SLAVERY AND ABOLITION
A Journal of Comparative Studies
Vol. 5: No. 1 (May 1984)
Black into White in Nineteenth Century
Spanish America: Afro-American
Assimilation in Argentina and Costa Rica*
In his masterful study of racial attitudes in Brazil, Thomas Skidmore has shown how the Brazilian elite consciously preferred, and pursued through the foment of European immigration after 1850, a "whitened" society in which the African element would be progressively reduced.' Given the historical realities of Brazilian society such a policy could perhaps be implemented, but only with great regional variability and never fully eradicating what the elite saw as the "inferior" African element represented by Negro and colored (mulatto) Brazilians.
Such a semi-official policy of whitening was common to both Luso- and Hispanoamerican elites of the late nineteenth and early twentieth centuries, with the curious symbiosis of a paternalistic acceptance of race mixture and its "beneficial" impact (unlike segregationist or apartheid views in the United States and South Africa at that time), and a belief in the innate inferiority of those of African descent (in this they shared the basic racism of the abovementioned societies).2 However, most Spanish American societies, just as northeastern Brazil, would not receive the mass of European immigrants who inundated the Brazilian southwest, or such Spanish American nations as Cuba, Uruguay, and Argentina. And yet, except for those areas continuing the African slave trade (Cuba and Brazil in particular) nearly everywhere there was a long term proportional decline, over the nineteenth century, of the Afro-American population, whether through race mixture and "passing", or simply as a result of a decreasing Afro-American biological component within the general population. Thus, the desired goal of the Brazilian and Spanish American elites - "whitening" or bleaching" of the population - did not always require massive European immigration for its realization.
While modern Spanish American society was whitened in general, regional experience was extremely diverse.3 The Indo-American areas of Mexico, Guatemala, and the Andean republics produced a mestizo more often than a mulatto population, with all such admixtures coming to be referred to as "casta", "ladino", or simply "mestizo", with some mention of phenotype appended for clarification if need be. In these Indo-American areas the numerical predominance of the Indian population during colonial times, especially in the Guatemalan and Andean countrysides, meant that whitening of the general population would proceed very slowly there if at all during the nineteenth century, although Afro-American assimilation took place much more rapidly.4
In those few areas of heavy Afro-American population within mainland Spanish America, coastal Venezuela and Colombia (including Panama), as well as parts of the Spanish Caribbean (Dominican Republic, Puerto Rico and, in particular, Cuba), whitening would proceed, depending upon the size of the Afro-American population, its location in the socioeconomic system, and the importance of European immigration (here only Cuba would have had a major nineteenth century European influx). Finally, in those peripheral mainland areas where colonial indigenous populations were small or nonexistent and where Afro-Americans formed a substantial but minoritarian part of the population, whitening would be most rapid and complete. In these areas, the elite ideal of whitening could eventually be realized, with a degree of racial homogeneity along Euro-American lines sufficient to allow for a veritable Social Darwinist national ideology of whiteness and superiority. Perhaps the two Spanish American nations most identified with this image are Argentina and Costa Rica, or rather Buenos Aires and its hinterland and the Central Valley region of highland Costa Rica.
A comparison of the precise means by which assimilation was achieved in these two contexts may illuminate not only the cases at hand, but also and more importantly, the structural factors behind the process of Afro-American assimilation or "disappearance" throughout much of Spanish America. While all Latin American elite groups may have held racist views regarding Afro-Americans, an important point which is highlighted by this study is that under certain conditions the goal of whitening may not have required massive European immigration. The colonial structural heritage of pervasive race mixture and lagged or lower replacement capacity of Afro-American populations, as well as the nineteenth century phenomena of European immigration and agro-export expansion, furthered this elite cherished goal of whitening. Quite without the intense introspection of the Brazilian and, to some extent, Argentine elites, many of their Spanish American counterparts presided over a structural "conspiracy of silence" which led to essentially similar outcomes. However, particularly in Argentina and Costa Rica, Spanish American national ideologies developed which proclaimed the superiority of a presumed European biological heritage, often denying any overtly racist basis of this mythology and historical process as well. The collective psychological mechanisms of contemporary suppression may be fairly simple, but historically there was a counterpart demographic structure which progressively diluted what Afro-American population existed and insured a continuous decline in this group's relative size. This demographic structure and its lower Afro-American replacement capacity, inherited from colonial times but affected as well by nineteenth century international and internal migratory movements, will be a major focus of what follows.
Afro-Americans in late colonial and early independent Argentina and Costa Rica occupied structurally similar positions, as indeed was true of Spanish America in general, excepting perhaps the few plantation sugar economies which employed extensive rural slave labor (Cuba, the coastal areas of Venezuela, Colombia, Ecuador and Peru, etc.). Afro-Americans accounted for about one-fourth to one-third of the population of some 50-l00,000 in Buenos Aires in the first half of the nineteenth century, and perhaps 10-20 per cent of the 50-80,000 souls inhabiting highland Costa Rica in the same time period.5 They were a preeminently urban, artisan, and domestic servant population. Those who worked in the countryside were usually male slaves or their descendants who served as cowboys or overseers of seasonal wage labor or forced Indian labor on ranches and large farms. The number of Afro-American women in the countryside lagged behind that of males, and the converse was equally true of the cities, where women often outnumbered men by one and a half or two to one after the early nineteenth century suppression of the African slave trade.
