1Annabel Desgrées du Loû, Michèle Ferrand, Hermann Brou, Annick Tijou-Traore, Nathalie Bajos

2A pregnancy that ends with a live birth is a moment of passage. The first pregnancy is a particularly important one; it brings a woman to motherhood and so can be a major marker of the passage to adulthood. But becoming a mother (or having another child) is not the only transition a pregnant woman goes through. The end of the pregnancy also involves several “stop-resume” processes that have been less often studied and described but can have important consequences, especially for her sexual life and fertility. A woman who has just given birth is not fecund and does not immediately start to have sex with her partner again. She will go through a succession of transitions that signal the return of fecundity and sexual activity. Her economic activity follows the same path: a pause in late pregnancy and a return to work after the birth, though the exact pattern varies.

3This chapter examines these different transitions.

4Pregnancy, although it concerns a woman first and foremost, usually involves a couple (of whatever kind − stable or not, legitimate or not, etc.). So analysing the factors around pregnancy is more complicated than analysing transitions that affect just one person. The researcher must take into account the complex states of two individuals who interact closely with each other. Their interests in the matter may coincide, as with a stable couple who both want a child, or a temporary liaison where both agree that theirs is just a passing affair. In many other cases, their interests may differ.

5To study the transitions connected with pregnancy, the researcher must take into account the prevailing system of norms regarding sexuality and reproduction. These systems vary widely from one society to another. While the separation between sexual activity and reproduction has been one of the great changes in industrialized societies in recent decades, in most other societies sexual activity is still closely bound to reproduction. So today there are two different systems of norms. In Southern societies motherhood at an early age is valued and marks a girl’s passage into adulthood. In Northern societies the first pregnancy and maternity are only one stage in the passage to adulthood. Others are economic independence, in many cases the move away from the family home, and the formation of a stable couple, though not necessarily a marriage (see Chapter 5 on the passage to adulthood). The pattern of an individual’s transition depends on the way the various stages are linked in their particular society, which may have a single system of norms or may allow for a wide range of possibilities.

6This chapter draws on experience in both Northern and Southern countries to highlight the diversity of the norms and systems regarding pregnancy and their implications for the modalities of the different passages involved. It has been written by two teams in collaboration, one working on France (see Appendix, Box 2) and one presenting work conducted in Abidjan, Côte d’Ivoire, where fertility rates are still high although they have been falling for the past 15 years (according to the Demographic and Health Survey (DHS) of 1998 the total fertility rate (TFR) in that year was 3.5 children per woman). Fertility ideals are also high: in 1998 the ideal number of children declared was 4 children on average (1998 DHS) (see Appendix, Box 3).

1. Pregnancy: Planned or Involuntary? Complex Perceptions of the “Right Moment” to Fall Pregnant

7It is not easy to define what is meant by a pregnancy that is wanted, expected and looked forward to, and one that is unintended or unwanted. The very definition of how a pregnancy is perceived depends on the prevailing norms with regard to pregnancy and varies from one society to another.

8In some sociocultural contexts, pregnancy is something that can happen as a result of any sexual intercourse event. It is seen as likely to happen any time a couple has sexual intercourse. But Western societies, over and over, reaffirm that sexual activity and reproduction are separate things, and that procreation depends on the explicit will of the sexual partners. The GINE survey (see Appendix, Box 2), which looked at unintended pregnancies (Bajos, Ferrand and the GINE group, 2002), showed the extent of polysemy in the term imprévu (unforeseen, unexpected) itself. The interviews show that some women whose pregnancies were not entirely unintended, said for tactical reasons that they were so. In some cases this was to make their parents accept a relationship they had so far refused to accept. In others it was to attain motherhood status even in the knowledge that this birth would be socially disapproved of and so could not be presented as a willing choice, or planned. In others again it was to test the dependability of a partner or a relationship by stopping contraception. The explanation for these atypical responses must be sought in the particular life course of each of the women concerned.

9The same polysemy shows up in the Cocon quantitative survey (see Appendix, Box 2), where it emerges that the sociodemographic, relational and contraceptive profile of women who said their pregnancy was “not at all expected” differed from those of women who said they would have wanted to fall pregnant, “but later”. The first group were mainly women in situations where their pregnancy was not in line with prevailing norms regarding the proper conditions for having a child. Many were women whose partner had been a casual one, or whose relationship was just starting, whereas most women in France today agree that a couple ought to be stable before planning a baby. Also in the group who said their pregnancy was completely unexpected were women who thought they were too young or too old to have a child, and women who felt vulnerable in some other respect (financially, as regards work, etc.). But women in stable relationships, who had not yet had all the children they wanted and were in a favourable situation to have their first baby or an additional one, were more likely to say their pregnancy was intended, but had happened too soon, if they were regularly using contraceptives. Putting women in both these groups in the same category of “grossesse imprévue” is not an obvious choice, but it does avoid limiting the “unexpected pregnancy” category to those women who absolutely did not want to become pregnant. Bajos et al. (2004) recently showed the strength of the norm that births should happen only in the best conditions for the child. The steady rate of induced abortions in France can be explained by a combination of two contrary trends: a drop in the number of unintended pregnancies due to better distribution and use of contraception, and an increase in recourse to induced abortion in the event of an unintended pregnancy (Bajos et al., 2004).

