Parishes and Haciendas: Geography and Statistics

According to a treasury census of 1829, the city of Lima had retained its colonial characteristics.^[2] It housed 58,326 inhabitants, of whom 4,602 were slaves living on estates and haciendas of various sizes on the outskirts of the city. The rural realm represented roughly the actual area of the province of Lima or the so-called valley of Lima.^[3] In this area approximately 200 production units classified as haciendas competed for the available cultivable space, a situation that would not change very much between 1780 and 1910.^[4] In this relatively small rural sphere I intend to correlate the size of the productive unit to its slave laborers, the slaves' opportunities for accumulation, and the links these haciendas had with the city.

There is no information about slave numbers or conditions of slave labor for all the haciendas on the outskirts of Lima. Nonetheless, the

1813 census provides some clues, and for Surco and Chorrillos, two rural areas located about ten miles from the city's main square, a second census allows us to compare the evolution of slave distribution between 1790 and 1813. Supplementary information from notarial registries for subsequent years has been elaborated by Aguirre (1990, 17 ff.; 1993, 51–52).

The rural environment was not uniform. Each parish had its most proximate rural region. And a little farther off was the area occupied by the haciendas, small conglomerate villages, and towns. One of Lima's six parishes, San Lázaro, may illustrate the rural environment's diversity. San Lázaro housed the largest black free and slave population. After independence this parish—like all others—was divided into cuarteles (districts) and the cuarteles into barrios. The small plots, farms, and estates were spread out behind the barrios. In cuartel one, formed by four barrios and several small plots of land scattered in the two areas of Guia and Amancaes, an examination of the distribution of people called citizens or slaves allows us to decipher the rural component of the environs of the urban centers (Table 2). If, based on these figures, we contrast the distribution of the Spanish and the slave populations in the rural and urban spheres, we obtain some interesting indicators. The average ratio of slaves to Spaniards in the four barrios

amounted to 1.3:1, while the same calculation for each barrio denotes a high degree of disparity (4.0:1, 0.7:1, 0.4:1, 1.0:1 respectively). Thus, we can see that the proportion of slaves to Spaniards was much higher in barrio VII than in the other three barrios. The dividing line between the rural and urban spheres was imprecise. This inequality probably reflects the more or less rural character of the barrios and their greater or lesser proximity to the estates or the city center. In the rural area of this parish (i.e., on the estates) the ratio of slave to Spaniard was even higher: 19.0:1.^[5] If we exclude the persons listed as female Spaniards on the estates, the result is a ratio of thirty-five slaves to each owner. The ratio would be only slightly higher if we included the fifty-nine priests and nuns in barrio X; they were probably also Spaniards. This is a low density per unit of production if we compare it to similar figures from the plantation zones of the northern and southern coast.^[6] Similar calculations suggest an equivalent density (as we shall see) for nearby areas of San Lázaro parish such as Surco, Chorrillos, Magdalena, and Miraflores. On the level of laborers per productive unit this figure indicates the existence of relatively small haciendas in Lima's hinterland. In a rural register of 1837 that took in the province of Lima—that is, all of Lima's hinterland—were 152 haciendas with 2,004 slaves (Aguirre 1993, 52). In a broader area encompassing the coastal valleys we see a similar distribution of the slave population (Figure 1). Seemingly, despite the great disparity among these haciendas, between 1813 and 1837 the average number of slaves per hacienda had dropped from 35.0 to 13.2. Only two years later, in 1839, the register listed 189 haciendas with an average number of 16.6 slaves (Córdova y Urrutia 1839, also quoted in Aguirre 1993, 52).