The implications of this peculiar position in the colonial society and economy were many and varied. As artisans, laborers, and servants Afro-American family formation patterns differed not only from those of the mass of village agriculturalists but also from those of their non-colored city brethren, rich and poor. As urban dwellers and descendants of a mobile slave labor force, they were characterized by even greater sexual imbalance (feminine predominance in cities; masculine in the countryside) than already was the norm in such pre-industrial cities. This skewed sex ratio fomented race mixture within or outside of marriage, while limiting and delaying access to same, thus significantly lowering group fertility and replacement rates vis-á-vis non-Afro-American populations. Moreover, changes identified with nineteenth century developments, from military conscription and European immigration in Argentina to rapid out-migration in Costa Rica, aggravated this relative demographic handicap further, leading to an even more rapid proportional decrease of the Afro-American population.
The assimilation or disappearance of Spanish American populations of African descent was predicated upon two interrelated factors: generalized race mixture and a lower than average replacement capacity of Afro-American populations. This lower or lagged replacement capacity was itself owing to the combination of three possible factors: (1) most importantly perhaps, more limited and later access to marriage (or common-law union), based on both the pronounced sexual imbalance within urban AfroAmerican society and the limited social attractiveness of Afro-American women as wives for urban white and mestizo males; (2) higher infant and general mortality rates; and (3) lower married fertility rates among this generally impoverished and urban population group. Our discussion of these interrelated variables, inherited from colonial society, will be followed by a brief analysis of the impact of later nineteenth century developments.
While miscegenation has been characteristic of all multiracial societies to one degree or another, Latin American experience has been notable in both the pervasiveness of this phenomenon and, more importantly, in the position accorded to those of mixed origin. Hoetink most clearly expressed this point, regarding the dual features of widespread "passing" across an ill-defined "color line" and the social acceptance of those of light color by local Iberoamerican whites as marriage partners. In these societies the local defmition of whiteness tended to include many of those of light color and, just as importantly, marriage or long term common-law unions (as distinct from more informal or surreptitious concubinates and liaisons) across what in other contexts would be perceived as racial lines (white vs. colored) was far more frequent.6
Thus, added to the quite frequent extramarital unions spanning racial lines, Latin American societies also witnessed the growth of a significant population born to both Church-sanctioned and common-law unions between Afro-Americans and the non-colored.7 The frequency of this latter phenomenon varied widely by time and place perhaps, but it was an ubiquitous feature of Latin American societies and could reach quite substantial levels in some cases, as we shall see below.
In both Argentina and Costa Rica there is abundant evidence of the existence of such a relatively flexible "color line", subject to surprisingly rapid redefinition over time, even in the case of individual lifetimes. Moreover, it is worth noting that exactly the same terminology is used to describe cases and individuals in Argentina and Costa Rica in the freeing of "white slaves" or in describing the physical appearance of these individuals when still enslaved. Andrews notes the use of terms such as "white mulatto, white, white slave" in manumission documents, as well as descriptions emphasizing "blond" or "straight" hair and white color.8 In Costa Rica references were repeatedly made to "whiteness" or "amber" coloration ("trigueño", exactly as in Buenos Aires and other Spanish American countries somewhat later)9 as well as "burnt blond hair", etc. Moreover, the use of color identification as a means of implicitly raising or lowering an individual's social rank was also a common feature of contemporary discourse, in reference to those of high and humble social, albeit racially suspect background.
Perhaps one of the clearest possible indications of the decided tendency of Iberoamerican society to classify light-coloreds as white can be found in the late colonial Costa Rican censuses.10 Therein the population is divided and enumerated as "Spanish", "Mestizo", or "Mulatto and Negro". However, no clear and binding descent rule is used in order to assign the children of mixed unions. Most often, when the mother was "mestiza" or Spanish and the father Afro-American, the children would be registered in the mother's racial category, although there were exceptions to this rule as well. In the case of Afro-American women married to or living with Indian, mestizo or Spanish males their children would usually be listed with them as "mulatos y negros", but even here exceptions could be found, logically enough since their listing as mestizos could have been socially and administratively advantageous for them.
Miscegenation may have been most common outside of formal unions such as these, but more stable, recognized relationships were very frequent as well, involving all racial groups in Spanish American society. As we shall see below, Afro-Americans' urban location and the feminine predominance which resulted from this fact, when added to pervasive racial preferences in the selection of marriage partners, assured that this group would have the most difficult and delayed access to marriage. In societies in which concubinage was the rule rather than the exception at all social levels, this could only foment extramarital miscegenation as well. Indeed, nearly all of the studies of Afro-Americans in urban Latin America would indicate "whitening" in the selection of both marriage and liaison partners to have been the norm." In Costa Rica illegitimacy among the Afro-American population was approximately double the average, reaching the level of a third to a half of all Afro-Americans baptized and a fifth to a quarter of all illegitimate baptisms at the end of the colonial period; this without taking into account those children not baptized and likely illegitimate as well.12 Important here too was the urban location of Afro-Americans, raising illegitimacy levels regardless of race, contributing to the differentiation of the community from mestizo villagers and lowering its replacement capacity. Presumably, a large number of these illegitimate children were the result of race mixture tending toward whitening.
Racially mixed and socially recognized unions were, however, not isolated occurrences in the societies we are dealing with. In late colonial Costa Rica, for example, fully 33 per cent of the 182 married Afro-American males and seven per cent of 137 married females in the colonial capital of Cartago were to be found within declared multiracial unions.'3 In the area of the later national capital of San Jose' to the west, the figures were even more striking: 55 per cent of the 134 married Afro-American males and 35 per cent of the 65 married females, typically with mestizos as their partners. Moreover, in Cartago itself two Afro-American males were enumerated with "Spanish" and three with Indian wives, while nine women were married to Indian males. The only cases of "Spaniards" cohabiting with mulatto women were to be found in the cattle range region bordering upon Nicaragua to the north. There, as well, two Spanish women were living with Afro-American males.