10Although the French study revealed the wide mix of situations covered by the term “grossesse imprévue”, the survey in Côte d’Ivoire revealed a different ambiguity around the same term. In the Abidjan survey, a high proportion of women who fell pregnant while the cohort was being monitored said they did not want this pregnancy. And yet they had not used the contraceptive means that were offered free of charge by the monitoring team (Desgrées du Loû et al., 2002). This raises a different kind of question. In a country where fertility is highly valued, might being pregnant be a way for a woman to reassure herself that she is fecund, even if she then has an abortion because the child is not wanted? In other words, perhaps wanting a child and wanting a pregnancy are two different things.

2. From the Personal to the Conjugal

11The above section mentions that a women’s perception of her pregnancy is closely dependent on her conjugal situation as well as her personal situation. To analyse the transitions that occur in relation to pregnancy, especially as regards sexuality and fertility, one must take into account not only individual life courses but also a higher level of complexity: the couple.

2.1. “The Right Moment to Have a Baby”: Considering the Social and Family Circle

12Regarding the decision to have a baby, one must consider what the woman, her partner, her family and the society she lives in consider the right time for a child to be born. In Southern societies, in Africa particularly, early childbearing is the norm. Becoming a mother is still the landmark that gives a woman a social status and social value. The first child is usually born very soon after the marriage, and may precede it if the young couple’s formal union has to be postponed for economic reasons (Delaunay, 1994). In France and other Northern societies, contraceptive use is increasingly the norm (Bajos and Ferrand, 2004) and women are expected to think about the right time for becoming a mother. While contraception and abortion allow them to “decide”, men and women have different views of the right age for becoming a parent, as the social and biological constraints are not the same for men and women.

13Whatever her situation, a woman must think about pregnancy and childbearing from three angles:

• Her own situation. In France, women must consider the right “social” age for becoming a mother: not too soon, not too late. She must also perform a juggling act between work and family responsibilities. Pregnancy while still at school or in higher education is considered to endanger the completion of studies, but interrupting a career or finding herself unemployed after the baby arrives is less limiting for the mother than for the father. Pregnancy and studies are equally incompatible in Côte d’Ivoire, where students are the most frequent users of abortion services, to avoid having to interrupt their studies (Guillaume and Desgrées du Loû, 2002). But in France and Côte d’Ivoire alike, young women who have not invested much in studies or a career may see early childbearing as a way of achieving social status and a social identity.
• The partner’s situation. As a rule, in all societies a man is expected to have the means to feed a child before he becomes a father. This means he is expected to have joined the workforce. In France that means not being a student, or unemployed, and preferably having a stable job (rather than a temporary job or a fixed-term contract). In Abidjan, fatherhood is as socially desirable as motherhood and the man’s desire for a child is often cited as the reason for a pregnancy. It is common for women who already have several children from a first spouse and do not want any more to get pregnant again because their new spouse has no children of his own, or thinks he does not have “enough”.
• The couple’s situation. The couple’s material and affective conditions, and above all their social existence, are taken into account. A casual or too recent relationship is not considered suitable for making a baby. The contraceptive norm in France highlights the extent to which women do not have children just for themselves. It is not so much the stigma of unmarried motherhood that stops them as their conviction − which they share with the rest of society − that a child needs two parents. In West Africa, the type of conjugal relationship is also decisive in the decision to have an additional child or not: while spouses rarely discuss sexuality, they do discuss how many children they want to have (Andro, 2001). Having children, and having enough of them, is a factor in keeping couples together. In polygamous households, the co-wives compete to have children so as to enjoy the husband’s favour (Antoine, 2002).