In San Lázaro's rural area there were 950 slaves, and in its urban center 1,874 (that is, rural slaves lived in the heart of the urban parishes). However, the urban presence alone was by far more important. According to the important 1792 census, Lima relied on 13,482 slaves. In 1818, the date of the next census after that of 1813, there were 8,589 slaves in Lima (Jacobsen 1974). If we assume that the two censuses covered the same geographical regions (though the later census did not refer to this matter), we notice that 32.9 percent of Lima's slave population lived in the parish of San Lázaro and that this slave population was concentrated in the urban zone of the parish, in the barrios. Thus, approximately one-third of the population within the parish of San Lázaro had occupations in the rural sector.^[7] The

[Full Size]

Figure 1.
Distribution of Slave ownership in five Lima valleys: 1837. Source: Aguirre (1993,52).

parish's hinterland, in accordance with the boundaries established by decree in 1626 and apparently still valid in 1884 (Clavero 1885, 32 ff.), embraced a vast area whose borders were the present zones of Lurin, Hauchipa, Naña, Carabayllo, Surco, Pachacamac, and Magdalena. The rural space and the urban zone were extremely interconnected. San Lázaro was a parish located outside the city walls and a parish where—as we will see—a critical component of the daily life of the black and slave population evolved.

For Surco, Chorrillos, and Magdalena, three minor locations next to San Lázaro, we rely on separate figures that allow us to look more closely at the situation of slaves on some of the haciendas, the estates named in the next three tables. The essentially rural physiognomy of these sites farther removed from the urban center is evidenced by the much lower proportion of inhabitants than in the barrios of San Lázaro.^[8]

The distribution of the slave population on the haciendas, as well as the ratio among men, women, and children in the productive unit was quite unequal. In Magdalena the average number of slaves per pro-

duction unit amounted to 38.1, which closely corresponds to the average (35.0) found for the rural setting in the parish of San Lázaro. If we leave aside the population of children, we obtain a figure of 32.9 (Table 3). Before hazarding what these figures mean, we should look at the equivalent data for the parishes of Surco and Chorrillos (Table 4, which also permits us to verify the variations between 1790 and 1813) and for the site of Miraflores (Table 5).

In Surco and Chorrillos the average number of slaves per productive unit was 88.8 in 1790, and 102.0 in 1813. We can attribute this rise to the existence of the Villa hacienda, the largest in either area.^[9] If we exclude it from the calculation, the average drops to 35.4 in 1790 (approaching the average of San Lázaro and of Magdalena in 1813) and to 56.3 in 1813 (still a high average given the enormous expansion of

the San Juan hacienda between 1790 and 1813). For Miraflores the average number of slaves per productive unit was 53.7 (excluding the case of Brabón), and 45.5 (omitting children). From these rough averages we can perhaps infer that the further from the urban nucleus, the greater tended to be a hacienda's ratio of slaves to productive unit and the ratio of slaves to Spaniards, if we consider Spaniard to be equivalent to owner. Here is the very small difference between the estates of Amancaes and the haciendas of parishes that bordered the urban center (Miraflores, Magdalena, Surco, and Chorrillos). Also reflecting this concentric gradation from the urban center is the size of the haciendas (which is not necessarily the same as the number of slaves). The closer to the city, the more dispersed the land-tenure pattern, and the greater the number of small-size productive units. The pattern held true for a later period as well. If we examine what was called the Lima-región between 1825 and 1840, we find that 53.4 percent of all production units in five valleys were garden plots. The Rí-

mac valley, the valley closest to the urban nucleus, in the parish of San Lázaro, contained by far the greatest number of small landed properties: 172 garden plots, 79 larger plots, 23 farms, and only 9 large estates (Table 6). Of the total of 183 garden plots, thus 172 (94 percent) were in the Rímac valley.

The size of a hacienda had a great deal to do with the degree of relative freedom slaves had within the hacienda, the mechanisms of control, and also the types of production that would eventually allow slaves to sell basic foodstuffs in the urban markets. Lima's haciendas were definitely much smaller than other coastal properties.

