-
La Falta de Brazos
Land and labor in the coffee economies of nineteenth-century
Latin America
WILLIAM ROSEBERRY New School for Social Research
A persistent problem for anthropologists and historians
attempting to understand social change in rural Latin America is
the placement of local regions within wider - global and "national"
- economic, social, and political frameworks) One temptation is to
subsume the local within the global, to make the "system" -
"capitalist" or "modern" - determinative, as in the more extreme
versions of dependency and world-system theories that dominated the
literature in the 1960s and 1970s. Another temptation is to avoid
the problem altogether, to reject any discussion of global
political and economic pressures as totalizing, reductive, or
teleological. This view, increasingly popular over the past decade,
would have us reject the ''fiction of the whole "2
Although both perspectives, as extremes, can point to
respectable intellectual pedigrees and can attract the sympathetic
attention of theo- retically inclined scholars, the student
examining substantive problems and aspects of social change in,
say, Silo Paulo or Antioquia of the 1920s must remain skeptical.
Confronting the global extremists, she or he will agree that
Antioquia was dominated by a coffee economy that had drawn the
region toward the centers of world economy; yet the very shape of
that economy, its most basic social relations and contra- dictions,
were fundamentally different from other coffee economies that
emerged at roughly the same time. Trying to understand why
Antioquia looked different, she or he will begin to explore the
settle- ment of the relatively open frontier, the prior emergence
of a gold pan- ning movement, the accumulation of capital by urban
merchants buying up gold, and their investment of accumulated
resources in land and coffee. In short, the "global" begins to
recede from view and the "local" seems predominant. Yet it hardly
seems helpful to dismiss the whole as a "fiction?' Our student of
1920s Antioquia cannot ignore the
Theory and Society 20: 351-382, 1991. 9 1991 KluwerAcademic
Publishers. Printed in the Netherlands.
-
352
massive investment of North American finance capital during the
"dance of the millions" directed toward the construction of roads
and railroads and the acquisition of controlling shares in local
banks and exporting firms. And she or he cannot forget that the
1920s were fol- lowed by the 1930s, the general depression and the
collapse of the world coffee market. The confident assertions of
the postmodern theo- fist, telling us that we can relegate the
world-system to the back- ground, 3 begin to lose some of their
seductive appeal in the face of such events and movements. A more
careful reading of global and local histories is necessary.
One form of sociological understanding that needs to be
recovered if we are to understand the contradictory formation of
human subjects at the conjunction of global and local histories is
that sketched by E H. Cardoso and Enzo Faletto in their call for
studies of the "internaliza- tion of the external" in Latin
America. Surveying the emergence of capitalism in various Latin
American countries, they argue:
The very existence of an economic "periphery" cannot be
understood without reference to the economic drive of advanced
capitalist economies, which were responsible for the formation of a
capitalist periphery and for the integration of traditional
noncapitalist economies into the world market. Yet, the expansion
of capitalism in Bolivia and Venezuela, in Mexico or Peru, in
Brazil and Argentina, in spite of having been submitted to the same
global dynamic of international capitalism, did not have the same
history or conse- quences. The differences are rooted not only in
the diversity of natural resources, nor just in the different
periods in which these economies have been incorporated into the
international system (although these factors have played some
role). Their explanation must also lie in the different moments at
which sectors of local classes allied or clashed with foreign
interests, organized different forms of state, sustained distinct
ideologies, or tried to implement various policies or defined
alternative strategies to cope with imperialist challenges in
diverse moments of history. 4
What we need, according to this view, is a "history of...
diversity," a sense that "the history of capital accumulation is
the history of class struggles, of political movements, of the
affirmation of ideologies, and of the establishment of forms of
domination and reactions against them. ''5 A history of diversity
is necessarily comparative. One way of approaching such a
comparison is to examine the various regions that produced a
particular export commodity during a particular period - coffee,
say, during the late nineteenth and early twentieth centuries. All
such regions would be subject to certain common global pressures;
they would be experiencing the same "world historical fact "'6 Yet
an understanding of the particular forms and dynamics of social and
eco-
-
353
nomic relations in the various regions would require careful
attention to local contexts, local fields of power. This essay is
directed toward such a comparative history.
The comparative problem
The nineteenth century (that is, roughly, from 1830-1930) was
the cof- fee century in Latin America. It was a period that
witnessed a dramatic increase in world trade (from 320 metric tons
in 1770, mostly from Asia; to 90,000 metric tons in 1820, with half
coming from Brazil; to 450,000 metric tons in 1870 and 1,600,000
metric tons in 19207 ) and per capita consumption (in the United
States, from 3 pounds in 1830 to 10 pounds in 1900 and 16 pounds in
19608). And it was a period in which coffee production was
associated with a profound transforma- tion of landscape and
society in several Latin American regions. In most cases, the
expansion of coffee cultivation coincided with terri- torial
expansion, the movement of settlers into frontier zones where
tropical forests were destroyed, "new forests ''9 of coffee and
shade planted, towns established, roads and railroads built,
regional identities forged.
It is not surprising, then, that we find some of the same
processes and themes repeated from coffee-producing region to
coffee-producing region - the incorporation of regions within an
expanding world mar- ket, the establishment of outwardly focussed
development strategies with the export of a primary product the
price of which fluctuates sig- nificantly but is beyond the control
of local producers and exporters, the building of roads and
railroads (generally with foreign capital) to carry the coffee from
the newly settled interior to port cities, the ambiguous question
of land ownership in frontier zones and the con- flicts between
rural settlers and urban investors, the related legal revo- lutions
in landed property and labor regulations, and the ubiquitous
concern for the labor problem - the "falta de brazos"
What is perhaps more surprising is the remarkable variation in
social, economic, and political structures and processes among
coffee- producing regions, the radically distinct structures of
landed property and the different resolutions of the labor problem
encountered in Brazil or Costa Rica or Colombia. We need to
consider this variation as an interpretive problem: how is it to be
understood? Each of these regions turned toward coffee at roughly
the same time (that is, within a
-
354
few important decades of each other: Brazil, Costa Rica, and
Vene- zuela had important coffee economies by mid-century;
Guatemala, El Salvador, and Colombia turned to coffee several
decades later - the 1870s, 1880s and beyond). Each was producing
the same primary product for export to the same European and North
American ports (though one might export primarily to London,
another to New York, another to Hamburg). The structure of trade
(that is, the relation between local exporters and international
firms) was roughly the same (though important differences developed
in Brazil as it came to domi- nate the market). Each of the regions
became "dependent" on a single export commodity, suffering the same
reverses and enjoying the same booms.
Despite the commonalities in their incorporation within the
world market, however, their most basic social relations, including
those associated with labor mobilization and "the specific economic
form, in which unpaid surplus-labour is pumped out of direct
producers "'l~ were fundamentally different. Easy assertions about
the dominance of the "latifundia-minifundia complex" are out of
place, as are more com- plex arguments that recognize variation but
subsume the variation within a common emergence of two "large nodes
of decision-making bodies" with the incorporation of regions within
the world economy - one based on the "plantation" solution and the
other based on the "merchant" solution (in which merchants dominate
and capture the production of small farmers). 11 Such assertions
explain away difference rather than confronting it.
Let us, then, confront these differences in the coffee-growing
regions in Latin America. Let us place Silo Paulo next to the
Central Valley of Costa Rica or Antioquia and ask why such
fundamental differences in landed property and labor mobilization
occurred and what effects these differences might have had for the
respective societies in which they occurred. A variety of easy
resolutions are closed to us. None of the distinctions in timing or
markets noted above was decisive. Nor do we have access to a
mechanical opposition between closed and open frontiers or to
different land and labor ratios. If our only contrast was one
between El Salvador and Costa Rica, such oppositions and ratios
might be convincing, but most of the regions were open frontiers,
with different results of settlement that are too important to
gloss over with grids and causal diagrams. A more considered
examination of the societies in which the frontiers were opened and
settled, the social, eco- nomic, political, and cultural contexts
in which coffee became an important export crop, is necessary.
-
355
This essay represents a preliminary examination of such
contexts, the aim of which is not to explain difference but to
begin a comparative dis- cussion. I develop the comparison with a
discussion of the manner in which coffee elites in different
regions resolved one of their most press- ing problems - the
mobilization and reproduction of labor. Although other aspects of
the respective coffee economies (e.g., commercializa- tion and
politics) deserve detailed attention, and will be treated else-
where, the labor problem - "la falta de brazos" - was central to
each of them, inflecting all aspects of social, economic, and
cultural life. It therefore constitutes our necessary starting
point.
However much this may seem to resurrect the labelling
controversies associated with the mode of production debates, the
crucial difference in the present exercise needs to be stressed.
The purpose of the present comparison is not to outline distinct
modes of production, and I do not consider here the capitalist or
non-capitalist character of the labor regimes examined. Indeed, one
of the problems with earlier labelling exercises was that they
directed our attention toward labels and away from a consideration
of wider economic, social, political, and cultural fields of
power.
