"Remaking frontiers"

Material Information

"Remaking frontiers" land tenure recovery, collective action, and survival strategies of agro-extractive peasants in the "Babassu Zone," Brazil
Porro, Roberto ( Dissertant )
Oliver-Smith, Anthony ( Thesis advisor )
Wood, Charles H. ( Reviewer )
Hildebrand, Peter E. ( Reviewer )
de Almeida, Alfredo Wagner B. ( Reviewer )
Place of Publication:
Gainesville, Fla.
University of Florida
Publication Date:
Physical Description:
xvi, 365 leaves : ill. ; 29 cm.


Subjects / Keywords:
Agricultural land ( jstor )
Agriculture ( jstor )
Alto ( jstor )
Cassava ( jstor )
Crops ( jstor )
Forests ( jstor )
Land use ( jstor )
Pastures ( jstor )
Peasant class ( jstor )
Rice ( jstor )
Babassu -- Brazil ( lcsh )
Dissertations, Academic -- Latin American Studies -- UF ( lcsh )
Land use -- Brazil ( lcsh )
Land use -- Government policy -- Brazil ( lcsh )
Latin American Studies thesis, M.A ( lcsh )
bibliography ( marcgt )
non-fiction ( marcgt )


The survival and social reproduction of peasant societies encompass changes in land and resource use resulting from socioeconomic and political transformations. since the 1960's peasants in the Brazilian Amazon have dealt with state policies promoting concentration of land ownership, pasture conversion, and the expropriation of their means of production. In Maranhao, the most "rural" of the Brazilian states, land scarcity forced a large peasant society to "remake frontiers", through the redefinition of survival strategies. By having the "remake of frontiers" as a central element for survival, peasants simultaneously introduce new aspects in their social and political identity. I have studied peasant communities in areas of babassu palm secondary forests in Maranhao. Expereiencing disputes with antagonistic social segments seeking conflictive land uses, peasants identified specific political ecologies consisting in the combination of economic alternatives, natural resource management, political expression and participation in state and market oriented initiatives. The processes associated to the recovery of tenure security resulted in changes in peasants' social and political trajectory, mainly expressions of collective action and the new forms of social organizations. Empirical references for this study are from five communities of "agro-extractivist, shifting cultivator" peasants in the Mearim and Grajau valleys, experiencing processes of tenure rights recovery in the 1980s and 1990s. Ethnographic observation, unstructured interviews, and 135 questionnaires addressing demographic, socioeconomic, and resource-allocation aspects served as the basis for the qualitative and quantitative analysis. Predominant outcomes in terms of socioeconomic standards, income composition, consumption patterns and allocation of labor were identified, attesting to the adoption of multiple and heterogeneous survival strategies among and within communities. the results obtained in the study indicate two major characteristics of peasant societies in their "search for new frontiers" in transformed economic and political realities. first, the establishment of a complex and multiple network of social relations. Second, their adoption of survival strategies that constantly reinterpret the various components affecting their livelihood, attributing differential weights according to specific correlation of forces and configurations of biophysical, socioeconomic, and political elements. by "remaking frontiers" peasant s in the so-called "babassu zone" look toward a more harmonic future despite social conflicts, linking the roots of their community's formation with their agency in the emergence and assimilation of new aspects in their identities.
Thesis (M.A.)--University of Florida, 1997.
Includes bibliographical references (leaves 356-364).
General Note:
General Note:
Electronic resources created as part of a prototype UF Institutional Repository and Faculty Papers project by the University of Florida.
Statement of Responsibility:
by Roberto Porro.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
The University of Florida George A. Smathers Libraries respect the intellectual property rights of others and do not claim any copyright interest in this item. This item may be protected by copyright but is made available here under a claim of fair use (17 U.S.C. §107) for non-profit research and educational purposes. Users of this work have responsibility for determining copyright status prior to reusing, publishing or reproducing this item for purposes other than what is allowed by fair use or other copyright exemptions. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder. The Smathers Libraries would like to learn more about this item and invite individuals or organizations to contact Digital Services ( with any additional information they can provide.
Resource Identifier:
002327768 ( ALEPH )
38848115 ( OCLC )
ALT1388 ( NOTIS )

Full Text

Copyright 1997
Roberto Porro

To Noen Felipe, Pedro, and Ana

I would like to thank two institutions for the support they provided during my graduate studies at the University of Florida. I held an Inter-American Foundation fellowship from September 1995 to August 1997. In January 1996, I received a grant from the Brazilian office of the World Wildlife Fund in order to carry out research and field work. I am deeply grateful to these institutions and to the staff associated with them.
My special thanks are due to the members of the supervisory committee with
whom I have learned and shared ideas during the research and writing of the thesis: Dr. Anthony Oliver-Smith, Dr. Charles H. Wood, Dr. Peter E. Hildebrand, and Dr. Alfredo Wagner B. de Almeida. This study could not have been done without their theoretical and methodological support, and constant encouragement. Extensive thanks go to professors that were a constant source of advice and inspiration. I give my gratitude to Drs. Marianne Schmink, Steven Sanderson, Clyde Kiker, Leslie Anderson, and Abe Goldman. I am also grateful to John Dain, Stephen Perz, and to my colleagues and the staff at the Center for Latin American Studies and Tropical Conservation and Development Program.
Over the years since my initial work in 1986 in Maranho, there were many people and institutions that contributed to the realization of this study. I am grateful to the Ford Foundation; to Drs. Robert Wilson, Lawrence Graham, and Carol King at the University of Texas; and to Vilmar Faria, Eugenio Diniz, and Angela Alonso at CEBRAP. Kaye Pyle, iv

Anthony Anderson, Peter May, and Jorge Zimmerman were close advisors in the transition between work in Maranhio and study and research in Gainesville.
The completion of the field research in the Summers of 1996 and 1997 was only possible due to the hospitality and efforts of my hosts and hostesses in five communities in the Mearim and Grajai valleys. My sincere gratitude to the families of Jos6 S. Sobrinho, Raimundo R. de Araujo, Domingos G. da Silva, Martim dos Santos Silva and Maria de Jesus Bringelo, and Cicero Pedro da Silva. Additional support was provided by the staff and board of directors of two non-governmental organizations: ASSEMA and ACESA.
Rather than the product of two periods of field research, this thesis is the result of eight years of work with peasants in Maranhdo. Out of these years of work, discussion and friendship, I would like to specially thank Luciene Figueiredo, Jaime Conrado, Valdener Miranda, Jos6 Carlos Florencio, Teresinha Alvino, Ildeth Sousa, Joaquim Shiraishi, Karl Wirtz, Domingos Dutra, and Adolfo Themme, Eriberto Rembecki, and Klaus Finkan.
For the privilege of having learned from the struggle of their lives, I am sincerely grateful to Manoel da Conceigio, Jos6 M. Carneiro, Manoel Rodrigues Sousa, Manoel de F.Ferreira, Lourengo Teles, Francisco de Paula Rocha, Francisco A. Lima, Ildo Lopes, Maria Alaides, Antonia Brito, Maria Adelaide, Juarez Dias, Antonio Linhares, Raimundo V. dos Santos, Jos6 L. Pereira, Raimunda Gomes, Atanagildo Matos, and Julio Barbosa.
My gratitude is also due to the support of Antonio and Ada Porro, Eugenio V. Caffarelli, and Shiro and Kazuco Miyasaka. Noemi Porro shared with me college, work, and graduate studies experiences, challenges, frustrations and realizations. I thank Noemi and our children Felipe and Pedro for the everyday lessons I learn from them.

