8 Family, Kinship, and Community

Flight, rebellion, resistance, and violence could not be the only responses possible to slavery in Brazil. The majority of Africans and their descendants tried to survive the slave experience and create as normal a life as possible in the context of this harsh regime. Thus, family and community were a fundamental part of the Afro-Brazilian experience and in turn helped to mold and define the larger society outside of slavery as well. Forced to work for others and with little control over their lives, slaves began to learn skills, form families, and create kin and friendship networks that would survive the institution of slavery. They also found solace and community in their religion and brotherhoods.

Although they spoke a multitude of different languages and came from different cultural systems and nationalities, the African slaves arriving to Brazil sometimes shared a number of commonalties that may have helped bind them together in the New World. The opening up of Atlantic trade, for instance, contributed to the development of interregional contacts within Africa by creating ever-larger market areas. Meanwhile, the fairly constant expansion and reorganization of African states and societies over time brought many different groups into contact. Although the process of cultural integration on the African continent should not be overemphasized, neither should the diversity of Africa be exaggerated. Not every ethno-linguistic group comprised a distinct culture that was totally different from others. Moreover, many Africans of the slave trade era were multilingual and at least minimally conversant in the dialects and customs of their neighbors. Significantly, the circumstances of the Atlantic slave trade did not completely disrupt this pattern. Slave ships often extracted slaves from regions where intense multinational contact had occurred, and many of the slaves whom traders purchased had also slowly migrated to the coasts where they were sold, having passed through numerous territories and interacted with a variety of populations along the way. All of this may help explain why newly arrived Africans, when setting foot in the Americas, were sometimes able to create bonds of friendship and community with slaves from different nations. Finally, the long and intense nature of the African slave trade to Brazil guaranteed that there were large groups of slaves from the same regions and even the same ethno-linguistic groups clustered within regions of Brazil, and there are numerous cases in the Brazilian documents of slaves from the same ethno-linguistic group living and working together, or even being selectively purchased by forros of the same African origin.¹

Certainly, some New World masters often expressed preferences for acquiring one or another group of Africans on their estates, thinking it beneficial to cultivating a productive work environment. Others actively sought to diversify their slaveholdings, based on the belief that having too great a concentration of a particular group might foment resistance and rebellion. Yet there was no systematic attempt by slave traders to diversify their sources, and New World demand had no effect on the selection of Africans who arrived in America. Shippers purchased slaves where they could find them, and their availability for sale depended exclusively on African supply conditions and the forces of European competition. There may have been some ability for masters to selectively choose the ethnicities of the slaves they wanted within the internal slave trade that occurred in the Americas, as slaves sent directly from Africa to one colony or region were transshipped to others. Moreover, planters and masters often discovered that slaves whom they purchased had special skills in mining or agricultural tasks that they had not expected and thus tried to purchase more slaves with these skills. But overall, planters had no real ability to determine the African sources of their slaves and could only complement their Yoruba slaves, for example, by purchasing them on the internal slave market.

Specific events in Africa may have led to the clustering of some ethnic groups in the Americas. For example, a series of wars in the middle of the sixteenth century brought about the decline of the Wolof Empire in Upper Guinea, which in turn increased their availability on the slave markets of this period. Another very special case is the collapse of the Oyo Empire, which brought streams of the Yoruba into captivity in the late eighteenth century. Among arriving slaves, in cases where their ethnic identity was clearly given and recorded in the historical record, one can begin to explore the possibilities for the construction of culturally coherent slave communities based on mutually shared ethnicities. Yet such efforts must be carefully made and to some extent must remain speculative. Many slaves were simply listed as originating from the ports where they were sold rather than being assigned a more specific ethnic identity based on clan lineages and birthplace. The use of these generic port designations often makes it difficult to assert that certain groups of slaves may have definitively shared a language and culture and, consequently, could form a coherent ethnic “cluster,” as some have argued.

Although many slaves recorded their actual ethnic identities with great clarity in testaments and sometimes in other official documents, not all Africans were correctly listed by origin in American registers, and relatively few left such wills and testaments.² It is difficult to determine the extent to which the ethnicities declared by officials and owners for their slaves formed a conscious part of everyday slave identities, and even more, it is unclear to what extent illiterate slaves actually claimed these ethnic identities recorded for them by such masters and officials. A number of slaves were either slow to assert an African ethnicity, or, if they came to the New World as children, never really grasped the meaning of their African ancestry.

Finally, there remains the question, much discussed by Brazilian historians, whether there were not hostile relations between native-born slaves (Creoles) and African ones, and even between African groups that sometimes reflected antecedent hostile contacts in Africa itself. Marriage endogamy for both Creoles and Africans suggests a preference for association with one's own group, the often-hostile role Creoles played in African-led revolts, and finally the discriminatory admission practices of the African-based brotherhoods all suggest that there was no unified slave culture and that tensions existed along origin boundaries. At the same time, historians have shown cooperation, marriage across these boundaries, and hostility declining over time as the slave population became ever more Creole dominated. There also appear to be very wide variations by place and economy as well as by time, in these patterns of conflict or cooperation and identity.³

The cultural relevance of African ethnicities in the Americas was not always related to demographic strength. Within a given colony or estate, some slaves who may have been few in number originated from ethnic groups that enjoyed primacy over others – either culturally, politically, or both – or in turn were the first to arrive and define the rules of acculturation. Equally important, some aspects of African cultures were simply more transferable across the Atlantic than others or were found to be more popular among the broader slave population. Africans in the Americas had no state apparatus, no political classes, and their clan organizations were severely ruptured; thus, African beliefs associated with all these activities had difficulty crossing the Atlantic. On the contrary, those beliefs relating the individual to health and well-being, interpersonal relationships, and relation of the self to the cosmos were most likely to retain their power in the New World.

It is also clear that elements of the emerging Afro-Brazilian culture were influenced by European beliefs. Variants of European Christianity became the dominant religion among slaves, even if syncretized with large elements of African beliefs and deities, and even American Indian beliefs were found to have penetrated Afro-Brazilian rituals.⁴ Second- and third-generation slaves were raised speaking the language of the local master class. In their adaptation to peasant agricultural practices, Africans and their descendants, although sometimes using African technologies, were often found adopting European tools, technologies, and ways of life. The structures of social stratification within the emerging Afro-American community may also have represented an amalgam of two different worlds or even a response to the special conditions that slaves encountered under New World slavery. In some cases, differentiation in original African social status successfully made the transition across the Atlantic. In others, differences in slave status derived from struggles against the varying social positions that were assigned to slaves by their masters, which frequently did not correspond to the notions of stratification that emanated from within slave communities themselves. Consequently, although a hierarchy of status in terms of occupation and skin color was imposed on the slave population, internal slave divisions often did not necessarily replicate white standards.

Certain features of this slave culture were common to all slave societies in America, whereas others were more especially developed in the Latin American context. It is now generally accepted that in the slave periods in Cuba, Haiti, and Brazil, powerful movements of proscribed religious practices developed that were most heavily influenced by a syncretic arrangement of African religious deities. Santaria in Cuba, Vodoo in Haiti, and Candomblé in Brazil were the major manifestations of these new religions.⁵ These movements came fully to light in the postabolition period in these Catholic countries, but never arose to any significant extent in the Protestant societies. These essentially non-Christian religions were among the more significant features that distinguished Brazilian Afro-American culture.

Although whites viewed all slaves as equal before the law, the differential prices paid for skilled slaves as opposed to field hands clearly suggest that whites recognized important variations in aptitudes, abilities, and other individual traits. As for the slaves themselves, there were obviously some levels of stratification within their own commonality of bondage. The traditional definitions of social status among the contemporary free persons, however, are not totally applicable when examining slave society. Positions with control over resources or other persons – highly prized in the free community – were not necessarily those that guaranteed higher status within the community of slaves, or even those recognized by the price differentials given by whites. Autonomy and knowledge often played an equally important role. Autonomy was clearly related to independence from the control and supervision of whites, whatever the job, just as knowledge could be both of the African culture of the past or of the white culture of the present.

The life of the slaves in Latin America was primarily defined by work. With the exception of the very young and the very old, everyone spent most of their time engaged in manual labor. More than any other segment of the society, slaves were both the least sexually defined by their labor and had the highest rates of economically active employment of any class or group. For this reason, work dominated the life of the slave more than others in the society, and questions of work autonomy or dependency were of vital concern to slaves. In plantation societies, supervision of the strictest kind was the lot of the majority of slaves, but even here, relative control over one's time was available to a surprising number of them. On an average sugar or coffee plantation, gang labor involved only half of the slaves. Another third or so were craftsmen or had occupations giving them freedom from direct white or overseer supervision. In the rural slave populations that were not on plantations, there was also a distribution of jobs under close supervision on family farms as well as relatively independent families of slaves tilling lands on their own, or skilled artisans or muleteers who could escape direct white control. In the urban setting, domestics made up a large share of the labor force, came into close contact with whites, and were most tightly controlled. But all who worked on a self-hire basis or as independent craftsmen tended to have the most free time for themselves outside the normally controlled work environment.

Control over their time and labor permitted some slaves to achieve a fuller development of their talents and abilities. Short of total freedom, this was considered a highly desirable situation, and slaves who held these jobs had a higher status within the slave community. It was also no accident that many of the leaders of slave rebellions and other political and social movements came from these more autonomous slaves. Interestingly, some of these jobs were highly regarded by the whites as reflected in their prices, and some were not. Commentators on slave occupations noted that these jobs created an independence not found among the field hands or even the domestic slaves. In the coffee plantations in early nineteenth-century Brazil, for example, the muleteers who carried crops to market were considered a particularly lively group and were thought to be the elite of the slave force. The same could be said for the itinerant miners or even fishermen and truck garden slaves who supplied the food markets in the cities and who often worked with minimal supervision.

Knowledge was also an important granter of status within the slave community. This could be an ability to read and write the local European language, or even Arabic and a reading knowledge of the Koran, just as it could be an understanding of the dynamics of the master class and the socioeconomic realities of the free world. These types of knowledge would often be associated with skilled occupations, those possessing autonomy, or those in which contact was had on a frequent basis with the master class and other nonslave groups, such as in domestic service. It was also more commonly found in urban settings and could be discovered even at the lowest level of the occupational skills ladder. However, knowledge of African ways and customs, or even in some rare instances, of prior noble or elite status transferred directly from Africa, gave some slaves leverage in their community in contrast with their occupational status within the American economy. Thus, the leaders of the 1835 Malê rebellion in Bahia were all alufás, or Muslim religious leaders, but their status in the nonslave world varied from a slave water boy to a free skilled artisan.⁶ The same occurred with many of the male and female Africans who were part-time religious, health, and witchcraft specialists, most of whom had a status inside the community completely unrecognized by the master class. Finally, there are even cases of lower-status slave women becoming beatas, or Catholic lay religious figures, who emerged with nonslave and even white followers.

