Slave Women, Family and Property, Part 3

— Cecilia Green

WE HAVE ALREADY identified two tendencies within the slave community, both of critical social significance and impact, but actually involving minorities within the slave community: (a) concubinage with white men, resulting<197>especially in extra-residential or tenuous co-residential situations--in a species of extended matrifocal family (socially elevated by biological and symbolic, even if not social, white fatherhood), and (b) patriarchal, male-headed (extended) families, whose chief principals were elite (and therefore dominant) slave men.

But what does an examination of a full range of family practices, which is mindful of the possibilities of gender and ethnic contestation, have to tell us? What about the majority of slave women, who were neither concubines of white men nor partners in a relatively "formally sanctioned" polygamous set-up? These questions cannot be properly answered here, as such a project would require another paper; however, a critical if cursory assessment of selective attempts by others to answer the last question may stimulate a onsideration of the fundamental principles at stake.

Below, with this in mind, I review aspects of the work of B.W. Higman, perhaps the most prominent Caribbeanist in the field of slave family reconstruction (see also Reddock, 1988: 125-31). (Whatever one may think of his conclusions, Higman has earned his reputation just on the basis of the incredibly painstaking and comprehensive demographic reconstruction that he has done for the late period of slavery in the British West Indies, and for which we are all deeply in his debt.)

Slave Family Reconstruction. Household reconstitution analysis has become one of the most respected means used by historians for providing an interpretation of the residential and family arrangements of the slaves. However, it has become increasingly clear that the data upon which this type of analysis rests have been forced to carry a far heavier burden of proof than they are adequate to bear.

Nominal data from estate account books and slave registration returns, which should provide at best complementary or secondary information, have been forced to tell us everything we need to know about slave families. Furthermore, an interpretative bias is clearly in operation when the significance of certain values is over-inflated, in the midst of a prevailing picture of wide variation and a lack of ascendancy of a single hegemonic form.

Household reconstitution methods and analysis, as executed by their best known protagonist, Peter Laslett, have been severely criticized for their qualitative and historical limitations. Laslett (1972), using those methods for analyzing household composition, had found that households in pre-industrial England and Europe were typically small and nuclear in structure, and saw this as evidence of the transhistorical existence of the nuclear family and the myth of the pre-industrial extended family.

His approach has been criticized on several grounds by Colin Creighton (1980) and others. Most obviously, Laslett confuses the categories of "family" and "household"; moreover, he deduces household structure from co-residential composition at a single point in time; he therefore ignores the "developmental cycle of the peasant household," and fails to take into account relations of property (and the modes of property transmission) as fundamental in the constitution of family forms; finally, in basing his definition of family form solely on the criterion of residence, he fails to distinguish between "residential rights" and "residential behavior," based on survival strategies and demographic imperatives (Creighton, 1980; Berkner, 1976).

In a series of seminal and oft-cited articles, Anglophone Caribbeanist and historical demographer, B. W. Higman (1973; 1975; 1978) attempts to tease out certain patterns of "family type" from data listed in estate account books, estate reports and slave registration returns for Jamaica and Trinidad mainly. Higman himself is clear that the 1825 estate reports (from three commonly owned adjoining properties with a total of 814 slaves) that he uses for his Jamaican case (1973; 1975) "contain no information about kinship," and the slave registration returns give only the mother's name (and that, only when she lived on the property) and "say nothing about paternity" (1973: 528).

The co-inhabitants of the particular households that Higman sees as containing "elementary families" are entered variously for the three properties as "male, woman, her children," "woman, male, the female's children," or "male, the woman's male children, woman, her daughters" (ibid: 538). This particular "household type" (Type 4 of 13 types), automatically assumed by Higman to also constitute "the elementary family" type, accounts for 19.8% of all households and 25% of all slaves (ibid.: Table 2, 535). By adding in households of Type 7 ("woman, her children, and others"), which he rather nonchalantly further assumes as "being in many cases essentially a sub-type of Type 4," Higman is able to conclude that "almost 50% lived in households approximating the elementary family" (lbid.: 534).

