Beyond Masters and Slaves: Subsistence Agriculture as a Survival Strategy in Brazil During the Second Half of the Nineteenth Century

The accumulation of capital has been a preferred framework for choice of topics of study in Brazilian historiography. This tendency is found not only in the emphasis on regional studies that concern principal exporting areas and, later, industrial centers,¹ but also in social history. Studies involving sociohistorical analysis are concerned predominantly with the type of work force which contributed to capital accumulation: slavery until the nineteenth century, immigration during the First Republic, urban workers after 1930.

I do not intend to deny the obvious relevance of these topics of analysis, but only to emphasize that this tendency has led to neglect of the demographic variable as a fundamental factor in the historical study of any society. For example, the second half of the nineteenth century, the period of specific interest here, has been studied preferentially from the viewpoint of agricultural exportation based on slave labor, whereas free and freed people composed 41 percent of the Brazilian population by 1818, a proportion which grew to 84 percent by 1874.² Notwithstanding its numerical significance (comprising a majority by the second half of the past century), the poor free population in slave society has almost always been only marginally considered by scholarly literature. Classic authors such as Caio Prado Júnior and Celso Furtado have already noted the significant demographic weight of free people in Brazilian slaveholding society. Yet, at the same time, they underscored the exclusion of free people from the market economy and their personal dependence on large landowners in such a way as to endow them with only secondary social and historiographic interest.

According to Celso Furtado, free people who arrived in Brazil “as artisans, soldiers, or simply adventurers, in one way or another became dependants of the large landowning class.” They engaged in subsistence agriculture of “very low productivity,” almost always on land belonging to someone else, utilizing “minimal capital” and “the most primitive techniques.” In some cases, however, they would have appreciable demographic importance, as, for example, in Minas Gerais after the decline of mining, where, according to Furtado, a “relatively numerous population” found “space for expansion within a subsistence regime,” and came “to constitute one of the principal demographic nuclei of the country.”³ For Caio Prado, they were “the embryonic form of a middle class between the large landholders and the slaves,” composed of mestizos and “degenerate whites,” dedicated to a “miserable subsistence agriculture,” which became the occupation of nearly 50 percent of the population of the empire.⁴

After reading the above two authors, one can have little doubt that there were many small, rural, free producers in both colony and empire, a number which continued to grow until the advent of the Golden Law. Yet, because poor free people had no clear place in the export plantation system that was held to be “dominant,” they were regarded as irrelevant as a specific topic meriting further study. In the end, the reification of a structure of domination to explain the essence of the social process turns references to this social group into disconnected parenthetical comments, which contribute nothing to a comprehension of the historical developments that one is seeking to analyze.

Until a few years ago, there was one work, Homems livres na ordem escravocrata, by Maria Sylvia de Carvalho Franco,⁵ that undertook an analysis “of the figure of the poor free man in the social system” of nineteenth-century Brazil. I disagree, however, with the classification system adopted by Franco for two basic reasons:

1. Under the rubric “free poor men,” her analysis includes social categories which from my point of view are quite diverse and not always clearly distinguished by “poverty.” For example, tropeiros (muleteers), vendeiros (general store merchants), and sitiantes (small farmers) are grouped with agregados (rural dwellers living on property that belonged to someone else) and camaradas (farmhands).

2. Personal dependence on large landholders is used by Franco more than “poverty” as a unifying characteristic of the free population studied. Such dependency held quite diverse meanings for each social category on the list, and this diversity appears to be at least as important as the traits they had in common.

Referring to the term tropeiro (understood to be an owner and trader of livestock), it is difficult to classify him as a “poor free man” in view of the importance of supplying cattle to fazendas and cities—especially the capital. Muleteers were a significant segment of property owners. Often, in other regional settings devoted to animal husbandry, they influenced and defined political and social control.⁶

Only in the analysis of small farmers does the term “personal dominance” appear to really define the category. To be sure, according to Franco, the ongoing relationship between fazendeiros and these farmers was in many ways one among equals. “Both fazendeiros and sitiantes were landowners, both holding an interest in landed property ownership and living on it, which thus conferred permanence and continuity on their relationship.” This formal equality was solidified in the bond of godparenthood (compadrio), which concealed the principle of personal dominance, manifested in the form of “economic assistance from fazendeiro to sitiante in exchange for political loyalty.”⁷ The sitiantes are characterized as voters and small landholders without being precisely defined in the text as to whether they were also slaveowners or what kind of agriculture they were engaged in. Instead, they are defined mainly in terms of their relationship to the large landholders. Having been recognized as equals by the fazendeiros on whom they depended economically and politically, the sitiantes are not easily clustered with other categories characterized most often by frequent migration, such as farmhands and agregados.

In the case studied by Maria Sylvia Franco, the physical dimensions of the area legally appropriated by the large slaveholders surpassed the area needed for production plus the area set apart for expansion. This, in turn, permitted the survival of the independent peasant (caipira independente) as a “resident on another’s property.” The same author points out, then, that the abundance of land and the spatial mobility of these poor freemen imposed limitations on the personal control exercised by the large landholders. In particular, this situation prevented large landowners from utilizing them in the regular work force “even when the demand for workers in large-scale agriculture became pressing,” thereby preserving the caipira tradicional from being “changed into a free worker.”⁸ Without denying the importance that has been attributed to “favors” in exchange for “services rendered,” as a regulating device in the relations among free men in nineteenth-century Brazilian society, I would argue that the fluidity of these relations needs to be more clearly emphasized, and that the role which land ownership plays as a foundation of personal domination must be challenged.

I have intentionally left for final consideration Franco’s treatment of the vendeiro. This figure is not to be confused with the dealer of the large commercial centers, but is one of the thousands of proprietors of general stores directed toward local consumption. What deserves to be noted is not only that they remained largely independent of the fazendeiros, who were not even their customers, as Franco argues, but, principally, that their presence in the local economy reveals consumption standards and economic activity closely integrated into the world of those “free poor men” who lived at the margin of the dominant social sectors. Although Franco suggests that the vendeiros were a “quasi” dispensable element, their numerical presence suggests that the matter should be analyzed in greater depth.