Sometimes this knowledge provided leadership and status potential, and sometimes it offered a potential for power as a cultural broker. Many domestics, for example, might not be considered elites within the community but could provide the kind of brokerage knowledge or contacts of aid to slaves more isolated from the dominant society. Thus, house servants often held a special ability to mediate demands between the slave quarters and the master's house. But this role often left them with little leadership possibilities on either side. Some slave leaders did come from domestic service, but usually they had occupations outside full-time master control.

Thus, it was no accident that urban slaves and artisan or transportation workers were usually to be found at the head of rebellions or were persons who were most likely to purchase their freedom. However, such leaders in and of themselves did not define the culture of the slave world. Afro-Brazilian culture as it emerged tended to develop in the small black villages that made up the world of the large plantations and in the common social spaces that slaves shared in cities. Thus, markets, taverns, watering spots by urban fountains, riverbanks used for washing of cloths, and even churches were sites for meeting and reinforcing relationships. Moreover, for a minority of slaves, formal religious associations were also an important component of their lives, and Catholicism for many was an accepted religion.

Whatever the effectiveness of the Church may have been in relation to protection of the slaves and their rights to celebrate the sacraments, there is no question that all slaves were baptized. Although the Church pressured the Crown to baptize all slaves in Africa or aboard the Portuguese-owned slave traders and laws were written to that effect, this was only modestly accomplished.⁷ Thus, a common occurrence in all Brazilian parishes was the baptism of adults. Of the 4,775 slave baptisms for which age and type was known, which occurred in the urban Rio de Janeiro parishes of Candelária São José and Santa Rita between 1751 and 1760, some 26 percent were for adults.⁸ In the São Paulo town of Mariana between 1700 and 1750 in the matriz parish, some 2,756 slaves were baptized, of whom 59 percent were adults.⁹ Of some 7,217 slaves baptized in four rural parishes on the frontier of the province of Rio Grande do Sul from the mid-eighteenth century to 1835, 11 percent were adult slaves.¹⁰ In the Vila Rica parish of Antonio Dias, of the 2,494 slave baptisms between 1759 and 1818, one-third (818) were for adults. Moreover, they were more than half of such baptisms in the earlier years when the African arrivals to Minas Gerais were heaviest. Thus, of the 960 slave baptisms that occurred between 1759 and 1773, some 62 percent were given to adults.¹¹ In fact, in the internal slave trade, officials defined “novo” slaves as those who did not have a Christian name and did not speak Portuguese (boçales in the traditional literature).¹² All these Africans were in fact baptized as soon as they arrived in their permanent place of residence, and no records exist in Brazil of any African first names applied to slaves – a key indication of universal baptism.

At the heart of the new black culture was the family unit. Although masters experimented with every type of communal arrangement for their plantation slaves, most slaves lived in family units. These households would define the emerging Afro-Brazilian culture and would socialize children to these beliefs and behavior. Black culture involved everything from sexual mores and kinship arrangements to language, religion, and the arts. It was a culture whose prime task was to create a coherent and reproducible community that would provide a social network of resources and support for the individual slave. Without this culture, slaves could not have functioned, and even white planters recognized its essential quality of providing social stability in an otherwise chaotic and hostile world.

Although the Church played an important role in legal marriages, especially in the southeastern region of Brazil, the majority of slaves lived in family units not formally sanctioned by the sacrament of marriage. Nevertheless, both legally sanctioned marriages and those composed of free unions played a fundamental role in slave society, culture, and identity. The best data currently available on slave family composition in terms of origin come from the legal marriage data. The sex ratio among the Africans guaranteed that initially, many males would not have access to slave women of their own background, and they would marry out more than the African women did. In the city of Rio de Janeiro in the early nineteenth century, higher legal endogamy rates were noted for African females than for African males. In three urban parishes from 1790 to 1837, African female slaves legally married African male slaves in 96 percent of the 444 marriages in which such women were legally married, whereas African male slaves married African women slaves in only 85 percent of the 501 marriages in which they were partners.¹³ In the Minas Gerais district of Barbacena between 1721–1781, some 717 Africans were married; of this number, only 68 percent married other Africans, and here, too, African males married native-born slaves far more often than African women married native-born male slaves. But it is worth noting that within all these African marriages, there was a very high degree of endogamy. In this study of Barbacena, which is one of the few such detailed and extensive studies of ethnicity currently available, it was found that among the 488 Africans who married other Africans, there was a very high rate of ethnic endogamy, with 96 percent marrying persons from the same nation, region, or language group as themselves.¹⁴ In the legal marriages of African-born slaves in Bahia and Rio de Janeiro, there were also evident patterns of endogamy about ethnic origins. Thus, for example, in 253 legal marriages of African slaves registered in Rio de Janeiro in the first half of the eighteenth century, there was an extraordinarily high rate of men and women marrying within their own ethnic group.¹⁵ Interestingly, the high endogamous rates among African groups were reversed in Bahia and Rio de Janeiro, which suggests that when any regional group was less well represented among the slaves, there was less endogamy. Consequently, the majority of Angolans were highly endogamous in Rio de Janeiro and far less so in Bahia, where they were a distinct minority among the Africans.¹⁶ In the coastal district of Angra dos Reis in Rio de Janeiro, 75 percent of the slave marriages were endogamous in terms of African origin, and only a quarter involved Creoles marrying Africans.¹⁷ In another survey of 160 marriages in the Bahian Recôncavo region in the eighteenth century, only 41 percent were shown to be endogamous marriages based on origin, but a more detailed look at origins of Africans showed that in each case, the single largest grouping of their marriages were to others of their own ethno-linguistic grouping.¹⁸

Over time, the ratio of men to women became more balanced as the native-born slaves came to dominate the resident population, and probably the rate of adult bachelors declined and free and legal unions of two slaves with children became the norm. Almost all studies suggest that the majority of slaves lived in family units. Within Brazil, there was a wide range of family arrangements. Like most of Catholic Latin America, Brazil was unique by contemporary European standards in its extraordinarily high incidence of free unions and illegitimate births among the whites and free population in general. In few European societies prior to the nineteenth century were births among free persons close to the very high illegitimacy rates commonly found in Brazil, as in all of Latin America, and in none were the levels of free union so high. Even among the elite whites, where formal marriage played such a crucial economic and political role, the rates of illegitimacy and free unions were higher than found in any corresponding European elite, including those of Portugal. Thus, for example, an 1855 census of the city of Salvador de Bahia found that, of those couples with children, some 59 percent were not legally married.¹⁹ In the parish of Sé of the city of Salvador between 1830 and 1874, only 38 percent of the more than nine thousand births to free persons were listed as legitimate children.²⁰ These urban rates of low marriages and low legitimate births seem rather special to the urban environment, especially compared to the more rural districts of the empire in the eighteenth and nineteenth centuries. Here the rates of marriage seem to have been much higher for the noncolored population of free-born persons.²¹

Given these varying patterns of legal marriage among the free population, it is rather surprising to find legally married slaves from the earliest days of slavery in Brazil. In fact, Brazil was fairly unique among slave societies in having a relatively important group of such couples. Moreover, these marriages were legally binding, and not only do married slaves appear in other documents as being married or widowed, but there are also numerous cases showing the Church and state intervening to prevent the disruption of co-residence of married couples. But slave marriage rates were not uniform across time and space. Although married slaves could be found everywhere, the zones of the Southeast showed more such legally guaranteed unions than other regions of Brazil, and some regions, like Rio de Janeiro, saw severe declines in these marriage rates during the nineteenth century. Other regions, however, seem to increase or at least to maintain their high marriage rates until the last years of the slave system. Also, somewhat unexpected, it would appear that there were more married slaves in the rural areas than in the urban centers.

The percentage of slaves who were legally married on the coffee estates of São Paulo was higher than in any other region of the country, whether plantation area or not. Whereas some 12 percent of the slaves were recorded as “ever married” in Brazil in the first census of 1872, in the coffee counties of São Paulo, some 29 percent of the adult slaves were so designated. As was to be expected given the bias toward men among the slaves, the ratio of slave women ages 15 and older who were married or widowed in São Paulo in 1829 was 39 percent compared to 23 percent of the adult slave males.²² This difference among the sexes in marriage rates was also found in the Minas Gerais district of Santa Luzia in 1831 – which was a zone of relatively low marriage rates for slaves. Here, 17 percent of the males older than age 15 were married, compared to 22 percent of the women.²³ A complete survey of the available census records for the province of São Paulo in the late eighteenth and early nineteenth centuries

TABLE 8.1. Percentage of Slaves Age 15 and Older Who Are Married or Widowed

Source: Francisco Vidal Luna, “Observações sobre Casamento de Escravos em Treze Localidades de São Paulo (1776, 1804 e 1829),” Anais do Congresso sobre História da População da América Latina, São Paulo, ABEP/SEADE (1989), tabela 3.

shows an overall rate of 28 to 30 percent of all adult slaves being legally married, a rate that could be considered the norm for this institution (see Table 8.1).

As might be expected, given the problems of maintaining married slaves in co-residence and the need for master approval, the overwhelming majority of marriages of slaves marrying other slaves were within the confines of the same slaveholding. Thus, in the Minas Gerais district of Catas Altas do Mato Dentro, in selected years between 1742 to 1834, some 98 percent of the marriages were between slaves owned by the same master, and only 2 percent involved slaves marrying nonslaves.²⁴ This was the pattern found for marriages in the coastal area of Rio de Janeiro in the nineteenth century²⁵ and for a frontier town in São Paulo as well.²⁶ It was also common for most Rio de Janeiro slave marriages in the seventeenth and eighteenth centuries as well.²⁷ All indicated that more than 90 percent of the slaves marrying other slaves were selecting their partners from within the same household. Numerous eighteenth- and nineteenth-century studies of slave plantations in Minas Gerais, Rio de Janeiro, and São Paulo also show that women on larger plantations were more likely to be living with partners than those on smaller units. Thus, to give just a few examples, in the sugar and coffee plantation region of Campinas in São Paulo in both the provincial census of 1829 and the national census of 1872, slave women living on estates with ten or more slaves were twice as likely to live in family units as were women being held on estates with fewer than ten slaves.²⁸ In the well-studied province of São Paulo, an analysis of some 200,000 slaves from provincial censuses stretching from 1775 to 1850 show that there was a consistent increase of married couples the larger the size of the slaveholding unit. This was more pronounced for women, for whom there was a systematic correlation between increases in the ratio of married women and the size of the slaveholding, but also showed for men, although there was a plateau reached for men for slaveholdings that exceeded ten slaves per owner. Thus, slaves who were owned by masters who had just one slave were legally married in only 6 percent of the cases for men and 4 percent for women. Meanwhile, in estates holding more than forty slaves, the ratio of married slaves was 21 percent for men and 39 percent for women.²⁹ Given these findings, it is not surprising to discover that in Campinas in 1872, only 29 percent of slave children under age 9 were declared legitimate on plantations holding fewer than nine slaves, compared to an 80 percent legitimacy rate for children on estates with more than ten slaves.³⁰