ACKNOWLEDGMENTS .............................................. iv
LIST OF TABLES .................................................. viii
LIST OF FIGURES ................................................... x
KEY TO ABBREVIATIONS .......................................... xiv
AB STR A CT ........................................................ xv
1.1 Introduction ................................................. 1
1.2 Survival Strategies of Contemporary Peasants ....................... 3
1.3 Central Question: The Remaking of Frontiers ....................... 19
1.4 Approach and Thesis Organization ............................... 21
1.5 Biophysical and Demographic Dimensions of the Babassu-Zone ......... 29
2.1 Introduction ................................................ 41
2.2 Revisiting Economic Strategies of Peasant Households and Communities.. 42 2.3 Social Organization and Collective Action as Peasant Survival Strategies .. 54 2.4 Formation and Consolidation of a "Shifting-Cultivator Peasantry".. ...... 61
3.1 Introduction ................................................ 88
3.2 Limited Land and Urban Influence: Seasonal Moves, Wages, and Retirees 92 3.3 "Mutirlo": Peasant Collective Action in the Babassu Zone ............ 108
3.4 Agricultural Projects in a "Black" Community ..................... 124
3.5 From Abundance to Scarcity: Internal Struggle in Former Indian Lands .. 134 3.6 Landless Occupation of Forested Lands in the Graja6i Valley .......... 154

4.1 Introduction ............................................... 180
4.2 Community Composition ..................................... 187
4.3 Measuring Distinct Expressions of Changes in Livelihood ............ 198
4.4 Assessing Differentiation through Consumption .................... 212
4.5 Shifting Cultivation within an Agro-Extractive Peasant System ......... 220
4.6 Analyzing Households Individually: Patterns of Income Composition .... 246
5.1 Introduction ............................................... 273
5.2 Significance of Peasants "Remaking their Frontiers" ................ 275
5.3 Final Considerations ......................................... 296
A SAMPLE QUESTIONNAIRE .................................... 300
VALLEY S ................................................... 306
COM POSITION .............................................. 310
"BABASSU ZONE" ........................................... 351
LIST OF REFERENCES ............................................. 356
BIOGRAPHICAL SKETCH ........................................... 365

1.1 Babassu stands and fruit production according major ecological zones ....... 33 1.2 Specific features of babassu ecological zones in Maranhio ................... 33
1.3 Area, population, and demographic density of counties in the Mearim and
Pindar6 micro-regions, Maranhio ................................. 38
3.1 Situation of the 94 plots after the 1981 land conflict and the 1984 agreement with
the rancher in the community of Alto Alegre, Lago da Pedra (MA) ......... 97
3.2 Whereabouts of peasant families who in 1984 participated in an agreement with
the rancher in Alto Alegre ........................................ 99
4.1 Area, location, community (C) and sample (N) composition of the five
communities in terms of tenure rights ............................... 186
4.2 Demographic data of five peasant communities at the Mearim and Graja6 valleys
in Maranho (a): household size and incidence of age groups ............. 188
4.3 Demographic data of five peasant communities at the Mearim and Grajafi valleys
in Maranhio (b): residence in the area according to age group ............ 190
4.4 Demographic data of five peasant communities at the Mearim and Grajau valleys
in Maranh~o (c): origin of the households according to age group ......... 192 4.5 Average number of children and infant mortality according age group ...... 194 4.6 Average consumer-worker ratio according to age group ................ 196
4.7 Indicators of household socioeconomic status (A): housing conditions, land and
material assets (objects) ......................................... 201
4.8 Collective goods owned by producers' associations .................... 208

4.9 Indicators of household socioeconomic status (B): semi-perennial crops, fruit
trees, and livestock ............................................. 209
4.10 Indicators of household socioeconomic status (C): average monthly consumption
of basic staple (rice, beans, manioc flour, sugar, coffee, and soy oil) ........ 215
4.11 Relative weight of consumption products in the average Sao Manoel's household
budget (1991 and 1997) ......................................... 217
4.12 Average annual consumption and cost of food, health, clothing, and transportation
for households in the Mearim and Grajafi valleys, according to age group... 218
4.13 Three-year average area, production, and productivity of agricultural plots in five
communities in the Mearim and Grajafi valleys, according to age group ..... 222
4.14 Three-year average area and production of agricultural plots per consumer unit in
five communities in the Mearim and Grajaii valleys, according to age group .. 223
4.15 Effects of land availability, stage of the settlement, life cycle position, and welfare
improvements in rice and manioc production and consumption ............ 238
4.16 Features of babassu extraction among sample households in four peasant
communities in the Mearim and Grajai valleys ........................ 240
4.17 Average fallow period and number of palms cut in agricultural fields on five
peasant communities in the Mearim and Graja:i valleys .................. 244
4.18 Occurrences of previous mining, urban work, and ranch work among sample
households of five peasant communities in the Mearim and Grajafi valleys ... 247
4.19 Incidence of fishing and hunting habits among sample households of five peasant
communities in the Mearim and Grajaii valleys ....................... 250
4.20 Household distribution according to 1995/1996 rice production, consumption
and purchases in peasant communities in the Mearim and Grajafi valleys ..... 255
4.21 Age group distribution according to 1995/1996 rice production, consumption and
purchases in communities at the Mearim and Grajaii valleys .............. 258
4.22 Distribution of peasant households according to strata of monetary income for
the 1995/1996 agricultural year ................................... 265
4.23 Summary of monetary income composition (in percent values) of sample
households from peasant communities in the Mearim and Grajaii valleys .... 269

Figure PASII
1.1 Peasant allocation of resources when access to land is not a constraint ....... 12
1.2 Peasant allocation of resources constrained by impediments in access to land,
forcing technological changes (agricultural intensification) ................ 14
1.3 Negative (1) and positive (2) land use changes resulting from peasant resource
allocation upon reaching a minimum pool of resources ................... 16
1.4 Maranhao state, Northeast region, and Brazilian Legal Amazon ............ 31
1.5 Four ecological zones in the area of babassu occurrence in Maranhao ....... 32
1.6 Mearim and Pindar6 micro-regions in Maranhao, and location of the five peasant
villages included in this study ..................................... 36
1.7 1985 distribution of number and area of rural holdings in 13 counties in the
Mearim and Grajaii valleys according to producers' category .............. 39
1.8 1985 concentration of land ownership in 13 counties at the Mearim and Graja6i
valleys, according to size of the holding .............................. 40
1.9 Number of holdings and total area for the 1985 distribution of land use classes in
13 counties at the Mearim and Grajafr valleys, according to the predominant
economic activity .............................................. 40
3.1 1985 aerial-photography of Alto Alegre, showing the city and neighbor ranches 96
3.2 1997 situation of land ownership in Alto Alegre: peasant plots and lands sold to
ranchers ..................................................... 98
3.3 Luis Vieira Mesquita cultivates vegetables in his small plot in Alto Alegre, selling
produce in Lago da Pedra's market ................................ 100
3.4 Landscape in Alto Alegre. The MA-245 road crosses village land and connects
it with Lago da Pedra and Bacabal ................................. 101

3.5 Francisco Zeferino, born in Ceara, bought a small cattle herd with his savings
and with the monthly retirement benefits he and his wife receive ........... 104
3.6 Alto Alegre's young babassu extractors contribute with family income ...... 105
3.7 Processes and occurrences associated with major facts, economic activity (EA),
land use (LU), and land cover (LC) in Alto Alegre (1925 1997) .......... 108
3.8 1985 aerial-photography of the area of Sao Manoel village, showing the allocation
of land for annual crops (shifting cultivation), pastures, and forested reserve 113 3.9 Raimundo Erminio da Silva is the manager of Sao Manoel' s "cantina" ...... 117 3.10 Field storage of rice in area of shifting cultivation in Sio Manoel .......... 118
3.11 Labor allocation for manioc flour production in Sao Manoel ............. 120
3.12 View of Sao Manoel's pastures. Raimundo de Lima, the association cowboy for
one month, brings cattle from pasture to the stall ...................... 122
3.13 Processes and occurrences associated with major facts, economic activity (EA),
land use (LU), and land cover (LC) in Sao Manoel (1925 1997) .......... 124
3.14 Martim dos Santos Silva's family of Monte Alegre: the couple, nine children,
son in law, and Martim's mother .................................. 126
3.15 Partial view of Monte Alegre's landscape: banana grove and babassu ....... 129 3.16 Dos Santos and two children at their field in Monte Alegre .............. 131
3.17 Processes and occurrences associated with major facts, economic activity (EA),
land use (LU), and land cover (LC) in Monte Alegre (1888 1997) ........ 134 3.18 1985 aerial-photography of the Sao Jose dos Ricardo area, westward the river 140 3.19 Female labor in manioc flour production in Sao Jos6 dos Ricardo .......... 146
3.20 View of Sao Jos6 dos Ricardo's dwellings at the Graja6 river ............. 148
3.21 Domingos Gomes and his son harvest rice in Sao Jos6 dos Ricardo ........ 149 3.22 Dioclidio Mota in front of his house, observes the swine he raises ......... 151

