situation stems from multilateral

communication, which is much reinforced. In other words, the respondent cannot ask for a break in her

relation with her mother without a reaction from her aunts, uncles and cousins. A conflict with one's partner

will also be commented upon by one's parents and parents in-laws. This particular organization of ties is

likely to increase the overall level of ambivalence. Individuals are forced to maintain a high level of

interdependencies with others due to the social control embedded in the overall structure of relationships

within the configuration. Therefore, conflicts cannot break the family configuration into pieces and are

transformed into ambivalent relationships. Table 12 reports the average density of support, conflict and

ambivalence according to family configurations of young adults.

INSERT TABLE 4.3 ABOUT HERE

Table 12. Support, conflict and ambivalence according to types of family configurations
** significance below a one percent risk of error; * significance below a five percent risk of error

In terms of support, the beanpole family configurations are above average, as their density of supportive
interactions is higher. Other family configurations all are at the same level, with a smaller density of
supportive interactions in friendship and father-oriented family configurations. In terms of conflict, the
results are similar. Beanpole family configurations have a higher density of conflicts compared with other
configurations. Post-divorce configurations are average; the other configurations have a much lower density
of conflict. Thus, beanpole family configurations have the highest rate of ambivalence, while the friendship

76
and father oriented family configurations have the lowest rates of ambivalence. Family configurations built
on intergenerational ties therefore produce at the same time a high level of supportive and conflict
relationships, which makes them highly ambivalent.

Conclusion

The link between the larger family context and conflict developed in partnerships or parent-child

relationships is often concealed by standard survey methods because of the cultural desirability of stressing a

high level of autonomy in one's life 51. When alternative methods, such as network analysis, are used, one

realizes that individuals in adulthood remain emotionally dependent to a large set of family members. This

dependency is expressed in conflict as well as in support. A large share of conjugal or parent-child conflict

are motivated by the overall issue of how to share resources of time, love, money and care in family

configurations. Individuals get into fights in families because they depend on each other: Resources are

scarce and their distribution within family configurations creates conflicts in key family dyads such as the

conjugal dyad and the parent-child dyad, which in many occasions influence the overall configuration in

endless retroactive loops.

Far from being a private matter insulated in specific dyads without dependencies on the larger family

contexts, conflicts and ambivalences developed in dyads relate to the overall organization of

interdependencies within family configurations. Dense family configurations make the occurrence of

conflicts in key family dyads more likely. The impact of emotional closeness on conflict stresses the

importance of understanding conflict in families as consequences of solidarity and commitments. Indeed, one

fights with people one is interdependent with. Conflicts develop with people who matter. Acquaintances and

distant family members are not the focus points of enough interdependencies for providing significant

opportunities to conflicts. The concept of ambivalence permits a deeper understanding of the roots of family

conflict by emphasizing the contradictions that many individuals develop while being interdependent with

family members. Family ties remain a primary source of emotional support in late modernity. Therefore,

they also are a primary source of conflicts. Interdependencies between adults and their parents create many

situations in which individuals are torn apart by attempts to fulfil contradicting reciprocity and transitivity
51
In the survey based on 1534 couples living in Switzerland (Widmer et al., 2003), we found that only very few
individuals, when asked directly, report a high level of dependency on others. Indirect questions such as the ones
proposed by the Family Network Method are more appropriate to measure interdependencies in a cultural context in
which the social desirability of being autonomous is so widespread.
77
with various family members. Receiving or providing support has deep meanings for individuals' identity, as

it modifies their position in the family configuration and their understanding of their own role in the family.

Conflicts and ambivalences in parent-child dyads are likely to spread within each family configuration, as

they involve reallocations of resources that are not easy to make. Rather than stemming from a thrust for

personal development and realization of self-centred individuals, family conflicts develop in the complex

webs of interdependencies that characterize intimate relationships in late modernity. In that sense, family

conflicts should not be regarded as signs that the Family is on the verge of collapse, but rather as an

expression of the vitality of family interdependencies.

78
 CHAPTER 5 - Post-divorce Families

“Blended” families participate to the fears and interrogations regarding the contribution of the

Family to society in late modernity. Research using the nuclear family as a stick yard underlines a

lack of social integration stemming from divorce and remarriage 52, which are held responsible for

problems in child development and parenting. Scholars have explained the problems of children in

stepfamilies as the results of a reduced involvement of parents and stepparents in their life. One may

however question the generality of this deficit in the light of the variability of family configurations

revealed by the previous chapters.

There are many ways in which one may recompose a family after splitting with a partner.

Individuals who go through divorce and remarriage may at the same time present distinct family

configurations from those who do not, and a diversity of family models. Indeed, the pluralization of

families does not stop with divorce. This chapter focuses on the issue of family diversity following

divorce, using two distinct and complementary datasets, one on young adults who refer to the divorce

and remarriage of their parents, the other on older adults with children. We shall first reconsider

existing results on social support following divorce and remarriage in the light of the configurational

perspective. We proceed by showing that the resources provided by families depend to a large extent

on the way in which family configurations are reshaped. Family configurations following divorce and

remarriage greatly vary, as several factors other than divorce influence the ways in which support and

conflict intermingle. The chapter ends with an empirical comparative study on the cultural meanings

of stepfamily statuses, which highlights the cultural framing of interpersonal relationships in the

United States and Switzerland.

52
For the sake of simplicity, divorce and remarriage are used in this chapter to refer to the termination of a
conjugal relationship and partnering after this termination while having children from a previous partnership.
Not all individuals actually married and then divorced by legal standards.
79
Social Support, Divorce and Remarriage
Research on parenting following remarriage has produced rather disturbing results. The

investment in new partners by parents is associated with a decreased attention towards their children.

Likewise, stepparents invest less in stepchildren because of their focus on their new partnership or on

their children from a previous marriage (Coleman, Ganong & Fine, 2000). Role definition is also

critical, as the stepparent's role is ambiguous and incompletely institutionalized (Cherlin, 1978).

Relationships with non-custodian parents, in most cases fathers, have been described as significantly

less close, less trusting and more conflictual in stepfamilies compared with relationships with fathers

in first families (King, 2002). Remarriage influences the amount and the type of support provided by

previous partners. When mothers remarry, fathers reduce their financial support. When fathers form

new relationships, they reduce support to their children from prior relationships.

Changes do not only concern the parent-child dyad but the larger family configuration as a whole.

Divorced parents receive lower levels of support from grandparents (Umberson, 1992). Support that

mothers receive for their children from their ex-partners' relatives declines after divorce and

remarriage. In addition, mothers receive less support from their own kin when they remarry and gain

stepchildren. Adults report considerably weaker obligations to stepgrandchildren than to grand-

children (Coleman, Ganong & Cable, 1997). While remarriage expands kinship networks, it also

makes them more diffuse and more ambiguous in their lines of responsibility for providing support

(Harknett and Knab, 1997).

Research also stresses that various factors make the experience of family recomposition

heterogeneous. For instance, the cooperation between the biological parent and the stepparent in

raising the child, as well as the level of the stepparent's direct involvement in education, are

significantly associated with child outcomes in various domains. Relationships between the custodian

parent and the non-custodian parent and the relationships between children and their non custodian

parent also have a strong influence. But at the same time, the impact of the stepfather may be more

ambiguous if the child’s biological father maintains regular contacts (Bray, 1999; Buchanan et al.,

80
1996). Finally, whereas the quality of marital relationship is the main factor of family adjustment in

nuclear families, the picture is more complex in stepfamilies as a good relationship between the step-

child and the step-parent becomes a powerful predictor of family harmony and duration of the new

marriage (Papernow, 1996). Stepparents enact affinity-seeking and affinity-maintaining strategies to

build a relationship with their stepchildren (Ganong, Coleman, Fine & Martin, 1999), according to a

specific dynamic: a successful bond is created when stepparents develop a friendly relation with their

stepchildren, and take a “quasi-parental” position in a second stage only.

Social Capital after Divorce

Overall, the available evidence point to a series of relational factors that make post-divorce

family configurations less able to provide support a guidance to their members, especially their

children. Research also points at moderating variables, which make the experience of post-divorce

families heterogeneous. Does that mean that less social capital is available to children and adults in a

majority of post-divorce family configurations or rather that a diversity of situations exist? Recall that

bonding social capital is not distributed equally among family configurations: Beanpole family

configurations provide more bonding social capital than friendship family configurations, which are

oriented towards a bridging type of social capital. Such alternative to bonding social capital are likely

to develop in family configurations of individuals who have experienced divorce and remarriage either

directly or indirectly, because of their parents. Scholars have underlined the more voluntary nature of

relationships in stepfamilies, which are less institutionalized than first-time families (Carsten, 2004;

Castren, 2008; Cherlin, 1978; Cherlin & Furstenberg, 1994). Bridging social capital, which is

associated with a position of intermediary, has indeed various positive consequences such as

enhancing individual autonomy and one’s sense of responsibility 53.

The results of the study on young adults (Widmer, 2006) confirm that having parents who

experienced divorce and remarriage is associated with a peculiar kind of social capital. Recall from

53
See chapter four.
81
chapter two that one type of family configurations was described as post-divorce. Respondents in this

family configuration include a large number of step-relatives: 1.38 of their family members are

steprelatives, compared with .21 in the whole sample. Stepmother and stepfather are over-represented,

with 62 percent of respondents including a stepfather and 41 percent a stepmother. Stepparents are

also included first in post-divorce family configurations, especially stepfathers, who come before

stepmothers. Fathers are much less frequently included than in any other family configuration: only

half of respondents in post-divorce families include them and they come almost one rank after fathers

in other configurations. It is also the case of siblings, who appear less often than in other

configurations. As for more remote kin, fathers’ relatives are under-represented, including paternal

grandparents. Overall, there is a deficit of blood relationships and an over-representation of alliance

relationships in the family configurations of young adults who have faced the divorce of their parents.

INSERT PICTURE 5.1 HERE

Figure 9. A post-divorce family configuration

As a consequence, their centrality in their family configurations is higher, although their direct

interdependencies are more limited in number. As exemplified in Figure 9, there are many holes in

post-divorce configurations, in which children develop unique ties with other family members. When

respondents are removed, post-divorce configurations split into a large number of disconnected

components: the configuration is separated between the father and his new partner, her father's

daughter (half-sister) on the one hand, and the mother, the mother's new son (half-brother), her new

partner and her own mother, on the other hand. On the father's side as well as on the mother's side,

individuals are often disconnected. For instance, the father's daughter has no connection with her aunt,

whereas on the mother's side, the mother's partner has no connection with the mother's mother.

Interestingly, the respondent has developed no direct interdependencies with both of her stepparents.

Therefore, she has got an intermediary position in her family. Because former partners often do not

keep close connections, a partition is also created in their children’s family configuration. A
82
widespread feature within post-divorce family configurations is a lack of transitivity in the triad child–

biological parent–stepparent (Widmer, 1999b). In Figure 9, this is expressed in two ways: first, the

mother’s partner and the father’s partner are disconnected from the respondent. Second, there are holes

between the parents’ new partners and their children, and some of the parents’ blood relatives. This

makes the overall family configuration highly sensitive to the removal of various family members,

among whom the respondent and her biological parents. As a result, the number of family members

directly connected with the respondent by supportive interdependencies is small, and the family

members supporting her are not connected with each other.

Overall, interdependencies between stepparents and stepchildren, and between half-siblings or

stepsiblings, cannot be taken for granted. In this respect, post-divorce family configurations do not

resemble the star-like pattern of the friendship family configuration, but come close to individualized

chains of interdependencies. This is likely to be a peculiar situation for children, as they have an

intermediary position between family members who also have an intermediary position between other

family members. In other words, interdependencies with family members do not provide access to

well-bounded and dense subgroups, but to a continuum of individualized interdependencies. In post-

divorce families, children are indeed part of divorce chains or remarriage chains (Bohannan, 1970), in

which interdependencies link a large set of persons living in different households.

What happens to family configurations when mothers facing divorce and remarriage rather than

children are interviewed? Because families have become individualized unbounded networks of

interdependencies, divorce and remarriage have distinct consequences for the family configuration of

parents and children. In other words, a divorcing mother and her son are embedded in distinct sets of

interdependencies. Therefore, studying families requires to consider the issue of stepfamilies from the

perspective of several family members. This is of course not empirically possible at all times as

multiplying linked interviews of family members is extremely costly. At the very least, one should not

take the validity of results found on one set of family members (for example, children) for granted for

83
another category of family members (for instance, mothers) 54. Comparing the profile of post-divorce

family configurations of mothers with the average profile for social capital, we found that they have

on average a lower density of relationships and that their centralization is lower as well. Neither

bridging nor bonding social capital is especially strong in those families. Therefore, post-divorce

families in older adulthood as in earlier adulthood withdraw from bonding social capital without

having a strong grisp on bridging social capital.

The results on both samples of young and older adults explain why there is a much more active

work required of “doing kinship” in order to create and maintain interdependencies in post-divorce

family configurations than in other family configurations (Schneider, 1980; Cherlin & Furstenberg,

1994). To summarize, young adults in many post-divorce families have a small number of

interdependencies embedded in long chains of connections that are sensitive to truncation. Therefore,

bonding social capital is not available in comparable quantities as in other family configurations.

Individuals in post-divorce family configurations are embedded in networks characterized by holes

and weak ties, without the centrality experienced by adults embedded in friendship or sibling family

configurations55. Because of the individualization of interdependencies, children raised in such

contexts might be more able to develop autonomous behaviours early on. This is not only a possibility

but a necessity for them, as the structural organization of such families requires from their members to

work their way through contradicting interdependencies

Conflict and Ambivalence in Post-divorce Families

Support and social capital are not the only dimensions of interdependencies different in families

stemming from divorce and remarriage; this is also the case of conflicts and ambivalences. Post-

54
In that sense, talking about families as if they exist beyond individuals as independent groups is similar to
splitting individuals and society. Whereas society does not exist without individuals and their interdependencies
(Elias, 1991), families do not have an existence beyond the interdependencies linking individuals.
55
As we saw in chapter 3, bridging social capital, which is associated with a position of intermediary, has various
consequences such as enhancing individual autonomy and one’s sense of responsibility, connecting individuals
to a variety of social experiences, multiplying channels of information and communication, and increasing the
heterogeneity of one’s social world.
84
divorce family configurations develop structural ambivalence 56, stemming from contradictions in

triads rather than in dyads. One may develop supportive relationships with family members while still

being in ambivalent triads if the interdependencies existing between two dependent family members

contradict each other. This imbalance seldom occurs in social networks, as relationships are usually

organized transitively (Davis, 1979; Davis & Leinhardt, 1972; De Sotto & Albrecht, 1968; Killworth

& Bernard, 1976; Kumbassar, Romney & Batchelder, 1994). But post-divorce family configurations

have to deal with structural contradictions that make such case more likely than in other family

configurations. We stressed above, following the American sociologists Andrew Cherlin and Frank

Furstenberg (1994), that a large proportion of children do not acknowledge their co-resident stepfather

as a family member. This may be due to indifference in some cases, to conflict in others. Because

relationships among stepparents and stepchildren are less intimate, less supportive, and are associated

with more conflicts than relationships between parents and children (Coleman & Ganong, 1990;

Pruett, Calsyn & Jensen, 1993), they are likely to create structural ambivalence in the family

configuration. Consider the case of this post-divorce family configuration (Figure 10).

INSERT PICTURE 5.2 HERE

Figure 10. Conflict, support and ambivalence in a post-divorce family configuration

Plain arcs: only support provided. Bold arcs: conflict

In Figure 10, supportive dyads are presented by light solid lines and conflict dyads are represented by

strong solid lines. Contrary to what happened in the beanpole family configuration presented in the

previous chapter, all dyads are either positive or negative. As an example, there is a supportive

relationship between the respondent and her mother but a conflict relationship between the respondent

and the new partner of her mother. In this case study, conflict relationships are not generated by

emotional support. They are rather the expression of a lack of support between two members of the

family configuration. Contrary to the expectation of balance theory (Heider, 1958), triads in that case

are structurally ambivalent, as they link individuals in an intransitive way: The respondent feels

56
See chapter 4.
85
supported by her mother who is emotionally dependent on her new partner but the respondent is in

conflict with the new partner. There is a contradiction in this triad explained by the fact that all three

individuals experience a structural ambivalence in their personal relationships: the mother has to deal

with her commitment to her daughter and her new relationship with her partner. The partner loves a

woman whose child despises him. The daughter is still faithful to her father and therefore cannot get

closer to the new partner of her mother despite her love for her mother. Even though such patterns are

usually avoided by individuals because of their discomfort (Freeman, 1992; Heider, 1958; Newcomb,

1961), the complexity of family interdependencies in late modernity makes them likely in some

instances57.

Ambivalence in Older Adulthood

So far, we have considered young adults. Because of their life stage, those who belong to post-divorce

family configurations are children with divorced parents and have not divorced themselves. We saw

that divorce of their parents put them in an intermediary position for both conflict and support. This is

not necessarily the case of parents who may have more leverage than children to reshape their set of

interdependencies after divorce. What happens when mothers who experienced divorce and remarriage

rather than children are considered? Let us consider a few telling cases. Alina is a 35 years old woman,

holding an unskilled cleric office and living in a small rural town. She has a son, Theo, 10 years old,

from her previous marriage which lasted two years. Since then, after some years spent as a single, she

met Laurent with whom she currently lives. Laurent has two sons, one below 10 and the other, a young

adult who does not live with them. Alina has one sister; her own parents are divorced and her father is

remarried. Of her grandparents, only her grandfather on the paternal side is still alive. Her family

configuration is a mix between her current partnership and her previous partner, the father of her son.

She first includes her son in her family configuration, followed by her current partner, his first son, her

previous partner, her mother, the second son of her current partner, the mother of her previous partner,

57
Paradoxically, the level of dyadic ambivalence is lower in post-divorce family configurations than in beanpole
family configurations.
86
her father, her sister and three friends. The composition of her family configuration is therefore very

heterogeneous. The order of inclusion of family members is mixed, with the ex-partner being included

within the first ranks, which tells his importance in the family configuration.

INSERT PICTURE 5.3 HERE

Figure 11. Conflict and support in two post-divorce family configurations (Alina and Dora)

The graph of support in Alina's family configuration in Figure 11 provides interesting

information. First, it is dense, which indicates a bonding social capital. Many individuals are

interconnected, which shows that information and support have a collective nature in this family. The

likelihood of collaborative work in crucial occasions, such as when the respondent requires some help,

is increased. This density however raises serious issues as unusual interdependencies are exhibited in

the graph: for instance, Alina keeps on being an important source of emotional support for her

previous partner. She also keeps emotional interdependencies with her previous mother-in-law.

Overall, the high connectivity of her family configuration makes it hard to insulate conflict within

specific dyads. The ex-partner has a direct access to the new couple's business as he is emotionally

connected to Alina. He also has indirect sources of information and of leverage by his son and his own

mother. The current partner's children therefore directly depend on what happens between Alina, her

ex-partner and her ex-partner's mother.

The peculiarities of this family configuration are confirmed when one takes a close look at the

graph for conflict. Indeed, its density is high, showing that ambivalence rules over this family. The

lines of conflict follow a rather unexpected path. The current partner is central, as he is involved in

conflict with Alina, his sons, and Alina's child. What is still more surprising, the negative

interdependencies with the ex-partner and the ex-partner's mother in several triads involve Alina's son.

Likewise, Alina and her partner are included in conflicts with her father and her mother. Overall,

87
rather than being circumscribed in a limited number of dyads or subgroups, conflict relationships join

several dyads that are usually kept separated.

Negative interdependencies connect individuals in intransitive triads, which makes the whole

family configuration unbalanced. As a matter of fact, the current partner may see with quite some

concern the ex-partner still being emotionally close to Alina. He may as well wonder about the

connection existing between Alina and her previous mother-in-law, as, by contrast, Alina does not

acknowledge his own parents as family members. In addition, the ex-partner may have serious

concerns about his son being raised and in daily contact with a man that he does not like and with

whom he is in conflict. Interestingly, conflicts do not overlap with support. In other words, in this

configuration, people in conflict are most of the time not supportive. The relationships are

unidimensional, as they either focus on support or conflict. This structural imbalance is likely to create

a large number of tensions, as conflict is not weakened by supportive behaviours.

This intricate network of conflicts rising from high emotional interdependencies crossing the

boundaries of various triads deteriorates relationships within key dyads such as the partnership

between Alina and her current partner, and the various parent-child dyads (Alina, her partner, her

previous partner and all their children). As a matter of fact, Alina reports a variety of serious conflicts

occuring between her and her partner's child, or between her partner and her child. Those conflict are

related, in her own words, with abusive parental behaviour of her current partner and with a lack of

recognition of her parental status by his children. When asked about who in the family configuration

increases the problems associated with those relationships, she cites her current partner and her

mother. The relationship with her current partner is rather problematic as well: serious problems of

communication, misunderstanding in sexual relationships, falling out of love and personality issues are

cited as problems, while Alina has already seriously considered divorcing. Again, members of the

family configuration participate in rising the tensions within the current couple: the ex-partner, the

son, the children of the current partner and the mother contribute to increasing the problems existing

88
between Alina and her partner. In comparison, only the two friends have a positive impact according

to her.

Overall, this family is representative of configurations that branch out in a large number of

directions after remarriage, with vivid interdependencies linking households and individuals, in

contradictory interdependencies. Although the network of supportive ties is rather dense, it is linked

with conflicts and tensions. The low level of resources and the localism of the family impose to all

members a type of collaboration which is linked with ambivalence. The maintenance of relational

closeness has a price that is paid by all family members, who are embedded in non transitive triads, in

which two individuals with whom they share emotional closeness are in conflict. In other words,

becoming a post-divorce family in which a large number of interdependencies are present has a cost

for individuals who have to cope with structural ambivalence.