Nevertheless, all zones recorded marriages exogenous to the slaveholding, and many that crossed the legal boundaries of the couples. In the mineiro town of Vila Rica in the eighteenth and early nineteenth centuries, for example, some 20 percent of the slave marriages were to nonslaves – in this case, all of them forros.³¹ In the paulista frontier town of Santana de Parnaíba, there were 569 marriages of slaves in the period from 1726 to 1820, but only 70 percent were between two slaves. There were 52 marriages between slaves and Indians and 117 between slaves and free-colored persons.³² One of the few systematic studies of such mixed marriages was carried out for the Paraná district of São José dos Pinhais for the period 1759 to 1888. Of the total of 179 legal slave marriages, surprisingly, 79 of them involved a slave marrying either a free person or a forro. Some 84 percent of these mixed marriages were between slaves and free persons (color unknown) and again, a very surprising 56 percent of all of these marriages involved male slaves marrying free females. Among the few forro/slave marriages, in general both male and female slaves equally married forro partners.³³ In the Rio de Janeiro sugar zone of Campos dos Goitacases, 96 percent of the 4,507 marriages of slaves whose partner's status was known were with other slaves, but there were 60 marriages of slaves to free persons and 102 were with forros. Female slaves were far more likely to marry free men (72 percent of these mixed marriages), whereas male slaves were more likely to marry forra women (in 61 percent of these cases).³⁴ The Paraná and eighteenth-century Minas data seem to be quite unusual, and the norm here is more like that found in Campos, with fewer than 10 percent of slave marriage occurring between slaves and nonslaves, although even this is a significant number given the social status of the slaves.³⁵

The free persons marrying slaves were almost always the poorest element among the free population and were predominantly pardos, pretos, or Indians, and very rarely whites. For free women, marriage to a slave could provide economic support from the master, who in all cases had to approve the marriage, or was the only available marriage partner if the free person had been living in a free union or had sexual relations with the slave. Such was the case of Antônia de Escobar, a liberta, who in 1749 appealed to the ecclesiastical authorities saying she lost her virginity to a slave – which he admitted – and as such she needed to marry him because her case was well-known in the community and no free person would marry her.³⁶ For males it was primarily a question of poverty. Many of them already lived and worked as agregados in the same households as their slave spouses or were so poor that marrying a slave permitted them to have their wife's sustenance paid for by the master as well. It would appear that free males marrying slave women was moderately more common than free women marrying slaves. Of the 125 marriages contracted between slaves and free persons that reached the ecclesiastic courts in eighteenth-century São Paulo, 60 percent were free males marrying slave women.³⁷ It was thus most common among slave couples, even those married with free persons, for their partners to be working for the same master.³⁸

Given that master approval was required in all cases of slave marriages, it is not surprising to find that the majority of slaves were forced to select marriage partners from the same plantel (or slaveholding), largely for the convenience of their owners who demanded stability and control over their workers. There were thus, clear demographic constraints on available marriageable partners and thus marriage rates and the selection of partners were greatly influenced by the size of slaveholding and the available pool of marriageable-age persons.³⁹ This would seem to be the norm in larger slaveholdings and was obviously less common when only one or two slaves were owned by the same individual. Thus, in rural holdings, for example, the size of the marriage pool was much influenced by the size of the slaveholding. In small rural holdings and in urban settings where the average number of slaves held was much smaller than in the fazendas, the potential pool of partners obviously was much smaller. This meant that coupling and marriage of slaves would occur across slave ownerships, and even between slaves and nonslaves in the same or other households in these smaller units. Because most slaves in fact married other slaves within their own slaveholding group, this inevitably meant that the larger the slaveholding, the more likely were slaves to find a marriageable-age partner and thus most studies that analyze marriage by size of slaveholding find that marriage rates increase with the size of the slaveholding.⁴⁰ This is the case not only of legal marriage, but of free unions as well. A study of slave families in the Rio de Janeiro district of Angra dos Reis shows that the larger the unit, the more likely the families were to be dual-parent households and the less likely they were to be matrifocal families.⁴¹

Finally, there is the issue of the change of marriage patterns over time. Many studies seem to suggest that during the course of the nineteenth century, the ratio of matrifocal, single-parent households increased at the expense of dual-parent households, of both legal and free union origins. This seems to have been the pattern in Lorena in São Paulo,⁴² and was the case of the coastal zone of Angra dos Reis in Rio de Janeiro in which dual-parent households, already a minority of families in the early nineteenth century, dropped to half their rate by the second half of the century and accounted for only 17 percent of all families.⁴³ The rural zones of Rio de Janeiro also seem to have experienced this decline, with nuclear families predominating in the first half of the century and matrifocal, single-headed households dominating in the middle and later part of the century, although the trends are not consistent over time.⁴⁴ Some scholars seem to suggest that the timing of this change is related to the end of the Atlantic slave trade in 1850,⁴⁵ whereas others propose that the U.S. Civil War of the 1860s suggested to planters that the Brazilian slave system was no longer viable and this was reflected in prices for young female slaves, no longer taking into account their reproductive potential and in turn their disinterest in promoting legal marriage.⁴⁶ Whether these or other factors can be said to exist in given regions is still questionable, given the lack of longitudinal data from enough regions to see if these trends in Rio de Janeiro and São Paulo were consistent across all slave regions in Brazil. Moreover, it would be useful to see if these patterns were also found in nonslave marriages. That these may have declined as well can be seen in the data on legitimate births in the Minas district of São João del Rei. Here the rates of legitimate births (among both the free and among the slaves) declined in the course of the nineteenth century.⁴⁷

Whether married or not, it is clear from all recent studies that slaves lived in family units. Although formal marriage was only of significance in the eighteenth and early nineteenth centuries in Rio de Janeiro, it was always a major aspect of slave life in the province of São Paulo. However, these two provinces were the exception, and in most regions slaves lived in consensual unions unprotected by the law or the Church. In fact, it would seem that in many regions, even the majority of free persons lived in consensual unions unrecognized by the Church.⁴⁸ Almost all plantation records show that most slaves in the rural areas lived in family groups and the same appears true for urban slaves. Church records, often used in conjunction with local census mapas, and postmortem inventories in which all slaves are listed for a given owner, along with a few surviving plantation lists, have been used to analyze slave family structure, residence patterns, family size, and naming patterns of children across generations. Wide variation has been seen in many of these regional studies, but it is clear that the dominant pattern is dual-parent households with resident children. In a few regions, however, female-headed households or single-person residents have dominated. As Slenes has noted, many of the local studies on family structure and slave marriages have shown wide regional variation and this has made generalizations difficult until more work is done.⁴⁹

It is difficult to estimate the fertility of the slave population based on these parish records. However, these baptismal records are crucial legal documents not only for free persons and their births, marriages, and deaths, but also for slaves. These registers were the only way for owners to legally record the progeny of their slaves as their own possession. Moreover, all inheritance issues for children depended on whether they were registered as legitimate or illegitimate. What is surprising in the Brazilian figures is the relatively high level of legitimate births among slaves, even compared to free persons. Given the low levels in general of legitimate births among the free population, the slave populations do not look that different. Although slaves always had the highest such rates of illegitimate births, in many regions they did not differ too much from the white population, whose rates were often quite high.⁵⁰ Given the higher rates of marriage found in the larger slaveholdings, it is not surprising that at least in one major coffee zone, that of Juiz de Fora in Minas Gerais, there was a high correlation between legitimacy of slave births and size of plantel in 1872, and more than half of slave births were registered as legitimate – a rather high rate.⁵¹ But the dominant pattern for slaves was that the majority of their births were out of wedlock. Probably closer to imperial patterns was the rate of one-third of all slave births being legitimate, which was found in a survey of more than nine thousand slave births registered from 1736–1854 in the parish of Nossa Senhora do Pilar de São João del Rei, also in Minas Gerais. In contrast, the legitimacy rate for the births of free persons was 71 percent (in more than 13,000 births) and 40 percent for the 2,400 or so births to forras mothers.⁵² Legitimacy rates in rural parishes of Rio de Janeiro in the eighteenth century show very high rates – with some 48 percent of 11,580 slave births in this category.⁵³ This pattern of relatively high legitimate slave births does not appear in the Bahian rural parishes or even in urban Rio de Janeiro zones in the nineteenth century – where rates varied from 10 to 33 percent.⁵⁴ However, these eighteenth-century rural carioca rates are quite similar to those found in Minas Gerais in the nineteenth century. In the Minas Gerais town of Santa Luzia in the early nineteenth century, an impressive 45 percent of the 1,006 slave births whose status was recorded were listed as legitimate children, this despite a relative low rate of just 19 percent of the adult population being married⁵⁵; and in Juiz de Fora and Muriaé in 1851–1888, the rate was 47 percent for slave births in the former district and 21 percent in the later one (for an overall 37 percent rate), which compared to a legitimacy rate of 89 percent among the free non-Indian population.⁵⁶ In the paulista district of Franca, some

TABLE 8.2. Legitimate Births for Total Populations and for Slaves in Selected Districts

Source: Praxedes, “A teia e a trama da ‘fragilidade humana',” quadros 5, 6, and 7, pp. 10–12; * for Vila Rica 1763–1773, Kátia Maria Nunes Campos, “Antônio Dias de Vila Rica: Aspectos demográficos de uma paróquia colonial (1763–1773),” XVI Encontro Nacional de Estudos Populacionais, ABEP (2008), tabela 3, p. 14.

45 percent of the 2,270 slave births in the first half of the nineteenth century were legitimate.⁵⁷ These mineiro and paulista patterns of high legitimate birthrates do not seem to have held for the northeast region of the empire. In Paraíba, the slave births that were considered legitimate in some three districts in the first half of the nineteenth century ranged from only 9 to 29 percent of all slave births. It is also interesting to note that the free non-forro population had relatively significant rates of legitimate births ranging from 64 to 79 percent in selected regions.⁵⁸ As can be seen in the summary data in Table 8.2, there is little question that legitimate slave births were consistently lower than that of the total population, and that there was considerable variation of legitimacy rates even among the free and forro population.

Looking at these births in more detail from the same region (see Table 8.3), we can see that, as might be expected, the forro population

TABLE 8.3. Rate of Illegitimate Births in Sabará (1776–1782) and Raposos (1770–1806)

Source: Praxedes, “A teia e a trama da ‘fragilidade humana',” quadros 6 and 7, pp. 10–12.

of ex-slaves were in the middle in terms of having their children baptized as legitimate.