Pando (the hacienda where I placed the Lasmanuelos family), located in the district of Magdalena, was a typical hinterland hacienda with its forty-five slaves. Since I contend that family bonds were one important ingredient of manumission, let us analyze the gender composition of the slaves on haciendas. Here our central objective is to clarify how the distribution of men, women, and children inside the hacienda shaped families' and individuals' strategies of accumulation and the way in which slaves linked themselves to urban life and to other social groups. What determined the ratio of men to women in

the haciendas is unclear. The type of crop might have influenced the ratio and conversely, changes in the gender composition might have altered cultivation. Otherwise gender ratios may have been either completely arbitrary, as women also performed typically male tasks, or a conscious strategy in order to encourage slave reproduction. There is no evidence for either of these arguments. It is surprising, nonetheless, that in both Surco and Chorrillos the increase of the slave population between 1790 and 1813 exhibited an almost perfect balance between the female and male populations, even if on the level of each unit of production this balance was quite skewed. Despite the fact that marriage was not proof of a couple's union—much less in the world of the hacienda—in Surco and Chorrillos a significant portion of the women who lived on haciendas were married to slaves working on the same hacienda. On the Villa hacienda in 1790, 180 of the male and female slaves (out of a total of 277 male and female slaves) were married couples (65.0 percent); in the case of San Juan, of 44 slaves, 26 were married (59.1 percent); on San Borja there were 32 married slaves out of a total of 58 (55.2 percent). On the smaller haciendas, Valverde and Chacarilla, only three married men appeared in each case, and on the estate of Porras out of twelve slaves, six slaves were married (50.0 per-

cent). These figures indicate that an average of 60 percent of slaves were married to other slaves on the same hacienda. These percentages were slightly higher on the larger units of production.

The 1813 census of Magdalena gives us a good idea of some of the gender ratios on its haciendas.^[10] Above all, the largest haciendas tended to maintain a balance between the female and male populations. This makes sense as these percentages corresponded to those slaves who were actually married. Perhaps many more just lived together. This assertion echoes the findings of Cushner (1972, 193) for the Jesuit haciendas in Peruvian territory at the end of the eighteenth century. Among Magdalena, Miraflores, Surco, and Chorrillos, Magdalena and Miraflores exhibited quite similar total percentages of men and women, with about 30 percent more slave men; in Surco and Chorrillos the total number of male slaves almost equaled that of female slaves.^[11] Insofar as we believe that both family life and a strategy for freedom were important to slaves, the landed properties that contained the greatest conflict and instability must also be those in which the greatest gender imbalance prevailed. When slaves on one hacienda could not foster family ties there, they often moved from place to place and established contact with neighboring haciendas and towns. Marriages among slaves from different haciendas and transfers through purchase from one hacienda to another explained such mobility. Higher mobility in turn reflected (at least in slave owners' eyes) weakening control over slaves; for the latter it certainly meant broader experience. Mobility made them aware of what was happening on other haciendas and allowed them to adjust or strengthen their arguments before owners or administrators. Thus, although slave marriages were a desired goal for masters who wished to augment an increasingly scarce labor force, when marriage or sexual contact meant higher mobility it also widened slaves' perceptions, anxieties, and possibilities.

There is no evidence that any hacienda except Palomino in Magdalena lacked female slaves. To compensate disparities or imbalances in the distribution of men and women throughout a broader geographical area without the purchase of new slaves, possibilities and opportunities for movement beyond the edges of the hacienda were needed, to give the slave population a certain amount of emotional satisfaction. Contemporaries were quick to interpret the slaves' greater circulation as idleness and even arrogance and boldness or worse, when slaves ran away. It diminished the slave owners' control over their labor force, in spite of the fact that on occasions—as was the case

with Manuel—slaves were severely punished because they left the hacienda. Such punishment was arbitrary and depended on specific slave-owner relationships; contemporaries usually interpreted punishment as an act that contravened established customs.

Marriage was a serious matter, even on the haciendas. Owners and slaves alike took into account and honored the matrimonial rules valid for the rest of society. Beyond respect for marriage itself, and for ecclesiastical rituals, we find that in slave society it was possible to annul a marriage on the grounds of a "second degree of illicit copulation," or opposition from parents or newlyweds.^[12] Even slaves had to declare that they were voluntarily consenting to marriage. These factors further weakened masters' ability to make marital decisions on behalf of their slaves and demonstrated the extent to which the sacrament of matrimony could interfere with notions of private property.