2.2. “Fuzzy Conjugality”: the Evolving Reality of the Couple

14Taking account of the couple as a social entity is very relevant for understanding how women handle things when they want to fall pregnant, but conjugality is a fuzzy state with ill-defined boundaries (see Chapter 2 on unions and separations). A couple is an evolving reality and can change even during the course of a survey, without the changes involved being made explicit. In the Abidjan survey on behaviours in face of the risk of HIV/AIDS, women were asked to talk about their regular sexual partner and their casual partner or partners, if any. The quantitative survey sought to establish the sociodemographic characteristics of the regular partner (age, educational level, cohabitation with the respondent or not), whether and how the couple discussed sexual matters (AIDS, STIs, protected sex), the partner’s sexual behaviour (sexual relations with other women if any). If the respondent said she had a “casual” partner, she was only asked whether sex with this partner was protected or not. But the qualitative survey conducted alongside this one showed that a partner described as “casual” at the start of the survey sometimes became a regular partner, although the quantitative survey had no way of recording such a change of status in the same relationship.

15So if a respondent (a woman, in this instance) is interviewed on more than one occasion, it is important to check whether, when they speak of their partner, they are talking about the same one as last time, and if so whether their status has changed. To avoid any confusion about partners’ identities, this check must be made systematically, at each stage of gathering data (quantitative or qualitative) about the nature of the relationship between the respondent and each partner they mention.

16Who are they talking about and what exactly is the status of their relationship with that person? It is also essential to identify the event or events behind the couple’s transition to a new state, not only to make sure the change of state is properly understood, but also to reveal the meaning of the couple’s restructuring. To search for the meaning of changes in the conjugal relationship, the qualitative approach seems the more appropriate, as it gives more space to what the respondents have to say, in their own words.

2.3. Understanding the Reproductive Decisions of HIV-Positive Women. Reflections Based on the Abidjan Example

17Where one or other partner in a couple is infected with HIV, and knows it, the situation surrounding decisions about childbearing is even more complicated. The AIDS virus, HIV, can be transmitted during sexual intercourse or from mother to infant during pregnancy, during the birth or via breast milk. So when a person finds they are infected, in principle this should lead them to rethink their decisions about sexuality and reproduction. As HIV is a life-long infection, decision-making moments will arise again and again during the life course, and this can have serious consequences for the future of the couple. Because of this, women who discover they are HIV-positive are torn between conflicting desires. One is to protect sexual intercourse so as to avoid falling pregnant or infecting their partner and/or baby. The other is to conceal their HIV status from their family, social circle and partner and to start a new pregnancy as soon as they can, to “prove” that they are in good health and to consolidate their couple.

18The HIV infection rate in Abidjan is high (about 11% of adults − Msellati et al., 2006, screening is not available to all (Desgrées du Loû and de Béchon, 2001) and those screened are mainly pregnant women who are offered the test under a programme designed to reduce mother-to-infant transmission of the virus.

19So in most cases it is the woman who learns of the infection first. And it is near the end of their pregnancy that infected women find out they are HIV-positive. The question of conceiving again, or not, will arise a few months later.

20During the Ditrame programme, which ran from 1995 to 1999, a high rate of new pregnancies was observed among these HIV-positive women in Abidjan. The medical teams monitoring the women advise them to avoid any further pregnancy because of the risk of transmitting the virus to the baby. But few women follow this advice. In Abidjan between 1995 and 1999, there was a high incidence of new pregnancies in the two years following the positive test and the consequent advice. Yet these women already had at least one young child and had access to free condoms and contraceptives. The same result was observed in Rwanda (Allen et al., 1993) and among a group of African women in France (De Vincenzi et al., 1997). Qualitative analyses explained this paradoxical finding. Far from abolishing the desire to get pregnant, the knowledge that they were HIV-positive strengthened it in these women, at least among those who thought they did not yet have “enough” children (the number of children considered to be “enough” was about four in Abidjan (DHS, 1998)). A new baby was synonymous with life and a future, in contrast to the “threat of death” that HIV represents (Aka Dago et al., 1999). Moreover, even if they know they are infected with HIV, women are expected to take into account their partner’s desire for children, that of their mother-in-law and even their own mother’s.

21In Africa, even more than elsewhere, the decision to have a child is rarely one that a woman can take on her own.

22For a woman carrying the AIDS virus, apart from the number of children already in the household, two factors will be decisive in the decision to have another child or not. One is whether or not she has told her partner about her infection. In Abidjan, fewer than half the HIV-positive women dared to tell their spouses, for fear of a negative reaction (Desgrées du Loû, 2005). For those who did tell their spouses, this seems to have made it easier to negotiate over the initial procreation project, and in some cases the men seemed readier than the women to reconsider it (Tijou-Traore, 2006a). The second decisive factor is the state of health of the couple’s existing children. If one of the children is HIV-positive and is sick, the mother is very much taken up with the struggle against her child’s illness and also, if the child was infected by mother-to-infant transmission, feels very guilty. The desire for another child is much less frequent in cases like this than in families where all the children are healthy (whether HIV-positive or not) (Aka Dago and Cacou, 2004).