Moreover, as could also be expected, neither slaves nor forros had abandoned children (expostos) listed in a far more complete survey of births in the same community of Sabará. In the former case, no owner would allow a slave to be listed as exposto, because that would mean the child was free and he would thus have lost a slave, and in turn, the forro population could easily absorb all children into the household workforce and thus did not abandon their illegitimate children (see Table 8.4). Interestingly, if illegitimate children were listed as “natural,” it signified that their parents, although not married, had no impediments to marry, as opposed to bastard children with one or more of whose parents was married to another person other than their parent.

Given the wide variation in legal slave marriages, it is not surprising that some regions would even show variations in legitimacy rates for slave

TABLE 8.4. Legitimacy of Births by Legal Status of Parents, Sabará 1723–1757

Source: Ana Luiza de Castro Pereira, “A ilegitimidade nomeada e ocultada na vila de Nossa Senhora da Conceição do Sabará (1723–1757),” Anais do XI Seminário sobre a Economia Mineira (2004), tabela 1.

children by time. In contrast, in two different periods in the Minas mining town of Sabará, rates were much lower – just 8 percent of the 1,627 slave children born between 1723 and 1757 were declared legitimate,⁵⁹ and 22 percent for the 529 slave births were so considered in the period 1776–1882.⁶⁰ This later rate is roughly the same as reported for some 1,118 slave children born in the parish of Inhaúma in Rio de Janeiro between 1817 and 1842, only 21 percent of whom were legitimate.⁶¹

But high illegitimate birthrates did not mean instability or lack of familial relationships among slaves. Most studies of slave families have argued for the existence of trial marriages among slaves, with couples eventually settling down into a familial pattern after the first or second child. Moreover, the complex Church rules on legitimate births made it difficult to find the family connections among the slaves if they were not formally married. Yet all studies of plantation records find most slaves living in family units.⁶² Even more surprising, these studies suggest a high rate of stability of these families even after the death of the owners. It has been argued, in fact, that the larger the slaveholding, the more likely it was that families would not be broken up through sales. It is the poorer slave owners, those with the fewest slaves, who are most likely to have slaves who are married or living with a partner outside their own plantel, and also most likely to suffer a breakup of the family unit because of sales.⁶³

The majority of slaves lived in free unions that in effect were formally sanctioned and recognized family units, albeit without the powerful support of the Church. Importantly, slaves themselves recognized these units, and once such families were established, slave communities went out of their way to ensure their internal stability by using the usual mechanisms of community control. Errant spouses or nonresponsible parents were condemned by the community and were made to conform. This conformity could be enforced by normal social pressure, witchcraft, or even violence to guarantee community peace and welfare. This did not mean that the slaves kept up a Victorian-style morality, but it did mean that once a slave family was firmly established, it was given legitimacy and sanction by the community.

The dominance of organized family living can be seen, for example, in detailed studies of the coffee plantation zone of Bananal in São Paulo in the early nineteenth century, where in 1829 some 83 percent of the 2,282 slaves were found living in family units.⁶⁴ Equally important, a study of the small northeastern São Paulo town of Batatais in the period 1850–1888 showed that more than three-quarters of all slaves whose residence and family arrangements were known lived in family units, with very few children living alone.⁶⁵ A similar study of some 2,245 slaves in the São Paulo plantation towns of Lorena and Cruzeiro in 1874 showed 55 percent of the slaves living in family units.⁶⁶

Of course, the high death rates of slaves and the impact of sales and forced separations on free unions sometimes led to serial marriages. These processes also led to the emergence of stepfamilies, as well as families with legitimate and illegitimate children. However, even in high mortality regions, on average most slave marriages, legal or otherwise, lasted for a long time. One of the few studies of marriage longevity that has been conducted for Brazilian slaves comes from the sugar and coffee plantation region of Campinas in São Paulo. In 1872, for women slaves between the ages of 35 and 44, living in units of ten or more slaves, the median length of marriage was 16 years and 8 months. Moreover, it was found that there was surprising stability of residence for these married women, with almost all breakups of families coming as a result of the death of one partner rather than sales off the plantation.⁶⁷ From the same census of Campinas, it was estimated that in six of ten marriages of young slave couples (in their mid-20s or mid-30s) in 1872, the average marriage lasted eleven years.⁶⁸ The average length of marriage in Batatais was ten years.⁶⁹ It was found that even after the division of the property of a deceased slave owner, the overwhelming majority of the slave families remained intact. Between 1850 and 1888, in the same west Paulista district of Batatais, only 5 out of 112 families whose fate was known were broken up after division of the properties of the owner.⁷⁰ A study of several hundred postmortem inventories of slaveholdings in Rio de Janeiro between 1790 and 1835 also showed that only 19 percent of the slave families whose fate was known in these lists were broken up by the death of their owners.⁷¹ There are also numerous cases of elderly children living with their slave parents, suggesting quite stable families. Such was the case with two slave families owned by the richest slave owner in the Paulista community of Cotia in 1798. Padre Rafael Antônio de Barros owned eighty-four slaves who were distributed into six nuclear families, four matrifocal ones, and two organized around widows. The slave Agostinha, a widow of fifty-five years of age, lived with her three children ages 16 to 30 years, while Eugênia, a sixty-four year old widow lived with her six children aged 17 to 26 years of age.⁷²

Although there was a moderate inverse relationship between the size of the holding and the ratio of families separated through sales, overall the data from Campinas and the study of the Batatais slave families showed that the death of the master had far less of an impact on family breakups than the death of the slaves themselves, even among smaller units of slaveholdings. Of course, these studies are for prosperous plantation zones of major slave concentrations and continued slave in-migration even after the end of the Atlantic slave trade in 1850. Thus, they were zones with limited outmigration with consequent low levels of family breakup. It might be expected that studies from the Northeast or of the southern provinces, which were subjected to an important outmigration of slaves into the internal slave trade after 1850, may show more family unit breakups through sales even prior to the death of the slave owners. Yet in general, all the current research suggests a surprising longevity of slave marriages and a relatively low incidence of separation through the death of slave owners or through sales. The overwhelming form of breakup of slave families was from the death of one of the partners.

In terms of internal marital stability, there currently exist few viable studies given the lack of major fazenda records over several generations with both parents being named. In studies of slave births in the United States, it was found that slave mothers began having children quite early, and it was common for them to engage in premarital intercourse rather freely. This continued until the birth of the first child. At this point, women usually settled down into a relationship that might or might not be with the child's father. Usually, except in cases of widowhood, the father of the second child was the father of all later children.⁷³ Unfortunately, the detailed reconstruction of births needed to evaluate this question of familial stability has only been partially carried out in the case of Brazil.⁷⁴ Until better data are provided, this theme cannot be fully explored at this time.

There is, however, a recent debate that has emerged in the literature about the impact of the end of the Atlantic slave trade on resident slave marriages, legitimate slave births, and even slave fertility. Although current studies of legal marriage and legitimate births show wide regional variation and great change over time, there seems to be no secular trends evident in the data, whereas the study of slave fertility is still a relatively recent area of concern and diachronic data are still unavailable. Yet some have argued that the end of the slave trade pushed planters toward encouraging births and marriages, whereas others have taken just the opposite position.⁷⁵ To date, there seems to be little evidence to suggest that planters themselves were more or less pro-family or pro-natalist before and after 1850. Studies of positive natural growth rates in Paraná and Minas Gerais even before the end of the slave trade and the steady growth of their resident slave populations until the last decades of slavery would suggest that there was no inherent bias among slave owners against promoting slave children and slave marriage.

With parenthood and cohabitation came kinship arrangements. Slave families several generations deep sometimes lived in extended families. These extended families, whether cohabiting or not, in turn developed clear rules about acceptable marriage partners. These rules included such universal human taboos as sibling incest prohibitions and even discrimination about marriage partners from along collateral cousin lines. Slave families may also have observed internal rules regarding naming patterns, property inheritance, and even place of residence, as newlywed couples negotiated whether they would live alone or with the “bride's” or “groom's” parents. In Brazil, the study of the slave family is still in its early stages, and we have just a few studies that have recently been undertaken that indicate detailed kinship networks, naming patterns, and marriage organization. On the Resgate coffee plantation in Bananal (São Paulo), well-preserved records for 436 slaves in 1872, along with their baptismal records from earlier years, provide detailed evidence of kinship on this long-lasting plantation. They show that almost 90 percent of the slaves were kin to other slaves or free persons of color. Those not having kinship ties were primarily males and either older Africans or recently purchased Creole slaves born in other provinces. All but two of the mothers were legally married on the plantation. The predominant form of family organization was the nuclear family, with only one-quarter being extended families. Of the 116 married couples, only a third involved Africans marrying Creoles; the rest were endogamous for either group. Among the mixed unions, African men married Creole women. When Africans intermarried (twenty-nine couples), the age difference between partners was only four years; for couples in which both were Creoles (forty-six in number), the difference was seven years, and in the mixed African-Creole couples, it was fifteen years. These figures suggest that African women – always in the minority to African men – were able to marry quite quickly with men close to their age, but when African men could not obtain African women, they had to be older and of more status in order to marry Creole women. Based on the baptismal records for 568 children born on this plantation from 1860 to 1872, only 67 were named for living or dead relatives.⁷⁶ This same low usage of the names of living or dead relatives was found in a study of some 130 slave families and their children in the baptismal records of several parishes in Rio de Janeiro from 1790–1830. In this case, only 5 percent of the children were named for either parent.⁷⁷ In both of the Brazilian cases, the patrilineal naming practices that were apparently the norm in the United States did not appear. The naming of children in Brazil came from both the paternal and maternal relatives. On the Resgate plantation, as in the region of Campinas, these slave marriages were quite stable and of long duration.

