Revisiting the Casa-grande: Plantation and Cane-Farming Households in Early Nineteenth-Century Bahia
2004; Duke University Press; Volume: 84; Issue: 4 Linguagem: Inglês
10.1215/00182168-84-4-619
ISSN1527-1900
Autores Tópico(s)Urban Development and Societal Issues
ResumoA few years ago, a book reviewer described Gilberto Freyre as “a favorite straw man” among scholars interested in the history of the family in colonial and nineteenth-century Brazil. That may or may not be a true and fair statement. But, if true, it merely indicates the lasting influence of Freyre’s views and arguments on the historiography. In effect, Freyre’s first major work, published in 1933 with the title Casa-grande & senzala (literally, The Plantation Big House and the Slave Quarters but translated into English as The Masters and the Slaves), and its sequels remain basic points of reference for research not only on the family in Brazil, but also on a whole range of other topics and issues in Brazilian history.1This essay, which also takes Freyre’s work as a point of departure, uses rare early nineteenth-century census materials from a major sugar parish in the Northeastern province (now state) of Bahia to investigate the structure and composition of households belonging to plantation owners and wealthy cane farmers. It examines, in other words, the casa-grande—or planter household— that not only figures in the title of Freyre’s most influential book but also holds a central place in his and other interpretations of Brazil’s past. Indeed, for Freyre, the family structures that would shape Brazilian society from the colonial period into the nineteenth century were to be found within the casa-grande. “The social history of the Big House [casa-grande],” he wrote in one of his best-known passages, “is the intimate history of practically every Brazilian, the history of his domestic and conjugal life under a slaveholding and polygamous patriarchal regime.”2At least as they have been commonly understood in the secondary literature, Freyre’s arguments about the planter family and household closely matched those put forward by Antônio Cândido in his 1951 essay on “the Brazilian family.”3 Scholars have often taken both authors jointly as the major source for what might be called the traditional view of planter households and families or—even more broadly—the traditional view of “the Brazilian family.”4 According to most scholars’ interpretations of that view, the plantation household was dominated by a white male patriarch, the planter, and brought together not only his immediate family—wife and legitimate children—but also his nonwhite mistress or mistresses (either free or, more often, slave women) and their illegitimate offspring. It further included numerous extended kin, free retainers of various sorts, and also, of course, slaves. The patriarchal casa-grande household, in turn, supposedly had its origins in the pattern of large-scale slave-based plantation agriculture that first emerged in the sugar-producing areas of colonial Northeastern Brazil and then later spread to other regions of the country. In short, within this traditional view associated with Freyre, the large, extended, complex, and polygamous patriarchal planter household best typified “the Brazilian family” in the colonial period and during much of the nineteenth century.Since the late 1960s, however, the historical literature on family and household in Brazil has undergone enormous expansion. Much of that expan sion has come from research using late eighteenth- and early nineteenth-century manuscript censuses, or household lists, from various districts in the provinces of São Paulo and Minas Gerais in Southeastern Brazil. The newer census-based research convincingly shows that complex extended households were not the norm in that region. In most districts, on the contrary, a majority of households were organized around a simple nuclear family unit or were headed by a lone individual (in both cases, with or without slaves or free nonkin dependents). It is not surprising, then, that households in those districts tended to be small. Moreover, women (often unmarried mothers) headed a significant share of those households. Drawing on such findings, scholars have challenged the traditional plantation-centered view that the “patriarchal family” characterized colonial and early nineteenth-century Brazil.5Yet, despite all its valuable contributions, the more recent revisionist literature has at least one major weakness: most of the census-based research focuses on São Paulo and Minas Gerais. Comparable studies dealing with Northeastern Brazil are rare; as a result, little at present is known about household structures in a vast and varied region that around 1820 sheltered nearly half of Brazil’s population.6 The lack of comparable studies on the Northeast represents far more than a regional lacuna. In the late eighteenth and early nineteenth centuries, neither São Paulo nor Minas Gerais ranked as well-established and major centers