3.23 Processes and occurrences associated with major facts, economic activity (EA),
land use (LU), and land cover (LC) in SAo Jos6 (1935 1997) ............ 153
3.24 Rice for daily consumption is peeled in an Estrela's household ............ 158
3.25 Partial view of Lagoa Nova's temporary habitations built in 1995, after the
occupation of lands in the CIGRA farm ............................. 162
3.26 Partial view of landscape and recently built habitations in Estrela .......... 164
3.27 Jos6 de Jesus arranges rice in a temporary storage in his field ............. 171
3.28 Household and exchange labor harvesting rice for Dioclcio Oliveira (first on
the right) .................................................... 173
3.29 "Settlers" in Estrela hunt only for household consumption ............... 175
3.30 Processes and occurrences associated with major facts, economic activity (EA),
land use (LU), and land cover (LC) in Estrela (1960 1997) ............. 176
3.31 Synthesis of processes and occurrences associated with economic activities in five
communities in Maranhio ..................................... 179
4.1 Three-year average frequencies for rice, maize, manioc, and bean fields according
to household age group in five communities in the Mearim and Grajafi valleys 225
4.2 Three-year average area of shifting cultivation per household and consumer unit,
according to age group, in five communities in the Mearim and Graja6i valleys 227
4.3 Three-year average size of fields of rice, maize, manioc, and beans for the total
sample's age groups of household in five peasant communities in the Mearim and
G rajaii valleys ................................................ 227
4.4 Three-year household average field size of rice, maize, manioc, and beans, in five
peasant communities in the Mearim and Grajai valleys .................. 228
4.5 Three-year average rice, maize, manioc, and beans' productivity among sample's
households in five communities in the Mearim and Grajafi valleys .......... 229
4.6 Annual production (three-year average) and consumption (averages per consumer
unit) of rice, manioc flour, and beans according to household age group in five
peasant communities in the Mearim and Grajafi valleys .................. 232

4.7. A Annual production (three-year average) and estimated consumption of rice,
manioc flour, and beans according to age group in sample households in the
communities of Alto Alegre and Sao Manoel .........................233
4.7.B3 Annual production (three-year average) and estimated consumption of rice,
manioc flour, and beans according to age group in sample households in the
communities of Sao Jos6 dos Ricardo and Monte Alegre.................234
4.7.C Annual production (three-year average) and estimated consumption of rice,
manioc flour, and beans according to age group in sample households in the
community of Estrela, and for the total household sample ................235
4.8.A 1995 / 1996 rice balance, Alto Alegre .............................. 253
4.8.B 1995 / 1996 rice balance, Sao Manoel .............................253
4.8.C 1995 / 1996 rice balance, Monte Alegre............................254
4.8.D 1995 / 1996 rice balance, Sao Jos& dos Ricardo .......................254
4.8.E 1995/ 1996 rice balance, Estrela ................................. 255
4.9.A 1995/1996 distribution and composition of monetary income of peasant
households in the community of Alto Alegre .........................259
4.9.B 1995/1996 distribution and composition of monetary income of peasant
households in the community of Sao Manoel .........................260
4.9.C 1995/1996 distribution and composition of monetary income of peasant
households in the community of Monte Alegre .......................261
4.9.13 1995/1996 distribution and composition of monetary income of peasant
households in the community of Sao Jos6 dos Ricardo .................262
4.9.E 1995/1996 distribution and composition of monetary income of peasant
households in the community of Estrela ............................263
5.1 Synthesis of processes included in the "remaking firontiers" by peasants in
Maranhao ................................................. 298

ABY13A Associagdo Brasileira das Ind6strias dos Derivados de Babaqu
ACESA Agdo, Comunitiria de Educagao em Safide e Agricultura, da Diocese de
ACR Animaglo, dos CristRos no Meio Rural
ASSEMA. Associaglo, em Areas de, Assentamento, no Estado, do Maranhao
BNB Banco do Nordeste do Brasil
CEB Comunidade Eclesial de Base
CEPAL Comissao Economica de Planejamento para a Am6rica Latina
CGCGM Companhia Geral de Com6rcio do Grio-Pari e Maranhao
COLONE Companhia de Colonizagdo do Nordeste
COMARCO Companhia Maranhense de ColonizagAo
COPPAU Cooperativa dos Produtores Agro-Extrativistas de Lago do Junco
EMAPA Empresa. Maranhense de Pesquisa Agropecuiria
EMATER Empresa de Assistencia T6cnica e Extens&o Rural
FNE Fundo Constitucional do Nordeste
TBGE Instituto Brasileiro de Geografia e Estatistica.
INCRA Instituto national de ColonizagAo e Reforma Agriria
ITERMA Instituto de Colonizaqdo e Terras do Maranhao
MEB Movimento de EvangelizagAo de Base
MER Movimento de Evangelizaoo Rural
mc Ministerio, da Indfistria e Com6rcio
MST Movimento dos Trabalhadores Rurais Sem-Terra
PCAT Programa de Colonizagao do Alto Turi
PIC Projeto Integrado de ColonizagAo
PNRA Programa Nacional de Reforma Agriria
PPM Programa de Povoamento, do Maranhao
PROCERA Programa Especial de Cir6dito para a Refonna Agriria
SUCAM SuperintendEncia de Campanhas e Safide P6blica,
SUDENE Superintendeincia para o Desenvolvimento do Nordeste

Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Arts "REMAKING FRONTIERS": LAND TENURE RECOVERY, COLLECTIVE
Roberto Porro
December 1997
Chairperson: Dr. Anthony Oliver-Smith
Major Department: Center for Latin American Studies
The survival and social reproduction of peasant societies encompass changes in land and resource use resulting from socioeconomic and political transformations. Since the 1960s peasants in the eastern Brazilian Amazon have dealt with state policies promoting concentration of land ownership, pasture conversion, and the expropriation of their means of production. In Maranhio, the most "rural" of the Brazilian states, land scarcity forced a large peasant society to "remake frontiers", through the redefinition of survival strategies. By having the "remake of frontiers" as a central element for survival, peasants simultaneously introduce new aspects in their social and political identity.
I have studied peasant communities in areas of babassu palm secondary forests in Maranhio. Experiencing disputes with antagonistic social segments seeking conflictive XV

land uses, peasants identified specific political ecologies consisting in the combination of economic alternatives, natural resource management, political expression and participation in state and market oriented initiatives. The processes associated to the recovery of tenure security resulted in changes in peasants' social and political trajectory, mainly expressions of collective action and the new forms of social organizations.
Empirical references for this study are from five communities of "agro-extractivist, shifting cultivator" peasants in the Mearim and Graja~i valleys, experiencing processes of tenure rights recovery in the 1980s and 1990s. Ethnographic observation, unstructured interviews, and 135 questionnaires addressing demographic, socioeconomic, and resourceallocation aspects served as the basis for the qualitative and quantitative analysis. Predominant outcomes in terms of socioeconomic standards, income composition, consumption patterns and allocation of labor were identified, attesting to the adoption of multiple and heterogeneous survival strategies among and within communities.
The results obtained in this study indicate two major characteristics of peasant
societies in their "-search for new frontiers" in transformed economic and political realities. First, the establishment of a complex and multiple network of social relations. Second, their adoption of survival strategies that constantly reinterpret the various components affecting their livelihood, attributing differential weights according to specific correlation of forces and configurations of biophysical, socioeconomic, and political elements. By "remaking frontiers!' peasants in the so-called "babassu zone"' look toward a more harmonic future despite social conflicts, linking the roots of their community's formation with their agency in the emergence and assimilation of new aspects in their identities.

