There is a continuing debate in anthropology about the analytical boundaries between sex, sexuality and gender, and the extent to which gender is biologically or culturally ‘constructed’ (Moore Reference Moore and Moore1999: 151–71). Without entering those debates, this article approaches gender relations from the very particular perspective of research on rural political economy. Most generally, I use ‘political economy’ to mean the total relations of production, exchange and power in which a people is embedded, leaving open (as did Marx and Engels Reference Marx and Engels1974 [1846]: 58) the reciprocal influence between political economy and other social spheres.Footnote 1 This article focuses on the ways in which changing rural political economy and changing gender relations affect each other.
In 1976 and from 1977 to 1979, and again for two weeks each in 1985, 1996 and 1998, I did fieldwork on rural political economy in the Hausa hamlet of Marmara. It is situated in Malumfashi District, in the economic region of southern Katsina (a part of Katsina Emirate). Marmara, and southern Katsina more generally, are part of that enormous area known popularly as ‘northern Nigeria’, where the jihad of 1804–8 established the Sokoto Caliphate – severely altered, but not abolished, by British and post-independence governments.
Society in Marmara is polygynous. Divorce and re-marriage are easy and common. Conversion from the ancient indigenous Maguzawa religion to Islam occurred from the mid-1950s to the early 1970s, while the few remaining Maguzawa families converted by the late 1980s. More recently, a purist form of Islam led by reformers known throughout northern Nigeria as Yan Izala has replaced an earlier eclecticism over the relative merits of different brotherhoods, the Tijaniyya and Qadiriyya.
Gender relations are profoundly influenced by the organization of social economy. At the same time, changes in religious belief have affected the social economy and thereby influenced ideas of gender. This mutual causation leads to a balance of continuity and change in notions of ‘manhood’ and ‘womanhood’. Rural people articulate the moral realm in a particular way. The obligation to marry, to have children, and, when income allows, for men to increase the number of their wives, is absolutely imperative. The crucial position of women in social reproduction – in the maintenance over time of production relations based on the family – has qualified the impact of Islamic conversion on local gender relations.
How fieldwork both limited and generated my questions on gender relations
My fieldwork was conducted mainly among men. I was primarily concerned with economic relations between household heads of varying levels of income; and all households were headed by men. Moreover, since the conversion of households to Islam, adult married women had to a large extent withdrawn inside the house, whence they engaged in the variety of economic activities described in this article. But, for a male researcher like myself, close conversation with or intimate observation of women was limited to two households which had been my entrée to the hamlet; and to one further household with whose head I had a close friendship. It is also worth mentioning a personal circumstance of research: I did not live in a household, which would have given me day by day so much direct experience of gender relations. Rather, I lived in the compound where the main grain trader stored his supplies and negotiated many of his transactions. My location eventually led many men to my room, where they could discuss their economic affairs with, perhaps, more privacy than would have been possible had I been living with a particular family. More deeply, the chosen focus of my research – processes of accumulation and the measurement of economic inequalities – committed me to interaction with men, who organize most farming and trading in the Hausa society of rural southern Katsina. This is not to deny the economic centrality of women that is outlined below. But it is to say that most research time was necessarily spent with men. What follows is a perspective on gender relations heavily influenced by the public predominance of men.
From 1976 to 1998, the domestic seclusion of women in the hamlets of southern Katsina was partial in the sense that they did move from house to house and from hamlet to hamlet; I could watch and listen to them chatting with men when they did so. While, after conversion to Islam, women gradually withdrew from most farming operations, I was able to chat with them during planting and also during the harvesting of cotton and maize, in which their role remained central. Nevertheless, I developed my experience of the Hausa language mainly by listening to men talk to each other or to me. Although there were many occasions when I listened to women, they were usually responding to men or taking part in a marriage ceremony. In summary, the nature of my fieldwork led me away from the intimacy of domestic space. At the same time, this very limitation generated the question that guides this essay: how are gender relations among Muslim rural Hausa to be understood from the standpoint of political economy? Concisely, this article explores the politico-economic reasons why men were preoccupied with women.
CHANGING POLITICAL ECONOMY
My fieldwork focused on the nature of economic change in rural Hausaland. Marmara is a roadside hamlet eight miles from the town of Malumfashi. It is part of the economic region of southern Katsina, long known for its marketed surpluses of grains, cotton and groundnuts. Agriculture is organized around the work unit of fathers and their married sons (gandu) – even when, as in the case of many rich village farmers, large numbers of casual or seasonal labourers are employed. Most farmers are innovative, commercial, entrepreneurial. Through universal surveys of land distribution and labour practices, and case studies of the range of credit relations, I explored how the commodification, first of crops, then of land and finally of labour, had changed the relations of production and exchange in Marmara. I was equally interested in the ways in which long-distance traders in Marmara linked up southern Katsina with other economic regions of Nigeria. These various elements of rural political economy provide an important context through which to understand changing gender relations.