In terms of the characteristics of economic development at the end of the colonial period and the significance of the Bourbon reforms, what is salient in this microscopic examination is that Lima's haciendas experienced an overall increase in the slave population. Only two haciendas, Valverde and Porras, reported a reduction in slave population; in 1790 each had twelve slaves. We might infer that the slave population on smaller haciendas decreased, even if Chacarilla, which between 1790 and 1813 augmented its slave labor force from three to twenty-three, contradicts this fact. Perhaps an additional argument that substantiates the decrease of slaves on small haciendas is that many of these recorded the presence of a non-slave labor force at the beginning of the nineteenth century. Such was the case with Brabón and Lobatón (in Miraflores).^[13]

The objectives and the organizational logic of the labor force are less clear if we include in our analysis slaves who were minors. We simply assume—but have no way to determine the reality—that all the under-age children were born of slaves living on the hacienda. The documentation available does not differentiate them from minors purchased outside the hacienda.

To evaluate the extent to which attempts to reproduce slaves within the hacienda were successful, we will construct two indicators. One records the average number of children per married couple and the other, of children per woman over sixteen years of age. Once again the only censuses that include these figures are the 1790 census of Surco and Chorrillos and the 1813 census of Miraflores and Magdalena (Table 7).

The five factors that researchers at various sites use to explain the low numbers of slave children are: (1) rates of low fertility and high infant mortality; (2) suspected (but not proven) choice of slave women to abort in order not to bear slave children; (3) matrimonial age and composition by age of the entire female slave population in each unit of production; (4) sale of children by an owner outside the unit of production; or (5) simple lack of concern among owners, a reason that could underlie all the others. Other researchers argue, on biological grounds, that a woman's procreative capacity might diminish as a result of excessive work or mistreatment. Some evidence suggests even more dramatic explanations for the cases in which masters, family members, or compadres were the parents of slave children. There were masters who, wishing—for a multitude of moral reasons—to erase traces of paternity, subjected pregnant slaves to harsh treatment and beatings. Their objective was to provoke an abortion or secure a confession that would blame the pregnancy on another man. In the case of Josefa Aparicio, for example, who in 1842 worked on the Retes hacienda, the owner had her locked in the pillory to be "subjected to a beating session every 24 hours ... compelling her with these punishments to ascribe paternity to Don Manuel Andrade."^[14]

It is true also, however, that the thesis of a low fertility rate is hardly surprising in a context where the fertility rate in general was

extremely low.^[15] Although the proof is yet far from convincing, we observe a decrease in the number of children per slave woman in Lima, to the point that the number of children dropped below the already very low birth rates on eighteenth-century Jesuit-owned haciendas (Cushner 1975, 190n.). Other factors beyond those we noted help explain why children were absent from haciendas. The closer to Lima proper, the more frequent the physical absence of owners; sometimes we might infer that slave children were taken into the city as house servants, were inherited by family members, or were sold. Combining the theses of an extremely low fertility rate, geographic proximity, and the particular characteristics of a hacienda's operation, we can infer a sixth reason for the low numbers of slave children: parents managed—even in the arduous context of the hacienda—to liberate their children and thus children appeared only randomly in inventories or censuses. Confirming this hypothesis are several different cases and some evidence that depicts slaves' capacity to accumulate on the haciendas, in order to pay or negotiate their children's purchase price. This is exactly what Manuel and Manuela did on behalf of their two sons, Manolo and Manolito.

Thus, even though a high degree of differentiation existed among haciendas in Lima's hinterland, on average these productive units were much smaller than the sugar and winery haciendas on the northern and southern coasts, where concentrations of six hundred or more slaves were common. The size of the property often determined a slave's success at procuring freedom. The smaller the property, the closer the connections between slaves could be; in a parallel manner the greater the chances of negotiation with owners, mayordomos , and caporales . Surely the capacity of these latter individuals to control the lives of their slaves was also greater, but greater control did not automatically bring harsher treatment. It is not necessarily true that larger productive units offered greater levels of liberty or provided slaves with any more anonymity. The slave's potential for negotiation—where channels of negotiation existed—increased when proximity between individuals was higher, that is to say, on the small and mediumsized units of production.^[16]