More recently, some other studies have considered the poor free population in Brazilian slaveholding society as a topic of investigation. However, they approach the topic by turning to the discourse among elites of the period as their main source. This type of information, grounded in the negative stereotype of the “vagrant,” is used almost exclusively as the empirical basis for theorization on socioeconomic dislocation.⁹ For these authors, poor free people were stigmatized as a labor force in an exclusive (slaveocratic) socioeconomic system which, ironically, both denied them better work opportunities and labeled them vagrants. Thus, beginning with the premise that only the accumulation of capital—made possible primarily within the confines of the slave system—conferred historical intelligibility on Brazilian colonial and imperial society, they sentenced anyone who lived outside that axis to a state of limbo destitute of meaning or purpose.

Taking the Paulista case as a reference point, it is still considered that slavery thoroughly degraded the notion of work for former slaves and poor free people, thereby marginalizing them as a labor force in immediate postslavery Brazil. Victims of the system, these people became a “reserve army,”¹⁰ first to the slaves and, later, to the immigrants, temporarily passed over by the agrarian elites as a viable work force in the eyes of the agrarian elites.

In a recent study, I set out to revise this type of interpretation, seeking to consider the poor free person in the second half of the nineteenth century as a socioeconomic agent. Using empirical material that could shed some light on this group, I sought to examine the ways in which it was included (not excluded) in the dominant slaveocratic order. For this purpose, I abandoned any pretense to wider generalization in favor of the case study of a region where there was a significant presence of the free population in the second half of the last century, and where this group could be analyzed through recourse to documentary series such as notarial registers which might more directly reveal local life.¹¹

The regional differentiation of Brazil in the nineteenth century must be a key factor in reevaluating the role of the poor free national in the transition away from slave labor. In the Northeast—at least in Pernambuco¹²—the older settlement of the region, which precluded anything like a moving agricultural frontier, and the hardening of clientelistic relations inherited from the colonial structure, determined the replacement of the slave by the free national long before the definitive extinction of slavery. In the south central region (Minas Gerais, Rio de Janeiro, and São Paulo with the exception of the West), the spectre of an insufficient work force, denunciations of “indolence” and “laziness” of the free and freed national, as well as laws which compelled them to work, all persisted until 1888.¹³ Paradoxically, the free population of these provinces grew during the entire second half of the nineteenth century, with a tendency to concentrate in areas where access to open land was still obtained with relative ease. The availability of free, or only nominally appropriated, lands made alternative survival strategies feasible, and, at the same time, protected poor free people from transformation into a labor force. Nossa Senhora da Lapa de Capivary, a município of the province of Rio de Janeiro, the topic of our study, was one of these areas. In this article, I set forth some of the conclusions from the investigation, and, based on them pose some questions concerning the terms which governed the utilization of both free and freed people in the rural south central region.

The Case of Capivary

At the end of the eighteenth century, the current município of Silva Jardim in the state of Rio de Janeiro formed the parish of Nossa Senhora da Lapa da Capivary in the município of Cabo Frio—an area practically devoid of inhabitants and covered by tropical vegetation.¹⁴ An integral part of the so-called Baixada Fluminense,¹⁵ adjoining the Serra do Mar in the present município of Nova Friburgo, Silva Jardim possesses a broken topography which rises steadily from low altitudes (40 meters) in the direction of the Serra escarpment (1,000 meters). The configuration of the land limits the amount of alluvial soil in the area suitable for the cultivation of sugar cane, which up to the end of the eighteenth century determined the nonindigenous occupation of the lowland (Baixada) region.

The expansion of coffee production in the Fluminense province in the first half of the nineteenth century would radically alter the initial appearance of the countryside. During its expansion period, coffee cultivation initially followed two roads—that of the highlands, extending all the way from Rezende and Vassouras to the Paraíba Valley, and that of the “hot lands,” which began on the slopes covered with sugar cane fields in the eastern Baixada Fluminense at São Gonçalo and Itaboraí, and finally integrated into the slave economy all the part of the lowlands touched by the foothills of the Serra do Mar, i.e., Silva Jardim, Rio Bonito, Cachoeiras de Macacu.¹⁶

The first occupation of the area was based on a concentrated appropriation of great expanses of available land, by producers who could count on a supply, however modest, of capital resources and slaves.¹⁷ In the valley and the lowlands, coffee was planted by slaves on the sides of the hills, on soil originating in decomposed rock formations. Meanwhile, the fields planted in the Paraíba Valley rapidly outdistanced those of the Baixada. It was thus discovered that, for optimum production, coffee should not be planted below altitudes of 400 meters, at least where important economic enterprises were concerned. Coffee from the hotter regions designated capitânia¹⁸ was for the internal market. With the rise in food prices starting in the second half of the nineteenth century,¹⁹ producers around Capivary would diversify to include large-scale manioc flour production.

Both coffee and manioc flour were grown for commercial purposes and were initially sent to neighboring marketplaces,²⁰ from there to be redistributed to the markets of Rio de Janeiro and Niterói, as well as to fazendas engaged in specialized agricultural production in the eastern Paraíba Valley. At the same time, the opening up of new highways and roads, along with the slow expansion of coffee cultivation (by virtue of the limitations of the internal market in comparison with the international market), attracted growing numbers of free people who set up farms and small garden plots on uncultivated lands situated within and without the already legally appropriated lands. By 1856, Capivary, already elevated to the status of município, contained the seventh-smallest concentration of slaves in the province. Of a population numbering 15,584, 38.5 percent were slaves. The normal rule in other municipalities was that the number of slaves equaled or exceeded the free population.²¹

The relatively high numbers of free people in the population, as well as the continued decrease in Capivary’s stock of slaves,²² initially appeared to indicate a precocious deterioration of the local slaveholding structure, as Capivary became a supplier of slaves to more dynamic areas after 1850. According to slave transactions registered in local notarial records for the entire period under study (1850-88), more than half (56.8 percent) corresponded to the sale of slaves destined for areas outside the town.²³ The importance of this participation in the so-called “intraprovincial slave traffic”²⁴ can only be fully appreciated, however, when it is placed within the context of the social stratification of local agricultural producers and patterns of slave ownership existing up to that point. In order to establish the stratification I shall use as a reference the list of agricultural producers’ register in the Almanack Laemmert²⁵ as an indication of the inclusion of local producers in a wider commercial circuit. Those who paid to have their names cited in the publication could not have been the producers dedicated to subsistence agriculture.