If we are still lacking residence patterns of families and more detailed studies of kinship in most regions of Brazil, there have been excellent recent studies on a secondary kinship system that developed among slaves in Brazil. Known as godparenthood, it was a major fictive kin system used by all classes – including slaves. Although few marriages were legally sanctioned, all births were recorded by the Church. In such Church recordings, a fictive kinship pattern of compadrio, or godparenthood, was established. This was a formal relationship between adults that bound them through their children. The godmother (madrinha) or godfather (padrinho) was supposed to be a close friend and one to whom the child could turn as a parent if their own parents died or even if they remained alive. The co-parent was obligated to provide for that child on all special occasions and incorporate that child into their household if the other household ceased to exist. Equally, the friendship relationship among the fictive and real parents was further cemented by these ties so that special claims could be made between them for support and services.⁷⁸

Among the elite and for the Indians and black slaves, this institution was heavily based on friendship and respect, with either close personal friends or community-recognized elders and morally sanctioned persons as the most likely candidates for such a role. Thus, white planters had fellow white planters as their godparent relations, just as Indians and slaves had fellow Indians, slaves, or freedmen from their own communities. Free colored, mestiços, and other middle groups were known to have sought godparents from higher-status individuals and thus used compadrio as a means of establishing more formal patron–client relationships, an important but alternative development of the compadrio system. Slaves, too, were not beyond using compadrio as a tactic for solidifying such vertical ties, especially with the world of free colored. Thus, in the zona de mata Minas Gerais district of Senhor Bom Jesus do Rio Pardo, some 1,970 slaves were baptized between 1838–1887. Of this number, padrinhos appeared in all but 11 percent of the baptisms, and only 19 percent of the madrinhas were missing on these parish registers.⁷⁹ Surprisingly, only 31 percent of the padrinhos were fellow slaves, whereas 69 percent were free men of color. The case of godmothers was similar, with only 38 percent of them being slaves and the rest free. Despite their minority status in the slave population, Africans dominated Creoles among padrinhos, whereas Creole women slaves dominated the group of madrinhas. Moreover, even when slaves were padrinhos or madrinhas at slave births, only two-thirds of them came from the same slaveholding.⁸⁰ In the Curitiba parish of Nossa Senhora da Luz dos Pinhais in the nineteenth century in the Bahian sugar zone of Iguape in 1835, and in selected Bahian parishes of the eighteenth century, similarly high rates of free godparents assisted at the births of slaves, with very few of these being their masters.⁸¹ This remarkable figure of having close to 60 to 70 percent of all godparents for slave children being free persons, the overwhelming majority of whom were not their masters, was also encountered in the Bahian sugar parishes of Monte e Rio Fundo in the period from 1780–1789, and in the small scale slaveholdings in Cabo Frio in Rio de Janeiro in the eighteenth century.⁸² However, this pattern was not consistent across all of Brazil's districts. In Rio de Janeiro, for example, in the parish of Inhaúma, between 1817 and 1842, only one-third of the 1,557 godfathers were free persons, whereas 64 percent of the slave godparents came from the same household or slaveholding as the child being baptized. Moreover, in contrast to other regions, the majority of the free persons serving as godfathers (almost three-quarters of them) were ex-slaves who had been freed within their own lifetimes (that is, they were forros). In the case of the godmothers, the ratio of slaves was even higher, with 78 percent of the 656 madrinhas being slaves and 72 percent of these slave godmothers belonging to the same master as the baptized child.⁸³

Whatever the variation from parish to parish, which at this point is difficult to explain, it should be stressed that few owners were godparents to their slaves, and that in the overwhelming majority of cases even the free persons were of color, were poor, and as the case of Inhaúma suggests, they were themselves recently freed persons who may have known the child's parents when they themselves were slaves. Brazil also shows slaves acting as godparents to children born free. In the aforementioned Minas parish of Senhor Bom Jesus do Rio Pardo, roughly 35 percent of the 979 births of free children had a slave godfather, and among the 904 births of free children who had a godmother, slave women stood up for these children more than 43 percent of the time.⁸⁴ Clearly there was not a rigid barrier between poor free persons of color and slaves, and friendship routinely crossed this boundary.

Despite the importance of compadrio as a structuring element in the extended family networks of Brazil, the poorest elements in the society from which godparents were drawn could not always fully honor such obligations, and sometimes slaves were baptized with only a padrinho present. Although this was against Church practice and custom, it did reflect the weaker ties of the institution of compadrio at the lowest level of the society. Nevertheless, all accounts seem to indicate that it was an effective support system that became an essential part of Afro-Brazilian culture, just as it was of free society. This fictive kinship system went to further the growing bonds of friendship and community among slaves, and given the status with which godparentage was held by the governing elite, it even provided white legitimacy to slave community-building efforts.

Equally important for the development of a community was the creation of a coherent belief system that would provide slaves with a sense of self, of community, and of their place in the larger cosmological order. The growth of a belief system would be a hard and slow task. One of the first areas where this evolved beyond the family level was in those practices that bound the community together. As in any peasant village, there were inevitable interpersonal conflicts among the slaves over resources. Sometimes these involved garden lands, personal effects, conflicts over potential spouses, sexual fidelity, or just personality clashes. These, plus the common problems of curing and divination, all led to the emergence of part-time specialists in witchcraft and curing. Given the importance these crafts had within Africa, it was inevitable that African influences would influence their evolution in the Western Hemisphere. It was usually older and single African males and females who provided the white or black magic that was an indispensable part of any community structure. Such individuals prepared herbs for curing and for influencing desired emotional or physical states in given subjects. They also provided recourse to a system of rough justice that guaranteed a limit to the amount of personal violence that the community could afford in fights over resources. Aggrieved adults who could not directly confront their opponents often had recourse to witchcraft to harm their rivals. This use of witchcraft and the knowledge that it was effective kept conflicts within acceptable limits within communities that had little policing powers of their own or any type of communal self-government.

Although the spiritual content embedded within the practices of these black New World ritual specialists integrated multiple non-African elements, slave beliefs began to evolve into ever-more elaborate cosmologies, and complete African-influenced religions began to develop by the late eighteenth and early nineteenth centuries. Masters were opposed to such formalized religious belief systems, which they held to be antithetical to their own Christian beliefs. All such formal cults were ruthlessly attacked, just as the less-threatening simpler forms of witchcraft were left unmolested. However, so powerful did these religious systems become that they were able to survive under the guise of alternative forms of the folk Catholicism developed under slavery, and occasionally, even influenced the practices of the dominant white society, whose members sometimes turned to the practitioners of these religious systems for guidance and assistance in their affairs. Although it often took several generations after abolition for Christian society to accept their legitimacy, the cults were finally able to establish themselves as independent religions in the twentieth century.

The most important of these cults in the era of slavery in Brazil was Candomblé (sometimes called calundus in the colonial period). It appeared in various guises throughout Brazil. A small initial group often established the basic cults that later massive migrations from entirely different areas in Africa adopted in their new environments. Even where many national candomblés existed – as in Bahia, for example – it was the Nago (Yoruba) candomblé which provided the basis for the theology, ritual, and festival activity of all other candomblés, even those named for Dahomean, Angolan, and Congolese tribes or nations. Thus, a process of acculturation went on among the slaves themselves, even in terms of the proscribed African cults and practices.

This process of syncretization and acculturation among the African religions in turn helps to explain why these cults found it relatively easy to accept and integrate parts of Christian religious belief and practice into the local cult activity. Initially, this integration was purely functional, providing a cover of legitimacy for religions that were severely proscribed by white masters. However, after a few generations, a real syncretism became part of the duality of belief of the slaves themselves, who soon found it possible to accommodate both religious systems. In the Catholic societies, dogma of the elite Church was not affected, but a rich tradition of folk Catholicism with its saints and local cults provided a perfect medium for syncretization of African deities. Also, the elaborate structure of lay religious societies and local community saint days was extended to the slaves and free colored by the white authorities in their desire to integrate and control slave beliefs. They also hoped these associations, many of which in the early days were based on African tribal origins, would guarantee internal divisions among the slaves and prevent the development of a coherent racial or class identity. Although moderately successful in this aim, these associations and local festival activity proved of vital importance in both legitimating and spreading African religious practices and giving blacks and mulattoes important communal organizations.

The Catholic Church was already well organized for a syncretic approach to religious conquest and conversion even before the full-scale development of American slavery. The Latin American Church had worked out most of the norms of this activity in its evangelizing of the American Indians. Local gods were to be destroyed, but sacred places were to be incorporated into the Christian cosmology through the erection of churches and shrines and the miraculous appearance of the Virgin. A brown-skinned Virgin appeared in all the traditional pre-Columbian religious centers, and her devotion took on many aspects of pre-Columbian rites and beliefs. Although the intellectual and upper-class Catholics fought the reduction of their monotheistic religion into a pantheon of virgins and saints who took on the role of local deities, they never succeeded in cleansing the Church of its folk aspects, either in Europe or America. Moreover, the Church in early colonial Latin America was unusually open to the rise of popular nonclerical religious figures, the so-called beatas, or lay religious individuals who developed local followings and even founded religious institutions, many of whom came from the lower classes and castes. One of the most extraordinary of such beatas was the mystic slave Rosa Maria Egipcíaca da Vera Cruz. Born in West Africa, she arrived in Brazil from the Gold Coast at age 6 in 1725, and was eventually forced into prostitution by her owner in Rio de Janeiro. Literate, a significant writer, and a mystic, she developed such an extraordinarily following and support, including that of the provincial of the Franciscan Order in Rio de Janeiro, that she and her final owner and confessor Padre Lopes were eventually taken prisoner by the Inquisition in 1762 and shipped to Lisbon after the Church attacked her teachings.⁸⁵ Although beatas had been openly tolerated in the sixteenth and seventeenth centuries, the Church became more resistant and hostile to their activities over time, and Rosa, like others in the eighteenth century, were more typically incarcerated and condemned if they developed important followings.

Into this system of syncretic absorption were implanted African belief systems. Very quickly, each of the major African deities took on an alternative saint identification. In Brazil, for instance, local Brazilian saints had a dual identity in the minds of the slaves, if not in those of the whites. Church leaders in the colony encouraged local slaves to stress the cult of Our Lady of the Rosary (Nossa Senhora do Rosário), which was reserved exclusively for the special devotion of blacks. Though all slaves were taught to accept the feasts, holidays, and saints of whites, they were also expected to celebrate their own saint days and holidays on an exclusive basis. In the urban centers this meant that slaves were to be grouped into religious brotherhoods (known as Irmandades) whose major purpose was to act as a mutual-aid society and prepare an annual celebration of the black-related religious figures. There were also special welfare societies (Santa Casa), which sometimes had black and mulato branches. There were even well-known dance groups or batuques in some of the northeast cities of Brazil that were grouped along African nationality lines.

How important these irmandades were in the lives of urban and rural slaves is difficult to estimate. There is no question that slaves participated in these irmandades everywhere. In fact, in the Rosario brotherhoods, they predominated as the major group of new or continuing members, well exceeding both the forros and livres, and sometimes making up more than half the membership (see Table 8.5).

In the brotherhood membership lists so far available that include both legal status and the sex of members, there is wide variation. Women represented two-thirds of the total slaves in Cachoeira Bahia, 51 percent in Vila Rica, and 42 percent of the slave members in São João del Rei.⁸⁶ However, they were only 23 percent of the slave members in the Rosario brotherhood of Cuiabá.⁸⁷ In a study of burials for the Antonio Dias parish of Vila Rica, of the 524 slaves buried as members of the irmandades, more than three-quarters were women.⁸⁸ Given these disparities and the random quality of overall female membership ratios in other irmandades,⁸⁹ it is difficult at this point to determine if the dominance of women among members was the norm throughout Brazil.

Although it could be expected that slaves would play a lesser role in the administration of these brotherhoods, what is impressive is that they did participate as active leaders. Thus, the executive board (mesa

TABLE 8.5. Membership in Rosário Brotherhoods, Eighteenth and Nineteenth Centuries, by Legal Status

Notes: * Livres not listed separately, but probably included in forros.