It is toward such a consideration that the present study of
labor regimes in Latin American coffee economies, and the larger
comparative study of which it is a part, are directed. We might
briefly outline three dimen- sions of the labor problem that
illuminate wider social and political relations and processes.
First, in places such as Brazil, Colombia, and Guatemala, large
landholders attempting to attract laborers were not acting in
isolation. They might be competing with growers from other regions,
with urban entrepreneurs, or, in the case of immigration schemes,
with planters or entrepreneurs in other countries. This is not to
say that landholders were powerless and a free market prevailed:
the monopolization of land in some regions was the most effective
means for securing a labor force. It is to say that planters acted
within particu- lar contexts, particular sets of constraints, and
that the systems they devised to attract workers in the first
place, or to assure more careful tending of coffee trees, or to
feed the working population, often created further constraints.
Structures of decision making and control could become quite
diffuse as coffee groves and food plots were let out to
tenants.
Second, labor mobilization schemes were never static. It would
be insufficient to set up a comparison in simple spatial terms with
large
-
356
estates and the colonato in Brazil, peasants and processors in
Costa Rica, haciendas in Cundinamarca, and peasants in Antioquia -
the large-estate regions being characterized by "oligarchic"
domination and the peasant regions seen as more "democratic" In
each of these regions, labor regimes changed over time. Haciendas
in Cundinamarca began to disintegrate in the 1920s and 1930s, for
example, partly due to economic problems encountered much earlier
and partly due to the increasing organization and militance of
their tenants. Careful attention to the fault lines created by
hacendados' resolution of labor-mobiliza- tion problems in previous
decades is essential for an understanding of their problems in the
1920s. In the peasant regions, in turn, we need to be sensitive to
changes over time. In Costa Rica, for example, small farmers faced
increasingly difficult pressures from the middle of the nineteenth
century to 1930, as open lands closed off or as relations with
processors became more exploitative or as household heads found it
increasingly difficult to provide an inheritance for all of their
children. 12
Finally, if we think about labor mobilization in terms of
contexts, con- straints, and fault lines, and if we consider the
way particular resolu- tions of the labor problem change over time,
we open up a most inter- esting area for investigation. One of the
interesting developments that emerges in the literature on coffee
in Latin America is the frequency with which elites experiment with
different strategies. The most famous is probably the Vergueiro
experiment in Brazil with immigrant share- croppers in the
mid-nineteenth century, four decades before the end of slavery (see
below)) 3 But we also find other experiments in, for exam- ple,
Cundinamarca in the 1920s TM or Guatemala in the 1920s and 1930s.
Indeed, careful attention to such experiments and debates can
illuminate the most profound economic, political, and cultural
dilem- mas confronting coffee elites. As we examine the kinds of
solutions that are attempted and the solutions that are not even
considered, we are able to sketch the limits of the possible (which
include the limits of the socially constructed mental and cultural
horizons of the elites at a par- ticular time) in various
coffee-producing regions. An apparently simple "economic" question
(how was labor organized), then, need not lead to a labelling
exercise. A discourse about labor is seldom "just" about labor.
Examining one such discourse, we may begin to unpack the sociology
of racism in Guatemala or Brazil or Costa Rica; in examining
another, we may begin to understand the particular features of
liberal- ism in, say, early twentieth-century Colombia.
-
357
In what follows, I concentrate on Brazil, Costa Rica, and
Colombia, which represent a range of resolutions to the labor
problem in nine- teenth-century coffee economies. In this
discussion, my aim is to devel- op a more detailed understanding of
the dimensions of difference. I then suggest an interpretive
framework in terms of which we can devel- op further comparative
discussions.
Land and labor in Latin America's new forests
Although the primary focus of this essay concerns land and labor
regimes and does not consider commercialization schemes in any
detail, certain basic features of the coffee trade in
nineteenth-century Latin America deserve brief consideration. For
those newly independ- ent countries with exploitable subtropical
soils, coffee served as a prin- cipal point of linkage to an
expanding world economy, the means by which they could turn toward
an "outwardly focussed" model of devel- opment. It could be stored
for long periods with relatively little spoil- age; it had a high
value per kilogram, making transport costs relatively low and
making inland territories valuable in a way they could not be for
crops such as sugar; 15 and it enjoyed a growing and lucrative
acceptance in European and U.S. markets. For merchants and trading
firms from countries entering the new Latin American markets,
coffee became a focus of trade.
Throughout the nineteenth century, coffee production and
marketing followed classic free-trade patterns. Control of
production was highly dispersed, both among coffee-producing
countries and among pro- ducers within countries. Although
international trade was controlled by merchant houses in London,
Hamburg, and New York, there was no significant concentration among
the houses until the early twentieth century. As concentration
began to occur, it responded at once to changing processing and
marketing structures in consuming countries and to crisis periods
in producing countries, during which foreign firms might take more
direct and active roles.
Foreign coffee firms would establish credit and commercial
relations with exporters and merchants in particular Latin American
countries, loaning funds to exporters with which the exporters
would acquire cof- fee - often by means of further loans to local
producers and merchants. The general features of such arrangements
can be briefly sketched. First, despite the close connection
between European or North Amer-
-
358
ican firms and Brazilian or Costa Rican exporters and producers,
local merchants and exporters were not subsidiaries of European or
North American firms for most of the period under discussion. Even
where the exporters were German or English expatriates, they were
expatri- ates acting as individual entrepreneurs and adventurers,
often with a privileged and preferred relationship with a
particular London or Hamburg house, but the tie that bound them was
one of credit and shared nationality rather than ownership. Second,
exporters, acting with their own funds or with borrowed funds from
abroad, were the principal sources of credit for local producers
and merchants. For most of the period that concerns us, national or
international banks were not involved in the coffee trade. Third,
with purchasing and credit arrange- ments linking particular
international firms and exporters, producers and merchants alike
were subject to price fluctuations. Exporters lacked the means to
withhold coffee in periods of low prices. There were no local
exchanges, and states were not involved in coffee trade. It was
only with the onset of the first general overproduction crisis in
the 1890s that discussions began finally resulted in Brazil's
valorization scheme of 1906. With this scheme, the first chinks in
the free-trade armor appeared? 6
Furthermore, the market was not homogeneous. In general,
European consumers have preferred the "quality" milds produced in
Costa Rica, Colombia, and Guatemala, and European markets were the
principal outlets for the milds. These export markets were cemented
with long- term arrangements with particular foreign houses.
Indeed, during the free-trade period much of the quality coffee was
exported not as Colombian or Costa Rican coffee but, "like French
wines, ''17 under the mark of a particular Costa Rican processor
(beneficiador) or Colom- bian hacienda. The United States, on the
other hand, has served as a market for the harsher, less expensive
coffees, especially from Brazil, but also as a subsidiary market
early on for the other countries. As with all generalizations, this
one requires some temporal specification. In the first place, no
producing country exported to a single consuming country. Second,
during the twentieth century the U.S. market became increasingly
important throughout Latin America, especially during the two World
Wars. Nonetheless, the segmentation of the coffee mar- ket is an
important feature, especially when we consider the quality end of
the scale. Given a marketing environment permeated by a discourse
of quality and a pricing system based on the grading hierarchies,
this provides an important point of control for coffee processors
and mer- chants, especially in relation to small producers. But
detailed consider-
-
359
ation of such relations remains beyond the scope of the present
essay. We need to turn now from the structure of commerce and
investment to the transformation of landscapes and societies.
The frontier character of many of the coffee regions has often
been stressed in regional studies? 8 If the frontier has impressed
historians and social scientists, it has also impressed historical
actors, both at the moment of frontier settlement and in memory.
The memory of cutting down the forest (tumbando montes), or the
image of a people forged in settlement and transformation (for
example, "the ethos of the hacha" 19 [an ax] in Antioquia) is
strong.
Indeed, as we see in detail below, most of the areas converted
to coffee cultivation attracted population migration and
settlement. Guatemala and E1 Salvador serve as counterpoints in
this story, in that both were densely populated. Even in Guatemala,
however, the microregions that were to become the most dynamic
production zones - the piedmont of Amatitlan, Suchitepequez,
Solola, Quezaltenango, San Marcos, and the Alta Verapaz - contained
much unused land. Only in E1 Salvador did the coffee zone
correspond with a region of relatively dense colonial settlement,
and only in El Salvador did the expansion of coffee and the
transformation of landed property that accompanied it involve a
wide- spread displacement and expropriation.
Despite the frontier character of much of the coffee expansion,
how- ever, most of the "wildernesses" into which coffee farmers
moved were already encumbered by people, overlapping and competing
claims to land, conceptions of space, time, and justice - in short,
"history" - before the coffee expansion began, and these
encumberances shaped their respective coffee economies even as the
regions were transformed by the move toward coffee. In each of the
regions considered in this essay, then, we begin with a brief
discussion of the occupation of space and the transformation of
landed property. With this foundation, we can then turn to a
consideration of the labor problem in various sorts of production
regimes.
