Feudalism, Capitalism, and the World-System in the Perspective of Latin America and the Caribbean: "Ever More Solitary"

by Steve J. Stern
Feudalism, Capitalism, and the World-System in the Perspective of Latin America and the Caribbean: "Ever More Solitary"
Steve J. Stern
The American Historical Review
Start Page: 
End Page: 
Select license: 
Select License

Reply: "Ever More Solitary"

DIRECTDEBATE BETWEEN SCHOLARS CAN SERVE TO ENHANCE understanding when the contending parties resist the temptation to talk past one another rather than engage the alternative viewpoint. For this reason, I welcome debate of my assessment of Immanuel Wallerstein's world-system intepretation and my pro- posed alternative from the angle of Latin American and Caribbean history. Wallerstein's commentary is so riddled with errors and misrepresentations, however, that it does not facilitate the kind of constructive debate that advances knowledge. A point-by-point rebuttal of each such item would serve little purpose for attentive readers and would exceed the space allotted to this brief rejoinder. I will therefore limit myself to illustrating the profusion of errors and misrepre- sentations in the case of Wallerstein's comments on silver. I will then turn to his comments on sugar as an example of the reductionist method that makes his analysis one-dimensional and mars its explanatory value. Finally, I will take up the broader issues at stake in the debate.


DISCUSSION statements and presentations of data the better to fit his thesis. Let us observe, for example, what he does to the chronology of share arrangements at Potosi. In his rendition of my presentation, the second phase of labor relations (from the 1570s to around 1600) was "pure mita,"' and the ore shares of the mitayo draft laborers were insignificant.' Only in the third phase of labor relations (the 1600s) did there develop in Potosi a mixed system of forced and voluntary labor. By Wallerstein's version of my own account, however, not until the late seventeenth century were the minga volunteer laborers "reasonably successful"3 in their efforts to assert customary share rights. Under the stimulus of "an expanding world market"4 beginning in the mid- eighteenth century, the mineowners reacted effectively by intensifying the work regime imposed on mitayos in the late eighteenth century.

' Immanuel Wallerstein, "Comments on Stern's Critical Tests," AHR, 93 (October 1988): 875.

Wallerstein bases his dismissal of the share practices of forced laborers on my comparison of the relative power of forced and voluntary laborers to assert share rights. See Wallerstein, "Comments on Stern's Critical Tests," 874, 875; Stern, "Feudalism, Capitalism, and the World-System in the Perspective of Latin America and the Caribbean," AHR, 93 (October 1988): 852-53.

" Wallerstein, "Comments on Stern's Critical Tests," 875, compare 874.

Wallerstein, "Comments on Stern's Critical Tests," 876.


Stern's Reply 887

This rendition of the chronology presented in my essay is mistaken on every crucial point. The second stage of labor relations, I observed, was not "pure mita" but also witnessed the rise of "a more spontaneous system of voluntary hiring." I even went so far as to suggest that the minga volunteers, certainly a majority of the work force in Potosi's mines and refineries by the 1600s, "perhaps" reached majority proportions earlier.5 The mixed labor system and the share customs accompanying it had therefore already developed by the late 1500s. Moreover, Wallerstein ignores the evidence that even the mztayos established customary share rights that mineowners could not eliminate and that the shares constituted their most important remuneration.6 This effort to slight the impact of share arrange- ments equally distorts Wallerstein's summary of the third phase. In Wallerstein's rendition, the mingm' efforts were not "reasonably successful" until the late seventeenth century. Actually, I pointed out that the corpa ore shares claimed by Indians loomed large in the picture as early as the 1580s. At that time, they accounted for about a fourth of total ore production, enabled Indian women and men to organize their household economies around the production, smelting, and trade of corpa ores, and triggered an abortive effort by the mineowners to eliminate the Indian ore market.' What happened sometime during the late seventeenth century, when the mingm added weekend mine raiding to their established practices, was a further expansion of a share system that had already operated with reasonable success (from the mingas' perspective) for about a cent~ry.~

Finally, while it is true that mineowners responded to this extreme situation, after a delay of at least half (perhaps three-quarters of) a century, by brutally intensifying the labors imposed on mitayos, and that this response provided the crucial marginal difference in Potosi's profitability in the late eighteenth century, it is also true that customary share rights nonetheless continued to deprive mineowners of well over a third of the refined silver they might otherwise have claimed.9 The latter point drops out of Wallerstein's chronology.