of plantation agriculture.7 That most households in São Paulo and Minas did not match the traditional model of the plantation casa-grande reveals a great deal about regional and social diversity at the time, but it tells us very little about household structures in Brazil’s main plantation regions.8 For that very reason, studies focusing on households in nonplantation areas can only indirectly challenge Freyre’s arguments or, more generally, the traditional view of family and household in Brazil.9The lack of research on household structures in the Northeast stems, in large part, from the extreme scarcity of manuscript censuses for the region. Given that scarcity, the existence of an 1835 census of Santiago do Iguape, a major sugar parish in the province of Bahia, presents a rare opportunity to examine planter households and to revisit, so to speak, Gilberto Freyre’s casa-grande. Obviously, a single census of one parish cannot sustain broad generalizations about household structures in the main sugar-producing areas of Northeastern Brazil during the more than three hundred years between the start of colonial settlement and the abolition of slavery in 1888. Yet, the results of the census of Santiago do Iguape are highly suggestive. In effect, a considerable degree of diversity characterized the households headed by sugar planters and wealthy cane farmers in the parish. But, within that diversity, households that displayed a complex structure and that included large numbers of extended kin and free retainers were the exception, not the rule. At least in this regard, plantation and cane-farming households in Santiago do Iguape resembled households in nonplantation areas of Southeastern Brazil far more than might be expected.10 It is, nevertheless, doubtful whether the findings from the 1835 census are truly incompatible with Freyre’s arguments and whether they instead support revisionist interpretations. The doubts arise in part because conceptual problems sometimes characterize the more recent census-based scholarship and also in part because Freyre was vague in spelling out his views about the composition of the casa-grande household and the patriarchal planter family.The 1835 census of Santiago do Iguape is one of the few surviving results of a failed attempt by the provincial government of Bahia to carry out a provincewide population count. Organized by fogo (hearth, that is, household), the census lists the parish’s inhabitants, both free and slave, by name and provides a range of information, such as birthplace, color (qualidade), age, legal status, and so on, for each resident. The census thus makes it possible to examine both household composition and individual household members.11Santiago do Iguape (or simply Iguape) represents, in many ways, an ideal parish for testing arguments about planter households in Northeastern Brazil. The parish was located on the western side of the Bay of All Saints, roughly 60 km from Salvador, Bahia’s capital, and, hence, in the region known as the Bahian Recôncavo. Like the Zona da Mata in Freyre’s native Pernambuco, the Recôncavo stood out, by the 1830s, as one of the oldest and best-established centers of slave-based plantation agriculture in Brazil. Sugar production in the Recôncavo dated back to the mid-1500s and would remain a mainstay of the region’s economy for at least the next three hundred years. Indeed, the region’s engenhos (sugar plantations with mills for grinding cane) may have supplied as much as one-third of all the sugar exported from Brazil at the start of nineteenth century.12 By then, Iguape (where the first plantations were established as early as the 1580s) had already won a reputation as one of the wealthiest and most productive sugar parishes in the region.13According to the 1835 census, Iguape had 21 operating engenhos, each of which was listed as a separate household. With an average workforce of approximately 123 slaves, those engenhos were among the largest sugar estates in Bahia, and in Brazil as a whole.14 Also living in the parish were more than 80 lavradores de cana (or simply lavradores)—that is, sharecropping cane farmers.15 Lavradores sometimes owned their own farms, but more often they cultivated cane on land rented from a senhor de engenho (engenho owner; i.e., sugar planter). In either case, they surrendered a portion of their crop (generally at least one-half) to a nearby senhor de engenho in exchange for having their cane milled and manufactured into sugar. Stuart Schwartz, in writing about early colonial Brazil, has described cane farmers as “protoplanters,” a group “drawn essentially from the same social origins” as the plantation owners. During the first centuries of settlement, as Schwartz shows, Bahia’s lavradores were predominantly white, often related by kinship or marriage to planters, and generally owned slaves. But, by the early nineteenth century, cane farmers in Bahia and elsewhere in the Northeast had become a far more socially heterogeneous group.16 That was