Portal de Periódicos da CAPES

Gender and the Manumission of Slaves in New Spain

2006; Duke University Press; Volume: 86; Issue: 2 Linguagem: Inglês

10.1215/00182168-2005-005

1527-1900

Frank “Trey” Proctor,

Colonialism, slavery, and trade

On October 6, 1677, Juan López, a widower from Mexico City, went before a local notary to free five of his slaves. Juan testified that he was having an amorous relationship with Isabel de la Cruz, his black slave, which had produced four children, all under the age of five. "Considering that it is unjust and unconscionable that my children are slaves," Juan explained, "they, and their mother Isabel, are hereby freed from this day forward."1Stories such as this have certainly captured the attention and imagination of scholars who try to understand patterns of manumission throughout the Americas. Many scholars have looked to sexual relations between male masters and female slaves, and the paternity of children resulting from those unions, to explain why most slaves were freed. Such explanations arise within a conception of Latin American slavery as distinct from its harsh northern neighbor, due to high rates of racial mixture between whites and blacks. For many, liaisons between masters and their slave women served to reduce the social distance between slave owners and their chattel.2 Thus, descriptions of masters liberating their slave lovers or illegitimate children explain patterns in slave manumissions, reinforce the image of the unique racial trajectory of Latin American history as a whole, and undergird accepted distinctions between slavery in Latin America and British North America.Yet the historical record is replete with examples that do not fit this mold. On July 27, 1675, doña Teresa de Alba y Corea, a widow also from Mexico City, appeared before another notary to manumit of two of her slaves, executable upon her death. Doña Teresa promised freedom to Isabel de la Asunción, a 20-year-old mulata, and María de Anzures, her three-year-old daughter, "in remuneration for the care and punctuality with which Isabel had served her, and because of the great love and goodwill that she had for Isabel and her daughter."3Which of these examples was more typical for manumission in New Spain? The similarities between the cases of the two Isabels and their children discussed above are exemplary of traditional historiographical understandings of manumission, but their differences highlight the need to push those treatments further in order to discern patterns within the process of freeing slaves in the New World. That in both cases a woman and her children were freed reflects larger hemispheric patterns (see tables 1 and 2). Yet only one can be explained through the logic of male master/female slave liaisons. How can we frame both cases as distinct manifestations of more general processes that can explain patterns of manumission in colonial Mexico?This study attempts both to explain the process of manumission and to illuminate the operation of slavery as a social institution and set of relationships in the colony. To do so, manumission must be theorized as an inherently gendered social process. "Gender," Joan Scott argues, "provides a way to decode meaning and to understand the complex connections among various forms of human interactions."4 Approaching individual manumissions as the result of the interactions not only between male masters and female slaves but also among masters and slaves of both sexes — relations determined and influenced by colonial gender norms — provides an expanded understanding of the process of freeing slaves, the institution of slavery, and the society in which they took place.5This consideration of the gendered nature of manumission is drawn from a larger work focused on slavery in colonial Mexico from the end of the regular slave trade in 1640 through the 1760s. I chose this end point to avoid the ideological implications of the Atlantic revolutions (American, French, and Haitian) for slavery and the sweeping economic and social changes brought about by the Bourbon reforms during the second half of the eighteenth century. Before 1640, the primary demand for African slave labor in New Spain came from the silver-mining, sugar, and urban economies. However, most historians assert that demand for slaves underwent a steady decline after the close of the slave trade. My larger study compares slavery in Mexico City, Guanajuato (the most important mining center in colonial Mexico), and the sugar zone of central Mexico in order to test that assumption of decline and to explore the changes brought about by the creolization of the slave population after the end of the regular slave trade.6This study draws evidence from several sources. Frederick Bowser's work on manumission in Mexico City provides the data for 1580 – 1650, while manumissions recorded in cartas de libertad (manumission deeds) and wills from all extant notary books in Mexico City (1673 – 76 and 1723 – 26) and Guanajuato (1699 – 1750) form the heart of my analysis.7Most studies of manumission begin by highlighting the importance of women and children among freed slaves. New Spain proves no exception, as slave women and children accounted for between two-thirds and three-quarters of all freed slaves with known ages. Female slaves dominated, receiving between 55 and 65 percent of all manumissions. Children under the age of 15 also represented a substantial portion of liberated slaves in colonial Mexico (table 1). These figures for New Spain are very consistent with those seen throughout the New World (table 2).Historians have offered numerous possible explanations for the prevalence of women and children in manumissions: the lower market value of women and children compared to adult men or the increased economic opportunities for urban slave women to accumulate money to buy freedom for themselves or their children. Others suggest that these patterns were perhaps part of larger slave family strategies for achieving freedom. Freeing