The net effect of Wallerstein's distortions of the chronology I presented is to compress severely the duration in time of share arrangements that had a great impact on labor relations and silver production at Potosi. This is why, in addition to misrepresenting my depiction of the second and third stages of labor at Potosi, Wallerstein is also eager to claim that I overstate the impact of share relations in the first phase (1545 to the 1570s).l0 This compressed chronology of share

'Stern, "Feudalism, Capitalism, and the World-System," 851.

'See Stern, "Feudalism, Capitalism, and the World-System," 852.

Stern, "Feudalism, Capitalism, and the World-System," 853.

"Expansion" may be slightly misleading insofar as the declining quality of ore may have reduced the benefit of "traditional" share customs in the late seventeenth century and therefore driven the mzngns to compensate for this erosion by pressing hard for "new" customary rights that granted them even more systematic opportunities to find and appropriate good pieces of ore.

Stern, "Feudalism, Capitalism, and the World-System," 854, and note 65.

lo To cite in response to my argument, however, Don Silvio Zavala's classic essays, published in 1943, when the historiography focused more on legal rules, juridical philosophy, and policy intent than on the on-the-ground social relations that have preoccupied historians since the 1960s, is a rather weak way to question the analysis of labor relations at Potosi before the 1570s. Nor does it help Wallerstein's case when my discussion of yanacona and encomzenda laborers in the early years is reduced, in his

relations, which decrees them insignificant during the first 125 years or so (1545 to, say, 1670) of Potosi's colonial history, neatly correlates, in Wallerstein's view, with the arguments of The Modern World-System. The mineowners "permitted" sharecropping in the late seventeenth century to minimize risks as the world market price for silver declined, then, under the incentive of world market expansion, intensified forced labor to reverse the damage." Small wonder that Wallerstein cannot see how the data on share relations contradict his interpretation ''.in any way."l2

The problem of chronology hardly exhausts the list of distortions, errors, and key omissions that pervade Wallerstein's comments on silver. A short sample will suffice to illustrate the depth of the problem. In Wallerstein's account, relative statements are transformed into hard-and-fast absolutes. Thus "overlapping" phases of mine labor become more starkly delineated,I3 and a division of labor that "tended" to concentrate mitayos in more primitive or repugnant tasks and mingas in more "skilled" and rewarded work is hardened into a purer, more rigid formulation.14 In Wallerstein's account, a somewhat more refined and modulated version of his tripartite division of international labor appears,Ij and having agreed that each region of the world-system presents a complex mix of labor relations, he asks us to look "for emphases, for the appearance of more rather than less." In contrasting the mixes in different zones of the world-economy, he observes triumphantly that "even Stern does not claim that most of the workers derived most of their income from sharecropping in Potosi."l6 The only problem is that I do make this very claim.17

response, to a discussion of encomienda Indians only. See Wallerstein, "Comments on Stern's Critical Tests," 874-75.

"Wallerstein, "Comments on Stern's Critical Tests," 876.

l2 M'allerstein, "Comments on Stern's Critical Tests," 876.

l3 Compare Stern, "Feudalism, Capitalism, and the World-System," 850, and Wallerstein, "Com- ments on Stern's Critical Tests," 873-74, 875.

l4 Compare Stern, "Feudalism, Capitalism, and the World-System," 851, and Wallerstein, "Com- ments on Stern's Critical Tests," 874,875. For Wallerstein, the correlation of "skilled" work with waged or free labor forms, and "unskilled" work with coerced labor, squares well with his own thesis. (His comments on the point downplay, however, the geographical thrust of this distinction in the Modern World-System volumes. See especially The Modem World-System: Capitalist Agn'culture and the Origzlls of the European World-Economy in the Sixteenth Centu~y [New York, 1974],87.)Wallerstein takes the distinction between skilled free labor and unskilled coerced labor so rigidly that he hypothesizes that Spaniards may have performed, in the late 1500s, much of the "skilled" labor later performed by minga volunteers. This was not the case. In the silver tunnels of Potosi mountain, distinctions between "skilled" and "unskilled" work referred mainly to distinctions between ore-cutting and ore-hauling work, the labor force was overwhelmingly Indian, and the division of labor between mitayos and mingas was blurred by the overlap between them.

'' This version is presented by M'allerstein not as a refinement of his views in light of subsequent thinking or research but as a restatement of them.

'' Wallerstein, "Comments on Stern's Critical Tests," 876.

l7 See Stern, "Feudalism, Capitalism, and the M'orld-System," 852-55passim. Along the way to this conclusion, Wallerstein states that yeoman farmers and sharecropping were insignificant in the colonial agriculture of Peru and Bolivia. The evidence is more mixed than this, depending on the time period, in some critical agricultural regions. In the valleys of Peru's central coast, small-scale farms and truck gardens (chicaras)were at first important in the provision of foods to Lima but were then eclipsed by the growth of large haciendas in the 1590s and early 1600s. See Robert G. Keith, Conquest and Agrarian Change: The Emergence of the Hacienda System on the Peruvian Coast (Cambridge, Mass., 1976),esp. 64-72, 8 1-97; compare Eduardo Arroyo, La Hacienda costelia e9a el Perk Mala-Caliete, 1532-1 968 (Lima, 198 1).

