Portal de Periódicos da CAPES

Free Colored in a Slave Society: São Paulo and Minas Gerais in the Early Nineteenth Century

2000; Duke University Press; Volume: 80; Issue: 4 Linguagem: Inglês

10.1215/00182168-80-4-913

1527-1900

Herbert S. Klein, Francisco Vidal Luna,

Race, Identity, and Education in Brazil

Brazil was traditionally depicted as a plantation economy dominated by slaves and slave owners. However, all recent studies have denied the picture painted so ably by Gilberto Freyre over a half century ago of a dichotomous society dominated by the plantation; in fact, most scholars have stressed that Brazil looked more like the United States than the West Indies in the relative weight of slaves and slave owners in the population.1 Our survey of São Paulo indicates that on average roughly a third of the population were slaves and roughly a third of the free population were slave owners. These figures are representative of most of Brazil, and compare favorably with those ratios in the United States in the nineteenth century.2Although nineteenth-century Brazilian slave society differed little from the contemporary southern United States in terms of the size and relative weight of slaves and their masters, it differed substantially in the color of its free population. Whereas the free population was over 95 percent white in the United States, whites tended to be less than half the free population in most of Brazil.3 By the early nineteenth century, Brazil had the largest free colored population of any slave society in the Americas. At the time of the first national census in 1872, some 16 years before final abolition of slavery, the free colored—all of them had slave origins—numbered 4.2 million persons, compared to just 1.5 million slaves. These free colored were, in fact, the largest single racial/status group within Brazil itself.4 In sharp contrast to the United States where less than one percent of all slave owners were non-white, Brazilian freed persons of color were well distributed throughout all the provinces of the Brazilian empire, as much urban as rural in their settlement pattern, and a significant number of them were heads of slave-owning households.5 Yet that was a time when the slave-based coffee economy was reaching its maturity and the price of slaves was on a long-term rise.6Despite their importance in Brazilian slave society, few of all the recent investigations of African slavery in Brazil have focused on the economic and social role of the free colored population within Brazil before the end of slavery.7Brazilian society like all other American slave regimes was by its very nature racist and the white elite discriminated in various ways against its free persons of color, even as it permitted a very active level of manumission. But how effective and important was that discrimination in controlling economic and social mobility? We have had little sense of how these free persons of color were integrated into the world of free persons and the market economy: Were the free colored largely cut off from normal avenues of economic and social mobility through legal means, as occurred among the free colored in the United States?8 Or were the barriers to mobility largely economic? Were free non-whites able to respond to market incentives and integrate themselves into the larger free society despite the white elite’s racist expressions of hostility to blacks?9 In Brazil there were no legal constraints to geographic or economic mobility, but to what extent did the free non-whites succeed in maintaining the skills they acquired as slaves, or compete with whites for land or slaves.Though we cannot fully answer these questions, our analysis of 2 major counties (municípios) in Minas Gerais and 15 counties in São Paulo provides a reasonable picture of the role of these free colored people within early-nineteenth-century society. Like the rest of Brazil, a majority of the population in these provinces consisted of free people, with a significant percentage of free blacks and mulattoes. Most of our data comes from unpublished provincial censuses that form the basis for much of the newer social and economic histories of the nineteenth-century economy and society of south-central Brazil.10 These so-called mappas provide detailed annual censuses and production data in the provinces of São Paulo and Minas Gerais from roughly the 1770s until the early 1850s, with the most complete runs coming from the 1820s and 1830s.11 It should be noted that our data do not provide the distinctions found in some Brazilian primary sources of this period between free persons of color born free (livres) and those who were manumitted (forros or liber-tos), though it would appear from the few data that we have that the manumitted persons tended to have far higher ratios of women than those of the free colored population born free.12 We use the term free colored to refer to both groups, without distinction of their status at birth. Moreover, we have kept the color terminology as given in the São Paulo censuses, which breaks down the free colored population into pretos (or blacks), and pardos (or mulattos, or browns).13We selected several counties in Minas because that was the region with the largest number and ratio of free colored of any province in Brazil. São Paulo was one of the newer plantation regions that expanded rapidly although African slavery arrived relatively late there; consequently, São Paulo had a smaller ratio of free colored. In the 1830s the free colored probably surpassed the total number of slaves in Minas Gerais, but in São Paulo they still remained less than half of the total colored population in the province; the free colored of Minas numbered several hundred thousand, whereas there were only about 66,000 in São Paulo at this time.14 Thus a comparison of these two regions provides important end points for the spectrum of roles and possibilities to which the ex-slaves had access during the slave period.The 15 São Paulo counties we selected are