My conclusions concerning economic change in Marmara are relevant to understanding the position of women not only in economic organization, but also in the more general thinking of villagers about gender (Clough Reference Clough1995, Reference Clough and Harriss-White1999, Reference Clough, Clough and Mitchell2003). The development of a flourishing market economy has not led to the development of capitalist relations of production and exchange. Rather, a complex, historically formed and indigenous ‘trajectory of accumulation’ is at work. It structures three specific forms of accumulation:
household accumulation, through the polygynous marriages of villagers and their children, and procreation, thus increasing the family work force;
cliental accumulation, whereby farmers who combine agriculture with trade expand their circles of clients and trading friends, with the concomitant financial burdens;
and capital accumulation, through which farmer-traders purchase labour power and means of production, appropriate surplus product, and profit from the appropriated product through commodity exchange.
Capital accumulation both serves, and is constrained by, household and cliental accumulation. The complex interconnections between these three forms of accumulation reinforce social ties of kinship and friendship between richer and poorer villagers. They limit the economic power of the rich.
The local practice of Islam has acted as a barrier to capitalist accumulation, and, more generally, to the division of rural society into two distinct classes – capitalist farmers and landless or semi-landless proletarians. Among its elements, local Islam confirms and encourages the ancient practice of polygyny which pre-dates Islam. Men expend a significant share of increases in their material assets on increasing the number of their wives and, therefore, their children. In Marmara, among the lowest tenth of land users, the great majority had one wife, and mean household size was approximately three; among the top decile of land users, the great majority had at least two wives, and mean household size was almost fifteen (Table 1). In the effort to develop polygynous households, women – as wives, potential wives and mothers – remained central to the ways in which Hausa men conceptualize the goals of economic life.
CHANGING RELIGIOUS BELIEFS AND THE POLITICAL PROCESS
Local Islam and political conflict: 1976–9
When I began fieldwork in 1976, Malumfashi District still had one of the largest remaining concentrations of Maguzawa, or non-Muslim Hausa, in northern Nigeria. However, their numbers had been declining and continued to do so. Formerly, they had flourished because they had occupied a rural hinterland, on the marches between the Emirates of Katsina and Kano. There they had farmed large surpluses of cotton and of grain, partly for sale, partly for use in brewing guinea-corn beer – a major focus of Maguzawa sociality. The development of roads and transport, through the colonial and post-colonial periods up to the petroleum-induced prosperity of the 1970s, had reduced their relative isolation. It had exposed them to the cosmopolitan world of West African Islam – marked by travelling Muslim teachers (malamai) and students (almajirai), awareness of religious debates in the towns, and the commercial attraction of entry into Muslim chains of trading patronage.
The roadside settlement of Marmara was established in 1960 out of dispersed rural compounds. It consisted of Maguzawa indigenes and of Muslim immigrants flowing in from Kano Emirate.Footnote 2 During the 1960s, most indigenes converted from the Maguzawa religion, with its pantheon of spirits, to Islam. Various causes, mutually reinforcing, led to Muslim conversion in Marmara. First, Islam itself, as the main literate religion in the area, attracted the mental assent of many. Second, an economic factor was at work. For successful older farmers and enterprising young ones, access to trading networks and the commercial loans circulating within them required conversion to Islam (see also Last 1979: 240–2). Third, the co-existence within the new roadside settlement of indigenous Maguzawa and immigrant Muslims set up its own dynamic of religious change. Indigenous households were exposed to the ideas and ritual habits of the immigrant families from Kano, with a much longer tradition of Muslim belief. Life in the concentrated roadside settlement itself speeded up the process of conversion.Footnote 3 For increasingly there were mosques, and mallams teaching children the Qur'an. And there was proximity by road to the nearby district capital of Malumfashi, its Friday mosque attracting thousands to prayer from within the town and the surrounding countryside. These influences increased people's sense of membership of an Islamic society and polity.4
In 1978, only four of the 120 household heads in Marmara continued to drink guinea-corn beer, and their wives to brew it. By 1996, the hamlet had grown to a village of over 200 households, all of whom were Muslim. Nevertheless, Islamic conversion did not mute rural people's awareness of the wider social divisions in southern Katsina, or of their political differences with the local hereditary ruling class. Rather, Islam as taught and practised provided moral legitimation for their sense of grievance.