23To take better account of the complexity of these situations (and in this precise case, of the observation of women who have discovered they are HIV-positive at a pre-natal clinic and are then monitored for several months in a cohort study), one must take into consideration what can literally be called a psychological transition. After the shock of the announcement that she is infected with HIV, it takes some time for a woman to accept the new situation. This period varies in length from woman to woman. It is only after this transition period, which sometimes involves denial of the infection, that these women will really take their infection into account when making decisions about their lives. Telling the spouse is a parallel factor which may either speed and facilitate the transition to acceptance or make it more difficult. Although only psychologists can study this psychological transition as such, it should be taken into account in interpreting the choices made and the behaviours adopted. This additional dimension may lead researchers to take a different view of time.

2.4. What Information Should be Gathered to Grasp the Complexities of a Conjugal Situation?

24Where a decision is taken about whether to start, or keep, a new pregnancy, a major key to understanding the behaviours adopted is the type of relationship between the partners in the couple. Whatever the society in which the survey is conducted, to analyse the processes surrounding pregnancy it will be useful to gather information not only about the man and woman concerned but also about the precise context of their social, family and sexual lives.

25Communication between the partners seems to be an important factor for understanding their conjugal relationship, elucidating their decision making on reproductive issues and understanding the resulting behaviours (Tijou-Traoré, 2006b). Dialogue between spouses is a component of the conjugal tie and to varying degrees enables them to negotiate agreements and take into account each others’ complexities. It can also be considered a vector of transitions.

26In other words, depending on whether it is conceivable, desired, impossible or unthinkable, and depending on the form it takes, dialogue within the couple plays a part in the passage from one state to another (or others). Researchers should try to gather the necessary data for analysing whether there is discussion/negotiation over the decision to become pregnant again and if so what forms it may take. This means gathering not only information that reveals the respective positions of the two partners (particularly through the capital and resources they can mobilize), but also information from which to establish whether the two partners discuss these matters and, if so, how.

27Part of this information is already collected in the surveys on sexual and reproductive health. This includes:

• the reproductive and conjugal history of each of the partners,
• each partner’s aspirations as regards reproduction and dialogue about the desire for pregnancy,
• each partner’s behaviour in matters of sexuality and fertility (contraceptive use, sexual behaviour, etc.),
• each partner’s social and economic capital,
• the division of roles between the partners (domestic tasks, the children’s education, etc.).

28To investigate more deeply, it would be useful to find out, and in greater detail, about:

• the conjugal relationship (the history of the couple, communication between them in general, what negotiations are possible, the woman’s degree of independence from the man and the couple’s degree of independence from their family and social circle,
• any pressure on the couple (exerted by either partner’s mother, their peers, etc.),
• particular factors that may alter the couple’s relationship, for example, in the case of an HIV-positive person, discussion or lack of discussion with their partner about HIV, the partner’s attitude with regard to HIV, etc.

29Here the qualitative approach and the quantitative approach complement each other very well. The quantitative approach serves in particular to determine the influence of the number of children each partner already has, or their respective ages, etc., on the decision to have start a new pregnancy. But to take account of communication between the spouses and its impact on actual behaviours, the qualitative approach is more useful.

30As the couple is a constantly changing reality and not a stable state, the best approach to conjugal processes leading to a decision about pregnancy would be to gather the same information at intervals over time, as is possible with a cohort study. This way one can apprehend developments in the couple and their effects on decisions about procreation. In this prospective analysis it should not be forgotten that socially, men’s and women’s life courses are not constructed symmetrically. While for women, attractiveness, reproductive potential and the likelihood of marriage decline with age, men may become better placed to marry as they get older because better established socially, while still having the potential to father children. This is especially true in societies where a wide age difference between spouses is common.

3. Analysing Imbricated Processes: Seeking the Source

31Sexuality, a new pregnancy and weaning are processes that take place over time, and whose starting point is difficult to date. In this section we take three examples to show (i) that it is difficult to pinpoint the start of these processes, (ii) that several choices may have to be made and (iii) that the research topic will determine which precise event the researcher must take as the initating event of a process or change of state.

3.1. Dating the Start of a Pregnancy

32Pregnancy starts with a biological event, conception, but a woman will not know she is pregnant for some time, as is demonstrated by the fact that some women go into denial of their pregnancy, for whatever reason. In research on unintended pregnancies and the use of induced abortion, the timing of the moment when the woman finds out she is pregnant is important, since (in France) abortion cannot be carried out any later than the 14th week of amenorrhoea.