Brazil. Let us first consider Brazil, which stands alone in the
dimen- sions of its forest. The extent of territory available and
suitable for cof- fee cultivation, first in the Paraiba Valley of
Rio province in the early and mid-nineteenth century and then into
Minas Gerais and the S~o Paulo west from the mid- to late
nineteenth century, dwarfs whole countries in Central America, not
to mention the much more restricted
-
360
zones suitable for coffee. The destructive nature of this
expansion, in which tropical forest would be cut and coffee
planted, setting in morion a 30-40 year boom during which the soil
would be depleted, the boom region set into decline, and then the
coffee grove revert to pasture or waste as new regions to the west
were opened up, is well known. 2~ In one respect, these interior
regions were "new," "untouched," "virgin" The declining sugar
complex of the northeast was quite distant. South- ern developments
during the colonial period had centered around the administrative
center in Rio and gold mining in Minas GerMs in the eighteenth
century, which in turn stimulated cattle and agricultural complexes
in the coastal and more accessible areas. Colonial claims to
interior lands nonetheless emerged. During the colonial period,
land belonged to the Crown unless it had been ceded by a personal
grant (sesmaria), generally one square league (44 sqaure
kilometers), in return for services to the Crown and on condition
of cultivation. With the building of roads between Rio and the
mines of Minas, sesmarias were granted, as the discovery of gold in
Mato Grosso led to trail and road blazing and the establishment of
way-stations. The lands encom- passed by these grants were
underurilized in the absence of commercial opportunities, however,
and they were settled by squatters who would displace Indians (who
were not protected and who were written out very quickly, both in
practice and in histories of settlement) toward the west. Squatters
might engage in subsistence agriculture or service the way-stations
along the proliferating mule tracks, but their lands (posses),
which could be quite extensive and might overlap with under-
exploited sesmarias, were not recognized in colonial land law. With
the westward expansion of coffee, these conflicting claims became
impor- tant as grant holders or the entrepreneurs to whom grants
had been sold turned their claims into extensive plantations with
vague bound- aries. With independence, sesmarias were no longer
granted, but both sesmarias and posses were bought and sold in a
conflictful rush to con- trol the land. The land law of 1850
resolved the conflict in favor of grant holders and those posseiros
wealthy enough to purchase their claim from the state. That is,
colonial grants were recognized as titles but the right of
possession was not. Land could only be rifled by means of
registration, survey, and the payment of a tax. In practice, this
dis- placed small squatters toward the west, and the expansion into
the Paraiba Valley or the western plateau of S~o Paulo was
characterized by a series of displacements: squatters displaced
Indians toward the west, only to be displaced by estate owners as
roads or railroads stretched further into the interior. 21
Nonetheless, while the land law had the effect of displacing
squatters, its desired effect of establishing a
-
361
land registry with carefully surveyed properties was not
realized. Indeed, one of the remarkable features of the coffee
economy through- out the period we consider is the resistance of
large landholders to land surveys and registries. Such resistance
within the particular field of power in which they operated allowed
them to avoid taxes but also al- lowed them to extend the effective
domain of their estates. 22
The spread of the large estate should not be treated as
unproblematic, however. No Latin American frontier of settlement
was larger than the Brazilian interior in this period. A mechanical
application of a frontier thesis might lead us to expect a more
"democratic" landholding pattern to emerge. Yet here, as elsewhere,
the importance of the larger field of power, the political,
economic, and cultural context of frontier settle- ment, needs to
be stressed. Again and again, historians point to this context and
the mental and cultural horizon it produced. Commenting on the
failure of smallholding in the vast frontier, Warren Dean notes,
"Unfortunately, the royal administrators could never entertain
serious- ly a reform that would bring about not only the desired
increase in revenues but also what would appear to them to be a
social revolution. The only organization they could conceive for
the immense colony had to be a society precisely as aristocratic as
that of the metropolis "'23 Of the spread of slavery to the
frontier, Stein observes, "Free labor as an alternative hardly
existed in the minds of the settlers,' 24
Such conceptions and minds have historical and social armatures.
While as a first approximation it might be useful to distinguish
between the sugar-growing northeast and the expanding coffee
provinces of the south, to see the one as conservative and
aristocratic and the other as more liberal, "less wedded to the
past," and holding "more adaptable economic and social views "'25
their liberalism took on a special, Brazil- ian character. Viotti
da Costa stresses that despite a late eighteenth- and early
nineteenth-century fascination with Enlightenment thought, which
led to the formation of secret societies and pro-independence
conspiracies, the liberalism that dominated in Brazil by
independence was one that had been purged of its more radical
social content:
In Europe, liberalism had originally been a bourgeois ideology,
an instru- ment in the struggle against the absolute power of
kings, the privileges of the nobility, and the feudal institutions
that inhibited economic development. But in Brazil, liberalism
became the ideology of rural oligarchies, which found in the new
ideas arguments they could use against the mother country. These
men were primarly concerned with eliminating colonial institutions
that restricted the landowners and merchants - the two most
powerful
-
362
groups in colonial society. When they struggled for freedom and
equality, they were actually fighting to eliminate monopolies and
privileges that bene- fited the mother country and to liberate
themselves from commercial restric- tions that forced Brazilians to
buy and sell products through Portugal. Thus, during this period,
liberalism in Brazil expressed the oligarchies' desire for
independece from the impositions of the Portuguese Crown. The
oligarchies, however, were not willing to abandon their traditional
control over land and labor, nor did they want to change the
traditional system of production. This led them to purge liberalism
of its most radical tendencies. 26
For both liberal and conservative planters, the monarchy became
a means for preserving an aristocratic society in the postcolonial
era. Indeed, a pact between northeastern sugar planters and the
expanding Rio elite was crucial in the ascension of Pedro II to the
throne in 1840.
With this nineteenth-century monarchy, unique in Latin America,
we might understand something of the political and cultural context
that would attempt to extend into the frontier the system of
production and privilege that had served as the basis for colonial
society. We can understand the political context in which royal
land grants were recog- nized as legitimate but not the rights of
possession. We can understand the attempt to recreate a whole
society and way of life, in which both land titles and aristocratic
rank could be granted, in which a personal empire could rest on the
labor of slaves.
But the attempt to expand and reproduce such a society took
place in new contexts. In the first place, planters viewing
abundant land and a dependent labor force adopted production
techniques that made for quick profits and long-term destruction.
Initial productivity depended upon the natural fertility of the
forest. Whole sections of forest would be cut and burned, and
coffee trees planted in vertical rows up hill- sides, to facilitate
access to the trees by slave gangs. At harvest, trees would be
stripped of cherries and leaves. In a classic and oft-repeated
description, this harvesting method (unique in Latin America) is
pic- tured: "Each branch was encircled by thumb and forefinger, the
hand then being pulled down and outward, thus 'stripping the branch
in one swift motion' and filling the screen with leaves, dead
twigs, and coffee berr ies . ''27 Such methods assured the
productivity of labor but not of land; indeed, with the erosion
caused by the vertical rows, they assured that the land would be
exausted at the end of the 20-30 year cycle of the coffee trees
themselves.
Second, Brazil's new trade relationship with Britain threatened
the
-
363
slave trade, and by mid-century the trade had effectively
stopped. Thus, while the initial expansion into the Paraiba Valley
had been facil- itated by the easy extension of slavery, the boom
of the 1850s and 1860s brought with it an increasingly costly labor
force. Slave prices doubled in the early fifties as an
international trade was replaced by an interregional trade, with
Rio planters buying slaves from the declining northeast. By the
1870s, central legislation began to limit slavery. The Rio Branco
Law of 1871 freed slave children born after passage of the law,
while the Sexagenarian Law of 1885 freed sixty-year-old slaves,
e8
Behind the picture of great wealth and aristocratic privilege
created by estate agriculture and slave labor, then, lay a social
reality of waste, decay, and impending crisis. Yet one of the
features that impresses the reader of Stein's study of Vassouras in
the Paraiba or Dean's study of Rio Claro in Sio Paulo is the
inability of most planters to respond to that crisis, to envision
anything other than the slavocracy that had been the basis of their
wealth and was decaying around them. Their opposi- tion to
abolition, their attempt to put it off for another generation, is
striking. Even so, other planters, especially in Silo Paulo, could
foresee the end of slavery and experimented early on with
alternative forms of labor - alternatives that could not be
realized as long as slavery con- tinued. Nicolau Vergueiro's
experiment beginning in 1845 on his Silo Paulo plantation has
received considerable attention. 29 Under this sys- tem, Vergueiro
financed the immigration of German and Swiss workers who were to
settle on his plantation, sharecrop an unspecified number of coffee
trees, and pay off the debt incurred by their passage. Their
compensation was to be half of the coffee yield (from which half
was to be deducted to retire the debt), a house, and a food plot.