representative of São Paulo during this period, in terms of their distribution by region and economic activity, and account for about 60 percent of the provincial population. For Minas Gerais, we chose two extreme examples of free colored participation: the traditional mining region of Sabará in central Minas, with its predominantly black population; and the agricultural settlement of Campanha in the southeastern frontier with São Paulo, with its predominantly white residents.15 Each region defines the end points of the free colored and its position within Minas society.The paulista counties include three largely urban populations: Santos, which was the primary port of the region; the capital of São Paulo province; and Curitiba. At the same time these counties represent the five major regions within the province. This includes Areias, Cunha, Taubaté from the rapidly changing Paraíba Valley region that was the first center of coffee production in the province, an important sugar exporting region and the home to a major food crop economy which supplied the imperial capital at Rio de Janeiro. The expanding agricultural frontier region of the west Paulista zone with its important sugar exporting centers is represented by Piracicaba, Taubaté, and Jundiaí; the coastal counties by Antonina, Paranaguá, Ubatuba and Santos, while Bragança and the city of São Paulo represent the capital region and the special environment of the southern region of the province is represented by the counties of Sorocaba, Curitiba, and Castro. Both counties of Minas Gerais are roughly similar in the structure and size of their population and in their dedication to agricultural and craft activities. Although some mining activity still occurred in Sabará (gold mining had made the region famous in the eighteenth century, although it declined after 1750), it was no longer the predominant sector of the local economy. By the end of the eighteenth century the province of Minas Gerais had become a very complex agricultural, cottage industry and mixed mining economy of which gold production was a minor element. The Minas Gerais economy with its exports of sugar, cane alcohol, food staples, and low-quality woven cotton cloth, more resembled the economy of neighboring São Paulo. Both Campanha and Sabará were rather typical of the province as a whole, in their concentration in agriculture, with a minor but important share of activities in crude textile manufactures, some metal and wood working, and some mining activity. In the two zones there was a significant sugar refining industry producing cane alcohol (aguardente) for local consumption and both also had a very active commercial sector. Though there was considerable self-sufficiency, both zones were closely tied into a larger zonal economy that included very active trade with the neighboring coastal provinces.16Controlling for differences in size and relative importance of the resident free colored population, our examination of the population census data shows that the free colored, except at the elite level, were found in all the occupations practiced by their contemporary white neighbors and had much the same social, occupational, and demographic characteristics as their non-slave originated peers. Regardless of whether they lived among predominantly Afro-Brazilian populations or among predominantly white ones, there was relatively little difference for the free colored in their patterns of work and social organization from their white neighbors. In addition, our analysis illustrates that free persons of color were significant slave owners in their own right.Our selection of counties was determined by their differences in racial composition as well as economic activity. Using the extremes in racial composition represented by these 17 counties in Minas Gerais and São Paulo, we can control for racial density as a key factor in determining integration or rejection of the free persons of color into non-slave society. By the racial standards of south-central Brazil in the first half of the nineteenth century, these regions incorporate the extremes. Among the 15 counties, 7 had a higher ratio of whites than Campanha, which was one of the “whitest” counties in Minas Gerais. In turn, none of the paulista counties had as low a ratio of non-colored as was found in Sabará (see table 1).All these counties from São Paulo and Minas Gerais had approximately the same ratio of slaves, that is, just over a quarter of the total population. Combining the slave and free colored population resulted in all but 7 of these counties having a majority of the total population being non-white, though because of their size these 7 meant that the 15 paulista counties overall had a slight majority of their population being white.We did not find a negative correlation between slaves and free coloreds or between the presence of large numbers of whites and freedmen. Free-born colored and liberated slaves and their offspring were an important element everywhere; in fact, their distribution was highly correlated with the distribution of whites and also with the distribution of slaves across these 15 São Paulo counties.17 Roughly they accounted for a fourth of the total population in all the paulista counties. The relative importance of the free colored in the total population (about a fifth of the total) again showed the São Paulo counties closer to the quarter of total population standard of Campanha than to the majority situation of Sabará. This was the case, even though slaves were on the whole a more important share of total population in São Paulo then they were in Minas Gerais. Although free persons made up 43 percent of all non-whites (again a figure quite close to the Campanha norm) they were but 28 percent of all non-slaves, thus well below even the Campanha model and very far from accounting for four fifths of the population as in Sabará. Thus for all their importance within the São Paulo regions, the free colored had still not reached the level of importance