The other households are variously divided into those containing a woman with her own children or, in addition, her grandchildren or her nieces and nephews and their children, those with apparent heterosexual couples by themselves, single-member or same-sex households, groups of apparently unrelated men and women, and other (extended) forms. The households apparently headed by women alone (containing, in various permutations, their children, grandchildren, nieces and nephews, and grand-nieces and -nephews), excluding the category appropriated by Higman for his so-called "elementary family" (i.e., the category of "woman, her children, and others"), constitute 13.9% of households and almost 16% of all slaves. "Others" in the appropriated category is apparently assumed by Higman to refer to a male spouse and his children perhaps? (Interestingly, he does not include Type 5, "male, woman, her children and grandchildren," in his elementary family group.)

If we include the appropriated category among woman-headed households (which does not seem unreasonable, given the recorders' obvious attention to perceived distinctions), such households constitute 27.8% of all households and a little over 40% of all slaves. Higman (1973: 534), however, feels confident enough of his categorization to speedily offer the suggestion that "the woman-and-children household type was far from dominant, whatever the importance of the mother-child link"; elsewhere, he is more unequivocal about its relative insignificance. It is a conclusion he makes much of, along with the one that the "elementary family" (consisting of a man, a woman, and her, his, or their child/ren) was modal, i.e. the single most frequently occurring type.

To strengthen his case for orderly heterosexual monogamy and parenting among the slaves, Higman relentlessly searches for proof of proper conjugal coupling in other less obvious "household types," using closeness in age as positive evidence of it. On the basis of equally questionable evidence, he is also anxious to demonstrate male initiative and male-status dominance in family formation.

But in terms of "structure," "function," "developmental cycle" and underlying "property relations," simple mating and co-residence, even when proved, do not necessarily a family make. It is therefore hard not to conclude that Higman is forcing and prejudging his data in the direction of a preferred "ethical" and cultural model. In fact, he makes a number of statements which show that he regards any recurring pattern of "matrifocality" (sometimes used interchangeably, and incorrectly, with "matriarchy") or "extended households" as evidence of instability, disorder and promiscuity, and a regular appearance of the "elementary family," "nuclear family" or "single family household" (again, all used somewhat interchangeably) as evidence of social and moral order and stability.

The confusion is repeatedly--though not consistently--a three-way one among "household composition," "family type," and a particular social and moral value. Higman does not consider any explanatory model that might attribute an overarching historically specific order and rationality to the clear evidence of non-"elementary" and non-nuclear forms (those that he tries to downplay), especially from the perspective of certain subject-positions within the slave community. Non-nuclear forms are seen either as failed nuclear families or as the result of extreme and extenuating circumstances. (Higman does confirm the regular, if limited, incidence of matrifocal colored domestic-slave families and "elite" male-headed polygynous families.) About the U.S. case, he notes that:

"[R]ecent studies of the black family both before and after the Civil War, based largely on detailed census sources, have challenged the traditional view of it as disorganized, unstable, and matriarchal. In fact it can now be argued that the slave family was not matrifocal but was characteristically a stable nuclear unit, the single-family household predominating." (Higman, 1975: 262-3)

Higman is aware that the Caribbean evidence cannot be so neatly packaged, and he notes that "in the Caribbean the slave family was subject to greater stresses than in the United States" (ibid.: 263). This qualification appears particularly confirmed by the data from Trinidad, a late-settled and still new sugar colony by 1813, when "[of] the 50% of slaves living in families... 44.2% were in mother-children units" (Higman, 1978: 170). But the discussion of the relationship between "stress" and family form is both inconsistent and revealing.

Higman has noted for Jamaica and especially for Trinidad, both with high proportions of African-born among the slaves, that the latter tended either to have no family connections at all or to be found in simple couples or nuclear groups. The creole (locally-born) slaves, who were in a more settled situation and had achieved generational depth in their kinship patterns were relatively more likely to be in matrifocal and extended family groups and less likely to be in simple nuclear units.

