From "father" and "mother" to the "parental universe" : the origins of the questionnaire

8The first versions of the Biographies et entourage questionnaire merely asked the parents’ date of birth and, if applicable, date of death, without defining the word "parent" in any way, but implicitly assumed the unique and unambiguous meaning of what was asked. From the outset, however, this framework was found to be unsuitable and too narrow because several parents or "pairs" of parents emerged, such as adoptive parents following the death of deported biological parents, for instance. No one parent appeared to be more important or legitimate than the other, and therefore no selection could be relevant or acceptable, especially in the framework of a biographical approach that intended to explore the influence of the respondents’ close family circle throughout their life course.

9In the following test we broadened the scope of the question to "all persons who had a parental role" which enabled us to include adoptive parents and confirmed the existence of additional and/or alternative figures to the biological parents. Now 37 respondents out of a total of 116 mentioned the existence of at least one person who played a parental role in their lives. A total of 59 persons were referred to in this way. Nevertheless, this approach also revealed the complexities that lay behind the notion of "parent", from established framework to personal reality and individual perception. Parents’ partners (i.e. step-parents) who might be thought to belong to this category, were not necessarily identified by respondents as having "played a parental role", even when they lived together for many years. Conversely, other unexpected figures were mentioned, whether or not they had lived with the respondent. In short, the test revealed that certain socially-instituted parent figures were not necessarily viewed as such by respondents, whereas other persons not usually perceived as having these roles were included in the parental universe. This introduced a notion of choice in the identification of a parental relationship, which is usually instituted and not chosen (at least from the child’s viewpoint).

10Clearly, from the respondents’ viewpoint, a daily presence in the home was not sufficient [2] to confer a parental role upon a co-resident adult, or for the adult to be accepted as a parent figure, while individuals who did not co-reside were not disqualified from assuming this role. This disconnection between co-residence, the "nurturing" function and the parent figure led us to look in greater detail at the ties of affinity that are probably implicit in recognizing or attributing a parental function and a parental relationship.

11In this respect, while not producing statistically reliable data, this same test, which systematically explored the affective ties between the respondent and his/her close family circle, [3] produced some interesting results. Only 4 respondents out of the 37 who referred to a person having played a parental role reported that they were not "close" to that person, while 22 respondents out of 116 reported, in the same precise terms, that they were not "close" to their biological father and/or mother. A juxtaposition of these observations would suggest that while feeling close to one’s biological parents – rarely contested – is not, paradoxically, self-evident, conversely, appointing a parent figure presupposes a certain quality of relationship and affection (even if not automatic). This comparison of socially-instituted parents and elective parents recalls the purely elective relational model analysed by Sabine Chalvon-Demersay in her study of television films broadcast in France in 1995. The author demonstrated that all the screenplays dealt with "the relational upheavals that have occurred in our society over the past 30 years, and the cognitive resources available to manage the consequences and explore the implications of an ideal ‘world of elective relationships’ in which all relationships are based on ties of affection, freely consented and not institutional or constraining" (Chalvon-Demersay, 1996, pp. 82-83). In our very different context of cohorts born between 1930 and 1950, our exploration of respondents’ parental universes led us to include parent figures who were recognized as such on the basis of both affinity and choice.

12Faced with the diversity of situations we found in the tests and the unexpected complexity behind the notion of "parent", the final version of the questionnaire included different combinations of figures who were biologically or socially instituted in parental functions, along with more elective figures. First, biological parents were systematically included, along with adoptive parents and parents’ partners (when they had lived with the respondents). Then we explored any other persons identified by respondents as "having played the role of a parent". [4] In further tests, data collected from a large-scale survey confirmed the relevance of this category. Far from causing any surprise or ambiguity among respondents who had just related in detail the trajectories of their biological parents, or of their adoptive and/or step-parents, these parent figures were mentioned (as we shall see) by a non-negligible proportion of individuals.

13Finally we should stress that the data concerning these parent figures and the roles they played were provided by the respondents looking back over their life course since childhood. This question thus provided information on the way in which respondents, now on the brink of old age, perceived their childhoods and the adults who took a part in their education.

