Delphy: Continuities and Discontinuities in Marriage and Divorce


A foundational work of materialist feminism, Christine Delphy's Close to Home: A Materialist Analysis of Women's Oppression is out now in a new edition as part of Verso's Feminist Classics series. 

Below, we present Delphy's "Continuities and Discontinuities in Marriage and Divorce," first published in the 1976 anthology Sexual Divisions and Society: Process and Change, edited by D. Leonard Barker and S. Allen, and included in Close to Home. Translated by Diana Leonard.

All the books in our Feminist Classics series are 40% off until October 2ndClick here to activate your discount. 

Studies devoted to divorce in the past have presented it as the sum of individual divorce situations, they have not defined it (e.g. Goode 1956; Kooy 1959; Chester 1973). This is doubtless because the definition of divorce and its sociological significance are taken for granted; divorce means the breakdown and failure of marriage. These are the words used by the individuals concerned and sociologists have implicitly approached the problem from the same point of view. Even if they have apparently (but not always) refrained from direct value judgements and emotionally laden terms such as ‘failure’, they have still considered that the definition of divorce as the end of marriage, its revocation, or as the opposite of marriage, was a satisfactory one.

By contrast, a great deal of attention has been paid to the individual causes of divorce, and here it is evident that sociologists have not limited themselves to the reasons advanced by the protagonists, nor to their psychological ‘motivations’, but have included in their studies more objective data: for instance, social characteristics such as class origin and educational level. They have, however, always directed their attention to the ‘couple’ or the individual union. This method may have enabled them to pinpoint the differences (if indeed there are any) between couples and/or individuals who are divorced and those who are not; but it cannot teach us about the institution of divorce, for this is not just a multitude of individual accidents.

Were a similar method of analysis to be applied to marriage as has been with divorce (and indeed this has  unfortunately often been the way sociologists have approached marriage, unlike anthropologists) we would look for – and would in all probability find – differences between married and non-married individuals. But marriage is an institution and merely to look at those who enter or leave it, cannot shed light on the institution or why it exists. Similarly with divorce. Divorce is an institution which follows certain rules, it is codified and subject to control, ranging from implicit but unformulated social control to penal control.

Furthermore divorce is organically related to the institution of marriage. In an old American film the heroine asks what the grounds for divorce are in the state where she lives, and the lawyer replies, ‘being married’. But I would go further and argue that not only is marriage the necessary condition for divorce; but also that divorce is not inconsistent with marriage. For while a divorce signifies the end of a marriage (marriage meaning here a particular union), it by no means implies the end of marriage as an institution. Divorce was not invented to destroy marriage since divorce is only necessary if marriage continues to exist. Indeed, it is often argued that the increase in the incidence of divorce can be interpreted as proof, not that the institution of marriage is sick, but on the contrary that it is thriving.

Further, divorce reveals and throws into relief certain institutional aspects of marriage, and it makes clear what is otherwise latent. Conversely marriage sheds light on divorce. Not only do certain aspects of marriage make the institution of divorce more intelligible; what is more noteworthy is that they are carried over and perpetuated in divorce.

The institution of marriage is, of course, complex and it is imperative to specify which aspect and which function is being studied. This paper will focus attention exclusively on the economic aspect of marriage, and to make clear what this means, I will first summarize briefly the approach that is used. 

A theory of marriage

My proposition is that marriage is the institution by which unpaid work is extorted from a particular category of the population, women-wives.1This work is unpaid for it does not give rise to a wage but simply to upkeep. These very peculiar relations of production in a society that is defined by the sale of work (wage-labour) and products, are not determined by the type of work accomplished. Indeed they are not even limited to the production of household work and the raising of children, but extend to include all the things women (and also children) produce within the home, and in small-scale manufacturing, shopkeeping, or farming, if the husband is a craftsman, tradesman, or farmer, or various professional services if the husband is a doctor or lawyer, etc. The fact that domestic work is unpaid is not inherent to the particular type of work done, since when the same tasks are done outside the family they are paid for. The work acquires value – is remunerated – as long as the woman furnishes it to people to whom she is not related or married.

The valuelessness of domestic work performed by married women derives institutionally from the marriage contract, which is in fact a work contract. To be more precise, it is a contract by which the head of the family – the husband – appropriates all the work done in the family by his children, his younger siblings, and especially by his wife, since he can sell it on the market as his own if he is, for example, a craftsman or farmer or doctor. Conversely, the wife’s labour has no value because it cannot be put on the market, and it cannot be put on the market because of the contract by which her labour power is appropriated by her husband. Since the production intended for exchange – on the market – is accomplished outside the family in the wage-earning system, and since a married man sells his work and not a product in this system, the unpaid work of women cannot be incorporated in the production intended for exchange. It has therefore become limited to producing things which are intended for the family’s internal use: domestic services and the raising of children.