Sources: Cachoeira do Campo to Barbacena in Elizabeth W. Kiddy, Blacks of the Rosary: Memory and History in Minas Gerais, Brazil (University Park: Pennsylvania State University Press, 2005), tables 6–9, pp. 256–7; São João del Rei to Vila Rica, Borges, Escravos e libertos nas irmandades do Rosário, tabelas 6, 7, and 8, pp. 230–1; Cuiabá, Christiane dos Santos Silva, “Irmãos de fé, Iramos no poder: A irmandade de Nossa Senhora do Rosário dos Pretos na Vila Real do Senhor Bom Jesus do Cuiabá (1751–1819)” (Cuiaba: diss. de mestrado, UFMG, 2001), p. 115; Bahia, Reginaldo, “Os rosários dos Angolas,” p. 202.

diretora) of the Mariana Rosário brotherhood had 303 slaves serving as officers out of 984 who served in the administration in the period 1748–1819.⁹⁰ Of course, they were underrepresented, being 71 percent of all members enrolled in this period, and only 31 percent of the board members (mesários), but not only were they an important minority, but the officials were overwhelmingly African born, whether free or slave. Of all the forros and escravos who served as officers of this association, those of African birth represented an impressive majority of these officials – of 723 whose origin was known, some 88 percent of them were born in Africa.⁹¹

Clearly then, even if only a minority of slaves in any district belonged to an irmandades, such organizations did play a fundamental role everywhere. There was not a region that contained slaves that did not contain a Rosário brotherhood, and these existed throughout the slave period of Brazilian history. Burials and festivities were their prime activity, but they also supported African identity elements as much as possible in this hostile environment. Many of the brotherhoods were predominantly of one ethnic origin, status, or color and had exclusive membership.⁹² Others denied access to Brazilian-born colored.⁹³ Therefore, in many ways they helped perpetuate conflict among slave groups and often among Africans and Creoles and slaves and free persons of color. However, with their increasing admission of crioulos over time and the fact that all of them contained increasing numbers of forros and livres, these data suggest that these institutions could also serve as integrative associations as well and help acculturate the Africans to their new environment.⁹⁴ Also, they were one of the few institutions legally available to them. Although many irmandades were based on occupations, most of these, especially of the mechanical trades, sometimes excluded slaves from membership.⁹⁵ Thus, the irmandades based on identity or legal status proved to be of fundamental importance to the slaves if not to the free colored, who had more options of association opened to them.

In Brazil, these legal and state and church-recognized associations were famous from the colonial period until the end of the Empire for their annual festive activities and, equally, for their constant conflict with the white authorities. Slave and free-colored demands for brotherhood self-government and control over their own churches and cemeteries were constantly opposed by the fraternal organizations of the whites. They were often in tense relations with ecclesiastical authorities over funding, because the irmandades paid for masses for the dead and funeral expenses of its members as well as the costs for religious festivals, and were often chastised for sacrificing the former obligation (with its funds supporting the Church) for the latter activity.⁹⁶ Yet despite white fears of their autonomy, in the majority of cases, the black and mulato brotherhoods were accepting of the dominant culture and were primarily integrative in nature. They did foster both self-pride and also legitimated African religious activity. In contrast, the African cults were forced to create independent organizations to survive. In so doing, they became essentially rejective and opposed to the values of the master class. These cults which competed with, reinterpreted Christianity for a slave audience, and most aided the development of an autonomous aspect to Afro-Brazilian culture.

From the plantation villages and the colonial cities came a distinctive Afro-American culture that provided the slaves with a self-identity and community, which allowed them to survive the rigors of their forced integration into the white society. This Afro-American culture was not homogeneous. Some of its elements were integrative and merely expressive of a subculture within the Western norms established by the white society. Others, however, were unique to blacks and provided an alternative value system to that of white society. Such a pattern was almost inevitable given the very hostility and ambiguity that the white culture expressed toward them. On the one hand, white society incorporated Africans into Christianity as co-equal members of a universalistic church. Among the Latin American legal codes, there was also a basic assumption that Africans would eventually become freedmen in these same slave societies. But at the same time, these were inevitably racist societies that rejected black self-identity and self-worth and often created a second-class citizenship for those who achieved their freedom. Social ascension and mobility were possible for enough blacks to give a majority a sense of hope, but the terms were always rejection of their Afro-American cultural identity and their blackness. In such a situation, it was inevitable that the cultures that were established by the slaves in America would serve two often-conflicting purposes: that of integrating the slaves into the larger master-dominated societies while providing them with an identity and meaning that protected them from that same society's oppression and hostility.

¹ This idea of interethnic linguistic and cultural communities developing in the Americas has been proposed by John Thornton, Africa and Africans in the Making of the Atlantic World, 1400–1800 (2nd ed.; Cambridge: Cambridge University Press, 1998). There is also no question of long and intimate contact between Brazil and Africa, even in both directions. There is even the case of João de Oliveira, an African-born Brazilian slave who returned to his home in the Bight of Benin in 1733 and then sent back money to Brazil to pay for his freedom. He subsequently became a major slave trader on the Slave Coast. Robin Law and Kristin Mann, “West Africa in the Atlantic Community: The Case of the Slave Coast,” The William & Mary Quarterly, 56, no. 2 (April 1999), p. 317.

² Only Africans who were free and who held some property actually made wills, and many of them were literate. They thus form a minority of such Africans who arrived in America.

³ Among the leading positions taken on this are those by Robert W. Slenes, Na senzala uma flor: Esperanças e as recordações na formação da família escrava: Brasil, Sudeste, século XIX (Rio de Janeiro: Editora Nova Fronteira, 1999), who has stressed the unity and cooperation; and Manolo Florentino and José Roberto Góes, A paz das senzalas: Famílias escravas e tráfico atlântico, Rio de Janeiro, c. 1790 – c. 1850 (Rio de Janeiro: Civilização Brasileira, 1977); and Hebe Maria Mattos, A Cor Inexistente: Os significados da Liberdade no Sudeste Escravista (2nd ed.; Rio de Janeiro: Nova Fronteira, 1998), who have argued for a constant conflict among Africans arriving in Brazil and basic conflict over resources between Creoles and Africans. A thoughtful reflection on these issues and their complexity in the Brazilian context is found in Sheila de Castro Faria, “Identidade e comunidade escrava: Um ensaio,” Tempo, 11, no. 22 (2007), p. 122–46.

⁴ See, for example, Seth Leacock and Ruth Leacock, Spirits of the Deep: A Study of an Afro-Brazilian Cult (Garden City, NY: American Museum of Natural History, 1972). For a useful review of the debate among Anglo-Saxon scholars of “creolization” versus “afrocentrism” and the possible applications of some of these ideas to the Brazilian context, see Luis Nicolau Parés, “O processo de crioulização no recôncavo baiano (1750–1800),” Afro-Ásia, 33 (2005), pp. 87–8.

⁵ The classic survey of these religions is Roger Bastide, Les religions africaines au Brésil (Paris: Presses Universitaires de France, 1960), which has been translated into Portuguese and English.

⁶ João José Reis, Rebelião escrava no Brasil, a história do levante dos malês (1835) (São Paulo: Editora Brasiliense, 1986), pp. 115, 156ff.

⁷ A.J.R. Russell-Wood, The Black Man in Slavery and Freedom in Colonial Brazil (London: Macmillan Press, 1982), pp. 130–1.

⁸ Mariza Soares, Devotos da cor: Identidade etnica, religiosidade e escravidão no Rio de Janeiro, século XVIII (Rio de Janeiro: Civilização Brasileira, 2000), tabelas 19–21, pp. 302–3.

⁹ Moacir Rodrigo de Castro Maia, “Por uma nova abordagem da solidariedade entre escravos africanos recém-chegados a América (Minas Gerais, século XVIII),” III Encontro de Escravidao e Liberdade no Brasil Meridional (2007), p. 4.

¹⁰ Silmei de Sant’Ana Petizl, “Considerações sobre a família escrava da fronteira Oeste do Rio Grande de São Pedro (1750–1835),” III Encontro de Escravidao e Liberdade no Brasil Meridional (2007), tabela 4.

¹¹ Iraci del Nero da Costa, Vila Rica: População (1719–1826) (São Paulo: IPE-USP, 1979), tabelas II.2 and III.2, pp. 220, 226.

¹² Fábio W. A. Pinheiro, “Tráfico atlântico de escravos na formação dos plantéis mineiros, Zona da Mata, c. 1809 – c. 1830” (Dissertação de mestrado, UFRJ, Rio de Janeiro, 2007), p. 73.

¹³ Based on the analysis of appendix tables 1, 1.1, and 1.2 in Janaina Christina Perrayon Lopes, “Casamentos de escravos nas freguesias da Candelária, São Francisco Xavier e Jacarepaguá: Uma contribuição aos padrões de sociabilidade matrimonial no Rio de Janeiro (c. 1800 – c. 1850)” (diss. mestrado, UFRJ, 2006), pp. 85–7.

¹⁴ Ana Paula dos Santos Rangel, “A escolha do cônjuge: O casamento escravo no termo de Barbacena (1781–1821),” Revista Eletrônica de História do Brasil, 8, nos. 1–2 (Jan.–Dez. 2006), tabelas 7.

¹⁵ Soares, Devotos da cor, tabela 8, p. 125.

¹⁶ Soares, Devotos da cor, pp. 123–4. The significance of this finding is also discussed in Faria, “Identidade e comunidade escrava,” p. 143.

¹⁷ Marcia Cristina Roma de Vasconcellos, “Famílias escravas em Angra dos Reis, 1801–1888” (São Paulo: tese de doutorado, USP-FFCH, 2006), tabela 40, p. 127.

¹⁸ Parés, “O processo de crioulização no recôncavo baiano (1750–1800),” tabela 3, p. 114.

¹⁹ Katia de M. Queirós Mattoso, Família e sociedade na Bahia do século XIX (São Paulo: Corrupio, 1988), quadro IX, p. 82.

²⁰ Mattoso, Família e sociedade na Bahia, quadro XIII, p. 90. The rate of legitimate slave births in this parish is lower than any other recorded rates, being only 19 out of 3,747 births. A second study of births whose parents’ origin was known in this parish from 1801 to 1888 found that 19 percent of the liberto children were legitimate, and only 33 percent of livre children were in this category. Isabel Cristina Ferreira dos Reis, “A família negra no tempo da escravidão: Bahia, 1850–1888” (tese de doutorado, História UNICAMP, 2007), quadro 1 (tabela 12), p. 105.