33Taking a broader perspective, whatever the society concerned, the clinical start of the pregnancy is the most relevant date for the social scientist to find out, depending on the technical possibilities of the situation (in Northern countries, pregnancy tests can be performed earlier and earlier). The date of the last period can be recorded. But this indicator may not be meaningful for the woman interviewed, especially if she has irregular periods or thinks she is protected (by contraceptives or breastfeeding). The date when the woman became aware she was pregnant seems to be essential.

34Also significant is the date the woman told her partner she was pregnant, and the interval that elapsed between these two dates. When abortion is intended, the date when the decision to abort was taken is also important, although the results of the qualitative survey show that women take this decision very quickly. The date she contacted the medical services, and the date of the abortion, can reveal how easy or difficult it is to access abortion providers − in the North the health services and in the South, often, underground networks. Every society organizes the social conditions for announcing a pregnancy. This happens very early in the North and much later in West Africa, where the family and contact circle is only told once the pregnancy is visible.

35The usefulness of gathering these data in a quantitative survey is that they enable time comparisons and point to the particular significance of each date.

36All in all, it seems most useful to record a succession of dates: the clinical date, if known, should no doubt be recorded specifically because it provides a precise measure of the duration of gestation until the birth or abortion. However, in some societies it is not easy to identify this date, or that of the last period (for example, if a woman becomes pregnant while breastfeeding and before her periods have returned). But the date when the pregnancy is announced to family, social circle or society at large (in some countries it is compulsory to declare a pregnancy) can also be meaningful as it enables interesting comparisons between societies. Pregnancy is a process whose unfolding over time is perceived by the woman in her body, and the question “When did you first feel that you were pregnant?” sheds light on the individual and social nature of this temporal process.

3.2. Dating the Return to Sexual Relations After a Birth

37For some research topics one may want to identify the exact date when a woman resumed sexual intercourse after a birth. For a study of fertility and family planning, this is the date from which a woman is again “at risk” of pregnancy (though this also depends on whether or not she is breastfeeding the baby and whether or not contraception is being used). For a programme to prevent mother-to-infant HIV transmission, and in which a woman has learned towards the end of her pregnancy that she is HIV-positive, it is the moment from which, unless condoms are used, she is once more at risk of infecting a sexual partner.

38But it appears that the dating of postpartum resumption of sexual intercourse is not a simple matter. In a survey in Abidjan in 1997-1998, new pregnancies were found among women who had answered “No” to the question “Have you resumed sexual intercourse since the birth?” in a quantitative questionnaire.

39A qualitative survey found the explanation for these apparent contradictions (Desgrées du Loû and Brou, 2005). It emerges that in Abidjan, postpartum resumption of sexual intercourse is governed by several conflicting norms. Under Islamic rules, the couple must have sex forty days after the birth, for the health of the mother and baby. But under the traditional birth-spacing rules of many West African societies, sexual relations should be avoided for as long as the mother is breastfeeding the infant (Van de Walle, 1988). Or until the infant is “big enough”. The notion of “big enough” is vague, based on imprecise indicators such as whether the infant can hold its head up, or can walk (Bledsoe and Hill, 1998). In urban West Africa, people juggle with the various traditional norms in a new situation where contraception and abortion are increasingly common (Guillaume and Desgrées du Loû, 2002) and the fear of AIDS is widespread. They use particular norms to justify their decisions, and those decisions can vary widely from one couple to another, depending on a range of demographic, behavioural and cultural factors − how many children the couple already have, the age of the partners, co-residence, existence or not of a co-wife; whether the couple use contraceptives or condoms or not; the opinions of mothers and mothers-in-law, advice from peers, religious or traditional dictates, etc. Behaviours therefore vary very widely. Some women may practice strict sexual abstinence for 18 months after a birth. Some may have intercourse once at forty days according to the Islamic tradition, then abstain again for some months and resume regular sexual activity 12 months after the birth. Some may have sex at widely-spaced intervals over several months before resuming more regular sexual activity. Women often say they have “not resumed sexual activity” as long as it is only occasional and not yet a regular activity.

40So efforts to date the resumption of sexual activity after a birth reveal a transitional phase that starts with the first sexual intercourse after the birth and ends with the resumption of regular sexual activity.

41This transitional phase may include just one or two incidences of sexual intercourse. Its duration varies from one couple to another. It may last for months or for two or three weeks. If regular sexual activity starts with the first intercourse, it is non-existent. It seems to respond to a range of different needs − to reassure one of the partners, partly satisfy an impatient partner, or comply with custom. It may also be a moment of broader negotiation between the partners, as with one woman who refused to resume regular sexual relations until her partner had split up from a mistress (Desgrées du Loû and Brou, 2005).

