1People do not marry just anybody. This was how Alain Girard summed up one of the main results of the survey he conducted in 1959 on "The Choice of a Spouse" [1]. Since then, many authors have tested the validity of this statement and attempted to assess the extent of homogamy in France by comparing the positions and social origins of spouses [2]. But demonstrating the homogamous nature of mate selection is insufficient to elucidate the mechanisms which bring about this outcome. An investigation into the different paths leading to the altar, rather than merely taking stock of transactions once they have occurred on the marriage market, is necessary to understand the relationship between love and homogamy.
2Many intermediary factors lie between individual emotion and statistical trends, but two fundamental ones can be distinguished. Firstly, the elements which define the setting in which interaction will take place and the probabilities that the future spouses will meet, depending on their characteristics and social background. Secondly, the categories of perception and judgement which structure the choices made within this framework. In Erving Goffman’s terms, this could be expressed as distinguishing between the "scene of judgement" and the judgement itself. We may assume that mate selection, like many other forms of selection, follows principles of which the individuals concerned are not necessarily conscious, since they involve individual dispositions with deep social roots, what Pierre Bourdieu has termed habitus [3]. According to this author, what brings a man and a woman together is mostly the affinity between their habitus, "produced by similar social conditions and conditioning" [4]. Within this framework, the possible existence of more specifically marriage-oriented strategies calls for investigation.
Who marries whom ? A comparison of the social origins of spouses, among French couples aged under 45 (married or not)
Who marries whom ? A comparison of the social origins of spouses, among French couples aged under 45 (married or not)
3To throw light on these questions, INED carried out in 1983-1984 a "Formation of Couples" survey on 3,000 individuals less than 45 years of age who were living as married or unmarried couples (see Annex 1), which complemented the 1959 survey on The Choice of a Spouse. The Formation of Couples survey confirmed that, during the last 25 years, couples have not been formed at random. Now, as in the past, individuals tend to prefer partners with a similar social background (Figure 1). Although empirically established, this major fact and its underlying mechanisms remain a mystery. In the present paper, we shall attempt to describe the first of the two factors mentioned above, the setting in which love-interaction takes place. Where and how did the respondents meet ?
How the places where people meet have changed since 1914
4The data from the two surveys conducted by INED were combined to form a continuous time series, with the same treatment throughout, so that the different forms of mate selection can be followed from 1914 to 1984. People nowadays meet their spouses in a wide range of places (Figure 2). In addition to those where people work or study, there exists a great variety of public places and leisure activities. Such a range is very recent : in the 1920s, two out of three marriages were the outcome of a meeting at a dance ("bal" [5]), at work, in the neighbourhood or during a private visit. Fifty years later, these four forms of encounter only accounted for one third of all meetings.
Where do spouses meet ? Distribution of meeting-places from 1914 to 1984
Where do spouses meet ? Distribution of meeting-places from 1914 to 1984
5The most notable feature is the steady decline in the role played by neighbourhood meetings during these seventy years. In the 1920s, these were the primary matchmaker, accounting for more than one out of five marriages, whereas today they have practically disappeared. Rural depopulation and the loosening of village ties cannot by themselves account for such a dramatic decline in the incidence of neighbourhood meetings ; this also reflects an increasingly pronounced reluctance to meet under the scrutinizing stare of the community.
6When the period as a whole is considered, the most popular meeting-place is indisputably the bal. After a large increase during the 1930s, the proportion of couples who first met at a bal peaked in the 1960s (no less than one out of five), then fell by half (one out of ten around 1980). It is true that night clubs and discos were increasing in importance in the meantime, partly offsetting the bal’s lower score until the mid-1970s. However, this is no longer true of recent years, and public dances have now lost weight on the marriage market. Through its various ups and downs, the bal has gradually changed both its form and its public. In the 1920s, it played an important role for both urban and rural couples, from all social groups, except the upper classes. After World War II, the bal became increasingly successful in country areas, since the increasing rural depopulation meant that traditional village ties were no longer adequate for facilitating the formation of couples. Manual workers, clerical and service staff, numerous in the country, often resorted to this source of meeting, much more than farmers. In the 1960s, the latter finally had little option but to turn to this new type of meeting-place and started going to bals further afield, throughout the district, in search of spouses which the immediate neighbourhood could no longer provide. The bal has now become a specifically rural meeting-place, popular mainly among farmers and rural manual workers.
7Since the 1960s, meetings in a public place or in the course of outings to the cinema, theatre, etc. have made considerable headway. Couples meet in the street, neighbourhood, housing estate, café, baker’s, park, shopping centre, hospital, public transport… The novel aspect of this mode of encounter should be noted. These impromptu meetings between strangers, where getting together requires an effort on the part of the individual, were rare in the past.
8People also meet their future spouses when visiting or entertaining friends or relatives. This private socializing is more frequent among urban couples than are neighbourhood meetings, and is more upper-class. Its importance for the formation of couples decreased continuously until the 1960s, and then regained ground. In the meantime, there has been a change in the type of meetings covered by private socializing. Until the 1950s, half of these were clearly arranged by a third party, generally from the older generation. Matchmaking now follows other paths, first and foremost, parties. Christmas and New Year’s Eve parties, the whole range between formal parties and having a few friends in, with or without dancing, etc. are all occasions which, over the last half-century, have become increasingly important for meeting one’s future spouse. Family gatherings, on the other hand, generally marriages of friends or relatives, have slowly lost ground. However, as these are events which are much rarer than dances or parties, their "marriage yield" per event is in fact high. Parental intervention, generally concerning daughters, has not altogether disappeared, but has become more discreet and indirect. Girls are now escorted by the group they go out with to dances or discos. The inquisitive solicitude of parents and neighbours has given way to the less strict and generally more tolerant intervention of peers. Judgement from on high has been replaced by more gentle advice from a jury whose members judge one another.
9Surprisingly, meetings at work or study have remained stable over time, as though the search for a spouse were not affected by later accession to the labour force or the raising of the school leaving age to 16 years, introduced in the late 1950s. Contrary to popular view, the incidence of meetings at school or university has remained low, increasing only from 1 out of 20 in the 1950s to 1 out of 12 since the end of the 1960s. The secondary school years are studded with love affairs, but these very seldom lead to union. The "serious affairs" start after school. Even among graduates, only 14 % met their spouse at university or college. In this field, the school environment seems to have less influence than might have been expected.