While the ini- tial success of the experiment led to expanded
immigration and share- cropping in the early 1850s, enthusiasm for
the project had waned by the late 1850s, partly due to strikes and
desertions of 1856-57, and partly due to decreased labor
productivity. The central problem, according to Stolcke, was the
initial debt. The indenture required the sharecropper to work off
his debt, but the deduction for debt encour- aged the sharecropper
to concentrate on the food plot rather than the coffee plot. The
planter therefore had to enforce an indenture contract in a
situation in which desertion was possible, and to stimulate produc-
tivity in a situation in which control over the labor force was
much more diffuse than with slavery.
Despite the demise of the Vergueiro experiment in the 1850s and
the continued dominance of slave labor until 1888, some planters
con-
-
364
tinued to experiment with free-labor regimes. 3~ They faced two
prob- lems. On the one hand, the initial debt associated with
planter financ- ing of immigration created an immediate obstacle.
On the other, the planter needed more control over the coffee
production process - and by extension over the productivity of
workers - than sharecropping allowed. The first problem was to be
addressed by the transfer to the S~o Paulo state of the entire cost
of immigration; the second was addressed with the adoption of a
"mixed task and piece-rate system "'31 Together, by the 1880s,
these two innovations became the distinctive features of the
colonato. Beginning in 1871, Silo Paulo began to take over limited
subsidization of European immigration for the coffee farms. In
1886, the Sociedade Promotora da Imigracao, a private organization
under contract to the state, was formed, producing a 60- page
booklet promoting Silo Paulo, published in Portuguese, German, and
Italian, and opened European offices, promoting and organizing the
immigrant stream. With the fall of the Empire and the establish-
ment of a republic in which the states had significant power and
auton- omy, the immigration program was taken over by S~o Paulo's
Depart- ment of Agriculture. 32 "From 1889 to the turn of the
century," Hollo- way writes,
nearly three-quarters of a million more foreigners arrived in
Silo Paulo, of which 80 percent were subsidized by the government.
From abolition to the Depression nearly two and one-quarter million
immigrants came in, com- pared to a population base in S~o Paulo in
1886 of one and one-quarter mil- lion. Some 58 percent of all
immigrants in that period were subsidized by the state. 33
The vast majority of the immigrants were Italian, although Italy
prohib- ited further subsidized emigration to Brazil in 1902. 34
Furthermore, the state engaged in a remarkable coordination of
planter needs and labor supply. Immigrants would be transferred
from Santos to a hostel in Silo Paulo, where the state would serve
as labor contractor. While at the hostel, the immigrant family
would sign a contract to work on a particular plantation and would
then be given railroad passage from Sgo Paulo to the interior.
35
The contracts they signed represented a unique form of labor
mobiliza- tion. First, they received a fixed wage per thousand
coffee trees weeded and maintained during the year, regardless of
yield. Second, harvest labor was compensated on the basis of yield
(so much per 50 liters of cherries). Third, they received a house,
and fourth, they received a food plot. Variations might appear in
regions where coffee was being plant-
-
365
ed, allowing colonos to plant food crops between rows of
recently planted coffee. The system preserved some of the
advantages of a sharecropping regime (some of the risk was reduced
with the harvest compensation tied to yield; costs were reduced
with the provision of a food plot) but eliminated some of
sharecropping's disadvantages (the set wage for tending a number of
trees allowed more space for planter control of the labor
process).
Because of state subsidization of immigration and the
elimination of debt as a social and economic relation between
planter and colono, there was extraordinary movement of persons in
the S~o Paulo West at the close of each annual cycle. Colonos on
the plantation might leave and move farther west, especially to
zones of expansion, where con- tracts were perceived by colonos as
being more lucrative. As long as the immigrant stream was
maintained, however, the instability in terms of personnel was of
little concern for the planter. A dependable, state subsidized and
controlled mass of cheap and replaceable labor re- mained
available, a6
Once implanted, the colonato system dominated coffee production
in Sao Paulo throughout the period that concerns us here, lasting
until the 1960s. The combination of incentives to individual
laborers, cost- reducing features, and a structure of labor
discipline, proved a powerful source of planter power in the early
decades of this century. Stolcke emphasizes, for example, that
planters were able to weather increas- ingly frequent periods of
low prices because the provision of food plots allowed planters to
reduce wages and compensate for decreased prices, a7 Nonetheless,
we need to look to the fault lines in any labor regime. A labor
regime that provides flexibility in response to one set of
pressures may create obstacles in others. The planters' dependence
on an ever-flowing immigrant stream was one such obstacle. Another
lay in the attraction of contracts in zones of expansion, providing
a built-in incentive to increase production as planters entered
decades of overproduction. Thus, while the combination of food and
coffee production provided planters with flexibility during
low-price periods, the incentives built into the colonato could
exacerbate the overproduc- tion problem, making price troughs more
frequent and severe.
Costa Rica also moved toward coffee cultivation early in the
nineteenth century, but the occupation of space and titling of land
differed markedly from the Brazilian example, a8 In the first
place, the land suit- able for coffee is restricted, concentrated
in the Central Valley from
-
366
Alajuela in the west to Ujarrfis in the southeast. Throughout
the colo- nial period, Costa Rica was a periphery of a periphery.
Part of the Audienca of Guatemala, most of Costa Rican territory
lay beyond the area of dense Mesoamerican indigenous settlement,
and the Spaniards who settled in the frontier colonial outposts
found little in the way of exploitable resources or population. At
the end of the colonial period, 40,000 out of a total population of
about 50,000 lived in the Central Valley. The most important
colonial commercial crop, cacao, had not been grown there but on
the Atlantic Coast, and the bulk of the Valley's population lived
in towns such as San Jos6, Heredia, and Cartago and villages,
practicing a "village economy. ''39
With independence came a search for a viable commercial crop. In
1821 the municipality of San Jos6 distributed coffee plants among
indigents and conceded land to anyone who would plant and fence
cof- fee groves. In 1831 the national assembly declared that anyone
who planted coffee in national lands (terrenos baldios) for five
years would be granted title to the land. n~ This was the first of
a series of relatively open and generous (though not always
conflict-free) legal instruments granting national lands to
settlers who would cultivate them. 41 It also led to the early
establishment of a land registry and survey, through which small
holders could protect and defend their holdings.
The expansion of coffee cultivation in the Central Valley can be
distin- guished among three regions: 42 (1) the nucleus around San
Jos6 and Heredia and surrounding villages, which was the first to
move toward coffee, which had the most fertile lands for coffee
cultivation, and which was the most densely settled center of
coffee production and commercialization; (2) the Alajuela/San Ram6n
region to the west, along and near the road from San Jos6 to
Puntarenas, toward which migrants from San Jos6/Heredia began to
move from the 1840s but especially during the last half of the
century, practicing a mixed coffee/ sugar cane/cattle and other
crops regime; 43 and (3) the Reventaz6n and Tundalba Valleys to the
east, which did not develop coffee farms until the completion of
the Limon railway, which passed through the Valleys. Unlike the
other regions, Turrialba did not attract peasant migration and
settlement but was characterized by large coffee, sugar cane, and
banana farms. Unlike the development of coffee in other countries,
regional expansion in Costa Rica was not accompanied by the decline
of earlier centers of production. The San Jos6/Heredia nucleus
remained the center of coffee production and commercializa- tion
even as new regions were opened up.
-
367
The expansion of coffee cultivation in Costa Rica's Central
Valley occurred within a social, political, and cultural context
that represents a stark contrast to the Brazilian example. Colonial
society had produced a town and village aristocracy who were not
far removed, in social and economic terms, from the rest of the
population. Slavery was virtually nonexistent (no more than 200
slaves at any point during the colonial era) and was outlawed in
1824. 44 Nor were other forms of servile labor widespread. As
Gudmundson notes:
Political and religious office went hand in hand with the
generation and pre- servation of wealth, just as in other, more
dynamic Spanish colonial societies. Ownership of land was not the
surest or quickest road to enrichment in this society, however much
it may have been both a form of security and a neces- sary element
in securing elite status and acceptance. Unlike other Central
American societies, landownership in central Costa Rica (excluding
Guana- caste) did not bring with it a servile labor force, a fact
that meant that there was even less interest in landholding among
the elite . . . . [lln Costa Rica land- ownership was not the
distinguishing feature of the elite; instead it was a combination
of commerce, office holding, and diverse investments in urban and
rural real estate. 45
Many authors have painted a picture of a widespread rural,
peasant population fanning subsistence crops. 46 For these authors,
the dis- persed peasantry serves as a starting point for their
analysis of develop- ments during the coffee era, a minority view
arguing that coffee led to an accumulation of landholdings and a
proletarianization of the peas- antry and a majority view arguing
that the widespread peasantry served as the social basis for
smallholding commercial production with the expansion of coffee
cultivation. 47 On both sides, we encounter silences. In Seligson's
case, for example, the analysis seems to follow from a theoretical
model of the effects of commercial agriculture, along with a
presentist reading of nineteenth-century census categories like
jornalero. In Hall's case, her most vigorous argument against land
con- centration and proletarianization stresses the contrast with
Brazil. Unlike the huge landholdings in Brazil, that is, large
coffee farms in the Central Valley were relatively small - 60
manzanas on average, with 60,000 trees. By the 1930s, these farms
held perhaps 25 percent of the coffee land in the San
Jos&Heredia nucleus. 48 While this does repre- sent a stark
contrast with Brazil, once the contrast has been made the Costa
Rican estate needs to be placed in a Costa Rican context. A 60,000
tree farm is not a peasant farm and cannot be worked with family
labor. We need to move beyond the contrast, then, and explore a
specifically Costa Rican field of power.