For the Afro-Argentines of Buenos Aires, Andrews found that in 1810 only 2.2 per cent of men and 2.5 per cent of women were married to the non-colored ("whites" according to the author). In 1827 the figures were 3.0 per cent for men and 6.0 per cent for women.'4 These figures, while sociologically significant, would hardly indicate any rapid assimilation through race mixture. However, additional data suggest that race mixture, within and outside of marriage, was much more widespread than the racial declaration of these household heads allows for. Andrews reports that between 1810/20 and 1850/60 the mulatto ("pardo") component of the Afro-American population enlisting in the army grew from 19.9 to 51.1 per cent. Moreover, in the 1827 census, some 20 per cent of the children of black ("moreno") household heads were listed as mulattoes ("pardo"), and suggestively, 2 per cent of the children of "white" family heads were listed in this category as well, despite the social disadvantage of "passing" thus denied.'5 Later, in the wake of massive Italian immigration, unions involving Afro-American women and non-colored males became even more common, leading one caustic Afro-American male editorial commentator to quip that, given the sexual imbalance in the community, black women who "could not get bread would have to settle for pasta".'6
II. Lagged Replacement Capacity
Nearly equal to race mixture in importance in the disappearance of the colonial Afro-American population in both Argentina and Costa Rica was this group's lower rate of replacement. This was owing, primarily, to the skewed sexual distribution of the community. Secondarily, this group often suffered a combination of higher general and infant mortality and lower marital fertility rates.
The feminine predominance within and racial discrimination without the urban Afro-American community meant that these women were far less likely to marry at an early age than mestiza and white women, particularly these latter women in the outlying villages and towns. This, in turn, led to fewer children per woman within the community, a lower general fertility rate and replacement capacity.
Women outnumbered men among Afro-Argentines in Buenos Aires shortly after the heavily masculine slave trade was formally abolished in 1813, reaching the level of 58.5 men for each 100 women by 1827. 171n Costa Rica we have no surviving nineteenth century census data which specifically records race. However, the last colonial census (1777-78) did list nearly 139 women per 100 men in Cartago and 113 in San José, in this former case the highest rate of feminine predominance of any of the three population groups categorized (Spanish, Mestizo, Mulatto and Negro).18 Moreover, we do have documentation which distinguishes the principal mulatto section of the colonial capital of Cartago (La Puebla de los Angeles, or more colloquially, La Puebla de los Pardos or mulattoes) in 1820, 1844 and 1864.19 Not surprisingly, La Puebla had the highest female to male ratio, oldest age at marriage, and the lowest child/woman ratio of the entire Central Valley region. In what follows, La Puebla and its approximately 1,500-1,600 residents over the first half of the past century will serve as a surrogate for the Afro-American population of nineteenth century Costa Rica.
Afro-American women outnumbered men in Buenos Aires by about one and a half to one over the nineteenth century and a similar pattern obtained in La Puebla in Costa Rica. Fully 137 women resided there for each 100 men in 1820, 144 in 1844, and 129 in 1864. Some 51 per cent of households were headed by single and widowed females in 1820, and 43 per cent in 1844, figures nearly double the average for outlying mestizo villages and significantly higher than other urban populations as well. The approximate age at marriage for women, based on the 1844 census, was an extraordinarily high 29 years of age, compared to a national average of 19.5-21 and a center city average of 21-22.20 Only 34 per cent of La Puebla women aged 1544 were or had ever been married, compared to values of 43 per cent for Cartago province, 56 per cent for Heredia province, 57 per cent for Alajuela province, and 58 per cent in rural Escazú, all to the west of Cartago.
These factors led to a substantially lower general fertility rate, as indicated by child/woman ratios. There were only 218 children aged 0-4 per 1,000 women in La Puebla in 1844, compared to 254 in Cartago province, 320 in San José, 315 in Alajuela, 318 in Heredia, and 364 in Escazú. More specifically, there were only 501 children aged 0-4 per 1,000 women aged 15-44 in La Puebla in 1844, compared to 522 in the province, 693 in Heredia, 654 in Alajuela, and 817 in Escazú.21 Similarly, in Buenos Aires in 1810 there were only 256.5 children under five years of age per 1,000 Afro-American women aged 15-44, compared to 400.8 among white women, while the figures for 1827 were even worse, 183.1 and 365.9 respectively.22 The negative impact of such a severely reduced general fertility and growth rate, this lagged replacement capacity, rivaled even miscegenation and reclassification in facilitating gradual whitening of the population and Afro-American assimilation.
Afro-American infant mortality was likely significantly higher than average in both cases. In Buenos Aires between 1827 and 1831 the mean annual infant mortality rate (deaths before age 1 per 1,000 live births) for Afro-Americans was 350.4 as against 284.3 for whites.23 No racially specific data on infant mortality yet exists for nineteenth century Costa Rica, but whenever clear distinctions between rich and poor are made, the child/wife ratios would seem to suggest higher infant mortality rates among the poor. In the city of San José, Afro-Americans were also a preeminently artisanal and impoverished population. At mid-century those who declared either laborer or propertyless status reported approximately 15-20 per cent fewer children under five when compared to wives and families of the propertied.24 San José's southern artisanal ward of "El Hospital" reported fully 50 per cent higher infant mortality than the other city wards in the municipal census of 1905, without distinguishing by race and without comparison to the still more healthful climes of the outlying villages where most of the non-colored population lived.25 While the Costa Rican evidence is both imperfect and indirect, the implication remains the same: higher than average infant mortality for Afro-Americans, as a function of their urban and relatively impoverished position.
