THE CONJUGAL FAMILY 'OPEN' OR 'CLOSED': A PRELIMINARY ANALYSIS OF AN ASPECT OF CHANGES IN THE PRESCRIBED NORMS FOR CONJUGAL RELATIONSHIPS OF A SAMPLE OF GHANAIAN UNIVERSITY STUDENTS* by Christine Oppong** Introduction This paper is concerned with some of the effects of one or more generations of schooling upon educated young men's ideas, as to what should be the domestic rights and duties of husbands, wives and kin. It has been observed in Ghana, and frequently stated that schooling modifies relationships in the conjugal family (e.g. Caldwell, 1968: 107, 180). An instance of these effects is the significant difference in attitudes to family size found amongst educated, urban dwellers (ibid., p.38). Lloyd (1966: 30) has recently suggested too, that in Africa in general, levels of education attained in the ascendant generation will have some effect on the pattern of nuclear family rela- tionship observed among the educated. The aim here is to show what some of these effects are upon conjugal family norms, in particular to see whether there appears to be a shift towards norms associated with the functionally individuated conjugal family, with increasing generations of education, in a section of the educated, Ghanaian population. The Classification of Data Much recent writing on the contemporary conjugal family, especially in societies undergoing rapid social change, has stressed its increasing functional individuation and isolation (e.g. Goode, 1963). The documentation and comparison of conjugal family types observed in such social settings has however been considerably hampered hitherto by the lack of adequate categories for the classification of data. The use of such terms as 'the extended family' and 'the isolated nuclear family' may gloss over facts rather than providing illumination (e.g. Aldous, 1962). The result may be a fruitless discussion as to whether or not "the extended kin family' persists in a parti- cular type of urban society, or whether the 'nuclear family' does or does not exist in a particular community. What is crucial in classifying and comparing systems of family relationships is the discovery of which kin-based units are held to be responsible for the performance of particular domestic functions, such as rearing and educating children, financial • The data under discussion were collected during a survey which was one aspect of the field-work carried out by the author in 1967-8, and formed part of a wider study of the conjugal norms and behaviour patterns, both reported and observed, amongst a sector of the urban, educated population of Ghana. The study was jointly sponsored by the Institute of African Studies, Legon, and the Department of Anthropology, Univer sity of Edinburgh. The field-work and subsequent data-processing were partly financed by a Social Science Research Council grant, administered by Professor K. Little of Edinburgh University, to whom grateful thanks are due. •• A paper ready by Christine Oppong, Research Fellow of the Institute of African Studies, Legon, at the Ghana Sociologies^ Association Conference in April, 1970. 18 support of dependents and so on. It is, I suggest, vital to discover who are the peo- ple who may, and habitually do, act as role substitutes for each other. (This I have discussed in more detail elsewhere: see Oppong, 1971). This kind of analysis has already been carried out in rural kinship systems in Ghana and elsewhere. In studies such as those of the Tallensi and the Ashanti we are provided with a detailed analysis of matters such as who may act as a parent sur- rogate, co-wife, financial provider and joint property owner. The kinship groups, within which domestic rights and duties are shared and kinship roles to some extent held in common, are clearly defined. In these two particular societies unilineal descent groups, at various levels of segmentation, form closed kinship groups, within which certain domestic rights and duties are shared. The nuclear family in these and other similar societies is not a functionally individuated unit for most domestic purposes, in the sense that conjugal, parental and filial roles are habitually played across its boundaries to kin in other nuclear families. The difficulties inherent in comparable analyses of urban kinship systems have scarcely begun to be met. Unfortunately a number of people, who have carried out surveys of large samples of urban dwellers, with a view to obtaining data on the conjugal family and the kinds of relationships maintained between kinsfolk in the town have concentrated upon such features as social visiting patterns and leisure activities, using these as criteria for judging the extent to which the conjugal family is an independent, solidary or isolated unit. The vital areas of family functioning, which are of universal significance have been to some extent neglected. What we should examine are such activities as child- rearing and education, financial-provision and management, procreation and the division of domestic labour and decision-making, in order to see to what extent the conjugal family is a functionally individuated unit, rather than such items as