²¹ For comparative urban and rural parish data on legitimacy of births among the free population in the eighteenth and nineteenth centuries – which shows consistently higher rates of legitimate births among rural free persons – see Sheila de Castro Faria, “Legitimidade, estratégias familiares e condição feminina no Brasil escravista,” Anais do VIII Econtro de Estudos Populacionais, ABEP (1992), vol. I, quadro 6, p. 312. For a listing broken down by the color of the children, it is evident that whites (brancos) had a rate of legitimacy in a series of selected southeastern parishes of 90 percent compared to slightly lower rates for pardos livres (in the 80–90 percent) and forros at normally half of births. Sheila de Castro Faria, A Colônia em movimento: Fortuna e familía no cotidiano colonial (Rio de Janeiro: Nova Fronteira, 1988), quadro II.14, p. 157.

²² Francisco Vidal Luna and Herbert S. Klein, Slavery and the Economy of São Paulo, 1750–1850 (Stanford: Stanford University Press, 2003), table 6.2, p. 142; and recent postpublication calculations from our data sets.

²³ Carolina Perpétuo Corrêa, “Aspectos da demografia e vida familiar dos escravos de Santa Luzia, Minas Gerais, 1818–1833,” XIV Encontro Nacional de Estudos Populacionais, ABEP (2004), tabela 8, p. 18.

²⁴ Tarcísio Rodrigues Botelho, “Família escrava em Catas Altas do Mato Dentro (MG) no século XVIII,” Anais da V Jornada Setecentista (2003), tabelas 4 and 5, p. 16.

²⁵ Vasconcellos, “Famílias escravas em Angra dos Reis, 1801–1888,” p. 116.

²⁶ Alida C. Metcalf, Family and Frontier in Colonial Brazil. Santana de Parnaíba, 1580–1822 (Berkeley: University of California Press, 1992), p. 165.

²⁷ Faria, A Colônia em movimento, quadro V.9, p. 316.

²⁸ Slenes, Na senzala, uma flor, tabela 6, p. 104.

²⁹ Luna and Klein, Slavery and the Economy of São Paulo, 1750–1850, table 6.3, p. 148.

³⁰ Slenes, Na senzala, uma flor, table 5, p. 102.

³¹ Of some 188 slave marriages whose partner's civil status was known in the period of 1727 to 1826, some 38 involved slaves marrying forros. Costa, Vila Rica, tabela 1, p. 35.

³² Metcalf, Family and Frontier in Colonial Brazil, table 19, p. 167.

³³ Cacilda Machado, “Casamento & Desigualdade Jurídica: Primeiras notas de um estudo em uma área da região paulista no período colonial,” XIII Encontro da Associação Brasileira de Estudos Populacionais, ABEP (2002), tabela 5, p. 10. There is even a well-known case of one of the leaders of the Malé rebellion of Bahia, the liberto storeowner and Muslim religious leader Elesbão do Carmo (Dandará) of Haussa origin who was married to a slave woman whom he did not own. Reis, Rebelião escrava no Brasil, p. 163.

³⁴ Faria, A Colônia em movimento, quadro II.9, II.10, pp. 143–4.

³⁵ Such marriages were usually 10 percent or less of all slave marriages. See, for example, Silvia Maria Jardim Brügger, Minas patriarcal, família e sociedade São João del Rei – séculos XVIII e XIX (São Paulo: Annablume, 2007), tabela 4.1, p. 224 (which shows a higher ratio of male slaves marrying across legal lines than female slaves); and Janaina Christina Perrayon Lopes, “Casamentos de escravos e forros nas freguesias da Candelária, São Francisco Xavier e Jacarepaguá: Uma contribuição aos padrões de sociabilidade marital no Rio de Janeiro (c. 1800 – c. 1850),” Anais do 1 Colóquio de LAHES (2005), tabelas 1–4.

³⁶ “Ficava…perdida e não haverá quem com ela queria casar por ter tido muita fama naquela vila.…” Elena María Réa Goldschmidt, Casamentos mistos de escravos em São Paulo Colonial (São Paulo: Annablume, 2004), p. 115.

³⁷ Elena María Réa Goldschmidt, “Casamentos mistos de escravos em São Paulo Colonial,” (dissertação de mestrado, FFLCH-USP, 1986), tabela 6, p. 248.

³⁸ A major survey of Rio de Janeiro parishes in the eighteenth century of slave marriages found that more than 97 percent of them were between slaves in the same plantel. See Faria, “Legitimidade, estratégias familiares e condição feminina no Brasil escravista,” p. 300, quadro 2; also Faria, A Colônia em movimento, quadro V.9, p. 316.

³⁹ See, for example, Machado, “Casamento & Desigualdade Jurídica.”

⁴⁰ See, for example, Renato Leite Marcondes and José Flávio Motta, “A família escrava em Lorena e Cruzeiro (1874),” Anais do XII Encontro Nacional de Estudos Populacionais. ABEP (2000), tabela 6, pp. 10–11; and Luna, “Observações sobre Casamento de Escravos em Treze Localidades de São Paulo,” tabela 13, p. 14. This is the conclusion as well of Robert Slenes after reviewing numerous studies; see Slenes, Na senzala, uma flor, pp. 78–9.

⁴¹ Vasconcellos, “Famílias escravas em Angra dos Reis, 1801–1888,” tabela 33, p. 109.

⁴² Marcondes and Motta, “A família escrava em Lorena e Cruzeiro (1874),” tabela 7, p. 13.

⁴³ Vasconcellos, “Famílias escravas em Angra dos Reis, 1801–1888,” p. 107.

⁴⁴ Florentino and Góes, A paz das senzalas, pp. 141–5.

⁴⁵ Florentino and Góes, A paz das senzalas, pp. 141–5. This is also the position taken by Sheila Faria, who also stresses the impact of the first failed prohibition of the slave trade in 1831. See Faria, A Colônia em movimento, p. 339.

⁴⁶ Slenes, Na senzala, uma flor, pp. 90–3.

⁴⁷ Brügger, Minas patriarcal, família e sociedade São João del Rei, tabelas 2.24 and 2.25, pp. 109, 116.

⁴⁸ In the census of the city of Bahia in 1855, some 52 percent of the families registered lived in consensual unions. Mattoso, Família e sociedade na Bahia, p. 82.

⁴⁹ Slenes, Na senzala, uma flor, p. 45. Along with the major study of Slenes on Campinas slave families, also see the work on Rio de Janeiro families by Florentino and Góes, A paz das senzalas. Other important original research in this new area of slave families and family structures include José Flávio Motta, Corpos escravos e vontades livres: Posse de cativos e família escrava em Bananal, 1801–1829 (São Paulo: Annablume, 1999); Elena María Réa Goldschmidt, Casamentos Mistos – Liberdade e Escravidão em São Paulo Colonial (São Paulo: Annablume, 1999); Heloisa Maria Teixeira, “Reprodução e famílias escravas, em Mariana, 1850–1888” (Tese de doutorado, USP-FFLCH, 2001); and the previously cited work of Rômulo Andrade, “Casamento entre escravos na região cafeeira de Minas Gerais,” loc. cit. The age of marriage for slaves is examined in Alida C. Metcalf, “Searching for the Slave Family in Colonial Brazil: A Reconstruction from São Paulo,” Journal of Family History, 16, no. 3 (1991). A recent attempt to look at slave marriages in terms of origin between native and African born is found in Parés, “O processo de crioulização no recôncavo baiano (1750–1800),” pp. 87–132.

⁵⁰ See for example, Rafael Ribas Galvão, “Bastardia e legitimidade na Curitiba dos Séculos XVIII e XIX,” XIII Encontro da Associação Brasileira de Estudos Populacionais, ABEP (2002); Vanda Lúcia Praxedes, “A teia e a trama da ‘fragilidade humana’: Os filhos ilegítimos em Minas Gerais (1770–1840),” Anais do XI Seminário sobre a Economia Mineira (2004). A good overall review of illegitamacy rates for free persons is found in Maria Adenir Peraro, “O princípio da fronteira e a fronteira de princípios: Filhos ilegítimos em Cuiabá no séc. XIX,” Revista Brasileira de História, 19, no. 38 (1999), pp. 55–80; Donald Ramos, “Single and Married Women in Vila Rica, Brazil, 1754–1838,” Journal of Family History, 16, no. 3 (1991). Worth looking at is the attempt to measure multiple illegitimate births for the same mothers in Maria Adenir Peraro, “Mulheres de Jesus no universo dos ilegítimos,” Diálogos. Revista do Departamento de História da Uem, Maringá, 4, no. 4 (2000), pp. 51–75; and the study of parents who mentioned illegitimate children in their wills, Sonia Troitiño, “Números da bastardia: Os ilegítimos nos testamentos paulistas oitocentistas,” Anais do XII Encontro Nacional de Estudos Populacionais ABEP (2000). A basic set of indices on the structure of families and age of married partners in this period is given in Donald Ramos, “City and Country: The Family in Minas Gerais, 1804–1838,” Journal of Family History, 3, no. 4 (Winter 1978).

⁵¹ Rômulo Andrade, “Ampliando estudos sobre famílias escravos no século XIX: Crianças cativas em Minas Gerais: Legitimidade, alforria e estabilidade familiar,” Revista Universidade Rural, Série Ciências Humanas, 24, nos. 1–2 (2002), p. 103, table 2. The rates of legitimate slave births for Campinas in São Paulo in 1872 were even higher and also correlated with the size of holdings, whereas the zone of Muriaé had far lower rates and they were not correlated with the size of holdings. Ibid., p. 104, table 3.

⁵² Silvia Maria Jardim Brügger, “Legitimidade, casamento e relações ditas ilícitas em São João del Rei (1730–1850),” IX Seminário sobre a Economia Mineira (2000), p. 45, tabela 3. This was not that different from nineteenth-century rates found in large samples of births. Thus, some 45 percent of the 4,760 slave births registered in the São Paulo district of Franca in the nineteenth century (1805–1888) were legitimate births. Maísa Faleiros da Cunha, “A legitimidade entre os cativos da Paróquia Nossa Senhora da Conceição de Franca – Século XIX,” XIV Encontro Nacional de Estudos Populacionais, ABEP (2004), p. 7, tabela 4.

⁵³ Calculated from data given in Faria, A Colônia em movimento, quadro V.14, p. 325.

⁵⁴ Faria, A Colônia em movimento, quadro V.13, p. 324.

⁵⁵ Corrêa, “Aspectos da demografia e vida familiar dos escravos de Santa Luzia,” tabela 9, p. 19.

⁵⁶ Andrade, “Casamento entre escravos na região cafeeira de Minas Gerais,” tabela 1, p. 102.

⁵⁷ Cunha, “A legitimidade entre os cativos da Paróquia Nossa Senhora da Conceição de Franca,” tabela 4, p. 7.

⁵⁸ Solange Pereira da Rocha, “Gente negra na Paraíba oitocentista: População, família e parentesco espiritual” (Recife: tese de doutorado, UFP, 2007), tabela 3.2, p. 205. The districts were Nossa Senhora das Neves (9 percent of 458 slave baptisms); Livramento (28 percent of 303 baptisms), and Santa Rita (29 percent legitimate of 294 baptisms).