certainly true in Iguape, where, in 1835, free pretos (blacks) and pardos (mulattoes), including five freed slaves, headed 44 percent of all cane-farming households. Likewise, more than one-fifth (23 percent) of all cane farmers in Iguape did not, according to the census, own a single slave. Of those who did employ slave labor, nearly 17 percent owned fewer than five slaves. Nevertheless, the average holding among cane farmers who did own slaves was slightly more than 14. That average reflects the presence in the parish of several well-to-do and even wealthy lavradores.Iguape also displayed the highly concentrated pattern of landholding so often associated with Northeastern sugar-producing districts. Although the census provides almost no direct information on landholding, local records from the 1850s indicate that a few dozen planters and wealthier cane farmers owned nearly all rural property in the parish.17 A high degree of land concentration did not, however, preclude the presence of small farmers; on the contrary, in 1835, more than two hundred such farmers found room in Iguape (generally on land owned by planters) to cultivate food crops such as cassava, beans, and maize for home consumption and sale. A significant number of free artisans, fishermen, and seamstresses, together with their families, also resided in the parish.The 1835 census records a total population of 7,410 inhabitants, more than half of them (54 percent) African- and Brazilian-born slaves. Individuals classified as whites accounted for no more than 8 percent of the parish’s residents; the remainder consisted of freeborn and freed pretos, pardos, and cabras (individuals of mixed preto-pardo ancestry). The census further indicates that the parish’s population was distributed among 966 inhabited fogos. The great majority of those fogos (over 95 percent) belonged to small farmers, fishermen, seamstresses, poorer cane farmers, artisans, and the like, who more often than not were classified as free pardos and pretos in the census. This essay sets aside most of the parish’s fogos to focus on a much smaller group of 37 households headed by senhores de engenho and wealthier cane farmers.The 37 households examined here include 15 of the 21 engenhos in Iguape. When the census takers surveyed the parish, the other six lacked resident proprietors. Among the six were the engenhos Vitória and Buraco, both of which belonged to the recently deceased and never formally married Pedro Ro-drigues Bandeira, whose estate still had not been settled. The owner of the Engenho Acutinga lived elsewhere in the Recôncavo, and the engenhos Caim-bongo and do Desterro were rented to absentee tenants.18 The sixth plantation without a resident proprietor was the Engenho Santa Catarina—one of two sugar estates owned by colonel Domingos Américo da Silva, who lived at his other estate, also located in Iguape. Thus, the 1835 census yields information on the households of 15 resident sugar planters.19The number of slaves these planters employed on their estates varied considerably, from 47 at the Engenho do Meio to 250 at the Engenho da Praia. Nevertheless, of the 15 planters, all but 2 could count on the labor of at least 65 slaves, 5 owned 150 or more, and the average holding was just over 125. These figures fall into perspective when compared with the results of Stuart Schwartz’s analysis of an 1816–17 slave registry from Santo Amaro and São Francisco do Conde, the two most important sugar-producing townships in the Recôncavo. Schwartz shows that the average engenho in those townships employed 65.5 slaves, or roughly half the average for the 15 estates in Iguape. Likewise, while only 15 percent of the engenhos in Santo Amaro and São Francisco do Conde could claim more than 99 slaves, 8, or more than half, of the 15 plantations in Iguape employed 100 or more enslaved workers. We might also compare the 15 with the engenhos in Jaboatão, a wealthy and well-established sugar parish in the Zona da Mata of Pernambuco. In 1857, Jaboatão’s plantations averaged only 50 resident slaves. Thus, by the standards of late colonial and early nineteenth-century Brazil, the 15 engenhos in Iguape ranked as large sugar plantations.20The other 22 households selected for analysis belonged to lavradores, all of whom could claim a resident workforce of at least 15 slaves.21 If we measure wealth by the number of slaves owned, then these 22 cane-farming households ranked among the richest 5 percent of all fogos in Iguape. That the 22 households were wealthy, or at least quite prosperous, becomes even clearer if we compare them with cane farms in Santo Amaro and São Francisco do Conde. In 1816–17, only 14 percent of all slave-owning lavradores in those townships had more than 19 slaves, fewer than 1 percent employed 40 or more, and the average holding was 10.5. By contrast, 18 (82 percent) of the 22 cane farmers examined here owned at least 20 slaves, 5 (23 percent) had 40 or more, and the average stood at 28.6—nearly triple the average in