that they had already obtained in neighboring Minas Gerais. Given the fact that the free colored in São Paulo would eventually reach levels of importance similar to that of Minas by the last quarter of the nineteenth century, it might be argued that this difference in 1829 was more a question of differing historical trajectories than any major difference in societal attitudes towards the free colored themselves. Minas Gerais amassed a large population of slaves quickly in the first half of the eighteenth century and then faced a declining or stagnating economy until the early decades of the nineteenth century, whereas São Paulo started from a much lower base of slaves at a much slower level of accumulation, which also explains the slower growth of a free colored class.Despite the lower ratios of free colored found in these paulista counties, what is still impressive is just how significant the free colored were among the total colored population this early in the nineteenth century. The decade of the 1820s was a major period of African slave importation to Brazil, and by the census of 1835 Africans made up 45 percent of the total provincial slave labor force, the same ratio that they were in these 15 counties. Yet even in these expanding zones with their ever-increasing ratios of African-born slaves, the free colored were 43 percent of the total colored population. Again this was closer to Campanha than Sabará. All this suggests that the role of free citizens was well defined for blacks and browns in south-central Brazil almost 60 years before the final abolition of slavery, and even an intensification of the slave trade to this zone did not reduce their importance.Moreover the demographic structure of the free colored would suggest that their potential for growth was quite high. On average, they were younger than the slaves, with the highest ratio of women. This differed sharply from the male-dominated slave population and the balanced sex ratio of the whites. The fact that there were 91 men for every 100 free colored women in the 15 São Paulo counties is most likely due to the systematic dominance of young women among manumitted slaves in all studies done on manumission.18 Moreover, this dominance prevails whether the emancipator is male or female.19 Unfortunately, our census material does not break down the free population into those born free and those manumitted in their lifetime, nor does there currently exist for any other Brazilian region a large sample of origins data for the free colored by which to generate a firm estimate of the relative importance of these recently freed persons in the total free colored population. But the fact that this overall imbalance of the sex ratio of the free colored population is exhibited even in the census of 1872, would suggest that the steady manumission of slaves, the majority of whom were women, meant that this population was unusual in its sexual balance compared to both the slave and free white population.20 It also meant that this population most likely had higher reproductive rates than the slaves, but probably even higher than the whites as well.The free colored were well distributed in the provincial population. Although they were 21 percent of the total population, they were a quarter of all heads of households. Moreover they were well distributed throughout the counties. They ran from 12 percent of all households in some of the smaller towns, to a third or more of the households in the larger counties of Santos, São Paulo, and Curitiba, and even reached three quarters of the population in the coastal town of Guaratuba, a rate quite close to that of Sabará (see table 2).Among the households that owned no slaves, the free colored were well represented everywhere. In the paulista counties they made up a third of such households; in fact, they were in the majority in a pretty diversified group of 6 of both primarily urban and rural counties (Cunha, Jundiaí, Santos, São Paulo, Antonina, and Guaratuba), reaching an extraordinary level of 85 percent of such non-slave-owning households in the small coastal town of Guaratuba. Here again, the paulista counties were closer to the 37 percent non-white norm of Campanha than Sabará, whose dominance of non-white households among the non-slave owning households was almost identical to that of Guaratuba. Not unexpectedly, given the relative importance of women in the free colored class, free colored women who headed households tended to control a larger share of such households than free colored men did among the men heading these non-slave owning units. Typically free colored women made up just under half of such female-headed households, compared to free colored men who represented just under a third of the total of such households.However, when we examined these same households by the criteria of slave ownership, this same level of importance and distribution was not evident. Free colored were significantly underrepresented as slave owners, being but 6 percent of the slave owners in the 15 São Paulo counties, and just 14 percent in the two Minas Gerais counties. Most free persons of color in both provinces did not head households that owned slaves. There was some modest variation by county, with a few of the urban centers sometimes having a higher ratio of slave owning families among the free colored population. Thus in the port of Santos, the free colored were 22 percent of all slave-owners (and a quarter of all free colored owned slaves), but even here the average slave holding was smaller than the norm in the rural agricultural communities. Only 283 free colored families owned slaves in the 15 counties surveyed, being only 6 percent of both all slave owners and all free colored households. This compared to a 35 percent slave ownership ratio among the whites, who were 96 percent of all slave owners. Even though the counties of Minas Gerais had twice the ratio of slave owning free colored households, even here whites had a higher representation, with 47 percent