Consider now a second family configuration, that of Dora, a 36 year old woman, first time

pregnant when she was interviewed. She had never been married before but has had several intimate

relationships. She has been with Don since she is 28. Don has two daughters, 11 and 9 years old, for

whom he has a shared custody with his previous partner. He has been divorced from her for several

years at the time of the interview. Dora has no grandparents alive and no uncles and aunts on either

her father's side or on her mother's side of the kinship network. But she has a sister who has played a

critical role in her life. As a result, her family configurations focuses on her parents, her sister, her

friends, her children and her partner. She has not maintained any connection with her previous

partners. The emotional interdependencies are organised in three distinct subgroups: her current

partner and his daughters, her parents and her sister, including her sister's partner and children, and her

friends. There are no direct connections between these subgroups and Dora plays an intermediary

position within the family configuration. Within each subgroups emotional interdependencies are

numerous and often reciprocal. In addition, the density of conflict is lower than in the previous

configuration. For the most part, conflicts are embedded in supportive ties, leading to dyadic

ambivalence. In other words, no relationship is built on conflict only and a mix between conflict and
89
support characterizes each of the three subgroups belonging to the family configuration. In that respect

the strain associated with negative interdependencies is lower than in Alina's family configuration.

This has an impact on her conjugal relationships, which have a low level of conflict and of

dissatisfaction. When asked about who may hinder her partnership, Dora only cites the children of her

partner, who are also sources of support for her.

Overall, this second example of a family configuration stemming from divorce and remarriage

looks quite similar to the sibling family configurations that was found in a large number of cases in the

sample of young adults58. In other words, the model is not peculiar to individuals who have gone

through a divorce and a remarriage. Of course, the two children from the previous marriage of her

current partner create a peculiar situation. But the family configuration that has developed based on

that situation share many features with family configurations found for women who have not yet

divorced: a couple, two children, with a series of relatives and friends considered as family members.

The respondent has built her family configuration by disregarding her partner's former partner and her

own former partners. Therefore, she took out much of the complexity that characterized the previous

example, with its conflicting triads.

Another, and more radical strategy is featured by Laura, a 38 years old woman who has a son of

three, born from a previous partnership that lasted five years and ended up right after the birth of her

son59. She met her current partner three years ago, soon after having broken up with her previous

partner and came with her son to live with him and his two children. Her family configuration is very

small and is focused on her nuclear family: her current partner, his two sons, and her own son. She

does not include her parents, her two brothers or any uncle or aunt. The definition of her family as

nuclear has significant consequences, as small groups are associated with a high density in all kinds of

networks (Wasserman & Faust, 1994). Again, as in the previous case, conflicts are embedded in

supportive ties, with many ambivalent dyads. Laura is central in her family configuration, both in

terms of support and conflict. Compared with other family configurations of that sample, the level of
58
See chapter 2.
59
This case is not represented graphically.
90
her conjugal and parent-child problems is low. But Laura said that she withdrew from the education of

her partner's children, which contradicts the ideal of the nuclear family in which the mother is in

charge of all children. Overall, this configuration features a focus on the household as the significant

family unit, in an attempt to conform to the model of the nuclear family even more than many

individuals who have not yet experienced divorce and remarriage.

There are various ways in which individuals build up their family configurations after going

through divorce and remarriage. Some focus on the maintenance of meaningful ties with their previous

partners, often for the sake of children. Others reinvest friendship and kinship ties in attempts to

compensate for the permanent or temporary weakening of their partnership. Others still try to build a

new nuclear family and focus on their current partnership and the children that they are responsible

for. In all cases, configurations of supportive ties that develop after divorce and remarriage have

consequences for conflicts and ambivalences. Many families in late modernity develop structural

ambivalence with their increasing complexity. Indeed, as the beanpole family configuration is

replaced by alternatives such as the post-divorce or the conjugal family configuration, ambivalences

gain ground. But some individuals resist and try to shape simpler family configurations despite the

circumstances.

Divorce, Family Structures and Configurations

It is now time to address the crucial issue of family configurations and family structures in more

depths. The previous sections revealed that individuals who went through a divorce and a remarriage

develop a diversity of family configurations; some of them create interdependencies that resemble

those of nuclear families. Family configurations are therefore different from family structures. Divorce

and remarriage do not necessarily lead to a post-divorce family configuration, especially when

individuals try to rebuild a new nuclear family by weakening or disregarding their ties with their

former partners. Getting back to the 101 cases of the study on older adulthood and parenthood, one

straightforward way of dealing with this issue is to compute the distribution of the configurations

91
according to family structures. Confirming the three cases that we have just considered, Table 13

shows that only a minority of individuals who went through a divorce and remarriage belong to a post-

divorce family configuration. Many have developed other sets of interdependencies, with their new in-

laws or with their blood relatives, that twist their family configurations into other directions.

INSERT TABLE 5.1 HERE

Table 13. Distribution of configurations according to the occurrence of divorce and remarriage

As a matter of fact, only 20 percent of individuals who have divorced and remarried belong to a

post-divorce family configuration. In other words, only some of them have an ex-partner who

continues to play a significant family role in their life, or children from the current partner who are

interdependent enough to be included as significant family members. Many have rather invested in

the development of horizontal ties with siblings and friends or, alternatively, have built a nuclear

family configuration while severing the ties with their former partners. Seldom have they however

been able to develop a beanpole family configuration or a conjugal configuration 60. Thus, having

experienced divorce and remarriage lead to a distinct probability of belonging to the various

configurations, while not imposing any of them.

In the eyes of sociologist Talcott Parsons, the Family was defined as the household. Little doubts

existed in the mind of researchers of Parsons' time that families could be objectively delimited and

corresponded to individuals who lived together and whose relationships were institutionalized by

marriage61. The pluralization of families since then has clearly stressed the inability of this perspective

to understand families in late modernity. In the previous chapters we have emphasized that actual

interdependencies rather than institutional criteria such as household membership should be the point

of departure of family research. This statement has strong implications for the study of post-divorce

60
See chapter two for a description of these types of configurations.
61
Even though the complexity of families in late modernity is frequently acknowledged by researchers , the
fuzziness of family boundaries and the intertwined interdependencies linking family members are not taken into
account empirically, especially in quantitative research.
92
families. How post-divorce configurations should be defined? As a matter of fact, not all divorces and

remarriages produce the same configurations of interdependencies. Research rather suggests a

diversity of custody arrangements and alimony practices. Children and parents, as well as former

partners, are embedded in a great variety of interactions, mixing various kinds of interdependencies.

Some former partners remain in contact and even maintain or develop some form of friendship. Others

become total strangers for each other quite quickly after splitting up.

The disconnection between the demographic fact of divorcing or remarrying and the

interdependencies developed afterwards reveals that family configurations and family structures are

distinct. Two realities should indeed be distinguished. The first reality is the family structure stemming

from a variety of hard demographic facts: did the couple divorce or not, how many children were born;

how many grandparents, uncles and aunts are still alive? The crude number of individuals living there

constitutes the backbones of family configurations, as it creates the potential for family

interdependencies to emerge. The second reality is the family configuration as such, which captures

the hard social facts of interdependencies between family members. Without a family structure, there

is no family configuration; but family configurations are not a full reproduction of family structures

either. Obviously, divorce and remarriage are only two dimensions that shape family structures among

many others, such as the distribution of individuals across generations and sexes, the localization of

individuals in the geographical and social space, the fertility rate and the life expectancy of family

members. This series of factors may make two family structures look very similar despite one

stemming from a divorce and the other from a first marriage. Indeed individuals who went through

marriage, parenthood, divorce and remarriage develop various family configurations. There is a great

diversity of possible paths to family life in late modernity, with quite heterogeneous consequences.

Two individuals having divorced and remarried may differ in other crucial life crucial experiences

that make their family configurations absolutely distinct. The number of generative mechanisms of

family diversity is great and one cannot expect to summarize this diversity by a single distinction,

namely that between first-time and stepfamily structures. In chapter two, we found that four types
93
were roughly enough to categorize the diversity of family structures in the sample of older adults:

avuncular, vertical, extended and step. When these four types of structures are compared between

individuals who have experienced divorce and remarriage, and those who have not, one realizes that

divorce is only one life transition among many others. The greatest differences concern the extended

family structure and the stepfamily structure. About half of women who remarried have a stepfamily

structure and only one quarter of them have an extended family structure. In comparison, 38 percent of

women without a divorce come from an extended family structure.

Overall, there is a first causal effect between events such as divorce and family structures. Not all

individuals who have experienced divorce and remarriage end up being in the same family structures.

A second causal effect concerns the link between family structures and family configurations. Not all

stepfamily structures lead to post-divorce family configurations. Being in a stepfamily structure, one

has about one chance over three, according to the study on older adults and parenthood, to be in a

post-divorce configuration. Indeed, many individuals who divorced and remarried weaken their

interdependencies with their former partner and do not activate interdependencies with their former

partners' new children or new partner. In other words, despite being in a stepfamily structure, with

stepchildren and stepparents around, individuals may focus on their new partner, their siblings, their

biological parents or their grandparents, reactivating the ties existing with members of their kinship

networks or focusing on their conjugal life.

The variety of family configurations that stems from stepfamily structures is fascinating. It is

even more fascinating when one realizes that the stepfamily structure is only one possible structure

stemming from divorce and remarriage. Although each stage of the causal chain linking the fact of

divorcing and remarrying to post-divorce family configurations is strongly channelled by the previous

stage, the combinatory power of events and structures is so great that it builds a variety of pathways to

current family interdependencies. It does not follow from the fact that one is remarried that one is

embedded in a stepfamily structure. It does not follow from the fact that one is embedded in a

stepfamily structure that one has developed interdependencies that match the post-divorce family
94
configuration. Of course, each stage increases the likelihood of the post-divorce configuration to rise.

But it never fully predicts it.

When one talks about stepfamilies, one actually refers to three distinct realities. First, and most

obviously, being a mother in a stepfamily alludes to the fact that the person went through a set of

transitions: starting a partnership, becoming parent, splitting up, starting a new partnership with

children involved. This is the most event oriented definition of stepfamilies. Second, this set of

transitions may lead to a post-divorce family structure, one in which there are children from a

previous partner, or children that the current partner has had with another person. In addition, other

kinship members, especially members from the older family generations on the male partner's side, are

less numerous, probably pending on the life stage at which remarriage occurs. Thirdly, the stepfamily

refers to a specific kind of configurations defined by active interdependencies, positive and negative,

with the former partner and his kin, and with the children and possibly ex-partner of the current

partner. Life events and transitions, family structures and configurations of interdependencies are

related but nevertheless distinct.

The Cultural Meaning of Stepfamily Statuses

Although research on post-divorce families has focused on interpersonal relationships, some

scholars have underlined the lack of an adequate language for dealing with stepfamily relationships

and the negative connotations of step-terms in the United States (Bohannan, 1970; Cherlin, 1978;

Coleman and Ganong, 1987), as in other cultural areas. It has been suggested that the role of the

stepparent is incompletely institutionalized (Cherlin, 1978; Cherlin and Furstenberg, 1994): social

norms do not tell individuals how to behave and what to expect in stepfamily relationships. Research

on stepfamilies in France has stressed the stereotypes associated with the stepmother status and the

difficulties of dealing with a relationship which is linguistically undefined (Théry, 1993). As a matter

of fact, in French, step-terms are confounded with in-law terms. Thus, one cannot distinguish

linguistically a “step-mother” from a “mother-in-law”, because both are referenced using the term

95
“belle-mère”. The lack of a specific vocabulary to deal with step-relationships was also underlined in

the North American context (Bohannan, 1970; Cherlin, 1978; Coleman & Ganong, 1987).

One basic assumption of the configurational perspective is that dyads are embedded in large

configurations of interdependencies and that it is impossible to account for them without considering

this larger relational context62. Since the nineteen sixties, cognitive anthropologists have been finding

similar interdependencies among meanings in a variety of cultural domains around the world

(d'Andrade, 1995). Meanings of birds, emotions and diseases are embedded in networks or maps.

Anthropologists ask their informants to make dual comparison between a variety of birds or emotions

in order to estimate the similarity of meaning existing between them. Those analyses have revealed

highly structured cultural patterns, as well as a large degree of sharing of those patterns around the

world.

The kinship realm is no exception. Since the beginning of cognitive anthropology, researchers

have been interested in American kinship. They have identified several key components that defined

the ways in which individuals think about the Family. They include gender (Nerlove and Burton,

1972), linearity and generational distance (Romney and D’Andrade, 1964; Wallace and Atkins,

1960). Indeed, studies done in the nineteen sixties showed that male and female family terms (mother

versus father) were perceived as distinct in meaning. Likewise, family members from the same lineage

(grandfather, father, son) were considered closer in meaning than individuals from distinct lineages

(siblings, first-cousins, other cousins). Interestingly the extent to which consanguinity influenced

meanings of family terms was not estimated at that time, maybe because blood relatedness was such

an obvious component of families in their perspective.

How do stepfathers compare with fathers, stepmothers with mothers, step-siblings and half-

siblings with siblings in terms of cultural meanings? Are there any differences between the United

States and Switzerland? To answer this set of questions, we set up three samples. In addition to the 89

subjects of a north American university used in a previous study 63 , we collected data from the French
62
See chapter one.
63
These results are largely drawn from a study published in American Anthropologist (Widmer, Romney &
96
speaking part of Switzerland and one sample from the German speaking part of Switzerland. The

Swiss French sample includes 52 college students from three schools of higher education in French

area. The Swiss German sample includes 38 students from one school of higher education from the

North Western part of Switzerland.

In order to compare cultural meanings of stepterms with terms refering to the nuclear family, a

subset of 14 family terms was selected. We focused on all family terms that exist in nuclear and post-

divorce families. These terms are father, mother, son, daughter, brother, and sister, which we compare

with stepfather, stepmother, stepson, stepdaughter, stepbrother, and stepsister. We also included half-

brother and half-sister. Seven female terms and seven male terms belong to this list. The generations

above and below respondents include four terms each. Respondents’ generation has six terms. There

are six terms that connote a full blood relationship and six terms that are not blood related, while half-

sister and half-brother have an intermediate position. Similar distinctions can be made in German and

French. The list of terms in the three idioms is presented in Table 14.

INSERT TABLE 5.2 ABOUT HERE

Table 14. Kinship terms in English, French and German

Respondents were asked to rate the similarity of each pair of terms in meaning, on a scale from 1

(extremely similar) to 7 (extremely different). For instance, they had to estimate how similar on that

scale were father and mother, father and stepfather, father and son. Their responses were analysed

using correspondence analysis, following well known procedure for the analysis of such data

(Widmer, Romney & Boyd, 1999). The first axis of correspondence analysis captures blood

relatedness, with a distinction between full blood relationships and stepfamily relationships, while

half-siblings are on the middle of the axis. Axis two is also highly structured, with family terms of

generations+1 (father, mother, stepfather, stepmother) and -1 (son, daughter, etc.) on one side, and

Boyd, 1999), and from an unpublished study applying the same research design on family terms in Switzerland
97
family terms of generation 0 (siblings, half-siblings, stepsiblings) positioned on the other side of the

axis. The third and fourth axes of correspondence analysis also show a highly structured pattern of

results. In all three contexts. axis 3 deals with generational distance but in a different way from axis 2,

ranking terms from generations +1, 0 to -1. In that case, rather than emphasizing the complementary

between parents and children as in axis two, there is a continuum across generations, from parents,

siblings, to sons and daughters. axis 4 deals with the gender of terms, with all the female terms on one

side and all the male terms on the other side.

Overall, four highly interpretable axes come out of correspondence analysis. Consanguinity

accounts for more than 50 percent of the total variance in the north American sample, while generation

accounts for about one-quarter of the variance and gender for less than 10 percent. The distinction

between stepterms and nuclear family terms appears on the first axis of the correspondence analysis

and accounts for more than half of the total explained variance of the average scores for the 91

comparison pairs. Thus consanguinity makes a tremendous difference in American kinship. The two

Swiss contexts only show slight variations of this pattern. The effect of generations is more

pronounced in the Swiss German sample than in the United States and Swiss French contexts. Gender

only has an effect in the American sample, which is still quite limited. Overall, there is a highly

structured and shared cultural view of family statuses in the three contexts. The structural methods

used in this analysis prove that a small set of components account for the judgements of similarity and

differences among family them.

Individuals in post-divorce family configurations have to deal with the implications of such highly

structured cultural meanings of family statuses in their actual interdependencies. Indeed, the conflicts

experienced in some family relationships may stem from the ambivalence created by the obvious

interdependencies existing between stepchildren and stepparents, and the concurrent distancing from

the non-guardian biological parent, in cultural contexts that define the stepparent as remote from the

parent. The realm of kinship and the family has not turned into a cultural fuzziness. Individuals have

to comply with cultural meanings of family statuses that still leave little space for alternative types of
98
family organization such as post-divorce families. This certainly constitutes one origin of the

ambivalence characterizing families in late modernity.

Conclusion

A person who went through divorce and remarriage may well develop a family configuration

quite similar to the family configuration of a person who did not experience these transitions, and

quite dissimilar from to the family configuration of another person who also divorced and remarried.

We expected much uniformity in post-divorce families; we found diversity again. There is no major

way of reshaping family interdependencies after divorce. Post-divorce family configurations are only

one alternative among others. This alternative is indeed marked by specific interdependencies: a

higher share of structural holes and weak ties, and more structural ambivalence than in other family

configurations. Keeping regular contacts with both parents for children in post-divorce families

implies a new kind of relational experience within families, which may have long term consequences

for the self64. Post-divorce family configurations are indeed one of the vanguards of such a

transformation in the family realm.

That said, many individuals who go through a divorce and a remarriage are embedded in other

types of configurations. They develop patterns of interdependencies similar to those of individuals

who did not experience this transition, by investing the field of kinship, of friendship and of other

pseudo-kinship ties, or by reinventing a nuclear family. Rather than being the vanguard of family

decline, individuals confronted with divorce and remarriage organize their family configurations in a

variety of ways, borrowing and adapting well-known kinship and friendship informal rules to the

peculiarities of their situation. Therefore, the fears and interrogation marks regarding families and

social integration are reviewed in a brighter light when one stops considering that labels such as “the

stepfamily” or “single parent family” correspond to homogeneous realities. Family configurations

steming from divorce and remarriage are diverse in their composition and interdependencies.

64
We once more get back to the Simmelian concept of the “intersecting circle” (Simmel, 1999) that describes
late modernity: individuals belong to individualized circles that define them in their uniqueness.
99
 One may wonder whether or not the stepfamily as a concept is truly useful, taking into account

this diversity. It is certainly true that individuals who went through divorce and remarriage develop

specific family experiences. It does not mean, however, that they all share a common family pattern,

characterized by a lack of social integration. If, in some cases, they are indeed characterized by a

deficit of bridging and bonding social capital, these cases are however a minority. Individuals develop

various ways of recomposing their family life, either by reinvesting their kinship network, developing

interdependencies with their partner's relatives, or with friends considered as family members. Many

individuals get back to family configurations quite similar from those that were found for individuals

who have not yet divorced. The development of such or such configurations has some important

implications for the availability of social capital in families, and probably, but this should be further

investigated, for the development of children. Therefore, we might be better off in our understanding

of families in late modernity to focus on interdependencies among family members beyond the social

labelling of family situations as “intact” or “blended”.

100
 CHAPTER 6 – Families and Psychiatric Problems

Family configurations of individuals with psychiatric problems are usually considered an anomaly

by sociological research. Their complexity is not well served by standard approaches of the Family

using survey designs and random sampling or in-depth qualitative interviewing. The strong link

existing between psychological problems and family relationships has often been interpreted as a sign

that these relationships only responded to psychological causes and processes and therefore that they

did not belong to the field of sociology. On the contrary, this chapter stresses that studying family

configurations of individuals with psychological problems from a sociological perspective leads to

new insights on the changing nature of families in late modernity.

So far, we have only considered families of individuals without a clinical record. Do individuals

with psychiatric problems stick with the nuclear family model and the well ordered view that it implies

on how social integration work in families? One may hypothesize that families of individuals with

psychiatric problems and a role of patient in either a private or public psychiatric practice also go

beyond the nuclear family. That has consequences for the understanding of the resources made

available by their family configurations, as well as for the conflicts that they develop with their family

members. Indeed, we saw in previous chapters, that the composition of families has an influence on

social capital as well as on conflict and ambivalence. The importance of family members beyond the

household creates a great diversity of family forms. This diversity may arise in families facing

psychological problems of one or several of their members. In other words, one expects diversity

again rather than a single model of family configurations, characterizing all individuals with

psychological problems alike.

This chapter first asks whether individuals with psychiatric problems have distinct family

configurations, compared with individuals without a clinical record. Then, the issue of the additional

impact of mental impairment on family configurations is addressed based on interviews of individuals

with psychiatric troubles and of one of their family members. Eventually, a third study was designed
101
to estimate the diversity of family configurations in which individuals under psychiatric supervision

are embedded.

Family Systems and the Configurational Perspective

Current empirical research regarding family ties of individuals with psychiatric problems focuses

on the impact of positive family relationships on psychological well-being and adaptation to crises, as

well as on the ability of family members to cope with the patient's troubles. Individuals who belong to

families in which relationships are gratifying and which adequately support their members, are less

prone to depression and relapses than others. Note that research has been mostly concerned with the

amount of family support provided to individuals with psychiatric problems. This focus has relegated

the interest for family configurations and their structural properties to a marginal position in the field.