visiting in-laws, attending the cinema or taking the children out.1 Closed and Open Kinship Units Two suggestions then are made as regards a method for looking at conjugal or nuclear family individuation. We need first to choose one or more universally relevant areas of domestic functioning and then to see in which kin-based groups or networks of kin the associated rights and duties are held. What we need to do is to look at the conjugal family to see whether it is a closed or open kinship unit with respect to a parti- cular domestic function.2 Here I quote Farber who has clearly stated what is meant by this classification, (1966: 79-80). "When (conjugal) family relationships are relatively open, the spouse and parent roles extend beyond the nuclear family to other kinship groups in the community. (The spouse and parent roles refer to rights and obligations about authority, family division of labour, child-care, and financial support). In societies characterised by open conjugal fami- lies the roles of uncle, aunt, grandparent, brother, sister, and cousin 1. For instance Caldwell (1968:63) writes "a more exacting test of family change is whether the parents also take children on outings" (eg. cinemas, parties and the beach). t. For earlier usage of the termi 'open* and 'dosed' in this way see Weber (1947:139-143) and Redfield (1947: 293-308.) 19 embody rights and obligations for enacting some spouse-paient roles in families of procreation other than their: own. The borderline between the individual's family of procreation is not well defined* ...i Societies differ in the extent that they prescribe closed relationships for the nuclear family." As we have already pointed out from two rural Ghanaian examples, there are many African cases documented in which the nuclear family is for many purposes an open unit and the lineage is in many ways the closed kinship unit. In other cases documented elsewhere the nuclear family is a relatively closed unit. Prescribed Norms Thus having defined what we mean here by open and closed conjugal families and having suggested some of the crucial areas of family functioning, which should be used for comparison, we can now go on to the next problem, that of defining what we mean when we talk about the 'prescribed norms' of a particular population, since the aim is to test the hypothesis, that with increased generations of educated forebears, people's prescribed norms for conjugal family relationships become increasingly closed. Not only what do we mean by prescribed norms in this context, but how can they be quantified and compared so that the hypothesis can be tested ? First of all by prescribed norms is meant the ideas in the minds of a group of people, which can be put in the form of statements specifying what people should do, ought to do, are expected to do, under given circumstances (Homans, 1950: 122); in other words, the formal rights and obligations felt to belong to a particular social position, such as that of spouse. Thus a questionnaire was designed which might elicit some of these prescribed norms by presenting statements for agreement or dis- agreement, statements which said what the rights and duties of husbands, wives and kin ought to be vis a vis each other. Ideally by gaining information on a whole battery of such prescribed norms or behavioural traits one may arrive at detailed role descrip- tions.3 Since evaluation of such a behavioural trait can vary from strong approval to strong disapproval through some middle point of indifference, an evaluative dimen- sion was included in the scheme for describing norms, by means of coding four types of responses from Absolutely Agree, through Partly Agree, to Absolutely Disagree.4 Method Armed with these stated aims and mode of approach and technique, a question- naire was administered to a random sample of one in four male undergraduates at Legon and Cape Coast Universities in 1968. Some of the results are presented below to see whether they appear to prove the hypothesis about generations of education. Ten different statements were used to obtain the students' attitudes as to whether a particular family function should only be performed by the conjugal pair or whether other people could act as role substitutes. Five activity areas were chosen, including 3. Bates and Cloyd (1956) have described the kind of method I attempted to use. 4. Jackson (1966) has discussed the dimension of evaluation implicit in almost every conception of the norm. "A norm that exists in a group involves shared tendencies to approve or disapprove a particular dimension of behaviour". 