1. 1 Introduction
The survival and social reproduction of peasant societies is a subject of growing policy concern and scholarly debate. Socioeconomic transformations at regional and higher levels and alterations in the biophysical environment play a decisive role in both the strategies to be adopted by peasants, and in the outcomes from those adaptative strategies. As a result of dynamic processes and changes, peasants will identify mechanisms to adjust themselves to these distinct domains, or otherwise be dispossessed and succumb.
In Brazil, the end of the slavery economy combined with the decline of the
plantation in the Northeast region fostered in the second half of the nineteenth century the consolidation of a free peasantry, independent rural cultivators no longer subordinated to the direct appropriation of their labor force. Initially occupying lands in the state of Maranh~o, and then moving westward to Parhi and other Amazonian states, in the last twenty five years these peasants have been increasingly struggling for their social reproduction through the identification of procedures that would ensure their survival in spite of increasing threats associated with the reduction of their resource bases brought about by the incorporation of land into the market.

This study addresses the survival strategies of contemporary Brazilian peasant societies confronting state policies resulting in concentration of land ownership, contributing to the "closure of the physical frontier" and promoting the expropriation of peasant means of production. Specifically, the study focuses on peasant economic practices and strategies of social positioning in the context of government-initiated settlement schemes established after land conflicts. Distinct forms of interaction with the environment changes in land and resource use associated with the adoption of alternative economic activities by households and communities -- are viewed as responses to socioeconomic and political transformations. Throughout the study, I intend to connect peasant economic practices with the constraints and scenarios imposed by public policies targeting the so-called rural development, and with the transformation in the relationship between peasants and natural resources.
This study was based on research in social situations (Gluckman, 1958) having as empirical references peasant communities in the Brazilian state of Maranhao. A number of aspects enhance the significance of policies affecting the economy of peasant societies in Maranhao and their utilization of natural resources. Maranhao is the most "rural" of the Brazilian states, and most of this "rural population"' are peasants. Compared to other Amiazonian areas, Maranhao has an older process of peasantry formation, providing more insights in terms of peasant economy and social organization. Also, important are Maranhao's transitional ecological features, with the predominance of secondary forests of babassu palm (Orbignyaphalerata), and the heavy adoption of state incentives and subsidies promoting cattle ranching and concentration of land ownership since the 1970s.

A large portion of the squatters, colonizers, and landless families currently on the agricultural frontiers of the Brazilian Amazon, either have their origin in Maranhio, or have been settled there for a significant period of their lives. As secondary forests of babassu palm are increasingly being observed in areas of later colonization, Maranhio presents a context in which land and forest use are undergoing significant changes that are likely to be experienced elsewhere in the Brazilian Amazon.
1.2 Survival Strategies of Contemporary Peasants:
"Peasants still form a major part of mankind, but their numbers are stationary while their shape in the population of developing societies is declining ... [and]
the livelihoods of those who survive as rural small holders include what has been
considered as non peasant characteristics"(Shanin 1976, Foreword: 22)
The last quarter of the twentieth century has represented the aggravation of processes referred to the modernization theory approach. Modernization principles to stimulate development were accepted and adopted by governments of developing countries and development agencies as references for the creation of policies to be implemented in those countries' rural sector. According to modernization postulates1, peasant societies are an impediment to development. Their agricultural practices do not meet the necessary changes to increase crop productivity, and promote agricultural and industrial integration. According to modernization theory, development policies should transform the so-called "backward" orientation of peasant societies. That would open the door for processes of social differentiation, favoring capital-based over labor-based units IFor a complete description of modernization postulates, see Rostow 1990. The five stages of growth: a summary.

of production, establishing a category of rural entrepreneurs, integrated in the market, able to incorporate values and objectives of the nation-state and contributing to capital accumulation, economic growth, and national development.
Although affected by "development" and "modernization", the ability and
flexibility of peasants to maintain their economy and adapt to critical environments and socioeconomic constraints exceeded the expectations of policy makers and scholars. Hence, peasant societies have been reconceptualized in their survival strategies and compared to other marginal segments of contemporary society (Kearney 1996). The extent to which peasants maintain their social reproduction has been increasingly a function of the adoption of non-conventional economic activities. These transformations, on the other hand, can generate distinct features in what could be called a peasant mode of agricultural production2. Among these distinct features, the opportunities to engage in capitalist relations of production to name the most significant ones do not simply replace functions formerly adopted by peasants. Peasant societies can be transformed as part of a broader process of change including subordination resulting through different entitlement relations applied to ownership (Sen 1981: 7-8)3, alterations in the availability of factors of production, and transformation in accessibility and composition of the resource bases.
One of the central issues in peasant studies is whether or not peasant economy constitutes a mode of production with theoretical status equal to capitalist mode of production. Throughout this study, I will assume that it does. For an overview of this debate, see Deere (1990: 3-11)
3 Sea refers to entitlement relations as rules of legitimacy determined by social relations (1981:1)

Political and socioeconomic processes affect peasant societies in different manners. Factors such as the characteristics of the hegemonic class, extent and historical facts related to peasant occupation, internal division of labor among the household, natural resource characteristics, and previous relations with state, market, and the broader society are only a few indicators of this variability. Whether peasants will become completely destitute and gradually disappear, deal with transformations and reshape their living strategies, or otherwise still be able to retain their basic characteristics, is a function of the interaction among these factors. Complex models could be formulated to assess the outcomes of a wide range of social situations. However, the objective of this study is to investigate peasant societies in areas of consolidated occupation, but still characterized by slash-and-burn, shifting cultivation, although showing initial signals of shortening fallow periods and agricultural intensification4.
Occupation and destitution: peasants on the frontier:
The trajectory of peasant cultivators in areas of free, unoccupied lands
incorporates to the extreme the dual perspective present in Schmink and Wood's definition of frontier, designating both "the (palpable) physical edge of a settled area and the (abstract) battle lines that marked the confrontation between competing
claims"(Schmink and Wood 1992:19). In fact, at the same time that peasants struggle for the physical occupation of new lands, they do so as their best hedge against expropriation and subordination.
' For the dynamics of land utilization associated with agricultural intensification, see Boserup (1965: 15-22); and Pingali and Binswanger (1988).

The importance of areas of free land for the configuration of a society's cultural identity was introduced by Tumer's work about the American expansion to the west (Turner 1921: 311-334; Taylor 1967). Transposed to peasant societies, Tumer's thesis recognizes that availability of land would confer ideals of individualism and personal advancement in the colonizer, leading to capitalist transformations subsequently affecting the society as a whole5.
Comparing Brazilian authoritarian capitalism with Tumer's American democracy, Velho defines as an "expansion front" the spontaneous process of penetration of hitherto unoccupied lands by marginal segments of a society. Different than what occurs in pioneer or agricultural frontiers (in which land is consistently commoditized assuming the form of private property), "expansion fronts'" are not only peripheral to the geographic center, they are also marginal to the predominant economic system and marginal to the notion of nation-state. Small producers occupying lands on the frontier, thus, are only indirectly subordinated to capitalist accumulation, initially finding a favorable milieu for the social reproduction of a peasant economy. As Velho pointed out, "the frontier, when it opens, seems to represent a privileged locus for the development of small agriculture. In the case of the peasantry, he is no longer a left-over from an early formation ... in this case he is ascending" (Velho 1973: 98).
5 For the applicability of Turner thesis to the context of peasant societies, see Velho 1973. Modes of capitalist development, peasantry and the moving frontier.
6 For the definition of expansion fronts in Brazilian Amazon, see Velho 1972. Frentes de expansion e estrutura agiria.