-
368
This is an area where Gudmundson's model of a colonial village
econ- omy and the ruralization of the peasantry with the expansion
of coffee cultivation is especially suggestive. It is a model that
helps us better to understand those nineteenth-century social
processes that historians have delineated: the privatization of
baldfos, the move from subsistence production to commercial crops,
a specific migration pattern in which a particular region would be
occupied and then the sons and daughters of a subsequent generation
would be faced with the choice of divided and reduced holdings,
occasional or permanent labor on nearby estates, or migration to
the western frontier of the Central Valley. The property-holding
commercial peasantry represented an obstacle to land concentration.
The expanding estate owner had to purchase small properties and
could not depend on generous land grants or the sale of extensive
bald/os or a structural space created by vague titles and non-
existent registries and surveys.
While this landholding peasantry represents a significant
contrast with other Latin American experiences, it should not be
romanticized. With the passing of generations and the increasing
shortage of land, small- holders were to be divided by growing
inequalities. 49 Further, the requirements of processing and
marketing their coffee placed them in direct contact with coffee
processors (beneficiadores), who were to become the coffee elite of
Costa Pica. Indeed, as Hall notes, the large- estate holders of the
San Jos6/Heredia nucleus were beneficiadores.
An examination of this commercial infrastructure lies beyond the
scope of this essay. 5~ For now it needs to be noted that the Costa
Rican field of power is inconceivable without it. As in the
colonial period, landholding was not the primary route to power.
Both the accumula- tion of land and access to labor depended on
one's position within and access to accumulated commercial wealth.
As Gudmundson con- cludes:
Coffee fundamentally transformed a colonial re#me and village
economy built on direct extraction by a city-based elite from a
peasantry that was as yet privatized to only a small degree. The
replacement of this direct extrac- tion by more subtle and
productive market-mediated mechanisms created a qualitatively new,
antagonistic relationship between the coffee elite of
processors-exporters and the thoroughly mercantile, landholding
peasantry. The road to agrarian capitalism in Costa Rica followed
along these lines, rather than those of an estate model based on
the rapid proletarianization of a formerly self-sufficient and
self-determined peasantry. 5~
-
369
Yet the elite's immobility in confronting the labor problem
needs to be emphasized. Given the situation of the large estate
within a peasant milieu, estate owners attracted permanent and
seasonal laborers with relatively high wages. The Costa Rican peon,
as Cardoso stresses, "was basically an employee, a wage labourer
and not a 'serf. '''52 Yet they were unable to mount any sustained
effort to attract additional laborers to Costa Rica. On the one
hand, this represents their more modest resources in a world in
which other countries - Brazil and Argentina - had begun massive
subsidized immigration schemes, not to mention the North American
zones of attraction for Italian migrants during the same period. On
another, it represents the limits of their own mental and cultural
horizons. The 1862 colonization law specifically forbade settlement
by blacks and Chinese, and Tomas Guardia rejected Chi- nese workers
in 1875, claiming they were "gamblers, thieves, and opium smokers?
'53 Moreover, even when contracts were signed for the construction
of a railway to the Atlantic Coast, the Costa Rican government
stipulated that the West Indian laborers brought in to work on the
railway were not to enter the Central Valley. 54
Colombia. The expansion of coffee production in Colombia began
much later than in Brazil and Costa Rica. Three branches of the
Andes divide the country into regions that, in the nineteenth
century, were iso- lated from each other and far removed from ports
that could be reached via the river systems of the Magdalena and
the Cauca. At the close of the colonial period, the bulk of the
population lived in the highlands, which had also been the site of
indigenous settlement. Around highland towns and cities, haciendas
developed alongside and often at the expense of indigenous reserves
(resguardos). But the haciendas and resguardos provisioned
regional, urban markets. Topo- graphy and demography combined to
hinder the development of an export economy and promote the
development of relatively isolated regional economies. Just as a
"national" market or export economy was weakly developed, the
central government was quite weak. Local hacendados held power in
particular regions, and though struggles between liberals and
conservatives concerned control of the central government, they
also, and often more importantly, concerned control of local
governments, their public offices and records, and their legisla-
tive power.
This is not to say that there were no exports at all, or that
the new mer- chants and free tradists in cities such as Bogotfi did
not organize proj- ects and attempt to establish closer ties with
world markets. Gold
-
370
mining in Antioquia was an important export activity and source
of capital, and the mid-nineteenth-century tobacco boom in the
Magda- lena valley, while short-lived, showed some of the
possibilities of the sub-tropical lowlands and slopes. But it was
only with the move toward coffee production, which began in earnest
after 1870, that firm links with world markets were established.
With the move toward coffee, the regional structuration of
Colombian topography, demography, and economy was important. The
development of the coffee economy fol- lowed three cycles, each of
which was concentrated in a particular region. Each region, in
turn, began its coffee cycle with a different colo- nial legacy in
terms of prevailing social relations and the occupation of
space.
Santander, in the northeast, near the Venezuelan border, was the
first Colombian region to turn toward coffee, after 1850. A region
of colo- nial settlement, hacendados were able to turn to coffee as
their tobac- co, cotton, or cacao markets collapsed. As the first
region to turn to coffee, Santander was to dominate Colombian
production throughout its first coffee cycle, accounting for some
60 percent of Colombian pro- duction at the end of the nineteenth
century. 55 An important percent- age of its coffee was exported
via the developing Venezuelan port at Maracaibo, as coffee
production was expanding in the Venezuelan Andes at roughly the
same time. By the turn of the century, Santander- ean production
was beginning to level off, and the region accounted for an
decreasing percentage of production in this century (only 8.9 per-
cent by 1943). 56 Although the move to coffee involved changes in
structures of production and landed property, with an accumulation
of large properties aided by regional liberal/conservative wars
(during which victorious forces would destroy local land
registries57), it did not depend on or attract strong population
movement.
In Cundinamarca and Tolima, however, the expansion of coffee
after 1870 took place along mountain slopes and on lands that had
not been important during the colonial and early post-colonial
period. Although there were important towns that were to become
centers of the coffee trade, the coffee expansion also opened up
new lands. The new lands, however, were encumbered by claims. From
the late eighteenth century, lands in the temperate slopes were
claimed in large latifundia. 5s The expansion toward coffee
involved the investment by Bogotfi and Medellin 59 merchants, a
"new class ''6~ looking for investments in export agriculture and
buying and dividing larger latifundia or buying public lands to
establish coffee haciendas. 61 It also involved the move-
-
371
ment of indigenous and mestizo peasants from the Sabana de
Bogotfi and the highlands of Boyac~i, who settled as renters on the
emerging haciendas. Although the Cundinamarca/Tolima coffee zone of
the western slopes of the eastern Cordillera was an important
coffee zone in that it served as the base for an oligarchic elite
that lived in Bogot~i and accumulated properties in the
Cundinamarca/Tolima slopes, it was never the most important
producing region in terms of volume of pro- duction. By the time
Santander entered into a prolonged decline, the western,
Antioquefio expansion was firmly established, and the Antio-
quia/Caldas coffee zone dominated Colombian production.
This western zone has been the subject of a powerful myth - the
Antio- quefio colonization, the establishment of a settler society
as colonists moved into the sub-tropical frontier, carved out
farms, and established towns and small-scale enterprises, with a
"democratizing" effect on Antioquefio and Colombian society. More
recent studies have empha- sized the less idyllic aspects of this
process, the appropriation of large tracts of land by a few, the
exploitation of small producers by urban merchants, the violent
conflicts over land and resources as public lands were privatized.