42In Western societies, there seem to be few dictates about the resumption of sexual intercourse after a birth or abortion (Bajos, Ferrand and GINE Group, 2002). The implicit norm seems to be that of consensus within the couple − when the woman “feels ready”. But the GINE survey showed that some women fell pregnant after a birth without wanting to, either because they thought breast-feeding protected them or because their doctor had not warned them of the risk and how to prevent it. It would seem that as regards medical prescription of the contraceptive pill immediately after a birth, doctors see their patients more as mothers than as fertile women and “naturally” downplay the risks involved in sexual intercourse, which they consider − consciously or unconsciously − somewhat premature. Here as in the above-mentioned surveys, a single sex act can result in pregnancy although the couple do not have the impression that they have resumed the rhythm of sexual activity they had before the birth.

43So it is important to gather data to identify this transitional period between the first intercourse and the resumption of regular sex. Then one can pick the most useful indicator for the particular question the research aims to answer. In one multiple-round survey among women after childbirth, to define the return to sexual activity the questionnaire included not one but two questions:

• Q1: Since the last visit [in the survey programme], have you resumed regular sexual relations? If so, how long after the birth?
• Q2: If not, have you had sexual intercourse at least once since the last visit? If so, how long after the birth?

44The question then arises as to what indicator to use: date of first intercourse after the birth or date or resumption of regular sexual activity. One must also take into account whether the sexual activity is desired by the woman or not. Here again, it is the research topic that determines the appropriate indicator. If the issue is HIV prevention, the date of the first sexual intercourse is relevant because HIV can always be transmitted during sexual intercourse. The first sexual intercourse can also be decisive for estimating the moment when infection occurred. If the issue is choosing the contraceptive method best suited to a woman’s needs, the two indicators should be combined: the contraceptive method best suited for occasional sex acts is not the same as that for someone who has intercourse on a regular basis.

45Prospective surveys (or cohort studies), though costly and difficult to conduct, are ideal for identifying isolated incidences of sexual intercourse and for assessing how “regular” the intercourse is. These surveys minimize the biases caused by forgetting and provide a series of observations over time, which is necessary for any study of regularity.

46The qualitative approach is also important for studying the brief or prolonged transition period during which a couple resumes sexual activity after a birth. It can reveal the negotiation that goes on between the partners during this period, which in turn is very informative about the nature of the conjugal relationship. The interview may also provide information about the way in which the woman respondent combines her life as a mother with her life as a spouse, and what place her sexuality will take in her life now that she is a mother.

3.3. Dating the Weaning of a Breastfed Infant

47Strictly speaking, weaning means gradually ceasing to breastfeed an infant and putting it on a diet of more solid foods. In common parlance it can also mean gradually switching from breast milk to manufactured baby milk. Either way, weaning is a process. For the purposes of a survey, it can be defined as the infant’s transition from the state of a breastfeeding baby to that of a child who is no longer breastfed at all. This process over time has a start and an end. If one wants to know the date when the child was weaned, the problem of definition criteria arises: is it weaned as soon as the process begins, or only once it is completed? One can distinguish three moments in the process: when weaning is begun, when it is under way, and when it is completed. To speak of the moment of weaning with no further explanation leaves a fuzzy reality that might correspond to any of these three moments.

48This fuzzy situation creates three problems:

• the first problem is to know what the respondent means by “weaning”. One woman may consider the infant weaned as soon as the process has started, another only when it is completed. The way in which the question is formulated is important. To reduce the fuzziness, one needs to know how the infant has been fed from the time of its birth to the time of the survey, asking the mother to explain each of the changes she mentions;
• the second concerns the researcher’s choice of the moment by which to define “weaning”: the start, the period when the process is under way, or the end? Taking only one moment into account in the survey is less useful for studying weaning itself, as a process interacting with other transitions in a new mother’s life − her resumption of sexual relations, the return of menstruation, starting work again, the possibility of leaving the child in someone else’s care. Care must be taken to identify each of these moments, and the appropriate indicators for each one. The subsequent decision to define one moment as the moment of weaning will depend on the purpose of the research. For example, for a study of contraceptive practices in Africa, one would choose the start of the process, as the frequency of breastfeeding conditions the length of the postpartum infertile period. But if one is studying the determinants of the resumption of sexual relations after a birth, the end of the process seems more relevant because this is when the “lactation taboo” comes to an end (this is the traditional belief that a lactating woman should not have sexual intercourse because sperm and milk will both be in her body and the sperm may “spoil” the milk). (Van de Walle, 1988). In Northern countries, there does not seem to be such a correlation between lactation and the resumption of sexual relations, especially with contraceptive pills that are compatible with breastfeeding. Total or partial weaning have more to do with the woman’s own desire to devote herself to other activities, including sexual activity.