State incentives for the occupation of the frontier, however, have historically targeted the retardation of processes of social diferentiation and class consciousness, serving as a safety-valve for the alleviation of social tensions resulting from demographic growth, distortions of economic development, and class conflict in areas of greater consolidation of the productive forces. In effect, peasant occupation of the "expansion front" fits into a number of assertions of the articulation of modes of production theory. First, peasants in the "expansion front" produce food surpluses of relatively lower standards and prices, to be channeled to local and regional markets. Second, the "frontier" absorbs labor surpluses expelled from areas of more intense economic activity, and at the same time consists of a reservoir of the same labor for eventual future needs, either on capitalist farms and ranches, or for urban centers' economic growth7. Third, they execute a great deal of work in clearing areas of primary forests, "domesticating the landscape"' for modernized systems of agricultural production.
Following initial occupation, capitalist economic interests tend to encroach upon the "expansion front". In waves of progressive surplus extraction, peasants are increasingly forced to sell or concede what they have built, sometimes to the better-off members of their original group, but mainly to outside elements of other societal strata. Primitive accumulation through unequal commercial exchange, the predominant type of extraction associated with initial stages of occupation, gives way to the direct destitution of peasants from their land and from their access to natural resources, ultimately threatening their survival. Contrasting with the previous period, this process of peasant
-'w features of the finctional dualism referred by de Janvry (1981: 39). The agrarian question.

destitution is facilitated by the presence of an institutional apparatus supporting social differentiation, capitalist class relations, and concentration of land ownership. Insofar as land will be available for new cycles of extensification and destitution, and especially when firm cultural roots were not established at that site, progressive moves tend to be adopted by peasants. When, however, resources are seen as exhausted (or when the distance and costs associated to out-migration become excessive), combined with the empowerment of peasant communities and the establishment of steadier links with the occupied site, the alternative is to remain. Remaining physically, however, does not assure maintaining their original peasant identity. Remaining could also mean subordination through proletarization. It could mean a progressive process of acquiring diverse economic practices and values. It could mean the loss of peasant attributes. This study seeks to provide empirical references that in the last two decades, peasants in Maranho have been aware of the imminent threat to their social existence. The loss of the physical frontier alternative (or at least a considerable reduction in its feasibility) resulted in the identification of a series of practices and mechanisms to neutralize or reduce the impact of increased integration within the capitalist economy, maintaining them as a peasantry. Peasant production_ resource use and environmental degradation in consolidated frontiers:
Contrary to the system of production of most of the so-called smallholder (Netting 1993: 2-3) type of peasant, land and natural resources and not labor, and to some extent, technology are the most important resources and constraints for peasant
I Netting defines smallholders as "rural cultivators practicing intensive, pennanent, diversified agriculture on relatively small farms in areas of dense population".

societies in the "expansion front", usually located within forested areas. For these societies, nature's provision of a variety of products and the practice of rotational agriculture through the adoption of adequate fallow periods is fundamental for survival. Given the almost complete lack of auxiliary means of production, relying basically on their labor and nature, peasants find in the slash-and-bum system in forested areas a very convenient practice. Whether this practice is to be considered sustainable or not9, however, depends upon demographic, socioeconomic, and political factors.
In fact, peasants have been considered, according to the source, either responsible for or victims of environmental degradation, or both"0. Especially when the frontier is closed, forms of access to land and to natural resources for subsistence constitute a dividing point in the characterization of peasant societies. This can be summarized by Scott Cook's statement that "peasants, like pre-industrial societies, are entirely constrained by the existence of satisfactory environmental factors to direct their productive effort on the provision of the material means of social reproduction"(Cook
1973: 46). Conditions determining the availability and forms of access to land will affect the outcomes in terms of peasant economic practices, and in these practices' predictability.
Alterations in land use and changes in land cover and in the global climate affect affluent sectors of the western society, motivating growing concern in scholarly research and in the work of international agencies dealing with rural development. This study
9 according to Kasperson, J.X., Kasperson, R.E., and Turner II, B.L. (1996: 6), a sustainable environment is one where resource-use systems and the level of human well-being can be maintained and options for future generations preserved over long periods of time. 10 Broad, R. 1994, gives an overview of the rural poor actions upon environmental degradation.

supports the view of environmental degradation in areas occupied by peasants and small farmers as an expression of the disequilibrium resulting from social inequalities and political economic determinants. Small cultivators' adoption of practices leading to degradation, however, is not considered here as an unintended consequence of this process. Pressured and threatened in their existence, peasants consciously reshape their survival strategies, and reconstruct their perception and interaction vis-a-vis natural resources and the environment. To the already poor small cultivators, but still able to control their way to make a living, the destitution of the means of production causes, as Annis argues, a disincentive to the care of their resource management (Annis 1992: 11). On the other hand, we argue that scarcity of resources; and the lack of economic alternatives, impel small cultivators to reconsider such a position in the long run, identifying more sustainable resource uses among the range of available options. Without generalizing Broad's viewpoint that peasants tend to act as environmental sustainers (Broad 1994: 812) when their livelihood is threatened by degradation of fragile ecosystems and once establishing firm roots and ties in their settlement, I argue that progressive reduction in resource availability, the loss of the image of the unlimited good", or the minimization of what Sahlins terms as the natural state of maximum dispersion (Sahlins 1972: 97-98), stimulates peasants to alter their strategies of survival.
To explain this argument, I depart from situations of open frontier, with no significant impediments in access to land, and socially regulated access to natural resources for those that I term the "agro-extractive, shifting-cultivator peasants". As "See Velho (1973: 100-101) forthe reversion of George Foster's argument.

socially regulated access I refer to forms of land tenure associating individually appropriated fields for annual crops with areas for the common use of resources essential for the economic realization of the households (Almeida 1989: 187). As Almeida points out, individual annual fields are established according to the wish of the households, with no appropriation of the land itself, and no obligation of territorial continuity. In such circumstances peasants will pursue a combination of activities maximizing returns for the performed labor (maximizing returns to labor), through engaging in a predominant petty class relation (ownership of the means of production and control over the labor process). When land and natural resources are abundant (or at least sufficient), this assumption regulates the economic allocation of peasant labor at diferent stages of the household life cycle. The allocation of factors of production targeting subsistence will result in long-term maintenance of ecological conditions, guaranteeing the conservation of natural resources, and consequently ensuring peasant social reproduction. L. Anderson stated that "knowing that they depend on nature for subsistence and survival, peasants watch their natural environment and protect it insofar as they can"(Anderson, L. 1994: 12). Empirical evidences show, however, that this is not an inherent and everlasting characteristic of peasants. The challenge is to understand the extent of each community's capacity to perform this protection, in the face of internal and external factors, and the momentum in which that conduct becomes no longer attainable. Figure 1. 1 summarizes the social situation resulting from socioeconomic and biophysical scenarios associated with no significant constraints in resource availability (open frontier).

no restrictions on land and natural resource access
mechanisms for social regulation (community)
_________V______control of the means of production adoption of a pool of activities
control over the labor process maximizing returns to labor
long-term ecological balance
conservation and rational use of resources
social reproduction of the peasant household
Figure 1. 1: Peasant allocation of resources when access to land is not a constraint.
As a consequence of unequal relations of power generally associated with restrictions imposed on peasants by capitalist class relations, Deere and de Janvry identified different forms of appropriation of peasant surplus labor. Such mechanisms are rents (in labor services, ind, or cash), the appropriation of surplus value through wages, the appropriation through prices, through usury, and state taxation. (Deere and de Janvry 1979). In addition, I would add that although not being directly appropriated by another social segments, restrictions on access to land and other resources reduce the productive capacity of peasant labor, violating basic principles of their economic calculation. Taussig exemplifies with occurrences in the Colombian Cauca Valley, how capitalist class 'surplus

accumulation relates to the reduction of the size of peasant holdings (Taussig 1978: 66). Particularly in social groups adopting shifting cultivation, restrictions in access to land will disorganize agricultural production due to the shortening of adequate fallow periods, reducing the recycling of organic material, and diminishing yields per area sown. In addition, an extra demand for labor will be required due to increased weeding. Reduced yields will provoke insufficient net household income, or inadequate levels of consumption, subsequently affecting the productive capacity of the household in the next agricultural cycle (Deere 1990: 267). Insofar as the "expansion front" exists, it will consist in the main alternative to the "shifting-cultivator peasant". When, however, the physical frontier is no longer a feasible alternative, there are set conditions for rearrangements in the mode of production, or for its modification. Unlike the process of demographic differentiation, derived from the very nature of the peasant farm, concentration of land ownership resulting from public policies, market mechanisms, and unequal entitlement relations, will force peasant families to reconsider their economic calculations and reallocate remaining resources. Up to a point, as illustrated in Figure 1.2, peasants will carry on transformations in their productive activities maintaining petty relations of production. Although altering productive practices, they however are still able to consider the utility function of maximization of returns to labor. In this case, intensification of labor and land use transform the "shifting cultivator-peasant" to a category increasingly dependent upon labor intensive technologies, the "small farmerpeasant".