20 child-rearing and education, sex and procreation, residence, inheritance of property, financial responsibility and family decision-making. First we shall examine each area to see whether each generation of education appears to affect views on closure and then describe how an attempt was made to combine all the ten, coded responses to produce an overall assessment of approval or disapproval of the closed nuclear family. Results Child Rearing and Education First of all with regard to child-rearing and education, there is ample evidence that the rearing and acceptance of educational responsibility for children of kin is a relatively widespread phenomenon in many different rural Ghanaian communities, both in the north and south (Goody, 1966, 1969; Nukunya, 1969; Oppong, 1974). We are well aware from our own experience, as well as from the documentary evidence available, that it is widely practised in the towns and among the educated.5 Statement:—-"A person should only be expected to educate his or her own children.' TABLE 1 Generation Educated %Absolutely Partly Agree 1 2 3 59 51 64 TABLE 2 Total Responses N °/ (215)100 (212) 100 (164) 100 Statement:—-"It is often beneficial for a child of needy parents to be reared by other relatives as they may provide a better home environment for the child." Generation Educated % Absolutely Partly Agree 1 2 3 72 75 73 Total Responses N % (216)100 (210)100 (166)100 S. For instance in a turvey of Senior Civil Servants I carried oat in Accra in 1968 I collected evidence of the frequency with which men in that occupation are refponsible for the education of children of kin (Oppong, 1969). 21 Tables 1 and 2 above show that two different statements had slightly different responses. While a considerable majority, almost three quarters in each category, thought it beneficial for a child to be reared by relatives, if its own parents were poor, (Table 2) over half thought that in general people should only have to educate their own children (Table 1). Among the third generation educated there was a slight increase in people who thought one should only have to educate ones own children. Parental role substitution obviously still continues to be a widely accepted norm among this sector of the educated, but more appear prepared to say that openness of the nuclear family in this respect may be a good thing, when poverty is a constraining factor, than to admit that others have a right to expect one to educate their children. Polygyny and Procreation When examining the issue of polygyny we know that in Ghana as a whole the majority of marital unions are monogamous, the proportion monogamous increasing throughout the population according to educational level. The vast majority of marriages are however both legally, and in fact, potentially polygynous and therefore the norm of openness for the conjugal family with respect to sex and procreation is institutionalized. TABLE 3 Statement:—'A wife cannot expect her husband to be always faithful to her. She should not think of divorce just because he begets an outside child." Generation Educated % Absolutely Partly Agree 1 2 3 54 58 51 Total Responses N % (214) 100 (211) 100 (165) 100 Table 3 shows the extent to which this norm appears current among potential graduates by their generation of education. As in the matter of parental role substitution there is no great cumulative change from one category to the next. The responses are fairly similar, over half in each group thinking that wives should accept the fact that their husbands may beget outside children without disrupting the marriage. Similarly about one in three think a wife of an ordinance marriage should accept the fact that her husband has one or more customary wives as well. The norm of openness with regard to sex and procreation in conjugal relationships is still widely accepted, though approval of legally sanctioned polygyny is less widely current than approval of extra marital liaisons. As has been suggested elsewhere it would appear that the decrease in polygyny, noticeable among the educated, is likely to be the result of changing economic pres- sures rather than changes in the normative system. 22 Inheritance The question as to whether the conjugal family should be an open or a closed group, with respect to inheritance of personally acquired property has been a vexed one for some time in Ghana, as witnessed by several recent abortive attempts to change the existing laws (Ollenu, 1966). One of the crucial factors is this, that for two in five or more people in Ghana, that is the Akan, the customary law, which may affect the disposal of their personally acquired property on death is that of matrilineal succession. According to its provisions not only is the conjugal family regarded as an open group, but the wife and children do not belong at all to the group of people recognised as potential heirs and in other systems the rights of the spouse and children may only be relatively limited. That people want change is continually evidenced by outcries in the press and the circula- tion of gossip about the plight of disinherited widows and orphans. Here we have quantitative evidence that ideas on this question are changing radically with increasing generations of education. A look at Table 4 shows the kind of change in attitudes which is taking place. With succeeding generations of education, fewer people think that a man's conjugal family should not be the unit for inheritince of his property. In other words the tendency is towards closure of conjugal family norms with respect to inheritance. TABLE 4 Statement:'—"It is not right for a man to leave the bulk of his property to his wife and children alone." Generation Total % Absolutely Partly Agree 1 2 3 65 58 43 Total Responses N % (214) 100 (213)100 (168) 100 Financial Responsibility and Decisions Now with regard to financial responsibility for dependents. We have been told elsewhere, again from a student survey, that a majority of educated young men (84 per cent) expect to spend between 10 and 30 per cent of their income on relatives outside their conjugal families (Caldwell, 1965: 191). And there is substantial evidence from a number of sources to demonstrate the frequency with which urban dwellers send financial help to kin in rural areas (e.g. Caldwell, 1969: 133). 