49In the particular case of HIV-positive women who are breastfeeding, the type of weaning has major consequences as regards the risk of HIV transmission. The WHO recommendations are as follows (WHO, 2004): “when replacement feeding is acceptable, feasible, affordable, sustainable and safe, avoidance of all breastfeeding by HIV infected mothers is recommended. Otherwise, exclusive breastfeeding is recommended during the first months of life” and should then be discontinued as soon as feasible. A study in South Africa has shown that children who are exclusively breastfed are less often infected with HIV than those on a combination of breast milk and commercial baby milk (Coutsoudis et al., 1999). So it is recommended that HIV-positive women who breastfeed wean their babies rapidly, to reduce the risk of HIV transmission. In this case it is important to know how long the process took, so it is not useful to take the date of completion of weaning as the sole indicator. The entire process must be taken into account.

• the third problem is linked to the difficulty of defining certain stages in the weaning process. It is easy to pinpoint the end of weaning when the child is no longer being breastfed at all; the start of the process is much harder to identify. The frequency and duration of breastfeeds may gradually diminish even though the mother is not intentionally weaning the child. It depends on the mother’s breast milk supply, the age of the infant and the socio--economic environment of the mother or parents. The start of a more varied diet for the child may reduce the mother’s breast milk supply, and so trigger the start of weaning independently of the mother’s wishes. In this case she may not perceive this as the start of weaning, though she may later take the decision to wean. The easiest indicator to identify is thus the end of weaning.

3.4. Taking Anticipation of a Process Into Account

50For the processes taken as examples above − pregnancy, postpartum resumption of sexual intercourse, weaning − the timing, conditions and duration of the process depend on how far the woman, or the couple, have anticipated it. This anticipation may involve a prior stage undertaken with a view to the final process. In Africa, where some still comply with the “lactation taboo”, a woman may wean her child earlier than initially intended in order to resume sexual relations, sometimes under pressure from her partner.

51To take another example, a woman who has learned from a screening test that she is HIV positive but dare not tell her husband, may soon tell him because she wants to wean the child early so as to reduce the postnatal transmission risk. Other women may tell their husbands once it is time to resume sexual relations, in order to justify using condoms as recommended by the medical team.

52These examples show how anticipation of one event may accelerate the occurrence of another (weaning in the first example, informing the partner of the HIV infection in the other two).

53The different processes should therefore not be analysed separately, and some thinking is required as to how best to characterize possible anticipation (Bocquier, 2006).

4. After the Birth: the Postpartum Trajectory

54The postpartum period after the end of a pregnancy is a time of many transitions. After a birth, a woman moves gradually from the postpartum state (amenorrhoea, no sexual relations, possibly breastfeeding her child, on maternity leave from work) to a state where sexual relations have been resumed, the baby is completely weaned, her periods have returned showing that she is fertile again, and she has gone back to work. The passage from one state to the other involves four transitions, though they may all happen simultaneously: resumption of sexual relations, weaning of the baby, return of menstruation, return to work. All this is a complex process and must therefore be analysed as a trajectory.

55The various processes occurring during the postpartum trajectory interact with each other, for biological reasons (the timing of weaning affects the duration of amenorrhoea) or cultural reasons (the resumption of sexual relations may depend on the return of menstruation or the completion of weaning; in some cultural environments it is difficult to continue breastfeeding once one has started work again). So to understand the trajectory it is important to find suitable methodological tools for analysing the sequence involved, the determinants of the process and the priority given to one determinant or another.

56As in the example of resuming sexual relations in Côte d’Ivoire, in such cases of multiple transitions people may justify their trajectory in terms of a number of different social norms that sometimes contradict each other, rather than constructing their trajectory according to one dominant behavioural pattern. In addition to this basic complexity there may be other transitions and a higher level of complexity. This is the case where the pregnancy has led to the discovery of an infection or illness. Many women in developing countries discover they are HIV positive through large-scale HIV screening of pregnant women under programmes designed to prevent mother-to-infant HIV transmission. For these women, the postpartum period is a time for integrating into their lives the new fact of their HIV infection. This, and the woman’s decision on whether or not to tell her partner, will have a major influence on the weaning of the child and the resumption of sexual relations, if indeed they have been interrupted during or after the pregnancy. In Abidjan, HIV infection proved to be the main factor for timing the resumption of sexual relations after a birth. HIV positive women resume sexual activity much later than others. They also breastfeed for a shorter period (in line with the advice they are given to help reduce the risk of mother-to-infant transmission of the virus via breast milk); as a result, their periods return sooner. The main element here is the problems they have in managing their sexuality in the light of their HIV status. Because they find it difficult to tell their husbands and to suggest he use condoms or persuade him to do so, one strategy they adopt is to postpone the condom problem by delaying the resumption of sexual activity (Desgrées du Loû et al., 2002). The complementarity between the qualitative and quantitative approaches is obvious here. Quantitative methods can be used to compare timings and detect changes in schedules, but it is through interviews that the information can be gathered for a detailed understanding of these changes of trajectory.