Stern5 Reply 889

Wallerstein's omissions betray a similar evasion of the argument. Let us recall that, among the range of major silver centers of colonial Spanish America, Potosi constituted the example most likely to support Wallerstein's thesis. Elsewhere, coerced labor gave way more rapidly to private arrangements in which the share system predominated.18 This aspect of the argument Wallerstein ignores alto- gether. The multiple (mis)representations of texts may delight deconstructionists but not readers interested in meaningful debate.19 Wallerstein simply seems unprepared to engage the crux of the argument. If, in the mix of labor relations at Potosi, share arrangements were dominant, if Potosi is nonetheless the most favorable "test" of the Wallerstein thesis among colonial silver centers, if silver was the highest American priority of the capitalist world-system during the "long" sixteenth century-if all this holds true, what does it imply about the tripartite international division of labor at the foundation of Wallerstein's Modern World- System, and about the explanatory power of his world-system paradigm?zO Surely, not that the capitalist world-system was irrelevant. But, just as surely, that it constituted only one of several great "motor forces" driving the history of labor and economy in colonial America.

LETUS TURN FROM A NARROW AND TECHNICAL DISCUSSION of "mistakes" to the larger question of method. Wallerstein's commentary on sugar serves well as an example of his penchant for reducing all significant phenomena bearing on the colonial economy to consequences derived from the world-system. The case of sugar is particularly interesting since Wallerstein and I are in agreement on much of the descriptive data to be explained, and his specific analysis of this data is more sensitive to the significance of social conflicts and conditions of production in America for an emerging labor system.

Wallerstein agrees that for colonial sugar planters in the greater Caribbean region (including Portuguese Brazil), the Amerindian populations proved insuf- ficiently exploitable, eventually, compared to African slaves. He agrees that epidemic disease as well as the Indians' violent resistance (in the case of Brazil) are

In Cochabamba, on the other hand, the strategic "granary" of Potosi, the early might of commercial haciendas was gradually worn down in the eighteenth century by peasant smallholders whose land rentals, sharecropping tenancies, and direct ownership of land enabled them to compete actively in the food market. See Brooke Larson, Explotacidn agraria y resirtencia campesina: Cinco ensayos histdricos sobre Cochabamba (siglos XVI-XIX) (Cochabamba, 1982); Larson, ColonialOm and Agrarian Transformation in Bolivia: Cochabamba, 1550-1900 (Princeton, N.J., 1988).

Stern, "Feudalism, Capitalism, and the World-System," 855-57.

l9 Unfortunately, errors and misrepresentations pervade Wallerstein's entire commentary, not just the discussion of silver (which I am using as an example). I take this opportunity to add, in fairness to M'allerstein, that it would be extremely regrettable to conclude from the errors marring his "Comments" that the Modern World-System books are unworthy of serious attention and reflection. As argued in my original essay, Wallerstein's work raises provocative and weighty issues and contributes specific historical and theoretical insights whose value should not be overlooked even if one concludes that the general paradigm is fundamentally flawed. Were his work not important and provocative, I would hardly have relied on a critical evaluation of it as a basis for reappraising colonial political economy in its wider world context.

'O The three patterns within the tripartite system "make sense primarily if we realize that they serve to maximize capital accumulation in the world-system as a whole"; Wallerstein, "Comments on Stern's Critical Tests," 877.

relevant to understanding the transition from Indian to African labor strategies. But his characterization of the issue of Amerindian vulnerability to epidemic disease caused by the spread of Old World microbes is revealing. To Wallerstein, "dying out" was a sociopsychological response by which hunter-gatherer cultures rejected the "disciplined agricultural labor" imposed by the world-system, and to which they were especially "inadapted." African slavery was unimportant, he points out, in the former Aztec and Inca territories (with their dense peasant populations and highly developed agriculture), and the background of imported West Africans as "settled agricultural peoples" was a significant factor in their suitability as slaves. "In the last 400 years," Wallerstein observes, "this process of 'dying out' has occurred wherever the capitalist world-economy has expanded into a zone inhabited by hunters and gatherers and tried to use them for disciplined agricultural work . . . It is a world-systemic pattern.""