Frequency of references to alternative or additional parent figures

14The frequency with which the various members of the parental and educational universe, alongside or in place of biological parents, were cited as being influential before age 15, [5] revealed their importance. A total of 1,008 persons were cited, whether adoptive parents, spouses of parents or non-family parent figures.

Table 1

Percentage of respondents who mentioned persons in their parental and educational universe other than their biological parents before age 15

Number of respondents concerned Percentage Persons who played a parental role (called parent figures) 585 21.0 Adoptive parents 28 1.0 Father’s spouses 103 3.6 Mother’s spouses 50 1.8 Note : Certain step-parents or adoptive parents mentioned in the parental universe may also have been cited as playing a parental role. Population : Residents of Paris region born between 1930 and 1950. Source : Biographies et entourage survey, INED, 2001.

Percentage of respondents who mentioned persons in their parental and educational universe other than their biological parents before age 15

15While few respondents lived with adoptive parents, family universes are more often reshaped by the repartnering of a parent, with the presence of a stepmother being relatively more frequent. A striking number of respondents mentioned a person who played a parental role for them before age 15 : more than one fifth referred to at least one such parent figure.

16We should add that no typical profile emerged from the socio-demographic characteristics of respondents who designated a parent figure. Although men and women did not describe parent figures in the same terms or attribute the same functions to them (as will be detailed later), they appeared in the children’s universe independently of the respondent’s gender.

17Similarly – and paradoxically one might say – an only child was more likely to mention a parent figure than a respondent with brothers and sisters, though the differences remain small. The number of brothers and sisters was not a determining factor and neither was social background or place of birth.

18Two-thirds of respondents concerned mentioned having one parent figure and 60% of those who mentioned several figures usually referred to couples. Individuals born before the war mentioned slightly more parent figures than others, especially in the specific context of the war, but few mentioned more than two people, and a total of 803 individuals were identified as "having played a parental role" in the respondents’ childhood and adolescence.

19Before analysing how all these parental figures combined to make up the respondents’ parental and educational universe, we will first look more closely at these more elective parents who complemented or replaced the biological or socially-instituted parents. Who were these parent figures? What ties did they have with the respondents? In what framework did they exercise their parental function?

When did they play a parental role and in what circumstances?

27Most parent figures, especially those who belonged to the family circle, assumed their roles in the respondents’ early childhood if not at their birth (Table 4). That was especially true of grandparents, 80% of whom played a key role before respondents reached 6 years of age.

Table 4

Respondents’ age at start of parent figures’ period of influence, by type of relationship (%)

Relationship From birth Ages 1-5 Ages 6-9 From age 10 Total Number Family circle 45 27 19 8 100 676 Grandparent 51 28 16 4 100 403 Uncle, aunt 35 29 23 11 100 165 Brother, sister 43 10 26 21 100 52 Other relative 32 34 18 13 100 56 Outside family circle 29 24 29 16 100 127 Childminder 39 33 20 7 100 54 Other non-relative 22 18 37 23 100 73 Population : Individuals mentioned by respondents as having played a parental role in their childhood. Interpretation : 45% of parent figures exercised "a key influence" from respondent’s birth and 8% from age 10. Source : Biographies et entourage survey, INED, 2001.

Respondents’ age at start of parent figures’ period of influence, by type of relationship (%)

28The starting point at which the non-related persons became influential was more variable, with contractual figures such as childminders and household employees mostly appearing early in respondents’ childhoods, and other non-related persons usually becoming influential around age 6, although they are not necessarily associated with a specific age.

29On average, respondents were 6 years old when they first came under the influence of parent figures and this influence often lasted beyond childhood and after age 15 for half the persons mentioned, or even a lifetime for 6%.

30Cohabitation during all or part of the period of influence often affected the relationship. Some 65% of parent figures lived with the respondents, and only friends, godparents and neighbours did not cohabit, although 44% lived close by (Lelièvre et al., 2005).

31While sharing day-to-day life under the same roof fostered parental influence (Gollac, 2003), it did not automatically lead to a parental relationship, since one third of the figures mentioned did not cohabit with the respondent. Conversely, other adults (related or otherwise) sometimes lived with respondents for several years without ever being designated as a parent figure.