While the African pattern would appear to be a result precisely of disruptive enslavement, individuation, and isolation within the new community<D> and the obvious need to begin bonding from the simplest building block or cell, Higman (1973: 536) equivocates in the Jamaican case, saying that "[this] might suggest that the Africans attempted to maintain nuclear families, while the creoles, dislocated by the experience of slavery, were unable to do so; or it may simply signify that the ramifications of creole kinship were that much greater." He makes no attempt to discover the full potential of those ramifications, but he continues, somewhat contradictorily, in a later article:

"Where there was a heavy predominance of male Africans it was improbable that all of them could find mates, and few of them could live with siblings, parents, or collaterals. Thus, as long as the slave trade continued there would always be a large number of slaves who had no chance of finding themselves in family households. This was the most immediate and most brutal feature of the slave system, and its impact on family structure is not in dispute. The issue is whether slavery and the plantation destroyed the family, making it impossible for these slaves who could find mates to establish any orderly form of family organization and inducing a growing creole section to become dominated by the matrifocal family. It is necessary to look at those family households which did emerge, for even though the creoles inevitably possessed more kin than the Africans, it does not follow that a large creole population necessarily meant greater order in family structure." (Higman, 1975: 270-1)

What is Higman saying? That the African-born were the most physically disrupted in the short term but the least morally corrupted in the long term? Again, when discussing the Trinidad data, he observes that "the African-born were more successful than the creoles in establishing families centered on co-resident mates," which "must have been seen by most of the African-born merely as the essential building-block of extended or polygynous family types rooted in lineage and locality" (Higman, 1978: 171). He continues (ibid.: 171-2):

"Their creole descendants, however, either lost sight of these models (as part of the process of creolization) or were prevented from achieving them by the brutality of the slave regime (heavy mortality, separation by sale, miscegenation, Christian proselytism, and so on)."

For Higman, therefore, it was the long experience of slavery<197>and distancing from cultural memory--that was most socially dislocating, evidenced by the high incidence of matrifocality among the creoles (even though, he insists, the nuclear family was always either modal or dominant). Matrifocality, then, becomes an indication not just of disorder, but also of de-Africanization. But Higman, in spite of sudden and unexpected contradictions of his own thesis here and there, has given no sustained consideration to the possible emergence of a relationship between matrifocality, on the one hand, and a certain kind of social order and reconstituted cultural tradition, on the other. (In fact, Higman devotes almost no independent explanatory space to accounting for the huge numbers of his population outside of apparent "elementary family" forms or any apparent--heterosexual--family form at all.)

Questions need to be raised about this. It could be, and indeed has been, suggested that increasing creolization expanded and routinized or systematized the (reconstituted) kinship systems of the slaves, and that extended continuity of ownership in conjunction with the large size of West Indian estates militated against the sudden disruptions of slave families through sale (though not through mortality). Higman himself points out the relationship between creolization, large estate populations, and extended families.

The evidence for the United States has been quite the contrary: much smaller plantations, greater disruptions through sales, and a noticeably higher level of nuclear family formation than in the Caribbean. The far higher retention of Africanisms and parallel reworking of African forms (among them, matrifocal family structures as possible derivations of mother-children polygynous cells) have also been repeatedly noted for the Caribbean, not the least by the often misrepresented Melville Herskovits (1941) and his followers (see also Sudarkasa, 1980).

Recently, moreover, shocking evidence of systematic and widespread family separations (and the instilling of psychological terror through the constant threat of such separations) on a scale unheard of in the Caribbean has come to light in the work of Michael Tadman (1989) on the internal slave trade ("Negro speculation") between the Upper and Lower South in the first six decades of the nineteenth century. Tadman (1989: 170-2) estimates that forcible separations through sales and non-market transactions (planter migrations, gifts of slaves and estate divisions between beneficiaries) "probably destroyed about one in three of all first marriages between Upper South slaves," and that almost one in two Upper South slave children aged fourteen and under were separated from one or both parents by sales and non-market transactions.

In light of all this, I propose that the time has come to re-think our notions of stability, stress and family dynamics. One could speculate that in the circumstances of externally imposed, far-reaching and traumatic demographic erosion, conjugal bonding, as the most immediately available means of cohesion and protection against vulnerability, takes on added significance. The subject needs to be pursued elsewhere. At this point, I wish to offer a number of leads for consideration in the study of slave family dynamics.

Matrifocality, or mother/woman-centeredness, must be tackled frontally, rather than residually, as a major tendency within the slave community. I see it originating from two directions: first of all as a ruling-class imposition but secondly as a re-appropriation and re-interpretation by slave women in keeping with their gender, class and ethnic consciousness and interests. Some of the households designated by Higman as comprising intact "elementary families" (a designation, incidentally, often reserved in anthropology for mother-child/ren units) might have easily been implicated in a matrifocal developmental cycle.