One detects here a rather strenuous effort to subsume the explanatory value of biological and epidemiological history under the rubric of the effects of a single higher force, capital accumulation by the world-system. While it is true that abortions, infanticide, suicide, and a loss of will to live figured in the sociopsy- chological response of some Amerindians (as well as African slaves!) to plantation labor, and that this response is significant to understanding population history, Wallerstein's way of posing the problem relegates the history of microbes, disease environments, and the biological contact and isolation of human populations to a minor and derivative explanatory role, something less than a force in its own right. The explanatory motor that subsumes biological and epidemiological factors is the world-system's labor demands, to which unsuitable hunter-gatherer cultures responded by "dying out." This approach presses historical causation and explanation into a single dimension: vulnerability to epidemic disease becomes "resistance"22 to plantation labor by inadaptable natives.

The formulation, however, does not work. The Arawak and Tupinamba peoples of Espafiola and coastal Brazil were not primarily hunters and gatherers; they practiced agriculture, albeit an agriculture whose slash-and-burn techniques perhaps facilitated physical mobility in a tropical environment.23 Moreover, the great sixteenth-century epidemics almost always unleashed greater devastation on the indigenous population of the coastal lowlands of Latin America and the Caribbean compared to those of the high mountains and plateaus, even when Europeans did not impose disciplined plantation regimes in the lowland areas.24

" M'allerstein, "Comments on Stern's Critical Tests," 877, 879.

''Wallerstein, "Comments on Stern's Critical Tests," 879.

23 See Mary W. Helms, "The Indians of the Caribbean and Circum-Caribbean at the End of the Fifteenth Century," in Leslie Bethell, ed., The Cambridge Histo9 of Latin America, volume 1 (New York, 1984), 49-52; Carl Ortwin Sauer, The Early Spanish Main (Berkeley, Calif., 1969), 51-58; Stuart B. Schwartz, "Indian Labor and New World Plantations: European Demands and Indian Responses in Northeastern Brazil," AHR, 83 (February 1978): 44-46. Whether one may refer even to a "semi- nomadic" lifestyle is uncertain. Sauer, Early Spanish Main, 52, argued that in the Greater Antilles shifting cultivation was unlikely.

24 The best introduction to the highland-coast distinction and sixteenth-century population history is Sherburne F. Cook and M'oodrow Borah, Essays in Population his to^: Mexico and the Caribbean, 2 vols. (Berkeley, Calif., 1971, 1974), esp. 1: 73-1 18, 41 1-29. The broad biological picture is admirably

Stern's K epb 891

This helps to explain why, for example, the Afro-American population of Lima and nearby coastal valleys constituted by the late sixteenth century a third or more of the region's total population, despite the absence of a plantation regime and despite an indigenous background of peasant agriculture on Peru's central coast.25

In short, the vulnerability of indigenous peoples to epidemic disease was a powerful force in its own right in the coastal lowlands, independent of the spread of sugar plantations or the cultural aversion of hunters and gatherers to disciplined agricultural labor.26 The point here is to illustrate Wallerstein's reductionist method and its limitations, not to argue that vulnerability to epidemic disease was the only important local variable. Local geography, conditions of health and work on the plantations and placer mines, and the Indians' cultural background and responses, including resistance, are extremely relevant to a full analysis of indigenous population decline and movement in the greater Caribbean and to their impact on plantation labor.27 Nor do I wish to imply the unimportance of the world-system. The "interplay," I argued, "between.. . local conditions of pro- duction and the interests and opportunities derived from the international market" accounts more powerfully than the world-system paradigm alone for the emergence of African slavery as an overwhelmingly dominant labor strategy on the plantations of the greater Caribbean.28

sketched in Alfred W. Crosby, Jr., The Columbian Exchange: Biologcal and Cultural Cor~sequences of 1492 (Westport, Conn., 1972); compare his Ecological Imperialism: The Biological Expansion of Europe, 900-1900 (New York, 1986).

25 I base this (conservative) estimate on Frederick P. Bowser, The African Slave in Colonial Peru, 1524-1650 (Stanford, Calif., 1974), 337--11,76-78,9 1;Noble David Cook, Demographic Collapse: Indian Peru, 1520-1620 (New York, 1981), 150-51, 156-57. To exclude coastal river valleys with numerous sugar and wine haciendas relying heavily on African slaves, I am using a restricted definition of "central coast" (the districts of Lima, Cafiete, and Chancay only) that omits the northern and southern reaches of more expansive definitions.