23 TABLE 5 Statement:—"A man's sole continuing financial responsibility is to his wife and child- ren, his relatives should not expect to get anything more than occasional gifts and help from him." Generation Educated % Absolutely Partly Agree 1 2 3 72 69 73 Total Responses N % (215) 100 (212) 100 (169) 100 The majority in this sample do however feel that a man's main financial respon- sibility lies within his own conjugal family and this feeling does not appear to vary significantly according to generation educated (see Table 5 above). At the same time there appears to be widespread agreement that guidance should be accepted by young couples from their senior relatives (see Table 6). The traditional norm concerning the decision-making power of senior family members continues to be prescribed. Whether this persists in actual practice lacks adequate documentation. TABLE 6 Statement:'—"A young couple should be guided by their elderly relatives when making important decisions." Generation Educated % Absolutely Partly Agree 1 2 3 Residence 69 63 64 Total Responses N % (212) 100 (209) 100 (166) 100 The co-residence of non-nuclear kin continues to be a widespread phenomenon in all sections of the Ghanaian population including, the urban educated (e.g. Caldwell, 1968: 61). The tendency among the latter is for such kin to be junior rather than senior relatives of the parental generation. This is an area of domestic organization in which the shift towards norms associated with closure of the conjugal family is most marked, for the sons of educated men are much more likely to think that the conjugal family should live alone, without other kin in the household, than are the sons of illiterates. 24 TABLE 7 Statement:—"After marriage a couple should live on their own with their children, without the continual presence of other relatives in the house." Generation Educated % Absolutely Partly Agree 1 2 3 59 72 74 Total Responses N % (214) 100 (212) 100 (168) 100 Variations in Norms If the responses of the third-generation educated are now compared, for the various areas of family behaviour, we find that closure is most marked for the norms of residence, financial responsibility, monogamy and inheritance and that openness still characterises to a greater extent the norms for rearing children, procreation and decision- making (all under 50 per cent approval of closure). The areas of most obvious change with succeeding generations of education are residence and inheritance. (The simple mode of tabulating data used here has however obscured the detailed shifts in degrees of approval and disapproval). An Ordinal Scale for Ranking Closed Open Norms It was assumed at the outset that a dimension of attitudes regarding the prescrip- tion of closed or open norms for conjugal family roles existed and that the position of each respondent and the population and segments of it could be placed at related and contrasting points on a continuum. Thus an ordinal scale was constructed for measur- ing the degree of closure approved of by scoring 0, 1, 2, 3 and totalling the responses to ten statements, some of which we have briefly examined here.6 The total scores were classified into three groups, according to degree of approval of conjugal family role substitution they indicated.7 6. The scale is called 'ordinal' since it, "ranks individuals according to the characteristic being scaled but carries no implications of distance between scale positions" (Moser, 1950:237). 7. A score of zero showed approval of the closed nuclear family and a score of 30 approval of the open nuclear family, similarly the range of scores in between showed gradations of approval. To facilitate comparison of segments of the population the scores were divided into three categories:— 0-8 9-15 denoting mixed or moderate views 16-30 denoting approval of open conjugal family noraos. denoting approval of the closed conjugal family 25 Prescribed Norms for the Conjugal family: Open or Closed TABLE 8 Generation Educated % Closed % Medium 1 2 3 14 20 27 54 51 47 32 29 26 Total Responses N % (216) 100 (213) 100 (169) 100 The results of this classification of responses is observed in Table 8, where we see that a marked difference is apparent in the views of the first, second and third generation educated people. An increase in approval of closed nuclear family norms is correlated with increasing generations of education. The association between having educated parents and approving of a more closed conjugal family system is