A second look at notarial records concerning the buying and selling of slaves reveals that of the total number of slaveowners listed as sellers to destinations outside the municipality, 63.2 percent were slaveowners not appearing in the classified lists in Laemmert. In the local transactions, 74.0 percent of the sellers do not appear in Laemmert, 12.3 percent of them being residents of adjoining towns (Araruama, Rio Bonito) which also appear among the areas that purchased slaves. At the same time, more than half of the buyers involved in transactions inside the municipality (58.4 percent) are classified in the almanac as coffee fazendeiros and “larger” (“mais fortes”) producers (see Table I).

The slave labor force initially appears to be widely spread throughout the local socioeconomic structure. Until at least the mid-1860s, the presence of one or two slaves is registered even in the smallest units of production. Nevertheless, many of these enterprises do not seem to have assumed strictly commercial status, being limited to the exchange of a small, unpredictable surplus at the local marketplace for a few consumer goods. And, with time, the increase in slave prices tended to concentrate ownership in the local agricultural society.²⁶

Taking as our source the postmortem inventories of property of deceased persons found in the local Segundo Ofício de Notas do Cartório, we can describe the social stratification more precisely.²⁷ Of the 230 documents under consideration, 99 described agricultural enterprises whose proprietors appeared in the classified lists of the Almanack Laemmert as “fazendeiros, lavradores e propietários mais fortes.” All the individuals listed in the inventories as owning more than ten slaves also appeared in the Almanack. I further subdivided the agricultural proprietors as owners of (1) fazendas and (2) “commercial farms” (sitios comerciais). Both were market oriented and owned slaves, but they were nevertheless differentiated in regard to their position within the socioeconomic hierarchy of the community.

In order to understand exactly what is meant by “fazenda” within the local agrarian setting, we must first take into account that various agricultural enterprises were given this classification in more than one of the sources analyzed, their owners being known as fazendeiros. The identification by name of some of these owners made it possible to select a sample of inventories of fazendeiros. The sample is too small for quantitative analysis, but it suggests some common characteristics capable of providing coherence to the analysis that follows.

The first point of fundamental importance that identified members of the sample is the number of slaves they owned—all had more than 20 in the inventories, in addition to other common traits in such matters as standard of living, amount of wealth, and size of producing units. On the other hand, many owners of relatively small amounts of land, but with numbers of slaves in excess of 15 showed similar characteristics; productive units possessing between 15 and 20 formed a transition group between fazendas and commercial farms. With the social concentration of slave ownership which occurred in the region in the second half of the nineteenth century, that transitional group tended to disappear, and those who owned that number of slaves came to rank as more and more “prosperous” in relative terms. Therefore, all productive agricultural units with more than 15 declared slaves (42 documents) will be considered fazendas. Likewise, agricultural enterprises with less than 15 slaves whose owners were nevertheless listed in the Almanack Laemmert will be considered commercial farms (57 documents).

When we use the term “fazendeiros,” we are very far from any real resemblance to the large coffee planters of the Paraíba Valley. If, in coffee cultivation generally, the scale of producing units was already much less in terms of land, slaves, and other comparisons with the model of the large-scale sugar plantation of the Northeast, the “rich” producers of Capivary could assume pretensions as large producers only in relation to their small local universe. If we take into consideration the parameters normally used to characterize the large landholdings in the Paraíba Valley, it is evident that in Capivary we are dealing with small units of production. References for Paraíba to more than 200 alqueires of land, 100,000 coffee plants, or 60 slaves²⁸ have no resonance for the area under examination here. In concrete terms, many of the agricultural enterprises described in this study actually are closer in many aspects to the productive units of sitiantes and agregados described by other authors for regions where the cultivation of coffee was synonymous with prosperity.²⁹

However, it appears that the key element for comprehending the dynamics of the community and its relationship to the surrounding society is not to be found in the actual productive capacity of the agricultural enterprises. The political and economic importance of “localism” in the Brazilian society of the period determined the composition of local or regional hierarchies in the configuration of political power and forms of social control. “Small large-scale farmers” became fazendeiros in Capivary simply by being the wealthiest group of local slaveholding producers, thereby giving credence to their claim of political and social control over the municipality.³⁰

In spite of the low income generated by their enterprises in comparison with other producing areas, these farmers succeeded in resupplying their establishments with slave labor until the 1880s, partly to the detriment of the other local producers. From postmortem inventories, we see that these slaveholders maintained an average of 20 to 30 slaves during the entire period, with at least 50 percent of them being of productive age.³¹ Acquisitions of slaves, even though restricted to the regional market, and the natural increase in their numbers appear to have allowed the owners to maintain, if not to expand, the functioning of the larger units of production. Likewise, they were able not only to replace the older coffee plantings on a regular basis but to diversify as well, increasing considerably their production of such commodities as manioc flour, in response to the growing demand for primary food products (see Tables II and III).

The persistence of slave labor on local plantations until the eve of the definitive extinction of slavery indicates the importance of slave labor in the commercial production of foodstuffs in this period. It also gives evidence of the possible existence of alternative strategies for survival for the rest of the population, since, in addition to the fazendeiros, there was a largely mestizo segment of the Capivary free population which only with difficulty could be converted into a suitable work force for commercial agriculture.