Data on marital fertility among Afro-Argentines and Afro-Costa Ricans offers, at best, a rnixed panorama. The data which Andrews presents would seem to indicate an equal or even higher than average age-specific fertility rate for early nineteenth century Afro-Argentines.26 However, in recent analyses of the 1810, 1827 and 1855 censuses, my colleague, Dr Mark Szuchman has found that the child/wife ratio by specific age groups was far lower for Afro-Americans than for the non-colored. This would be an indirect measure of age-specific marital fertility, still somewhat subject to the distorting influence of a higher infant mortality rate to be sure, but perhaps suggestive of a markedly lower married fertility rate among Afro-Argentines.27
In Costa Rica there was no clear indication of lower age-specific marital fertility among the colonial Afro-American population or among the La Puebla residents at mid-nineteenth century. In the 1778 census of Cartago we find the number of young children ("parvulos", hypothetically those under 6 years of age) within Afro-American households to be similar to that of mestizos and whites.28 Likewise, in La Puebla in 1844 there were roughly similar numbers of children under five per 1,000 wives of specific age groups compared to the provincial average. However, the peaking of the number of resident children under age five (as a surrogate for married fertility) in two-parent Afro-American households occurred somewhat later (wives 25-29 and 30-34) than the norm (20-24 and 25-29), a reflection of later age at marriage typical, to a greater or lesser degree, of all local urban and artisanal populations.29
Overall, the lagged replacement capacity of Afro-Argentines and Afro-Costa Ricans was principally due to the sexual imbalance (males in the countryside or absent entirely, females in the city) which limited and delayed access to marriage in the best of circumstances. Add to this the rejection of Afro-American women by mestizo and white males as suitable marriage partners, higher infant mortality due to urban location and impoverishment, and possibly lower marital fertility, and one can readily see how Afro-American disappearance, when not assimilation through race mixture, could result in relatively few generations. These basically colonial structures and processes were further accelerated by later nineteenth century developments, to which we will now briefly turn our attention.
III. Nineteenth Century Patterns of Change ...
While both Argentina and Costa Rica developed as agricultural exporters over the nineteenth century profound, if dissimilar, changes were taking place in the respective societies. In both cases the settlement and bringing into production of large areas of virgin lands was the fundamental task of the mid- to late nineteenth century. However, in Argentina this was accompanied by extensive civil and international strife before 1870 and massive Italian immigration thereafter. Moreover, Buenos Aires maintained its unrivalled control over the greatly expanded Argentine hinterland during the entire period. In Costa Rica, no such mass immigration or armed conflict accompanied the mid-nineteenth century transition to coffee culture. Nevertheless, a rapid and profound redistribution of existing population toward the western coffee frontier led to the breakdown of colonial urban control (Cartago recedes in favor of San José as the new capital) and a thoroughgoing "ruralization" of local society. In both cases these innovations tended to increase the rate of race mixture and, more often than not, accentuate the lagged replacement capacity typical of colonial Afro-Americans.
Argentina was intermittently at war with itself or with others from the wars of Independence beginning in 1808 to the War of the Triple Alliance in 1865-70. Moreover, Afro-Argentines were disproportionate participants in all of these conflicts, with the inevitable consequences for a group already plagued by a highly skewed sexual distribution. Andrews provides a wealth of information on the military exploits of Afro-Argentines of all social levels, and he concurs with local authors regarding the final disastrous blow to Afro-Argentine males resulting from their service and all too frequent death in Paraguay during the War of the Triple Alliance.30 Census data for the period 1810-1855 in Buenos Aires would suggest a disproportionate recruitment, male absence and death, leading to fewer children among Afro-American households.31 This impact of the military campaigns and conscription upon Afro-American males may have been imperfectly quantified as yet, but its generally negative direction would seem indisputable.
The impact of massive Italian immigration was predictable in its foment of race mixture, perhaps unnecessarily at such a late date. By 1887 urban Afro-Argentines, or more accurately those still designated as such in the city census of that year, amounted to only 8,005 souls, 1.8 per cent of Buenos Aires' population, the remaining 425,370 being classified as "whites".32 Of these eight thousand individuals 4,700 were women.33 Given such a sexual disequilibrium it was hardly surprising that unions of Italian immigrant males and Afro-Argentine women might become frequent occurrences. Andrews reports several editorial commentaries upon this phenomenon, including the above mentioned witticism involving "bread and pasta". However, not only do these commentaries recognize the near inevitability of such an outcome in light of urban demographic realities, they also display little concern for the long term consequences of race mixture for the future of the Afro-Argentine population. Clearly, the earlier nineteenth century process of assimilation had been nearly completed, so that the role of Italian immigrants was neither really critical nor much resented.
If Afro-American assimilation was carried forward in nineteenth century Buenos Aires by the combination of forces described above, the process in Costa Rica was accelerated precisely by the "ruralization"34 associated with coffee culture. As the Afro-American neighborhood of the colonial capital of Cartago declined along with the entire city, and these groups in San Jose' suffered similar if less drastic processes of change, the relative weight of non-colored populations increased. Moreover, Afro-Costa Ricans migrated as well, putting them in a context of mestizo village majorities, more likely partners in the ongoing process of race mixture.