Santo Amaro and São Francisco do Conde and more than enough slaves to operate a small or even medium-sized sugar plantation. That average, in fact, surpasses the 23 slaves owned in 1785 by the typical senhor de engenho in the important sugar-producing district of Campos (Rio de Janeiro).22 Indeed, the 22 included cane farmers with impressively large slaveholdings, such as Francisco d’Amorim Cavalcante and Manoel Estanislau de Almeida, who owned 46 and 64 slaves, respectively. As a group, the 22 lavradores might therefore be equated with the cane-farming “proto-planters” that Schwartz describes for early colonial Bahia.23In brief, the 37 households examined in this essay were all headed by wealthy, or at least quite prosperous, slave-owning agriculturalists in one of the oldest and most important sugar-producing regions of Northeastern Brazil.Perhaps the first question of interest here is who headed those plantation and cane-farming households. Table 1 provides basic demographic information about the 37 household heads and shows that they often shared a number of traits. For example, with the exception of three naturalized Portuguese immigrants, all were native Bahians.24 They also tended to be middle aged (the average age of cane farmers was 45.1, while that of planters was 47.3). Thus they tended to be old enough to have acquired wealth and to have built up their fortunes—whether through inheritance, their own efforts, or some combination of the two. Likewise, more than 80 percent were either married or widowed. By contrast, nearly half (48 percent) of all other household heads in the parish were listed as single.Color also distinguished the senhores de engenho and wealthier cane farmers from the bulk of Iguape’s free population. Although freeborn and freed pretos and pardos made up the majority of all free inhabitants in the parish, all but one of the 37 household heads were white. The sole exception was Fran-cisco Marinho e Aragão, a freeborn pardo cane farmer who owned 15 slaves. Local inheritance records reveal that Marinho e Aragão was the son of Antô-nia Francisca Marinho and Feliciano Rodrigues Godinho. Feliciano Rodrigues Godinho, in turn, was a freed pardo and fairly prosperous slaveholding cane farmer. But the sources fail to reveal how Godinho had gained his freedom and then established himself as a slave-owning cane farmer. He may have been the illegitimate son of a wealthy white male, perhaps a planter.25 In any event, as the examples of Marinho e Aragão and his father demonstrate, slaveholding was not an exclusively “white” privilege in the rural Recôncavo of the early nineteenth century. Free blacks and, far more frequently, free pardos in Iguape and elsewhere in Brazil did, in some cases, own slaves. As slaveholders, they might even achieve a measure of real prosperity.26 Nevertheless, no free black or pardo in the parish came remotely close to matching the wealth of the largest planters, all of whom were listed as white. Marinho e Aragão, with his 15 slaves, was in fact the wealthiest nonwhite slaveholder in Iguape. Clearly, the uppermost ranks of local society in this sugar parish remained largely closed to the free blacks and pardos who accounted for more than 80 percent of the free population at the time.Finally, table 1 also shows that men made up the majority of Iguape’s planters and wealthier cane farmers. Even so, 8, or more than one-fifth, of the 37 households were headed by women. The 8 female-headed households included some of the largest slave-owning fogos in the parish. Although, both in the older historiography and in popular imagination, large rural slave owners are almost always portrayed as male, it is not surprising that women headed a significant share of Iguape’s wealthy agricultural households. Recent research has shown that, in various parts of late eighteenth- and early nineteenth-century Brazil, women headed anywhere between 10 and 45 percent of all households. That research has also revealed that female-headed households tended to be poor, which was also true in Iguape, where one-third of all household heads were women (often unmarried mothers). Roughly 80 percent of those women were free pretas and pardas, most of whom did not possess even a single slave.27 Here, however, we are dealing with female headship within the wealthiest segments of local society. The chief explanation for female-headed plantation and cane-farming households in Iguape almost certainly lies in the inheritance laws. In Brazil, widows generally retained one-half of the couple’s joint property; the other half was divided equally among all surviving sons and daughters. Nevertheless, widows, it would seem, were far more likely than unmarried surviving daughters to find themselves heading a large slave-owning household. Of the eight women discussed here, seven were widows. Only dona Francisca Maria Vitória das Mercês e Aragão, a cane farmer in her mid-thirties, was listed in the census as single.28The households headed by these 8 women and 29 men