One is reluctant, however, to arbitrarily privilege matrifocality as a conceptual framework in the way that Higman and others have done for male-dominant nuclear family forms, the latter no doubt being a genuinely important tendency within the slave community. An examination of the system must stand on its own for now.

Ruling Class Imposition. Manyoni (1980: 93) points out that slaves were "considered outside the formal institution of legal matrimony" and "both State and Church shared a strong disapproval of slave marriages, a situation that continued unmitigated until the eve of emancipation." His is one of the few accounts that begins to look systematically at ruling class ideology and practices with regard to slave marriage and family in the context of ruling class interests.

For well over a hundred years, the West Indian planter class had firmly opposed any metropolitan or local efforts to Christianize the slaves and invest them with spiritual and legal matrimonial rights. As far as the planters were concerned, investing the slaves with any kind of officially sanctioned civil and human (especially spiritual, moral and intellectual) rights and freedoms would, in their own words, "impair their value and price" and interfere with the freedom to economically and sexually exploit their slaves in the enjoyment of absolute and unmediated rights of ownership.

The implications of this have not been fully explored from all the angles. Three of those are critical: (i) the race/class; (ii) gendered/patriarchal; and (iii) transnational or dually located character of the British West Indian planter class.

First, as Manyoni points out, Christianization, education and (clerical/legal) marriage for the slaves were the "three closely intertwined" bogeymen that haunted West Indian slave regimes. Whites found the idea of converted, literate, and "properly" married slaves to be offensive and dangerous. It eroded and contaminated the exclusivity and sacrosanctity of European privilege and identity by allowing slaves to partake of them, if only distantly and symbolically; it weakened the availability of the slaves as objects of unmitigated exploitation and of the murky colonial underside of European "civilization" and fantasy; and it potentially threatened the physical security of the whites and their property by inciting the slaves to visions of total freedom.

Manyoni (1980: 99) cites a contemporary observation that "when coloured concubines adopted Christianity, `their renunciation of base connections gave the greatest offence to the white community.'" It occurred to whites that "uppity" black people who went to the great lengths required to become literate, convert to Christianity or obtain a clerical marriage, might have pretensions to being something other than their sexual and economic tools. In the overwhelming black majority situations of the Caribbean, such pretensions were extremely threatening to the planters' race/class niche.

At the same time, having denied their slaves basic human rights and what they considered to be the finer spiritual and moral trappings of European civilization, whites claimed to be disgusted and repelled by slave women's alleged promiscuity, "shocking licentiousness," and lack of godliness, or at least its Christian variant. The alleged "culture" of the oppressed became a subterranean cesspool in which the whites disposed of and concealed their refuse of culpability, conferring it instead upon their victims as the latter's "native" heritage. As it turns out, even when they were extended legal and spiritual matrimonial "privileges" in the late eighteenth and early nineteenth centuries, few slaves were interested.

Second, Euro-Christian marriage did not confer equal rights on the partners, but of course gave men superior rights over women and the family patrimony. Thus, intimately bound up with the withholding of legal marital rights was the desire to avoid having to endow black male heads--perceived through patriarchal lenses as the chief potential rivals of ruling class men--with an independent and corporate (familial) base of authority and property, one that might, moreover, mediate and limit exploitative planter-class access to their women and children. This has already been discussed; what has not is the question of alternative arrangements that might have been followed.

We have already seen that the coding practices of the planters routinely recorded domestic units as defined by a mother and her children, regardless of the (usually duly noted) presence of a co-resident male and his biological relationship to the children. The other thing would be to find out how means of subsistence and subsistence production, most notably the slaves' personal provision grounds, were allotted; this would tell us something about class-induced relations of reproduction and property by gender, important imperatives in family formation. [For a discussion on the role and importance of the provision grounds see part 1 of this essay in ATC 40, especially pages 35-37--ed.]

The evidence from contemporary commentators appears to be genuinely mixed and sometimes indecipherable, due to the universal and indiscriminate use of the masculine pronoun and obvious masculinist bias in certain texts or parts of texts. Higman himself notes the recording in the Old Montpelier Account Book of detailed statements about gardens and provision grounds belonging to each household, but, puzzlingly, does not call upon this information to aid him in his assiduous family reconstruction exercise.