Cohabitation with biological parents

39Taking account of respondents’ cohabitation with their father and mother enabled us to refine our analysis by integrating breaks in co-residence with biological parents lasting at least one year, for whatever reason : death, marital breakdown, geographical separation due to war, parents’ work, respondents’ schooling, etc. (Lelièvre and Vivier, 2006).

40Thus when we take into account the actual presence of parents in the household (Table 6), we see that only 54% of respondents actually grew up with both their biological parents from age 0 to 14 without an interruption of more than one year, and that in 9% of cases, the parents’ presence did not exclude the influence of another person. Last, only 48.9% of respondents lived in the type of configuration in which the biological father or mother made up the sole, exclusive, parent couple in continuous co-residence with the respondent.

Table 6

Configurations of respondents’ actual parental universes before age 15

Composition Number % Mother + biological father in uninterrupted co-residence 1,384 48.9 53.9 + parent figure 142 5.0 Mother’s uninterrupted co-residence 389 13.7 19.0 + parent figure 96 3.4 + mother’s spouse 48 1.7 + parent figure + mother’s spouse 6 0.2 Father’s uninterrupted co-residence 37 1.3 2.4 + parent figure 20 0.7 + father’s spouse 9 0.3 + parent figure + father’s spouse 3 0.1 No biological parent or other person in uninterrupted co-residence 324 11.5 No biological parent in uninterrupted co-residence but : one parent figure in uninterrupted co-residence 274 9.7 24.6 parent figure + other person in uninterrupted co-residence 44 1.5 other person in uninterrupted co-residence (a parent’s spouse, adoptive parent) 54 1.9 Total 2,830 100.0 100.0 Source : Biographies et entourage survey, INED, 2001.

Configurations of respondents’ actual parental universes before age 15

41Conversely, a remarkably large proportion of respondents (46%) did not live continually with their two parents up to age 15 (and in come cases never lived with them) because cohabitation was interrupted with the mother or the father, or with both parents, before that age :

• 21.5% of respondents cohabited continuously with one of their biological parents. In the majority of cases, the father was absent from the household : 19% of respondents lived at least one year with their mother [10] but without their father. The reverse (father present, mother absent) was rare and concerned only 2.5% of respondents.
• 24.6% spent at least one year without their father and one year without their mother on a daily basis before age 15. In the vast majority of cases, the father and mother were absent simultaneously (at least some of the time). Only 7 individuals had never experienced the simultaneous absence of both parents, since their parents alternated their periods of absence.

42The modal duration of separation for respondents who had experienced simultaneous interruption of co-residence with both parents (the case for 689 respondents) was one year, [12] but just over half of respondents had in fact spent between 1 and 4 years without their parents.

Time spent in boarding schools, institutions and with third parties

43Persons living in an institution for at least one year before age 15 necessarily ceased to live with their two parents on a daily basis, and 336 (or 12%) out of the total 2,830 respondents were in this situation. Three-quarters of them were at boarding school but there were also periods in orphanages, sanatoriums, preventive health institutions and other medical structures, detention/correction centres, hostels for refugees or emergency shelters.

44Half of those who ceased to reside with both parents for at least one year (Table 7) were in boarding school. Periods in boarding schools were not just the result of an educational choice or the absence of a suitable local school. They often coincided with other events such as parents’ separation, migration etc. that affected family life (Lelièvre, Vivier and Clément, 2005). Boarding school may have been perceived by parents as a means to take charge of the children at a difficult time in their life course.

Table 7

Time spent in an institution and the parental universe of children whose co-residence with both parents was interrupted before age 15

Type of parental universe Number % Time spent in an institution 336 48.8 Father and mother continuously absent and at least one other person in parental universe 127 18.4 Father and mother continuously absent and no other person in parental universe 209 30.3 No time spent in an institution 353 51.2 Father and mother continuously absent and at least one other person in parental universe 238 34.5 Father and mother continuously absent and no other person in parental universe 115 16.7 Total 689 100.0 Population : Respondents who did not live continuously with their two biological parents before age 15. Source : Biographies et entourage survey, INED, 2001.