Interestingly, classical authors in system theory developed a deeper interest in structural

dimensions of family interactions (Broderick, 1993). They pointed at the various family conflicts and

tensions associated with mental health problems, and raised the issue of their spread throughout

various family subsystems, from the conjugal subsystem to the parent-child and the sibling

subsystems. Various instruments such as the Genogram, the Triadic Play and the FAST, have focused

on the impact of triads and of larger configurations of family ties over conjugal dyads, parenting and

psychological troubles (Corboz-Warnery et al., 1993; Fivaz-Depeursinge & Corboz-Warnery, 1999;

Fivaz-Depeursinge & Favez, 2006; Gehring, 1998; Gehring & Wyler, 1986). The configurational

perspective makes it possible to empirically deal with some of the concepts and hypotheses of system

theory, by formal methods tailored to analyse complex systems of relationships 65. It has many things in

common with the work of Murray Bowen and Salvador Minuchin (Kerr & Bowen, 1988; Minuchin,

1974) , as both of these prominent scholars of family system theory emphasize that family dyads are

interdependent. Minuchin's interest for the boundaries of family systems is close to the issue of family

65
Let us also again stress the contribution of Jacob Moreno (1934). Moreno was a psychologist and inventor or
the psychodram, which was related to his sociometric work. Although his influence on current systemic approach
of families in troubles is weaker than that of other renown scholars, his contribution to a structural approach of
families should not be underestimated.
102
composition, and Bowen's work on triangulation has many acquaintances with the interest of network

research for triads and transitivity. The configurational perspective draws much from their insight that

family dyads are interconnected. Following their lead, it stresses the complex patterns of negative and

positive interactions in which individuals with psychological problems are embedded, rather than

focusing on the amount of support provided by family members.

The organization of family interdependencies and the resources that they provide is likely to be

shaped distinctly in family configurations of individuals with psychiatric problems. Systemic research

on a variety of psychological troubles emphasizes the inter-adaptation to each other of the family

system and the individual in troubles. From the configurational perspective, three distinct but

nevertheless related issues should be addressed: the composition of the family configuration, the type

of social capital that it provides, and the conflicts and ambivalence created by family

interdependencies. We now address the two first issues and leave the third one for the end of the

chapter.

Family Composition and Social Capital

One may expect that while individuals with psychiatric troubles lack some important family ties,

they are likely to compensate this deficit by developing alternative family relationships such as those

stemming from their kinship network, or even by developing family-like interdependencies with

professionals from the care support system. Thanks to those ties, individuals with psychiatric problems

may have relational resources similar to those of individuals without psychiatric problems. An

alternative hypothesis states that individuals with psychiatric problems have a smaller number of

significant family members because they have a lower probability to experience a stable partnership

and parenthood (Burnand et al., 2004; Pescosolido & Wright, 2004). A deficit of family members may

mean that bonding social capital is less likely for individuals with psychiatric problems because of the

negative effect that psychiatric problems have on family integration (Olson, 1989). As for bridging

social capital, the deficit hypothesis states that it is lower in clinical populations because of the

103
difficulty for individuals with psychological instability to maintain significant ties with others, and

because of the feelings of dissatisfaction and incompetence of family members (especially parents and

siblings), sometimes leading to their withdrawal from interactions.

A first research that dealt with the issue of family based social capital took place in the

rehabilitation unit of a department of adult psychiatry in a large public hospital 66. This unit treats

outpatients suffering from enduring psychiatric problems with potentially disabling effects. It aims at

reintegrating individuals with psychiatric troubles within professional and social life. The clinical

sample includes a total of 54 individuals in young adulthood, with a DSM-IV diagnosis of mood or

personality disorder. They live on their own or in community housing. They were compared with a

non-clinical sample that comprised a total of 54 individuals from the sample of young adults described

in the previous chapters. Both samples were matched for age and sex. Figure 12 shows several

examples of emotional interdependencies in family configurations drawn from the clinical sample.

INSERT PICTURE 6.1 HERE

Figure 12. Family configurations of individuals with psychiatric problems

As the graphs of Figure 12 show, the composition and social capital of individuals with

psychiatric problems are distinct from those of individuals that we considered in the previous

chapters. First, family members are significantly less often resources of emotional support. On

average, respondents include 1.1 family members as support providers for them, compared with 4.3 in

the sample of college students67. On average, individuals with psychiatric problems have family

members who are less connected with each other. They are also significantly less central in their set of

supportive family members. Results are similar for family members to whom respondents provide

emotional support. Individuals with psychiatric problems provide emotional support to a much smaller
66
This section is based on a series of results published in the International Journal of Social Psychiatry (Widmer
et al., 2008).
67
In order to precisely estimate the impact of psychiatric problems, the twin sample technique was used. For
each individuals with a psychiatric background, we set it an identical twin without a clinical record as belonging
to a non-clinical comparison sample. Twins are the same sex and age than their alter-ego. Therefore, the impact
of these two variables is controlled when assessing the differences of family configurations between individuals
with and without a clinical record.
104
number of family members than others (1.9 compared to 5.5 in non-clinical samples). The

interdependencies among the family members whom they support are significantly fewer. Their

centrality in their set of dependent family members is also on average smaller. In other words, they

play a less active role as support providers and their status in the group is decreased as much.

To summarize, the amount and the structures of social capital are strongly influenced by

psychiatric problems. Individuals of the clinical sample have a smaller number of supportive family

members, linked by fewer significant relationships. They can count on only one to two persons, on

average, as support providers. Hence, many individuals with psychiatric problems do not have an

access to bonding social capital in their family configurations. This situation causes social isolation in

life transitions such as the transition to adulthood or the transition to old age. The negative impact of

parental divorce on care to individuals with psychiatric troubles was for instance underlined, as well

as the issue of care when support providers grow old. Interestingly, the issue is not only that of

receiving care but also of becoming caregivers and playing as such a meaningful and valued social

role. Individuals with psychiatric troubles indeed seldom play a role of support providers for other

family members, which raises some concerns regarding the acknowledgement of their significance in

the family.

In addition, they have a lower centrality both in their direct circle of supporters and in their family

configuration as a whole, which makes their bridging social capital low as well. The amount of

energy, relational competency and time necessary to be a bridge between otherwise disconnected

family subgroups is beyond the resources available to many individuals with psychiatric problems.

Bridging social capital, as the ability to develop personal interdependencies between family members

that are otherwise disconnected, is lacking. The deficit hypothesis is therefore empirically confirmed.

Overall, the family as a main support provider, as a place in which meaningful roles (such as helping

others) can be experienced, as a social control agency or as a help to connect with a variety of persons,

is much less so for individuals with psychiatric problems.

105
 This deficit of relational resources and integration within family configurations goes along with

the strong investment necessary to care for individuals with psychiatric problems. Because psychiatric

problems increase the level of family stress and reduce the quality of family life and well-being

(Miklowitz, 2004), care providers are frequently physically and emotionally overwhelmed by the

constant requirement of helping a psychiatric family member. Parents, in particular, often describe

themselves as exhausted, desperate and hopeless. For many of them, the difficulty to understand the

origin of the troubles leads to feelings of dissatisfaction, incompetence in the parental role and low

self-esteem ( Miklowitz, Goldstein & Nuechterlein, 1995). These feelings, in turn, lead to their

withdrawal from other family relationships. They are frequently the main reason for deciding for

institutionalization (McIntyre, Blacher & Baker, 2002). Indeed, it is difficult to maintain a relational

life of its own while facing psychiatric problems of a family member, as these problems become an

organizing principle of the family configuration. Interviews with parents of individuals with

psychiatric problems reveal that the demands and problems of their children let little space for other

relationships to develop. Interdependencies between the person with the troubles and her closest

family members (usually the parents, sometimes one sibling) are so important that other

interdependencies are left aside as less critical. The lack of social capital for individuals with

psychiatric problems has therefore indirect effects on their parents' social capital.

This lack of family based social capital is to a significant extent due to the composition of family

configurations. Individuals with psychiatric problems have family configurations of a much smaller

size68. They do not often include partners and, as a consequence, they also do not have in-laws. Blood

ties are also less frequently activated. Overall, individuals of the clinical sample less often include

their father and siblings, grandparents, uncles and aunts, and friends as significant family members.

Interestingly, the inclusion of mothers remains at the same level as in non-clinical samples. Do these

68
The average size of family configurations is 9.7 for the sample of college students and only 6.2
for the clinical group. In the clinical sample, respondents included a total of 50 family terms after
standardization of minor terminological differences, among which 25 were included by one
respondent only. In the non-clinical sample, respondents included a total of 76 family terms, among
which 36 were included by one respondent only.
106
results correspond to a lack of acknowledgement of existing family members by individuals with

psychiatric problems or to the absence of such family members in the family structure 69? Only five

respondents over 54 had a partner at the time of the interview (nine percent). Therefore, the lack of

inclusion of partners in family configurations closely matches an absence of partners in the family

structure. The difference for blood ties, however, is not related with family structures. On average,

respondents with psychiatric problems have 1.7 siblings, 1.8 grandparents, 6.4 uncles and aunts, and .5

step-relatives. This set of figures is similar to what was found in the non-clinical samples. In other

words, compared with individuals of the non-clinical sample, individuals with psychiatric problems

include fewer members with a blood connection in their family configurations, although they have as

many in their family structures. The difference concerning partners goes along with the low rate of

married or cohabiting individuals with psychiatric problems. In addition, there is no compensation of

the relative lack of blood connections or partners by the inclusion of alternative ties in family

configurations, such as friends or professionals from the institutionalized care system considered as

family members. Thus, the deficit hypothesis is empirically confirmed: compared with others,

individuals with psychiatric problems have much smaller family configurations, with less bonding and

bridging social capital.

This shortage of family members and of social capital provided by family configurations has

meaningful consequences for the social integration of individuals with psychiatric problems. As the

Family remains the central institution of social support and social control in late modernity societies,

the deficit of bridging and bonding social capital that it provides to individuals with psychiatric

problems casts doubt on its ability to deal on its own with their social integration. The possibility for

families to take care of their impaired members without significant institutional support is therefore

limited and attempts to heavily rely on them in order to decrease the costs of public health may create

serious problems and shortcomings.

69
For the distinction between family configuration and family structure, see chapter 5. A similar issue was raised
for individuals belonging to non-clinical samples. Indeed, the previous chapters showed that family
configurations depend on family structures but do not replicate them.
107
Intellectual Impairment

Little is known about family configurations of individuals with intellectual impairment in addition

to their psychiatric troubles. What kind of relational resources do their families provide to them?

Research underlines the imbalance that intellectual impairment creates for family interdependencies.

As it is the case for individuals with psychiatric problems, the presence of a person with intellectual

disability in a family often has a negative effect on family members, as it means a heavy load of

overwork. Even if a significant reward may also be found in care giving (Heru 2004) and even if

parents of people with intellectual impairment frequently report positive aspects and an overall

satisfaction in their quality of life concerning their involvement with their child’s education (Jokinen

and Brown, 2005), the existing literature mainly underlines how family organization, the function and

role of each member, and the family interactions, are negatively affected by the presence of a disabled

person (Lambert & Lambert-Boite, 2002).

First of all, the presence of a disabled individual modifies the affective regulation in the family

configuration. For instance parents frequently develop overprotection towards the person with

intellectual impairment, and they do not have the same availability for their other children. The focus

on the disabled individual in some case isolates her from the rest of the family and especially from her

siblings. Parents could also feel sympathetic with their disabled child in such a way that they reject the

handicap reality and the social reactions that it triggers. The expression of negative emotions

(aggressiveness, shame, anger, jealousy, rivalry and guilt) is often repressed as well as questions and

interrogations, each family member attempting to keep her distress secret (Meynckens-Fourez &

Tilmans-Ostyn, 1999). This is not without consequences for family interdependencies as it means that

rather than dealing with a large number of ambivalent relationships, parents may choose to reduce the

number of positive as well as negative ties, while increasing their own isolation 70. Scholars also

suggest that psychiatric problems associated with intellectual impairment have a stronger impact on

70
This may lead the family configuration to either atomization or captivation. See chapter 4 about the four
theoretical types of situation relating conflict and support.
108
the family than the intellectual disability itself (Maes, Broekman, Dosen & Nauts, 2003). Family

members are frequently physically and emotionally overwhelmed. Parents, in particular, described

themselves as exhausted, desperate and hopeless. These feelings, associated with psychiatric and

behavioural problems of intellectually impaired individuals, are often the main reason to decide for

their institutionalization, as family members consider that professional help providers are more able

than themselves to solve them (McIntyre et al., 2002).

This set of evidences explained why, compared with non clinical individuals, individuals with

intellectual impairment include fewer siblings and fathers in their family configurations in the sample

under study (Widmer, Kempf, Lanzi, Robert-Tissot, Galli-Carminati, 2008). Contrary to individuals

with psychiatric problems without mental impairment, they compensate these losses by including

professionals and alternate family roles provided by more remote kin. The resources made available to

them by family members are however smaller, as parents, partners, children and siblings are the

primary support providers in families (Widmer, 2004). This difference of composition of family

configurations has hence consequences for social capital. Indeed, individuals with psychiatric

problems and intellectual disability have only a small number of supported and supportive family

members, in very sparse circles. Family members are disconnected and respondents have a low

centrality in their family configurations. Therefore, individuals with intellectual impairment and

psychiatric problems do not benefit from the same amount of either bridging or bonding social capital.

The family as a main support provider is less so for them than for non-clinical individuals. As

individuals with intellectual impairment have a larger vulnerability to psychiatric illness than others

(Moss et al. 1998), this lack of family based social capital has an additional deteriorating effect on

their life chances, in making them less able to deal on their own with non-normative events associated

with their psychiatric troubles.

109
 So far, we have focused on the accounts of individuals with psychiatric problems, whose views on their

family configurations might be distorted by their psychiatric problems. Are results based on their interviews

confirmed by other sources of information? There is indeed support in the literature for the perception biases

of family relationships by individuals with intellectual impairment 71. Research emphasizes the cognitive

processes limiting their ability to adequately perceive their relationships with others. People with intellectual

impairment may be at risk of greater biases in perceptions of social relationships: other individuals’ feelings

are often misunderstood and interpersonal situations are perceived in a very subjective way ( Pescosolido &

Wright, 2004). Perceptions of interpersonal relationships is not only influenced by intellectual impairment

but also by its conjunction with psychiatric troubles. For instance, paranoid adult individuals perceive

family relationships differently than their parents (Rankin & al, 2005). Perceptions of individuals with

intellectual impairment concerning their position in the family are described as frozen, with no link with the

situation as described by their relatives (Nandrino & Doba, 2001). Note that the ability to adequately

perceive one's own social situation leads to the improvement of psychological well-being and is of great

importance for the individuals' caring (Kawachi & Berkman, 2001). Furthermore, a shared perception of

situation by family members reveals larger social abilities and larger availability of resources (Reiss, 1981).

An exploratory analysis of 17 interviews from the first family member included as a significant family

member by individuals with psychiatric problems and intellectual impairment provides some elements of

responses concerning the subjectivity of the accounts made by individuals with psychiatric problems on their

family configurations. For the most part, family members support their views 72. They confirm that

individuals with psychiatric problems and intellectual impairment are not central in their family, with only

few supportive ties available to them, and even fewer relationships in which they are support providers.

Interviewed family members support the views that individuals with psychiatric problems are marginal in

flows of support in their own families, either as support seekers or, and most of all, as support providers.

Therefore, individuals with intellectual disability with psychiatric problems are at risk of developing small

and sparse family configurations, with few ties and a higher ratio of conflict over support when compared

71
The issue of perception biases of social relationships has also received considerable attention in the literature on
social networks. Results on non clinical samples revealed that individuals bias the evaluation of their networks toward
higher reciprocity of relationships and a higher centrality for themselves in the network (Killworth & Bernard, 1976;
Krackhardt, 1987; Kumbassar, Romney & Batchelder, 1994).
72
This research is described in greater details in Widmer, Kempf and Galli-Carminati (2010).
110
with other families. Individuals with intellectual impairment and psychiatric problems have significantly less

social capital than individuals without a clinical record.

Social capital is strongly shaped by the presence of psychiatric problems. Compared with individuals

without a clinical background, individuals from the clinical samples have only a small number of supported

and supportive family members, who are more frequently perceived as disconnected from each other.

Therefore, individuals with psychiatric problems do not benefit from the same amount of bridging and

bonding social capital. This lack of social capital is likely to have an additional deteriorating effect on their

life chances in making individuals with psychiatric problems unable to deal with the numerous non-

normative events associated with their trajectories, with consequences for their adaptation to crises and

unexpected life events.

Variability Again

The previous chapters revealed that a great variability of family configurations characterize non-clinical

samples. The emphasis on variability was somewhat lost in the last sections, which underlined the

distinctiveness of family configurations in clinical samples compared with non-clinical samples. That

individuals with psychological health problems have on average fewer significant family members and much

poorer social capital does not imply that they all share a similar type of family configurations. To the

contrary, one expects that variability also characterizes them. Indeed, individuals with psychiatric problems

may have access to distinct relational resources depending on the composition of their family configurations.

To systematically address the issue of variability of family configurations in clinical samples, we used a

sample of individuals 61 individuals, all of whom were undergoing psychotherapy in Switzerland 73. On

average, respondents were in their mid-forties; their average age was 43 and 74 percent were women. They

had a variety of severe psychological problems, such as borderline troubles, psychosis, bipolar troubles, and

anxiety or mood problems. Respondents had to report about their family configurations every two to three

months during the period of a year and one half during which the follow-up lasted. Forty-two individuals

73
This sample includes individuals who are not institutionalized despite serious psychological problems. The larger
variability of housing arrangements and life trajectories, the fact that no individuals of this sample are institutionalized
in a psychiatric hospital during the study makes this sample more comparable to non-clinical samples.
111
over 61 participated in the five waves of interviews 74. A total of 251 family configurations were included in

the analysis, as up to five waves per individual were possible.

Five types of family configurations capture the variability of family configurations in that sample. The

first type of configurations (18 percent of cases) focuses on the nuclear family. They include children and, in

almost all cases, a partner. Other family members are systematically underrepresented. The average size of

those configurations is small with only 5.7 members. The kinship configuration (26 percent of

configurations) includes a large number of relatives from the kinship network. The mother, the father, the

uncles and the aunts, as well as their partners, are over-represented. Cousins and the partner’s relatives are

also included. In comparison with the first type, partners and children are less frequently included than in the

previous type. All other family members included are relatives by blood or marriage. Overall, this

configuration is the largest with 8.1 members included. The third type of configurations (22 percent of cases)

is similar to the beanpole family configuration of the previous chapters, although with a much smaller size

(4.4 instead of 8.1). It is focused on close blood ties from the family of origin (parents and siblings) and

excludes all relatives by marriage or partnership, unlike the previous type. Post-divorce Configurations (19

percent of cases) are characterized by the overrepresentation of the previous partner (included in half of the

cases) and the absence of a current partner. Relatives by blood or marriage are underrepresented, and the

family configuration does not include a nucleus that is constituted by one’s partner and children. This family

configuration is rather large and includes a great number of other relationships mostly associated with

divorce and remarriage. The fifth type (16 percent) includes configurations that focus on friends and care

professionals considered as family members. In this type, friends are as many as 2.5 persons, whereas blood

relatives, in-laws, and steps are almost absent. The partner and the children are also underrepresented.

There is indeed a large variability of family configurations in samples of individuals with psychological

problems. In other words, all respondents do not include the same family members. They put an unequal

emphasis on parents, siblings, aunts, uncles and cousins, friends and care professional considered as

significant family members. Despite their smaller size and their lower density, the composition of family

configurations of individuals under psychotherapy relates to that found in other samples. Indeed, the types

described above are similar, to a large extent, to those of the previous chapters. They have however

74
Dropouts were due to individuals quitting the therapy.
112
important distinct features because of the status of individuals as clients in psychotherapy. For instance, the

large number of individuals including a psychiatrist or a social worker as a family member is not found in

other samples. Figure 13 provide several examples of family configurations within which individuals with

psychiatric problems are embbedded.

INSERT PICTURE 6.2 HERE

Figure 13. Types of family configurations of individuals with psychiatric problems

113
 How does this variability affect the social capital available to individuals with psychiatric

problems? Although the level of density, the connectivity of interdependencies and the centrality of

respondents are much lower in this sample than in non-clinical samples, as exemplified in Figure 12,

there are still great variations of social capital according to the composition of family configurations.

In this regard, the results are quite similar to what is found in non-clinical samples. In beanpole family

configurations, there is a lower number of supportive family members and a higher density of

supportive interdependencies. The same happens in nuclear family configurations, which also

emphasize bonding social capital. In these family configurations, respondents are embedded in a dense

set of emotional interdependencies and have a low centrality in their families. Many connections

among their family members do not depend on them, and their family configurations are resistant to

their own removal.

By comparison, kinship family configurations provide a greater number of helpers and help

seekers, who are less frequently connected to each other. Overall, kinship family configurations are

associated with a binding type of social capital but within a larger and more pluralistic family context

than the nuclear and the beanpole family configurations. The friendship family configurations are

markedly different and include a large number of friends and care professionals who are considered

family members. Bridging social capital is dominant, as friends and blood relatives are kept separate

in several non-overlapping subgroups This implies that respondents benefit from a large structural

autonomy. Post-divorce family configurations are intermediate: Density and connectivity are on

average lower than in nuclear, kinship, and beanpole family configurations. Respondents in post-

divorce families, however, do not have the same centrality as respondents embedded in friendship

family configurations. Therefore they do not benefit from the same amount of either bridging or

bonding social capital as individuals with other family configurations.

These results are quite similar to those pertaining to non-clinical samples 75. They shed light on

the variability of family configurations and on the unequal amount of social capital that these

configurations make available to individuals with psychiatric problems. Family configurations vary in

75
See chapter three.
114
the extent to which they include friends or care professionals, children, parents, siblings, and other

relatives. Nuclear, kinship, beanpole, friendship and post-divorce family configurations trigger

distinct ways of building interdependencies with family members. These variations have an impact on

the social capital that they make available to individuals. In regard to bonding social capital, beanpole

and nuclear family configurations are optimal. But we already stressed above the burdens associated

with bonding social capital: family interference and family over-care have some detrimental effects as

they exert much normative pressure on individuals who already have difficulties to deal with social

expectations. As for bridging social capital, individuals with psychiatric problems embedded in

friendship family configurations may find some advantages. It is however dubious that the level of

activity that maintaining such a large number of disconnected family members implied is easily

achieved in their situation76.