10It is difficult to draw a general picture of the seventy years covered by the two INED surveys. Change does not boil down to a gradual suppression of restraints which supposedly limited mate selection in the past. The impression of being able to make free choice had been prevalent for a long time. The pattern of meetings has changed because the meeting-places themselves have changed and, even more, because the social environment of the two interested parties has changed. During their late teens, young people are divided between the family group and the peer group, each of which has its own particular influence on the way the future spouse is met.
11The diversity in the modes of encounter, the presence of parents or peer groups on the scene, also account for the variations observed in what follows that fatal first meeting (Table 1). A meeting in a night club or disco, at a party, on holiday, a place of entertainment, at work, university or college, is followed shortly after by sexual intercourse and, in most cases, unmarried cohabitation. On the other hand, meetings in more conventional places or where community control is stronger, such as the bal, public festivities, family gatherings, the neighbourhood, sports, clubs or societies, or school, take a more traditional course : the couple go out together for longer before "consummating" their union and generally marry without first cohabiting. This traditional behaviour, particularly prevalent in rural areas, contrasts with the "emancipated" attitude more common among urban dwellers. There is a clear correlation between first meeting-place (bal or night club, confirmation [6] or funfair…) and the ensuing behaviour. A study of the social characteristics of the couples is necessary to throw light on this correlation.
Timing of the mating-process, broken down by circumstances of encounter1,2,3,4
Timing of the mating-process, broken down by circumstances of encounter1,2,3,4
1. The man was a customer or a patient.2. The woman was a customer, a patient or a student.
3. With or without a meal ; but not a party. The home can be that of a friend, a relative, or of the interviewed person.
4. Unweighted figures (but means and percentages were calculated on figures weighted to take into account the true proportions of married couples and of unmarried couples). See Annex 1.
Closed and open : social categories and modes of encounter
12Close correlations are observed between social space and meeting-place (Table 2 and Figure 3). The form of representation adopted here brings to light a social hierarchy in meeting-places, which are all the more closed as they are related to the upper classes. In addition to the well-known selective role of education, we discover that holiday places are also highly selective (the indices of over-representation show that they are even more so than schools and universities for engineers and senior managerials).
13Informal parties are as selective as sports or cultural activities practised in a club, or at least involving regular participation. In other words, individual choices which step by step delimit a social circle constitute a sort of cooptation process which is every bit as efficient as older formalized procedures. A more careful investigation reveals, however, that these two modes of selection are not completely interchangeable : club activities are more successful among members of occupations with a more intellectual component (primary and secondary school teachers, social workers) than in groups defined by a more economic component (engineers, senior managerials, liberal professions…). Conversely, the latter are more prone to meet their spouses through private socializing.
14At the other end of the social scale, among the lower social groups, such "local" oppositions can also be found, but they do not challenge the general principle of a social hierarchy. For instance, unskilled workers have to resort more often than other categories of manual workers to the open-market type of meeting place, bals and public places. They make little use of family gatherings, night clubs or discos : these semi-open meeting-places (in a way, the most closed of the open places) are more easily available to the upper categories of the lower social groups. The publics of night clubs and bals respectively are observed to differ, and substitution of one for another can only be partial, as the historical presentation showed.
Where the members of different social groups meet their spouses
Where the members of different social groups meet their spouses
Population : see Figure 3.Distribution of encounters according to social group and social profile of meeting-places
Distribution of encounters according to social group and social profile of meeting-places
15Meetings at work have a different logic. They are particularly frequent among men working in sectors in which the ratio of women in the workforce is high : teaching, social welfare work, civil service or, in the private sector, clerical work. The case of primary or secondary-school teachers is somewhat particular. In their working world, differences between groups do not imply hierarchic subordination. This creates, more than elsewhere, a context of social homogeneity that in turn favours homogamy. Craftsmen and tradesmen may meet their spouses at work in a number of ways : they may be customers (the most common case), employees or, when the craftsmen/tradesmen are themselves salaried, colleagues.
16Faced with interpreting this series of profiles, the analyst might be tempted to proceed in one of two directions. He could try and establish a strict correspondence between social class and form of encounter, which is clearly contradicted by empirical evidence (if this correspondence did exist, it would already be well known and there would be no need for a sociological study to demonstrate it). Alternatively, he may use an approach which is every bit as unrealistic, by interpreting the different shades perceptible on most profiles as evidence of a social "continuum" which, step by step, would mean in the end that all types of encounter would be permissible for anyone. This argument is often put forward in social mobility studies, where the change from one generation to the next goes together with drifts of this kind, but it is totally inadequate for describing, at a given point in time and for individuals instead of lineages, the range of probable and improbable behaviours. In the analysis presented below, we propose an alternative to the traditional choice between complete segregation and a continuum.
17The central part of the graph (Figure 3) is far from perfectly regular, which is hardly surprising. Not only are the two classifications too detailed for that but, for purposes of cross-tabulation, groups need to be arranged in a linear hierarchy, whereas modes of encounter, like social positions, call for a multi-dimensional representation. But these very limitations mean that the results are all the more significant : the fact that the link between meeting-place and social space is so apparent, despite the linear presentation, is in itself a striking result. It will be seen below that the "deviant" cases situated far from the diagonal (such as the rare members of the upper classes who meet their spouses in open places), far from weakening the link observed between meeting-place and social milieu, actually strengthen it.
The triangle of encounters
18The apparent diversity of meeting-places covers a network of relations between functional equivalents. This is confirmed by the fact that an aggregated classification of meeting-places into only three classes already shows extremely pronounced social differences (Figure 4). For men and women alike, the whole social pyramid is apparent in this "triangle of encounters".