Of course, with the increase of industrial production (and hence the number of wage-earners) and the decrease in family production, many women-wives now work for money, largely outside the home. They are none the less expected to do the household work. It would appear that their labour power is not totally appropriated since they divert a part of it into their paid work. Yet since they earn wages they provide their own upkeep. While one could, with a touch of bad faith, consider the marriage contract as an exchangecontract when women work only within the household, with married women providing domestic work in exchange for upkeep, when married women earn their own living that illusion disappears altogether. It is clear then that their domestic work is given for nothing and the feature of appropriation is even more conspicuous.

However, the modes of appropriation differ depending on whether the woman has a paid job or not. When she does not, her total work power is appropriated, and this thus determines the type of work she will do – if her husband is a doctor she will make appointments for the patients; if he has a garage she will type the bills, etc. It also determines the nature of the relations of production under which she operates – her economic dependency and the non-value of her work – for while she may accomplish exactly the same tasks as her well-to-do neighbour, the upkeep she receives will be different if her husband’s financial status is not as good. When she has a job, however, she recuperates part of her labour power in exchange for the accomplishment of a precise and specific type of work: housework. Legally any woman can now choose the second solution, although in France the law requiring a husband’s authorization for his wife to work outside the home was abolished only some ten years ago. In point of fact, however, it seems reasonable to suggest that the only women who work outside the home are those whose husbands give their consent if they consider that they do not need all their wife’s time. Equally, in France, the obligation to do housework is not written in any law; all that is said in the Code Civil is that the wife’s contribution to the ‘household charges’ can be in kind if she has no dowry or independent income. But this obligation is inscribed negatively, so to speak, in the sense that failure to assume it is sanctioned.

Some of the possible sanctions are social worker intervention or divorce (see Dezalay 1976). When social control agents intervene, whether it be in the person of the children’s judge, the social worker, or the court, and if a divorce ensues or the family budget comes under the control of the social workers, the obligations of marriage are officially expressed and in particular the differential duties of the husband and the wife. This precision and differentiation contrasts markedly with the vague legal formulation of marriage contracts, which suggest an apparent reciprocity in the respective duties of the partners (notably the wife’s contribution in kind and the husband’s in money are represented as having the same value and producing a similar status for both partners).

Conclusions following from this theory of marriage

It is clear that the position of women on the labour market and the discrimination that they suffer, are the result (and not the cause as certain authors would have us believe) of the marriage contract as we have described it.2

If we accept that marriage gives rise to the exploitation of women, then it would be logical to suppose that pressure is brought to bear on women to persuade them to marry. Of course there are various sorts of pressure – cultural, emotional-relational, and material-economic – and one could argue that the last is not the most important, or that it is not perceived as a pressure at the time of marriage, or that it is not operational at this time. However, if we compare the standard of living to which a woman can aspire if she remains single and the standard which she can reasonably expect from being married, it seems certain that relative economic deprivation will be experienced by single women as time goes on. We are confronted with a paradox: on the one hand marriage is the (institutional) situation where women are exploited; and on the other hand, precisely because of this, the potential market situation for women’s labour (which is that of all women, not just those who are actually married (see Barron and Norris 1976)) is such that marriage still offers them the best career, economically speaking. If the initial or potential situation is bad, it will simply be aggravated by the married state, which becomes even more necessary than ever. The economic pressure, in other words, the difference between the potential ‘single’ standard of living and the actual ‘married’ standard of living, simply increases as time goes on.

Marriage as a self-perpetuating state

When women marry or have a child they often stop working or indeed studying; or even occasionally among the middle class – the American model is becoming general in France 3– they stop studying in order to put their husband through college, by means of a job that has no future, and they stop working as soon as their husband has obtained his degree. If they continue working, they do so at the cost of enormous sacrifices of time and energy, and even then they are still not as free to devote themselves to their work. As a result they cannot aspire to the promotion which they might have had if they had not had to look after a husband and children materially as well as themselves. Ten years after the wedding day, marriage is even more necessary than before because of the dual process whereby women lose ground or at best remain at the same place in the labour market, while married men make great progress in their work as they are not hampered by household obligations. Of course, individual husbands are not responsible for this situation, but all men benefit from a situation that is taken to be normal. A ‘normal’ day’s work is that of a person who does not have to do his own domestic work. But even though this is the norm, it is none the less made possible only by the fact that the household tasks are assumed by others, almost exclusively by women. It is evident that the career of a married man must not be compared for our purposes with that of other men, but with the life he would have led if he had remained single, or if he had had to share the household tasks including the raising of children. This dual process is particularly evident in the case where the wife gives up her own studies in order to finance her husband’s. Here, even though both begin in more or less the same position (not taking discrimination into account), marriage results in the wife moving down the economic ladder and the husband moving up, and these changes combine to create an important gap between the economic possibilities of the two partners.