This difficulty was especially felt by those producers whose mode of coffee and manioc flour production for the local and regional markets included the use of slave labor, but who were economically dependent on the larger local producers and were without sufficient financial resources to replace their slaves as needed. As a general rule, local farmers who used slave labor did not have enough disposable capital to initiate their own agricultural enterprise. In numerous cases, they depended on financing by fazendeiros of the region, in exchange for political loyalty. Until the 1870s, more than 70 percent of the negotiated mortgage and loan notes in the local registry consisted of advances provided without interest by the planters to small producers attempting to establish themselves.³² The terms almost always were counted in years, and were frequently deferred. Consequently, we are not dealing here with a simple diversification of investment on the part of fazendeiros with capital at their disposal, since the money invested did not produce any profit. Political loyalties became interwoven in a web of economic dependence among the voting citizens of the municipality.³³ Moreover, as small slaveowning producers for the regional market, local farmers were inhibited by low levels of capital accumulation from fully replenishing their supplies of slaves—something which they had been able to do in the ’50s.³⁴ As a consequence, they were not able to respond to the growing demand for foodstuffs at higher prices, nor were they able to effectively replace their plantings of coffee (see Tables IV and V).

Thus, we have an idea of the dimensions of the internal market of the province of Rio de Janeiro during the second half of the nineteenth century. It was able to both absorb the coffee production of the Baixada (though the latter was at a disadvantage vis-à-vis the Paraíba Valley in export terms) and stimulate some enlargement of productive structure through the growth in demand for primary food products in the period which followed the extinction of the slave traffic. Nevertheless, the low level of capitalization and profitability of local production meant that its expansion did not lead to a situation of land monopoly, nor were producers able to attract free workers by other means. The local fazendeiros were aware of this situation, so that by the mid-’70s they preferred guarantees of liquidity to new investment in the productive or political (loan-based) structure which they controlled. In the decade of the ’80s, more than 50 percent of the fazendeiros’ wealth was invested in public debt instruments.³⁵ Between 1872 and 1890, furthermore, the white population of Capivary decreased by 6 percent, reflecting a significant exodus of local planters, who tended to sell off the lands they controlled in small plots. Drastically compromised in their ability to accumulate capital, the small farmers in this area came close to approximating the standards of subsistence agriculture. The social stratification of the município as it was organized until the close of the ’80s simply could not subsist without slave labor. At the same time, the black and mestizo populations grew by more than 50 percent during the period, thereby reaffirming the attraction that the possibility of being an independent worker held for former slaves and poor people of the region.³⁶

Subsistence Agriculture as a Strategy for Survival in Capivary

According to the general census of 1872, almost one-half (48.0 percent) of the free population of the town was composed of persons without a profession or dedicated to domestic service. These numbers probably refer to women and children. The remaining one-half, which I shall call the active free population for this analysis, was composed 68.1 percent of lavradores (farmers). This reinforces the argument that the possibilities opened up by available land around the municipality produced a population of “poor farmers” not always in legal ownership of the land they were exploiting. This meaning of the term lavrador, associated with independent agricultural labor, is reinforced in the same census where the slave population of two essentially rural parishes is not found classified by profession as lavradors (a term reserved for the free population), but primarily as “servants and day laborers.”³⁷ The concept of the term lavrador held by the local census takers made it difficult for them to consider the inclusion in that category of field slaves, who were undoubtedly the predominant group in the local slave population.

Nevertheless, a significant number (1,514, or 21.9 percent of those whom I would designate as an active free population) are also classified as “servants and day laborers.” Probably these are a contingent of the population which could not gain its subsistence through independent agricultural work, and thus received a classification suggestively similar to that of the slave population (see Table VI).

In relation to urban occupations, only in the town of Capivary, seat of the município, does one find an elaborated socioprofessional diversification. The other nuclei (Correntezas and Gaviões) register only the presence of clergy (whom they needed in their condition of parish and curacy), notaries (who were responsible for the local registries created in the previous decade), pharmacists, teachers, and merchants, all in numerical inferiority to the municipal seat. In all cases, the parish of Nossa Senhora da Lapa de Capivary contained more than half of the persons dedicated to these occupations. The same disproportionate ratio can be found among seamstresses, carpenters, masons, and cobblers. This small network represents virtually the entire socioprofessional diversification to which the division between agricultural work and artisans in the municipality gave rise, as reflected in the census under study.

In the criminal records found in the Segundo Ofício de Notas of the município for the period beginning in 1840 and ending in 1890, are 932 witnesses from the free population resident in Capivary.³⁸ I have used this sample of the free population in the period under analysis as a source for a slightly more in-depth study of the means of support employed by residents of the community. These testimonies are classified according to occupation, sex, and place of birth in Table VII.

The small proportion of persons not classified in terms of occupation (10.3 percent) stands out, on first observation, as does the presence of a majority of men in the total assortment of recorded witnesses, which facilitates a more detailed analysis of the possible types of jobs held by the active local population. In addition, we can identify the noticeable presence, among the residents of the municipality recorded in the documents, of persons born in other towns or regions. The latter detail gives evidence of the extreme spatial mobility of the free population in that area during the period under analysis (50.2 percent), as well as of the municipality’s character as an economic and demographic frontier.

In considering the group of those involved in nonagricultural occupations, detailed presentation cannot go much beyond the information offered in the professional census of 1872. Public employees can be categorized more concretely as postal workers, legal staff, and jailers, among others. In addition to priests, the church employed sacristans. To the specialized occupations listed in the 1872 census can be added such occupations as “saddle maker,” “blacksmith,” “baker,” “coachman,” “cabinet maker,” as well as “horsebreaker,” which, in limited numbers, made up occupations necessary for the life of the community. There appears to be a significant number of commercial houses which did not advertise in the Almanak Laemmert, which indicates that they were involved strictly with the local market. At the same time, it is interesting to note that of the 33 persons classified as clerks (i.e., commercial employees), 29 were between the ages of 10 and 30. Likewise, in addition to ordinary tailors and seamstresses, we find 2 master tailors (officiais de afaiate), both under the age of 30. Starting in the second half of the ’70s, railroad construction in the município created new employment opportunities for mechanics, foremen, and manual workers—the latter (only 2) both under the age of 20. Unskilled employment and youth appear to have had some kind of connection. Of the 28 agricultural day laborers, 20 (71.42 percent) were unmarried and were less than 30 years old. Thus, it is plausible that this very same population of young people, often still linked to the productive units of their parents—who complemented the familial income with occasional salaried work in business establishments and commercial farms—could have made up the greater part of the category “servants and day laborers” in the 1872 census (see Tables VII and VIII).