The triangle of encounters. Distribution (%) of men and women from each socio-occupational category among three groups of places : public, private, select
The triangle of encounters. Distribution (%) of men and women from each socio-occupational category among three groups of places : public, private, select
19Among the lower social groups, public places in the broad sense seem to be most successful as meeting places. These are open to all, with no form of selection other than a modest entrance fee : public festivities, fairs, bals, the street, cafés, shopping centres, parks, cinema, public transport… Members of the upper classes, on the other hand, tend to meet their spouses in places to which access is limited. We have termed them select places : societies or clubs, work, school, university or college, restaurants, night clubs, concert halls, sports centres… This enumeration may appear somewhat heterogeneous, but its relative unity is based on the fact that admission to these places does not depend solely on payment of an entrance fee, but is subject to a numerus clausus which can be either established formally, through selection tests or cooptation procedures, or, just as surely, by more symbolic means, the deterrent effect of certain behavioural rules specific to the place (from this point of view, going to a restaurant is more exacting than going to the café, the concert more than the cinema, and so on). Because it is more cultural than economic, this form of selection tends to distinguish the intellectual professions (or more precisely, the professions whose capital is mainly intellectual). They are to be found nearest the summit representing the "select places", not the general managers or liberal professions, who feel more at home in a third form of sociability : the private social sphere, that is, among friends or the family (private places on the figure).
20It was not possible to use the same occupational sub-groups for women as for men. For instance, women holding (or who had held) a senior managerial post were too rare to be put in a group of their own. Most frequently the senior posts held by women were in education ; in the pyramid, these are the equivalent of secondary school teachers or senior civil servants for men. Considered as one group for the same reasons, skilled women workers manifestly shun the meetings in public places which are so popular among unskilled workers. Members of middle-level professions and clerical and service staff were, on the contrary, sufficiently numerous to be broken down into sub-groups. Large differences can thus be observed between domestic and personal service workers (hairdressers, cleaners), sales staff and clerical workers. Globally, it is clear that the social pyramid of forms of encounter is organized around the same principles for men and women alike.
21Thus through the multiplicity of scenes of first encounter pierces a form of social logic which is already that of homogamy. The primary opposition between public and select places, combined with the secondary opposition between select and private places, tends to segment the marriage market, even though this is not necessarily the result of specific marriage-oriented strategies. A large part of the selection process has already taken place upstream, through more general strategies of social behaviour ; in particular, those whereby the upper classes avoid the "madding crowd" and open spaces, i.e. all the circumstances in which the individual, because of the large number of participants, must often leave the arrangement of encounters to chance. Relying on this selection which has been gradually built into their everyday lives, members of the upper classes can afford to make fun of openly arranged meetings. And it is true that traditional matchmakers, in their attempts to pair people off, tend to "overdo things" by instilling more strategy into the system than is actually needed : they draw attention to calculated interests which love has already quietly integrated.
22Conversely, it would no doubt be wrong to maintain that the lower social groups lack all form of selective sociability. They, too, have their preferences, but these are generally for public places [7]. If they are the only ones to use spaces which are open to all for selecting a spouse, this is because the upper classes have abandoned the field to them. They have no need to fence in this social circle ; they need only count on the number of people present to increase the chance of an interesting encounter.
23A crowded place seems to go hand in hand with "love at first sight". The impression that this has been experienced, common to one out of eight respondents, is more frequent among couples who met in a public place (one out of five), together with the feeling that they have avoided the would-be manipulations of friends or kin. 75 % of respondents feel that they met by chance, and this is accentuated for relatively crowded places : bals, discos, public places or working places open to the public. This transition from a crowd of individuals to a crowd of couples is accepted and practised more frequently among the lower social groups.
The time of a dance
24Attitudes towards the matrimonial function of the dance also vary a great deal between different social groups. However, to characterize it, we must start by discussing the general nature of the dance. Speaking of the bal, Alain Girard has said that by paving the way for encounters between total strangers, or enabling people who had already noticed one another to meet again, it "broke down the barriers erected between individuals of different sexes by a variety of social restraints". The inexorable decline of the bal in favour of other forms of leisure activities does not invalidate this analysis, which also applies to dancing in general. This remarkable social institution forms the back-cloth to first meetings in a proportion of cases which by far exceeds the setting of the bal alone (Figure 5). Except during the Occupation (1940-1944), the matchmaking success of dances steadily increased between the 1920s and the 1970s, and although an appreciable decline has been observed in recent years for public dances, private dances have become increasingly productive. It is interesting to examine why the formation of couples owes so much to the dance.
Proportion of spouses meeting at dances since 1914
Proportion of spouses meeting at dances since 1914
25Since dancing in groups gave way to dancing in couples, the bal and its public or private substitutes (public dances, discos, night clubs) have offered an unhoped-for opportunity to pair off for a time with a member of the opposite sex. Nothing is more difficult than to make contact of this kind, and at such a crucial moment, dancing smooths things out by proposing a common language, a ready-made solution to the problem of accosting. The rythm of the music envelops and hypnotizes the actors, "tuning" their behaviour, an effect which is often reinforced by special lighting which marks off the dance floor. The dance cuts time into neat stretches, defined by the pieces played by the band or records. Alternating periods of fast dancing, generally more frequent, and slow numbers, set the two stages of the temporary pairing ritual : watching first, then getting closer.
26During the first stage, the dances that require a certain know-how (rock-and-roll and the like), reserved for the best couples, keep all the less skilled dancers on the sidelines ; usually accompanied by friends of their own sex, they take this opportunity to spot and look over prospective partners. In other dances, people move about the dance floor on their own (or near a group of friends) and can thus imperceptibly draw closer to those they wish to observe in greater detail [8]. A slightly different situation is offered by group dances such as the chain dance, which favour unexpected proximity. In all these cases, verbal contact is practically impossible, and it is observation of physical appearance and body posture that enables the participants to make their tentative choice.
27This choice can be tested in a second stage, during the slow numbers, where no expertise is required and even the poorest dancers can find a partner. This two-by-two dance brings bodies so close together that conversation becomes possible. The crucial moment, between the first and second stages, is the accosting (that is, the invitation to dance). The peer-group may at this point provide key support when one of its members decides to make a move [9]. Dancing then shows its enormous advantage as an institution with set rules and phrases : it frees individuals from the effort of inventing new accosting tactics with each attempt ; the invitation, like the refusal, follows a standardized, conventional procedure. In this sense, dancing represents an economy : it cuts both the cost of the first step, and the consequences of any slips.