The slave laws themselves, however, sometimes prove to be quite unambiguous and, embedded as they are in the heart of a fundamentally patriarchal project, startlingly and meticulously gender-conscious in their language. Indeed, one often encounters the irony in many discussions of slave reproduction and economy that the only gendered language present speaks to us directly from the text of the slave laws of the eighteenth and nineteenth centuries.

Late eighteenth-century legislation in Grenada directed that each adult slave (over fourteen years) should receive "his or her proper ground"; an 1831 law in Dominica stipulated the allotment of "not ... less than half an acre for each individual slave"; in 1834, it was "ground adequate, both in quality and quantity, for his or her support, and within a reasonable distance from his or her usual place of abode"; 1784 and 1786 revisions of the Code Noir in Martinique decreed that each adult slave was to receive a small plot of land to cultivate on his or her own account (Marshall, 1991: 52; Trouillot, 1988: 74-75; Tomich, 1991: 72). Of the provision grounds in Martinique, where practices were similar to those in the British islands, Schoelcher wrote, "[they] pass them on from father to son, from mother to daughter, and, if they do not have any children, they bequeath them to their nearest kin or even their friends" (quoted in Tomich, 1991: 81).

It seems clear that the slave masters individuated the adult status of the slaves across gender in the allocation of means of reproduction, and endowed women directly as individual adults and as mothers responsible for the welfare of their children<197>not through male heads. This did not mean that pooling of provision grounds by marriage partners did not occur and become a "de facto," socially recognized family enterprise; it meant rather that the allocation of provision land was not premised on or preconditioned by the male-headed or conjugal family as a corporate entity or as the object of endowment.

Insofar as a particular type of gendering occurred it was that which produced a direct equation between biological motherhood and adult status for women, especially in the later period of slavery, when motherhood increasingly became the occasion for direct endowment and "reward." Slavery may in fact have reinforced the "cult of the mother" tendencies present in many West African pre-capitalist cultures as well as their assumption of a direct conversion between motherhood and economic providership and responsibility. Parallels also exist with the role of women as the prime horticulturalists and marketers in many West African societies.

In plantation society, of course, the worker-breeder-sex object-provider female-role complex was the peculiar invention of an ethnically and socially alien ruling class and was predicated on its convenience and self-aggrandizement. But there is no doubt that slavery and racism were based not just on the Eurocentric creation and projection of (fictive) difference but also on the exploitation and displacement of (real) difference and the inversion of the culture of the other within a social relation of inferiority and subordination to European ruling class "civilization."

The project of making the slaves over into (first) little English men was a belated and relatively unsuccessful one, limited, at least in terms of a more general consensus within the planter class, to the later and increasingly desperate years of slavery. In directly endowing slave women with the means of subsistence, the planters no doubt felt they were confirming them in their status as means of re/production themselves and as beings located outside the requirements of civilized society.

Third, the social space of the West Indian planter class was transnationally split. Those of the old internal/external debate who assume that social determination is always necessarily confined within the internal boundaries of physically defined social formations understand little about colonialism, and even less about the West Indian case.

The schizoid social character of the West Indian planter class was represented precisely by the physical and social separation of certain fundamental conditions of their reproduction--slave-worked sugar plantations and metropolitan investment (and cultural) capital--into different worlds. While they were late "of" the center, the planters' economic interests were tied to enterprises physically and socially located in the periphery, and, based on varying degrees of residence and absence, their particular social, i.e. everyday, character was heavily invested in the little fiefdoms they had created and over which they presided.

No matter where they happened to be at a particular time, the creole planters suffered from a kind of transnational parochialism, wanting to be absolute despots "locally" (i.e., in the colonies), but to continue to maintain their leverage and life systems "transnationally" or through their metropolitan connections. In one recurring version, the transnational split was accommodated entirely by proxy, with the proprietors remaining in the metropole and their second-class surrogates presiding over their plantations and replicating the dualism of the system on their behalf.