"Let us note, in addition, that the distinction between the "dying out" of indigenous "hunters and gatherers" and the adaptability of West African agriculturalists is tempered for another reason as well. In the Antilles and Brazil, local conditions of health and diet, work, and domestic and sexual life were such that the Afro-American slave populations had great difficulty reproducing their numbers through natural increase. It was fresh imports that enabled the Afro-American slave populations to remain stable or grow.

"As Wallerstein observes, my essay devotes little attention to the colonial British experience with white indentured labor. I preferred (in the limited space of a single, albeit lengthy, article) to focus the detailed analysis on Spanish and Portuguese America because I am writing as a scholar of Latin America, and because the Iberian experience with sugar in America preceded the British experience. My comments on British America were therefore limited to showing the consistency of the British experience with the argument developed for Spanish and Portuguese America. See Stern, "Feudalism, Capitalism, and the World-System," 862-63. There is much truth in the cost-benefit and supply- demand analyses quoted in Wallerstein, "Comments on Stern's Critical Tests," 877-78, but for a splendid work that demonstrates the relevance, too, of local social and political conflicts not easily reducible to calculations of the comparative cost of labor, see Edmund S. Morgan, American Slavery, American Freedom: Tlze Ordeal of Colonial Virginza (New York, 1975).

28 Stern, "Feudalism, Capitalism, and the World-System," 862. The mirror image of Wallerstein's reductionist tendency is his belief that he must "come to the defense of the world market" in the explanation of the transition to African slave labor in sugar production; Wallerstein, "Comments on Stern's Critical Tests," 880. The defense is unnecessary, since my own analysis is careful to point out the importance of the world market and the dangers of a reverse reductionism. (See Stern, "Feudalism, Capitalism, and the World-System," 860, 861-63.) But the defense is understandable if one wants to grant the world-system a primacy and inclusiveness that subject other variables to a secondary and derivative status.

For Wallerstein, this "interplay" is an illusion because the forces I call "local" are


creations, or consequences, of the world-system itself. Conditions of health and mortality derived from the integration of the greater Caribbean into the capitalist world-economy that imposed plantations in the area. The proximity of poorly controlled frontier zones providing refuge to popular resistance is a function of world-system geography: "a 'frontier' is by definition a phenomenon of a 'world-system."' The real and feared effects of popular resistance may have mattered, but it is "the structure of the world-system" that explains why Indian resistance was more successful than that of African slaves. "In what sense," Wallerstein asks, "are any of these conditions 'local'?"29

The question is a fair one, but I believe that Wallerstein's answer to it rests on two mistaken judgments. First, Wallerstein seems unable to accept the possibility that even if, from the point of view of "origins," inclusion in the world-system dramatically altered the disease environments, social relations, and power conflicts of colonized American regions, these (changed) conditions of life and labor could quickly take on a dynamic or life of their own. The conditions of health, mortality, and physical abuse that framed life and options in the greater Caribbean may have had as much to do with epidemic disease crises beyond world-system control, and with conflicts and behaviors rooted in the antagonistic dialectic of social control and conflict on the plantations,30 as with social designs decreed or spread from the core of the world-system. The human geography of frontiers, and the uneven spread of social control, are not reducible to a mere phenomenon of the world-system. Human beings coping with life in America had an important hand in creating and defending "frontiers," in rendering physical environments "more" or "less" accessible to the forces of social control.31 The limits and complications imposed by Indians and by European indentured laborers on the early labor options of aspiring planters derived not simply from the structure of the world-system but also from the turbulence and uncertainties of the lives they made in America. The real and latent threats of Indian and European laborers heightened the relative attraction of slaves imported from across the sea (although the latter, too, stirred

29 All quotes are from Wallerstein, "Comments on Stern's Critical Tests," 879.

30 For example, given the realities of poor health and overwork on plantations of northeastern Brazil in, say, 1560, an Indian (or African) laborer might have an interest in feigning or exaggerating illness. But a plantation owner or manager, eager to resist signaling weakness of will by accommodating such "illnesses" too easily, and conscious that the insecurities of disease and mortality made long-term perspectives on health risky anyway, might have an interest in slighting all but the most serious health problems or even punishing those suspected of feigning sickness. This sort of dialectic could hardly be expected to improve the laborers' health, even if the long-term interest of elites at the level of systemic needs or benefits lay in such improvement.