of their households holding slaves. Although slave ownership was an option for the free colored, and differed by the sex of the head of household, it was clearly a possibility for only a minority, even in Sabará. In the latter town, were they accounted for three quarters of all free persons, the free colored headed 43 percent of the slave-owning households (see table 3).This very low participation of free persons of color among the slave-owning households could probably be explained by economic factors. To own slaves required a level of wealth that was greater than the average for even the majority of free white heads of household. Only 28 percent of all households owned slaves in the 15 paulista counties and the 2 in Minas Gerais, and even among paulista households headed by whites, only a little over a third owned slaves. Nevertheless, free persons of color were still far behind their white peers in owning slaves. This fact can probably be better understood as one of the economic legacies of slavery rather than as a result of race prejudice. Free persons of color had much lower levels of initial savings when they reached free status than those that most whites began life with. As we will see when we examine their occupations, the free colored tended to be less involved in farming than their white compatriots, another indication of their lesser initial income levels than was probable among whites who headed households.But greater historical and familial initial poverty may not account for the internal divisions within the free colored of all counties. The very high ratio of either browns or mulattoes to blacks is evident everywhere among the free persons of color in the total populations of these counties. Although pardos (or browns) represented 61 percent of all Afro-Brazilians slave and free in Sabará, they were 84 percent of the free colored. In Campanha 45 percent of all Afro-Brazilians were pardos, but they were 83 percent of the free colored class. In the 15 São Paulo counties in the census of 1835 the same pattern could be observed in an even more distorted fashion. Here browns were just under a quarter of the total population, and just under half of all non-whites, but they were only 16 percent of the slaves and 94 percent of all free persons of color.This discrimination by color among non-slave-owning households could be due to the residual racism in the society as a whole, which penalized blacks more than browns, or it could be that the origins of persons entering the free colored class was more likely to favor the mulattoes over the blacks. The racial prejudices of the white society did guarantee that manumission would favor those of mixed racial background as opposed to their non-mixed brethren. Blacks tended to purchase their freedom more than did their pardo peers.21 On the other hand, among the free persons of color, many of whom by now had been born free, mulattoes could be produced from relationships between whites and mulattoes or whites and blacks, or even whites and Indians, whereas all blacks came only from parents who were both black. It could be assumed at this level of abstraction that both miscegenation and prejudice were probably working jointly to favor the more rapid growth of the mulatto class of freemen. These same color biases also appeared when we examined the 6 percent of the free colored households that owned slaves. In the paulista counties pardos represented 94 percent of the free colored who headed these slave-owning households. There were 267 such pardo families in the 15 paulista counties compared to just 16 preto families who owned slaves. They are also evident in the two Minas Gerais counties where there were 898 pardo families that owned slaves and only 87 households headed by blacks that owned slaves.But race was not the only factor that stratified the households in these regions. Both for whites and non-whites, the sexual division among heads of households led to important differences in a series of characteristics, among which was marital status. In fact, the sex of the head of household was more important in determining marital status of the head than was race or ownership of slaves. Single-headed households of the unmarried or widowed tended to be more the norm for women than for men. In turn, these single unmarried or widowed female-headed households operated in different economic spheres from those headed by men of whatever color. It has generally been assumed from all studies of household types, that those headed by single or widowed women tended to be less stable in social terms and poorer than households headed by two adults. In this respect, color shows only moderate difference, and whites, blacks, and mulattoes are quite similar in their marital rates in both Minas Gerais and São Paulo regions for both sexes and for the two types of slave-owning and non-slave households. Women in these more heavily urban counties were more single or unmarried than men. It has been suggested that single women who headed these urban households migrated from farming regions as a result of loss of lands and economic livelihood in the rural parishes rather than because of increased economic opportunities available to them in the city centers. In a study of the capital region county of Santana de Parnaíba in this period, it was found that these urban women were more likely to have illegitimate children and poorer skills and occupations than their male cohorts.22The influence of sex on marital status of head is also present in the presumably wealthier slave-owning households. Whereas males heading slave-owning households were predominantly married, women heading such homes were overwhelmingly widows, not that different from the non-slave-owning households. But here it was more likely that the widows heading these wealthier households had inherited them from their husbands, and were thus not in the same poverty situation as their peers in the non-slave owning households. Nevertheless, it is still worth stressing that