Variability not only concerns supportive ties. Family conflict has been for years a central issue for

understanding individuals with psychiatric problems. In chapter 4, we found that conflict and

ambivalence were more frequent in beanpole family configurations than in other family

configurations. Is there a similar tendency in clinical samples? In other words, are some family

configurations associated with more conflict and ambivalence than others for individuals under

psychiatric stress? Results of Table 15 support this hypothesis: beanpole family configurations are

again associated with a higher probability of stress and ambivalence. Interestingly, the nuclear family

configuration also scores high in terms of conflict and ambivalence. Bonding social capital promotes a

state of conflict and ambivalence in families confronted with psychological problems. More horizontal

family configurations, such as post-divorce and friendship family configurations, with their emphasis

on bridging social capital, create fewer occasions of conflict and dyadic ambivalence for individuals

with psychiatric problems.

INSERT TABLE 6.1 ABOUT HERE

76
Chapter 7 gets back on the issue of stability and change of family configurations of individuals with
psychiatric problems.
115
Table 15. Support, conflict and ambivalence in family configurations of individuals with
psychiatric troubles

Conflicts that occur between individuals with psychiatric problems and their family members are

overrepresented in family configurations in which a high density of positive interdependencies also

exist. They are not, as such, the direct results of a patient's personality but rather the consequences of

long term interdependencies of family members with the patient. Over-care and over-concern for

individuals with clinical symptoms create a shrinkage of the family configurations that may either lead

to atomization and a lack of ties, or to an investment in a small number of helpers with a highly

ambivalent and conflictual family context. Dyadic conflicts involving the individual with psychiatric

problems eventually become the expression of a family system, a fact that is well-known in systemic

psychotherapy.

Conclusion

Compared with non-clinical individuals, individuals with psychiatric problems from the studies

reviewed in this chapter include fewer members from their nuclear family, especially partners and

children. This has profound consequences for relational resources available to them, as the conjugal

sub-system is a primary support system in late modernity (Widmer, 2004; Widmer, Kellerhals, &

Levy, 2004). Other family ties are also less frequently reported: the members of the larger kinship

network are under-represented. Therefore, the number of significant family members is lower. There

would be much to say about this shrinkage of family configurations triggered by psychiatric troubles.

Let us recall once again that the energy necessary to take care of family members with a clinical

background exhausts many individual resources. Parents, rather than connecting their child with their

larger kinship networks and friends, tend to withdraw from their social circles. This has consequences

for individuals with psychiatric problems, who are less able to build fruitful interdependencies within

their parents' family configurations. Intergenerational ties, in particular with grandparents, uncles and

aunts, but also with fathers, and the sense of continuity and the social control that they provide, are

severed. Ambivalence is therefore replaced in many cases by atomization 77. Mothers play a crucial
77
Atomization is defined as a lack of interdependencies among family members.
116
role for the social integration of individuals with psychiatric problems, as in many situations they are

the only stable family members in their child's life. This comes with a cost for mothers, as strong

interdependencies with their troubled child means a disengagement from other interdependencies. It

also has a cost for the child, as this situation makes him or her utterly dependent on a single person,

with obvious risk of loneliness on the long run. In two over three clinical samples, there is no

compensation of the deficit of partners, children and blood relatives by the inclusion of alternative

family members, such as those associated with the professional care system. Overall, the hypothesis of

a deficit of social capital for individuals with psychiatric problems is confirmed: they have a smaller

set of significant family members as compared to non-clinical individuals that is only partially

compensated by other sources of support. Hence, social capital is also much poorer.

That said, it would be inconsequential to state that all family configurations of individuals with

psychiatric problems refer to an homogeneous model of deficit. Individuals undergoing psychotherapy

define their family configurations in a variety of ways. They vary to the extent that they include

partners or in-laws, friends, parents, siblings, and other relatives. Some are fully isolated. Most of

them are not, however, and focus on either parents or siblings. Others also include care professionals

as family members. In some cases associated with divorce, partners and ex-partners intermingle.

Overall, researchers and therapists who adhere to the definition of the Family as a configuration

may become more able to understand the complex family contexts of individuals with psychiatric

problems because such individuals, as others, have experienced the consequences of the pluralization

of family life. The configurational approach overcomes the constraints associated with the definition

of families that matter as nuclear, as it does not predefine what the significant family configurations

are. It also makes it visible that various types of family configurations are associated with unequal

levels of relational resources for individuals with psychiatric problems.

117
 CHAPTER 7 – Short Term Changes in Families

When Gina was asked about her family configuration the first time, she included her partner, her

two children, her mother, her sister, the partner of her sister, her niece, several relatives of her partner

(his parents and siblings) and two friends that she considered close enough to be included in her

family. When she was again interviewed, six months after, she had split with her partner. As a

consequence, she did not include him, his parents and siblings in her family configuration any more.

Her sister was also going through turbulent times with her own partner, so that the ties between the

two sisters got reactivated, with much psychological support exchanged. Friends considered as family

members gained additional significance for a while. Another six months passed and Gina met a new

partner, who also had children from a previous relationship. Within a year, her family configuration

changed tremendously, due to a series of events that could quite easily be traced out. The story of

Gina is not uncommon, although personal stories usually feature more stable family configurations on

the short term. In many configurations, one hardly notices a change for years, although a closer look

might reveal subtle modifications of social capital and ambivalences.

How do family configurations evolve on the short term? By what mechanisms do they remain

stable or achieve new balances of support and conflict month after month? The issue of change in

family configurations is not easy to tackle. Quantitative research has addressed family change in

focusing on normative events, such as marriage or giving birth. In the family developmental

perspective, inspired by functional analysis (Parsons & Bales, 1956), expected transitions of life create

shift in family stages (Duvall & Miler, 1985). The importance of daily events that go along with life

was disregarded, even though participating to a family gathering, getting into a fight with a sibling or a

parent, leaving the country, putting an end to an intimate relationship or dating a new partner are

events that have consequences for family interdependencies. But how exactly do they modify family

configurations? Results described in the previous chapters are static. However, one assumption of the

configurational perspective is that family interdependencies are not once and for all set but that they

118
change overtime. This chapter considers some of the structural mechanisms that account for how

family evolve on the short term.

Focus Points

In order to understand the passing of time in a more detailed way, we interviewed Betty and her

family members nine months after the first interview. Figure 14 presents Betty’s family configuration

as defined by the second wave of interviews.

INSERT PICTURE 7.1 HERE

Figure 14. The family configuration of Betty nine months later

Changes happened during this nine month period: Compared with the first interview 78, Betty has

dropped her two friends as significant family members, and her sister-in-law Jane is not included in

the family configuration any more. But overall, a remarkable stability characterizes Betty’s family

configuration over time. The pattern of interdependencies has not changed much. Only Betty’s

mother, brothers and sisters are more loosely connected than on wave 1. From qualitative reports, at

least one major change has occurred in Betty’s family: her mother, Nina, who was living with Betty at

the time of the first interview, has moved out, following a series of disagreements with Betty. Nina

and Betty were no longer able to tolerate their living arrangement, which had lasted about one year.

Their understanding of life and their religious beliefs were too different for the two to be able to live

together, according to various family members. Betty also pointed at differences of perceptions of past

events related to her father. Therefore, after a “family vote” among all the siblings and Betty's

daughter, it was decided that Nina would live with Henry, one of the brother of Betty. Interestingly,

Nina had no real say in the vote. She was just told that she was to live there, since it was in both her

and Betty's best interest.

What kind of insights does the case of Betty provide? This case study suggests that in order to

understand why some individuals were added and other dropped in family configurations, one has to

take into account that concerns of individuals change over time. At time 1, Betty is deeply involved in
78
See chapter 2.
119
care and conflicts with her mother, which is not the case at time 2. This focus on her mother at the first

interview required some form of bridging capital to be present, which explains the inclusion of friends.

Indeed, Betty needed their support for counterbalancing the tensions created by the daily negative

interactions with her mother. On the second interview, Betty's focus on her mother is significantly

decreasing because of the new living arrangement decided by her and her siblings. This change has

created an opportunity for the configuration to evolve by a change of focus in Betty's life.

Family configurations change because individuals vary over time from one focus to another one.

Overall, sociologists have tried to understand how families change by the virtue of normative and non-

normative events. A more straightforward explanation of changes can be drawn from the fact that

individual lives are lead by shifting personal concerns for things, activities or persons. For some

individuals, work plays a major role and everything for a time is organized in their family life around

it. For other individuals, a child, a mother, the home, a leisure or some other activity, space or person,

becomes the main principle organizing family life. These focus points make individuals interact with

peculiar people and therefore create new family configurations in which they lead them to be included.

A conceptualization of this phenomena can be drawn from the work of North-American

sociologist, Feld (1981), with the concept of focus points. Recall that, according to Feld, social

relations develop as a consequence of joint activities. Interactions among individuals are organized

around specific meeting points such as workplaces, hobbies, hangouts, roles, etc. “As a consequence

of interactions associated with their joint activities, individuals whose activities are organized around

the same focus will tend to become interpersonally tied and form a cluster” (Feld, 1981, p. 1016). A

focus is defined as “a social, psychological, legal, or physical entity around which joint activities are

organized” (Feld, 1981, p. 1015). Family interdependencies come into being because some

individuals care for similar persons or for similar activities. In other words, individuals become

interdependent with each other in a family configuration because they share a focus point.

Consider the courtship process. Courtship goes through a series of stages by which individuals

develop common interests and activities and a shared culture that designates to them a series of objects

that need to be jointly invested. If the couple fails to create such joint concerns, the likelihood that
120
partners will fail to adjust to each other and to their family configurations is great (Berger & Kellner,

1964; Lewis, 1973). Moving together in a household creates a focus called “home”, a place with its

own values and spirit, which the two partners invest emotionally. When one has a child, the child

becomes one organizing principle of relationships in one's family configuration. Grandparents and

parents come together in order to provide a sense of identity, comfort and support to children.

Alternatively, a family configuration may split into several pieces when the focus couple divorces.

Because individuals are lead by the circumstances to care for similar objects or persons, they become

interdependent. Marriage, birth, health problems, residential moves, divorces, or the death of family

members are associated with the destruction of older focus points and the development of new ones.

Focus points are generative of interdependencies in family configurations 79.

Adaptation Within Stability

Numerous events influence the daily organization of family configurations. Family celebrations,

special meals, holidays and daily nuisances, problems or successes at work all have an influence on

the ways in which emotional and practical interdependencies are set up by shifting the focus points of

individuals. Persons are the object of a large number of events that change their focus points and thus

their need for others. Therefore, configurations are never static but evolve on the short term 80. Does

that mean that they have no stability?

In order to answer this question, a sample of college students filled the Family Network Method

(Widmer, 1999b; Widmer & La Farga, 2000) two times separated by a month interval (Monney,

2007). Overall, their family configurations are very stable. Ninety percent of the family members

included on either waves are present in both. Six percent were dropped from wave one to wave two,

and four percent were added on wave two. These figures indicate that there is a great stability in the

composition of family configurations within this period of time. Individuals who include their fathers

or siblings on wave one do so again on wave two. Interestingly, the family members who change the

79
Life in late modernity is a constant generator and destructor of focus points. New organizing life principles
frequently come into being in individual lives due to the complexity of the social fields in which individuals are
active. Indeed, the variety of social fields in which individuals participate is high in late modernity.
80
This does not mean that configurations are static. As the systemic perspective asserts, they oscillate around an
equilibrium from which they never go too far away.
121
most are those included last in family configurations. Therefore, they correspond to family members

who do not have a large functional importance. The more central family members, who are included

first in their configurations, are especially stable. Mothers and fathers for instance do not change

between the two waves81. This stability concerns social capital as well. The density, the connectivity of

family configurations and the centrality of respondents in them remain fairly stable between wave one

and wave two.

Figure 15 shows the family configurations of two young adults and how they change within a

month and a half. The composition of the family configuration remains for the most part the same,

with a large number of family members included in both waves. The mother, the father and the two

brothers occupy the first ranks of inclusion. They are followed by uncles, aunts, the cousins and

grandparents. The configuration is very dense and fully interconnected while the respondent does not

have an intermediary position in it. Overall, the pattern at the first interview (time one) is typical of a

beanpole family configuration82.

INSERT PICTURE 7.2 HERE

Figure 15. Change in family configurations over a month and a half

The composition of this family configuration at time two (second interview) is similar to the one

of time one. All family members of time one have remained in the configuration, and two new family

members were added, the respondent's godfather (a great-uncle on the father's side) and the

respondent's godmother (his partner). The configuration at time two is again dense and connected,

while the respondent fulfils a similar position as in the previous month. The family configuration of a

second individual, which graph is not included, is also extremely stable. It is classified by cluster

analysis as a post-divorce family because it includes a half-sister and a half-brother. Two half-siblings,

the two grandmothers, the father and the mother remain and the same kind of tightly and dense
81
There is a significant number of individuals who do not include their fathers as significant family members.
They do not change from one month to the other. Not including a father or a brother in the family is something
that has huge emotional and symbolic consequences and that is not done lightly. Mothers are the most stable (at
least at this life stage) as they are almost always included.
82
See chapter 1.
122
network of interdependencies is present in both times. Only the half-brother's partner is missing at

time two. The configuration is however not fully identical on times one and two. As a matter of fact,

some relationships have changed. The presence of the godparents, without altering the family

configuration and its interdependencies, have twisted the balance towards the relatives of the

respondent's father. This marginal change is of course not without explanation, as it responds to family

gatherings such as birthdays or Christmas, which are times in which relationships with godparents are

reactivated.

Most of the cases that we consider correspond to these rather stable patterns of relationships

across time. There are however a few cases where more dramatic changes occurred in family

configurations on the short term. Consider the second case in Figure 15. The respondent put an end to

her relationship with her partner between times one and two. The partner of time one therefore is not

included as a significant family member on time two; meanwhile, the respondent has reinvested in

interdependencies with her aunts and uncles on the paternal side, as well as in her interdependencies

with her grandmother. Note that her partner did not develop any interdependencies with her other

family members on time one; ending this partnership was hence not a too difficult matter as far as the

whole family configuration was concerned. The loss of the partnership lead to a reinforcement of its

overall logic already present on wave one, one typical of a father-oriented family configuration 83. Note

that what these cases exemplify is confirmed by further statistical analysis (Widmer & Sapin, 2008):

There is much stability in family configurations on the short term, although there are adaptations to

events in a significant number of cases.

Structural Instability

The question arises whether individuals with psychiatric problems benefit from the same stability

in their family configurations. The follow-up of the family configuration of Joanna in Figure 16

exemplifies the issue of stability of family configurations on the short term for individuals with

psychiatric problems (Widmer, Chevalier & Dumas, 2005). Joanna is a twice divorced woman in her

83
See chapter 2 for a description of the composition of family configurations.
123
forties, with two children under age ten of two distinct fathers. Her children are living in a community

home and are also undergoing psychotherapy. Joanna has recently started a new partnership with a

man that she met while doing a psychotropic treatment. Joanna' s parents live in a neighbouring

country but her sister resides in a close-by city. Several family members of Joanna are also taken care

off in various psychiatric treatments, such as her mother and her sister. Aside, she does not have the

custody of her children, who have lived in a special facility for a long time.

Joanna has been followed by the partner psychotherapist for one year when she is first

interviewed on her family configuration. She is under the supervision of a trustee following an

hospitalization in psychiatric settings for alcohol consumption and treatment of addictions, associated

with a long-term psychiatric follow-up for borderline troubles. On month one, Joanna is

institutionalized. She is in conflict with her psychotherapists and her trustee, and has severed her ties

with her parents because she thinks that she is systematically criticized by them. The family

configuration on month one is sparse and disconnected. It is composed by her parents, her sister and

her grandmother. Strikingly, Joanna is not including her two children as significant family members at

that time because she is feeling estranged from them. She is perceiving no supportive relationships

among her family members, which makes her family configuration utterly distinct from configurations

of individuals without a clinical record. On month three, however, things are changing; the family

configuration is now including her two children and the new partner that she met while she was

institutionalized. Various supportive links have developed in comparison with month one. A new

configuration has emerged from several concrete changes that have occured during that period. First,

Joanna has successfully built a trusting relationship with her new partner and is thus wishing to rebuild

a nuclear family comprising her two children. The links with her parents and her sister are reactivated

by the hospitalization of her grand-mother, which has given the family members a chance to meet

regularly for a while. Interestingly, Joanna is using this focus to introduce her new partner to her

parents. As a whole, this is a period of construction of new ties and a full reactivation of old ties,

which is confirmed on month five when her father and her partner meet each other while taking care of

removing the belongings of the grandmother from her apartment. On month five, the family members
124
are told about the cancer of Joanna's sister. This new focus is producing a dramatic change in the

family configuration, which is turning its resources on this sister. In reaction, Joanna is seeking

comfort and attention from her partner and her partner's mother. From the graph on month five, one

can see that the grandmother has lost her centrality in the configuration. Joanna is in a phase of

transition between her focus on her parents and her sister, and the focus on her partner and her

children.

INSERT PICTURE 7.3 HERE

Figure 16. Change in the family configuration of Joanna within a year

The following months confirm the attempt by Joanna of reconstructing her family configuration.

Professional caregivers and her parents and sisters are first crucial, but slowly are giving way to a

greater centrality of her children and of her partner. In summary, tracing the family configuration

throughout therapy reveals a steady process of change of the focus points of Joanna, from a lack of

focus other than herself, to a focus on blood relatives and psychotherapists, up to a focus on her

children and her partner. In this short period of time of a year, various events have occurred that have

changed the ways in which interdependencies are shaped. Meeting with a new partner, having her

grandmother move to a nursing home and her sister getting cancer, and being allowed to have her

children back with her are crucial moments of a rather swift transition between two types of family

configurations.

A second case study presented in Figure 17 concerns Beatrice, a woman of 25 years old. She

comes from a socially disadvantaged family, and has been under psychotherapeutic supervision since

childhood, due to borderline personality disorders. She has gone through a variety of problems such as

alcoholism, suicide attempts and regressive behaviours. At the time of the first interview, she is under

the supervision of a guardian because she has been implicated on several occasions in petty thief and

credit debts. Her younger sister has also suffered from personality troubles with a history of

depressions. In the first three months of the follow-up, Beatrice is living in her own apartment, after a

125
four month period where she lived in a community home, after a suicide attempt. The graphs show

that she has a very small and disconnected family configuration, which reveals her difficulty of

making new interdependencies and recovering old ones. She is trying to involve her psychiatrist and

her guardian in her family configuration and is asking them to spend week-ends together, something

that the two professionals consistently refuse to do. The family configuration of Beatrice is stable

during that time period, with its lack of relational resources except for the support provided by the two

professionals.

INSERT PICTURE 7.4 HERE

Figure 17. Change in the family configuration of Beatrice within a year

On month four, Beatrice is going to a camp organized by a cult. She is putting much distance with

her guardian, informing her that she is now feeling very well and does not need to be followed in

therapy as often as before. The family configuration of Beatrice on month five is in complete

contradiction with her family from the previous months. From months five to month seven, she has a

very large family configuration with many meaningful relationships. In contrast to previous times,

these relationships are reciprocal, as she is perceiving herself as giving and receiving support rather

than just receiving support and being dependent on others. Interdependencies are effective, warm and

reassuring. Interestingly, all members of her family configuration are well interconnected with each

other. Her new family configuration is a typical example of bonding social capital, with its normative

control and collective support.

Although this might be interpreted positively, the statuses of family members on month five

indeed show that the composition of family configurations is of prime importance to make sense of

their relational properties: the professionals and the blood relatives have been left out and replaced by

members of the cult as family members 84. Members of the cult are trying to disconnect her from her

other family members, and in particular to isolate her from her blood ties. They are overall successful:

Only Beatrice's sister and her aunt keep on being included as significant family members on the

84
Sectarian communities are indeed frequently described as large families that especially welcome fragile
individuals.
126
following months. However, they are not supportive any more. Therefore, Beatrice is falling into an

emotional dependency to the cult's members without any relational alternative. On month eight,

although the composition of her family configuration remains about the same as in the previous

months, the relational pattern of the family configuration is dramatically changing again. At that time,

Beatrice is feeling highly disillusioned: she is again in a state of unilateral dependency on her family

members of the cult, as the reciprocity of relationships has disappeared and the interdependencies

existing among family members are negated.

These two case studies exemplify the high rate of change of family configurations of individuals

with psychiatric problems, which is confirmed by other analyses (Widmer & Sapin, 2008). The

turnover of family members, as well as the ups and downs of social capital are central dimensions

associated with psychological problems. The follow-up of Joanna nevertheless illustrates increasing

positive interdependencies with family members, as a cumulative process of revival of old ties and of

creation of new ties. Reciprocal interdependencies slowly build up and are confirmed. The

acknowledgement that they should be invested in order for them to endure is acquired by Joanna. In

the case of Beatrice, quite to the contrary, the pile-up of interdependencies do not last and the

stabilization of the family configuration is not secured. The pattern of changes triggered by the

inclusion of Beatrice in the cult as a substitute family ends up being a repetition of what happened

previously in her life. After a short period of time in the cult, Beatrice is feeling utterly dependent

from her new family members, disregards her own role in the configuration and is blind to any

connection beyond herself. She is back on the same kind of relational patterns that she experienced

previously with her father, her mother and her sister.

Family configurations of individuals with psychiatric problems are more unstable than others,

changing in a rapid pace or, in other cases, being frozen for years 85. Their sensitivity to life events

occurring either to them, to their parents or siblings is great. Therefore, stability within adaptation is

less granted in those family configurations than in others. Individuals with psychiatric problems
85
We have focused on two accounts of unstable family configurations because this pattern is the most
widespread in the data that we collected. A few cases, however, present exactly the opposite tendency, with small
configurations, mostly Nuclear or Beanpole, with bonding social capital, which staid the same during the year
and a half in which they were studied.
127
develop a distinct pattern of change over time, as they face much more frequently non-normative

events in their life. Sudden changes of focus points are more frequent for them than for others. Their

capacity of keeping friends and partners for a long time is lower. One additional reason for the greater

occurrence of changes in such circumstances in the homophily structuring interpersonal relationships.