62
Unlike the Santanders or Cundinamarca/Tolima the area of western
colonization contained a good deal of unclaimed public lands
(baldios) at the close of the colonial period. The predominant
economic activity was gold mining, which was not characterized by
the servile labor rela- tions predominant in other regions. Most of
the gold had been mined by mazamorreros, descendents of slaves and
mestizos who had left other regions and worked independently by
mining gold along western rivers and streams, selling their gold to
urban merchants in Medellin. At the beginning of the Antioquefio
colonization, then, the gold econ- omy provided a social base for
independent settlement and activity (the mazamorreros) and a source
of capital accumulation allowing urban merchants to invest in new
enterprises. 63
The settlement of baldios in the nineteenth century fell into
two broad periods. During the first, from independence to the
1870s, public lands were sold as a source of revenue for a weak
central government and without regard to the occupation of land by
settlers, setting up the basis for the same kind of conflict that
occurred in Brazil. During this period, colonization took a
"collective" character, in which a whole settlement would be
granted title, including house and farm plots. In this form,
baldfos might be ceded by the state of Antioquia, or colonists
would
-
372
settle uncultivated forest land held in colonial land title
(tierras realengas), or merchants would organize settlement
projects and obtain title, ceding some land to settlers but
maintaining the bulk of the land for cattle haciendas. Thus a
mixture of large and small holding resulted, with large cattle
haciendas occupying the lowlands and small farms on the forested
mountainsides. In contrast with Brazil, the passage of laws 61 of
1874 and 48 of 1882 placed limits on the size of holdings that
could be titled from public lands and, more importantly, recognized
the rights of prior settlement and possession. In practice, this
did not represent a transfer of power from large landholders to
small, and sta- tistical analyses of public-land sales and grants
show a continued pre- dominance of large holdings. But it created a
legal terrain on which settlers could struggle, and through which
they could oppose the appropriation of their farms. 64
The laws of 1874 and 1882 were especially important as the lands
held by settlers increased in value with the expansion of coffee
production in the west from 1890s forward. Before this period,
Medellin mer- chants interested in coffee invested in haciendas in
the Cundinamarca/ Tolima region, especially around Sasaima. 65 The
first Antioquefio cof- fee farms were established on large
haciendas near MedeUin (Fredonia) in the 1880s. Further expansion
in the 1890s and 1900s occurred in the areas of small-scale
settlement, on the mountain slopes to the south. By 1913, Antioquia
and Caldas had displaced the Santanders as the most important
producing region, creating the basis for a prodi- gious
twentieth-century expans ion . 66
Within and among the three Colombian regions that dominated
Colombian coffee production, we find land and labor regimes that
approach the Brazilian and Costa Rican extremes and that cover a
range of intermediate forms and relations. A rough survey of the
three regions can be quickly sketched. 67 In Santander in its
period of expan- sion and establishment (1840-1900), a form of
sharecropping (aparceria) predominated in which a tenant would be
given a house, food plot, and coffee plot in return for a third to
a half of the coffee produced and a smaller portion of the food
plot's yield. Arango sug- gests that this system emerged after an
early use of wage labor and was a response to labor scarcity and
the need to fix labor on the land. 68 This, in turn, supports
Palacios' contention that Santanderean share- cropping was not
associated with servile social relations. In this view,
sharecropping represented a short-term economic contract that
implied neither a long-term relationship with the land nor a
servile
-
373
relationship with an absentee owner . 69 If outside labor was
necessary for the harvest, it was employed by the sharecropper.
In Cundinamarca, a form of labor rent (arrendamiento) emerged
during its period of expansion and establishment (1875-1900). The
large haciendas on the western slopes of the eastern cordillera
were owned by Bogotfi merchants who hired resident administrators.
Because the subtropical slopes had been relatively open at the
begin- ning of the coffee cycle, the haciendas depended on the
migration of highland Indians and mestizos from Boyacfi and
Cundinamarca who would settle on hacienda lands and be given a
house and access to land for food and livestock production. In
return, they would be expected to provide a contracted number of
days per month on the hacienda's coffee plot. While this was the
most servile of the labor relations to emerge in Colombia,
arrendatarios were in privileged positions in rela- tion to others
such as the casual laborers (voluntarios) hired from the region or
from highland Cundinamarca and Boyacfi for the harvest. Long-term
rental arrangements on haciendas gave the arrendatarios access to a
livelihood; their access to land for corn, beans, sugar, and
livestock production created a space for an alternative commercial
economy within the hacienda and with neighboring towns, of which
some arrendatarios were able to take advantage, hiring voluntarios
to do their obligatory work on the hacienda. TM
In Antioquia during the initial expansion (1885-1905), large
haciendas near Medellin and Fredonia used an intermediate system of
agregados, in which the house alloted to the worker was separate
from the land to be worked, minimizing the possibility that the
agregado could develop an alternative agricultural economy within
the hacienda. 7~ With the spread of coffee cultivation to the
south, however, small-scale commer- cial peasant production was
widespread, and a structure of production and commercialization
similar to that of Costa Rica's Central Valley emerged.
Although it is useful to make an initial distinction between a
structure of production dominated by a commercial peasantry in the
Antio- quefio west and one dominated by haciendas and dependent
tenants in the east, such an opposition needs to be modified by
more careful attention to spatial variation (the existence of
small-scale production in regions dominated by haciendas, and of
haciendas in regions domi- nated by peasants) and to temporal
development. Palacios suggests that the development of coffee
production in Colombia can be divided into
-
374
three broad periods: the hacienda phase (1870-1910), the peasant
phase (1910-1950), and the empresarial phase (small-scale commer-
cial farms, with much greater capital inputs for new strains,
fertilizers, and labor, 1960 to present). 72 On the one hand, the
"peasant phase" of the early twentieth century represents the
growing importance of Antioquia and the southward expansion of
settlement and coffee pro- duction. Yet it also reflects the
fragmentation of hacienda holdings in the center.
To understand the dynamics of this fragmentation, Palacios'
emphasis on the peasant character of Colombian coffee production on
haciendas is especially helpful. He begins by stressing the
frontier character of the Cundinamarca coffee zone, the
implantation of a hacienda regime that involved the investment of
commercial capital from Bogotfi and the immigration of highland
peasants from Boyacfi. But he suggests that it was "an entire
peasant structure" that migrated, meaning that servile relations
from the highlands were successfully implanted in the early decades
but also that household-based production was installed at the very
center of the hacienda regime. 73
This was to be increasingly important as arrendatarios
established commercial production in their food plots and pastures,
and as hacen- dados needed to renovate their coffee plots. Hacienda
administrators complained about the difficulty of enforcing labor
obligations. That is, the manner in which hacendados resolved their
labor problem created the structural space for an alternative
economy within the hacienda, which was increasingly important at
the close of the initial expansion phase. By the 1920s, the crisis
on central haciendas was acute, as peas- ant movements began to
organize, first against the arrendatarios and then in combination
with the arrendatarios against the hacendados. One response of
hacendados was to divide and sell off their estates to peasants and
outsiders, a long process that continued through and beyond the
depression. With this, the "cellular structure" that charac-
terized the organization of production within haciendas became the
basis for a new structure of landed property, and small-scale
property and production became central in the two most important
production zones of the country. TM As in Costa Rica, this
peasantry should not be romanticized: the exploitative relationship
between merchant proces- sors and small producers was crucial. But
the importance of the peas- antry, both within the hacienda regime
and in that regime's collapse, should not be forgotten. 75
-
375
Fields of power: Toward an interpretive framework
This essay began with a paradox - the common transformation of
Latin America's coffee republics in the late nineteenth century and
the radi- cally distinct experience that transformation engendered.
After a brief exploration of one dimension of those distinct
experiences, we need now to ask why these different forms and
relations emerged. My answer, which cannot satisfy those who prefer
their explanations to be more precise and "economical" is that
understanding can only be sought in the comparative discussion
itself. "The determinate 'cause' of such changes" writes Sidney
Mintz concerning another problem, "is a context, or a set of
situations, created by broad economic forces. ''76 In this case, I
have tried to sketch radically different social contexts into which
these broad economic forces were inserted, and I wish now to
suggest that these different contexts "determined" the different
direc- tions the coffee economies took.