The "dual-marriage system" must be understood in this context. Mintz (1989 [1974]: 302-28) points out that the non-Hispanic settler entrepreneurs of the Caribbean felt themselves under no compulsion to transplant or reconstitute systems of "social respectability," into which, moreover, they could integrate the slaves (albeit in a subordinate way). He contrasts this with the situation in the Hispanic Caribbean, where the settlers "came to stay" and totally separated themselves from Spain in a process of hybridization which rendered them, in the end, "Cuban" or "Puerto Rican," and which provided "familial and acculturational models" to which their slaves had access.

A contrast also existed with the Southern planters, who recreated resident cultures and singular hegemonic systems which virtually imprisoned their slaves within those orbits ("in the shadow of the Big House"), providing them with few escapes from the identity of the Euroamerican Subject's Other. Barbados came closest to this model in the West Indies, while South Carolina approximated the Caribbean model in many ways.

The planter class in the Caribbean did not<197>or could not--seek to erase or incorporate the culture of the slaves into a singular hegemonic design, but rather sought to condemn and marginalize it as the illegitimate or pagan underside of a bicultural mode, which corresponded to the dual and hierarchically related sites of the colonial project. Through their differentiated and contradictory involvement in both--one as the subject of their self-affirmation, the other as the object of their ab/use--the planters helped to reproduce and sustain the one as "legitimate" and the other as "illegitimate" across social and physical borders, but felt no need to reconcile them within a strong resident Euroamerican model.

Slave Women's Appropriation of Matrifocality. Some comments by Bush may well set the tone for this final discussion. She accepts that the nuclear family may have existed on a greater scale among the slaves than has been assumed, ("modeled on African rather than European cultural precedents,") but finds that:

"[M]ore flexible definitions of family and marriage are needed to account for the primary importance in slave society of consanguine (extended) as opposed to conjugal (nuclear) relationships. Complex extended family links existed in the slave community: In the absence of blood kin, "fictive" kin, often newly arrived slaves, were "adopted" to reestablish traditional African-derived kinship patterns and ensure cultural continuity. A strong sense of community existed among slaves, and interpersonal relations were marked by "warmth and sentiment." Children were cared for communally as were the elderly, who were afforded considerable respect in accordance with African practice. Within this broader structure, the polygynous unit probably only accounted for a minority of slave marriages. Marriage did not imply a lifelong relationship between two adults in a co-residential nuclear unit, and partners were often incorporated into wider extended family arrangements and may not have been continuously co-residential as in polygynous relationships.' (Bush, 1986: 120-1)

Bush is mostly concerned to challenge the "promiscuity" and "disorganization" thesis from an ethnic and non-bourgeois class perspective, so that she tends to reinterpret gender roles wholly within the context of (reconstituted) West African ethnicity. She is less concerned to consider ways in which slave women, acting autonomously and for themselves, may have seized hold of opportunities for self-determination inadvertently or indifferently passed on by the system itself.

Slave women obviously manipulated, in their favor, a certain convergence or accommodation between the demands made upon them by the system and the cultural resources brought with them from West Africa. Bush's observations are important in that regard, and point to critical elements in the reconstitution of rational cultural systems in the new setting. The new setting often allowed slave women to rework certain West African cultural principles in a way that was optimal for them.

Bush mentions the examples of the economic independence of women, ease of divorce, the relative separation between sex, childbearing, co-residence and marriage, respect for seniority, the relatively greater focus on consanguineous as opposed to conjugal relationships, late weaning of infants, and so on. The system accommodated the reconstitution of these ethnic features within a base which lacked a formally sanctioned and collectively organized patriarchal authority, communally speaking, that is. This relative political neutralization of gender presented women with certain opportunities. Indeed, it is instructive to us that others saw it differently:

"This economic independence of slave women was identified by Matthews as a causal factor of what he termed the crisis in the West Indian family. He argued that on slave plantations men and women each had a separate economic base, as each had an independent land grant (possibly provision ground). Women controlled their own money and according to Matthews, `tampering with the reputed wife's money or garden produce was a major crime, as disreputable as theft by a common stranger.'" (Reddock, 1988: 120-1)

"Reddock (1988: 122-5) confirms the rejection by both slave men and slave women of legal marriage in favor of consensual unions (and not only on account of its inaccessibility). Even today in the Caribbean, marriage, as the pre-eminently sanctioning and simultaneously all-inclusive institution for sexual relations, childbearing, conjugal community, and spousal and filial property rights, is a class-exclusive phenomenon. Between 60-80% of all children continue to be born outside of legal marriage, which is not absent from but occupies a different place in working-class cultural repertoires and biographies.