During 1976–9, the people of Marmara defined themselves as talakawa (commoners) in contrast to the sarakuna (rulers). They also referred to themselves as ‘mu, Habe’ (‘we, the Habe’ – that is, Hausa commoners) in opposition to ‘su, Fulani’ (‘they, the Fulani rulers’). These angry phrases expressed a difference which was political rather than ethnic, for they did not apply them to local Fulani pastoralists. People continually brooded over the former extortionate practices of the hereditary rulers, the Village and District Heads and Emirs, which had only been abandoned under the military regime in power since 1966. They called these practices zalunci (oppression).
In the 1950s and 1960s, local Muslim activists of the Northern Elements Progressive Union (NEPU), the opposition party in northern Nigeria, had been inspired by the old tradition of the jarumi, or brave man, who had the religious knowledge (ilmi) to challenge rulers.Footnote 5 Later, in 1979, I was in Marmara during the national elections, contested locally by the PRP (the populist People's Redemption Party) and the NPN (the elitist National Party of Nigeria). Most people in the hamlet spoke strongly in favour of the PRP. They referred to its national leader, Aminu Kano, by the Islamic term Shehu (leader and scholar). The PRP drew its local leaders from farmer-traders and Qur'anic mallams. The NPN was led locally by the hamlet head, the representative of the rulers. The chagrin of the hamlet head at the election results was the more intense when his wives told their friends that even they had voted for the PRP.
People were keenly conscious of new kinds of both division and integration. As Maguzawa, their economic differences had derived from the number of wives and children that a household head could deploy in farming. As Muslims, they saw themselves as bifurcated – between a small number of rich villagers who combined farming with trade, and the rest, who were dependent on the rich for credit. At the same time, they saw themselves as constituting a new kind of community, much larger and more abstract than before conversion. They were now part of the jama'a – the people of God. The jama'a have repented (tuba) and in consequence, they continually beseech or beg God (rokon Alla).
Islam in Marmara was collective and self-conscious. It gave detailed form to the day, week and year. All men gathered five times a day for prayer in one of the earthen mosques which had been built by villagers, and on Friday many travelled to the central mosque in Malumfashi for the early afternoon prayer (azahar). All boys – but not girls – gathered in front of the houses of mallams for several years of instruction in the Qur'an. Men would frequently gather in small groups to listen to preaching (wa'azi), or the recitation of and commentary on the Qur'an (tafsiri). Older boys from immigrant families often spent several years as begging Qur'anic students (almajirai) in far-away Borno or Hadejia. Then, again, there were ecstatic gatherings, at which both men and women shouted out their sense of the presence of God. All adults fasted during Ramadan (azumi). And when infants were given their Muslim names, most men would gather for the prayers at the ceremony (suna). Men would travel miles to attend the suna of a friend's child. Whereas marriage ceremonies were social festivities, with a core religious ritual, naming celebrated induction into the jama'a. It thus symbolized religious unity in a way that marriage ceremonies did not.
The local practice of Islam laid emphasis on three activities. Sadaka was the giving of alms. Infrequently, this took the impressive form of giving a daughter in marriage without marriage payments (auren sadaka). Frequently, people gave very small gifts of money or food to those less fortunate, or simply to groups of friends. More generally, sadaka referred to an attitude of mind – the duty to be generous. Zakka was the giving of one-tenth of each harvest crop to the mallams and poor households of one's choice, though, in practice, the proportion varied from 5 to 10 per cent. Hajji was the pilgrimage to Mecca. Before 1985, ten successful farmers or farmer-traders had been to Mecca. No women had yet been on the pilgrimage, though it had already become common for Muslim women in Nigeria to do so.
Seven household heads were malamai (singular malam) – religious teachers with the power of intercession through prayer. All were immigrants from Kano Emirate. With landholdings of less than four acres, they depended on harvest contributions and gifts from the rest of the hamlet. People gave more respect to the one mai tafsiri, who was able to translate the Qur'an from Arabic into Hausa and interpret it in the light of other holy books, than they did to the Liman, the official leader of prayer appointed by the Village Head. Most men had a particular mallam whom they favoured for advice and supernatural charms.
Yan Izala Islam and rural populism: 1985 to the present
Since 1985 the ‘purist’ or ‘reformist’ Islamic movement, known in northern Nigeria as Yan Izala – with its emphasis on doctrinal understanding, Arabic literacy, and strict ritual observance without Sufi accretions – has captured local support in Marmara, and become the dominant form of local Islam. Its discourse – the stress of its tafsiri – emphasizes the wider participation of all believers in the political order, and the necessity of an informed ‘public opinion’ (Gumi Reference Gumi1992: 92, 161, 189).