a woman meant that any future children would be freeborn (even if the father was still enslaved), as only children born to enslaved women were slaves according to Spanish law.8 In addition, children may have been more likely to free their mothers due to the matri-focal nature of most slave families.9 Lastly, many historians suggest that slave women's domestic occupations within the household may have provided more opportunities for daily interactions with their owners, which may have, in turn, translated into increased opportunities for freedom for themselves and their children.Yet, the argument that slave women were able to manipulate their sexuality to improve their own chances for manumission, and those of their children, remains a dominant explanation for the preponderance of these two groups among freed slaves. Orlando Patterson, for example, suggests that women dominated amongst freed slaves "primarily because of their frequent sexual relations with the master or with other free males."10 Similarly, Douglas Cole Libby and Clotilde Andrade Paiva argue that one of the most important factors in female dominance in manumission "was the possibility of building up relationships with free males, relationships which often produced offspring as well as levels of affection, both of which could lead to manumission."11 And, explaining high rates of child manumission in Lima, Peru, and Mexico City between 1580 and 1650, Frederick Bowser argues that Spanish fathers frequently freed their slave children, who accounted for a over half of all liberated slaves in those cities prior to 1650.12Despite all of these potential explanations for the preponderance of women and children among liberated slaves, few historians have approached the subject as explicitly gendered. Only Rosemary Brana-Shute, in her study of slavery in Suriname, has tried to connect all the different issues discussed above by placing them all under the rubric of slave women's traditional roles within the domestic sphere. Focusing on slaves, she urges scholars to theorize the advantageous position of women and children as a result of the gendered nature of manumission. However, the emphasis on slave gender presents only part of the picture.13Importantly, the sex of masters liberating slaves provides a slightly different perspective on the potentially gendered nature of manumission. Numerous studies calculate manumission rates based on the sex of the master, just as they do for the sex of the slave.14 Yet it is not enough to just count the numbers of men, women, and children freed or the number of men and women doing the freeing. The sex of the liberating master must be cross-tabulated with the sex of the freed slave to see which masters were freeing which slaves, a calculation few studies have made.15By including master's sex alongside that of the freed slave we can engage in a larger discussion of the implications of the gendered interactions between slave and master on the larger process of manumission and thereby treat it as an explicitly gendered social process. Furthermore, doing so will test the theory that amorous relations were the driving force behind the majority of slave liberations.To critically examine the question of gender, we must also analyze information such as age, the origin or method of acquisition of the slave (whether they were born or purchased into the household that freed them), and the impact of the intimate, but not necessarily sexual, relations between masters and the slaves they freed. These tests suggest that contact between adult slaves and masters of the same sex within gendered social spaces — particularly the domestic sphere for women — had more influence on manumission patterns than did contact between masters and slaves of the opposite sex, especially male masters and female slaves. Such a finding significantly undermines the paternity/amorous relations thesis.Lastly, the patterns associated with manumission must be understood within the context of the changing slave market in New Spain. After the end of the slave trade in 1640, the colony had to rely on natural reproduction to meet the demand for slave labor. In that context, the relative value of both slave women and children increased compared to that of adult men over time. Correspondingly, the advantages that women and children enjoyed in manumission decreased over time. In the end, such an inquiry will shed light on manumission patterns from multiple new perspectives.Although the tantalizing story of Juan López indicates that men occasionally acknowledged their sexual relationships with slaves and the resulting progeny in the course of granting them freedom, we find few such direct acknowledgements in the manumission records of New Spain. In fact, no male master explicitly acknowledged that he was freeing his lover or his children in any of the 347 manumissions by men included in this study.16To reexamine the thesis that paternity and sexual relations were a significant factor in patterns of manumission, let us first look at the sex of manumitting masters. Surprisingly, women freed a significant proportion of slaves in colonial Mexico and throughout the Americas. Women were responsible for 39 – 55 percent of slaves liberated in New Spain (with the highest rates found in Mexico City) during the seventeenth and eighteenth centuries, representing one of the highest known rates of manumission by female owners in the Americas (table 3). In short, male masters freed far fewer slaves than might be expected if we accept the hypothesis that patterns within manumission reflect the impacts of sexual relations and paternity.Furthermore, the percentage of slaves freed by women seems to have been much greater than the overall percentage of slaves owned by women. Men largely dominated economic life in colonial Spanish America, and thus we might expect to find both more male than female slave owners and larger slaveholdings among male slaveholders. The lack of census materials frustrates any