28But perhaps this functional interpretation is too narrow. It is just as important to note that the dance, as a rite of interaction, is like a double entendre, in that the degree of involvement of each party remains open [10]. On the one hand, it is possible to take advantage of this introduction to enter as deeply as possible into the relationship ; on the other, it is possible to put an end to it straight away by hiding behind its institutional definition – it was only a dance. In its simplest form, this dilemma boils down to the choice between keeping the same partner for the next dance or changing partners. The dance accomplishes the feat of drawing bodies together and rousing feelings "on trial", with no commitment : a game with no stake. Approaching someone is not the whole story, it is also essential to be able to withdraw from the game when necessary, respecting social conventions. This possibility of withdrawing tactfully also exists when it comes to smaller gatherings than the traditional bal, such as private parties : in this context, the dance remains a social institution. It is because of this self-service type of functioning, where one can always change one’s mind, that the dance has become what it is, a socially acceptable means of exploring the marriage market. The dance might not play this role if dancing with someone meant being a "buyer", as in the traditional type of shop where the customer always feels obliged to purchase something [11].
Conversation or dancing ?
29Just as cohabitation has become a sort of private marriage, the small get-together type of dance is a private bal which has the advantages of the institution without its collective aspect. Far from vanishing, institutional control is internalized, and no longer requires formal guarantees. This process may ultimately lead to the total disappearance of the dance in some circles. Indeed, the proportion of meetings at dances triples from one social group to another. It is highest among manual workers and the small self-employed, but drops to one-tenth among the upper classes, even when private dances are included (Table 3). It is as though the further up the social ladder they are, the more reluctant people are to rely on pre-determined forms of mediation, which are too openly collective, and perhaps also too physical and emotional. The dance-floor is not the place where the upper classes feel most at home in showing off their assets. This may seem an unusual expression of "upper-class blues" ("malaise des cadres" [12]), but merely illustrates how each social group is inclined to opt, consciously or not, for the meeting-places which ensure the best yield for their specific form of capital.
Proportion of meetings at a dance, according to social group (%) (increasing order)
Proportion of meetings at a dance, according to social group (%) (increasing order)
30The dance is not the only form of ritualized encounter. There may be another catalyzer : the fête [13], which accounts for more than one out of four meetings. However, dance does not necessarily mean fête, nor conversely (Figure 6). If the upper classes all tend to spurn the dance, there are variations as regards the fête : liberal professions and general managers favour it in one out of three cases, scarcely less than farmers and more than manual workers ; in general, these are private parties among friends. Secondary-school teachers and senior civil servants, on the other hand, are extremely sober and ascetic : they appreciate neither dancing nor fêtes. They prefer other means of making acquaintance, such as at study, and the type of conversation they can cultivate there. The same applies to the middle-level professions whose social existence depends on the possession of certificates : primary-school teachers or middle-level civil servants (social workers, postmasters, supervisors and other B-grade civil servants). Between these two poles, economic capital on the one hand and cultural capital on the other, engineers and senior managerials occupy, as might be expected, an intermediate position.
The role of meetings at dances and fêtes in mate selection. "Formation of Couples" Survey
The role of meetings at dances and fêtes in mate selection. "Formation of Couples" Survey
31The fête also splits up dancing fans into craftsmen and tradesmen on the one hand and industrial workers on the other, while artisanal workers are in between. If members of the small self-employed categories (with the exception of farmers) frequent public festivities, they cannot muster the same network of relations as the self-employed upper-classes for organizing private parties. Whether public or private, the dance and the fête are not interchangeable settings which are natural to one and all. They are fully accepted by a given social group only when they are well suited to its particular resources.
Small investment, high returns
32Even places which favour direct contact less than dances and fêtes can afford many opportunities, at least to people who know how to make the most of them. It is true that accosting someone without a good reason, in a public place or elsewhere, tends to be interpreted as ostentatiously revealing one’s intentions. The first meetings reported in our survey show that this is neither the most frequent nor the most successful method. But a minor incident, not at all premeditated, can often give Cupid’s arrow a helping hand : a collision, a headache, an umbrella taken by mistake, a pedestrian drenched by a passing car, a slight misunderstanding… Working or studying in the same place affords many opportunities of this kind. The incident is followed by a "corrective" ritual [14] in order to restore the situation to normal, excuses are offered and accepted, ice is broken and a warmer and deeper relationship is born. Although originally accidental, the encounter respects conventions and follows a certain ritual. "Chatting-up" (« l’art de draguer ») is generally improvised, a tactical capacity to make the most of incidents or situations which arise ; more exceptionally, it can be the art of creating the incident which will trigger a relationship.
33There is no lack of opportunities in everyday life, but people do not all have the same dispositions for using them, or the same wish to do so. Some make the most of opportunities provided by their familiar working environment, while others, particularly among the lower social groups, make good use of unexpected situations which lead to encounters in open places. Members of the upper classes generally refuse to be enticed into the game of unpremeditated encounters, as they have not mastered its rules.
34There are certain similarities between the "corrective" ritual and the dance ritual. Its success is also due to the fact that a double entendre exists : an act of kindness or helpfulness can be taken to be simply a normal response to the situation, or as the prelude to a new relationship [15].
The stamp of social origin, or anticipation of future destiny
35In the above analysis, we have taken the social position of the individual at the time of the survey as reference. We could just as well have taken his/her social origins, as determined by the father’s social class. In fact, whatever the index, the structure of the differences between groups remains unchanged, intergenerational social mobility being on the whole too low to modify it. However, it seemed interesting to examine whether certain deviant individual life-courses were more specifically related to certain types of meeting-place, whether individuals whose social life-course diverged from the modal type for their groups of origin were not inclined to encounter their spouses in specific places.
36The example of farmers’ children shows the limits of this assumption (Table 4). For instance, those who remain on the land show little difference in behaviour and choice from those who become manual workers, as the latter remain culturally too close to the farming world to diverge much. On the other hand, when farmers’ children become clerical workers or rise to the middle or upper-level professions, their scope increases, if only on account of their occupational, and often residential, mobility. Places of study and work then compete with the traditional types of encounter in this group. Yet even in this socially mobile subgroup, bals and public festivities remain important, which confirms the persisting influence of social origin. The fact that members of this group more frequently made the acquaintance of their spouses through work or study does not reflect an upward mobility strategy, but more "automatically" derives from the growing diversity of their social network.