The most learned local mallam in 1979 and 1985, the only mai tafsiri, was still the intellectually dominant force in Marmara when I returned in 1996 and 1998. He was proud of his association with the deceased founder and leader of the Yan Izala, Sheikh Abubakar Gumi. In Ramadan of 1998, he was leading daily tafsiri for two hours in the main village mosque which had been greatly enlarged and rebuilt. It was noticeable that whereas, in 1979, only a small group of villagers would gather to listen to his Ramadan commentary, in 1998, at least fifty villagers would daily devote their afternoons to this. Men of all ages attended. Most of the malamai of the hamlet were seated around the mai tafsiri. Many villagers were carrying the Qur'an – unusual in 1979 – and were trying to follow his reading. Local Muslim practice had become more literate and self-conscious.
The concern with understanding Islamic discourse in the original Arabic does not just reflect the local prestige of ‘literate knowledge’ (Western learning is also prestigious). More importantly, it is seen as a key to power. Here, I have in mind access to supernatural power as much as to social power. The Yan Izala have always emphasized that understanding Muslim law, through the holy books of Islam, provides protection for the common people against the demands of the powerful and the rich. To appreciate this political dimension, I turn to the impact of Izala discourse in the sphere of marriage relations.
Yan Izala and marriage payments
Throughout northern Nigeria, one of the key criticisms levelled by the Izala movement against the ‘traditionalist’ Islam of the religious brotherhoods has been that indigenous custom encourages high, burdensome marriage expenditures, which tell on the resources of the poor (Gumi Reference Gumi1992: 162). While, in some parts of Hausaland, this criticism has been embraced by poorer farmers and by young men anxious to limit socially necessary expenses without contracting debts or selling scarce land (Loimeier Reference Loimeier, Westerlund and Rosander1997: 296), in others it has been condemned by rural people as a sign of the rupturing of kinship obligations (Masquelier Reference Masquelier, Werbner and Ranger1996).
Intriguingly, since the Izala interpretation of Islam became dominant in Marmara, changes in the pattern of marriage expenditure have been more apparent than real. In 1979, sadaki, the money gift required in Muslim law as part of the marriage transaction, was a tiny part (less than 5 per cent) of the total marriage expenditure undertaken by the bridegroom or his father.Footnote 6 Other categories of marriage expenditure had different names, and their money burden was easily enumerated (Clough Reference Clough1995). In 1998, I recorded expenditures on marriage by four smaller farmers – one of whom was in the process of becoming a mallam of the Izala persuasion – and by the richest trader in the village. In 1998, the word sadaki included not only the small cash gift required in Muslim law, but also money given to the father or guardian of the bride to use on her behalf, and given to the bride for buying various personal items. It ranged from 10 to 27 per cent of total expenditure by the groom or his father. Essentially, a ‘renaming’ of the various categories of marriage expense had occurred between 1979 and 1998.
Two continuities were apparent. First, the variety of goods brought into a marriage were still donated by the family of the bridegroom and that of the bride in roughly equal measure. Second, the real burden of expenditure on marriage goods had remained constant. In 1979, a smaller landholder would spend a total in the region of N500 on the marriage of himself or his son to a previously unmarried girl (budurwa) – the money equivalent of twenty-five sacks of guinea-corn in the local marketplace. A rich farmer's marriage expenses on a budurwa extended to N1,000 (fifty sacks). In 1998, the normal marriage expenditure of a bridegroom or his father was the money equivalent of from 19 to 31 sacks of guinea-corn (N30–50,000, in a much-changed currency), whereas the richest trader in the village spent on the goods in sadaki alone N100,000 (62 sacks). I found no evidence of a decline in the burden of marriage expenses.
Villagers had negotiated their adherence to ‘purist’ Islam in such a way as to avoid a radical change in their sense of kinship obligations. In this regard, it is crucial to bear in mind three features of the local social economy. First, rotating marriage gifts between biki (marriage feast) partners had always defrayed a significant portion of marriage expenses which appear on the surface to be excessive. Second, these partnerships linked richer and poorer villagers – both men and women. Third, women had always been the key accumulators of small livestock. In their capacity as mothers, they very often took on the responsibility for providing their sons with the money to buy the bride's cloth and special clothing (lefe) – a substantial category of marriage expenditure. Crucially, then, expenditure on the marriage of children was a joint undertaking by husbands and wives. In conclusion, the preoccupation of Marmara villagers with reformed Islam was much less concerned with its impact on domestic or gender relations than with its implications for political economy. Essentially, their concern was in a discourse of social justice (adalci).
In summary, we can see that conversion to Islam was a process spanning decades, in which the very ideas of what constitutes Islam were subject to change. Second, public discussion of right Muslim behaviour, and movement to a new form of Islam, were intertwined with debate about the morality of the larger political economy in which the community was embedded. Last, deep notions concerning the nature of moral behaviour were to a considerable degree continuous from a pre-Muslim past – particularly in the sphere where intra-household and inter-household relations are reproduced over time through marriage.