3' The south of Chile, with its gold mines, offers a particularly fascinating and dramatic case of human struggles to define, control, and conquer frontiers. For splendid analyses, see Robert Charles Padden, "Cultural Change and Military Resistance in Araucanian Chile, 1550-1730," Southwestern Journal of Anthropology, 13 (Spring 1957): 103-21; Alvaro Jara, Guerra y sociedad en Chile: La Traruformacidn de la guerra de Arauco y la esclavitud de 10s indios (Santiago de Chile, 1971); Andrew Daitsman, "The Dynamic of Conquest: Spanish Motivations in the New World" (M.A. thesis, University of Wisconsin-Madison, 1987). On the arduous efforts of runaway Afro-American slaves to make frontier hideaways more "remote" from social control, see Richard Price, "Introduction: Maroons and Their Communities," in Price, ed., Maroon Societies: Rebel Slave Communities in the Americas (2d ed., Baltimore, Md., 1979), 5-8.

Stern's Reply 893

fears and brought complications). America-centered dynamics such as these bore on the productivity of labor and the limits of social control, thereby exerting a powerful reciprocal impact on the options, constraints, and opportunities of the world-system and its agents in America.

Wallerstein might object that the world-system's impact on American life and dynamics did not stop at the point of "origins." If the world-system continuously framed the American conditions I call "local," would not "interplay" again turn into an illusion? Such an objection would rest on a second mistaken judgment, distinct but related to the first. Wallerstein seems to me to confuse the geographical reach of a relevant framework with its power of historical determination and explanation. This leads him to reject the possibility of a dynamic "interplay" between local conditions that provided a context, or framework, within which world-system actors operated, and a world-system which simultaneously set a context, or framework, for local actors. For Wallerstein, one framework subsumes the other.32 The spatial reach of a relevant framework, however, does not prove its all-embracing power as the relevant framework for understanding human history. Such an equation would be a bit like saying that, because the structure of the solar system sets the environmental and ecological framework of human life on earth, it follows that "local" human actions on earth have no independent impact on the environmental and ecological framework of human life on the planet.

LET US CONCLUDE WITH THREE BROAD QUESTIONS of "mythistory," a topic Wallerstein rightly and perceptively introduces. I want first to ask whether the charge of "reductionism" is inherently unfair, since we all of necessity engage in it. I will then consider the problem of continuity and change raised by Wallerstein. Finally, I will introduce what is perhaps the most important "mythistorical" issue at stake in the debate: the historical agency and "solitude" of Latin Americans.

Wallerstein (citing William H. Mch'eill) points out that we humans recognize historical, social, and natural patterns by selecting among the endless streams of data or stimuli rushing toward our consciousness. The point is valid, and in this larger sense we are all "reductionists." Selectivity and simplification are essential to survival and intelligence. If, in Mexico City street traffic, I give undue attention to air traffic and sidewalk vendors, my vehicle and I will not last long. Historians and social scientists, like urban street drivers, must select what merits more, less, or no attention to discern the relevant patterns and relationships. "Reductionism" in this broadest and most general of senses is essential to the historian's task and therefore an unfair measuring stick for sizing up a historical interpretation. We must turn to a narrower, more refined sense of the term. The urban street driver necessarily engages in "reductionism" to survive. But if, in Mexico City street traffic during the rainy season, I pay scant attention to cross-street traffic when the traffic light on my street is green, I am headed for trouble. I am unaware of, ignoring,

32 "The world-system is neither an 'actor' nor a 'factor.' It is the whole historical framework within which the actors act and the factors exist"; Wallerstein, "Comments on Stern's Critical Tests," 881.

or bending into a wishful dimension the relevant knowledge that the street lights are not reliable during the rains and that the city's drivers do not always defer to the lights, especially when the lights are unreliable. When "reductionism" ignores or twists relevant data or patterns into a misleading analysis of the topic at hand, it may become a strategy for disaster, not survival.

This narrower, more precise and restricted meaning of "reductionism" is the context for my critical evaluation of Wallerstein's method and intepretation. Reductionism in this narrower sense occurs when a historian or social scientist is unaware of or ignores data, or twists its analysis into a single "mold" or dimension, in ways that fundamentally distort or impoverish our ability to understand or explain the problem under study. Critical evaluation is fairer and more telling if one examines the key building blocks of the argument in its own terms. For this very reason, my analysis focused on silver and sugar rather than minor exports, and the study of silver focused more on Potosi (the "most favorable case" among the major silver centers) than on, say, Oruro or Zacatecas. For this reason, too, my analysis of the American periphery focused on the tripartite international division of labor that is at the foundation of Wallerstein's edifice rather than, for example, on the details of European politics and administration in America. The "reductionism" of a historical interpretation matters little unless one demonstrates how new data or a distinctive line of analysis affects the essential substance or reasoning concerning topics vitally important to the original interpretation. Readers will decide for themselves, of course, whether my evaluation of the silver and sugar sectors is convincing and whether Wallerstein's "reductionism" is in that context a valid characterization.33