public policies I state intervention asymmetrical market exchange unequal entit Y ent relations
restrictions in the access land and natural resources
lower land / natural resource availability: shorter fallow periods, lower recycling higher labor and, diminish yields
reduction of the productive exasinfro
capacity of peasant labor: exasinI
insufficient household income changes in productive
inadequate consumption IIactivities / practices
affecting next cycle of (intensification of use
production of land / labor force)
Figure 1.2 Peasant allocation of resources constrained by impediments in access
to land, forcing technological changes (agricultural intensification)
The situation summarized by Figure 1.2 occurs when forces promoting land
incorporation in the market are not yet completely manifested, and there is a relatively weak opposition of interests between peasants and rural entrepreneurs. More often, however, the outcome of this process (of concentration of land ownership) consists of peasants remaining with a quantitative and qualitative minimum pool of resources, forcing transformations that go beyond technological change. When that occurs, mainly if

combined with increases in the size of the social group, remaining resources will not be enough to sustain consumption needs if peasants only sustain petty relations of production. Falling agricultural yields, when this stage is reached, already provoked in the "Cagro-extractive, shifting cultivator-peasant" the loss of confidence in his/her previous survival strategies. Moreover, worsening economic conditions will activate a variety of more intense responses, including seasonal dislocations, the establishment of coalitions with sectors of the civil society targeting political empowerment, or the adoption of a variety of forms of collective action. Most important, however, is their parallel engagement in other class relations, mainly through selling their labor-force in the market, if existent. Social differentiation among the peasantry, in this case, is encouraged, as well as the individual internal differentiation.
Maximumn returns to labor will no longer be achievable at this stage, resulting in higher levels of self-exploitation, and lowering household well-being. In these circumstances, Chayanov's postulates will not be adequate to analyze peasant economic practices, and the concepts of household and class relations, as well as Kearney's critiques of the applicability of decision-making models will prevail."2
Figure 1.3 illustrates a scenario in which the stock of land available for peasants to engage in shifting cultivation decreases to a minimum. When is reached this point (which varies according to characteristics of nature and of the social group), and when peasants do not establish proper coalitions for alternative provisions for their livelihood, allocation of resources in production results in environmental degradation. The development of this '2ayanov, Deere, and Keamney theories are explained in Chapter 2.

scenario nourish one of the predominant paradigms within the field of conservation and development, expressed by the World Commission on Environment and Development: "those who are poor and hungry will often destroy their immediate environment in order to survive (WCED 1987: 28).
restrictions in the access to land and natural resources It
quantitative and qualitative minimum pool of resources
increasing size of the social group / more pressures exhaustion of remaining resources
environmental degradation land use change (1)
engagement in multiple class relations establishment of coalitions
internal social differentiation seasonal migration
collective action
environmental consciousness" 12 nd use change (2)
Figure 1.3 Negative (1) and positive (2) land use changes resulting from peasant
resource allocation upon reaching a minimum pool of resources.

Nevertheless, such a degradation, as L. Anderson mentions, is usually part of a long process in which many factors have preceded and have pushed peasants to a point where they have no choice (Anderson, L. 1994: 196). Political economic analysis of natural resource degradation, in fact, presents perspectives moving beyond blaming the poor. According to Broad's analysis, these perspectives include three major components. First, by focusing on the inquiries into the ultimate causes of poverty rather than on poverty or the poor themselves. Second, through evidences and arguments of the creation of an "environmental consciousness" among poor people, acting not as environmental degraders but as sustainers. Third, through the appearance of new social movements in which the poor become environmental activists (Broad 1994: 813). Resource allocation and peasant survival in the babassu zone:
In the area of babassu secondary forests, in the Brazilian state of Maranhio, trends of the political economy have been illustrating the arguments expressed above. A large peasant society, initially responsible for most of the land occupation, confronts from the late 1960s a series of policies aiming in the name of development to replace small cultivators by large holdings for cattle ranching. As agro-extractivists, combining agriculture with extractive activity of babassu palm products, these peasants were not only able to settle and cultivate the land, but also to elevate the state of MaranhAo to the position of one of the largest producers of rice and manioc flour two of the country's major food staples. A sustainable type of shifting cultivation, benefitted by babassu's provision of biomass and fire resistence, occurred so long as land was not considered a commodity. Although in poverty and subordinated through commercial exploitation,

peasant societies seemed secure and self sustainable in their holdings. In a parallel to what Taussig shows for the Cauca valley in Colombia (Taussig 1978: 66), the reduction of peasant holdings in Maranhio was a consequence of the political economy and social antagonisms rather than from demographic growth. Immediate impacts on peasant societies resulted from subsidized credit for pasture conversion and the state government's intent to incorporate Maranho's lands into the market. Indirect, but critical transformations resulted from agricultural policies related to the development of soybean production, and advances in the generation of synthetic raw materials, both for the substitution of industrial utilization of babassu oil. Recent changes in commercial and tariff policies vis-a-vis the import of southeast Asian palm, palm-kernel, and copra oil were added to this picture of apparent hopelessness for Maranhio's agro-extractive peasants.
Facing these and other constraints, peasants encountered a temporary alternative in out-migration, either to the westward expansion front or to urban centers. Since the mid 1980s, however, the occupation of most of the nearby available frontier has been discouraging such strategy. Moreover, peasants are now increasingly facing land scarcity, and forced to identify new survival strategies. This process includes their repositioning towards plural society, the engagement in capitalist relations of production, changes in household division of labor, and a reinterpretation of their relation with the natural environment. Besides dealing with environmental constraints, and defining a pool of economic activities maintaining their livelihood and generating surpluses for the larger society, the outcomes of this search also depend on the specific strategies peasants will adopt to achieve broader political objectives, specially the acquisition of citizenry. In this

context, the establishment of new forms of social organization and adoption of collective action are critical for the social reproduction of this peasantry within a transformed reality.
1.3 Central Question: The Remaking of a Frontie
Facing many pressures and threats, peasant societies in developing countries have demonstrated their ability to identify and engage in specific political ecologies represented by the combination of economic alternatives, rules regarding management of natural resources, and growing political expression. In the Eastern Amazon, this flexibility addresses the "closure" of the physical frontier with an increased participation in state and market-oriented initiatives. The eventual engagement in capitalist relations of production, the adoption of collective action, and changes regarding natural resource management are components of these political ecologies.
In the case of babassu areas in MaranhAo, in the Mearim and Grajau valleys,
intense disputes with antagonistic social groups seeking conflictive uses for their lands resulted in peasant undertaking, from the late 1970s, onwards, of an effective "remaking of the frontier". Rather than the frontier defined as the physical edge of unexplored lands, peasants are building abstract spaces where they are able to counterbalance strong political economy threats and to reproduce themselves socially. The "remaking of abstract frontiers"' by peasants materializes through a series of processes of social mobilization, including the balance of seemingly passive forms of resistence with eruptions of collective action in response to land conflicts, the active manifestation of their class representatives towards the state, the formalization of new categories of social organization, and the

constitution of proper alliances and/or partnerships with other sectors of the civil society. Peasant social mobilization in the babassu zone provoked an agrarian transformation that if not massive, demonstrated their stake in refusing to give up the limited parcels of land, a resource that realizes their social condition. Beyond that, the recovery of tenure security through a number of distinct instruments served as a basis for the start of another phase in their economic trajectory, incorporating what Albert Hirschman refers to as the "principle of conservation and mutation of social energy", which he defines as the phenomenon through which "early participation in public action of one kind leads later to involvements in collective endeavors of a very different nature"(Hirschman 1988: 8-11).
In this trajectory of dealing with a transformed reality, peasants consciously adopt multiple strategies to survive. Heterogeneous economic calculations, and distinct attitudes toward the plural society, towards the state and the market, towards their resource base, and towards their own communities and households, are important factors that maintained them as peasants in a transformed universe. Peasants thus are not free riders in the process of environmental change associated with socioeconomic transformations brought by the implementation of development policies and programs. In contrast, being constantly threatened in their survival, their economic strategies targeting the provision and reproduction of the household have to address their resource base in a flexible way. Their engagement in multiple strategies of survival adequately deals with the restriction or loss of accessibility to land and natural resources. Consequently, ecological alterations will be manifested to a lesser or greater extent according to the resilience of each ecological site, and the resilience of the peasant society itself.