THE IMPACT OF ISLAMIC CONVERSION ON GENDER RELATIONS
Though in 1977–9 Marmara had been a Muslim community for some ten years, there were still five Maguzawa households with thriving work units (gandaye, singular gandu) to make possible comparisons. I was also familiar with a large Maguzawa household three miles from the hamlet in the still heavily Maguzawa area where I farmed.
Conversion to Islam brought deep changes to rural economy and, as a result, to the relations between men and women. In Maguzawa households, the male head of the work unit farmed not only with his married sons in gandu, but also with his wives and the wives of his married sons in gandu. The gandu assembled for three – at most, four – days per week. On other days, wives and sons farmed their own fields and those given to them by the household head. The household granaries were used for five months during the farming period to provision the farming unit. For the remaining seven months, the household head depended on food supplied by his wives, and sons tended to form a common unit for grain provisioning with their respective mothers.Footnote 7 After conversion, women gradually withdrew from all farming operations except planting and harvesting, in which they remained heavily involved. The household head became responsible for provisioning his dependants throughout the year. Islamic conversion completely restructured the household as a production and a consumption unit. It reduced the power and autonomy of women in agriculture. Equally important, it created a huge demand for male hired labour. Thus, the commodification of labour was largely the result of Islam (Clough Reference Clough1995).
Maguzawa social life centred on beer parties revolving between households. Women bought the grain, brewed the beer, and sold it to men. The image in my mind is of married women relaxed at the door of their huts, chatting amiably and informally at these parties with married men from other households. And there was no sight more rueful than men sitting disconsolately when the women of a household told them, ‘We are not brewing today’.
After conversion to Islam, married women still moved freely from house to house provided that they wore a very open shawl, but their public position was much reduced. However, although women withdrew from most farming operations, other forms of economic action expanded after conversion. They became more involved in the preparation of cooked food retailed by their unmarried daughters in the village. They continued to farm through their husbands and hired labourers. Revolving savings-and-credit associations (adashi) linked most women.Footnote 8 The income from all of these activities they invested in small livestock and sometimes in cattle. They reserved the sale of livestock for expenditure on the naming, childbirth and marriage ceremonies of their sons and daughters. Covertly, husbands often turned to them for loans, in order to reduce their dependence on male creditors. Mainly, however, women's assets constituted a social fund reserved for the needs of their children. It is more revealing to speak of a change after religious conversion in the nature of the economic power of women rather than of its net reduction.
The shift from Maguzawa religion to Islam was consistent with a high degree of continuity in gender relations. The main preoccupation of men and women remained the development of a stable relationship between husband and wife, and the pursuit of new marriage partners. These goals are not as contradictory as they seem. Incorrigibly, a man with one wife sought a second and, if he had the income, a third (see also Table 1). Divorce was easy, if legally tedious. It entailed wearying trips to the Alkali Court, in which the alkali urged reconciliation (‘Ku daure! Ku daure!’ – ‘Tie yourselves together!’), followed by a final trip at which the divorce was settled. Most men and women had been through several marriages.
The thinking of both men and women was dominated by the prestige of polygynous marriage. In 1978, 48 per cent of all household heads had two or more wives. The prevalence of polygyny was in marked contrast to Batagawara in northern Katsina, where Hill found that only a third of heads of farming units had more than one wife (Hill Reference Usman1972: 23).Footnote 9 Polygyny must also be seen from the women's perspective. Almost 70 per cent of the wives of household heads had co-wives (Table 2).
After conversion to Islam, the withdrawal of women from most forms of farm labour actually enhanced the value of polygyny among men. For it signified a man's ability to attract more than one wife. In women's eyes, entry into a polygynous marriage for the first time often increased their economic security and social prestige, since a man's capacity to have more than one wife depended on his income. Moreover, it meant that they were sharing domestic duties with co-wives. Partial seclusion did not prevent wives from visiting their women friends in other households or hamlets.
Easy divorce engendered a crucial distinction – between the previously unmarried girl (budurwa) and the previously married woman (bazawara). Zawarawa were much sought after by men seeking a new wife. For a woman, the period between marriages was a time of great freedom. Even in later life, a bazawara could be subject to attention from various suitors.
Mothers and sons remained close, regardless of the mother's divorce or re-marriage or removal to a distant place. The close bond engendered an important social distinction – between full brothers (yan daki daya – sons of the same marriage room) and half-brothers (yan riga daya – sons of the same robe). Full brothers retained a strong sense of social and economic obligation to each other, whereas relations between half-brothers were more distant. To overcome individual disabilities, full siblings often formed a unit of joint investment in trade, livestock or oxen for ploughing.