Most working-class women marry, if at all (a large minority never marries, legally or consensually), after extended periods of childbearing and sexual relations, sometimes co-residential, sometimes not, often involving serial partners/genitors who may or may not become significant and active social fathers. (I say "significant and active" because in the Caribbean the physical fact of paternity always has social meaning.) What should be judged as a rational (if still gender-discriminatory) way of pacing entry into different union forms and parenting arrangements from a matrifocal base<D> continues to be deemed "illegitimate" (although no longer in law).

The history of marriage among the slaves and ex-slaves has shown that occasional surges in its popularity have been induced by a direct link with economic opportunity. For example, those ex-slaves who chose to leave the plantations and join the "free village movement" under missionary leadership adopted marriage as part of their new status as free peasants and parishioners on land that they paid for and to which they had clear legal title (see Mintz, 1989 [1974]: 157-79).

Speaking of Jamaica, however, M.G. Smith (1962: 261) points out that although the majority of the senior population of those villages adopted marriage, "their juniors ... did not, and by 1861, the missionary impetus had spent itself without displacing the traditional mating forms." None of the externally induced marriage "spurts" strewn across post-emancipation history managed to bring working-class Afro-Caribbean family forms into close conformity with the European pattern on which West Indian marriage laws were modelled. This is true even though Christianity gradually increased in popularity to eventually attain mass dimensions.

A more substantive erosion, however, took place with regard to the property relations that had evolved within the slaves' sub-economy, shaped in accordance with their gender, ethnic and class interests. A system of "family land" had emerged around the provision grounds of slavery, based on principles not unlike those of West African unilineal descent groups, according to which all the descendants in a determined family line had rights in the communal property, which could not be alienated (except by consent of all the heirs). These rights could be actively exercised at any time and "[included] all the family in perpetuity" (Clarke, 1966 [1957]: 60).

A little over one hundred years after the abolition of slavery, Edith Clarke's Jamaican informants explained to her the difference between "family land," which many of them traced back to their male and female African slave ancestors, and "bought land," which conferred individual freehold title and could be alienated at will. ("Bought land," however, could be converted into "family land.") In her famous study, My Mother Who Fathered Me (1966 [1957]: 48-9), she explains:

"A woman, whether wife or concubine, has no customary `rights' to land inherited or bought by her spouse. She is not, according to the traditional system, as distinct from the legal code under which she has the same rights as in British law, in any circumstance his heir. If there are no children of their union, the property reverts to his brothers and sisters and their descendants. At the same time it is generally conceded that the wife or concubine has a right to live in the family home and on the land during her lifetime. Such provision may be made in the man's will or verbally. Not infrequently there is the direction that the right of use lapses if she remarries or takes a concubine. But even where there is no will, it would be rare for the family to dispute her right, whether as the mother of his children or not, to live on the family land."

Sisters and brothers were equally endowed with rights in family land, and wives had lifetime rights of residence by privilege (and might well have access to their own family land). Clarke's informants explained that the land could be transmitted "through the blood" (through the mother) or "by the name" (through the father), and it was apparent that for those sisters who stayed and became the occupiers and trustees of the land (and in one of the villages of her study that was precisely the case), their "concubines" (so named in the text) or consensual partners had only tenuous rights at best. Such (matrilocal) cases provided the clearest examples of matrifocality at work.

But the system of family land, which was less discriminatory to women or in fact favored them, was considered to be an inferior and less economically productive species of property. The post-emancipation system increasingly promoted freehold title and male cash-crop farming and marginalized family land and female provision farmers.

Conclusion

A number of factors eventually conspired to effectively evict women from the land in the post-slavery period: the introduction and official promotion of cash- crop farming, which favored men and reinforced official neglect of provision farmers who were mostly women; the increasing introduction of individual land titles and male line inheritance; male-centered colonial institutions of agricultural extension and finance as well as land settlement schemes; the increasing redefinition of women as dependents and housewives rather than as farmers and workers; and later on, mechanization and rationalization on the larger estates which favored skilled or semi-skilled male workers.