Where sons of farmers and sons of higher social groups meet their spouses, according to their own social trajectory(1),(2)
Where sons of farmers and sons of higher social groups meet their spouses, according to their own social trajectory(1),(2)
(1) Craftsmen, tradesmen, clerical, middle-level or upper-level professions.(2) Clerical, manual workers, farmers.
37The social life-courses of people with upper-class origins show differences of a similar kind (Table 4). Children of upper-level professionals, whatever group they eventually enter, show a strong aversion to bals and public festivities, but are attracted by private parties. The popularity of places of study is very variable, reflecting divergent social pathways, downward mobility being clearly related to shorter periods of study. Certain patterns of mate selection are thus apparently borrowed from the family environment (bals and public festivities for farmers’ children, private parties and aversion to bals for children of upper-level professionals), while others are less a distinctive feature of original social class than markers of a social position which is in the process of being attained (places of study, of course, but also in some cases clubs, societies and workplaces).
The choice of meeting-place : a marriage-oriented strategy ?
38"Opportunity makes the thief". If upper-level professionals tend to marry a fellow student and manual workers a dancing partner, is this not simply because the former more frequently pursue their studies and the latter more often go dancing ? Is it surprising that people become acquainted with their future spouse in their familiar social circle ? Before considering this view, it should be noted that, however rudimentary this approach may seem, it represents an improvement compared with models which claim to explain most of the trends and regularities observed in the process of mate selection by the mechanical effect of "propinquity". For instance, the opportunities theory put forward by certain American sociologists stresses that the possibility that a couple will be formed varies with the inverse ratio of the distance (or the square of the distance) between the homes of the two partners [16].
39This neglects the fact that the many social institutions which contribute to defining the environment in which encounters take place do not all have the same recruiting areas (the bal draws youths from the canton, the university gathers students from the whole regional education district – covering several départements in France –, skiing resorts attract holidaymakers from several regions, etc.), and that members of different social groups, however close they may be situated to these places, have very unequal access to them. Within the social continuum, the institutions create gradients, set boundaries, burrow new distances or forge new proximities, deeply modifying the probabilities of meeting. The concept of a "circle" proposed by Louis Henry was interesting in that it was not explicitly defined in terms of physical distance, but that the possibility of a definition in terms of social variables was kept open [17].
40What is true for the space continuum also holds for the time continuum. Here again, society sets its own rhythm, with cycles of peak times and slack times which orchestrate the timing of first meetings. This limits the scope of the "exposure to risk" model which assumes that the more time a person spends in an activity or attending leisure events, the more he/she is likely to meet the kindred spirit there. The marriage market has its slack and its busy seasons, and functions more intensely at times when social networks are being renewed : mainly during the summer, when there are holidays but also bals and public festivities ; then during the autumn, with the beginning of the school and working year : meetings at school, university or work occur most often in September and October, when new faces appear on the scene. This logic of renewed social circles also applies to meetings at Christmas and New Year parties, birthday parties and family ceremonies (weddings, communions) which often occur in May or June, as well as part of the meetings at dances. Early spring, March and April, is an unproductive time for romance, unlike the following months. The essential role of these periods of renewal, and digressions in everyday social life (holidays, new school/working year, fêtes, dances…) suggests that the time of exposure matters less than the turnover of participants and, above all, than having access to the place concerned and knowing how to use it to the best purpose. This puts the focus on the social properties of individuals.
From everyday sociability to meeting one’s spouse
41To account for the distribution of meeting-places among the different social groups, a study based on sociology or the history of sociability is not sufficient. It is clear that people are more likely to marry someone who frequents the favourite haunts of their own social group, but the relation is not at all mechanical, since meeting-places vary as regards "marriage yield". Take dancing for instance : according to a survey on leisure activities carried out by the French National Institute for Statistics and Economic Studies (INSEE) in 1967, among single men aged under 22 it was equally popular in all social groups, from the working to the upper classes. What actually differentiates the social groups is where the dancing takes place : in a private setting for upper-class youths, a public one (the bal) for the working classes. This uniform rate of participation in dances is not echoed by the results of the "Formation of Couples" survey, which showed that, compared with upper-class youths, almost three times as many youths from working-class or farming families met their future wives at a dance. This means that it is in the lower classes that dances yield their highest marriage score.
42The description given by Pierre Bourdieu thirty years ago of the awkwardness of farmers faced with urban dances [18] may make this result seem somewhat surprising. Does this mean that manual workers and farmers now feel more at home than the upper classes in an activity which involves one’s body and presentation to such an extent ? In fact, two successive attitudes can be discerned among young working-class dancers : the younger ones, less well established in their occupations, "are only out for a good time", and may change partners quite often ; then, in a second stage, when they are settled in a job, they start to look round more seriously and methodically for a stable relationship. The serious-mindedness of the latter is just as much part of the dance, bal or night-club as is the carefree Saturday-night attitude of the younger members. It is the ritual aspect of the dance and the fact that it takes place in public which account for its success among working-class youngsters as an ideal place for watching and being watched, a parade ground where they can show off their appearance and character [19], and one of the rare stages where they feel at home, because they are with their own kind and self-assured.
Proportion of graduates of a given social origin who met their spouses at college or university (%)
Proportion of graduates of a given social origin who met their spouses at college or university (%)
43Encounters at college or university provide another example. This is not an intermittent type of institution, but a permanent living environment. Spending a few years there might seem sufficient to find a spouse, but this is by no means the case. On average, only 15 % of men and 13 % of women who had studied in college or university met their spouses there (Table 5). Globally, the marriage yield is therefore low, but it varies with social origin and sex : 18 % of children of upper-level professionals marry a fellow student against 10 % of those of working-class origin. Children of manual and clerical workers have been subject to a severe selection process at school, but as they come from large classes, they nonetheless account for a considerable proportion of students [20]. But being less familiar with the university environment than upper-class students, their chances of making lasting relationships there are more limited. The children of upper-level professionals, on the other hand, followed by those of craftsmen and tradesmen, are the group with the highest marriage yield : more at home in this world, which for them represents a normal stage of life, they are capable of obtaining a better score.