attempt to measure the extent of female slaveholding in New Spain, but it is clear that the percentage of slaves freed by women is much higher that the percentage of slaves bought or sold by women. Women purchased approximately 19 percent, and sold nearly 29 percent, of slaves in Mexico City in both 1675 and 1725, but accounted for 48 – 54 percent of manumissions. In eighteenth-century Guanajuato, women bought 24 percent and sold 37 percent of slaves but freed 39 percent.17 These rates of purchase and sale support the assumption that men were more important as masters than were women. If we can use the sale data as a proxy for ownership, then it appears that women were more likely to free the slaves they owned than were men. Thus, slaves owned by women, although in the minority, were more likely to achieve freedom than those owned by men.In addition, the ratio of female to male slaves was more skewed among slaves freed by women than among those freed by men, further challenging the theory that the sex imbalance in manumission was due to amorous relations between slave women and their masters. In Guanajuato, 1699 – 1750, 62 percent of slaves freed by women were female, compared to only 51 percent of those liberated by men. And in Mexico City, two-thirds of slaves freed by women were female, and female slave owners liberated the majority of slave women in both the 1670s and 1720s (table 4).18 Conversely, in both contexts male masters liberated the majority of male slaves. In Mexico City in the 1670s, men and women freed nearly the same number of male slaves, but by the 1720s slaveholding men freed nearly two-thirds of all male slaves. The connections between the sex of the master and that of the slaves they freed indicate that patterns within manumission primarily resulted from the contacts between women (slaves and mistresses) and between and men (masters and slaves), rather than between female slaves and their male owners.19 These patterns suggest that gender was critical in the manumission process, but not for the reasons that we once thought.The breakdown of child manumission by master's sex reveals similar patterns. Children were proportionally more important among slaves freed by women than among those freed by men (see table 5). For example, in 1720s Mexico City, children represented 36 percent of slaves manumitted by women, compared to 19 percent of those manumitted by men. In fact, in Mexico City women slaveholders freed six out of every ten children manumitted in both the 1670s and the 1720s.20Putting slave age and sex in conversation with slaveholder sex leads to a slightly different perspective on manumission in New Spain. Women and children did receive the lion's share of manumissions, but the association of those patterns with sexual or paternal relationships between male slave owners and female or child slaves is suspect. Rather, the stronger associations were between female slave owners and slave women and children, on one hand, and between male slave owners and adult male slaves, on the other.Why women and children were more important among slaves freed by women will be addressed below, but at present our interest lies in the relatively few women and children liberated by male masters. It is possible to estimate the impact of amorous relations and paternity on the motivations of men who liberated slaves by focusing on the types of manumissions they gave slave women and children. Following the historical literature, which holds that gratis (immediate, without cost) manumissions granted by men were "almost exclusively limited to cases in which masters were, in fact, liberating their own children or sexual partners," we can calculate the potential importance of amorous relations to overall patterns of manumission.21 By including every gratis manumission granted by men, we find that less than one out of every seven (13 – 17 percent) manumission cases could possibly fit the description above.22 Similarly, if sexual relations were a significant motivation to grant gratis manumissions to women and children, we might expect men to do so at higher rates did women. Such was not the case, however, as male and female masters granted gratis manumissions at similar rates.23 Even if sexual relations between masters and slaves influenced manumission patterns by male masters, it accounts for only a small proportion of freed slaves overall. Therefore, we must put to rest the overarching association of access to freedom with the presumed sexual contacts between male masters and their slave women, at least in New Spain.24If the impact of sexual relationships and paternity upon manumission was not as great as many assume, then what other factors determined which slaves were more likely to be freed? Manumission was the result of very complex human interactions. Thus, we must focus on the relationships between freed slaves and the masters who freed them, both individually and collectively, to begin to understand the process of liberating slaves as a gendered phenomenon. First, the origins of slaves — purchased versus born into the household (criado) — went a long way toward determining which slaves were actually freed. But, as we will see, it was not simply whether a slave was raised in a household, but also the types of relations formed within the domestic sphere, and with whom they were formed, that determined relative access to liberty. Similarly, we can and must decode the meaning of the actual explanations of motives given by masters in cartas de libertad. Doing so leads to a very different vision of manumission in colonial Mexico, and perhaps throughout the hemisphere.In New Spain, criado slaves were liberated more often than purchased slaves. In Guanajuato, 65.8 percent of slaves manumitted in cartas de libertad were either born or inherited into the household that freed them.25 In Mexico City, criados enjoyed a similar advantage, but they witnessed it decline over time. In the 1670s, two of every three freed slaves were criados, but by the 