Between 1844 and 1864, the height of the early coffee expansion, the population of the Central Valley was rapidly shifted toward the west and the smaller villages. The province of Cartago declined from 31 per cent to 22 per cent of the region's population. The western province of Alajuela increased from 14.6 per cent to 26 per cent in the same period. Over the course of the entire nineteenth century the percentage of population living in the four provincial capitals declined in all cases, with the partial exception of the capital of San José. Rapid movement of colonial "urban" populations toward the agricultural frontier typified the period of coffee's expansion in Costa Rica.
This process of ruralization in Costa Rica had several implications for Afro-American assimilation. To the extent that urban Afro-Costa Ricans migrated this would tend to increase the rate of race mixture, given that the areas they entered were more heavily mestizo than their neighborhoods of origin and that socially recognized mixed unions were more frequent there in any case.35 As the outlying village population experienced a decline in average age at marriage and an increasing fertility rate,36 already divergent from central city patterns (as with infant mortality) in any event, this largely mestizo population would grow faster than previously and leave urban Afro-Costa Ricans even further behind. If this migration, however voluntary, was heavily masculine then, just as in Argentina, such a withdrawal of males from the community would further aggravate an already difficult access to marriage and lower replacement capacity. To the extent that Afro-Americans remained "trapped" in their colonial urban settings and avoided any inherent increase in the rate of race mixture they would still be negatively affected by coffee-based ruralization. The increasing growth rate of the village mestizo population would lead to a proportionate reduction in Afro-American population in any event. Indeed, even if the village population's intrinsic rate of growth did not increase, the simple redistribution of largely non-colored populations from city to countryside, with its relatively higher growth rate, meant that Afro-American populations would decline as a portion of the general population over the nineteenth century. Moreover, a marked process of pauperization and proletarianization associated with the decline of artisanry under pressure of imported English goods, as well as with growth of the capital city of San José late in the nineteenth century, likely led to an increasingly wide gap in general and infant mortality rates between city and countryside, rich and poor.37 This, in turn, reduced even further the competitive demographic position of the largely urban Afro-Costa Rican population inherited from the colony.38
Overall, nineteenth century developments in both Argentina and Costa Rica contributed to the disappearance of the colonial Afro-American population. However, the relative importance of these national period changes should not be overstated. Race mixture and lagged replacement capacity were the structural constants behind this process, and these were clearly features of colonial society.
The above analysis has demonstrated how parallel outcomes, the assimilation and virtual disappearance of the Afro-American population, obtained in widely divergent contexts of nineteenth century Spanish America. Argentina, inundated by white European immigration and overwhelmed by agro-export wealth; Costa Rica, untouched by massive European immigration but equally profoundly transformed by export agriculture and ruralization, were both led toward national ideologies of "white superiority" amid racial homogenization and Afro-American disappearance. This was perhaps facilitated somewhat by late nineteenth century European influx in Argentina, but often at the cost of heightened racial animosity as well. Much more critical in determining such an outcome were the longstanding Spanish American colonial patterns of Afro-American assimilation: generalized race mixture (whether through marriage, common-law union, or concubinage) and sexual disequilibrium and impoverishment among this heavily urban segment of the population, which resulted in limited and later access to marriage, higher infant mortality rates, and lower replacement rates, race mixture itself notwithstanding.
We have seen how nineteenth century changes accelerated colonial processes of assimilation, whether by decimating Afro-Americans through conscription, warfare and swamping those few remaining with Italian immigrants as in Argentina, or as was more often the case, by increasing the pace of race mixture and the rate of rural, mestizo population growth by the redistribution of settlement, as in Costa Rica. The colonial heritage was essentially accelerated in its work by the great transformations of the nineteenth century. However, as we have seen in both cases, but particularly in Costa Rica, assimilation was not really dependent upon massive European immigration. The racist solution advocated and adopted by the Brazilian and Argentine elites - among others, that of European bleaching to allegedly "improve" the local gene pool - was not always necessary, even where both Afro-American assimilation and a national racist mythology were to be fully consolidated. Costa Rica and its silent resolution of this presumed problem eventually led, as in Argentina, to self-exaltation based upon a purportedly more pure Spanish racial background than its Central and Latin American neighbors. This Costa Rican experience surely demonstrates the strength of underlying structural factors, both social and demographic, which nearly predetermined the assimilation so anxiously sought after by the Argentine and Brazilian elites.
In effect, the colonial order provided the most reliable, albeit racist, mechanisms for resolution of the "racial problem" which early national elites often considered, paradoxically enough, the curse of that very colonial legacy and one which some naively believed remediable only by radical measures. There is a certain irony in the fact that, in advocating radical solutions such as massive immigration, panic-stricken liberal modernizers throughout nineteenth century Latin America always assumed the continuation of colonial social patterns (race mixture, concentration of Afro-Americans in urban laboring occupations, urban hierarchical control, etc.). These were perhaps more important conditioners of Afro-American assimilation than any such nineteenth century innovations, although admittedly somewhat more gradual in their effect than the one or two generation process of whitening demanded by such Europhile elites.
Just as in so many other aspects, the nineteenth century post-abolition order only developed and accelerated social patterns which had already emerged in the late colonial period. The disappearance of Afro-Argentines and Afro-Costa Ricans may have been completed in the national period, but if the timing of this process was national its pattern was eminently colonial. Paradoxically, the so-called "caste system" and the utterly ineffectual segregationism associated with the legal framework of colonial Spanish America, rather than nineteenth century European immigration, agro-export expansion, or Social Darwinist theory, were at the heart of this process of Afro-American assimilation. Subsequent rationalizations may have had recourse to the terminology of a late nineteenth century Europhile Social Darwinism, but the colonial heritage provided both the structural and, in part, the ideological mechanisms which assured whitening and racial homogenization.