44Differences are observed between men and women. In small groups, such as children of farmers or middle-level professions, the proportion of encounters at university which lead to a union is higher among men than among women, as though they anticipated a future social promotion. For women in these categories, the student years seem to be devoted above all to obtaining a degree which will form (as François de Singly so aptly put it) their "educational dowry" [21]. This is not the case for daughters of craftsmen and tradesmen, who meet their future spouses there more often than do the sons of members of this group. It is as though some of these daughters went to university with the idea of meeting their future husband there, which would mean that, in this case and to a lesser extent in the case of daughters of men in upper-level professions, the term "marriage-oriented strategy" would apply.
45College and university life represents a melting-pot in which the "cadres" [22] social group is formed, partly by self-reproduction and partly by integrating non-members. Women who married a fellow student chose a son of "cadre" in only 38 % of cases, but, when their social destinies were considered, these were future "cadres" in 60 % of cases (whereas the probability of becoming a "cadre" for a son of "cadre" is only 42 %). In comparison, women graduates who chose a spouse outside the university or college sphere were less inclined to marry sons of "cadres" (31 %) and much less inclined to marry future "cadres" (only 43 %). It is therefore true that "good catches" are more frequent at university or college than elsewhere and also true that certain women know this. However, most of the differences observed between social groups, and between men and women, are explained by differences in the social use that is made of university life.
Deviant choices
46Is it possible to assess the contribution of each meeting-place to overall homogamy ? Or to compare, for a given social group, the "homogamic yield" of the different places where its members meet their future spouses, so that a list of the places that should be frequented, or on the contrary avoided, could be drawn up to ensure the best-matched unions ? The size of our sample does not make it possible to go into such detail, but by grouping certain categories together, the dominant features can be studied.
47When the meeting-place is deviant, so is the chosen spouse. Logically, the chances of meeting a spouse of one’s own social origin decrease considerably when acquaintance is made outside the usual haunts of one’s own group : globally, outside closed settings for the upper classes and outside public spaces for manual workers. A public festivity, a fair, a local bal or a disco are all disreputable places (even dens of iniquity) for upper-class youths, which cuts their chances of meeting their own kind by half. In the same way, at the other end of the social scale, children of skilled manual workers (which form a large enough category for comparison) are more likely to meet upper-class spouses in closed places such as clubs, societies, sports, parties or during their studies, than in open places : the chances are doubled for girls and tripled for boys.
48Deviant behaviours, commonly termed "exceptions", are by no means random or arbitrary, nor do they contradict the overall correlation between meeting-place and social group. There are two reasons for this. First, "places", like "social groups", are slightly hazy notions. In reality, despite the existence of certain institutional barriers, they often consist of a core and flexible contours which may overlap with other groups or places. Part of the distinction between typical and deviant cases stems from the "classification effect" (as does part of the distinction between homogamous and heterogamous unions). But even if the (social) spectrum shows only very gradual changes in colour, this does not imply that colours should not be distinguished. However arbitrary the limits of age-groups may be, the existence of young people and of the elderly cannot be questioned, and it is not because dawn and dusk exist that there is no point in making a distinction between day and night.
49The second reason, related to the first, is that the deviations observed in certain social groups, what may be called "out of bounds" meetings, can often be explained by the fact that they occur in a subgroup, or among people who follow social life-courses that could not be identified at first. For instance, the children of "cadres" who meet their spouses at a bal are not the children of just any "cadres" : the social and geographic origins of their parents are distinctly below average for that group [23]. A more precise definition of the social background, obtained by using a more detailed classification or complementary variables (mother’s occupation, grandfather’s ocupation, residence…), enables a large fraction of the exceptions to be recovered. It is difficult, even impossible, to measure this proportion exactly, since classifications can always be improved and, at the same time, the size of the sample does not allow the analysis to be carried much further. But as long as better results can be obtained by more detailed breakdowns of groups or places, this course should be pursued, even if this offends the convictions of those who believe in chance or free choice.
50* * *
51Cupid’s arrows do not strike the social chess-board at random, but form a diagonal line, perfectly visible in the cross-tabulation of social origins of spouses. How can the multitude of individual love choices converge to give such a result ? This could never happen through strategies applied consciously by the social actors, even less with specifically "marriage-oriented" strategies. The statement that "when a woman marries, she is interested in choosing and being chosen by a man with the highest possible social value" or that she "tries to get a man whose social value is at least equivalent to her own in order to protect her interests" [24] cannot be taken at face value, as this would imply searching for the active principle of the process in its outcome. The homogamy table may show that capital marries capital, but nothing entitles us to state that individuals marry in order to unite or exchange their capital [25]. Other mechanisms are at work, much further upstream, which spare the actors a substantial part of the effort of selecting a spouse. They take the form of socially structured sociability behaviours, mainly based on the opposition between closed and open places, which tends to parallel the opposition between the upper and the lower social groups.
52This pre-selection itself may take on two separate forms. On the one hand, there are social strategies that are not necessarily marriage-oriented, but aimed at reserving admission to a given meeting-place to members of one’s own group (this is more often achieved by avoiding places frequented by other groups than by a positive move). Advantage may also be taken of certain familiar places which are already limited by existing social boundaries, thus ensuring meetings only with one’s fellows. This situation may, in turn, be the remote result of what used to be a deliberate strategy for investing a place or keeping others at bay, which was perfectly conscious at the time, but which has now become fully assimilated into the social structure (residential or school choices made by the previous generation, for instance).
53This objectified calculation is so deeply rooted in the everyday environment that it is no longer seen as a strategy, even if parents are more prone than their offspring to reactivate its strategic aspect by maintaining a dichotomy between recommended and out-of-bounds places, anticipating the effects on future unions. As for the actors themselves, they can generally only adopt this attitude once they have reached an age when views on society become more objective and disillusioned. An extreme example is provided by prospective spouses who are forced to put into words what should go without saying, that is, to state clearly their intention to marry by going through specialized intermediaries, the personal columns of the newspapers or the marriage bureau. There is, moreover, a widespread distrust of this form of encounter. When asked "If it had been necessary, would you have used the personal column or gone through a marriage bureau ?", 80 % of respondents said no.