Mothers provided their daughters with clothing, cosmetics and spending money. A daughter's sense of obligation to her mother was so strong that the mother usually remained throughout her lifetime the arbiter of her daughter's marriages. Her consent to a daughter's marriage marked a new development in her social and economic position. The son-in-law had to give her a continuous stream of gifts, and defer to her advice, if his marriage were to last. The mother-in-law was an important figure in rural society. She not only made the marriages. She broke them.
Because marriage was central to economy and culture, the reciprocal exchange of gifts (gudummawa) at marriage ceremonies was an important institution linking people related by agnatic or affinal ties. An important kind of friendship also flowed from the centrality of marriage – biki partnership. Every man and woman had a number of biki partners. Biki partners contributed money at each other's marriage and naming ceremonies. Over the long term, the relationship was symmetrical regardless of income differences: equal amounts had to flow in both directions. In the short term, one partner might be giving more because of the disparity in the number of their respective ceremonies.
The institution contained an element of redistribution. As a person's income increased, the number of his or her biki partners rose – often poorer people. At the same time, biki provided insurance. The total donations which a person made in the interval between two of his or her own ceremonies were returned by the various partners at the second ceremony. Since smaller farmers generally had smaller households (Table 1), several years usually elapsed between their own ceremonies. The total donations which they had made to biki partners in the interval – returned to them at their second ceremony – therefore defrayed a considerable share of its cost.
Beneath changes in religion and political economy, the men and women of Marmara manifested profound continuity in their ideas of what was ultimately desirable and morally obligatory. The achievement of a stable polygynous household arrangement remained the major goal of life. New Muslim forms of sharing – sadaka and zakka – had been added to the old reciprocity of gudummawa and biki partnership. As wives or potential wives, mothers, sources of household saving, and mothers-in-law, women were still the central focus of social exchange and economic life.
‘Man’ and ‘woman’ among the Hausa of Marmara
From these continuities and changes in relations between men and women I construct my understanding of gender identities in this part of Hausaland. ‘Man’ (namiji) and ‘woman’ (mace) were not highlighted linguistically in terms of ideal sexual or physical characteristics. Rather, they were conceived of by reference to parenthood and offspring. Gender identity needs also to be related to social ‘age’. Notions of social age are complex in this polygynous society. The apex of ‘manhood’ was the social status and moral responsibility attached to being a head of household (megida). But successful men often sought their third or fourth wife among women much younger than themselves. This sometimes placed them in a position of deference to a much younger man – the father or senior brother of the bride, who had given her in marriage. And women, as mothers, looked forward to a trajectory of increasing power from the circulation of their daughters in marriage.
The organizing concept behind notions of masculinity and femininity was hidima. Dictionaries have translated this as ‘serving a person’ (Abraham Reference Abraham1962) or, much earlier, as ‘service, administration’ (Robinson Reference Robinson1925). In Marmara, hidima seemed to have the special connotation of ‘responsibility’ and was used of a range of personal responsibilities to kin – especially the responsibility to provide oneself and one's children with marriage partners, an onerous burden where increasing wealth leads to more marriages in an increasingly polygynous household. The notion of hidima is other-oriented, cross-gendered, and conveys the sense of ‘obligation’.
CONCLUSIONS: A LONG-DISTANCE STRUCTURAL COMPARISON
In their political aspect, the changing relations of rural political economy in Marmara are still dominated by men.Footnote 10 In their economic aspect, however, the persistence of polygyny at the centre of accumulative process ensures that women remain the critical preoccupation of men in their thinking about personal economic growth. Since divorce is easy, men must satisfy wives in many ways for their marriages to endure. This gives the mothers of wives their influence over marriage negotiations and, more generally, over marital stability.
Islam in Marmara has been dependent on the mallams who have emigrated there from regions of Kano Emirate with a much longer tradition of Islamic knowledge. Both the dissemination of this knowledge and public ritual have been skewed heavily toward males, almost to the point of female exclusion. In this particular sense, Islamization has reduced the public involvement of women in religious experience. This may change: in other parts of northern Nigeria, Abubakar Gumi and his followers in the Yan Izala have worked for a modern Islamic school system incorporating the education of women (Loimeier Reference Loimeier, Westerlund and Rosander1997: 294). Certainly, the pervasively communitarian edge in Marmara to thought about the divine and about moral obligation is consistent with my experience that notions of both manhood and womanhood were anchored in an idea of the responsibility involved in performing the functions of parenthood.