Out of a total of 484 testimonies of individuals dedicated to agricultural activities who are absent in the Almanak Laemmert, we find 61 persons listed as black or mulattos (pardos). Twenty-four of them were identified as lavradores, 5 of them free Africans; 28, including 1 free African, were “living from farm work”; 4 were called former slaves and “field workers”; 2 were “day laborers,” and 3 “brushwood workers” (trabalhadores de mata). Taking color as an indicator of social position, these data allow us to perceive, on the basis of 52 of the 61 cases, the access gained by freed people and their descendants to independent agricultural production, even if on land belonging to someone else (as specified in two cases). This group does not necessarily represent all blacks or mulattos found in the sample: often no classification based on color was given in the documents.

In order to analyze the conditions of production and consumption of the agricultural poor in Capivary, we will consider, from the group of postmortem inventories, those referring to individuals who did not appear in the lists found in Laemmert and whose production methods did not suggest any large-scale commercialization of agricultural products (mills moved by water or steam).³⁹ The type of agricultural production reflected in the 79 inventories appeared as one of the principal elements of homogeneity, while the range of crops produced in these units did not differ greatly from those establishments classified as slaveholding and commercial: coffee, manioc, fruit orchards, and a selection of small periodic harvest crops, principally corn and beans. Coffee was planted by 70 of the 79 inventoried, who, however, never exceeded the small number of 2,000 coffee plants per agricultural unit, nor did they present any tendency toward increasing or decreasing that level of production, which could easily be tended by one person. At the same time, the descriptions of the farms themselves rarely gave the age of the coffee plantings, in contrast to the group of inventories whose title holders appeared also in the lists of producers in Laemmert. There was also no mention of any types of processing equipment in these establishments.

Manioc, although not a permanent crop, could remain in the ground for two years after maturation without being harvested, thus allowing for later storage, literally in the earth itself. It was regularly harvested according to the producers’ needs. One can find in any of the inventories analyzed approximately 400 plants (equivalent to ten bags of flour) “stored in the soil” per agricultural establishment, over and above any harvested manioc or sacks of prepared flour. According to data compiled by Schmidt,⁴⁰ a harvest such as this might be completed by one person in just 12 days, which indicates what this reserve could represent in commercial terms.

Yet, 53.1 percent of these farm units were not equipped for simple production of so-called dry flour or commercial flour, which, refined or crude, was sold widely in regional markets.⁴¹ Found uniformly throughout the registry for the entire period was the mention of one or two burros used for transport of foodstuffs and of small animals for domestic consumption (pigs and chickens), or installations for the raising of such animals (chicken coops and hog houses).

The analysis of documents of deceased persons enables us also to perceive something about the living conditions of the group under study, as well as about the role of commercial relations in their social reproduction. For example, Dona Maria Joaquina de Oliveira, deceased in 1877, lived with her husband and two small children in a thatched-roof house with a living room and two bedrooms. Annexed to the house was a small building, also covered with a thatched roof. Furniture was limited to a “well-worn sofa,” a table also “well used,” “an old bench,” two benches “already very worn,” and an “old bathtub,” valued at 25 milréis.⁴² In the cases studied, the homes never had wooden floors, the ground being covered instead with straw or, more rarely, with tile. The only items of furniture entered on the lists were those objects which could have some monetary worth. Furniture or wooden objects were systematically valued, revealing possessions of extremely limited nature, such as the example of the home of Dona Maria Joaquina mentioned above. Notable in the lists by their almost total absence are pots and pans, dishware, cups, tableware, and other domestic utensils. From a comparison with inventories of more well-off farmers, which systematically mention iron pots and pans, metal tableware, and dishware in general (almost always with comparatively high appraisals), it can be deduced that the household objects of this group were, in the majority of cases, domestically produced, almost always of clay, and had no commercial value worth mention in the reports of inventoried goods.

The individuals featured in these documents would acquire sparse but regular amounts of some products sold in the local general stores. Mention of these products was often accompanied by numerous current accounts annexed to the debt entries of their inventories. During 1864-66, João Rodrigues de Figueiredo⁴³ acquired with some regularity from the store of the firm Braz Paulo Muniz e Cia. pieces of cloth, soap, bacon, salt, tobacco, and small amounts of wine. He asked the firm directly for money at least two times each year, for a total amount of 85$260. During that time he further acquired a hoe, a scythe, three pairs of socks, a pair of trousers, a shirt, and a straw hat. Payments were made with coffee sacks full of coconut or manioc flour, for a total of 147$010. In 1869, the balance in favor of the merchant totaled 188$965, 16.18 percent of the total value of the inventoried goods (1:167$500). This type of indebtedness rarely totaled more than 20 percent of the total value of an inventory and would be paid annually, partly in kind. Moreover, the small merchant appears to have been a true “financier” of this type of production. Many times his role resembled that which one expects to see played by larger landowners or important neighbors.

Stores were found scattered in almost all areas of the município, on the side of roads, within the fields of plantations, and in villages. They almost always carried on business in small houses “covered with tile” and rarely having wooden floors, with “business goods” valued at between 100 and 300 milréis. They sold practically everything needed in a rural world that was in fact profoundly specialized in production of a few basic products. Listing of the small and diverse stock of goods of those businesses, which oscillated between 700 and 2,500 milréis in value in the cases studied, almost always began with the dry goods on hand, which were one of the things most in demand. Pieces of cotton, calico, sailcloth, among others, summed up, as a rule, almost one-half of the total value of their listed merchandise. More rarely, the large stores sold ready-to-wear clothing, such as coats, shirts, and cloth hats. From the analysis of the store goods inventoried, one can see the common practice of making clothing domestically from cloth easily found at any small roadside store. Such stores also provided the necessary tools for seamstress work, such as trim, lace, silk ribbons, hooks and clamps, needles, thread, and pins. In addition, they offered the entire gamut of useful objects, principally those destined for production and transport of agricultural products, such as hoes, scythes, hatchets, and cutting tools. They did not lack stationery articles, nor articles of adornment, such as earrings, small pieces of gold, pens, pencils, or paper. For foodstuffs, the establishments carried dried beef, salt, olive oil, jars of oil, wines and cachaça, cans of sweets, or fish acquired from local fishermen. The list is practically infinite and impossible to reproduce in all its unimaginable details. In general terms, these stocks could be classified as “articles of clothing and farm equipment,” “groceries and cupboard supplies,” and “hardware,” in addition to coffee and manioc flour, which were almost always taken in as payment from the farmers of the area. Representing the second most valuable category of goods in stock, coffee and manioc flour tied up almost one-third of the capital in the lists.