Thus, both Latin American white supremacists (Argentine and Costa Rican elites in particular) and those who would, on the contrary, exalt its miscegenous heritage (theoreticians of the "cosmic race" and "indigenismo" in Mexico and Peru, or "racial democracy" in Brazil) reacted to an ongoing structural process which they could perhaps affect but certainly not control, and did not always clearly comprehend. Indeed, what is remarkable is that such widely divergent racial ideologies should have arisen from essentially similar processes. Afro-American assimilation through race mixture and lagged replacement could be exalted, decried or, as in Argentina and Costa Rica, studiously ignored, but it nevertheless formed a common pattern throughout nineteenth century Spanish America.
1. Thomas Skidmore, Black into White: Race and Nationality in Brazilian Thought, Oxford University Press, 1974.
2.The basic difference between Latin and Anglo-American race relations is now generally recognized to be limited to the question of race mixture, the social position of mulattoes, and the precise location of the color line and passing between white and light colored, rather than in their attitudes towards blacks per se. This point was first clarified by H. Hoetink, The Two Variants in Caribbean Race Relations: A Contribution to the Sociology of Segmented Societies, Oxford University Press, 1967, and Slavery and Race Relations in the Americas: An Inquiry into Their Nature and Nexus, New York, Harper & Row, 1973. This forms the basis for Carl Degler's award winning study, Neither White nor Black: Slavery and Race Relations in Brazil and the United States, New York, 1971, and its central argument for the so-called "mulatto escape hatch".
3. The literature on the colonial Afro-American in Latin America is indeed voluminous, but with the exceptions of slaveholding Brazil and Cuba, much less is known about 19th century and post-abolition Afro-American populations. An early source book for this period is the edited collection of Magnus Morner, Race and Class in Latin America (Columbia University Press, 1970). George Reid Andrews' study of the Afro-Argentines of Buenos Aires, 1800-1900 (University of Wisconsin Press, 1980), cited extensively throughout this paper, is one of the first to deal systematically with post-abolition Afro-Americans in Spanish America. This lacuna can be traced, in large part, to the scarcity of documentary sources for the 19th century where, unlike the colonial period, racial designations were frowned upon if not outlawed in most fiscal, censal and parochial records. The classic synthesis of the colonial experience is that of Frederick P. Bowser, "The African in Colonial Spanish America: Reflections on Research Achievements and Priorities", Latin American Research Review, VII, 1 (Spring 1972), pp.77-94.
4. Christopher Lutz, Historia sociodemográfica de Santiago de Guatemala, 1541-1733 (CIRMA: Antigua, Guatemala, 1982), a revised version of the author's 1976 doctoral dissertation at the University of Wisconsin presents a richly detailed analysis of the very rapid assimilation of the colonial Afro-Guatemalan population, emphasizing "whitening" in the selection of marriage partners and reclassification toward "higher", or at least more ambiguous, status groups. A similar argument for the 19th century Mexican mestizo is convincingly made by John K. Chance, "On the Mexican Mestizo", Latin American Research Review, XIV, 3 (1979), pp. 15~8, despite that same author's earlier insistence upon the importance of race in determining socioeconomic position in his Race and Class in Colonial Oaxaca (Stanford University Press, 1978). For a fascinating contemporary study of the overwhelming social content of such purportedly biological classificatory terms (mestizo, cholo, etc.) in the Andes, see Pierre L. Van den Berghe and George P. Primov, Inequality in the Andes: Class and Ethnicity in Cuzco (University of Missouri Press, 1977).
5. The principal data for this article come from Andrews' work on Buenos Aires and my own on Costa Rica. The latter includes a study of colonial race mixture in Estratificación socio-racial y económica de Costa Rica, 1700-1850 (Editorial Universidad Estatal a Distancia, San José, Costa Rica, 1978), 19th century demographic data presented in my doctoral dissertation, "Costa Rica Before Coffee: Economy and Society on the Eve of Agro-Export Based Expansion" (University of Minnesota, 1982), a few recently discovered sources, and the reworking of some of the censal data used in the dissertation. Wherever possible citations will be made to the published materials and will omit extensive archival reference. For a recent overview of the local literature, to the mid-1970s, relating to colonial Afro-Costa Ricans, see Michael Olien, "Black and Part-Black Populations in Colonial Costa Rica: Ethnohistorical Resources and Problems", Ethnohistory, 27:1(1980), pp. 1-29. For further information on blacks in colonial Buenos Aires see Lyman L. Johnson, "Manumission in Colonial Buenos Aires", Hispanic American Historical Review, 59,2, May1979, pp. 25-79; and "The Impact of Racial Discrimination on Black Artisans in Colonial Buenos Aires", Social History (England), 6, 1981, pp.301-16.
6. Hoetink, op.cit.; particularly Slavery and Race Relations.
7. Throughout the text we will be referring to both Church-sanctioned and common-law unions as "marriage", in part because the census data indicates cohabitation rather than formal legal status. On the other hand, there is a certain more compelling logic to such a socio-demographic equivalency. In Latin American societies informal or common-law unions usually accounted for either a majority or a very large minority of couples. "Illegitimacy" (i.e. birth outside of church-sanctioned unions) was pervasive at all social levels, so much so that in fact a three-tiered system of "legitimate", "natural" and "illegitimate" or "bastard" children developed. So-called natural children were those born to single mothers whose consorts openly recognized paternity and were to be included as legal heirs of their father. Children of common-law unions fell somewhere between fully legitimate and "outside" or "natural" children, tending toward the former in most cases.