54When analysing events such as the first meeting with one’s future spouse, it is understandably difficult to draw the line between deliberate option and constrained behaviour. For observers and actors alike, it is extremely easy to climb a few rungs up or down the ladder which connects subjective strategies to objective constraints by way of objectified strategies. All this forms a spectrum in which each type of behaviour is represented not by a single point, but by a fairly broad band, open to several interpretations. Within these gradually shifting bands, connections can be established step by step between openly marriage-oriented behaviours and others that were apparently far removed from such ideas, such as choosing friends of the same sex (whose brother or sister turn out to make a fine husband or wife…), pursuing one’s studies, joining a sports club, enjoying public dancing or parties, or the propensity to approach strangers in a place open to all (which implies some sort of social training). It would be a mistake to tip the scales in favour of either unconscious social structure, or conscious strategic involvement. This would shackle the process of finding a spouse with a univocal, predetermined chain, which would leave little room for the very possibility of love-choice. In fact, the "marriage market" would have much difficulty in functioning if the interested parties saw it as the market-place that others (parents, sociologists, economists) want to see in it for them.
Technical characteristics of the "formation of couples" survey
55The survey was preceded by a series of 30 semi-structured interviews, which served as a pilot survey for the preparation of the questionnaire, and was followed up by a second series of 75 interviews with volunteers selected from among the respondents of the main survey.
562,957 respondents were interviewed. The quota sampling method was used to obtain a sample of individuals of both sexes, either married (for the first time or not) or having cohabited for at least two months, of French nationality (not necessary for the spouse). The fieldwork was carried out by 237 interviewers from the Survey Department of INED, under the supervision of Benoît Riandey, who also established the quotas in such a way as to overrepresent cohabitants. The quotas were defined as follows :
Age of respondents on January 1st 1984 :
5718-24 years : 21.5 % ; 25-34 years : 43.4 % ; 35-44 years : 35.1 %.
58Mean age of sample : 31.3 years.
Sex structure :
59Males : 46.7 % ; females : 53.3 % (there are more females than males among under-45s living as couples, because of the age difference between spouses).
Marital status :
60Married : 73.5 % ; cohabiting : 26.5 % (784 individuals). At the data processing stage, the true proportion of cohabitants was applied : roughly 9 %.
Socio-occupational categories of men :
61No occupational activity : 1.4 %
62Senior civil servants, senior managerial staff, higher intellectual professions : 12.7 %
63Middle-level professions : 19.8 %
64Clerical and service staff : 10.7 %
65Manual workers : 40.1 %
66Farmers : 6.7 %
67Craftsmen, tradesmen and general managers
68(i.e. heads of private firms) : 8.3 %.
The questionnaire :
69It comprised 250 questions and interviews took about an hour on average. In addition to the standard questions designed to locate the two spouses precisely in both geographical and social space and to determine the chronology of events, numerous open-ended questions were included, as these prompt respondents to give accounts or descriptions which make it possible to follow the process of formation of their couple very closely.
70This type of approach was applied to the circumstances of the first meeting ; the role of peer-groups in the encounter ; how further contact was made after the first meeting ; the social life the couple led ; the impression made by the future spouse at the time of the first encounter and subsequently his/her physical appearance and way of dressing ; the respondent’s ideal partner prior to the meeting and the opinion of what is an acceptable difference in height between men and women ; the first meeting with the future parents-in-law. For most subjects, an open question was associated with a set of closed questions. For instance, the open question on the circumstances of the first encounter introduced many questions on the date and exact place, the role attributed to chance, the existence of people who knew both partners before the first meeting.
The French classification of occupations and socio-occupational categories
71The new French Classification of Occupations and Socio-Occupational Categories (PCS : Professions et catégories socioprofessionnelles) was used to identify the social origins and position of respondents. The PCS classification, conceived and elaborated by the French National Institute for Statistics and Economic Studies (INSEE), was built as a reasoned empirical system. It takes into account empirical statistical and historical evidence of the French socioeconomic scene (e.g. the existence of a group of "cadres", deeply rooted in French social history – and with no equivalent term in English – which covers senior civil servants, senior managerial staff and the higher intellectual professions). It is therefore very much a reflection of French society, but in addition, is based on practical and theoretical criteria and distinctions of more general relevance.
72The construction of this classification followed the principle that society should not be considered and analysed on a one-dimensional scale, but as a social space, whose description should bring into play hierarchic and general criteria as well as local oppositions which do not necessarily have to apply throughout the entire social space.
73The main divisions of the PCS are the following :
- the division between wage-earners and self-employed ;
- the division between civil servants, employed by local and central governments (public sector), and workers in industrial and commercial firms (private sector) ;
- the occupational hierarchy adopted for collective agreements (unskilled workers, skilled workers, foremen and supervisors, technicians, etc.) ;
- functional classifications were also used to differentiate people working in a given social sphere : artisanal workers and industrial workers form two distinct groups among manual workers, occupations specifically involving a certain cultural capital (university and other teachers, researchers, etc.) are always clearly distinguished from others.
74In conclusion, the PCS classification may be used at several levels : at the most aggregated one, it is reduced to six economically active categories, the standard one to 19 categories and the detailed one to 32 categories. [****]
Notes
-
[*]
Translated by Linda Sergent (with the participation of Michel Bozon).
-
[**]
INED.
-
[***]
See those of Jean Sutter.
-
[1]
Alain Girard, Le choix du conjoint. Une enquête psycho-sociologique en France, new edition with an added preface, INED, "Travaux et Documents", no. 70, Paris, Presses Universitaires de France, 1974, p. XVI. The first edition ("Travaux et Documents" no. 44) was published in 1964. The survey carried out in 1959 covered 1,650 couples who had married between 1914 and 1959.
-
[2]
See L. Roussel, Le mariage dans la société française : faits de population, données d’opinion, INED, "Travaux et Documents", no. 73, 1975 ; J.-C. Deville, "Mariage et homogamie", Données sociales, INSEE, 1981, pp. 21-30 ; A. Desrosières, "Marché matrimonial et classes sociales", Actes de la recherche en sciences sociales, 20-21, 1978, pp. 97-107 ; C. Thélot, Tel père, tel fils ? Position sociale et origine familiale, Paris, Dunod, 1982, chapter 8 : "Les alliances".