Time spent in an institution and the parental universe of children whose co-residence with both parents was interrupted before age 15

45Periods in boarding school were half as frequent for those who mentioned at least one person in the parental universe other than their biological father and mother : 34.8% went to boarding school compared with 65.2% of those who had no other person in the parental universe between ages 0 and 14.

46Apart from boarding school, another form of separation with parents consisted of placing children with third parties [13] thereby mobilizing different resources and educational figures. Half the children who were separated from both parents simultaneously before age 15 were in the care of a third party. [14] The identification of the "resource persons" who took charge of these children again reveals the potentialities of blood ties : nearly half the children were sent to live with their grandparents (for at least one year) who thus constituted the primary source of backup childcare (Table 8). Remember that grandparents are also the people most often mentioned by respondents as having played a parental role before age 15, frequently in the context of co-residence. After grandparents came uncles and aunts and, less frequently, other family members, brothers and sisters, cousins, etc.

Table 8

Relationship with the reference person in the household where respondent was placed before age 15

Relationship with respondent Number % Grandfather and/or grandmother 160 44.7 Uncle and/or aunt 63 17.6 Foster family 58 16.2 Adopted parent 28 7.8 Brother and/or sister 12 3.3 Other relative (cousins, father’s or mother’s spouse etc.) 16 4.5 Friend 6 1.7 Godfather / godmother(a) 4 1.1 Other non-related person 11 3.1 Total 358 100.0 (a) The number of godparents could be underestimated as they could also be designated by another relationship (uncle and aunt for instance). Population : Respondents who lived for at least one year before age 15 in a household other than that of one or both of their parents. Source : Biographies et entourage survey, INED, 2001.

Relationship with the reference person in the household where respondent was placed before age 15

47Last, children were not only fostered out to informal care, but also officially to foster families or adoptive parents, who in some cases were related to the respondents.

Historical and family contexts : from war to working mothers

48The childhoods of these cohorts (1930-1950) were marked by major social and political events, such as the the French Front Populaire, World War II and reconstruction, rural exodus, all of which affected children’s daily lives. Events cited include the first children’s holiday camps, the trauma of the civilian exodus after the German invasion and the war-time evacuation of children (see for example, Downs, 2002). While the context in which the parent figures were mentioned varied a great deal in this period (1930-1965), our data shows no significant difference in the propensity to cite a parent figure according to the context, including the war years (1939-1945) which one might suppose to be an important period for parent figures. Certain figures were typical of the war era, namely the so-called "other non-related persons" (neither godparents, nor neighbours, nor domestic servants) who were mostly mentioned during this period, alongside foster families. For children who had lost parents in the war or whose families were destroyed or dispersed, the impact of the war inevitably continued into peace time. Most foster families [15] are mentioned by "war orphans" after the war. Strikingly, many of the statements about parent figures during the war did not reveal a traumatic vision of the period, but referred to memories of daily life that were often quite happy. This viewpoint contrasts with today’s descriptions of the period, dominated by collective memory and by government child protection policies that have crystallized memories of a traumatized war generation (Downs, 2005). While war played a role in the emergence of certain family configurations, it does not alone explain the frequency and diversity of the parent figures that were mentioned. The majority of these parent figures exerted an influence outside the war period, in the 1930s for one quarter of them, and after the war for the majority (44%). [16]

49Before the war, the majority of persons in parental roles were the grandparents (59%); but this was no longer the case in 1940 (below 50%). Only uncles and aunts were mentioned regularly over the whole period. After the war and in the 1950s, brothers and sisters appeared more frequently, as did people from outside the family such as friends, whose role was increasing (especially parents’ friends), and neighbours. This shift in the profile of parent figures reflects the changes in the family support network taking place in a context of the rural exodus, rapid urbanization and the housing crisis that characterized 1950s France. Throughout the period from 1930, female employment, while statistically invisible, was impacting the daily organization of households. As respondents explained, the difficulty of reconciling work and family life justified the need for a third party to take over from parents, especially from unavailable mothers (Lelièvre, Vivier and Clément, 2005). Methods used varied from leaving children at the grandparents’ home on a daily or weekly basis, to the regular presence of a person in the child’s home, whether under a contractual agreement (household employee, childminder) or not (grandparents, neighbours, etc.).