the ratios between the sexes remain constant even as wealth increases. It would seem from these data that women, whether slave owning or not, whites or free colored, were more likely to remain unmarried than men.But what role did color have in relationship to the occupations of these heads of household.23 Rural São Paulo and Minas Gerais in the nineteenth century, just as contemporary North America, had a large proportion of its population even in the rural areas engaged in non-agricultural activities.24 Just over half of all non-slave households were dedicated to farming in all these communities, but a substantial number of households were engaged in non-agricultural activities. After farming, crafts and commerce were the most numerous primary occupations in São Paulo and in this distribution between farm and non-farm occupations it turns out that color was important. Though free colored made up a third of all households whose occupation was known, they were only a quarter of the farming households. They did far better in the non-agricultural activities, being just under half of the major categories of artisans, day laborers and poor, and they dominated in the few service jobs and strangely among the few who lived on rents. They were just a little over a quarter of the farmers and liberal professionals but were underrepresented, in terms of their importance as heads of households, in the merchant class and the army, and surprisingly did quite poorly among the seamen and fisherman (being only a quarter of the total). When examined in terms of their own internal divisions, farmers made up well over half of the total of such free col-ored classes, followed by artisans and the poor as the next most numerous occupations. Thus if one were to summarize their occupational position among the households not owning slaves, it would be to say that they were found everywhere, and in no occupation less than a fifth of the total group. They were overrepresented in the poorer occupations and underrepresented as farmers, merchants and liberal professionals, the more well to do occupations; in fact, farmers and artisans made up 63 percent of all free colored households (see table 5).However, when it came to slave-owning households, they were in a far more limited role. All free colored heads only represented 5 percent of the households owning slaves whose occupation could be determined. These richer free colored households were overrepresented in the same professions as their non-slave-owning peers and did as poorly among the farmers and merchants. Within their group an almost identical ratio of them were dedicated to farming, but the next most important category was merchants and together these two categories accounted for four fifths of these 262 slave-owning households. Thus the patterns found among the non-slave holders were not that different from the free colored who owned slaves. Moreover, in both slave- and non-slave-owning households in São Paulo, they were overrepresented among the poor and day laborers. Even in Minas Gerais, being a jornaleiro was a very common occupation for the free colored in non-slave-owning households, while in Cunha it even stood out as the majority occupation for both black and mulatto males. These lower social and economic status occupations were less frequently found among the white households, even in those owning no slaves, though there was wide variation by county. Only in the very poorer regions of the province, such as Castro, Curitiba, and the mixed farming zone of Cunha was it at all significant. In Cunha and Curitiba over a quarter of the white heads of non-slave-owning households were in these two categories, but this rate was half that among the free colored non-slave-owning heads of households in these same counties, while in the southern county of Castro two thirds of both the free colored and white households were listed as poor and jornaleiros.Evidently the free colored, while found everywhere and participating in all occupations, were nevertheless at the lower end of the social and economic scale in their majority. This was to be expected given their slave origins and ultimately poorer background, more limited education, and lack of capital. But these free colored, however poor, were not alone at the bottom of the social system, as the very significant participation of whites in the poor and day laboring categories indicated. Thus free colored could be found everywhere and in sometimes surprisingly significant numbers in sometimes more elite occupations.But color was not alone in defining occupational distribution. Sex was also an influential factor determining occupational activity of the head of household and this crossed color and slave holding lines. Clearly in both slave- and non-slave-owning households more women were dedicated to crafts than were the men and were far less likely to be engaged in either commerce or the liberal professions. In both free and slave households, among these 15 counties, only 35 percent of the women who ran these households were doing farming (as opposed to 60 percent of all male households) and 27 percent of them were crafts persons (as opposed to just 7 percent of the male households). They were also far more likely to be listed as poor and beggars than the men (29 percent versus 8 percent of all households) and finally less likely to be day laborers than men (1 percent as opposed to 8 percent of male households). They were far more likely to be engaged in the textile related activities of spinning and weaving and this was consistent across color lines and to a lesser extent across the slave ownership boundary.Although color did have some influence over occupation, non-whites tended to follow the same trades as their white male or female compatriots even if at lesser rates and those who owned slaves were more likely to follow the trends among white slave owners than they were to follow the patterns of their fellow free colored who did not own slaves. Color, of course, was influe