Overall, individuals tend to interact with others similar to them in terms of social class, education

level, sex and ethnicity (McPherson, Smith-Lovin & Cook, 2001). As being a “patient” constitutes a

master status (Hughes, 1945) in many social settings, individuals with psychiatric problems tend to

interact with individuals with similar problems, who have experienced identical non-normative events,

and therefore a high turnover of focus points. Therefore, homophily has a multiplicative effect on

instability because not only respondents but also the focus points of their family members, are often

unstable.

Conclusion
Changes in family configurations closely match events that alter one's focus in life. Events change

the shape of family configurations by imposing new focus points on individuals. They concern a wide

range of family members; therefore, the potential of change in family configurations is great. For

instance, a former spouse who is dating with another person, becomes less significant and is dropped

from the family configuration; or, contrastingly, individuals may get involved in a new intimate

relationship and thus they may rebuild a nuclear focus. Some individuals rediscover ties going back to

childhood, which are reactivated following some special circumstances, bad or good. The variety of

events that trigger such changes of focus does not only depend on the respondents' experiences, but

includes events that occur in the personal life of their family members as well. In this regard, family

configurations have a high potential for change.

Despite this potential, there is a great stability of family configurations on the short term. That

confirms the ability of individuals to maintain meaningful interdependencies with a large number of

family members, despite the randomness of life on the short term. One's birthday, a short-term

sickness, meeting by chance or for some purpose with this one or that one, are however occasions in

128
which marginal interdependencies are reshaped following some changes of focus points triggered by

the circumstances. In non-clinical populations, those focus points do not usually gain much

importance: they mostly explain who is included in the periphery of family configurations, that is last

in the list of family members86. Less functionally significant family members are more likely to be

added or dropped. Family configurations adjust in their margins to those changes, while their core

interdependencies remain the same.

Individuals with psychological problems obviously experience a distinct type of family changes

on the short term, as many of them develop a high turning rate of their family relationships. A variety

of non-normative events, which happen to them or to their family members, modify their family

interdependencies. Falling into drug or depression, being abruptly left by a partner, meeting

individuals with similar problems as theirs or having to deal with decisions taken by a variety of

professionals modify their focus points. Because individuals with psychiatric problems are exposed

to a larger number of non-normative events, and are more sensitive and less efficient than others in

their response to these events, many of their family configurations are turned upside down regularly.

This is likely to create further difficulties for them, as circles of cumulative disadvantages occur

between family configurations and psychiatric troubles. In some circumstances, it may be regarded as

functional from the point of view of individuals with psychiatric problems as the alternatives are even

less liable87.

Family configurations are ever changing, some slowly, other rapidly, a minority frantically.

Neither fully predictable nor random, family configurations are patterned by the occurrence of events

that create and destroy focus points. Because such events are not randomly distributed, they impact on

some individuals more than on others. As family configurations are constituted by individuals who are

86
Note that these family members are still significant, as they are the basis on which more central
relationships, such as the conjugal or the parental relationships, develop. The importance of
acquaintances and weak ties for families and individuals were rightly underlined by British
sociologist David Morgan (2008).
87
One alternative is, for some individuals, to fully withdraw from their family configurations and keep
the interdependencies with family members as minimal as they might be. This self-exclusion of family
configurations is certainly costly as it means a lack of bridging and bonding social capital, a situation
not at all uncommon as we saw in chapter six.
129
exposed to similar events, and that respond to those events in quite similar ways, there are cumulative

effects of events on family configurations.

130
 CHAPTER 8 – Family Trajectories

Following a secular trend, family trajectories achieved a high level of uniformity by the nineteen

sixties in most Western countries. At that time, a large majority of individuals went through an

identified set of ordered and age-graded family stages with very few of them getting out of sequence

or skipping transitions (Kohli, 1986; Modell, Furstenberg & Hershberg, 1976). The move towards

standardization of life courses was replaced in the late nineteen sixties by an inverse tendency leading

to a pluralization of family trajectories. This trend toward a greater diversity of family pathways is one

of the most profound changes of societies in late modernity. Family stages and family transitions have

characterized a smaller part of the population and have occurred at increasingly dispersed

chronological ages in younger cohorts than in older cohorts (Brükner & Mayer, 2004).

This result is often interpreted as stemming from the larger autonomy and agency left to

individuals about how to proceed with their own life in late modernity (Beck, 1986: Beck & Beck-

Gernsheim, 1994; Sennett, 1998). Is it really so? Recently, the hypothesis of the pluralization of life

courses was critically examined in various empirical analyses (Brükner & Mayer, 2004; Elzinga &

Liefbroer, 2007), which revealed that this trend has been less pervasive than its supporters claimed.

Important national differences in levels of pluralization were found, depending on historical and social

policy continuities, with the contrast in union formation between Mediterranean and Northern Europe.

The interpretation of pluralization as a sign of the greater freedom of individuals to master their own

life in new cohorts was questioned by various results stressing the great social and gender inequalities

that shape pluralization processes: women and individuals from the working class have experienced

much more constrains on that matter than men and professionals (Widmer & Ritschard, 2009; Widmer

et al., 2003). Overall, the empirical evidence suggest that rather than being a general trend that has

concerned all individuals and all life domains uniformly, the pluralization of the life course has taken

distinct shapes and has followed distinct paces in various social groups.

If the spread of pluralization processes needs more empirical work before being rightly estimated,
131
it is beyond reasonable doubts that family trajectories have achieved a higher complexity in late

modernity. Various demographic changes that have occurred since the nineteen sixties cast doubt on

the universal ordering and sequencing of family transitions. The tight set of transitions prevalent then,

from leaving the parental home and becoming financially autonomous to marrying and becoming a

parent (Modell et al. 1976), was progressively replaced by a set of less chronologized and less

sequenced life changes. Young adults in most Western countries have postponed leaving the parental

home, marriage, and parenthood (Lesthaeghe, 1995; Shanahan, 2000), with various complex living

arrangements characterizing this prolonged transitional stage in younger cohorts. Rising rates of

divorce have triggered a large pluralization of the second part of adult life. In their forties, many

individuals enter for a second time an establishment phase when they remarry, or a preschool stage if

they have a child from a second partnership. Accordingly, the variance of age at key transitions of

family life has increased. In summary, family trajectories have become more complex and diverse

since the nineteen sixties (Aldous, 1996; Mattesisch, & Hill, 1987).

Why do should we care about family trajectories? Based on results of chapters two and five,

showing the connections between family configurations and family structures, one expects that family

configurations directly depend on the life trajectories of their members. Therefore, the pluralization of

life trajectories may explain the diversity of family configurations in late modernity. So far, we have

considered how family configurations evolve on the short term. But they may also result from

sequences of events and transitions on the long run of a life time. This chapter first reviews the

evidence existing about the pluralization of family trajectories. It stresses the shortcomings of the

developmental approach of families, which played a leading role in family research for years before

being abandoned by most researchers. The configurational perspective offers an alternative to the

developmental approach which, while keeping the focus on orderly family changes in the life course,

does not assume that family trajectories stick with a standard model. Empirical analysis, rather than a

highly abstract and normative model, should be given prominence in the understanding of the

development of families. Various innovative statistical techniques (Gabadinho, Ritschard, Studer &

Müller, 2008; Gauthier, Widmer, Bucher & Notredame, 2009) shall be used in order to uncover what
132
types of family trajectories best describe individuals from non-clinical and clinical samples. Finally,

the importance of the results for a deeper understanding of long-term family changes shall be stressed.

The Family Life Cycle Revisited

Current research is left with various evidence that point at the greater diversity and complexity of

family trajectories in younger cohorts. The developmental model has made the understanding of

family trajectories an uneasy task. Up to the early seventies, sociologists and demographers stressed

the prevalence of a universal family cycle featuring a set of pervasive stages. These stages were

ordered, associated with the chronological age of their members, specialized functionally and distinct

for the composition of the household unit. All families were supposed to go through a beginning

stage, a childbearing stage, a preschool stage, a stage with adolescents, soon to be followed by the

launching of young adults and a post-parental stage (Duval, 1957; Hill & Rodgers, 1964). The family

life cycle ended with the so-called “later year”, “ageing” and “retired” families.

Accordingly, research dealing with family changes throughout the life course has been based on

the constitution of family stages a priori, using criteria such as the age of the oldest child and the

retirement status of the husband father (Aldous, 1996). These criteria raise a number of empirical and

analytical difficulties. Between 1965 and 2004, the "Sociological abstract" database reveals that more

than 1000 publications referred explicitly to the family developmental perspective. The number of

publications has however greatly decreased since then. This underlines the shortcomings of the model

for investigating the complexity of family life courses in late modernity. As a matter of fact, a major

weakness of the developmental model is that it focuses on nuclear families and disregards other family

configurations.

The gap between the theoretical model of development and the empirical approach to studying

empirical family trajectories has been recognized (Oppenheimer & Lewis, 1999). The developmental

model has been criticized for its assumption of the universality of the family stages, which impedes

the inclusion of alternate configurations to the nuclear family, such as cohabitation outside wedlock,

childlessness, divorce and remarriage (Laszloffy, 2002). In other words, the family developmental

133
model is not able to account for critical social changes that have occurred in the family realm within

the last 30 years. Indeed, the universality of family stages, their ordering as well as their link with

chronological ages have been seriously questioned and, with them, the idea that most families follow a

regular and ordered life circle (Furstenberg & Spanier, 1987).

Several critics were particularly hard to deal with by the family developmental perspective. The

number of household structures to take into account when dealing with the family life cycle has

dramatically increased even if only a small number of criteria are considered, such as the sex of the

co-resident parent and the presence of a stepparent , of half-siblings or step-siblings in the household

(Pasley, 1987). The ordering of stages is also misleading, as an individual may enter a second time an

establishment phase when he remarries, or a preschool stage if he has a child born from a second

partnership. Although a series of attempts have tried to make the developmental perspective more

differentiated in the number and the characteristics of stages (Glick, 1947; Mederer & Hill, 1983;

Duvall & Miller, 1985; Mattessisch & Hill, 1987), one is still wondering in what family stage does this

individual really stand. The lack of a strong correlation between transitions such as divorce and

remarriage with the chronological ages of both parents and children is another evidence questioning

the reality of a developmental family cycle. The family stages proposed by the developmental

perspective are neither ordered nor chronologized, and the model is hardly “developmental”. This

makes this perspective unable to account for a large share of the variance of actual living

arrangements and other family interdependencies over time. The emphasis on a single and typical

family life cycle model assumes that all families abide strongly by normative constraints and therefore

do not follow alternative pathways. Referring to such a model forces the researcher to exclude many

individuals or family configurations, such as widowed, divorced, or separated individuals, as they

simply do not fit within the formulation.

The ambiguity and limitations of the family life cycle model and its developmental perspective

lead to serious analytical problems, that have cramped family sociology for a long while. However, it

is, despite its various shortcomings, the only perspective that seriously considered families in relation

with time, structures and functions. Sadly, if the majority of family and life course sociologists have
134
acknowledged the increasing complexity of family trajectories, models that account for change of

families over the life course are deceptively few. Is there any alternative to the view of the family life

course as a standardized, ordered and chronologized development of stages?

Empirical Family Trajectories

The pioneer work of American sociologist Andrew Abbott (2001) opened a new analytical

perspective for the understanding of family development in late modernity. In a series of papers

unrelated to family sociology, he advised researchers to study empirical sequences of events rather

than to define stages, transitions and the causality that link them on theoretical grounds. Abbott

stressed that the percentages of cases that empirically, in a longitudinal perspective, follow any

predictive model, is likely to be small in a variety of situations. Rather than trying to see how to make

the empirical data fit within the theoretical model of the family life cycle, why not using statistical

techniques that reveal their patterns? One may later see how those patterns relate with criteria such as

the ordering of stages, and link them with the chronological ages of family members.

The alternative inspired by Abbott's stance responds to critics raised against the family

development perspective. It implies a change of focus from the family considered as a bounded small

group, to individuals as centres of large sets of family interdependencies evolving through time. This

fits well within the configurational perspective, which focuses on individuals and their interdependent

family members. The configurational perspective does not understand families as comprehensive holes

with a life of their own, as it does not believe that families are living organisms that follow a finite set

of stages in their development. Individuals, however, experience various family configurations during

their life that are to some extent developmental: these configurations are indeed structured, functional,

chronologized and ordered in a variety of ways. In other words, the configurational perspective

stresses that referring to individuals and their changing family interdependencies over time rather than

to families as developing units provides a deeper understanding of the family life course.

In a series of analyses on the biographical survey of the Swiss Household Panel (Gauthier,

2007; Gauthier et al., 2009; Widmer et al., 2003; Widmer & Ritschard, 2009), we have encapsulated

135
family interdependencies over time by focusing on whom individuals have lived with between age 20

and age 4588. The cohabitational trajectory describes the composition of the interviewed person’s

household with ten categories: living with both biological parents, with one biological parent only,

with one biological parent and her or his partner, alone, with a partner, with a partner and a child, with

a partner and a non-biological child, alone with a biological child, with friends, and other. The

trajectory of each individual is described by a sequence of states such that each state corresponds to

the age of the person expressed in number of years. The time during which the person stays in each

state is thus accounted for. Five types of cohabitational trajectories that best describe the variety of

existing alternatives were drawn from innovative sequence analysis (Gabadino et al., 2008; Gauthier et

al., 2008). Figure 18 presents a series of bar-charts by trajectory types. The bars represent the

distribution of response categories by chronological age. For instance, about 70 percent of individuals

in the parental type live with their two biological parents at age 20, and only two percent at age 30.

INSERT PICTURE 8.1 HERE

Figure 18. Types of family trajectories

The parental type is overly dominant (62 percent of the sample). It is in line with the

developmental model of the family, as it features an ordered sequence of stages from leaving the

parental home to creating a couple and having children. A second type (17 percent of the sample)

includes trajectories centred on partnership. Individuals belonging to this type have spent most of their

adult life (19.2 years over the 25 years considered) living with a partner, but without children at home.

The three remaining types are quantitatively less prominent. Seven percent of individuals have always

lived with their parents until age 45 (parental home). Another type (8 percent of cases) includes a

variety of sequences stemming from life as a single, living with a partner without children, living with

a partner and the partner’s children or without a partner but with biological children (mixed

cohabitation trajectories). Finally, five percent of individuals have not yet formed a stable cohabitation

88
This is obviously a crude and imperfect measure of family interdependencies throughout the life
course, as coresidence is a poor criterion to define family configurations, as we have noted earlier.
Using household membership is certainly not sufficient to measure interdependencies, especially
when emotional and cognitive interdependencies are considered of prime importance.
136
with a partner during their adult life.

Overall, there is a limited number of types that account for the diversity of individual

trajectories. Interestingly, these types depend to a significant extent on the birth cohort that individuals

belong to, as well as on their level of education. As a matter of fact, individuals of younger cohorts

and those with a higher level of education are under-represented in the parental type, and

overrepresented in the mixed and single cohabitation trajectories. They have a greater likelihood of

developing a slow transition to parenthood than individuals of older cohorts. Individuals in late

modernity go through a greater number of family configurations and focus points throughout their life

than individuals in previous times. The transition to parenthood is for instance extended and associated

with a larger number of partnerships. The variety of family trajectories have indeed been greater in

recent decades, and younger generations have been the main recipients of this trend towards

pluralization. Rather than belonging to a single type of family trajectories, individuals in late

modernity follow a variety of life paths. This plurality is however bounded, as the number of trajectory

types is limited.

Life Trajectories and Family Configurations

This pluralization of trajectories have had important consequences for family configurations.

In the previous chapter, we stressed the impact of events on short-term changes of family

configurations. Because events impose new focus points on individuals, they promote a reorientation

of family configurations, which become frantic for some individuals with psychiatric problems. What

is true for events is even more true for life trajectories and the development of family configurations

on the long run. Having a long-time partner and children or developing more temporary relationships

and remaining childless change one's family configuration. The multiplication of focus points in the

family trajectories of younger cohorts account for the complexity of family configurations. Family

configurations are indeed marked by the relational history of individuals. Consider again the two

family configurations stemming from divorce and remarriage presented in chapter 5. Alina is this 35

years old woman with a ten year old son from a previous marriage. Her partner, Laurent, has two sons

137
from a previous marriage. Her family configuration includes her ten year old son, her current partner

and his two sons (10 and 20 years old), her previous partner, her mother, the mother of her previous

partner, her father, her sister and three friends. The composition of her configuration is therefore very

heterogeneous, with a mix of blood relatives, friends, steprelatives and previous in-laws. As we saw in

a previous chapter, her former spouse plays a central role in her family configuration, with a high

level of structural ambivalence which makes family life quite uneasy.

This peculiar family composition to a great extent is a consequence of Alina's relational history

as well as her current partner's history. Alina left her parents' home at age 22 to live with one of the

female friend that she currently includes as a family member. Soon after, she met the man with whom

she had her only child at age 23. She lived with him during two years and they split the same year in

which she delivered her child. After that, she lived with her son alone for ten years. During that time,

she only had one other intimate relationship that lasted between age 25 and 27. She met her current

partner at age 31 and has lived with him ever since. Of the new partner, we know little except that he

has two sons with a great age difference, from two distinct partnerships. His youngest son is partly

living with him and Alina; therefore strong dependencies have developed between her and this son,

which is not the case for the oldest son. On the other side of her family, the long period of time spent

by Alina has a single mother has created enduring interdependencies with the father of her child and

his parents, especially his mother. Such tendency of single mothers without a new partnership to

remain dependent on their in-laws is not uncommon. The extended period spent by Alina without a

stable partner explains to some extent why the emotional and cognitive dependency on the previous

partner has remained so strong. Ten years after the divorce, he is still included in the family

configuration and used as an emotional resource by Alina. A sediment of the past has remained alive

because of her focuss on her child and on her mourning of this intimate relationship. This sediment is

mixed with her current partner's own complicated relational history in which dependencies on his own

two children create a series of difficulties for the current couple. Indeed, Alina acknowledges many

communication problems occuring in her family configuration and difficulties for her to establish a

trusting relationship with her co-resident stepson. The various tensions that are experienced in the
138
family configurations obviously stem from the difficulty of making personal histories of

interdependencies work together.

Now consider the second post-divorce family configuration that we reviewed in chapter 5, that of

Dora, this 36 woman, first time pregnant when she was interviewed. She has been with Don since she

was 28. Remember that Don has two daughters, eleven and nine years old, for whom he has a shared

custody with his previous partner. He has been divorced from her for eight years at the time of the

interview. Dora has had a very distinct intimate trajectory than that of Alina. She did not develop a

partnership before age 25. Later on, she had three short-term affairs, each lasting about six months and

none involving a cohabitation. Indeed, she has kept on living with her parents until she met Don at age

28. Unsurprisingly, her family configuration focuses on her parents and her sister, in addition to her

current partner and his two daughters who live with them. Contrary to the previous case, she is not

connected with her previous partners, who were mainly casual encounters. The development of a

strong connection with her sister, whom she includes among the first family members as a source of

support for her, lead to the inclusion of her sister's partner and her sister's daughter in the family

configuration. Overall, this family configuration, although, stemming from divorce and remarriage, is

kinship oriented, because Dora's relational history is one of strong and lasting interdependencies with

her parents and her sister. The birth of her child to come shall undoubtedly reinforce this orientation

towards her own kin. On the other side of her family, her partner has the custody of his daughters and

Dora has lived with them since they were two and four. This also supports the conclusion that a set of

interdependencies have gradually emerged linking Dora, her partner and his children, and her parents

and sister. The focus of Dora for a number of her adult years on her parents and her sister has been the

basis on which her couple relationship has developed.

The various ways in which individuals build up their family configurations depend on a large

series of events, transitions and stages that have happened in their life course. Some focus on the

maintenance of meaningful ties with their previous partners, often for the sake of children. Some

reinvest friendship and kinship ties in attempts to compensate for the permanent or temporary

139
weakening of their partnership. Others try to build a new nuclear family and focus on their current

partnership and the children that they are responsible for.

Family Trajectories of Individuals with Psychiatric Problems

What kind of family trajectories individuals with psychiatric problems develop? Are they more

unstable than those of other individuals? The previous chapter revealed a much larger turnover of

family members on the short term in such cases, which have implications for family configurations on

the long run. In their summation article, Cook et al. (1997) emphasized the distinct experiences of

psychiatric frailty that individuals have developed depending on their birth cohorts. Since the nineteen

sixties, there have been great changes in the ways in which individuals with psychiatric problems have

been taken care off by society. Their concentration in ``total institutions'' that sought to control all

aspects of their life for long periods of time, which dominated the social spectrum until the nineteen

sixties, gave way to a ``desinstitutionalization'' of large numbers of individuals with psychiatric

problems, who came to reside in the community (Cook et al., 1997). The spread of psychotropic

medication reinforced this trend by making it possible for large groups of individuals to resume life

with their family members and regular social participation. This process was however not without

concerns, as many individuals without strong family support and financial resources were left by

themselves in dealing with the consequences of their problems.

Changes across cohorts in the ordering and chronologization of life sequences have consequences

for trajectories of individuals with psychiatric problems. The weakening institutional constraints

imposed on individuals with psychiatric problems as well as the decreasing impact of social norms

relative to the chronologization of key life transitions such as leaving the parental home, becoming

economically independent or becoming parent, may have lead to a higher complexity and a

pluralization of the life trajectories of individuals with psychiatric problems. The overall trend towards

a pluralization of the life course may explain why clinical populations have also achieved a great

variability of family configurations in late modernity as chapter six showed.