What respondents said about their parent figures

51Generally speaking, respondents answered in three types of narrative registers : the functions register (57%), some of which relate to the biological parents, the circumstances under which the parental role was exercised, and the personal characteristics of the parent figures in question.

52The functions attributed to the parent figures were categorized into four main themes : material tasks (feed, nurse, meet the needs of, mind, take care of, etc.) learning, socialization and education (learning about things, contributing, etc.), affective and psychological roles (loving, cosseting, being a role model, admiring, etc.) and lastly, exercising responsibility and authority (Table 9). The terms used by respondents evoke practical, day-to-day parenting. We also observed that the five universal functions of parenthood identified by anthropologists (Goody, 1982) which consist of bearing and begetting descendents, endowing them with birth-status identity (legitimation), nurturing them, training them and introducing them socially into adulthood (sponsorship), all appeared in the corpus at different levels, with functions sometimes taken on by persons or institutions in addition to the biological parents.

Table 9

The main explanatory registers of the role of parent figures

Register Themes Frequency % Number Sub-themes Material role 9% 732 Material support (help, provide for…) Health (illness, hospital, care for…) Feed (feed, meals, cake…) Look after (look after, watch over…) Care (take_care_of, shopping, sleeping, laundry…) Development Socialization 8% 666 Early learning (learn_about_things, poetry, humour, culture…) Leisure (cinema, holidays, outings…) Contribution (contribute…) Nurturing 6% 499 (raise, take_care_of…) Functions carried out Welcome support 6% 499 Mobility (coming, going, taking…) Making welcome (welcome, put up, foster…) Affective role 5% 455 Love_affection (love, cherish, tenderness, cosseting…) 57% Respect (respect, admire, recognition…) Relationship and psychological framework 5% 396 Role model (reference, model…) Dialogue (talk, listen, say, share…) Psychology (confide, trust, affinity, closeness…) Influence Advice Educational role 4% 347 Studies (school, mathematics…) Education (educate, education…) Responsability, authority 3% 264 Administrative (guardian, adoption, social services…) Responsibility (decision, take_on, in_care…) Authority (impose, discipline, obey, fear…) and compared Comparison 7% 618 Comparison of adults (more_than, less_than, better_than…) Complement / Substitution 4% 295 Replace Second_ (2nd_ mummy, 2nd _family, 2nd _father…) Space Temporality 16% 1316 Time (years, every_day, often, Thursday…) Place (house, Paris, France…) War (world_war, exodus…) Context 28% Stages of life 6% 488 (child_childhood, throughout _life, first communion, marriage…) Parental environment 6% 532 Accident (death, disappearance, divorce…) Work (i.e. the fact that parents worked…) Absence (alone, away, miss…) Description of parent figure 14% Personality and status 14% 1203 Quality (funny, nice, stern, strong, present, cheerful…) First name Occupation or activity Indulgence (mischief, whims, give in, rascal…) Nature (rabbit_s, earth, flower_s…) Total 100% 8310 Field : Statements about individuals mentioned by respondents as having played a parental role in their childhoods. Interpretation : Several registers may occur in a single reply. We identified 8,310 thematic clauses grouped in this table under themes and sub-themes. 57% of the lemmatized corpus concerns functions exercised by parent figures, including for instance, the "nurturing" function which appears in 6% of the corpus. Source : Biographies et entourage survey, INED, 2001.

The main explanatory registers of the role of parent figures

53The respondents also supplied contextual elements that placed the event in context and specified the circumstances (historical or family) under which the parental role was exercised. Most indications concerned time and space, i.e. when, with what frequency, on what occasion and place, the parent figure first exercised his/her influence. Respondents thus supplied elements of understanding, reasons that explained or legitimized the involvement of a third person and that revealed the absences, constraints and difficulties to which their families were subjected and which led to (or "justified") the involvement of another "parent". Tying in with the functions assumed by parent figures, respondents thus defined the various adults who made up their parental universe using terms of comparison, complementarity or substitution.