would, at first view, seem to be implausibly large: the 1835 census enumerates for them a total of 2,701 residents. Thus, although they represented less than 4 percent of all fogos in Iguape, they sheltered 36 percent of the parish’s population. More than two-fifths (46 percent) of the 37 households claimed at least 50 members, while 8 had between 100 and 258 members. The average size was 33 for cane-farming households and just under 133 for plantations. These figures are in some ways misleading, because they result directly from a census-taking convention. In Iguape and elsewhere in late colonial and early nineteenth-century Brazil, census takers listed slaves as members of their masters’ households. Slaves, in fact, accounted for 93 percent of the residents in the 37 fogos.Not surprisingly, if we exclude slaves for the moment and focus on the free component in those fogos, household sizes change dramatically. Indeed, as table 2 shows, the free component was not especially large. None of the 37 contained more than 12 free members. Less than one-sixth had between 9 and 12 free residents, while over half sheltered no more than 5. In fact, five households (including three engenhos) had only one free resident. To be sure, the households of senhores de engenho tended to be larger than those of cane farmers; even so, fully 80 percent of the plantation households had fewer than 9 free residents, and one-third ranged in size from 1 to 5 free members. The result is a fairly modest average of 6.2 free members for planter-headed fogos and an even more modest mean of 4.3 for those belonging to cane farmers. These averages are roughly comparable to those found in various parts of Southeastern Brazil. Donald Ramos, for instance, reports a mean of 4.4 free inhabitants for fogos in five districts in early nineteenth-century Minas Gerais. Likewise, according to Elizabeth Kuznesof, the average rural household in the municipality of São Paulo had 5.6 free members in 1802.29 Thus, while the traditional view (and even some of the more recent revisionist literature) would lead us to expect planter and wealthy lavrador households to be not only large, but also much larger than fogos in nonplantation areas of Southeastern Brazil, the 1835 census fails to bear that out. Moreover, by itself, the information on household size (excluding slaves) also suggests that most planters and wealthy cane farmers in Iguape did not permanently reside with large numbers of extended family members and free retainers.But that does not mean that more distant relatives, retainers, and other free dependents were absent from their households. On the contrary, as table 3 indicates, such individuals numbered 84 and accounted for nearly 45 percent of all free residents in the 37 fogos. Although extended kin figured among the 84 nonnuclear household members, they were not especially numerous. In fact, they represented only 8 percent of the free population in those fogos. Free employees were even less common; in Iguape, the great majority (86 percent) of all free employees on sugar plantations and cane farms maintained their own households.30Table 3 instead shows that the single largest group of nonnuclear members consisted of 44 men, women, and children identified as domésticos (literally, “domestics”); they made up more than one-fifth of the total free population in the engenho and cane-farming households. The term doméstico, which is unusual in late colonial and early nineteenth-century Brazilian censuses, requires some explanation. Although some domésticos may have done housework, they should not be confused with servants; rather, doméstico served as a local synonym for what was more commonly known as an agregado (someone added on or attached to a household). The 1835 census takers in Bahia used both terms, more or less interchangeably, as catchall categories to classify individuals who had become in some way dependent on a household head and hence subject to the head’s domestic authority.31 Domésticos can therefore be equated with the free retainers who, according to traditional interpretations, inhabited, in large numbers, the typical casa-grande.Doing so, however, tells us very little about the men, women, and children whom planters and cane farmers brought into their households. The 44 do-mésticos did, in many cases, share common traits. For instance, with the exception of one widow, all were listed as single. That is perhaps not surprising since, although the elderly could be found among their ranks (including one 90-year-old doméstica), they tended, as a group, to be young and often quite young: over half (55 percent) were under 18, and nearly two-fifths (39 percent) were aged 10 or younger.32 Most were also female (33 females versus 11 males). Furthermore, regardless of age or sex, the majority were classified as pardos and pretos (77 percent and 9 percent, respectively); the 38 nonwhite domésticos included 21 freed slaves.Despite those common traits, it is also clear that not all domésticos