140
 In order to further estimate whether or not individuals with a clinical record follow similar types

of family trajectories, we considered 100 individuals with severe psychiatric problems 89. The average

family trajectory in this study is distinct from that found in non-clinical samples. Indeed, the

normative sequencing of transitions characterizing life courses in non-clinical samples is significantly

lower in this sample, with many individuals going back and forth between family stages that are

usually considered irreversible. Also, for many individuals, some key transitions, such as leaving the

parental nest, forming a stable partnership or having children, did not occur. The independence

provided by having a separate household is in some cases not secured, as individuals with psychiatric

problems continue to live with their parents or reside in public facilities. Overall, when compared with

individuals from their cohorts, individuals with psychiatric problems have distinct trajectories.

Does that mean that all individuals belonging to clinical samples share the same family

trajectories? By using advanced statistical techniques dedicated to the study of sequence data, such as

optimal matching (Gabadino et al., 2008; Gauthier et al., 2008), we found several types of family

trajectories, which are similar in some respects to the types found for the non-clinical sample, but

distinct in other respects. A first type of trajectories is composed mostly of individuals who grew up

with both of their biological parents before living alone (17 percent of all individuals). In a second

type (16 percent), individuals who have spent most of their childhood and adolescence with one

biological parent only, before living alone as individuals with the first type of trajectories. The third

type (35 percent) features individuals who left the parental home much earlier than in other types to

live in institutional settings for the most part of their lives. The fourth type includes individuals who

have lived a long time with a partner but have never experienced parenthood. Only a small minority of

individuals (12 percent) belong to the standard type of developmental family trajectories , with a long

lasting partnership and the presence of biological children. Overall, individuals with psychiatric

problems, as other individuals, present a diversity of family trajectories. Therefore, averaging out

these cases in a single logic of development would provide poor estimates of their family trajectories.

89
The data are again taken from the two studies on individuals followed in a private psychiatric facility and in a
day care hospital (Widmer, Chevalier & Dumas, 2005; Widmer et al., 2008 ).
141
However, this diversity again, as in non-clinical samples, does not equate to a full individualization of

life courses. Indeed, a limited number of trajectory types account for it.

This pluralization of life courses has had an effect on individuals in clinical samples. It explains

why individuals with psychiatric problems develop a variety of family configurations. Family

trajectories that stems from an early family recomposition are associated with family configurations in

which friends and therapists, or relatives, are central. Family trajectories characterized by living as a

single lead to a family configuration in which parents are dominant. Family trajectories in which a

stable relationship and children are present obviously triggers a much greater likelihood of being

embedded in a nuclear family. Not only do recent events have an influence, but the overall focus of

family life has one as well. As for non-clinical samples, life trajectories quite directly translate into

family configurations. Overall, life trajectories impose long-term focus points at individuals for the

development of their family configurations: Their children, their partners, or themselves as singles

become the organizing principles of their lives, along with their psychiatric problems. A couple of

qualitative examples further illustrate the connection among life trajectories and family configurations.

Michel, a man in his sixties, moved to Switzerland from Southern Europe in his early adulthood.

He came from a disadvantaged family and had no formal education. In Switzerland, he was hired as an

unskilled labourer and has remained so for the past 40 years. At age 20, he met his wife, who

originated from the same country as him. They had two children and have remained together ever

since. Therefore, Michel's trajectory has been centred on his children, his wife and his home.

Interestingly, Michel has managed his various phases of depression by taking short interruptions from

his work while keeping his integration within his family. The maintenance of a continuity in his family

life fully translates into the organization of his family configuration: its small size, its focus on his co-

resident children and his partner, as well as its bonding social capital are directly linked with his

family trajectory.

By contrast, Gabriel, a man in his fifties, has had a highly unstable family trajectory. He has

changed partners (and jobs, usually simultaneously) every five years, while keeping in touch,

emotionally and socially, with many of his former partners, especially when they are the mothers of
142
his children. Therefore he has developed a post-divorce family configuration. Within the year and the

half in which he took part to the study, he has gone from a friendship family configuration to end up in

a kinship family configuration. Those changes are not random either, as the sudden death of one of his

ex-partners, the mother of his youngest child, created much turmoil in his life. After this death

occurred, he had to turn to his own parents and siblings to solve the problems associated with his

orphan child, and thus, his family configuration changed. Finally, the case of Bernard, a young man in

his thirties, is characteristic of a person that has never experienced the transition to a stable partnership

and to parenthood. Although he has a formal education that includes some college, he has never been

able to find a job. He was not successful in intimacy either, as severe psychotic episodes in his early

twenties made the development of an intimate relationship hardly possible. Therefore, he has focused

on his parents ever since, without any change during the time of the follow-up.

Conclusion

Family trajectories remain highly patterned and ordered in late modernity. In sum, a large

majority of family trajectories belong to identified types, even in younger cohorts. Families in late

modernity are not characterized by randomness and a lack of predictability but by the emergence of a

few additional types to the stock already present in the nineteen sixties. Using a configurational

perspective strongly supports the thesis that family trajectories are far from being disorganized or free

from structural and normative constraints. Indeed, transitions continue to shape family trajectories.

Most individuals start their lives living with their two biological parents. Then, during childhood or

adolescence, some of their parents divorce, which account for a first factor of variability. Later

transitions create additional faithful moments in which one's life may go along a few alternative

pathways. Family development, from a configurational perspective, is constituted by a limited set of

turning points, such as divorce and repartnering of parents, leaving home, moving to live with a

partner, becoming a parents, divorcing and starting a new partnership. The cumulation of such turning

points creates a bounded pluralization of life trajectories. In parallel, it makes many individuals

believe that “everything is possible” and could be different. What is realized in empirical cohabitation

143
trajectories is actually much more homogeneous than often expected. The family life cycle has

increased in complexity in younger cohorts; it remains however patterned and structured.

The paths followed in life have an importance for family configurations that individuals develop.

Family configurations keep several sediment layers in them despite time passing by: parents and

siblings come early on and they often survive as meaningful family members in older adulthood.

There are other representatives of childhood, such as uncles, aunts, cousins and, of course,

grandparents. In young adulthood, partnerships are developed, and with them, a whole new series of

relatives are acquired. This is another layer of relationships that now often comes a decade earlier than

children. The transition to parenthood involves a change of focus for intimate relationships, with

friends dropping in significance. When children, nephews and nieces become parents, new

interdependencies are created that again reconfigure family contexts. Divorce and remarriage reshuffle

family interdependencies once more, while creating another sediment of family members. Family

configurations are constituted indeed by various layers associated with the various stages of family

trajectories.

That life trajectories have become more heterogeneous during the last forty years means that a

greater variety of configurations have emerged, since additional sediments provide the ground on

which focus points develop. The diversification of life trajectories is not synonymous of lower

constraints on individuals, however. Leaving the home of ageing parents, developing a long time

partnership after a history of short term affairs, taking care of one partner's children while not being a

mother, raising a child without the other child's parent around, are indeed constraining situations that

highly structure the development of family interdependencies. Family configurations stem from the

conjunction of a variety of events that occurred to a large number of interconnected family members.

Indeed, family configurations depends on personal histories that span over several decades.

The family developmental perspective as it was developed until the nineteen eighties has drawn

many criticisms upon herself because of her inability to account for the complexity associated with the

pluralization of family lives in the last decades. The configurational perspective may do better because

144
it does not stick with a definition of families as nuclear and because it focuses on empirical trajectories

at hand rather than on normative models stating how in theory families should develop over time.

145
 CONCLUSION – Individualized Families

Ernest Watson Burgess (1960) developed already before World War II an agenda of research

dedicated to study the transformations of families, by the way of the integrative concept of the

Companionship Family. Although many scholars and the general public of his time believed that the

Family was on the verge of collapse due to rising figures of divorce, the transformation of

intergenerational relationships and the spread of juvenile delinquency, he stressed the emergence of a

new organization of family ties based on democratic arrangements, a functional specialization of

husbands and wives, a privatization of family life and a focus on the nuclear family. The intellectual

project of defining the family model of modernity was taken over and further systematized by Talcott

Parsons and Robert Bales (1956), who a bit carelessly stressed the nuclear family as the only

functional model available in modern societies, making it a norm and disregarding its alternatives. The

Family was indeed, according to them, one in which a married couple with a gendered division of

household and paid work, found the necessary emotional resources to support the stress of modern life

while raising its legitimate and resident children to fulfil the normative standards of the larger society.

As the nuclear family has lost quantitative and functional grounds during the last decades, various

scholars have announced the decline of the Family. This view surprisingly associated the decline of a

model of families with the decline of families (Bengston, 2001). Indeed, increasing numbers of well

functioning family contexts do not meet its normative expectations regarding their duration, their

boundaries and their organization. Therefore, reactions against that normative view of the Family

have raised since the nineteen seventies, supported by a series of trends at various levels,

demographic, social, political and, not least importantly, scientific. Faced with the diversity of family

configurations as well as with increasing evidence that the model of the nuclear family has a strong

normative component, family sociologists have been more and more critical to the attempt of defining

a single model of family. A symbolic decision was made in the year 2000, when the National Council

for Family Relations changed the name of the premier research journal in the field from Journal of

146
Marriage and the Family to Journal of Marriage and Family in order to acknowledge the wide

diversity of families.

The configurational perspective on families goes beyond the nuclear family model, which has

limited the understanding of family diversity, while keeping the project, proposed by Burgess (1925),

Parsons and Bales (1955), of stressing the structural and functional principles to which families

respond. New tools and new concepts were necessary to reveal some of those principles because

families had become more diverse than they were in the near past. This is not to say that they have

become less meaningful or fully individualized. The results presented throughout this book indeed

suggest that a limited set of formal principles makes it possible to account for a large number of the

variations that characterize family contexts in late modernity.

Individualization
By focusing on the nuclear family, the functional perspective imposed a normative definition of

the Family. In a time in which individuals are the stick yards of society and of families, researchers

need to let them define family members by themselves. Doing that, we astonishingly found that a few

alternative rules set the boundaries of families. Although families of late modernity do not fall within

the model of the nuclear family, neither are they the fluid and non committed relationships refered to

by some sociologists or the mass media. The diversity of demographic structures stemming from

divorce, remarriage, fertility patterns and migration made researchers assume that family life has

become fully pluralized and individualized. This was without taking into account the necessary

organization by individuals of their family interdependencies. Because individuals are limited in the

emotions, cognitions, time and money that they can invest in the development of their family ties, they

focus on a limited number of persons and follow some informal rules regarding how to deal with them,

even if they are not, in most cases, fully aware of those rules.

Indeed, family configurations do not vary to infinity. For once, the number of activated family

models is rather limited. As a matter of fact, similar types of family configurations come back over

and over again in all the samples considered in this book: Beanpole, friendship, post-divorce,

147
conjugal, kinship (either on the father or mother's sides) and sibling family configurations set the

boundaries of family configurations, with a few differences according to the life stage and the type of

samples, clinical or non-clinical, considered. Each type features a well defined logic, with an emphasis

on kinship or friendship ties, on parents, in-laws or step family members, on the women's side or on

the men's side. Sex, generation, blood, partnership or friendship, along with genealogical distance,

account for distinct family configurations. Individuals may focus on male or female members and

invest on their father's side or on their mother's side of their kinship networks. They may focus on

blood ties by maintaining strong interdependencies with their biological parents and siblings. Quite to

the contrary, they may make links with their in-laws prominent. They may choose to maintain or

develop interdependencies with steprelatives. They may develop relationship beyond the realms of

blood and marriage by investing in friendship ties defined by them as family ties; they may follow a

logic of genealogical proximity by promoting interdependencies with the closest blood or marriage

ties, or rather pick more distant relatives in genealogical terms as significant family members.

Interestingly, those configurations exhibit a remarkable consistency throughout the life course, as they

make generational shifts rather than truly change when individuals grow older 90. For instance,

individuals with a matricentric family build their configurations on a link between sisters (that is

between their mother and her sisters) whereas older adults focus on their own sisters and their brother

in law, their nephews and their nieces in a sibling configuration. The logic is the same, although the

generations involved are different because their uncles and their aunts have passed away or have lost

their functional importance.

The complexity of individualized family configurations relate to the ways in which a limited

number of informal rules are put to work. Actual family configurations are at the same time diverse

and quite homogeneous. In that sense, the attempt of defining The Family Institution characterizing

late modernity should not be regarded as impossible or dangerously normative. Indeed, if one defines

the Late Modern Family as a small collection of informal rules that are meant to be organized and

reshaped by individuals in a variety of ways, according to their resources and the constrains that shape

90
This point should be verified by the use of longitudinal data that are currently unavailable.
148
their action, one keeps at the same time the emphasis on the homogeneity of the Family and the

diversity of actual families. Let us say that much personal work and collective influence is involved in

the development of configurations in late modernity. Power issues and conflicts are at stack, as the

resources that one can invest in family life are limited and the competition among various needs is

fierce. Therefore, doing family is not an easy task, especially when discordant investments are

necessary.

The Fabric of Diversity

What mechanisms create family diversity? For many scholars, individual agency first comes to

mind as the ruling principle of such diversity: individuals, they say, are more than ever driven by

autonomous values and their own lifestyle in a less constraining social environment; therefore, they

are likely to develop fluid, ever changing family commitments. This book rather emphasized the

structural genesis of family diversity. The pluralization of life courses in various social fields since

the nineteen sixties has increased the likelihood for individuals to experience a variety of turning

points. The number of job shifts, the increasing spatial and social mobility as well as the cultural

pressure of living a worthy life create the ground on which family diversity builds up. These events are

to a large extent imposed by social, cultural and economic forces associated with world globalization.

Individuals react differently to them but they still have to cope with them, one way or another. Events

impose new focus points on which individuals develop their family configurations. Indeed, human

beings evolve in investing their resources in focus points that they share with others. When an

individual becomes a parent, a child suddenly steps on the foreground of her life as an organizing

principle of relationships in her family configuration. Grandparents and parents come together in order

to provide a sense of identity to the child. Because individuals care for similar persons or things, they

develop new focus points and hence change their set of interdependencies in key transitions of the life

course.

The diversity of contemporary families comes from the interaction between the various events

that occur in the life course of individuals linked by family interdependencies. In other words, a

149
combination of events concerning various individuals imposes a set of alternative family structures,

within which family configurations develop. The causality linking life events, family structures and

family configurations is best illustrated by post-divorce families. Indeed, there is no such thing as an

homogeneous stepfamily model. Individuals who experienced divorce and remarriage present a

diversity of configurations. Individuals reorganize their family relationships in various ways after

divorce, as they may do after migration or death of a family member; many of them develop similar

configurations as individuals who have not yet experienced these transitions. Family structures change

in response to the variety of events that create new focus points for their members. There are long term

changes associated with normative and non-normative events such as birth, marriage, death or divorce.

Daily events such as meeting someone by chance, being sick or having a family member back from

another country also have an impact on family configurations. A few of those events are enough to

produce a variety of family configurations, as they combine in trajectories in which processes of

cumulative advantages or disadvantages play out. The family path that one followed has an enforcing

effect on one's family configuration. As life trajectories have become more heterogeneous in younger

cohorts, a greater variety of configurations has emerged.

This diversity is nevertheless not synonymous of lower constraints on individuals and of a greater

freedom of action in late modernity. Finding the proper partner, having children at an older age, taking

care of them while being employed, moving together in a common residence when one is separated by

jobs in distinct places are indeed structures that limit individual agency in the development of a family

configuration, or rather shape it. Basic socio-demographic events impose serious limitations to the

ability of individuals to recompose their family lives as they wish. Keeping strong interdependencies

with parents or grandparents obviously depends on where they live and on the regularity of contacts

which has been maintained throughout the years. Developing a sibling family configuration is not

possible for an individual who migrated alone, as it is not possible for young adults whose parents are

still together to belong to a post-divorce family configuration. Family structures open up a range of

possibilities while setting up heavy structural constraints on individuals' ability to construct family that

fully correspond to their expectations.

150
Social Integration
This book started by discussing the lack of functions supposedly characterizing the Family in late

modernity. The deficit of social integration brought by individualization was stressed by various

authors supporting the thesis of family decline. In most studies reviewed in this book, there is no lack

of social capital but a variety of social integration models mixing social capital and conflict. Family

members constitute a large share, both in the United States and in the Swiss contexts, of support

networks. With the rise of modernity, as Burgess and his colleagues stressed (Burgess et al., 1960),

emotional and cognitive dimensions of family life have become central. Meaning, identity, relatedness

and commitment have in part replaced the material functions of family transfers that characterized

families in earlier historical periods. The focus on emotional interdependencies among family

members reveals that contemporary families are not characterized by a lack of commitments but rather

by several distinct patterns of interdependencies.

Using the concept of social capital, we stressed the individualized resources that individuals get in

large, unbounded and quite heterogeneous family contexts. Overall, the normative expectations

associated with family life have changed with the development of bonding and bridging social capital

in families; support imposed by institutionalized family roles is being replaced by relational structures

in which individualized resources come to play a prominent role. Even though the dyads of the nuclear

family play a key role in one’s life, they are embedded in a larger set of family relationships that

provide alternatives and backups. Relationships with parents, siblings and their partners frequently

remain active after those individuals stoped living with each other. They appear as first support

providers and are often linked in dense subgroups. Ties with grandparents, uncles, aunts, cousins, but

also friends or neighbours considered as family members constitute a larger set of usually weaker ties

that also play a role as they have an influence on the more central family dyads in their daily

functioning. Indeed, if mothers are central in one's family configuration, aunts, as sisters of mothers,

keep on being significant for mothers, and, therefore indirectly, for respondents. The transitivity

created by blood ties participates to continuously wire up family members beyond their immediately

significant others. Overall, family configurations in which a large number of blood ties exist, feature a

151
high density of both supportive and negative ties. They therefore provide a bonding type of social

capital, with much collective support but also some interference that erodes conjugal satisfaction and

parent-child relationships. Dyadic ambivalence is at its highest in such family configurations because

dyads are bound by a structure that force them to be active. In a time of individualization, being

embedded in a close group creates obvious tensions in beanpole and nuclear family configurations.

Quite distinctly, family configurations that include friends or steprelatives provide alternate forms

of social capital, in which bridging resources are present. This alternate form of social integration is

likely to have profound, and not necessarily negative, consequences. Individuals benefiting from a

high level of bridging social capital develop more structural autonomy: they are not under the

supervision of a great number of third parties. Of course, this comes with a cost, as being a bridger

requires a great personal investment in time, energy and sociability in order to create or maintain

discrepant family interdependencies. Therefore, individuals in such family configurations are less

motivated by maintaining long term strong ties with specific family members, as they have more

alternatives to consider, and less time, money and emotional energy to invest in each of them

separately. In other words, the way in which individuals experience their family lives may become

more individualized without loosing its integrative function.

Overall, scholars who emphasize a lack of long term commitment as defining families in late

modernity may have underestimated the resources that individuals draw from family members beyond

the nuclear family. It is true that conjugal dyads have become more fragile since the nineteen sixties.

But that does not prove that the Family as a whole has become less resourceful, as a variety of other

family ties than those between spouses or parents and their resident biological children, have remained

meaningful and lasting, while alternative family ties, such as those with friends considered as family

members, have developed.

We also stressed that conflict and ambivalence in family relationships should not be regarded as

proofs of family decline. Indeed, conflict frequently comes along with positive interdependencies. In

individualized families, each individual is embedded in a specific set of interdependencies. The

uniqueness of each individual position leads to a higher probability of structural ambivalence:

152
individuals are supported by family members who compete with each other for love, support, attention

and other psychological resources. This situation does not correspond to what is traditionally expected

of the Family as an homogeneous solidarity group. Indeed, individuals with bridging social capital do

not think about their families as small groups. Their experience nevertheless belongs to a kind of

social integration that is widespread in social fields such as work, leisure, or the politics, one in which

individuals have become more structurally autonomous. The fact that many families become

responsive to the overall emphasis on the individualization of social ties should not be interpreted as

ending the contribution of families to social integration. We have seen that in many families

individuals develop personal ties that often compete with each other. Giving prominence to the nuclear

family model as an homogeneously solidarity group lead scholars to over-emphasizes social

atomization and family decline. By asking individuals to define their significant family contexts, the

configurational approach proposed in this book shows that in a large majority of situations individuals

have developed an active set of emotional interdependencies with family members. This demonstrates

the key role played out by families in the realm of social integration in late modernity.

The tenants of the decline of families might however be right in a sense. There is indeed a serious

concern to be raised about the ability of some families to face the dilemmas and tensions imposed by

late modernity because of their small size and their closure. The only true loosers of social capital in

late modernity may be, paradoxically, individuals who stick with the definition of their families as

nuclear. A minority of respondents of the various samples considered in this book indeed exclusively

focus on their co-resident partner, parents or children to define their families. Interestingly, they are

over-represented in the clinical samples, in which individuals have greater difficulties to build a

variety of ties. Indeed, a large majority of individuals go beyond the nuclear family when defining

their family configurations. Those who do not do that, have, as a matter of fact, a much lower number

of family members to count on, in either extremely dense or extremely sparse configurations. They

may have to choose between a fusional understanding of family and the vacuum of frozen

relationships that do not relate to ties dating back to childhood or created by divorce and remarriage.

The nuclear family as the ultimate institution providing individuals with support is in that sense on the
153
verge of collapse. Due to the high rate of divorce and the overall difficulty of dealing with internal and

external issues within its boundaries, creating significant interdependencies with additional family

members makes parental and conjugal work much easier. Does the decline of the nuclear family mean

the decline of the Family? Various results presented in this book rather stress that if the nuclear family

is on the verge of collapse, family configurations still do well.

Rather than belonging to a zombie category that has definitely lost its integrative functions, family

configurations provide meaningful support in a variety of ways in late modernity. If the great changes

that have happened since the nineteen sixties have had significant consequences for families, they

have not destroyed the presence of a functional stable model of family life. Individualization did not

make the family group decline but rather changed it into various models of interdependencies,

characterized by support and ambivalence.