The mere listing of goods bought and sold by the general stores goes far in clarifying the commercial functions of these stores in the rural areas. At the same time as they sought to cover almost all the local consumer necessities, the minimal level of monetarization prevailing in that rural society meant that their cash reserves came essentially from their commercialization of the commodities that they received in exchange for their merchandise—and whose production they in large part financed.

The numerous current accounts found in the small inventories of the general stores attest to their significant intermediary role between local subsistence agriculture and regional markets. In these accounts there is not one case in which payment was listed in the form of money given by the debtors; on the contrary, cash was furnished by merchants to their clients with greater frequency than any of the consumer goods habitually purchased (such as salt, sugar, cloth, etc.). At the same time, these businesses did not perform the same functions as the wholesalers of the regional commercial centers with whom the fazendeiros and more well-off farmers maintained regular commercial relations. Their accounts receivable mentioned amounts of coffee and manioc flour that only for the group of merchants as a whole represented appreciable quantities, and which more than likely they commercialized in the regional centers exactly as the large-scale farmers did with their own products. Moreover, the accounts reveal an impressive level of specialization in this type of “subsistence” agricultural production, in that local residents made payment in coffee and manioc flour even for beans and corn. Although the merchants also dealt with the wealthier farmers of the region (who without abandoning the use of credit were also able to pay in currency), the importance of the commercialization of this subsistence-type small agricultural production is clearly revealed in the accounts receivable of the business proprietors, which were never entirely paid off and in effect guaranteed them a steady supply of agricultural produce to be sold in the larger commercial centers.

Only 11 inventories referring to proprietors of general stores were found and analyzed in detail for this initial study. This represents a small but significant sample by virtue of its uniformity in respect to the items considered up to now. Moreover, the current accounts appearing in these inventories represent almost one-half of the small landholding farmers under study. On the other hand, if the stores, their stocks, and functions appear to be uniform in nature, the same cannot be said about their proprietors. (Nor does the group of owners of general stores whom I was able to identify through the inventories and complementary sources correspond to the retail merchant, as that social type is normally understood.) Of the 11 store owners whose cases were analyzed, 5 appear in the lists published by the Almanak Laemmert. While these showed in their business larger stocks and more capital than the others, their “sales” did not appear qualitatively different from the group as a whole. They managed, however, to accumulate larger stores of goods to be commercialized in the regional marketplaces. Aiming at the wholesale merchants more than the local consumer, they made themselves known in the Almanak. Yet of the total number of merchants whose names appear in the criminal testimonies, only 20 percent (14 names) are listed in the Almanak. It is thus clear that the network of general stores in the municipality was considerably greater than could be perceived by reading that one source.

The retailers in a strict sense, i.e., the individuals supporting themselves exclusively by income from their business, as a rule had few or no slaves. In addition, they had in the total of their entries of supplies, installations, and debts collectible more than 40 percent of their liquid assets. In general, though, they owned small tracts of land, and they never gave up cultivation of subsistence crops. Eventually, the breadth of and profits from their business transactions might allow them to diversify their investments, buying land and slaves, and thereby become wealthy farmers in the regional context. Francisco Xavier Espínola, proprietor of a general store for several years in the parish of Correntezas—a business large enough to justify an entry for him in the Almanak Laemmert as a general store proprietor ever since 1850—at his death in 1883 left 27 slaves and one of the largest agricultural properties recorded in the documentation. According to all indications, he formed this from the profits of his business, in which he continued to work until the time of his death.⁴⁴

There are also cases, however, of planters who diversified their investments by opening their own businesses. In these cases, the agricultural properties represented more than 75 percent of the inventoried wealth. Two of the inventories are typical examples of this possibility, which illustrates the fundamental role in the concentration of capital and financing of local agricultural production played by a particular social type, the fazendeiro-merchant.⁴⁵ Moreover, these examples give evidence of the absence of strict barriers between farming and commerce in the region. From the small retailer with a business located on someone else’s land to the large fazendeiro-merchant, the shop always served a double commercial function: to answer the consumption needs of local farmers, especially those who did not have access to the region’s wholesale markets, and to provide an outlet for the surplus of the município’s subsistence farmers in those larger markets, just as was done directly by the more well-off farmers in regard to their own production.

This analysis still has not taken into account, however, the basic factor which has generally defined the frontiers between rich and poor, socially and economically, in nineteenth-century agrarian Brazil—the ownership of land. Neither has it taken into consideration perhaps the major lacuna extant in the studies and analyses of the so-called poor free people in slaveowning Brazil—their access to and relations with slave property.

Francisco Pinto Pereira, deceased in 1869,⁴⁶ left the following goods to be divided among six sons: “a slave by the name of Felício, 15 years old, field work, valued at 1:300$000; one bed frame without value, a leather basket worth 1$000, and a wooden-framed house covered with straw, on land owned by Sr. João Gomez da Cunha,” value illegible. There is no doubt concerning the “poverty” of Francisco Pinto Pereira, which appears to have been shared with Felício, his only slave and the most valuable of his belongings. There are many Franciscos and many Felícios among the 79 inventories of poor farmers. Until 1875, only 23 percent of the cases studied did not register ownership of slaves. This index increases to about 40 percent after the second half of the 70s, consonant with the process of regional and social concentration of ownership of slaves throughout the course of the extinction of slavery. It is assumed, of course, that slave labor was the basis for “wealth” in Brazilian society of the last century. Equivalent to capital, the surplus labor extracted from the slaves comprised the basic foundation of all the process of valorization of agricultural production destined for export or for regional markets. What was the significance, then, of the utilization of that source of labor by farmers whom we could consider to be both empirically and conceptually outside this economic circuit?