significant at the five per cent level (X2= 10.17, df=4) Among the first generation educated, more than twice as many approve of open nuclear family norms than of closed nuclear family norms. closure and openness are similar. In the third generation educated category the proportions approving of The Hypothesis We have thus given quantitative evidence which supports the hypothesis that with succeeding generations of education there is a noticeable shift in the prescribed norms for some areas of conjugal family functioning from an open to a more closed system in the chosen population. By examining individual behaviour areas of universal domestic importance we have shown that this shift is not uniform. It is more marked in some areas than others in particular in the areas of property inheritance and co- residence. In other areas however such as degree of openness thought proper with regard to the decision-making power of senior relatives or polygyny, there has obviously already been a marked shift in opinion away from traditional norms by the first gene- ration educated respondants, since norms institutionalised in the regions from which they come now appear to be less widely current. Conclusion In re-examining the material however, it needs to be remembered that what have been analysed here are some prescribed norms of a sample of educated young men, with respect to a number of aspects of conjugal family relationships. They are not reports of actual behaviour patterns, modified by a multitude of conflicting pressures from wives, kin, colleagues and so on and restricted by the lack of means. It was accepted as axiomatic by the author at the beginning of the study, from which these data items are extracted, that in analysing aspects of a situation characterized by wide spread individual mobility, both social and spatial and by marked change in domestic institutions, it was not only admirable but imperative, to collect data on both norms and behaviour. Thus an important aim of this paper as well as the testing of a parti- 26 cular hypothesis has been the demonstration of the present need for more detailed empirical studies of the changing norms, as well as behaviour patterns in the contem- porary family in Ghana. A further intention has been methodological, that is to indicate the utility of the classification of conjugal family relationships as open or closed in the way defined above, and to begin to evolve an appropriate technique for the quantification and comparison of degrees of openness and closure. It was with this aim in view that an ordinal scale was constructed, for ranking degrees of approval of conjugal family closure in a number of domestic activity areas. Adams, B. N. 1967 Aldous J. 1962 Bates, A. P. & Cloyd 1956 Caldwell, J. C. 1965 Caldwell, J. C. 1968 Fortes, M. 1949 Fortes, M. 1950 Goode, W. J. 1963 Goody, E. N. 1966 Goody, E. N. 1969 References Kinship in an Urban Setting Markham Publishing Co. Urbanization, the Extended Family and Kinship Ties in West Africa. Social Forces Vol. 41, 1, 6-12. Towards the Development of Operations for Defin- ing Group Norms and Member Roles. Sociometry Vol. XIX No.l, p. 26-39. a Extended Family Obligations and Education: study of an aspect of demographic transition amongst Ghanaian University Students. Population Studies Vol. XIX, No. 2. 183-199. Population Growth and Family Change in Africa. The New Urban Elite in Ghana. Australian National University Press, Canberra. The Web of Kinship Among the Tallensi London. Oxford University Press (for I.A.I) Kinship and Marriage Among the Ashanti, in African Systems of Kinship and Marriage, (eds.) Radcliffe Brown, A. R. & Forde, D. London. Oxford University Press, pp. 252-284. World Revolution and Family Patterns, Free Press. New York. Collier-Macmillan Ltd., London. Fostering in Ghana: A Preliminary Survey. Ghana Journal of Sociology 2. Kinship Fostering in Gonja, in Socialization. The Approach from Social Anthropology. (A.S.A.8) (ed.) Mayer, P. Tavistock publications. 27 Homans, G. C. 1950 Jackson, J. M. 1966 Fortes, M. 1970 : The Human Group. New York. Harcourt Brace. : Structural Characteristics of Norms. Chapter 9 in, Role Theory: Concepts and Research (eds.) Biddle, B. J. & Thomas, E. J. J. Wiley & Sons, New York. : Kinship and the Social Order. Routledge and Kegan Paul. Lloyd, P. C. 1966 (ed.) : The New Elites of Tropical Africa. Oxford Univer- sity Press (for I.A.I.) Moser, C. A. 1958 Nukunya, G. 1969 : Survey Methods in Social Investigation. Heineman London. : Kinship and Marriage Among the Anlo Ewe. The Athlone Press. Ollennu, N. A. 1966 : The Law of Testate and Intestate Succession in Ghana. Sweet & Maxwell. Redfield, R. 1947 : The Folk Society. Am. J. Soc. L II Weber, M. 1947 Oppong, C. 1969 Oppong, C. 1971 Oppong, C. 1974 : The Theory of Social and Economic Organization. Oxford University Press. : Education of Relatives' Children by Senior Civil Servants in Accra. Ghana Journal of Child Develop- ment Vol. 2 No. 2 December. : "Joint" Conjugal Roles and "Extended" Families. Journal of Comparative Family Studies Vol. II No. 2. : Growing up in Dagbon. Ghana Publishing Corpo- ration. 28