1720s they accounted for only 51 percent of manumissions. Not only were criado slaves more likely to be liberated; there also was an association between the manner of acquisition and the form of manumission. Criado slaves were more likely to receive gratis manumissions than were purchased slaves. Conversely, purchased slaves were more likely to purchase their own freedom than were criados.The relation between where a slave was raised (the form of acquisition) and the possibility of manumission points to many factors. First, the form of acquisition may indicate the urgency of an owner's need for the labor of a particular slave. More than likely, purchased slaves were acquired to fill specific labor needs in the owner's home or business, perhaps decreasing their master's subsequent willingness to free them. Secondly, slaves born and raised in a household were more likely forge long-lasting, personal relationships with their owners, which may have positively impacted their opportunities for manumission. Lyman Johnson, for example, hypothesized that criados "would be more likely to rely on the paternalistic interest of their masters and develop those skills, habits, and behaviors that were favorably evaluated within the limited world of the household," which may have translated to increased opportunities for freedom.26In explaining the dominance of criados among freed slaves in Bahia and Buenos Aires, Stuart Schwartz and Lyman Johnson both note the tendency of slave owners to express a sense of responsibility and obligation for slaves they freed, particularly for those raised in their household. They suggest that sense of responsibility might be called "surrogate paternity or maternity."27Schwartz, however, despite his own conclusion that women were twice as likely as men to use such expressions when manumitting a slave, and although he included surrogate maternity in his initial discussion, immediately turned his attention to the possibility that "surrogate paternity" was a mask for the real thing. In fairness, Schwartz does conclude "that is it not clear" if surrogate paternity was a mask for real paternity. But, focusing on that possibility and overemphasizing paternalism reinforces the assumption that men were liberating the great majority of slaves, thereby inferring that the important relationships that prompted manumissions were forged between male masters and their slaves.In New Spain, as Schwartz found for Brazil, women were much more likely to use such expressions when freeing slaves. The specific explanations given by female slave owners as to why they freed particular slaves highlight the impact of the shared social universe of mistresses, their female slaves, and the children of both on manumission. One important pattern that emerges from such statements is that the motivations behind particular manumissions were clearly related to the caregiving roles of women — both slave and free — particularly childrearing.Numerous mistresses, married and widowed, freed slave women or their children in remuneration for service as wet nurses or nannies. For example, Bernarda Fernández de Guevara freed Rosa, her 37-year-old mulata slave, for her assistance in raising her daughters. In some instances, as in the case of Rosa, mistresses freed their female slaves, but more often they freed the children of their slave wet nurses or nannies. For example, Magdalena de Soria, a widow, freed Agustín, her 1-year-old slave, in compensation for the good service of his mother Juana de San Antonio. Similarly, doña Catalina Rosal y Lugo and her husband don Fernando Rubiera de Valdez rented the services of Juana, a 26-year-old black slave, to serve as a chichigua (wet nurse) for their 10-month-old son. In return for the "loving manner" with which Juana had cared for her son, doña Catalina paid Juana's owner, Melchora de los Reyes, one hundred pesos for the freedom of Juan Lorenzo, Juana's own infant son.28 Serving as surrogate mothers to masters' children did increase the likelihood of liberation for slave women and their children.Male masters could and did reward slave women for fulfilling maternal roles within their households as well, but did not do so with the same frequency as did mistresses. Juan López de Rivera freed a 2-year-old slave named Francisco as compensation for the care and love with which Francisco's mother had raised his daughters.29 Examples like this should not, however, obfuscate the fact that in most cases, domestic service and childrearing was the shared social space of mistresses and slave women, which could have resulted in increased opportunities for freedom for those particular slaves and their children.Childrearing was but one of myriad possible types of caregiving relationships between slave women and their masters that Brana-Shute includes under the umbrella of the influence of slave gender on manumission patterns. For example, masters and mistresses often freed female slaves for the care given to adults, especially during illnesses. Isabel de Olivares, a widow from Mexico City, freed her 35-year-old mulata slave, María de los Reyes, in return for "coming to her aid and healing her while she was sick."30 This is but one example of the increased levels of intimate, but clearly not sexual, contact between slaves and their masters that positively impacted access to freedom.An exploration of childrearing from another perspective further underscores the importance of the shared social universe of mistresses and their slaves within the domestic sphere. Mistresses often expressed notions of surrogate maternity when manumitting slaves, particularly children, and did so much more often than did men. For example, Nicolasa de Guevara y Orellano, a widow from Mexico City, freed her 11-year-old slave María de Cristo, because "she was born into my hands and grew up in my home." Similarly, María BenítezTamayo freed 6-year-old María de la Trinidad and her 2-year-old sister, Catarina de San Ildefonso, because the "girls had grown up in her