Thus it can be said that, from the woman’s standpoint, marriage creates the conditions for its own continuation and encourages entry into a second marriage if a particular union comes to an end.

In this respect statistics are ambiguous, or, more precisely, are difficult to interpret. There are generally more divorced women at work than married women (annual statistics from the Ministère de la Justice 1973). This could be taken as confirmation that their economic situation, notably the absence of an independent income, discourages full-time housewives from getting divorced. But on the other hand many women begin to work just because they face a divorce – they start the moment they decide to get a divorce, long before the decree is issued. This explains why they are registered as ‘working’ at that particular time. Having a job enables some women to envisage divorce, while others in the same situation but lacking a job have to ‘make a go’ of their marriage. A large number of women who are divorced or about to be divorced come on the labour market in the worst possible conditions (as do widows), with no qualifications, no experience, and no seniority. They find themselves relegated to the most poorly-paid jobs. This situation is often in contrast with the level of their education and the careers they envisaged, or could have envisaged, before their marriage, the social rank of their parents, and not only the initial social rank of their husband but, more pertinently, the rank he has attained when they divorce, some five, ten, or twenty years after the beginning of their marriage. In addition, those with dependent children have to look after them financially, and this new responsibility is added to the domestic work which they were already providing before divorce. For the majority of women, the contrast between the standard of living that they enjoy while married and that which they can expect after divorce simply redoubles the pressures in favour of marriage or remarriage depending on the circumstances.

The state of divorce as a continuation of the state of marriage

The fact that the material responsibility for children is assumed by the woman after divorce confirms the hypothesis concerning the appropriation by the husband of his wife’s work, but it suggests as well that the appropriation which is a characteristic of marriage persists even after the marriage has been dissolved. This leads me to contend that divorce is not the opposite of marriage, nor even its end, but simply a change or a transformation of marriage.

At the beginning of a marriage this appropriation is legally masked; it is a matter of custom in the sense that the legal framework which underlines it is vague and unused and even useless. It only begins to operate – by means of the intervention of the judicial system – when the marriage comes to an end. Even then its apparent purpose is not to burden the wife with the entire responsibility for the children nor to exempt the husband totally. It permits such an outcome, but by omission rather than by a positive action. There is positive action, however, in the official guideline of considering ‘the child’s interest’.

Unofficially custody of the children is considered to be a privilege and even a compensation for the woman who may be left badly off in other respects. A real battle is staged to make the two spouses turn against each other and to keep them uncertain as to the outcome of the conflict for as long as possible. The custody of the children becomes the main issue 4, and at the end of the battle the spouse who obtains this custody considers that he or she has won the war. But in fact when the children are young they are almost always entrusted to their mother. Officially both parents share the responsibility for the cost of looking after the children, but the woman’s income after divorce is always very much lower than that of her former husband, and the allowance for the children decided by the courts is always ridiculously low.5The woman’s financial contribution is thus of necessity greater in absolute value than her husband’s, even though her income is lower. As a result her participation and her sacrifices are relatively much greater. Furthermore, 80 per cent of all allowances are never paid (Boigeol, Commaille, and Roussel 1975). Even if the official directives are respected and the allowance is paid, the amount agreed never takes into account the woman’s time and work in the material upkeep of the children.6

Thus the courts ratify the exclusive responsibility of women both by positive actions, granting custody to the mother and assigning a low allowance for the children; and by negative action, failing to ensure the payment of the allowance. The ‘child’s interest’ makes it imperative for him or her to be entrusted to his or her mother, be she poor, ‘immoral’, or sick, as long as he or she requires considerable material care: as long as there are nappies to wash, feeds to prepare and special clothes, toys, medicaments, lessons, etc. to pay for. As soon as the child reaches the age of 15 the courts usually regard the father more favourably than the mother:7 she is thought to be unable to provide the child with as many advantages as the father, who is better off (for very good reasons). A child who has been entrusted to his or her mother can then be handed back to his or her father, again in the ‘child’s interest’. And yet, curiously enough, this aspect of the child’s interest – the parent’s wealth – did not come into play when the child was younger. Objectively the child’s interest has served to make his or her mother poorer and his or her father richer 8, creating thereby the conditions in which it will be ‘in his/her interest’ later on to return to the father.