54Last, respondents provided descriptive elements characterizing the personality or the status (social, occupational) of the parent figure, thus giving clues as to why a parental role was attributed to that individual. These characteristics were related less to context (history or family) than to respondents’ personal perceptions. They shed light on the functions of these individuals and on the quality and affective dimension of the relationship.

Who said what about whom?

55Men and women referred equally to the eight parenting functions, whether with regard to caring, help, assistance, nurturing, responsibility, welcoming or supervising (Table 10). However, women specifically mentioned notions of psychological and affective care and functions such as intellectual stimulation, whereas the men, whose descriptions were less detailed overall, referred more to practical and educational functions (especially regarding family members) and often summarized the parental role as one of "replacing" parents without providing any specific details.

Table 10

Summary of main functions by respondent’s gender and relationship with parent figures (PF)

Functions Female discourse Male discourse Related PF Unrelated PF Related PF Unrelated PF Feed – provide for grandmother aunt godfather, godmother Take care of grandmother, sister childminder aunt childminder Help brother, sister godfather, godmother grandfather parents’ friend Nurture grandmother childminder grandmother childminder Educate grandmother godmother grandmother, aunt, sister, uncle godfather, godmother Responsibility uncle, brother, sister godfather grandmother, brother foster family Replace parents grandparents, brother, sister foster family grandparents, brother, sister godfather, godmother, foster family Welcome uncle, aunt, brother, sister parents’ friend foster family, parents’ friend Supervise childminder employer Affective & psychological role grandmother, aunt, sister godmother, foster family, childminder sister foster family Recreation / leisure grandmother, uncle, brother childminder, parents’ friend sister childminder, parents’ friend Early learning grandparents, aunt, uncle, brother godmother, respondents’ and parents’ friend uncle, sister Context and description : – of the person – of the circumstances personality, temperament, place, period, frequency occupation, activity, war, period Field : Statements about the individuals mentioned by respondents as having played a parental role in their childhoods. Interpretation : To feed/provide for is a function mentioned with respect to most parent figures, but for men this function is proportionally more often used in relation to aunts among related parent figures, and to godfathers and godmothers among unrelated parent figures. Source : Biographies et entourage survey, INED, 2001.

Summary of main functions by respondent’s gender and relationship with parent figures (PF)

56Women were also more likely to place the exercise of these functions in context and to describe the parent figures, whereas when men broached these aspects, they referred more to the occupational and social identity of the parent figures (worker, employee, in business, daughter of, brother of) and/or specified the context by pointing out its exceptional nature (alone, father prisoner, world war, absent, travel). Women were more likely to describe the person’s temperament (cheerful, open, strict, honest, available, sentimental) or their attributes (pipe, plants), and specify the frequency of the relationship (often, Sunday, Thursday, weekly, always, period, per year) and the place (Paris, Trouville, next door, opposite).

57These descriptive nuances seem to reveal gender differences in the perception of practical parenting. Men appear to be more marked by the social status of the parent figure and the context, when this was exceptional, whereas the women remained more attached to the description of the personality and the characteristics of the relationship (nature, frequency, place). This suggests that men tend to look for reasons why a third person should enter their parental universe (exceptional circumstances), whereas women appear to attach more importance to the concrete and factual description of the third person’s presence (practical aspects). These differences in perception are doubtless connected to the respondents’ later lives and their current position in relation to the practicalities of parenthood. It would be worthwhile to expand the comparative analysis of siblings’ discourse where studies have shown that brothers and sisters describe their biological parents differently (Clément 2002 ; Langevin, 1989).

Predominantly affectionate relationships

58Frequently, and in a wide variety of ways, the descriptions of parent figures from the children’s viewpoint broach another important aspect of the relationship, namely affection. The affection that occurs in the construction of a feeling of kinship for instance, in the context of Weber’s practical kinship (2003), was not identified among the universal functions of parenthood. Yet affection was frequently mentioned in the children’s viewpoints that we collected. Over and above the material and supervisory functions of parenting, most respondents mentioned affection and closeness as a decisive factor in the attribution of a parental role. Respondents expressed the happiness [18] of living with a grandmother who loved them more than their mother, they were grateful to the person who were devoted to their well-being or who took the trouble to intervene on their behalf. This relationship quality was particularly important in the women’s statements, showing the existence of this more affective function. Certain individuals were mentioned only because of the exceptional quality of the relationship.