The salience of political economy to our understanding of local gender relations is highlighted if we compare the changes in Marmara with those in a very different African society – the Giriama of Kaloleni District in coastal Kenya studied by David Parkin. Not only distance but also time separates my study from his, conducted in 1966–7. However, the relevance of the comparison is that both societies were undergoing intensified commodification and Islamic conversion, and both were polygynous. Given similar tendencies in commerce and religion, it becomes possible to isolate the local structure of their political economies, and test its impact on gender relations. Put differently, a comparative anthropology of different times and spaces can be useful where the existence of structural similarities throws into relief the significance of structural differences. Essentially, I will argue that, among the Hausa, men's accumulation of women as wives was primarily economic, whereas among the Giriama, this accumulation of women was political. This difference in the nature of accumulation shaped the profound difference between the two societies in their gender relations.
Among the Giriama of Kaloleni, Parkin showed, the greatly increased production of copra for the international market gave rise to a new social distinction between young ‘enterprisers’ and older ‘losers’. Islamic conversion enabled young enterprisers to achieve ‘ritual distinctiveness’, and thus legitimated their withdrawal from the sharing of palm wine with the non-Muslim majority, and from the financial drains of clientage. It therefore helped them to conserve resources for ‘accumulation’. Parkin's main argument was that ‘accumulation’ by young enterprisers was masked by a series of cultural devices which kept up the appearance of social continuity – the mediating role of elders in the sale of palm trees and land, the idiom of traditional inter-generational conflict in funeral ceremonies, and heavy expenditure on bridewealth. These customs mystified the radical transition from a redistributional to an accumulative mode of exchange (Parkin Reference Parkin1972: 98–9).
Compared with Marmara, Islamization in Kaloleni was a very partial process – at least in 1967. Only 6 per cent had converted. Even then, conversion was limited to the Ramadan fast and a ban on commensality with non-Muslims. The Giriama only accepted a few of the Muslim rituals – and these were added to the body of indigenous customs (ibid.: 46).
Under the redistributional economy, men had taken their wives from poorer men. By disposing of bridewealth they gained political support. Wives were the ‘symbol of pre-eminence’ (ibid.: 73). Under the new accumulative economy, enterprisers tended to take their wives from other enterprisers, and marriage transactions were quicker, so that bridewealth could circulate as capital for investment (ibid.: 74). This still occurred within a socio-political order of agnatic lineages, whose elders constituted the majority in councils which allocated heritable property and widows. Elders were also influential in government courts and controlled sorcery, in which everyone believed (ibid.: 84, 102).
While divorce was common in the period soon after marriage, thereafter conjugal stability was high. Women's choice of a marriage partner was limited to suitors approved by their fathers. Women had very few ways of earning money. Apart from small gardens to which they were entitled, they worked the land of their husbands, who controlled the proceeds. They received no portion of the bridewealth given to their daughters. Parkin characterizes their economic status as ‘ineffectual … transacted rather than transacting’ (ibid.: 71–2). Clearly, women were pawns in strategies of male alliance rather than assets in a process of economic accumulation. Partly in consequence, the rate of polygyny (1.5) was not significantly different among ‘enterprisers’ and ‘losers’ (ibid.: 74).
The position of women was radically different between the Giriama of Kaloleni and the Hausa of Marmara. In Marmara, divorce was common throughout the adult life of women, and, as often as not, was engineered by the woman. While their first marriages were often arranged by fathers, these were subject to the approval of mothers, who remained the arbiters in further marriages. Women had many ways of making money, and often contributed a significant share of the bridewealth for their sons’ marriages. These patterns variously echo those of pre-Muslim days, when women had close farming relationships with their sons. In today's more commodified economy, female labour is still crucial at harvest. This gives women bargaining power over male employers. In contrast to the Giriama, the rate of polygyny rose with the wealth of male household heads (Table 1).
In developing these comparisons as an analysis in which we might discern general connections between political economy, religion and gender, I wish to avoid the reduction of any one sphere to another. And I wish to retain the axiom of mutual or reciprocal causation. Of course, this introduces an apparent vagueness. Moreover, the positing of connections amounts to a striving for something like ‘conclusiveness’. Logically, there seem to be two escapes from these dilemmas. One is to follow in the path of Evans-Pritchard and Fortes, who argued synchronically that, while one sphere (for example, economy) may set constraints on another (for example, religion, or kinship), the social relations and ideas in each palpably retain their distinctiveness and autonomy (Evans-Pritchard 1956: 320; Fortes 1969: 229–31). The other is to argue diachronically that, while one sphere may indeed have more weight than another in their reciprocal causation over time, there remains a ‘strange … space … of human freedom’ based on the difference between ‘individuals’ and ‘society’ (Clough Reference Clough, Clough and Mitchell2001: 246; cf. Foucault Reference Foucault, Dreyfus and Rabinow1983: 208–26). This is a thesis of ‘qualified indeterminacy’. It is based on the power of individuals to occasionally alter the apparent structural fit between different social spheres.