However much this may look like an argument that the coffee
econo- mies were different because they were different, the
historical and anthropological understanding that informs it is
more complex and requires elaboration. I have referred at various
points in this essay to specific Paulista or Costa Rican fields of
power. I need now to make my meaning more explicit. Despite the
profilerating use of "power" as a concept in recent literature, my
most direct source for the phrase is Eric Wolf's Peasant Wars of
the Twentieth Century. 77 Characteristically, he defines what he
means by practice rather than explicit precept. The phrase appears
most prominently in his conclusion that, "Ultimately, the decisive
factor in making a peasant rebellion possible lies in the relation
of the peasantry to the field of power which surrounds it. ''78
While he goes on to use a definition of power offered by Richard
Adams, his understanding of the field of power is less susceptible
to codification. It clearly refers to the class structure of which
the peasants are a part - the landlords, merchants, state
officials, capitalist planters, and others who press claims upon or
otherwise threaten peasant liveli- hoods. But his understanding of
the class structure is one that is less dependent on a ready set of
sociological categories than on detailed anthropological and
historical investigation. Earlier, Wolf observed that
the anthropologist is greatly aware of the importance of groups
which medi- ate between the peasant and larger society of which he
forms a part. The landlord, the merchant, the political boss, the
priest stand at the junctures in social, economic, and political
relations that connect the village to wider-
-
376
ranging elites in markets or political networks. In his study of
peasant villages he has learned to recognize their crucial role in
peasant life, and he is per- suaded that they must play a
significant role in peasant involvement in polit- ical upheaval. To
describe such groups, and to locate them in the social field in
which they must maneuver, it is useful to speak of them as
"classes" Classes are for me quite real clusters of people whose
development or de- cline is predicated on particular historical
circumstances, and who act to- gether or against each other in
pursuit of particular interests prompted by these circumstances. In
this perspective, we may ask - in quite concrete terms - how
members of such classes make contact with the peasantry. In our
accounts, therefore, we must transcend the usual anthropological
account of peasants, and seek information also about the larger
society and its constituent class groupings, for the peasant acts
in an arena which also contains allies as well as enemies. This
arena is characteristically a field of political battle. TM
There is much in this statement that bears the marks of the
period, over twenty years ago, in which it was written; there is
also much in it that is extraordinarily refreshing in the context
of theoretical preoccupations that have dominated the literature in
the subsequent twenty years. What Wolf was marking out was less a
confining set of concepts and hypotheses and more a historical and
anthropological attitude, which he then took to his six case
studies of peasant rebellion. It is in these case studies that we
find, in practice, Wolf's concept of a field of power. In his study
of the Mexican Revolution, for example, he begins with the
formation of indigenous peasant communities during the colonial
period, their relations to haciendas, cities, and the colonial
state; the War of Independence and the social and political
transformations of the nineteenth century (the liberal reforms and
the expansion of haciendas, especially under the Diaz regime); the
development of mining and industry in the north. It was only in
this context that he ana- lyzed the various locally focussed
Mexican Revolutions and some of the initial consequences for
regional peasantries of the new Mexico that emerged. In each of the
case studies, an attempt is made to under- stand the formation of a
particular peasantry in terms of its internal relations, forms of
landholding and community, its relations with hacendados,
merchants, the Church, representatives of the state, etc., and to
examine how this complex of relations changes with, say, the
passage of new land laws in Mexico, the end of one colonialism or
the introduction of another, the imposition of a head tax or the
develop- ment of rice plantations in Vietnam. Although he does not
use the lan- guage, each of the case studies can be seen as an
attempt to capture the conjunction of local and global histories,
or to explore the internaliza- tion of the external.
-
377
Although it might have seemed to some reviewers that Wolf's case
studies are "too complex and vague" and that he writes the "least
theo- retically" of the authors who examined peasant rebellions in
the 1960s and 70s, 8~ it should be apparent that his approach to
fields of power is actually well informed by theory. Likewise, the
examination of coffee economies in this essay comes out of a
certain theoretical understand- ing, one that organizes our account
of the different social and cultural contexts in which coffee
economies developed in a certain way. To each of the regions
considered, I take a set of questions that fit comfortably within a
historical materialist framework: the occupation of space and the
transformation of landed property, the mobilization and reproduc-
tion of labor, and (in a discussion to be presented elsewhere) the
organ- ization and capitalization of markets, and the political and
ideological processes associated with state formation and the
emergence of hegemonic blocs.
In addressing these questions, I have tried to avoid the
temptation of filling structural boxes, by locating within each
theme real problems that confronted historical actors - obtaining
title to land, or resisting a land survey, recruiting a labor force
by experimenting with various forms of compensation, pressing the
state to pay for the transport of one's labor- ers, or agitating
for market control in a depression and finding that the control
scheme results in greater foreign domination. It is through
attention to these problems, their varying local solutions, and the
prob- lems created by those solutions that we can sketch the
structure of class relations in Silo Paulo or Antioquia in a way
that pays attention to the action of human subjects and to the
contradictory forms and results of such actions.
Notes
1. This article presents a portion of the summary and argument
contained in my intro- duction for a forthcoming volume on "Coffee,
Society, and Power in Latin Amer- ica," edited by William Roseberry
and Lowell Gudmundson. While the present essay concentrates on
questions of land and labor, the longer introduction explores these
questions in a wider range of countries and also treats questions
of coffee processing, commercialization and trade, as well as class
formation and politics, all of which are necessary for the
comparative interpretation suggested here. The introduction, in
turn, depends upon and was inspired by the essays by Michael
Jimenez, Lowell Gudmundson, Mario Samper, Hector P6rez, Marco
Palacios, Fer- nando Pic6, David McCreery, Verena Stolcke, and
Mauricio Font gathered in the volume. The conference that led to
the volume was generously funded by the Uni- versidad Nacional de
Colombia and the Social Science Research Council, with funds from
the National Endowment for the Humanities.
-
378
2. G. Marcus, "Imagining the Whole," Critique of Anthropology, 9
(3, 1989), 7. 3. G. Marcus, "Contemporary Problems of Ethnography
in the Modern World Sys-
tem," in J. Clifford and G. Marcus, editors, Writing Culture:
The Poetics and Politics of Ethnography (Berkeley, University of
California Press, 1986), 165-193.
4. F.H. Cardoso and E. Faletto, Dependency and Development in
Latin America (Berkeley, University of California Press, 1979),
xvii.
5. Ibid., xvii, xviii. 6. K. Marx and F. Engels, The German
Ideology (New York, International, 1970
l18461), 55-58. 7. J. de Graaf, The Economics of Coffee
(Wageningen, Netherlands, Centre for Agri-
cultural Publishing and Documentation, 1986), 26. 8. U.S.
Department of Commerce, Business and Defense Services
Administration,
Coffee Consumption in the United States, 1920-1965 (Washington,
D.C., 1961), 5. 9. M. Palacios, E1 Card en Colombia, 1850-1970, 2nd
ed. (Mexico City, El Colegio de
M6xico, 1983), 178. 10. K. Marx, Capital, vol. 3. (New York,
International, 1967 [1984]), 791. 11. I. Wallerstein, The Modern
World-System III: The Second Era of Great Expansion of
the Capitalist World-Economy, 1730s-1840 (San Diego, Academic
Press, 1989), 152-153.
12. See L. Gudmundson, "Peasant, Farmer, Proletarian: Class
Formation in a Small- holder Coffee Economy, 1850-1950," Hispanic
American Historical Review, 69 (2, 1989), 221-258; M. Samper,
"Enfrentamiento y Conciliaci6n: Comentarios a Prop6sito de las
Relaciones entre Productores y Beneficiadores de Caf6," Revista de
Historia, Ntimero Especial (1985), 207-212.
13. S. Stein, Vassouras, A Brazilian Coffee County, 1850--1900:
The Role of Planter and Slave in a Plantation Society, 2nd ed.
(Princeton, Princeton University Press, 1985); W. Dean, Rio Claro:
A Brazilian Plantation System, 1820-1920 (Stanford, Stanford
University Press, 1976); T. HoUoway, Immigrants on the Land: Coffee
and Society in S~o Paulo, 1886-1934 (Chapel Hill: University of
North Carolina Press, 1980); V. Stolcke, Coffee Planters, Workers,
and Wives: Class Conflict and Gender Rela- tions on $5o Paulo
Plantations, 1850-1980 (New York: St. Martin's, 1988).
14. A. Machado uses articles written by hacendados outlining the
benefits of new forms of tenancy that they had recently adopted.
Machado uses the articles as evi- dence of particular forms of
sharecropping, but they are also interesting as elite dis- courses,
as planters simultaneously trying to present themselves to each
other in a particular way and trying (publicly) to resolve
increasingly intractable problems as their tenants left the farms
and worked on public works projects. (A. Machado, El Cafd: De la
Aparceria al Capitalismo, [Bogotfi, Punta de Lanza, 1977]
179-199.)
15. See L. Bergad, Coffee and the Growth of Agrarian Capitalism
in Nineteenth Century Puerto Rico (Princeton, Princeton University
Press, 1982), 38.
16. With the depression of the 1930s and the closure of the
European market in World War II, the United States and 14 producing
countries signed the International Cof- fee Agreement of 1940,
setting export quotas for the various countries. The agree- ment
was the first of a series of international control schemes that
stabilized the market and facilitated a dramatic post-war price
increase. It also corresponded with (indeed required) the formation
of national coffee-marketing boards, marking the effective end of
the free-trade model of coffee marketing. Some of these boards had
been formed earlier, during the depression, or, in Brazil, to
administer the valorization schemes.
17. M. Arango, Caf~ e lndustria, 1850-1930 (Bogot~i, Carlos
Valencia Editores, 1977),
-
379
184. The analogy, while suggestive, is inexact. Coffee is
subject to a grading system, at first developed by traders in
consuming countries and in recent decades devel- oped by marketing
boards in producing countries as well. It has never been asso-
ciated with the sort of politically and commercially charged
designation of lands that produce grapes that can be processed into
wines with certain appellations, and within appellations,
designation of grapes and the lands that produce them into grand,
premier, and lesser crus, nor can it be. That a discourse of
quality can give to a coffee processor a control analogous to that
exercised by, say, a wine negociant is, nonetheless, an interesting
possibility.