1.4 Approach and Thesis Organization
This study analyzes the adoption of distinct survival strategies by peasants in contemporary Maranhilo, specifically in areas where they were able to recover tenure security. Social situations resulting from the dynamic of the political economy affecting peasants in the babassu zone are differentiated among and within communities. The understanding of this heterogeneity is essential for a correct assessment of socioeconomic and biophysical transformations at local, regional, and national scales.
Although the more significant occupation of the region dates from the 1950s, and acute peasant destitution began in the 1970s, it is possible to identify in a snap-shot taken in the present, the reproduction of generic social situations characterizing the various stages of this process. That can be verified even when restricted to the subset of areas in which peasants recovered or consolidated their land tenure.
In order to access such a diversity, the empirical references used in this thesis are based on social situations encountered in five peasant communities that have accessed land through different mechanisms. For certain purposes, additional stratifications are made in each of these situations according to household structure and life cycle position. Besides expressing a wide range of mechanisms used for the recovery of their tenure rights, the five situations to be focused on reproduce processes of initial peasant occupation (or settlement) on lands recently incorporated into a "remade frontier" (1); the destitution of peasants from most of their resource base (2); collective action through new forms of social organization (3); the effective recovery of tenure rights, either through strong

connections with state agencies (4); or through the support and mediation of nongovernmental institutions (5). Relevant socioeconomic aspects of this society such as growing social differentiation, the constitution of capitalist and communal relations of production, the subordination through commercial exploitation, and initial transformations in the system of production (agricultural intensification) are observed throughout both a discursive representation and the introduction of quantitative data.
After explaining the methodology used in this thesis, the remainder of this chapter introduces the biophysical dimension and a demographic representation of the research site, highlighting its concentration of land ownership. Chapter 2 consists of a presentation of the range of theories associated with the study of peasant economics and social organization in general, and the study in particular of the political economy associated with the formation of a free peasantry in Maranh~o, its dispossession, and resistence. Also, Chapter 2 relates the research questions of this thesis with the literature. Then, through the presentation of the social situations, I intend to characterize and analyze this phase in which "shifting-cultivator peasants" of the babassu zone in Maranhiio, although remaining on their land, are in effect challenging hegemonic sectors of the plural society through a series of social events, and through being able to "open a new frontier". This frontier, though not substantially improving their standard of living, fulfills the critical role of guaranteeing survival. I seek to demonstrate how the dynamic of the political economy is captured by peasant households and reflected in diverse allocations of their factors of production, in the reformulation of their economic alternatives, and through new forms of social organization. Chapters 3 to 5 describe and analyze how peasants in Maranhio

responded to the socio-historical context and adapted their practices to political economy constraints. Chapter 3 consists of an ethnographic analysis of social processes, relations of production, economic activities, and forms of land use encountered in those situations. Chapter 4, through the interpretation of recently obtained quantitative data, portrays communities' demographic and socioeconomic features, including the analysis of household income composition. The discussion provided in Chapter 5 relates peasant economic strategies and resulting patterns of land use and natural resource management to the political ecology of each social situation. The conclusion of this thesis addresses the practical and theoretical significance of its findings and relates its results to broader theories.
Methodol ov:
The major concern of the methodology used in this thesis was to integrate the
theoretical basis with empirical evidence from the field. Bibliographic investigation for this study covers the various approaches of peasant economics and decision-making. Particular attention was given to the works of A.V. Chayanov, the evolution of Marxist thought regarding peasant societies and rural development, and the debate between James Scott and Samuel Popkin concerning peasant's "moral or rational economy". In addition, this study refers to the theories of articulation of the modes of production (de Janvry, Taussig), household and multiple class relations (Deere), the political ecology of the contemporary peasant (L. Anderson), and the ultimate reconceptualization of the peasantry (M. Kearney).

The bibliographic study of socioeconomic and biophysical characteristics of the babassu area in Maranhao was conducted through the works of A.B. Anderson and P. H. May. In addition, valuable historical and geographic information were obtained from geographers 0. Valverde and M.C. Andrade, and authors of Maranhao's economic development. The study of the formation of a free peasant society in Maranhao had as its major sources the works of Velho, Musumeci, and Almeida. The latter author was also the reference for the study of the role of contemporary social movements in Maranhao.
Theoretical postulates and assertions of peasant economics were applied to field observations in rural Maranhao. My knowledge of rural realities in that region constituted a process of cumulative reflections and interactions since 1986, when I first started working in Maranhao as a practitioner in rural development. Most of the thoughts and questions included in this study had their origin in the eight and a half years of close contacts with peasant communities and rural grassroots organizations. Moreover, my commitment to graduate study was itself a response to the complexity of evolving processes connected to the impact of rural development in those peasant communities.
During the period of my work in Maranhao, I focused on technical support for
peasant organizations to become involved in improving their living conditions and to build alternative economic strategies. From this period, it was possible to identify villages and communities that could suitably provide insights in this study. In addition, auxiliary reports, data, and publications were obtained from the non-governmental institutions in which I have previously worked. Initially, for three years, I jointly-coordinated ACESA (Aglo Comunitiria de Educaqo em Safzde e Agricultura), the "Community Action for

Education in Health and Agriculture", an organization created by the local Catholic Church, through German Franciscan friars. ACESA assisted peasant communities having experienced land struggles in the Dioceses of Bacabal, and seeking technical support for the moments following the conflict. Subsequently, I participated in the creation of ASSEMA (Associagio em Areas de Assentamento no Estado do Maranhao), the Association in Settlement Areas in the State of Maranhao, an institution conceived and built by leaders of peasant communities in the Mearim and Grajafi river valleys, and that conjoins the features of grassroots mobilization and empowerment, with the provision of technical support in rural development programs. Beyond documents and data, this previous working experience served as an enormous facilitator for my acceptance among peasant households, and for the merging of theoretical frameworks with concrete situations.
Field research for the specific scope of this study was conducted in the summer of 1996, and in a three-week period in May of 1997. During residence in the five peasant communities, all but one of which I was already familiar with, I focused my work on unstructured interviews with community members (most of them recorded), and the application of a questionnaire. Unstructured interviews stressed aspects related to the household arrival in the area, conflicts over land tenure, resistence and consolidation, as well as perceptions of the conditions of the local resource base, and future expectations. The questionnaire was applied to a total of 135 households, with distribution ranging from 22 to 31 questionnaires per community. The questionnaire, included in Appendix 1 of this thesis, focused on quantitative data of the household, especially referring to demographic

aspects, socioeconomic conditions, tenure security, agricultural systems, consumption needs, and the role of the extractive activity and other economic strategies. In addition to the questionnaires and interviews, another goal of the field work was the actual recognition of the predominant landscapes associated with each community's system of production. For this purpose, a photographic documentation was conducted.
Parallel to the socio-historical analysis of each peasant community stage in the process of integration within capitalism, the survey provided quantitative data to be transferred to spreadsheets for the purpose of providing a comparative assessment of socioeconomic indicators among and within social situations. For the purpose of this study, I have identified predominant outcomes of decision making processes regarding allocation of labor and resources within alternative economic activities. As will be shown in Chapter 4, most of these outcomes were stratified by community and by the three major household groups within each community, according to their position in the life cycle, designated as "age groups".
Social situations, communities, and villages in MaranhAo's peasant society:
A methodological question that appeared during the field research was how to
focus the broadness of the survey. I intended to address the issue of peasant occupation of lands acquired through processes of collective action, which culminated in forcing the state to recognize their tenure rights through the initiation of agrarian reform measures. Therefore, it is necessary to finmit the object of this research by establishing boundaries to determine the physical sites (areas occupied by peasants, some of them formally called