-
[3]
See for instance P. Bourdieu, Outline of a theory on practice, Cambridge University Press, 1977.
-
[4]
P. Bourdieu, Choses dites, Paris, Editions de Minuit, 1987, p. 88.
-
[5]
The French word bal will often be used, as it is not quite the equivalent of a public dance held in a club, dance-hall, etc. With the traditional bal, there is often no permanent dance floor, but one that is transported from place to place (often from village to village) ; consequently, people not only attended dances in their own village, when it was their turn to have the dance floor, but also went further afield to attend dances when the floor was transported to other areas. Bals are often held in the open air, in connection with local fêtes or national celebrations : on Bastille Day, in particular, open-air bals are held in every neighbourhood.
-
[6]
Confirmation ("communion solennelle"), when a child is around 12 or 13 years old, is an important social event in French life, and provides the opportunity for a large family gathering.
-
[7]
The INSEE survey on "Social Contacts" shows that members of the lower social groups tend to shy away from clubs or societies (except for sports clubs). See F. Héran, "Un monde sélectif : les associations", Economie et statistique, 208, March 1988, pp. 17-31. For a detailed observation of forms of sociability among the working class, see M. Bozon, Vie quotidienne et rapports sociaux dans une petite ville de province, Lyon, Presses Universitaires de Lyon, 1984, in particular chapter 4 : "Lieux publics et groupes sociaux", pp. 73-99.
-
[8]
"I was dancing by myself. He came and danced by himself next to me. The next dance was a slow one, he asked me to dance with him". (She is a salesgirl in a supermarket, he is a farmer).
-
[9]
"In a night club, she was dancing with a girl-friend. So with a mate, we went and asked them to dance with us". (He is an assembler in the car industry, she is a hairdresser, in a medium-size town). "I met her at the "ducasse" (traditional local festivity in Northern France), then at the bal afterwards : I’d bet some friends that I’d ask her to dance a slow one". (Both wage-earning pump attendants, in a large town).
-
[10]
Erving Goffman, Interaction ritual, Doubleday, 1967 ; Penguin, 1971. See the chapter entitled "Alienation from interaction".
-
[11]
The best literary descriptions of dances/balls (such as those given by Jane Austen in Pride and Prejudice) have turned into a sort of romantic plot the institutionally ambiguous side of an invitation to dance : mere courtesy or advances ? Empty forms or meaningful forms ? Or, in the language of linguists, "demotivated" syntagms or "remotivated" ones ? (On the sociological import of this distinction, see F. Héran, "L’institution démotivée", Revue française de sociologie, 28 (1), 1987, pp. 67-97).
-
[12]
This set phrase refers to a "disease" prevalent among French upper-level professions ("cadres" : see Annex 2), which covers all sorts of social and professional frustrations.
-
[13]
The French word fête covers a variety of social events, and does not correspond to any single word in English. It has been translated, depending on the context, as party, gathering, festivity, get-together… The French word has been kept to refer to the concept, i.e. the festive nature of an event.
-
[14]
Erving Goffman, Interaction ritual, op. cit. See the chapter entitled "On Face-Work : An Analysis of Ritual Elements in Social Interaction", especially what is said about the corrective process.
-
[15]
A close idea is found in Goffman, "The Interaction Order", American Sociological Review, 1983, Vol. 48, February, p. 8 : "Friendship relationships and marital bonds (at least in our society) can be traced back to an occasion in which something more was made of an incidental contact than need have been".
-
[16]
See articles which have become classics in the sociological literature : A.C. Kerckhoff, "Patterns of homogamy and the field of eligibles", Social forces, 42, 1964, pp. 289-297 ; and W.R. Catton and R.J. Smircich, "A comparison of mathematical models for the effect of residential propinquity on mate selection", American Sociological Review, 29, 1964, pp. 522-529.
-
[17]
Louis Henry, Démographie. Analyse et modèles, Paris, INED, 1984 (1st Edition : Larousse, 1972), pp. 320.
-
[18]
P. Bourdieu, "Célibat et condition paysanne", Etudes rurales, 5-6, 1962, pp. 32-135.
-
[19]
By working on their appearance, body posture, attitude towards others, they can improve their ordinary public image, weakening, or on the contrary enhancing, certain typical features, or even completely inventing them.
-
[20]
Re-processing data from a longitudinal study conducted by INED on the educational attainments of a cohort of children followed from 1962 to 1972, from the 1st to the upper 6th form, Christine Ungerer has shown that the proportion of members of the lower social groups in the total school population contributes to concealing the consequences of social selection : whereas the rates of accession to the different levels of the school system vary considerably from one social class to another, the social composition of the overall school context is only gradually biased. See "La double vision de la sélection scolaire. Retour sur une enquête de l’INED". Revue française de sociologie, 28 (2), 1987.
-
[21]
F. de Singly, Fortune et infortune de la femme mariée, Paris, Flammarion, 1987.
-
[22]
On the "cadres" group, see Annex 2.
-
[23]
For a detailed study of these selection phenomena, see the original French version of our paper : M. Bozon and F. Héran, "La découverte du conjoint I – Evolution et morphologie des scènes de rencontre", Population, 6, 1987, pp. 943-986 ; and "La découverte du conjoint II – Les scènes de rencontre dans l’espace social", Population, 1, 1988, pp. 121-150.
-
[24]
F. de Singly, op. cit., pp. 27-28.
-
[25]
The interpretation of this model remains the same when areas of the marriage-tables that are distant from the diagonal are considered ; it is wrong to see strategies in all heterogamous marriages. In both cases, the desire for identifying the "invisible hand" which rules the marriage market is satisfied without great outlay.
-
[****]
For a more detailed presentation of French socio-occupational categories, see Mirjana Matejovic-Scott, "Notes on the French" New Classification of Occupations and Socio-Occupational Categories"", ILO Statistical Bulletin, 1985, 3, pp. 24-30 ; or Alain Desrosières and Laurent Thévenot, Les catégories socio-professionnelles, Paris, La Découverte, 1988.