arms, and she loved them like her own daughters."31 Obviously, real maternity was not at issue for these women, so their explanations should be taken as metaphorical language that conveyed the privileged position of slaves born and raised under the authority of their mistresses. Such expressions illuminate the need to expand our consideration of the role of gender beyond the narrow focus on the slave.This language serves to highlight the higher level of intimacy between mistresses, their slaves, and their children. On the other hand, men rarely, if ever, used such language to explain their motivations for freeing slaves. Women — mistresses and slaves alike — shared the experiences of bearing children, raising them, and providing care for other members of the slaveholding family, especially during illnesses. This created connections unattainable for slave women and male masters in most cases. Those shared experiences in the social reproduction of society created a sense of social intimacy that increased the likelihood of achieving freedom for those slave women and their children. Therefore, it appears more likely that intimate daily interactions between mistresses and their slaves represented the single most significant determinant as to which slaves were freed and why. That privilege was not extended only to slaves born in the home but potentially to any slave who operated within the same social space as their female owners.The importance of that shared social universe on the opportunities for slaves to achieve freedom is more clearly illustrated when the sex of the manumitting master and the method of acquisition of the freed slave are cross-tabulated. In Mexico City and Guanajuato, criado slaves were proportionally more important among slaves liberated by women than by men. For example, in Guanajuato nearly 90 percent of slaves freed by women were criados. On the other hand, just over half of slaves freed by men were criados. In Mexico City during the 1720s, nearly 63 percent of slaves freed by women were criados, compared to 31 percent of those freed by men (table 6).Remarkably, however, the advantage enjoyed by criados owned by women was not gender specific. As suggested by table 6, female criados (criadas) were much more likely to receive manumission from mistresses than masters, suggesting that the social universe shared by women positively impacted the opportunity for freedom. However, male criados owned by women shared a similar advantage. In the eighteenth century, male criados were much more likely to be freed by mistresses than by men in both Mexico City and Guanajuato.That male criados were proportionally more important among slaves freed by women suggests that interactions within the gendered space of the domestic sphere, rather than some sense of sisterhood that united mistresses and their female chattel, determined increased opportunity for liberation.32 Consider that mistresses were more likely to free their slaves than were male masters, that mistresses were more likely to express sentiments of surrogate maternity — as opposed to surrogate paternity — in their manumission documents, and that mistresses appear to show more favor toward slaves born and raised in their homes than male masters. All these pieces of evidence point toward the conclusion that the relationships formed within the domestic sphere were the primary motor behind manumission.This advantage was the result of shared social, although not necessarily physical, space. Studies of urban slavery have found that a high number of slaves did not, in fact, live with their owners. In cities like Salvador, Rio de Janeiro, Lima, and Mexico City, slaves were often forced to find housing for themselves and their families. Such a situation could, of course, render the argument regarding the high level of daily contact between slaves and mistresses a moot point. However, our conception of the domestic sphere should not be limited to physical space. Bianca Premo has argued that the types of bonds of social intimacy discussed herein were formed between masters and slaves through the bonds of childrearing under the master's protection and patronage in colonial Latin America, regardless of whether or not masters and slaves actually shared the same dwelling space.33 Furthermore, the cartas make it clear that a specific group of slave women — those occupied in domestic service and childcare for their mistresses — enjoyed the advantages discussed above. The young children of those slave women, we must assume, would also have spent a great deal of time within the domestic sphere.The level of intimacy created by the daily contacts between mistresses, their slave women, and the children of both within the domestic sphere was much higher than that between male masters and their slaves. Slaves who came into regular contact with their mistresses within the domestic sphere — regardless of their gender — had the greatest chance for freedom. Yet, those shared experiences that led to a sense of social intimacy, which served to increase opportunities for manumission, did not threaten the social order that reinforced the social institution of slavery in any real way.Compadrazgo (god-parentage) patterns for slave children further underscore the effects of the shared, gendered social universe on the relations of mistresses and their slaves and the increased intimacy resulting from it. In the mining regions of northern New Spain, the representation of Spaniards as godparents for slave children increased during the period between 1652 and 1749. During that time, Spanish women served as godmothers for nearly 30 percent of slave children baptized in Guanajuato and San Luis Potosí.34While male members of slaveholding families never sponsored one of their own slaves at baptism, and the godparent and listed owner were never the same person, young women within slaveholding families occasionally served as godmothers for infants born to a family slave. For example, in May of 1653, a slave infant name