⁵⁹ Pereira, “A ilegitimidade nomeada e ocultada na vila de Nossa Senhora da Conceição do Sabará,” tabela 1, p. 14.

⁶⁰ Praxedes, “A teia e a trama da ‘fragilidade humana’,” tabela 6, p. 11.

⁶¹ José Roberto Góes, O cativeiro imperfeito: Um estudo sobre a escravidão no Rio de Janeiro da primeira metade do século XIX (Vitória: Lineart, 1993), tabela V, p. 59.

⁶² In a study from São Paulo in the nineteenth century based on postmortem inventories, only 39 percent of the slaves whose familial relations could be identified were listed without a family tie. See Juliana Garavazo, “Relações familiares e estabilidade da família escrava: Batatais (1850–88),” Anais de XIV Encontro Nacional de Estudos Populacionais, ABEP (2004), tabela 1, p. 6.

⁶³ In a large sample of rural slaveholdings in Rio de Janeiro in the period 1790–1835, it was found that three out of four slave families remained united after the division of inheritance of the original slave owners. Florentino and Góes, A paz das senzalas, chapter 6. In her study of the São Paulo district of Batatais in the second half of the nineteenth century, Garavazo found that 86 percent of the slave families were together fifteen years or more. See Garavazo, “Relações familiares e estabilidade da família escrava,” p. 13. Of 288 families that existed on the death of their owners in the paulista district of Mariana between 1850–1888 and whose fate was known on the division of the properties, some 64 percent remained together, one-third were either totally or partially broken up among the inheritors, and 4 percent were freed. Teixeira, “Reprodução e famílias escravas, em Mariana, 1850–1888” (tese de doutorado, USP-FFLCH, 2001), tabela 50, p. 133 – based on my own calculations, which eliminated the unknowns.

⁶⁴ Motta, Corpos escravos e vontades livres, tabela 3.15, p. 137.

⁶⁵ Garavazo, “Relações familiares e estabilidade da família escrava,” tabela 2, p. 7.

⁶⁶ Marcondes and Motta, “Família escrava em Lorena e Cruzeiro (1874),” tabela 8, p. 14.

⁶⁷ Slenes, Na senzala, uma flor, tabela 3, p. 98.

⁶⁸ Slenes, Na senzala, uma flor, pp. 98–99

⁶⁹ Garavazo, “Relações familiares e estabilidade da família escrava,” p. 13.

⁷⁰ Garavazo, “Relações familiares e estabilidade da família escrava,” tabela 9, p. 14.

⁷¹ Florentino and Góes, A paz das senzalas, apêndice 6, p. 237. There were a total of 128 families, of which the fate of only 12 families was unknown.

⁷² Fabiana Schleumer, “Além de açoites e correntes: Cativos e libertos em Cotia colonial (1790–1810)” (dissertação de mestrado, FFCH-USP, 1999), p. 157. Of the four nuclear families, two involved a union between slave males and free women – one being a livre and the other a liberta.

⁷³ Herbert G. Gutman, The Black Family in Slavery and Freedom, 1750–1925 (New York: Vintage Books, 1976).

⁷⁴ One of the few such studies, but one still lacking crucial elements needed to answer questions of spacing and parentage, was carried out by Vasconcellos, “Famílias escravas em Angra dos Reis, 1801–1888,” chapter 6.

⁷⁵ For the former position, see Manolo Florentino and Cacilda Machado, “Sobre a família escrava em plantéis ausentes do mercado de cativos: Três estudos de casos (século 19),” XI Encontro Nacional de Estudos Populacionais da ABEP (1998), especially p. 1381. In contrast, Slenes takes the opposite view, suggesting far less interest in slave fecundity and slave marriage in all the plantation areas of the Southeast after 1850. Slenes, Na senzala, uma flor, p. 92. For an argument showing strong marriage ties from the earliest times, see Tarcísio Rodrigues Botelho, “Família e escravidão em uma perspectiva demográfica: Minas Gerais (Brasil), século XVIII,” in Douglas Cole Libby and Júnia Ferreira Furtado, eds., Trabalho livre, trabalho escravo, Brasil e Europa, séculos XVIII e XIX (São Paulo: Annablume, 2006), pp. 195–222.

⁷⁶ Manolo Florentino and José Roberto Góes, “Parentesco e família entre os escravos de Vallin,” in Hebe Maria Mattos de Castro and Eduardo Schnoor, eds., Resgate: Uma janela para o Oitocentos (Rio de Janeiro: Topobooks, 1995), pp. 140–64.

⁷⁷ Florentino and Góes, A paz das senzalas, p. 88.

⁷⁸ A good discussion of the rules of compadrio are contained in Stephen Gudeman and Stuart B. Schwartz, “Purgando o pecado original: O compadrio e batismo de escravos na Bahia no século XVIII,” in João José Reis, ed., Escravidão e Invenção da Liberdade: Estudos sobre o negro no Brasil (São Paulo: Brasiliense, 1988), pp. 33–59; and Stuart B. Schwartz, Slaves, Peasants, and Rebels: Reconsidering Brazilian Slavery (Urbana: University of Illinois Press, 1992), chapter 5.

⁷⁹ This same pattern of better representation of padrinhos than madrinhas is also found in baptisms of slaves and free colored children in several nineteenth-century northeastern parishes, and holds for both legitimate and natural-born children. Rocha, “Gente negra na Paraíba oitocentista,” chapter 4.

⁸⁰ Jonis Freire, “Compadrio em uma freguesia escravista: Senhor Bom Jesus do Rio Pardo (MG) (1838–1888),” XIV Encontro Nacional de Estudos Populacionais, ABEP (2004), tabelas 1, 5, 6, and 7, pp. 11, 17, 19.

⁸¹ Schwartz, Slaves, Peasants, and Rebels, tables 4 and 5, pp. 148, 151. Only in a sample from late nineteenth-century Paraíba do Sul did Schwartz find a majority of slaves serving as godparents.

⁸² For the Bahian data, see Stuart B. Schwartz, Sugar Plantations in the Formation of Brazilian Society (Bahia, 1550–1835) (Cambridge: Cambridge University Press, 1985), p. 409; and for the Cabo Frio materials, see Ana Maria Rios, “The politics of kinship: Compadrio among slaves in nineteenth-century Brazil,” The History of the Family, 5, no. 3 (2000), p. 289. From a sample in the late nineteenth century from the same district, she finds an increase of the use of slave godparents that she claims is due to an increased concentration of slaves on larger slaveholdings.

⁸³ Góes, O cativeiro imperfeito, tabela III, p. 57.

⁸⁴ Freire, “Compadrio em uma freguesia escravista,” tabelas 8 and 9, pp. 20, 21.

⁸⁵ Luiz Rosa Mott, Egipcíaca: Uma santa africana no Brasil (Rio de Janeiro: Bertrand Brasil, 1993).

⁸⁶ Lucilene Reginaldo, “Os rosários dos Angolas: Irmandades negras, experiências escravas e identidades africanas na bahia setecentista” (Tese de doutorado, UNICAMP, Campinas, 2005), p. 202; and Célia Maria Borges, Escravos e libertos nas irmandades do Rosário: Devoção e solidaridade em Minas Gerais, séculos XVIII e XIX (Juiz de Fora: Editora UFJF, 2005), tabelas 6 and 8, pp. 230–1.

⁸⁷ Silva, “Irmãos de fé, p. 115.

⁸⁸ Costa, Vila Rica, tabela 6, p. 235.

⁸⁹ For example, women represented 46 percent of all the members of the Cachoeira do Campo Rosario brotherhood, and only 36 percent of all members regardless of status in the Pilar Parish Rosario of Ouro Preto. Kiddy, Blacks of the Rosary, tables 4–5, p. 255.

⁹⁰ Fernanda Aparecida Domingos Pinheiro, “Confrades do Rosário: Sociabilidade e Identidade étnica em Mariana-Minas Gerais (1745–1820)” (dissertação de mestrado, UFF, Niterói, 2006), tabela XI.

⁹¹ Pinheiro, “Confrades do Rosário,” tabela XIII.

⁹² In the city of Salvador de Bahia, for example, “Os mulatos se reuniam nas Irmandades de Nosso Senhor da Cruz, na igreja da Palma, na de Nosso Senhor Bom Jesus da Paciência, na igreja de São Pedro, na Irmandade de Nossa Senhora da Conceição do Boqueirão e na igreja do mesmo nome. Os negros africanos agrupavam-se por nações de origem: os angolanos e os congoleses formavam a Irmandade de Nossa Senhora do Rosário, na praça do Pelourinho; os daomeanos, a de Nosso Senhor das Necessidades e da Redenção, na capela do Corpo Santo e os Nago-Yorubás, formada por mulheres e consagrada a Nossa Senhora da Boa Morte, na pequena igreja da Barroquinha. Os negros nascidos no Brasil se reuniam na Irmandade do Senhor Bom Jesus dos Martírios e em torno da devoção a São Benedito, seja na Igreja de Nossa Senhora da Conceição da Praia, seja na de São Francisco, ou ainda na de Nossa Senhora do Rosário, como também em quase todas as igrejas da Bahia.” Humberto José Fonsêca, “Vida e morte na bahia colonial: Sociabilidades festivas e rituais fúnebres (1640–1760)” (tesis de doutorado, FAFICH/UFMG, Belo Horizonte, 2006), p. 114.

⁹³ In the city of Rio de Janeiro, the Irmandade de Santo Elsebão e Santa Efigênia was exclusively for Minas (Western African) persons, and Central West Africans and Creoles were excluded from membership. Apparently these Angolans and pardos were concentrated in the Rosário brotherhoods. Soares, Devotos da cor, pp. 188–9.

⁹⁴ The Rosario brotherhood in Mariana in the 1870s was already 90 percent free, and half of new members to the Barbacena Rosario brotherhood in the 1850s and 1860s were free persons. They were, of course, increasingly native born as well after 1850. Kiddy, Blacks of the Rosary, pp. 152–3.

⁹⁵ This was the case with the Rio de Janeiro shoemaker brotherhood. Mônica de Souza Nunes Martins, “Entre a Cruz e o Capital: Mestres, aprendizes e corporações de ofícios no Rio de Janeiro (1808–1824)” (tesis de doutorado, Rio de Janeiro: U.F.R.J./I.F.C.S., 2007), p. 50.

⁹⁶ Alisson Eugênio, “Tensões entre os Visitadores Eclesiásticos e as Irmandades Negras no século XVIII mineiro,” Revista Brasileira de História, 22, no. 43 (2002), pp. 33–46, and her unpublished essay, “Solidariedade e tensões no simbólico das festas das irmandades de escravos e libertos em Minas Gerais no século XVIII.”