The relative priority one places on the study of historical continuity or change is closely related, in Wallerstein's view, to questions of "mythistory." Both concerns are "fundamental," but continuity-the discovery of similarity across time-"is harder and takes priority."34 It seems to me rash to place continuity and change, a priom, in rank order of intellectual priority and complexity. The relative importance of each, and the periodizations historians find most useful, depends on the issue or question under examination.35 For Latin America, at least, the

33 Because, in my view, the perception of "truths" may require a certain reductionism, and even the most flawed and one-dimensional models may expose or illuminate one or several "truths" among many, I also thought it important to consider whether Wallerstein's interpretation contributes valuable specific insights despite my rejection of the general paradigm. Readers will decide for themselves, too, the merits of the specific insights I highlighted from Wallerstein's work as especially illuminating for understanding the feudalism/capitalismIworld-systemtrilogy in an American context.

34 Wallerstein, "Comments on Stern's Critical Tests," 881.

35 In a classic article, Joan Kelly-Gadol demonstrated, for example, how shifting the focus of analysis to women recasts our sense of historical periodization and change: "Did Women Have a Renaissance?" reprinted in Renate Bridenthal and Claudia Koonz, eds., Becoming Visible: U'omen in European History (Boston, 1977), 137-63. In a recent study, I argued that to interpret peasant rebellions scholars need to incorporate multiple time scales simultaneously into the analysis, since both long-term ("structural") and short-term ("conjunctural" and "episodic") perspectives on continuity and change are essential to understanding why and when peasant rebellions happen and what such volatile moments mean to the rebels. From this perspective, a ranked choice between "continuity" and "change" seems too confining. See Steve J. Stern, "New Approaches to the Study of Peasant Rebellion and Consciousness: Implications of the Andean Experience," in Stern, ed., Resistance, Rebellion, and Cor~\ciousness in the Andean Peasant U'orld, 18th to 20th Centuries (Madison, Wis., 1987), 11-13.

illusion of false or superficial continuities has been as alluring as the illusion of false or superficial changes. One can see the grip of a sense of continuity on many levels-from the early twentieth-century intellectuals who saw the landed estates (haciendas) of their day as direct institutional descendants of the sixteenth-century encomiendm;3G to Alejo Carpentier's brilliant novel depicting a journey into the Venezuelan provinces and jungle as a journey backward in time;37 to contempo- rary social historians who observe that the Wars of Independence, violent and destructive as they were, left fundamental social structures intact;38 to tourist brochures extolling the folklore of exotic Indians living much as their pre- Columbian or colonial ancestors did centuries earlier.39 This sense of continuity is not, of course, a mere product of imagination, and it exposes something important about Latin America. To sort out, ponder, and analyze the meaningful continuities as well as changes in Latin American and Caribbean history is important and difficult, but I see no reason to reach a sweeping judgment, independent of the specific issues under study, of their relative importance.

Continuity, change-the perspective of each brings its own drawbacks and traps. Too firm and overriding a commitment to continuities may lead to questionable projections in time and casual juxtapositions of distinct historical eras. Note, for example, Wallerstein's response to my comment that a long-term view poses a theoretical paradox to the capitalist world-system intepretation of early colonial America because, in the early nineteenth century, some agrarian regions that had been dynamic in colonial times experienced an apparent "regression" into feudalism. Wallerstein's response is to quote at length his analysis of colonial haciendas and economic patterns in the seventeenth and eighteenth centuries.40 The analysis he quotes is perceptive (and I share his rejection of the feudalism thesis for the seventeenth and eighteenth centuries), but it is irrelevant. More precisely, the pertinence of this sort of juxtaposition is at best an open question unless one assumes, apriorz, that underlying structural continuities are so great that the analysis of one era is applicable to the other. Similarly, Wallerstein refers to the power of the world-system in the late twentieth century to buttress his view of its power in the early modern period. "It was no accident," he tells us, that labor unions in the 1970s were more formidable in countries such as Germany than in

36 This is why it became an important historiographical accomplishment to demonstrate that the encomienda did not confer legal land title upon encomenderos,and that the waves of hacienda expansion and consolidation in Latin America occurred after the age of the powerful encomienda had passed. For a thoughtful discussion of the encomienda-hacienda tangle, see James Lockhart, "Encomienda and Hacienda: The Evolution of the Great Estate in the Spanish Indies," Hispanic American Historical Review, 49 (August 1969): 41 1-29.

"70s Pesos perdidos (Mexico City, 1953).

38 See, for sophisticated examples, John Lynch, The Spanish American Reuolutiom, 1808-1826 (New York, 1973); George Reid Andrews, "Spanish American Independence: A Structural Analysis," Latin American Perspectix~es, 44 (Winter 1985): 105-32.