In the final analysis, therefore, while the relentless efforts on the part of the planter class and elements within the colonial state to severely limit and even abort the development of an independent Caribbean peasantry are well known, what is less well known is that a key element of this strategy involved the almost complete elimination of the concept and reality of the female food grower as an independent farmer, and the privileging of the male cash-crop grower, within the now properly subordinated peasantry.

Working-class men were drawn increasingly into the front line of class relations, prompting a significant shift in the contours of class and gender, both domestically and extra-domestically. Where they could, Afro-Caribbean women fought to hold on to the hard-won roles that they saw as worth having, and in certain areas--for example, "huckstering"--the lines of continuity are evident. They also aggressively sought to maximize the new opportunities that came their way. The result of this can be most clearly traced in the phenomenal strides they have made in the past thirty years in the area of education, prompting at least one male Caribbean scholar to talk about the "rise of matriarchy" and the "marginalization of the Black male" in the Caribbean (Miller, 1986; 1988).

Generally, however, as the division by class and race or color has become less absolute and more porous, gender has been increasingly manipulated as a factor in social dynamics, considerably complicating the articulations of color, class and gender. Certainly, there seems no better time than the present for Caribbean women to seek lessons from their past.

References for Part 3

Berkner, L., "Inheritance, Land Tenure and Peasant Family Structure: A German Regional Comparison," in J. Goody et al. (eds.), Family and Inheritance. Cambridge: Cambridge University Press, 1976.

Bush, Barbara, "'The Family Tree Is Not Cut': Women and Cultural Resistance in Slave Family Life in the British Caribbean," in Gary Y. Okihiro (ed.), In Resistance. Amherst: The University of Massachusetts Press, 1986.

Clarke, Edith, My Mother Who Fathered Me. London: George Allen & Unwin, 1966 [1957].

Creighton, Colin, "Family, Property and Relations of Production in Western Europe." Economy and Society, Vol.9 No.2, 1980, 129-197.

Higman, B.W., "African and Creole Slave Family Patterns in Trinidad." Journal of Family History, Vol.3 No.2, 1978, 163-180.

--------, "Household Structure and Fertility on Jamaican Slave Plantations: A Nineteenth-century Example." Population Studies, Vol.27 No.3, 1973, 527-550.

--------, "The Slave Family and Household in the British West Indies, 1800-1834." Journal of Interdisciplinary History,<D> Vol.6 No.2, 1975, 261-287.

Herskovits, Melville J., The Myth of the Negro Past. New York and London: Harper & Brothers, 1941.

Laslett, Peter, "Introduction: The History of the Family," in P. Laslett and R. Wall (eds.), Household and Family in Past Time. Cambridge: Cambridge University Press, 1972.

--------, "Mean Household Size in England Since the Sixteenth Century," in P. Laslett and R. Wall (eds.), op. cit.

Manyoni, Joseph R., Extra-Marital Mating Patterns in Caribbean Family Studies: A Methodological Excursus. Anthropologica, Vol.XXII No.1, 1980, 85-118.

Miller, Errol, The Marginalization of the Black Male. Kingston, Jamaica: Institute of Social and Economic Research, UWI, 1986.

--------, "The Rise of Matriarchy in the Caribbean." Caribbean Quarterly, Vol.34 Nos.3&4, 1988, 1-21.

Mintz, Sidney W., Caribbean Transformations. New York: Columbia University Press, 1989 [1974].

Reddock, Rhoda, "Women and the Slave Plantation Economy in the Caribbean," in S. Jay Kleinberg (ed.), "Retrieving Women's History. Oxford and New York: Berg, 1988.

Smith, M.G., West Indian Family Structure. Seattle: University of Washington Press, 1962.

Sudarkasa, Niara, "African and Afro-American Family Structure: A Comparison." The Black Scholar, Vol.11 No.8, 1980, 37-60.

Tadman, Michael, Speculators and Slaves. Madison: The University of Wisconsin Press, 1989.

Tomich, Dale W., "Une Petite Guinee: Provision Ground and Plantation in Martinique, 1830-1848." Slavery & Abolition, Vol.12 No.1, 1991, 68-91.

Trouillot, Michel-Rolph, Peasants and Capital. Baltimore and London: The Johns Hopkins University Press, 1988.

January-February 1993, ATC 42