59

On the subject of a grandfather [19] with whom the respondent had "spent his life" up to age 10 : "He had a little hose-drawn cart and took me out all day. He showed me all sorts of interesting things and explained them, and he would say ‘if you’re not the strongest you’ve got to be the smartest’. Not a day passes when I don’t think about him."

60More generally, psychological and intellectual aspects, as well as early-learning and leisure activities, mostly described by the female respondents, were also described by the male respondents but in relation to their sisters and uncles. There is a kind of crossed relationship between brothers and sisters from this point of view : elder sisters have a special relationship with their younger brothers and elder brothers have a special relationship with younger sisters.

61

On the subject of his elder sister, a teacher, a male respondent explained : "She took more care of my schooling than my parents did, helping me get from the 6th grade to the 3rd grade, and into second grade [secondary school classes] ; She took me out, or we relaxed, went to the theatre or on cultural outings and she took me on holiday with her."

62But for both men and women, the listening and confiding functions remained more in the female circle of sisters and aunts (Table 10).

63Finally, this affective aspect was occasionally contrasted with tensions in relationships with parents. One respondent remembered neighbours in the same building "where I stayed frequently when my parents were fighting. They taught me about family life and were very affectionate." This affective aspect reveals the links between the various adults who made up the parental universe. What is at stake for adults is exercising the parental role. Were the elective parents in competition with the biological ones or substitutes for them? To what extent were they assigned to their role by the parents themselves?

Distribution of roles : from complementary to surrogate

64Distributing parent functions between several persons was not just a practical domestic and family arrangement, carried out in the interest of the child. It also served to maintain social bonds, and to ensure their reciprocity and continuity by means of the relationships it generated between adults. Other societies, especially in Africa and Oceania, are more familiar with the idea that parental functions may be held by a variable number of individuals (Goody, 1999 ; Brady, 1976), but multi-parenting has always existed in our societies too (Burguière, 1993 ; Lallemand 2002). Whereas today families are reconstituted after divorce or separation, in the past they were reconstituted after widowhood, since death left countless spouses prematurely widowed and gave the surviving spouse the right, if not the social duty, to remarry. Today reconstituted families, assisted reproductive technologies, same-sex parents and adoption are challenging the fundamental criteria of unicity and exclusivity that dominate representations of paternal and maternal roles (Fine, 2001 ; Lallemand, 2002 ; Ouellette, 1998). Given the wide variety of parental configurations also observed in these earlier generations, it would be interesting to learn more about the interplay between the different complementary and surrogate parent figures.

65The distribution of tasks was often the result of an agreement between adults and organized in a complementary manner, as demonstrated in these comments about an aunt :

66

"She was there when I was born. My mother and aunt were so close that I can’t distinguish them. They brought us up together. My aunt had more time, she lived with us from the day I was born until I was four years old."

67Describing the personality of the parent figure gave respondents the opportunity to compare the various adults that made up their parental universe either in positive terms or in relation to their respective parenting skills.

68The notion of "replacement" appeared in several expressions (Table 9) relating to experience of both complementary and surrogate parenting.

69

On the subject of a maternal aunt : "There were a lot of us at home. Paula wanted to take me because I ate better. She was like a second mother."

70Replacement did not exclusively concern orphans but related to a temporary or permanent reassignment of parental functions to these third parties who "played the role" of parents to make up for absence, (during war or following separation or death), lack of availability when the family was large or when the parents’ occupations (especially those of the mother) were very demanding, [20] or inadequate care if the parents were "ill", "depressed" or had abandoned their children, either literally or figuratively. [21]

71Of course, the composition and functioning of the parental universe also depended on the available resources and each individual’s personal availability, a fact which respondents also referred to :

72

Thus on the subject of a maternal uncle : "He was a bachelor, we all lived on the same farm. He didn’t marry so that he could look after his parents and his sister (my mother) and her three children (us). He devoted himself to us, he was the head of the family and played the role of the missing father." (Quote from a male respondent)

73Here the uncle takes on the father’s role in a fatherless family because he is a man and his own history is marked by his role, but also because he is available : he lives on the spot, he is not married and has no children. Thus geographic proximity, gender relations, availability in terms of the parent figure’s own family and working life, are all important parameters affecting their availability as parental substitutes.