Two conclusions can be drawn: in divorce, as in marriage, the work involved in raising children is carried out by the woman unpaid and the husband is exempted from this charge as part of the normal process. Furthermore, the financial care of the children, which was shared by the couple or assumed by the husband alone in the marriage, is thereafter assumed predominantly or exclusively by the woman.

In compensation the woman no longer has to carry domestic responsibility for her husband. This casts a special light on the marriage contract. Indeed, when the married state is compared with the official as well as the real divorced state, it becomes clear that the material responsibility for the children is the woman’s ‘privilege’ in both cases; while in marriage, in contrast to divorce, the wife provides for her husband’s material upkeep in exchange for his contribution towards the financial upkeep of the children.

Marriage and responsibility for children: a question of theoretical antecedence

An overriding concern in this paper so far has been to rethink the economic aspects of the institution of marriage and to give them the definition that they have lacked. Comparing marriage to divorce, it seems that the material upkeep of the husband by the wife is related to the participation of the husband in the financial upkeep of children. This provides grounds for viewing marriage differently. This approach is consistent with the contention that whereas marriage sheds light on divorce, the reverse is also true. So far this has meant only that divorce reveals the nature of the marriage contract, but it can also be taken to mean that divorce can shed light on what made this contract possible in the first place.

I contend that these conclusions allow us to see childcare (from the analytical not the empirical point of view) as separate from the rest of domestic work. The obligation of childcare may have to be viewed as not so much perpetuating the husband’s appropriation of his wife’s labour, as making it possible in the first place. Or, to put it slightly differently, these conclusions compel us to consider the possibility that the continuation of the obligation of childcare is a continuation of the marriage contract, in so far as the appropriation of the wife’s labour includes the obligation of childcare; but that this obligation, while carried out in marriage, does not necessarily stem from it; that it might be antecedent to it, and might even be one of the factors that makes the appropriation of wives’ labour – the free giving by them of the rest of housework – possible.

If marriage is considered as giving rise to the appropriation of the women-wives’ work, the position of married women who work outside of the home suggests that this total appropriation can be transformed into a partial appropriation, bearing no longer on their time or work power as a whole but on a specific task, the household work, that can eventually be replaced by an equivalent sum of money.9This evolution of the system of appropriation of wives’ labour may at first sight call to mind the evolution of the appropriation of the labour of slaves between the Roman Empire and the late Middle Ages. The appropriation by the seigneur of the slave’s total work power became a partial appropriation, approximately half of his time, three days work per week (Bloch 1964), when the slave became a ‘serf’ and was ‘settled’. He then worked part-time for his own profit on a piece of land which he rented from the seigneur. The time debt to the siegneur was later itself transformed into the obligation to accomplish a specific task, the corvée, which later on could be commuted into a money payment.

However, this way of formulating the problem is perhaps false because the partial appropriation of the married woman’s labour on this analogy should be counterbalanced by the woman partially recuperating her work power, when in fact she pays for the freedom to work outside, and to have an independent income, with a double day’s work. It cannot be said that she recuperates either a period of time or a value. On the other hand she does partially escape from a relationship of production characterized by dependency.

Furthermore, if marriage as a state is characterized and differentiated from divorce by the ‘contract’ of appropriation, marriage and divorce can be considered as two ways of obtaining a similar result: the collective attribution to women of the care of children and the collective exemption of men from the same responsibility.

Seen from this angle, not only the married and the divorced states but also the state of concubinage, in short all the situations in which children exist and are cared for, have similar characteristics and are different forms of one and the same institution, which could be called X. The situation of the unmarried mother can be taken to be its extreme form, and at the same time its most typical form, since the basic dyad is the mother and child. Marriage could be seen as being one of the possible forms of X, in which the basic couple is joined by a man who temporarily participates in the financial upkeep of the child and in return appropriates the woman’s labour power.

This view is similar to that of those anthropologists (Adams 1971; Zelditch 1964) who criticize Murdock (1949) and say that the family defined as a trio proceeding from the husband and wife couple (taken to be the fundamental dyad) is not a universal type, whereas the mother-child association is. This point of view may become a new element in the study of western societies, where it has generally been taken for granted that the family is patrifocal. This element may be new, but it is not contradictory; for if the family, considered as the place where children are produced, can be viewed as matrifocal, even in our own societies, it remains none the less true that as an economic production unit (for exchange or for its own use) it is defined, as during the Roman era (Engels 1884), as the group of relatives and servants who give work to the head of the family: the father.