The Disadvantaged

Social integration in late modernity has a strong emotional and relational component. Therefore,

the disadvantaged are those who, for a variety of reasons, fall short of developing such ties with their

family members. Compared with others, individuals with psychiatric problems from the studies

reviewed in this book include fewer partners and partners' parents. Other family ties are also much less

frequently reported. Intergenerational interdependencies, in particular with grandparents, uncles and

aunts, but also with fathers, and the sense of continuity and the social control that they provide, are

severed. Family members are much more often disconnected from each other and respondents have a

low centrality in their own family configurations. The Family as a main support provider is less so for

them than for others.

Indeed, individuals with psychiatric problems develop much less bonding and bridging social

capital. They also face the challenges created by the family life course more than others. Their family

configurations do not follow the regular patterns found in non-clinical samples. Although this point

need further empirical enquiry, various evidence points either to a much faster pace of changes in

some cases or an inability of families, in other cases, to follow a developmental cycle. The provision

154
of emotional support nevertheless requires that long lasting interdependencies have built up among

family members and this cannot be achieved in a week or a month. In other cases, individuals with

psychiatric problems focus on a small set of family members, usually their parents, sometimes one or

two siblings, who hardly ever change of importance. Many individuals with psychiatric problems

cannot make adaptive changes in their family configurations according to the requisites of their life

stage or the overall social situation that they currently experience. In David Olson and colleagues'

words (1989), the way in which family change is experienced in those cases is either chaotic or rigid.

Obviously, both cases make the provision of emotional support by family members difficult. Overall,

the hypothesis of a deficit of social capital for individuals with psychiatric problems is confirmed by

research reviewed in this book. Family configurations have cumulative effects on psychiatric

problems, and psychiatric problems deteriorate family configurations, in reciprocal causality.

Psychologically fragile individuals cannot count on the same level of family resources than other

individuals. That makes them less likely to overcome their own problems.

That said, it would be inconsequential to define family configurations of individuals with

psychiatric problems as belonging to a single type of organization. A variety of family configurations

also arises there. Individuals with psychiatric problems vary to the extent that they include partners,

friends or care professionals, children, parents, siblings, and other relatives. Some build their family

configurations on a nucleus constituted by their partner and their biological children. Most of them do

not, however, and focus on either kinship or friendship. Others also include care professionals as

family members. Although the overall level of social capital provided by family members is much

lower than in non-clinical samples, distinct types of family configurations and social capital

nevertheless appear. Family disadvantages experienced by individuals with psychiatric problems is not

homogeneous. It is marked by the pluralization of life trajectories that creates much diversity in the

composition of family configurations and thus, indirectly, in social capital. Family diversity again

responds to the structural constraints imposed by the organization of life courses in late modernity

societies.

155
The Family in Late Modernity
Ernest Watson Burgess (Burgess, 1926; Burgess, Locke & Thomes, 1960) from the early nineteen

twenties to the nineteen sixties aimed at uncovering a model of family that accounted for the variety

of new organizations that went away from the traditional patriarchal and rural family. Contrary to the

tenets of the family decline hypothesis of his time, he stressed that there was still a functional family

model in modernity and worked hard empirically to uncover its specificities. This model was

constituted by a small and stable group of interacting personalities linked by meaningful social roles, a

feeling of shared belonging and a legitimate although unequal division of labour. A long term

romantic sexual relationship, legalized by the mean of marriage, was the very core of this model of

interdependencies. This is not the type of family that we currently face. Great changes have happened

since Burgess and his colleagues published “Family from institution to Companionship” (1960),

changes that have modified the face of the Family. Interestingly the debate of some 80 years ago has

repeated itself, although its current contributors somewhat lost track of the previous contest. Facing

family changes, a widespread tendency of scholars is to frame it into a narrative of decline, which does

not help us to make much sense of families. It does not help individuals to deal with the contradictions

imposed by their family interdependencies either. Science as well as social policy or private citizens

are not well served by stories of family decline. Uncovering resources and stressing the variety of

contemporary families was in Burgess' s time, and still is today, a more fruitful way of understanding

families. The configurational perspective stresses that families, as they stand, have much resources to

propose, even though they are diverse in their composition and in the social capital that they provide to

their members. This book indeed found strong family interdependencies among individuals beyond the

nuclear family.

A widespread alternative to the family decline hypothesis is constituted by the detheorization of

family research. Because the functional perspective of the nineteen fifties and sixties and its emphasis

on the nuclear family proved to be normative and ineffective at accounting for the pluralization of

families, why altogether seeking models and generative principles of family development beyond

individual cases? Astonishingly, much of sociology has gone away from the ambition of explaining

156
what constitutes the structures and functions of families in late modernity, while overemphasizing the

singularity and the fluidity of family experiences. Untimely, the configurational perspective stresses

that families in late modernity refer to a small number of informal rules that can be uncovered by

empirical research. There is a bounded plurality of family types by which The Family incorporates

into singular family configurations, under the influence of the diversification of social constraints. In

that sense, one may wish to reconsider the issue of the Family Institution. In their classical book,

Burgess, Locke and Thomes (1960) described the shift from the Family Institution to the

Companionship Family. Paradoxically, the Companionship Family also had institutional features, with

its set of normative expectations regarding conjugal love, long term commitments, socialization and

the gendered division of labour. This is also the case of families in late modernity. Individuals

organize their family configurations along a relatively small number of informal principles stressing

the importance of blood ties and friendship in the face of the uncertainty raised by conjugal ties and by

the overall complexity of life trajectories. There is still much to learn on these principles from a

configurational perspective.

157
References

Abbott, A. (2001). Time matters. On theory and methods. Chicago: University of Chicago Press.

Adams, B. N. (1970). Isolation, function and beyond: American kinship in the 1960s. Journal of
Marriage and the Family. Vol. 32, p. 575-597

Aldous, J. (1996). Family careers: Rethinking the developmental perspective. Thousand Oaks, CA:
Sage Publications.

Aschenbrenner, J. (1973). Extended families among black Americans. Journal of Comparative

Family Studies. Vol. 4, p. 257-268.

Beck, U. (1986). Risikogesellschaf : Auf dem weg in eine andere moderne. Frankfurt am Main:
Suhrkamp.

Beck, U., Beck-Gernsheim, E. (1994). Riskante Freiheiten. Individualisierung in modernen

Gesellschaften. Frankfurt am Main: Suhrkamp.

Beck, U., Beck-Gernsheim, E. (2002). Individualization. London - Thousand Oaks - New Delhi: Sage
Publications.

Beck, U., Beck-Gernsheim, E. (2002). The normal chaos of love. Cambridge: Polity Press.

Belsky, J. (1984). The determinants of parenting: A process model. Child Development. Vol. 55, p. 83-
96.

Belsky, J. (1990). Parental and nonmarital child care and children’s socioemotional development: A
decade in review. Journal of Marriage and the Family. Vol. 52, 4, p. 885-903.

Bengtson, V. L. (2001). The Burgess Award Lecture: Beyond the nuclear family: The increasing
importance of multigenerational bonds. Journal of Marriage and Family. Vol. 63, 1, p. 1-16.

Bengtson, V. L., Harootyan, R. A. (1994). Intergenerational linkages: Hidden connections in

American society. New York, NY: Springer Publishing Company.

Bengtson, V. L., Rosenthal, C. & Burton, L. (1990). Families and ageing: Diversity and heterogeneity.
In Binstock, R., George, L. (eds), Handbook of ageing and social sciences. New York, NY: Academic
Press, p. 263-287.

Berger, P. L., Kellner, H. (1964). Marriage and the construction of reality. Diogenes. Vol. 12, 46, p. 1-
24.

158
Berscheid, E. (1983). Emotion. In Kelley, H.H., Berscheid, E., Christensen, A., Harvey, J.H., Huston,
T.L., Levinger, G., McClintock, E., Peplau L.A., Peterson, D.R., (eds), Close relationships. New York,
NY: W. H. Freeman and Company, p. 110-168.

Berscheid, E. (1996). The “paradigm of family transcendence”: Not a paradigm, questionably

transcendent, but valuable, nonetheless. Journal of Marriage and the Family. Vol. 58, p. 556-564.

Bodenmann, G., Meyer, J., Binz, G., Brunner, L. (2004). Eine deutschsprachige Replikation der
Paartypologie von Gottman. Zeitschrift für Familienforschung. 16, p. 178-193.

Bohannan, P. (1970). Divorce and after: An analysis of the emotional and social problems of divorce.
Garden City, NY: Anchor.

Bourdieu, P. (1985). The Forms of capital. In Richardson, J.C. (ed.), Handbook of theory and research
for the sociology of education. New York, NY: Greenwood Publishing Group, p.241-258.

Bowen, P. A., Powell, D., Cowan, C. P. (1998). Parenting interventions: A family system perspective.
In Damon, W. (ed), Handbook of child psychology. Hoboken, New Jersey: John Wiley and Sons, Inc.,
p. 3-62.

Bray, J. H. (1999). From marriage to remarriage and beyond. In Hetherington, E.M. (ed), Coping with
divorce, single parenting, and remarriage. Mahwah, NJ: Lawrence Erlbaum Associates, p. 253-271.

Broderick, C. B. (1993). Understanding family process. Thousand Oaks, CA: Sage Publications.

Brückner, H., Mayer, K. U. (2004). The De-standardization of the life course: What it might mean?
And if it means anything, whether it actually took place? In Macmillan, R. (ed.), The structure of the
life course: Standardized? Individualized? Differentiated ? Advances in life course research, 9.
Amsterdam: Elsevier, p. 27-54.

Buchanan, C. M., Maccoby, E.E., Dornbusch, S.M. (1996), Adolescents after divorce. Cambridge,
MA: Harvard University Press.

Burgess, E. W. (1926). The family as a unity of interacting personalities. The Family. Vol. 7, p. 3-9.

Burgess, E., Locke, H. J., Thomes, M. M. (1960). The family: from institution to companionship. New
York, NY: American Book.

Burnand, Y., Duriaux, S., Brousseau, M.-E., Cailhol, L., Rubovsky, G., Damsa, C., Andreoli, A.
(2004). Three years follow-up of borderline patients with suicide attempt. Paper presented at the 12 th
European Psychiatry Association, Geneva, Switzerland.

Burt, R. (2001). The social capital of structural holes. In Guilléen, M.F., Collins, R., England, P.,
Meyer, M. (eds.), The new economic sociology. New York, NY: Russel Sage Foundation.

Burt, R. S. (1995). Structural holes. The social structure of competition. Cambridge, MA: Harvard
University Press.

Caplow, T. (1982). Christmas gifts and kin networks. American Sociological Review. Vol. 47, p. 382-
392.

159
Carsten, J. (2004) After kinship (new departures in anthropology). Cambridge: Cambridge University
Press.

Cartwright, D. and Harary, F. (1956). Structural balance: A generalization of Heider’s theory.

Psychological Review. Vol. 63, p. 277-292.

Castrén, A. M. (2008). Post-divorce family configurations. In Widmer, E.D., Jallinoja, R. (eds),

Beyond the nuclear family. Families in a configurational perspective. Bern: Peter Lang.

Chen, Z-Y., Kaplan, H. B. (2001). Intergenerational transmission of constructive parenting. Journal of

Marriage and the Family. Vol. 63, p. 17-31.

Cherlin, A. (1978). Remarriage as an incomplete institution, American Journal of Sociology. Vol. 84,
3, p. 634-650.
Cherlin, A. (2004). The deinstitutionalization of American marriage. Journal of Marriage and Family.
Vol. 66, 4, p. 848-861.

Cherlin, A. J., Furstenberg, F. F. (1994). Stepfamilies in the US: A reconsideration. Annual Review of
Sociology. Vol. 20, p. 359-381.

Coenen-Huther, J., Kellerhals, J., Von Allmen, M., Hagmann, H., Jeannerat, F., Widmer, E. D. (1994).
Les réseaux de solidarité dans la famille. Lausanne: Réalités Sociales.

Cohen, S., Wills, T. A. (1985). Stress, social support and the buffering hypothesis. Psychological
Bulletin. Vol. 98, 2, p. 310-357.

Coleman, J. (1988). Social capital in the creation of human capital. American Journal of Sociology.
Vol. 94, p. 95-121.

Coleman, M., Ganong L. (1987). The cultural stereotyping of stepfamilies. In Pasley, K., Ihinger-
Tallman, M., (eds), Remarriage and stepparenting: Current research and theory. New York, NY:
Guilford Press, p. 19-41.

Coleman, M., Ganong, L. (1990). Remarriage and stepfamily research in the 1980s: Increased interest
in an old family form. Journal of Marriage and the Family. Vol. 52, p. 925-940.

Coleman, M., Ganong, L., Cable, S.M. (1997). Beliefs about women's intergenerational family obliga-
tions to provide support before and after divorce and remarriage. Journal of Marriage and Family.
Vol. 59, 1, p. 165-176.

Coleman, M., Ganong, L., Fine, M. (2000), Reinvestigating remarriage: Another decade of progress.
Journal of Marriage and the Family. Vol. 62, p. 1288-1307.

Cook, J. A., Cohler, B. J., Pickett, S. A., Beeler, J. A. (1997). Life-course and severe mental illness:
Implications for caregiving within the family of later life. Family Relations. Vol. 46, p. 427-436.
Corboz-Warnery, A., Fivaz-Depeursinge, E., Gertsch Bettens, C., Favez, N. (1993). Systemic analysis
of father-mother-infant interactions: The Lausanne triadic play. Infant Mental Health Journal. Vol. 14,
4, p. 298-316.

160
D'Andrade, R. (1995). The Development of cognitive anthropology. Cambridge: Cambridge University
Press.

Davis, J. A. (1979). The Davis/ Holland/ Leinhardt studies: An overview. In Holland, P. W., Leinhardt,
S. (eds), Perspectives on social network research. New York, NY: Academic Press, p.51-62.

Davis, J. A., Leinhardt, S. (1972). The structure of positive interpersonal relations in small groups. In
Berger, J. (ed). Sociological theories in progress, vol. 2. Boston: Houghton Mifflin.

De Carlo, I., Widmer, E. D. (forthcoming). The fabric of trust in families: Inherited or achieved? In
Jallinoja, R., Widmer, E. D. (eds), Assembling families. In preparation.

De Soto, C., Albrecht, F. (1968). Cognition and social orderings. In Abelson, R. P. (ed), Theories of
cognitive consistency: A sourcebook. Chicago: Rand McNally and Company, p. 531-539.

Duck, S. (1996). Handbook of personal relationships: Theory, research and interventions. Hoboken,
NJ: John Wiley & Sons.

Dunning, E., Mennell, S. (2003). Norbert Elias. Sage master of modern social thought. London: Sage.

Duvall, E. (1957) Family development. Oxford, England: J.P. Lippincott.

Elias, N. (1983). The court society. Oxford, UK: Basil Blackwell.

Elias, N. (1984). What is sociology? New York, NY: Columbia University Press.

Elias, N. (1991). The society of individuals. Oxford, UK: Basil Blackwell.

Elias, N. (1994). The civilizing process. Oxford, UK: Basil Blackwell.

Elzinga, C. H., Liefbroer, A. C. (2007). De-standardization of family-life trajectories of young adults:

A cross-national comparison using sequence analysis. European Journal of Population. Vol. 23, p.
225-250.

Erikson, E. (1968). Identity: Youth and crisis. New York: W. W. Norton & co.

Fehr, B., Perlman, D. (1985). The family as a social network and support system. In L’Abate, L. (ed),
The handbook of family psychology and therapy. Homewood, IL: The Dorsey Press, p. 323-356.

Feld, S. L. (1981). The focused organization of social ties. American Journal of Sociology., Vol. 86, 5,
p. 1015-1035.

Festinger, L. (1957). A theory of cognitive dissonance. Stanford, CA: Stanford University Press.

Finch, J. (1989). Family obligations and social change. Cambridge, UK: Polity Press and Basil
Blackwell Inc.

Firth R., Hubert J., & Forge A. (1970). Families and their relatives. London: Routledge & Kegan Paul.

Fischer, L. R. (1983). Mothers and mothers-in-law. Journal of Marriage and the Family. Vol. 45,p.
161
187-192.

Fivaz-Depeursinge, E., Corboz-Warnery, A. (1999). The primary triangle. New York, NY: Basic
Books.

Fivaz-Depeursinge, E., Favez N. (2006). Exploring triangulation in Infancy: Two constrasted cases.
Family Process. Vol. 45, 3-18.

Freeman, L. C. (1992). Filling in the blanks: A theory of cognitive categories and the structure of
social affiliation. Social Psychology Quarterly. Vol. 55, 2, p. 118-127.

Furstenberg, F. F., Kaplan, S. B. (2004). Social capital and the family. In Richards, M., Scott, J., Treas,
J. (eds). Blackwell companion to the sociology of families. Malden, MA: Blackwell publishing, p. 218-
232.

Furstenberg, F. F., Spanier, G. B. (1987). Recycling the family: Remarriage after divorce. Beverly
Hills, CA: Sage Publications.

Furstenberg, Jr., Frank, F. (1987). The new extended family: The experience of parents and children
after remarriage. In Pasley K., Ihinger-Tallman M. (eds), Remarriage and stepparenting: Current
research and theory. New York, NY: Guilford, p. 42-61.

Gabadinho, A., Ritschard, G., Studer, M., Müller, N. S. (2008). Mining sequence data in R with the
TraMineR package, a user’s manual. Geneva: University of Geneva. Available at:
http://mephisto.unige.ch/traminer.

Ganong, L., Coleman, M., Fine, M., and Martin, P. (1999), Stepparents'affinity-seeking and affinity-
maintaining strategies with stepchildren. Journal of Family Issues. Vol. 20, 3, p. 299.

Gauthier, J.-A. (2007). Empirical categorization of social trajectories: A sequential view on the life
course. PhD thesis. Faculté SSP. University of Lausanne.

Gauthier, J.-A., Widmer, E.D., Bucher, P. & Notredame, C. (2009). How much does it cost?
Optimization of costs in sequence analysis of social science data. Sociological Methods & Research.
Vol. 38, 1, p. 197-231.

Gehring, T.M. (1998). Family System Test (FAST). Seattle: Hogrefe & Huber Publishers.

Gehring, T.M., Wyler, I.L. (1986). Family System Test (FAST): A three dimensional approach to
investigate family relationships. Child Psychiatry and Human Development. Vol. 16, 4, p. 235-248.

Giddens, A. (1990). The consequences of modernity. Stanford, CA: Stanford University Press.

Giddens, A. (1991) Modernity and self-identity. Self and society in the late modern age. Cambridge:
Polity Press.

Giddens, A. (1995) Transformation of intimacy: Sexuality, love and eroticism in modern societies.
Cambridge: Polity Press.

162
Glick, P.C. (1947). The family cycle, American Sociological Review. Vol. 12, p. 164-174.

Goodenough, W. (1965). Yankee kinship terminology: A problem in componential analysis.

American Anthropologist. Vol. 67, p. 259-287 (Special Publication).

Granovetter, M. S. (1973). The strength of weak ties. American Journal of Sociology. Vol. 78, 6, p.
1360-1380.

Guilley, E., Lalive d'Epinay, C. (2008). The closing chapters of long lives: Results from the 10-Year
Swilsoo study on the oldest old. Hauppauge, New York: Nova Science Publishers.

Harknett, K., Knab, J. (2007), More kin, less support: Multipartnered fertility and perceived support
among mothers. Journal of Marriage and Family. Vol. 69, 1, p. 237-253.

Harris, C. C. (1990). Kinship. Buckingham: Open University Press.

Heider, F. (1946). Attitudes and cognitive organization. Journal of Psychology. Vol. 21, p. 107-112.

Heider, F. (1958). The psychology of interpersonal relations. Hillsdale, NJ: Lawrence Erlbaum
Associates.

Heru, A. M., Ryan, C.E. (2004). Burden, reward and family functioning of caregivers for relatives
with mood disorders: 1-year follow-up. Journal of Affective Disorders. Vol. 83, 2, p. 221-225.

Hill, R., Rodgers, R. (1964). The developmental approach. In Christensen, H.T. (ed.), Handbook of
marriage and the family. Chicago: Rand Mc Nally, p. 171-211.

Holman, T. B. (1981). The influence of community involvement on marital quality. Journal of

Marriage and the Family. p. 143-149.

Hughes, E. C. (1945). Dilemmas and contradictions of status. The American Journal of Sociology. Vol.
50, 5, p. 353-359.

Hyman, H. H. (1942). The psychology of status. Archives of Psychology. Vol. 269, p. 5-94.

Johnson, M. J., Milardo, R. M. (1984). Network interference in pair relationships: A social

psychological recasting of Slater’s theory of social regression. Journal of Marriage and Family. Vol.
46, p. 893-899.

Jokinen, N. S., Brown, R. I. (2005). Family quality of life from the perspective of older parents.
Journal of Intellectual Disability Research. Vol. 49, 10, p. 789-793.

Julien, D., Markman, H. J., Léveillé, S., Chartrand, E., Begin, J. (1994). Networks’ support and
interference with regard to marriage: Disclosure of marital problems to confidants. Journal of Family
Psychology. Vol. 8, 1, p. 16-31.

Kantor, D., Lehr, W. (1975). Inside the family: Toward a theory of family process. Journal of
Marriage and Family. San Francisco: Jossey-Bass.

Kawachi, I., Berkman, L. F. (2001). Social ties and mental health. Journal of Urban health. Vol. 78, 3,

163
p. 458-467.

Kellerhals, J., Montandon, Cl. (1991) Les stratégies éducatives des familles. Genève: Delachaux et
Niestlé.

Kellerhals, J., Widmer, E. D., Levy, R. (2004). Mesure et démesure du couple. Cohésion, crises et
résilience dans la vie des couples. Paris: Payot.

Kerr M., Bowen M. (1988). Family evaluation. W. W. Norton & Company.

Killworth, P. D., Bernard, H. R. (1976). Informant accuracy in social network data. Human
Organization. Vol. 35, p. 269-286.

King, V. (2002). Parental divorce and interpersonal trust in adult offspring. Journal of Marriage and
Family. Vol. 64, p. 642–656.

Kohli, M. (1986). The world we forgot: a historical review of the life course. In Marshall, V. W. (ed),
Later life. The social psychology of aging. London: Sage, p. 271-303.