74In some cases the terms of comparison are more explicit and the biological parents are compared unfavourably with the complementary or surrogate parent figures :

75

On the subject of a childminder : "In terms of affection, she replaced my mother. [She was a] very good person, devoted, had a heart of gold. [I never received] any affection from my mother. I was very sad to return to her because Germaine fell sick. I called her Mummy, I couldn’t call my mother Mummy."
Referring to a maternal grandmother : "I lived with her, I was very happy, in the countryside, nice people, she was a mother to me, whereas I was one too many as far as my mother was concerned. My grandmother brought me up."

76Lastly, there was the legitimation function, whereby a substitute parent assumed the legal responsibilities of a biological parent in his/her absence. This was mentioned for legally-recognized foster families, as well as for certain figures (usually male) such as uncles, brothers and godfathers. It arose in relation to the respondent’s civil status or to the duties of adults as the respondent’s official representative in dealings with social services or schools or in precise situations sometimes outside the specific context of childhood (e.g. marriage), and likewise to their duties as the respondent’s legal guardian or as the person viewed as his/her protector. When the statement specified that the guardian was "legal" he or she was described in formal, administrative terms. For instance, one respondent said of a paternal uncle who lived locally, "He was not our guardian but he took care of the family and the business like a guardian." However, he specified that, "he didn’t live with us so the other brother was more important." This example reveals that the very selective recognition of a parental role from the respondents’ viewpoint was based more on practical day-to-day parenting than on the legal role. Guardianship and legitimation were minimized and reduced to an abstract legal function. Conversely, they were valued by respondents when attributed to a person recognized as being a "real" guardian, even without having any legal status, because those people were present and influential on a daily basis.

77Like legal guardians, although the godparents’ relationship was formally instituted in law and/or by the parents, they were not necessarily recognized as having a parent function by the child. As a result, this type of relation remained outside the register of practical parenting that prevailed in respondents’ statements, a fact which explains the limited presence of godparents among the persons mentioned as having played a parental role (Table 2 and note 6).

The special role of the grandmother

78Grandmothers are dominant parent figures in several respects. They were available, since their own children had grown up, they were often economically inactive, and were very frequently mentioned by respondents, particularly by women (Table 10). This calls for a specific examination of their role.

79Grandmothers were abundantly described and characterized by the diversity of the functions they held. In the material sphere (using the following terms : running [the household], providing, ingredient, kitchen, preparing, care, contagious) as well as in the psychological and intellectual sphere (share, cheerful, museum, learning about things, available). The grandmother is specifically associated with concocting tasty dishes. She is a cake-making granny but also the mistress of the house who runs the household. The grandmother pays a more important role for the elder children. She is the first figure to be mentioned as a replacement for an absent working mother. Her affective and psychological role is often described as making up for a less affectionate mother. In comparison, information relating to the grandfather, associated first and foremost with the image of the father and the status of the grandmother’s husband, is very scant.

80Among relatives, the grandmother was most often cited for nurturing functions in early-age experiences, while the childminder was most cited among the non-relatives. The grandmother was also at the top of the list of family figures having an educational role. [22] When respondents described this role they referred to an educational style (strict, or on the contrary, lenient and affectionate), in the precise framework of school work or broader intellectual discovery (museums, reading, etc.).

81Above all, the grandmother was referred to in connection with everyday life. However, the relationship with the grandmother, more than with other parent figures, was viewed as being part of a tradition. Her role often conveys a notion of family transmission. Reference to a female family culture is perceptible, and sometimes set against the relationship between respondents and their own mothers. The grandmother-granddaughter relationship, more than any other, may make up for a difficult relationship between respondents and their mothers, though it may also be in competition (an overbearing grandmother attempting to "squeeze out" the mother).