8. Andrews, p.88.
9. For a fascinating discussion of the use of the term "trigueño" as a euphemism for light mulatto, tending, however ambiguously, toward "passing" and reclassification as white in not only Argentina but also Puerto Rico, Nicaragua, and Peru, see Andrews, pp. 83-5, especially note 61. See, also, Leslie B. Rout, The African Experience in Spanish America (Cambridge University Press, 1976).
10. For examples of this in Costa Rica see Estratificación socio-racial..., pp.26, 56-8.
11. For three particularly revealing analyses of both marital and extramarital "whitening" of Afro-American populations in colonial Latin America see Edgar Love, "Marriage Patterns of Persons of African Descent in a Colonial Mexico City Parish", Hispanic American Historical Review 51:1, February 1971, pp.79-91; Stuart B. Schwartz, "The Manumission of Slaves in Colonial Brazil: Bahía, 1684-1745", Hispanic American Historical Review, 54:4, Nov.1974, pp. 603-35 especially p.622; and Verena Martínez-Alier, Marriage Class and Colour in Nineteenth-Century Cuba: A Study of Racial Attitudes and Sexual Values in a Slave Society (Cambridge University Press, 1974). See, also, Lutz, op.cit. All of these authors document a pattern of unions spanning "proximate" racial lines (white-light colored; light colored-dark colored, etc.) rather than across the entire color spectrum, in each case evidencing a disproportionate tendency to marry "upward", i.e. toward lighter colored groupings. The Moorish, Iherian background and logic of this process is made clear by Love in particular.
12. See Héctor Pérez Brignoli, "Deux siecles d'illégitimité au Costa Rica, 1770-1974", in J. Dupaquier et.al., Marriage and Remarriage in Populations of the Past (London: Academic Press), 1981, pp.481-93, especially Figure 1, Tables II and III, and note 7.
13. Estratificación socio-racial...,p.52. Of 182 married (cohabiting in fact) Afro-Americans in Cartago, 55 males listed "mestiza" wives, while only 10 married Afro-American women were listed with mestizos (1) or Indians (9). In San José all 51 males and 23 females were cohabiting with mestizos.
14. In 1810, only 4 of 160 women and 3 of 134 men; while in 1827, 20 of 334 women and 7 of 231 men. I would like to thank Professor Andrews for providing me with these figures, taken from his doctoral dissertation and not included in the work cited in note 3.
15. Andrews, p.89. The author's insistence upon denying the reality of Afro-American assimilation through passing remains a curiously obtuse argument within an otherwise excellent analysis. In being reclassified from "moreno" through "pardo" through "trigueño" to white in a lengthy process of race mixture, those who Andrews insists upon referring to and rescuing as "Afro-Argentines" would perhaps be unwilling partners to such an operation. Past a certain point of race mixture these so-called Afro-Argentines were generally considered, and thought of themselves, as non-colored, making the question of their African heritage a moot point for all practical purposes. It is perhaps important to remember here that "passing" was everywhere and always an individual, minoritarian avenue for "escape", perhaps subject to revocation by a racist white elite in critical social situations. Moreover, blacks were often simply misregistered as "non-colored", making censal figures always somewhat suspect. However, the general pattern of whitening amid racist but flexible norms remains wherever one looks in Spanish America. For an excellent study of white supremacist ideas among a biologically mulatto but socially "white" Spanish American elite see H. Hoetink, The Dominican People, 1850-1900 (John Hopkins University Press, 1982; translation of the Spanish edition of 1972), also his article in Morner (ed.), Race and Class..., pp. 96-121, on "Stratification, Immigration, and Race".
16. Andrews, pp.187-88. The paper later mentions an Afro-Argentine marrying a Basque woman, implicitly a form of gender revenge vis a vis black women in union with Italians.
18. Estratificación socio-racial ..., pp. 48-9 for the 1777-1778 census tabulations.
19. The manuscript citations for La Puebla's censuses are: Archivos Nacionales de Costa Rica (ANCR), Complementano Colonial, No.3629(1820); Gobernación, No.24906 (as part of Cartago province in 1844). The 1864 aggregate returns were published by the Dirección General de Estadística y Censos.
20. The child/woman ratios of La Puebla in 1820 were among the lowest reported. There were approximately 245-255 children aged 0-4 per thousand women, and 500-525 such children per thousand women aged 15-44.
21. The highest child/woman ratios obtained in the western provinces to which Cartago residents, La Puebla included, were then migrating. Thus, the interprovincial rather than the intraprovincial comparisons are more revealing of actual patterns of change society-wide.
22. Andrews, p.72.
23. Andrews, ibid.; also p. 37 where he cites (note 46) a British contemporary who reasoned that black women's practice of taking their children to work with them (as washerwomen on the marshy riverbank) was a major cause of high infant mortality.
24. "Costa Rica Before Coffee .~.", p.150. In the wives' age groups of 20-24, 25-29, and 30-34 in the city of San José, laborers' wives averaged 17.8 per cent fewer children under 5 than farmers' wives, and those without capital worth 21.5 per cent fewer children than those declaring capital. In the surrounding villages laborers' wives reported 12.2 per cent fewer such children than farmers' wives. Wives' ages were estimated for each group by subtracting 5 years from the husbands' age.