18. Stein, Vassouras, 3; Dean, Rio Claro, 1-23; C. Hall, El Caf~
y el Desarrollo Hist6rico-Geogrdfico de Costa Rica (San Jos~,
Editorial Costa Rica, 1976); Pala- cios, El Caf~ en Colombia,
passim; D. A. Rangel, Capital y Desarrollo: La Vene- zuela Agraria
(Caracas: Universidad Central, 1969); W. Roseberry, Coffee and
Capitalism in the Venezuelan Andes (Austin, University of Texas
Press, 1983), passim.
19. Palacios, El Caf~ en Colombia, 294. 20. The classic account
for Rio is Stein, Vassouras. 21. An excellent general treatment of
land policy is in E. Viotti da Costa, The Brazilian
Empire: Myths and Histories (Chicago, University of Chicago
Press, 1985), 78-93. For treatments of the conflicts between
squatters and grantholders in Rio and S~to Paulo, see Stein,
Vassouras, 10-17; Dean, Rio Claro, 11-20; HoUoway, Immigrants on
the Land, 112-114.
22. See Holloway, Immigrants on the Land, 113, 120-121. 23.
Dean, Rio Claro, 13. 24. Stein, Vassouras, 55. 25. B. Bums, A
History of Brazil, 2nd ed, (New York, Columbia University
Press,
1980), 189. 26. Viotti da Costa, The Brazilian Empire, 7. 27.
Stein, Vassouras, 35. 28. Stein, Vassouras, 65-67. 29. Dean, Rio
Claro, 89-123; Holloway, Immigrants on the Land, 70-72;
Stolcke,
Coffee Planters, Workers, and Wives, 1-9; Viotti da Costa,
Brazilian Empire, 94- 124.
30. Stolcke, Coffee Planters, Workers, and Wives, 9-16. 31.
Ibid., 17. 32. Holloway, Immigrants on the Land, 35-40. 33. Ibid.,
41. 34. Other important nationalities of immigrants were Spanish,
Portuguese, and Japan-
ese. See ibid., 42-43. 35. Ibid., 50-61. 36. This summary has
depended on descriptions in Stolcke, Coffee Planters, Workers,
and Wives; Holloway, Immigrants on the Land, and Dean, Rio
Claro. 37. Stolcke, Coffee Planters, Workers, and Wives, 28-34. 38.
The best analysis of the occupation of space in Costa Rica is Hall,
El Ca# y el
Desarrollo Hist6rico-Geogr6fico. See as well idem, Costa Rica: A
Geographical Interpretation in Historical Perspective (Boulder,
Westview Press, 1985).
39. L. Gudmundson, Costa Rica before Coffee: Society and Economy
on the Eve of the Export Boom (Baton Rouge, Louisiana State
University Press, 1986).
40. Hall, El Caf~ y el Desarrollo Hist6rico-Geogrtifico, 35-37.
41. J.A. Salas Viquez, "La Bdsqueda de Soluciones al Problema de la
Escasez de
-
380
Tierra en la Frontera Agr/cola: Aproximaci6n al Estudio del
Reformismo Agrario en Costa Rica, 1880-1940." Revista de Historia,
Nfimero Especial (1985), 97- 160.
42. Hall, El Card y el Desarrollo Hist6rico-Geogrdfico, 72-101.
43. See as well M. Samper Kutschbach, "La Especializaci6n Mercantil
Campesina en el
Noroeste del Valle Central: 1850-1900. Elementos Microanaliticos
para un Modelo." Revista de Historia Ntimero Especial (1985),
49-98.
44. M. Seligson, Peasants in Costa Rica (Madison, University of
Wisconsin Press, 1980), 8.
45. Gudmundson, Costa Rica before Coffee, 57. 46. M. Seligson,
Peasants in Costa Rica; E. Fonseca, Costa Rica Colonial (San
JosE,
EDUCA, 1983); C. Hall, El Cafd y el Desarrollo
HistErico-Geogrdfico. 47. For the first view, see Seligson,
Peasants in Costa Rica; for the second, see Hall, El
Caf~ y el Desarrollo HistErico-Geogrdfico. 48. Hall, El Caf~ y
el Desarrollo Hist6rico-GeogrEfico, 85-87. 49. L. Gudmundson,
"Peasant, farmer, proletarian." 50. See the Introduction, cited in
note 1, as well as Hall, El Caf~ y el Desarrollo His-
t6rico-Geogr(~fico, 47-49; G. Peters Solorzano, "La formaci6n
territorial de las grandes fincas de car6 en la Meseta Central:
Estudio de la firma Tournon (1877- 1955)/' Revista de Historia
(9-10, 1980), 81-167; V. H. Acufia Ortega, "Clases sociales y
conflicto social en la econom/a cafetalera costarricense:
productores con- tra beneficiadores: 1932-1936." Revista de
Historia (Nfimero especial, 1985), 181-212.
51. Gudmundson, Costa Rica Before Coffee, 152. 52. C. F. S.
Cardoso, "The formation of the coffee estate in nineteenth-century
Costa
Rica," in Land and Labor in Latin America, ed. K. Duncan and I.
Rutledge, (Cam- bridge, Cambridge University Press, 1977), 194.
53. Hall, El CafO y el Desarrollo Hist6rico-Geogrtifico, 57. 54.
Ibid. All the more interesting, then, the famous mural depicting
Costa Rican econ-
omy and society in San JosE's National Theater. The romanticized
picture of coffee and banana workers shows them all to be white -
an obvious misrepresentation of the banana zone, an accurate
representation of an elite's self-image.
55. Palacios, El Card en Colombia, 73. 56. Machado, El Caf~,
117; Palacios, El Caf~ en Colombia, 70-73. 57. Arango, Caf~ e
Indastria, 47. 58. Palacios, El CafO en Colombia, 169. 59. On
Medellin merchants in the early Cundinamarca coffee economy, see
Arango,
Card e Industria. 60. Palacios, El Caf~ en Colombia, 78-79. 61.
Ibid., 131-169. 62. The classic study of Antioquefio colonization
is J. J. Parsons, The Antique~o Colo-
nization in Western Colombia, 2nd ed. (Berkeley, University of
California Press, 1968). Recent reconsiderations include M.
Palacios, El Caf~ en Colombia, 293- 340; M. Arango, Caf~ e
Industria, 68-87. C. LeGrand's Frontier Expansion and Peasant
Protest in Colombia, 1850-1936 (Albuquerque, University Of New
Mexico Press, 1986) is not limited to Antioquia but offers an
innovative study of the appro- priation of and conflict over public
lands (baldios) in nineteenth- and early twen- tieth-century
Colombia.
63. Machado, El Caf~, 17-32; C. Bergquist, Labor in Latin
America (Stanford, Stan- ford University Press, 1986), 287-290.
-
381
64. This entire discussion depends on LeGrand, Frontier
Expansion, 10-18. See as well Arango, Caf~ e lndustria, 68-87.
65. Arango, Caf~ e Industria. 66. Machado, El Ca#, 117. 67.
Sources for this comparison include Palacios, El Caf~ en Colombia,
187-234;
Arango, Caf~ e Industria, 130-151; Machado, El Cafd, 33-85;
Bergquist, Labor in Latin America, 313-330.
68. Arango, Caf~ e Industria, 149-151. 69. Palacios, El Caf~ en
Colombia, 191. 70. Palacios, El Caf~ en Colombia, 206. 71.
Palacios, El Caf~ en Colombia, 193. 72. Palacios, El Caf~ en
Colombia, 342. 73. Ibid., 171-175. 74. Palacios, El Caf~ en
Colombia, 372-401. See also M. Jimenez, "Traveling Far in
Grandfather's Car: The Life Cycle of Central Colombian Coffee
Estates. The case of Viotfi, Cundinamarca (1900-1930)," Hispanic
American Historical Review 69 (no. 2, 1989), 216.
75. Here again, the limited nature of the present comparison
needs to be emphasized. A full understanding of the respective
fields of power sketched in this essay re- quires consideration of
the relations between small producers and merchants. But the bases
for merchant control, and the special characterisitics of coffee
that make such control possible, are sketched elsewhere. See the
Introduction, cited in note 1, as well as the Costa Rican sources
cited in note 50. For Colombia, see Palacios, El Caf~ en Colombia;
Arango, and Card e lndustria. For other countries, see Rose- berry,
Coffee and Capitalism; Bergad, Coffee and the Growth of Agrarian
Capital- ism.
76. S. Mintz, Sweetness and Power." The Place of Sugar in Modern
History (New York, Viking, 1985), 181.
77. E. Wolf, Peasant Wars of the Twentieth Century (New York:
Harper and Row, 1969).
78. Ibid., 290. 79. Ibid., xii. 80. T. Skocpol, "What Makes
Peasants Revolutionary?" in R. Weller and S. Guggen-
heim, Power and Protest in the Countryside (Durham, 1982), 166,
178.