Going a step further, the state of marriage-with-children appears as the meeting place for two institutions: on the one hand the institution relating to women’s exclusive responsibility for child-care, on the other the institution relating to the appropriation by the husband of his wife’s labour power.

Indeed if one considers marriage alone, it appears that the care of children, their upkeep, which is no different from the material upkeep of the husband by the wife and which is carried out in the same manner – the execution of work in exchange for maintenance (financial upkeep) – partakes of and flows from the appropriation of the wife’s labour power by her husband. As long as there are two parents it can be postulated that the children, in accordance with the legal terms, are their common property, possession, and responsibility. In this case, in the marriage situation half the work involved in the upkeep of the children is appropriated by the husband-father, and continues to be so after the divorce. But children do not always have two owners. In the absence of the father, their upkeep by the mother, or even half of this upkeep, is obviously of no benefit to any particular man. Besides, even in marriage or divorce it is doubtful whether the parents are the only ones, excluding society as a whole, to benefit from the children, and consequently it is not at all certain that the husband-father should be considered as the only one to benefit from his half of the work involved in looking after the children, or as the only one to appropriate his wife’s work, since he does not carry it out with her. If this is accepted, then the raising of the children will have to be considered apart from the woman’s family work (household or other) and the exclusive responsibility of women concerning the children will have to be treated as a relatively autonomous institution with respect to marriage.

If the relationship between marriage and divorce is viewed in this way, it appears slightly differently from what was suggested at the beginning of this paper. The husband’s appropriation of his wife’s work then ceases, in part or completely, as soon as the marriage comes to an end (depending on whether or not the husband is considered as continuing to benefit from the children, and from their upkeep, either partly or not at all). In this view divorce is not the continuation of marriage. However, the situation after divorce, in which the responsibility for the children is an important aspect, constitutes a strong economic incentive to remarriage for women.

When there are children, the responsibility for their care continues to be borne exclusively by the woman after divorce, and this burden is increased by the financial cost. However, rather than considering that this illustrates a continuation of the husband’s appropriation of his wife’s work, it would now seem more exact to say that it illustrates a new form of women’s responsibility for children, which exists before the marriage, is carried on in the marriage, and continues afterwards. This responsibility can be defined as the collective exploitation of women by men, and correlated with this, the collective exemption of men from the cost of reproduction. The individual appropriation of a particular wife’s labour by her husband comes over and above this collective appropriation. It is derived from, or at least made possible by, the collective appropriation which acts in favour of marriage, since if the husband appropriates his wife’s work power, in return he contributes to her financial upkeep and the children’s, and in this way he ‘lightens’ her burden by partially assuming a responsibility from which society exempts him. In other words, the institutional exemption from which he benefits allows him to claim his wife’s total labour power in exchange for his contribution to the children’s financial upkeep. 


1. I use the expression woman-wife to stress that the one is a person and the other a role. This ontological distinction is blurred by the fact that the social role is so widely associated with a biological category that they have become equivalent.

2. The thesis of Blood and Wolfe (1960), for example, is that no model exists, let alone a patriarchal one. If more married women do the housework than married men, it is because they have more time to do it and their husbands less since they work outside (!). And if married women are of less weight in making decisions, it is owing to the fact that since they do not work outside (this being compensated by the extra time they have to do the housework) so their contribution to the domestic economy is less important.

3. See, for example, couples where the husbands are at business school (Marceau 1976).

4. This is a legal notion which officially denotes official responsibility and, unofficially, the right to dispose of and enjoy as one may dispose of and enjoy any possession.

5. In a study I was involved in, we found in one provincial court that the ex-wife was awarded a mean of £10 per month per child. In general, courts in France will never instruct the ex-husband to pay more than one-third of his income to his ex-wife and children.

6. I distinguish the financial and material upkeep of a family. The first is the part of the consumption that is bought. The second consists of services, or labour applied to goods bought by the wage.

7. This is based on statistics from the Ministère de la Justice (1973) and oral communications from a lawyer.

8. That this is a mere legal fiction is clear if we consider the result to which it leads, and that from the very beginning it is the judges and not the children who talk of their ‘interest’.

9. When for example the woman buys off her obligation by paying for a nurse or a public nursery, etc. out of her salary.

The Feminist Classics edition of Close to Home: A Materialist Analysis of Women's Oppression is out now. All the books in our Feminist Classics series are 40% off until October 2ndClick here to activate your discount.