held the same status within the household or, more broadly, in local society. Thus, the six white domésticos would have ranked higher in the prevailing racial hierarchies than their pardo and preto counterparts and certainly higher than the 21 freed slaves. The six whites included the 19-year-old Benvinda de Cas-tro, who lived in the plantation household headed by colonel Rodrigo Antônio Falcão. Not only was she listed as white, but also, and perhaps more important, as dona Benvinda de Castro. The census takers, in other words, assigned to Benvinda the same title of respect that they gave to colonel Rodrigo’s wife and, as a rule, to the wives and daughters of the other planters and wealthy cane farmers. They also assigned that title to Maria do Nascimento, a 60-year-old white doméstica at the Engenho Maroim. Likewise, although most domésticos would have had little or no personal wealth, in two instances, domésticos owned their own slaves.The catchall category of doméstico may also hide other differences in status. For example, census takers may have used the term as a discreet way to classify illegitimate children born to planters and cane farmers by their mistresses or concubines. That would not be an unreasonable explanation for the 15 pardo domésticos under the age of 15 found in male-headed plantation and cane-farming households; 9 of the 15 were “libertos” and hence the freed sons and daughters of slave women. But, about such matters, the census allows for little more than speculation.Along the same lines, some parda and preta domésticas may have been concubines, whom the traditional view portrays as a typical feature of the casa-grande household. That possibility deserves consideration because, in surveying poorer households in the parish, the census takers did use the term doméstica to designate what were, in some cases, clearly consensual-union wives. Nevertheless, no more than 7 of the 29 male-headed households contained domésticas who might plausibly be regarded as concubines.33 And even for those 7 households, we cannot, without additional evidence, go beyond speculation or dismiss the possibility that the women in question were simply domésticas—that is, merely agregadas.Indeed, in only two cases does the census, by itself, strongly hint at the presence of a resident concubine. One of those cases involves Diogo Pereira do Lago, a prosperous 60-year-old widowed cane farmer. The census lists Leonor Maria da Piedade, a 45-year-old freed parda and the unmarried mother of three freeborn children, as a doméstica in his household. Among her three children was an 8-year-old son named Diogo. That, of course, does not necessarily prove that the Diogo Pereira do Lago was the father of Diogo and Leonor Maria’s other two children. The coincidence in names does, however, raise suspicions. Moreover, in 1857, the same Diogo, now bearing the surname Pereira do Lago, as well as his brother and his sister, would appear in local notarial records as heirs to their mother’s estate, which included 27 slaves (some of whom had belonged to the elder Diogo Pereira do Lago in 1835), the Fazenda Cassinum with its 124 tarefas (54 hectares) of good cane land, and a two-story house in the village of Santiago do Iguape. All in all, her estate had a net worth of more than 27 contos de réis (Rs.27:262$000, or £3,026 at current exchange rates). If Leonor Maria was not the mistress or consensual-union partner of Diogo Pereira do Lago (senior) and if he had not fathered her children, then it is hard to imagine how she (a freed slave) could have possibly acquired such a fortune and come to own his farm as well as several of his slaves.34The census hints at a similar relationship in the household of José Francisco Moreira, the 34-year-old unmarried owner of the Engenho do Meio. José Francisco resided with Maria Joaquina das Mercês, a 28-year-old freeborn parda and the single mother of four young children. The census does not classify Maria Joaquina as a doméstica; in fact, it fails to provide any information about her relationship with José Francisco as the head of the household. By itself, the lack of information might represent an effort at discretion on the part of either José Francisco, the census takers, or perhaps both. That did not, however, prevent the census takers from recording that one of Maria Joaquina’s sons was named Guilherme Francisco Moreira and that his older brother bore the even more telltale name of José Francisco Moreira.35Thus, it would seem safe to assume that male planters and wealthy male cane farmers in Iguape sometimes did maintain free mistresses within their households. But, contrary to traditional depictions of the polygamous casa-grande household, the census does not suggest that the practice was especially widespread in the parish. Even if we give free reign to speculation, only 7 (24 percent) of the 29 male-headed engenho and cane-farming households included a free concubine. It is also worth noting that the two most likely cases in
Referência(s)