25. Cleto González Víquez, Apuntes estadísticos sobre la ciudad de San José, (Imprenta de Avelino Alsina, San José, Costa Rica, 1905), p.29. This rare pamphlet was located in the Latin American collection of the University of Texas at Austin. Therein (p.10) the author, and later President of Costa Rica, reported that the southern district of "El Hospital" produced only 1/5 as many children as the most prolific northern district, proportionate to its population, likely as a consequence of the same skewed age and sex distribution analyzed above. This led him to speculate upon the role of high illegitimacy (10 per cent higher in the South), feminine predominance and widespread prostitution in this particular district as a partial cause of such low fertility per capita. Combining both southern districts for comparison with the two northern wards, they reported 152 infant (under 5 years) deaths per thousand, as against 98 for the northern half of the city, or over 50 per cent higher infant mortality!
26. For Buenos Aires in 1822, 50.5 births per 1,000 Afro-American women, compared to 48.3 among the non-colored. By 1837 the figures were 53.9 for Afro-Americans and 50.8 for the non-colored. Andrews, p.73, citing Marta B. Goldberg, "La población Negra y mulata de la ciudad de Buenos Aires, 1810-1840", Desarrollo Económico, 16 (April-June, 1976), p.87, note 36, and p.95.
27. After standardizing for age distribution the figures were..
Number of children (0-4)/1,000 wives (15-49):
Whites Colored Whites Colored
(N=353) (N=42) (N=798) (N=62)
639 376 640 288
One plausible explanation could be based upon disproportionate conscription and mortality among Afro-Americans in post-Independence warfare, thus leading to prolonged male absence and lower fertility. However, radically higher infant mortality could reconcile higher Afro-Argentine married fertility rate (note 26) with a lower child/woman ratio (above and note 22), without any logical inconsistency.
28. Whites would actually appear to have been the least fecund, with only 22 per cent (Cartago) and 25 per cent (San José) of its population in the "párvulo" category, compared to 33-34 per cent for mestizos and 30-38 per cent for "mulatos y negros". Estratificación socio-racial ..., p.48. However, since it may well be that classification as "white" was partially a function of age, wealth, and social status, the older the population the more likely it could be assigned to this category. Thus, the figures should be used with great caution.
29. The figures for Cartago Province and La Puebla were as follows:
1,000 children 0-4/wives aged:
20-24 25-29 30-34
Cartago Province, 1844 1,149 1,210 1,078
La Puebla, 1844 895 1,208 1,536
La Puebla, 1820 1,078 1,385 1,056
So few cases in La Puebla caution against any firm conclusion regarding comparative marital fertility levels.
30. Andrews, chap. 7 "The Black Legions". The devastating impact upon Afro-Americans of the Paraguayan campaigns (1865-70) and the yellow fever epidemic of 1871 is discussed on page 91.
31. Szuchman's data also suggest more frequent Afro-American widowhood, likely owing to male conscription and disproportionate mortality in the war years following Independence in 1810. Cf. note 27.
32. Andrews, p.66.
33. Ibid., p.92.
34. Andrews, p.75, argues that urban to rural migration would have been negligible for Buenos Aires, but the Costa Rican population was rapidly redistributed toward the periphery by coffee, heightening race mixture and accentuating the lagged replacement capacity of urban Afro-Americans. In effect, any urban or semi-urban population migrating from Cartago west to Alajuela would not only find an overwhelmingly mestizo local population, but also one in which average age at marriage for women was significantly (2-3 years) lower.
35. Consider the censal figures discussed earlier for Cartago and San José and the frequency of interracial unions in each case (note 13).
36. Age at marriage for women was approximately l9-20 for village women and 21-22 for city women. The village average may have declined slightly over the period as well. The child/woman aged 15-44 ratio was consistently highest for Alajuela, on the western fringe, throughout the 19th century.
37. C.f. note 24.
38. It goes without saying that the modern Afro-American population of West Indian origin in Costa Rica is not dealt with herein. This population, until recently concentrated in the Atlantic coast province of Limón, has experienced increasing assimilation through race mixture as interprovincial migration has increased. In Limón, the banana export economy of the early 20th century led to extreme male predominance and, thus, to a low replacement capacity. However, due to linguistic, cultural, and religious differences and the greater relative isolation and autonomy of contemporary black Costa Rican groups, assimilation has proceeded much more slowly and not without criticism of the racism implied by "whitening". Thus, the similarities with colonial and nineteenth century experience should not be overstated. For Limón and its West Indian populations see Carlos Meléndez and Quince Duncan, El negro en Costa Rica (Editorial Costa Rica, San José, 1972); Jeffrey Casey, Limón, 1880-1940 (Editorial Costa Rica, San José, 1979); Michael Olien, "The Negro in Costa Rica: The Ethnohistory of an Ethnic Minority in a Complex Society" (Ph.D. dissertation, University of Oregon, 1967); Roy Simon Brice La Porte, "Social Relations and Cultural Persistence (or Change) Among Jamaicans in a Rural Area of Costa Rica" (Ph.D. dissertation, University of Puerto Rico, 1962); and Paula Palmer, "What Happen": A Folk-History of Costa Rica's Talamanca Coast (Ecodesarrollos, San José, 1977).
*I would like to thank Thomas Skidmore, George Reid Andrews, and Leslie Rout for their hefpful comments upon an earlier version of this paper. My thanks as well to Mark Szuchman and Héctor Pérez for access to their research data. An earlier version of this paper was read at the Southwest Historical and Social Science Associations' meeting in Ft. Worth, Texas, March 21-24, 1984.