39 Anthropologists will recognize more scholarly and sophisticated examples of this sense of overwhelming continuity in indigenous life in the community studies removed from a sense of broader space-time contexts, and in the "ethnographic present" approaches to the indigenous past, that were once quite common. A masterful critique of this tradition by an anthropologist is Eric R. Wolf, Europe and the People without History (Berkeley, Calif., 1982); see also Padden, "Cultural Change," 103-21.

40 See Stern, "Feudalism, Capitalism, and the World-System," 866-67; Wallerstein, "Comments on Stern's Critical Tests," 882-84.

countries such as Malaysia and that German manufacturers would therefore move factories to Malaysia.41 Suppose, for the sake of argument, that one accepted this example and decided that the structure of the capitalist world-system was in the 1970s so decisive and powerful that it subsumes or overrides explanation based on "local" analysis of capital-labor relations in countries such as Germany and Malaysia. How would one know that the explanatory strength of the contemporary capitalist world-system may be projected with confidence across time as far back as the early modern era? The pitfalls of continuity are as great as those of change.

Perhaps the most serious difference in our contending "mythistories" lies not in our approaches to "reductionism," or to continuity versus change, but in my rejection of the consequences of Wallerstein's world-system paradigm for the historical agency and "solitude" of Latin Americans. If one may discern a "subtext" in my essay, it is that the peoples of Latin America and the greater Caribbean, even poor laboring peoples of color or humble background, have had major importance as historical agents and causes of their own experience. This agency should not be idealized or overstated, but it has not been limited to futile resistance against the onslaught of the capitalist world-system. A full analysis of such agency-its history, explanation, accomplishments, failures, and limitations-requires serious study of America-centered social dynamics and structures as well as world-system dynamics and structure. It requires, too, acceptance of the possibility that the peoples and economies of the American periphery sometimes fit into the larger world-system in unexpected, less than optimal, or even "dysfunctional" ways. A parallel subtext has been that Latin American intellectuals have often gone further than their North Atlantic counterparts (myself included) in developing the theoretical tools we need to understand the idiosyncratic history and realities of Latin American life.

This double agency, historical and intellectual, is swallowed and rendered invisible by the all-powerful world-system machine. In Wallerstein's big picture, the imprint of the laboring poor of Latin America and the Caribbean, in the early modern era, shrinks to futile resistance to the external force, on terms that do not change in any meaningful way the course of the world-system's impact on economic life and social relations in the American corner of the world. The contours of popular struggles, successes, and failures are so determined by the world-system framework that historical "agency" shrivels up. This disappearance of agency is all the more regrettable since the data and analysis now available make it possible to see a sharply different and more realistic picture. Lost in the world-system paradigm, too, is a good deal of intellectual agency. The idiosyn- crasies that make the dynamics of Latin American economic life distinctive, and about which Latin Americans have debated sharply and for which they developed sophisticated theoretical tools, are devoured by a capitalist world-system that mandates variants of capitalist labor relations within the peripheries of capitalism. The peculiarities and paradoxes, the theoretical tools, the debates fade into the background of a familiar picture.

41 Wallerstein, "Comments on Stern's Critical Tests," 880.

Stern's Reply 897

The disappearance of agency, historical and intellectual, is in my view closely related to the notion, made famous by Gabriel Garcia Marquez, that Latin Americans have been condemned to "one hundred years of solitude."4* In his 1982 acceptance of the Nobel Prize for literature, the great writer pressed his case before the Swedish Academy of Letters. For Garcia Marquez, Latin America is not solitary because it is disconnected from or unsubordinated to the wider Western world or the capitalist world-system. The solitude is rooted elsewhere: in the "outsized reality" responsible for "a lack of conventional means to render our lives believable," in the rejection of Latin America's originality "as if it were impossible to find another destiny than to live at the mercy of the two great masters of the world," in Europe's insistence "on measuring us with the yardstick that they use for themselves." For Latin America, one danger of a nearly omnipotent world- system paradigm is that, despite so much evidence to the contrary, it shrinks agency and reinforces solitude. "The interpretation of our reality through patterns not our own serves only to make us ever more unknown, ever less free, ever more solitary."43

SteveJ. Stern

42 I refer here, of course, to the novel that thrust Gabriel Garcia Marquez into international prominence: Cien arios de soledad (Buenos Aires, 1967). 43 All quotations are from Garcia Marquez's 1982 Nobel Prize acceptance speech, translated by Marina Castafieda, and reprinted in The New York Times (February 6, 1983).

  • Recommend Us