Krackhardt, D. (1987). Cognitive social structures. Social Networks. Vol. 9, p. 109-134.

Kuhn, M. H. (1964). The reference group reconsidered. The Sociological Quarterly, Vol. 5, p. 5-23.

Kumbasar, E., Romney, A., K., Batchelder, W. H. (1994). Systematic biases in social perception.
American Journal of Sociology. Vol. 100, 2, p. 477-505.

Lambert, J.-L. (2002). Les déficiences intellectuelles : actualités et défis. Fribourg: Édition
Universitaires de Fribourg.

Laszloffy, T. A. (2002). Rethinking family development theory: Teaching with the systemic family
development model. Family Relations. Vol. 51, 3, p. 206-214.

Lee, G. R. (1980). Kinship in the seventies: A decade review of research and theory. Journal of
Marriage and the Family. Vol. 42, p. 923-936

Lesthaegue, R. (1995). The second demographic transition in Western countries: An interpretation. In

Mason, K.O., Jensen, A.-M. (eds). Gender and family change in industrialized countries. Oxford:
Oxford University Press, p. 17-62.

Letonturier, E. (2006). Jeu, réseau et civilisation. Métaphores et conceptualisation chez Norbert Elias.
L'Année sociologique. Vol. 56, 1, p. 67 – 82.

Lévi Strauss, C. (1949). Les structures élémentaires de la parenté. Paris: Presses Universitaires de
France.

Levin, I. (1993). Families as mapped realities. Journal of Family Issues. Vol. 14, 1, p. 82-91.

Levy, R., Bühlmann, F., Widmer, E.D. (2007). Dual and single career couples in Switzerland:
Exploring partners’ trajectories. Journal of Family Research. Vol. 3, p. 263-289.

164
Lewis, R. A. (1973). A longitudinal test of a developmental framework for premarital dyadic
formation. Journal of Marriage and the Family. Vol. 35, 1, p. 16-25.

Lüscher, K. (2002). Intergenerational ambivalence: Further Steps in theory and research. Journal of
Marriage and Family. Vol. 64, 3, p. 585-593.

Lüscher, K. (2004). Conceptualising and uncovering intergenerational ambivalence. In Pillemer, K.,

Lüscher, K. (eds.) Intergenerational ambivalences: new perspectives on parent-child relations in later
life. Oxford: Elsevier, p. 23-62.

Lüscher, K. (2009). Ambivalenz: Eine soziologische Annäherung. In Dietrich, W., Lüscher, K., Müller
C. (eds) Ambivalenzen erkennen, aushalten und gestalten. Eine neue interdisziplinäre Perspektive für
kirchliches Arbeiten. Zürich: Theologischer Verlag.

Madsen, W. (1964). The Mexican-Americans of South Texas. New York: Holt, Rinehart & Winston.

Maes, B., Broekman, T.-G., Dosen, A., Nauts, J. (2003). Caregiving burden of families looking after
persons with intellectual disability and behavioural or psychiatric problems. Journal of Intellectual
Disability Research. Vol. 47, 6, p. 447-455.

Mattessich, P., Hill, R. (1987). Life cycle and family development. In Sussman, M.B., Steinmetz, S.K.,
Peterson, G.W. (eds), Handbook of marriage and the family. New York and London: Plenum Press, p.
437-469.

Mauss, M. (1997). The logic of the gift: Toward an ethic of generosity. New York, NY: Rotledge.

McIntyre, L.-L., Blacher, J., Baker, B.-L. (2002). Behaviour/mental health problems in young adults
with intellectual disability: the impact on families. Journal of Intellectual Disability Research. Vol. 46,
3, p. 239-249.

McPherson, M., Smith-Lovin, L., Cook, J. M. (2001). Birds of a feather: Homophily in social
networks. Annual Review of Sociology. Vol. 27, p. 415-444.

Mederer, H., Hill, R. (1983). Critical transitions over the family life span: theory and research.
Marriage and Family Review. Vol. 6, p. 39-60.

Merton, R.K. (1957) Social theory and social structure. New York, NY: Free Press.

Meynckens-Fourez, M., Tilmans-Ostin, E. (1999). Les ressources de la fratrie. Toulouse: Erès

Editions.

Miklowitz, D. J. (2004). The role of family systems in severe and recurrent psychiatric disorders: A
developmental psychopathology view. Development and Psychopathology. Vol.16, p. 667-688.

Miklowitz, D. J., Goldstein, M. J., Nuechterlein, K. H. (1995). Verbal interactions in the families of
schizophrenic and bipolar affective patients. Journal of Abnormal Psychology. Vol. 104, 2, p. 268-276.

Milardo, R. M. (1983). Social networks and pair relationships: a review of substantive and
measurement issues. Sociology and Social Research. Vol. 68, 1, p. 1-18.

165
Milardo, R. M. (1988). Families and social networks: An overview of theory and methodology.
Thousand Oaks, CA: Sage Publications.

Milardo, R. M. (1989). Theoretical and methodological issues in the identification of the social
networks of spouses. Journal of Marriage and the Family. Vol. 51, p.165-174.

Minuchin, S. (1974). Families and family therapy. London: Tavistock Publications.

Modell, J., Furstenberg, F. F., Hershberg, T. (1976). Social change and transitions to adulthood in
historical perspective. Journal of Family History. Vol. 1, 1, p. 7-32.

Monney, A. (2007). Les configurations familliales et le Family Network Method (FNM):

Aspects théoriques et étude de validation. Université de Lausanne.

Moore, G. (1990). Structural determinants of men’s and women’s personal networks. American
Sociological Review. Vol. 55, p. 726-735.

Moreno, J. (1934). Who shall survive? Foundations of sociometry, group psychotherapy and
sociodrama. Oxford: Beacon New York.

Morgan, D. (2008). Acquaintances: Their position within webs of relationships. In Widmer, E. D.,
Jallinoja, R. (eds), Beyond the nuclear family. Families in a configurational perspective. Bern: Peter
Lang, p. 345-363.

Moss, S., Prosser, H., Costello, H., Simpson, N., Patel, P., Rowe, S., Turner, S., Hatton, C. (1998).
Reliability and validity of the Pas-Add checklist for detecting psychiatric disorders in adults with
intellectual disability. Journal of Intellectual Disability Research. Vol. 42, p.173-183.

Murdock, G. P. (1949). Social structure. Oxford: The MacMillan Company.

Nandrino, J.-L., Doba, K. (2001). La représentation de l’organisation familiale chez les patients
schizophrènes institutionnalisés. Etude comparative du dessin de famille. Cahiers Critiques de
Thérapie Familiale et de Pratiques de Réseaux. Vol. 26, p. 121-137.

Nerlove, S. B., Burton, M. L. (1972). A further examination of cognitive aspects of English kin terms.
American Anthropologist. Vol. 74, p. 1249-1253.

Newcomb,T. M. (1961). The acquaintance process. New York: Holt, Rinehart and Winston.

Newton, T. (1999). Power, subjectivity and British industrial and organisational sociology: The
relevance of the work of Norbert Elias. Sociology. Vol. 33, p. 411 - 440.

Olson, D.H., McCubbin, H.I., Barnes, H.L., Larsen, A.S., Muxen, M.J., Wilson, M.A. (1989).
Families: What makes them work ?. Beverly Hills: Sage Publications. 2ème éd.

Oppenheimer, V.K., Lewis, A. (1999). Career development and marriage formation in a period of
rising inequality: who is at risk? what are the prospects? In Booth, A., Crouter, A. C., Shanahan, M. J.
(eds) Transitions to adulthood in a changing economy: No work, no family, no future? Westport:
Praeger Publishers.

166
Papernow, P. L. (1996), Becoming a step family. Patterns of development in remarried families.
Cambridge, MA: GestaltPress.

Parsons, T., Bales R.F. (1956). Family, socialization and interaction process. London: Routledge and
Kegan Paul Ltd.

Pasley, K. (1987). Family boundary ambiguity: Perceptions of adult stepfamily family members. In
Pasley, K., Ihinger-Tallman, M. (eds), Remarriage and stepparenting: Current research and theory.
New York: Guilford Publications, p. 206-225.

Pescosolido, B. A., Wright, E. R. (2004). The view from two worlds: The convergence of social
network reports between mental health clients and their ties. Social Science & Medicine. Vol. 58, 9, p.
1795-1806.

Peterson, D. R. (1983). Conflict. In Kelley, H. H., Berscheid, E., Christensen, A., Harvey, J. H.,
Huston, T. L., Levinger, G., McClintock, E., Peplau, L. A., Peterson, D. R. (eds), Close relationships.
New York: W.H. Freeman, p. 360-396.

Pierce, G. R., Poppe, L., Sarason, B. R., Sarason, I., Shearin, E. N., Waltz, J. A. (1991). Perceived
social support and working models of self and actual others. Journal of Personality and Social
Psychology. Vol. 60, 2, p. 273-287.

Popenoe, D. (1993). American family decline, 1960-1990: A review and appraisal. Journal of
Marriage and Family. Vol. 55, 3, p. 527-555.

Popenoe, D. (2005). War over the Family. New Brunswick, NJ: Transaction Publishers.

Pruett, C. L., Calsyn, R. J., Jensen, F. M. (1993). Social support received by children in stepmother,
stepfather and intact families. Journal of Divorce and Remarriage. Vol. 19, 3-4, p. 165-179.

Quintaneiro, T. (2005). The concept of figuration or configuration in Norbert Elias' sociological

theory. Teoria & Sociedade. Vol. 12, 1, p. 54 - 69.

Rankin, P., Bentall, R., Hill, J., Kinderman, P. (2003). Perceived relationships with parents and
paranoid delusions: Comparaisons of currently ill, remitted and normal participants. Psychopathology.
Vol. 38, p. 16-25.

Reiss, D. (1971). Varieties of consensual experience. Family Process. Vol. 10, 1, p. 1-28.
Reiss, D. (1981). The family's construction of reality. Harvard : Harvard University Press.

Romney, A. K, D’Andrade, R. G. (1964). Cognitive aspects of English kin terms. American

Anthropologis. Vol. 66, p. 146-170 (special edition).

Roussel, L. (1992). La famille incertaine. Paris: Odile Jacob.

Ruggles, S. (1996). The effects of demographic change on multigenerational family structure: Unite
States Whites, 1880-1980. In Bideau, A., Perrenoud, A., Lynch, K. A., Brunet, G. (eds), Les systèmes
démographiques du passé. Lyon: Programme Rhône-Alpes de Recherches en Sciences humaines, p.
21-40.

167
Sandler, I. N., Barrera, M. (1984). Toward a multimethod approach to assessing the effects of social
support. American Journal of Community Psychology. Vol. 12, 1, p. 37-52.

Sapin, M., Widmer, E. D. (2008). Social isolation or relational instability ? Family configurations of
women at risk of child abandonment. In Widmer, E. D., Jallinoja, R.(eds). Beyond the nuclear family.
Families in a configurational perspective. Bern: Peter Lang, p. 303-328.

Sarason, I. G., Levine, H. M., Basham, R. B, Sarason, B. R. (1983). Assessing social support: The
social support questionnaire. Journal of Personality and Social Psychology. Vol. 44, 1, p. 127-139.

Sarason, I. G., Sarason, B.R., Shearin, E.N. (1986). Social support as an individual difference variable:
Its stability, origins, and relational aspects. Journal of Personality and Social Psychology. Vol. 50, 4, p.
845-855.

Scanzoni, J., Marsiglio, W. (1991). Wider families as primary groups. Marriage and Family Review.
Vol. 17, p. 117-133.

Scanzoni, J., Polonko, K., Teachman, J., Thompson, L. (1989). The sexual bond. Rethinking families
and close relationships. Thousand Oaks, CA: Sage Publications Inc.

Schneider, D. M. (1980). American kinship. A cultural account. Chicago: The University of Chicago
Press.

Sennett, R. (1998). The corrosion of character. New York: W. W. Norton & Company.

Shanahan, M. J. (2000). Pathways to adulthood in changing societies: Variability and mechanisms in

life course perspective. Annual Review of Sociology. Vol. 26, p. 667-692.

Simmel, G. (1955). Conflict and the web of group-affiliations. New York, NY: Free Press.

Simmel, G. (1999). Sociologie. Paris: Presses universitaires de France.

Smart, C. (2007). Personal lives: New directions in sociological thinking. Cambridge, UK: Polity
Press.

Sprey, J. (1969). The family as a system in conflict. Journal of Marriage and the Family. Vol. 31, p.
699-706.

Stacey, J. (1990). Brave new families: Stories of domestic upheaval in late 20th century America. New
York, NY: Basic Books.

Stacey, J. (1993). Is the sky falling ? Journal of Marriage and the Family. Vol. 55, p. 555-559.

Stack, C. B. (1983). All our kin. Strategies for survival in a black community. New York: Basic
Books.

Stein, C. H., Bush, E. G., Ross, R. R., Ward, M. (1992). Mine, yours and ours: A configural analysis of
the networks of married couples in relation to marital satisfaction and individual well-being. Journal
of Social and Personal Relationships. Vol. 9, p. 365-383

168
Surra, C. A., Milardo, R. M. (1991). The social psychological context of developing relationships:
Interactive and psychological networks. In Kingsley, J. (ed), Advances in personal relationships. Vol.
3, 1-36.

Théry, I. (1993) Le démariage : justice et vie privée. Paris: Odile Jacob.

Treas, J., Widmer, E. D. (2000). Attitudes toward women labor force participation in 24 industrialized
nations. Social Forces. Vol. 78, 4, p. 1409-1436.

Umberson, D. (1992). Relationships between adult children and their parents: Psychological
consequences for both generations. Journal of Marriage and the Family. Vol. 54, p. 664-674.

Vaux, A., Phillips, J., Holly, L., Thomson, B., Williams, D., Stewart, D. (1986). The social support
appraisals (SS-A) Scale: Studies of reliability and validity. American Journal of Community
Psycholog. Vol, 14, p. 195-219.

Wallace, A.F.C., Atkins, J. (1960). The meaning of kinship terms. American Anthropology, 1946-
1970. Vol. 62, p. 58-80.

Wasserman, S., Faust, K. (1994). Social network analysis: Methods and applications. Cambridge,
UK: Cambridge University Press.

Weston, K. (1997). Families we choose: Lesbian, gays, kinship. New York: Columbia University Press

Wethington, E., Kessler, R.C. (1986), Perceived support, received support and adjustment to stressful
life events, Journal of Health and Social Behavior. Vol. 27, p. 78-89.

Widmer E. D. (2007). Social capital in wide family contexts: An empirical assessment using social
network methods. International Review of Sociology. Vol.17, 2, p. 225-238.

Widmer, E. D. (1997). Influence of older siblings on initiation of sexual intercourse. Journal of

Marriage and the Family. Vol. 59, p. 928-938.

Widmer, E. D. (1999a). Les relations fraternelles des adolescents. Paris: PUF

Widmer, E. D. (1999b). Family contexts as cognitive networks: A structural approach of family

relationships. Personal Relationships. Vol. 6, p. 487-503.

Widmer, E. D. (2004). Couples and their networks. In Richards, M., Scott J., Treas, J. (eds). Blackwell
companion to the sociology of families. Malden, MA: Blackwell publishing Ltd, p. 356-373.

Widmer, E. D. (2006). Who are my family members? Bridging and bonding social capital in family
configurations. Journal of Personal and Social Relationships. Vol. 23, 6, p. 979-998.

Widmer, E. D., Chevalier, M., Dumas, P. (2005). Le "Family network method" (FNM): Un outil
d'investigation des configurations familiales à disposition des thérapeutes. Thérapie Familiale. Vol. 26,
4, p. 427-445.

Widmer, E. D., Giudici, F., Le Goff, JM., Pollien, A. (2009). From support to control: A
configurational perspective on conjugal quality. Journal of Marriage and Family. Vol. 71, 3, p. 437-

169
448.

Widmer, E. D., Jallinoja, R. (2008). Beyond the nuclear family. Families in a configurational
perspective. Bern: Peter Lang.

Widmer, E. D., Kellerhals, J., Levy, R. (2004). Types of conjugal networks, conjugal conflict and
conjugal quality. European Sociological Review. Vol. 20, 1, p. 63-77

Widmer, E. D., Kellerhals, J., Levy, R. (2006a). Types of conjugal interactions and conjugal conflict: a
Longitudinal Assessment. European Sociological Review. Vol. 20, 1, p. 63-77.

Widmer, E. D., Kellerhals, J., Levy, R. (2006b). What pluralization of family relations ? Conflicts,
conjugal interaction styles and social milieu. Revue Française de Sociologie. Numéro spécial en
anglais, Vol. 47, 5, p. 131-156.

Widmer, E. D., Kellerhals, J., Levy, R., Ernst, M., Hammer, R. (2003). Couples contemporains:
Cohésion, régulation et conflits. Une enquête sociologique. Zürich: Editions Seismo.

Widmer, E. D., Kempf, N., Galli-Carminati, G. (2010). Exploring perceptions of family

relationships by individuals with intellectual disability and psychiatric disorders. Families in
Societies.

Widmer, E. D., Kempf, N., Lanzi, F., Robert-Tissot, C., Galli-Carminati, G. (2008). How central and
connected am I in my family? Family based social capital of individuals with intellectual disability.
Research in Developmental Disabilities. Vol. 29, 2, p. 176-187.

Widmer, E. D., La Farga L.-A. (2000). Family networks: A sociometric method to study relationships
in families. Field Methods. Vol. 12, 2, p. 108-128.

Widmer, E. D., La Farga, L.-A. (1999). Boundedness and connectivity of contemporary families: A
case study. Connections, Vol. 22, 2, p. 30-36.

Widmer, E. D., Legoff, J.-M., Hammer, R., Kellerhals, J., Levy, R. (2006). Embedded parenting? The
influence of conjugal networks on parent-child relationships. Journal of Personal and Social
Relationships. Vol. 23, 3, p. 387-306.

Widmer, E. D., Orita, A., Gauthier, J.-A., Sénac, N., Cucchia, A. T., Stekel, K., Grasset, F. (2008).
How central and connected I am in my family? Bridging and bonding social capital in family
configurations of young adults with psychiatric disorders. International Journal of Social Psychiatry.
Vol.54, 6, p. 550-561.

Widmer, E. D., Ritschard, G. (2009). The de-standardization of the life course: Are men and women
equal? Advances in Life course Research, Vol. 14, 1-2, p. 28-39.

Widmer, E. D., Romney, A. K., Boyd, J. (1999) Cognitive aspects of step-terms in American kinship.
American Anthropologist. Vol. 101, 2, p. 1-5.

Widmer, E. D., Sapin, M. (2008). Families on the move. Insights on family configurations of
individuals undergoing psychotherapy. In Widmer, E. D., Jallinoja, R.(eds). Beyond the nuclear family.
Families in a configurational perspective. Bern: Peter Lang, p. 279-302.

170
Widmer, E. D., Treas, J., Newcomb, R. (1998). Attitudes toward non marital sex in 24 countries.
Journal of Sex Research. Vol. 35, 4, p. 349-357.

Widmer, E. D., Weiss, C. C. (2000). Do older siblings make a difference ? The effects of older sibling
support and older sibling adjustment on the adjustment of socially disadvantaged adolescents. Journal
of Research on Adolescence. Vol. 10, 1, p. 1-27 .

Wolfinger, N. H. (2005). Understanding the divorce cycle: the children of divorce in their own
marriages. Cambridge: Cambridge University Press.

Yi, E. K. (1986). Implications of conjugal role segregation for extramarital relationships: a network
model. Social Networks. Vol. 8, 2, p.119-147.

171
Index of Tables
Table 1. Matrix of emotional support in a family configuration...............................................19
Table 2. Family members of Betty (relation to Betty, age, gender, education, occupation).....30
Table 3 Composition of eight sampled family configurations .................................................33
Table 4. Distribution of family configurations by family structures (in percent) ...................39
Table 5. Closeness in the family configuration of Betty (percent)...........................................49
Table 6. Types of social capital in family configurations.........................................................52
Table 7. First individual to be called for emotional support ( percent).....................................60
Table 8. Second individual to be called for emotional support ( percent)................................60
Table 9. Cluster of attitudes towards family members and friends as emotional support
providers (percent)....................................................................................................................60
Table 10. Support, conflict and ambivalence according to the level of closeness ...................71
Table 11. Typology of support and conflict in family configurations.......................................73
Table 12. Support, conflict and ambivalence according to types of family configurations....76
Table 13. Distribution of configurations according to the occurrence of divorce and
remarriage.................................................................................................................................92
Table 14. Kinship terms in English, French and German.........................................................97
Table 15. Support, conflict and ambivalence in family configurations of individuals with
psychiatric troubles.................................................................................................................116

Illustration Index
Figure 1. Perceived emotional support in a family configuration.............................................18
Figure 2. Relationships in the family configuration of Betty...................................................29
Figure 3. Transitivity in triads...................................................................................................46
Figure 4. A beanpole family configuration...............................................................................54
Figure 5. A friendship family configuration..............................................................................54
Figure 6. Supportive ties in Betty's family configuration........................................................72
Figure 7. Conflict ties in Betty's family configuration.............................................................72
Figure 8. Conflict, support and ambivalence in a beanpole family configuration...................75
Figure 9. A post-divorce family configuration..........................................................................82
Figure 10. Conflict, support and ambivalence in a post-divorce family configuration...........85
Figure 11. Conflict and support in two post-divorce family configurations (Alina and Dora).87
Figure 12. Family configurations of individuals with psychiatric problems ........................104
Figure 13. Types of family configurations of individuals with psychiatric problems .........113
Figure 14. The family configuration of Betty nine months later............................................119
Figure 15. Change in family configurations over a month and a half....................................122
Figure 16. Change in the family configuration of Joanna within a year.................................125
Figure 17. Change in the family configuration of Beatrice within a year...............................126
Figure 18. Types of family trajectories...................................................................................136

172