In the first place, one cannot forget that the number of slave workers associated with the effort to participate in the regional commodities market constituted a basic parameter for our classification of the empirical data gathered from inventory lists. This approach proved highly effective in analyzing the commercial and slaveholding production of Capivary. It is not a question, however, of denying the role of slave ownership in the socioeconomic ranking of the propertied groups in the Rio agricultural areas during the period under study, and its influence on their real possibilities for investment expansion and accumulation. It is just that, before the labor crisis brought about by extinction of the slave trade began to produce its full effects, slave ownership was diffused well beyond the frontiers of commercial agriculture, export oriented or not. Several other recent studies have pointed in the same direction.⁴⁷

Moreover, one cannot forget the very low levels of productivity which characterized the utilization of slave labor by the farmers analyzed up to this point. Until 1875, the 32 small farms (situações) which counted on the use of such labor possessed, on the average, three slaves per unit, of which, as a rule, only one was between the ages of 15 and 40. The presence of one female slave with minor-aged children was very frequent, as was that of aged slaves, very often infirm. After the mid-1870s, the average number of slaves per productive unit was less than two, many of them children or aged.

The opposite tendency seemed to operate with regard to ownership of land by the group of farmers studied. Francisco Pinto was the owner of one slave and of a situação on land belonging to João Gomes da Cunha. In much the same way, almost 60 percent of the cases analyzed up to 1875 refer to farmers on land not belonging to them, or, more rarely, terras devolutas (technically vacant government lands). However, this extremely high ratio begins to decline noticeably (to 33.9 percent) after that date.

The legal concentration of land ownership did not eliminate the pressure for unsettled land in the interior of the properties, which constantly attracted new poor farmers. In Capivary, in particular, where commercial agricultural expansion was not at the same pace as in the higher lands, this pressure became even more apparent, though at the same time the tensions between landowners and squatters were less explosive. The socioeconomic significance of legally owned property in the region during the period studied is thus couched in profound ambiguities. On one hand, the desire for a legal monopoly of land with a view to extending the scope of commercial farming operations was, as a rule, outside the range of possibilities for poor farmers. Yet, they did not have major difficulties establishing themselves in areas not being used by the more well-off farmers. Most of the time, these areas were inside the large landed estates, within which the expansion of coffee cultivation was taking place very slowly. The progress of such small-scale farming in the municipality tended, in truth, to limit the socioeconomic power derived from large landed property. The possession of farm plots and improvements on lands either owned by someone else or devolutas was legally recognized by the local judicial system, not only for purposes of inheritance, but also for registering commercial transactions in the notarial records. The squatter was transformed into a typical figure of the rural area under consideration, without being necessarily integrated into the exact type of subsistence economy that I have described. At the same time, not infrequently the poor farmer was actually a landowner, even though his lands were in the majority of cases of very low value. In a still slaveholding agrarian structure, where commercial agriculture was expanding only slowly, land ownership turns out to be an unsuitable parameter for separating wealth from poverty, or for defining in absolute terms the levels of social hierarchy.

What can be concluded from the above is that the so-called “poor free people,” an imprecise intermediate stratum of slaveowning Brazilian society, are difficult to define as “dispossessed” from a structural perspective. If they existed, and clearly they did, it becomes necessary to determine just what is meant by the term “poverty” in that society. Wholly different from the figure of the “slaveowning latifundiarist exporter,” they would comprise, as a category, almost all of Brazilian society outside the so-called “dynamic centers” identified by Celso Furtado. My own analysis is modest and localized, but it refers to a município where the free population was extremely significant during the entire nineteenth century. It appears that, at least in Capivary, an entire category of poor farmers became owners of small extensions of land or even of some slaves, and sought to protect their own survival space against the agricultural expansion proceeding under the aegis of commercial capital. Poor but not “dispossessed” even when they owned no land or slaves, they managed to gain their subsistence as a result of their labors. Hence, they were fundamentally different from the slave, who, first and foremost, worked someone else’s farm. They solved the problem of land scarcity by simply moving to another area; and the abundance of unoccupied land explains their significant presence in the município under study.

Conclusion

Precisely because of its local and monographic nature, the analysis of Capivary permits us to question a series of ideas concerning the poor free person in the era of slave labor, all of which have been established without very much concern for a consistent empirical foundation.

The increasingly backwoods nature (caipirização) and decapitalization of Capivary agriculture in the postslavery period illustrate quite clearly the difficulty of converting the freeborn poor and the former slaves into a source of labor for commercial agriculture. We can further discern something about their alternatives for survival. Thus, these “vagrants” comprised a group of small producers dedicated to subsistence agriculture, who, until the middle of the last century, could even count on one or two slaves. The surplus of their production, exchanged for goods in kind at the local general stores, could thereby reach the commodities markets (urban or not), complementing the production capacity of a slaveowning commercial agriculture dedicated to internal consumption and noticeably undercapitalized as compared to agroexport production. Their contribution to internal provisioning illustrates magnificently how little sense it makes to attempt to classify them according to their “functional” or “dysfunctional” role, and their integration or exclusion in relation to a logically organized slaveholding system. Not only the “indolence” of the free population, but also the general stores that made it possible, were the target of anger of those who saw them as a potential work force.⁴⁸

The case of Capivary permits one to question to what extent the models of land monopolization and coronelista dominance as they existed under the republic can be applied to the forms of local bossism during the period of slavery. Until quite late, agregados and occupiers of government lands were barely differentiated in terms of any type of normative obligation from legal landowners. This was the case in such legal proceedings as deeds of purchase, sale or mortgaging of improvements, or the drawing up of inventories of property of the deceased. But, apart from its extremely low level of capitalization, would Capivary have been an atypical or exceptional case in the agrarian world of south central Brazil? In many aspects it does not seem to have been, for other studies have been pointing in the same direction. In Campos de Goitacases, also in the province of Rio de Janeiro, Sheila Faria makes clear that large-scale sugar production was maintained overwhelmingly by slavery, even to the point of acquiring slaves from the provinces of the Northeast, while at the same time family labor was widespread among poor producers and squatters. More than this, immediately after abolition and throughout the decade of the ’90s there occurred a crisis without precedent in the region, which reduced to practically nothing the sugar shipments from Campos to Rio de Janeiro—a crisis that observers of the period said was directly due to labor shortage.⁴⁹