The wide order of political economy appears to be the single most important sphere, having a predominant influence on religion and gender. Here, we are comparing an order based on individual tax-paying farming families subordinate to an external ruling class (the Hausa of Marmara), with a system of agnatic lineages run by local elders (the Giriama of Kaloleni). Strangely enough, the Hausa system of external control (which extends back at least as far as the Sokoto Caliphate (Usman Reference Usman1974: Chapters 3, 4)) gives more prominence to the individual – as subject – than does the Giriama system, where adults are co-participants in rural lineages based on the mystical sanctions of elders. Denied political participation almost until the end of colonial rule in 1960, Hausa individual men have focused their energies on economic success. Women have played a crucial part in that success, and cannot be reduced to counters-for-exchange in a local game of political alliance, as among the Giriama. In the former order rural commoners, debarred from office beyond hamlet level, to all intents and purposes individually owned the land upon which they paid tax to outside rulers (for similarities and differences across northern Nigeria before and during the colonial era, see Hill 1977: 8, 211–12). In the latter, local elders ultimately controlled the use of lineage land and the exchange of women. Women have had greater freedom of manoeuvre in a system where all adults are defined in opposition and subordination to an outside class than in one where patriliny constrains people according to their social age and their gender.
Where (as among the Hausa of Marmara) people were until very recently restricted in their access to political power, the economics of polygyny become crucial to individual success. Where (as among the Giriama of Kaloleni) male elders are at the centre of local political power, the politics of polygyny predominate. In the former system, polygyny becomes central to a complex trajectory of economic accumulation; in the latter, the provision and receipt of women become symbols of relative political influence. In the former, males save over time in order to negotiate for women with a value that is primarily domestic and economic; in the latter, the value of women is domestic and political. As a result, in the former polygyny is likely to be associated with the gradual acquisition of material resources for bridewealth, and therefore, with the increasing age of male accumulators; in the latter, women are distributed for reasons of social alliance, and therefore the age of the accumulator is not necessarily a determining factor. In terms of a structural comparison, the externality of political power to the rural Hausa of Marmara has been paralleled by a high degree of socio-economic individuation (Tables 1 and 2 and Clough Reference Clough2003); whereas its internality to agnatic lineages among the Giriama of Kaloleni has led over time to a deep classification of human beings as elder male, junior male, or woman.
Seen in this perspective, the process of religious change is likely to be a secondary influence on gender relations. This is not to deny that, for example, Islamic conversion has an impact on gender. We have seen, for instance, that in Marmara conversion to Islam gradually but radically restructured the units of production. Quite separately, the journey into Islam is a journey from rural shrine to urban mosque, from local space to a hierarchy of religious learning across space (Gilsenan Reference Gilsenan2005: 27–54) – though some have seen it as an oscillation between rural and urban, particular and universal (Gellner Reference Gellner and Gellner1981: 1–86; Joffe Reference Joffe, Westerlund and Rosander1997: 55–77). In the very long term, this cosmopolitan enlargement of social space may, intriguingly, subject both males and females to a common identity rather than disparate identities (Evers Rosander Reference Evers Rosander, Westerlund and Rosander1997: 1–27). And the call to Islam is a divine call – from beyond gender. Thus, sharply separate notions of ‘man’ and ‘woman’ become altered, at least, in the direction of the uniformity of souls (Clough Reference Clough2006: 268).
Rather than focusing only on changing religious beliefs, we need to look more deeply at continuity and change in moral concepts. Following Howell (1997: 4–5, 11), we need to consider the principles whereby a people prioritize their various goals – all of them wished for – and thereby resolve dilemmas in action. Over the long term, the evolution of relations of production, exchange and power – both internal to and external for any society – requires a complex of ideas as guidance in any given present. Among the Hausa of Marmara, I have argued that hidima (responsibility to kin, especially the responsibility to provide oneself and one's children with marriage partners) gives purpose to their internal relations and adalci (justice or fairness) to their external ones. These notions structure their very understanding of changing political economy, religion – and gender.
These conclusions have compared social structures in which there has been continuing evolution. What might happen were there to be abrupt revolution in any structural sphere, is difficult to say. If there is a strange space of human freedom, then prophets – secular or religious – might rise up ‘unpredictably’ (Clough Reference Clough, Clough and Mitchell2001: 246).
ACKNOWLEDGEMENTS
I am grateful to Sibyl O'Reilly Mizzi for her constructive critique of an earlier and different draft of this article, read at the ninety-ninth annual meeting of the American Anthropological Association in 2000. I thank the convener, Ronald Lukens Bull, for inviting me onto the panel at which it was read. I am grateful for the comments of two anonymous reviewers on the penultimate draft of this article.
TABLE 1 Landhol ding, number of wives and household size: household heads, Marmara, 1979
* This includes the household head.
TABLE 2 Number of wives of household heads, Marmara, 1978