57In Northern societies the postpartum period raises different issues. The key factor in women’s timing decisions after a birth is not the resumption of sexual relations or weaning but the return to the labour market (although breastfeeding has made a major come-back and many women now want, or are determined, to continue breastfeeding for longer than the three-month legal maternity leave). With regard to the return to work, quantitative surveys are better for measuring the difference an additional child has on a woman’s economic activity. Numerous studies of this question (Barrère-Maurisson, 2003; Commaille, 1993; Fagnani, 2000; Ferrand, 2004; Garner et al., 2004; Singly (de), 2003; Testenoire, 2001) have shown that several factors must be recorded to assess the impact of motherhood on women’s life courses. The date they return to work is a prime marker, but is not sufficient. It is also necessary to grasp their new “working mother” state and the tensions created by the additional workload a baby entails. Has the mother gone back to the same job? Are her working conditions the same as before her maternity leave? Or has she changed her working hours, her post or her type of work (less responsibility, less mobility required, etc.)? Has she changed her workplace to be nearer home or to find a firm that gives access to a crèche, etc.? Data on any changes in the father’s working conditions should also be gathered, in a simple form: have they changed? In what respect? It is also necessary to record the type of childcare that fully closes the postpartum period, and identify the date this type of childcare began and its characteristics (full-time, part-time, flexible).

58But these data, even if very precise, do not in themselves explain the modalities of the change from being a “working woman” to being a “working mother”. Only the interview will provide an understanding of how the woman balances successful motherhood with her hopes for professional success, the choices she has had to make in view of the constraints on her time, and any negotiating she has had to do with her spouse. The interviews from the GINE research show that usually, the maternal identity takes precedence over the professional identity except in very particular cases (Bajos, Ferrand, 2006). The example of facilities provided by French family policy highlights the impact of childcare systems on defining the mother’s identity. It can be said that for women who choose the state allowance to pay for help at home, their identity is built on the “professional woman’s” passage to motherhood, while for women who decide to take a break from their work with the state allowance for parents of small children staying at home, their identity is built on the shift to being a mother rather than on the status of working woman. Recent research (Méda et al., 2003) seems to show that this latter choice has serious consequences for a woman’s professional future. They may think they are simply taking a temporary break from the labour market but they have no guarantee of finding a job later (especially in times of high unemployment). A movement regarded as temporary may result in a woman being continuously absent from the labour market and take her over the threshold into unemployability (Daune-Richard, 2001).

Conclusion

59This chapter has sought to reveal how the start and end of a pregnancy are complex transition phases that are worthy of study for several reasons. They do not only involve the woman; very often they involve a partner or an entire family around the couple. In this regard, the context in which the pregnancy occurs must be clearly identified. The patterns of communication and agreement between the man and the woman on the subject of the pregnancy, and around it, must be especially taken into account. Hence the importance of combining quantitative and qualitative approaches in this field of research. It may be worth stressing that it would be very relevant to consider men’s viewpoints on pregnancy as well as women’s; hitherto, most surveys about pregnancy have taken very little account of the men’s views.

60The process of the end of pregnancy (in this chapter, we have only discussed unproblematic pregnancies resulting in live births at full term) also involves several other processes: postpartum amenorrhoea, breastfeeding and then weaning, a temporary break from sexual relations and then resumption, and the matter of the mother’s economic activity. These aspects interact with each other more or less closely; no one of them can be studied without analysing the others. So it is appropriate to analyse the whole set of transitions around pregnancy by reconstituting a multidimensional trajectory rather than analysing each process on its own. This means gathering health and sociodemographic data from which one can, among other things, date the events concerned, though the dating may itself be a complex matter. Alongside this, the qualitative approach and the gathering of socio--anthropological data are essential for understanding the conjugal, family, occupational, social, affective and psychological settings that explain the sequence of processes that unfolds in the complex trajectory set in motion by pregnancy and its outcomes.