Research by Roberto Martins affirms the persistence of slavery in Minas Gerais and the purchase of slaves in noncoffee-producing areas of the province during all of the nineteenth century.⁵⁰ I disagree with his final conclusion when he attempts to describe a “slave-owning economy of Minas oriented toward internal consumption.” I believe that the highly aggregated nature of his analysis, covering the entire province, precluded access to local sources, especially notary archives which would have given evidence of a greater division of labor and of the accumulation of capital in the large slaveowning farms of the province. From my point of view, that accumulation is what made possible the new importation of slaves.⁵¹ The research is, however, extremely well substantiated in its treatment of empirical evidence, not leaving the smallest amount of doubt about its fundamental finding that slavery and the net importation of slaves were maintained in the noncoffee-growing regions of the province throughout the nineteenth century. Moreover, I cannot fail to agree with the author that the persistence of slavery in the province and the existence of a numerous free peasantry are different sides of the same coin.

Minas Gerais, the major slaveholding province in Brazil at the end of the nineteenth century (approximately 300,000 slaves in 1872) had, at the same time, a population of 1,700,000 free persons, more than half of whom were blacks and mestizos.⁵² The abundance of free lands or lands appropriated de jure but not de facto in the province guaranteed them the necessary resources for their existence. According to Martins, a large part of the province remained completely unsettled and was comprised of lands which were either abandoned or only nominally appropriated. In 1845, it was estimated that 45 percent of the area of the province was effectively occupied.⁵³

The English involved in mining in Minas Gerais during the nineteenth century felt very close to the problem. One of them, James Wells, observed that the free Brazilian, on being contracted for work, was an excellent worker but “the difficulty lies in inducing [him] to accept, for [he] would not work for a wage if it were not for being compelled by a desire for some small thing for himself or his family. On the contrary, he swings in his hammock, smokes his cigarette, and plays on his guitar or sleeps, saying that he is very busy and maybe he will come, God willing, next week or the week after.”⁵⁴

I do not intend to trace an idyllic picture of the life of these small producers, for in all senses of the word their existence was precarious, and, moreover, they could easily be expelled from their plots of land. The ineffectiveness of land monopolization in day-to-day terms never inhibited the expansion of commercial agricultural interests. In public lands, or in the interior of the private properties themselves, the extensive growth of large-scale slaveholding production forcefully imposed itself on the shacks and any cleared land it might find in its path. But it is important to note that only at the very moment when, forced off the land, the rural poor took to the highways or forests, did they become truly “dispossessed.” And they easily reproduced their former situation. Moreover, the structural precariousness of their access to land did not compromise the relative stability of their existence, for outside of the more dynamic areas of agroexportation, the economic frontier was expanding at a very slow pace.

Based on all these cases, one can state that the ease of access to alternative survival strategies, made possible by the abundance of land, determined the persistence of slavery in the Southeast, and not the other way around. I am not minimizing the impact of cultural and racial prejudice on the thinking of the period, and on the creation of societies for the promotion of immigration. But the practical option of subsidized immigration on the Paulista western frontier was based primarily on the concrete difficulties of obtaining a regular influx of a disciplined, national free work force in an area which had remained, until recently, for the most part forested and sparsely populated but had suddenly accelerated its demand for labor.

In the Legislative Assembly of São Paulo, arguments concerning the “laziness,” “indolence,” and “undisciplined nature” of the free national population are the same, whether used by the proponents of immigration or by those who favored utilization of free native Brazilians by means of legal incentives and/or coercion.⁵⁵ Many times, the division of opinion between the two options reflected regional differences within the province itself over the most effective substitute for slave labor. In areas where the demand for slave labor had stabilized, the use of free Brazilians in spite of all difficulties, was a more feasible solution than resorting to immigrant labor, in view of the competition for immigrants with western São Paulo. On the pioneer frontier, however, to utilize the free nationals as the basis of a work force would mean to “colonize” them. Some isolated voices became proponents of using for that very purpose the funds appropriated for subsidization of immigration, but such an ill-thought-out solution would have meant tying up voluminous amounts of capital to attain uncertain and less desirable results. Finally, subsidized immigration permitted the joining of “the useful to the agreeable,” since it peopled the province with Europeans at the same time that it solved the problem of an insufficient work force. Indeed, notwithstanding prejudice and fantasies of “whitening,” it was the difficulty of obtaining a spontaneous flow of regular, disciplined, and cheap labor for the coffee frontier that made the immigrant solution viable. The continuation of subsidization until the 1920s clearly demonstrates the difficulty of obtaining a self-sustaining supply of labor for the region, to allow it to keep up its rate of accumulation. But Capivary (where commercial farming rapidly declined) and the Paulista West are exceptions from the standpoint of the substitution of slave labor. The rule, based on available knowledge of the agricultural reality of Minas, Rio, and São Paulo during the First Republic, was the continuation of commercial agriculture through the use of Brazilian labor. This, in turn, became the basis upon which the system of coronelismo was established in those states.

Because of an abundance of unoccupied or only nominally appropriated land especially in the south central region, the difficulties in subduing, even in part, a work force of free or freed people were so great that agrarian elites preferred using slave labor until the eve of the definitive abolition of slavery. From this fact a question arises, as yet unanswered, and seldom even formulated by the historical literature: in what ways were the mechanisms of rural social control redefined, so as to maintain the viability of commercial agriculture and the class dominance of the rural producers who controlled it? I do not have an answer to this question. Nevertheless, I shall repeat an expression of Fernando Novais, which appears especially apt: “To pose a problem is, perhaps, a fruitful way to conclude.”⁵⁶

Author notes