settlement areas"3), and to identify peasants who (re)acquired tenure rights over this physical site. However, when one goes to the field, neither are the settlement areas the only sites in which these families (settlers = "assentados') are living in and working on, nor are the settlers' families the only ones working and living in the area. In addition, maintaining one of the essential features of this peasantry"., settlers in these situations seldom appropriate individual plots, in contrast to the usual pattern of colonization schemes. Neither is there a pattern differentiating the dwelling site for families who have and who do not have tenure rights. And finally, there are areas not targeted by official settlement policies, and in which other mechanisms for tenure recovery were assessed by peasants themselves, who in the sequence attempt to obtain official recognition (and benefits) as "land reform"' beneficiaries. Therefore, the concept of social situation, coined by Gluckmnan (1979 [1958]) is used to overcome limitations imposed by terms such as settlement area, village, or even community". it is a necessary concept to define the object of the survey, because even restricting the study to areas experiencing land struggles and the so-called settlements (official or not), there is a wide range of patterns of habitation and disposition in regard to the land/property in focus.
13 Settlements ("assentamentos") in this thesis refer to the official designation that the Brazilian agrarian agency (INCRA) and other government institutions attribute to sites for the instalatian of agrarian reform schemes.
14 See section 2.4 for details about forms of appropriation of land by peasants in MaranhIo. According to Gluckmnan, it is convenient to speak of a social situation when an event is studied in the field of sociology. Gluckmnan pointed out that the analysis of social situations "reveals the underlying system of relationships between the social structure of the community, the parts of the social structure, the physical environment, and the physiological Iffe of the community's members ." (Gluckman 1979 [1958]: 9)

In their initial stages of land occupation, peasants in the Mearim and Graja6 valleys lived in small hamlets (centros) (Velho 1973: 224-225), a limited agglomeration near their agricultural fields. Some of the "centers" became larger villages (povoados), that in addition to household dwellings may incorporate sites for commercial, religious, and educational functions. As the creation of settlement areas is relatively recent (less than two decades), patterns of settlers' habitation will be highly influenced by previous existence of villages inside or in the vicinities of the expropriated (or acquired) land. In the case of an existing village outside the settlement area (often close to its boundaries), it is frequent that a number of families maintain residency in this village, at least for a certain time. Peasant villages in this region vary in size, from a small, family-based neighborhood with less than a dozen households, to large villages, with a couple hundred families, especially in those geographically privileged areas close to a water source or to a major road. The average village in this part of the state, however, is composed of 30 to 45 households (SUCAM, 1990). The village is thus a situational concept associated with a physical site that includes all people living in relative geographic proximity. Within the village there is no exclusion due to cultural, social, economic, or political attributes. When the concept of community is introduced in this region of Maranhio, it reflects a process of diferentiation within the village. Although in most cases there is no common cultural and historical heritage, the concept of community is usually associated to the social group having common social positioning. A peasant community is thus a subset of a village. When there are only peasants in the village, both institutions refer to the same social segment. Hence, even in settlement areas, the concept of peasant community is wide enough to include,

besides families engaged in the settlement program, the landless, the smalholder, and even medium holders, as well as peasant specialists (artisans, carpenters, etc) and those involved in commercial activities. In sum, a peasant community might include all those that in spite of living in the village, do not participate in or are benefitted by settlement projects. In addition, community members participate in other institutions according to their interests. This is the case of associations of producers, religious based communities, cooperatives, rural union's representations, and women's groups.
1.5 Biophysical and Demographic Dimensions of the Babassu-Zone in Maranhao
This section introduces the biophysical and demographic features of the area of babassu secondary forests, and specifically, the geographic micro-regions where the five communities focused in this study are located: Alto Alegre, Sao Manoel, Sao Jos6, Monte Alegre, and Estrela.
Biophysical dimension:
The state of Maranhao contains biophysical features of the Brazilian Northeast (to which it belongs) and of the two surrounding geographic regions: North and Center-West. Figure 1.4 shows the geographic location of the state, and its dual categorization as it is included in the Northeast region, and considered part of the Brazilian Legal Amazon.
Maranhao's eastern portion limits and resembles the dryer climate and vegetation of its neighbor, Piaui state. Its southern part exhibits features of the "cerrado" of the Brazilian central plateau. Its western fringe contains the remnants of the tropical rain

forest originally existing in a large part of the state's 324,616 km2. As a transition of these three ecological zones, and containing characteristics of them, a large area of the state is covered by secondary forests of babassu palm, a dominant species in the regeneration of deforested areas in the eastern and southern borders of the Brazilian Amazon. Babassu occurs in almost 200,000 km2 of the Brazilian territory (May 1990: 49), but its main area of occurrence, for convenience termed here as "the babassu zone"', lies from 2 to 7 degrees of latitude south, and between 42 and 48 degrees of longitude west. Babassu kernels produced in this zone during the 1990-1993 period represented 98% of the national production (LBGE. 1991 1994). The "babassu zone" has an annual rainfall ranging from 1,000 to 2,000 mm, heavily concentrated in the December-June period.
Average annual temperatures in Maranhao range from 24 to 280 C, and most of the babassu stands occur in altitudes of less than 200 meters. A. B. Anderson associates high babassu densities with relatively fertile soils such as Ustalf, Udalf and alluvial soils from the Pamaiba, Itapecuru, Mearim, Graja6, and Pindar6 river basins. In Maranhao, the babassu region (Figure 1.5), has four distinct ecological zones, designated by a 1982 Brazilian Ministry of Industry and Commerce survey as "cerrado" (scrub-savanna), "baixada" (wetlands), "cocais" (palm-forest zone), and "Imperatriz" (designating areas at the margins 6f the Tocantins river) (Brazil. MIC/STI 1982: 22-29). Since that survey, the expansion of babassu occurrence is also noticed in Maranhao's western frontier, in the so16 The babassu zone corresponds to lands of both North and Northeastern regions. It includes the north-west of Piaui, the central, eastern, south-western, and northern parts of Maranhao, the very northern portion of Tocantins (named "Bico-do-Papagaio"), and a small fraction of the south-east of Pari. This denomination was adopted by major academic authors focusing babassu related economy, biology, and society (May 1990; Anderson, May, and Balick 1991; Almeida 1995).

called pre-Amazon, associated with human occupation and slash-and-burn agriculture in primary forests. Specific features of these zones are summarized in Tables 1.1 and 1.2.
Figure 1.4: Maranhio state, Northeast region, and Brazilian Legal Amazon

45. 44* 3' 42A31
secondary~~~~ foet m Maaho ore2I SI18

Table 1.1: Babassu stands and fruit production according major ecological zones
area (km2) 50,000 18,500 30,000 20,000 118,500
states MA-PI MA MA MA-TO
area e 2vered 18,000 7,200 18,600 9,000 52,800
area effectively 44%MAcovered (%) 25%PI 39/9% 62% 45% 44%
average fruit
production (kgha) 1,250 1,300 2,200 2,500 1,800
annual average 2
production (ton) 2,250,000 950,000 4,100,000 2,250,000 9,550,000
Source: MCT/STI, 1982
Table 1.2: Specific features ofbabassu ecological zones in Maranhao
annual rainfall
annual rainfall) 700-1,300 1,500-1,800 1,500-1,800 1,600-2,100
dry season 6-8 months 5-6 months 6 months 5 months
babassu major bottom lands elevated sites across landscape across landscape
incidence (islands)
soils of babassu alluvial gleys alfisols alfisols
primary gallery forests low savanna; evergreen to evergreen to deciduous
vegetation sedges, grasses deciduous
shifting shifting cultivation; shifting shifting cultivation; cultivation; cattle buffalo and cattle cultivation; cattle cattle ranching (planted
major land use ranching / natural ranching (natural ranchin pastures)
pastures cotton, s. pasture) (planted
cane pastures)
Sources: Anderson, May, and Balick, 1991; May,1990

Babassu still is one of the most important products for Maranhdo's primary economy, justifying the inclusion of aspects of species biology and ecology in the description of the area's biophysical dimension. When studying the structure of 1 ha of primary forest composed of a total of 63 species in the babassu zone, A. B. Anderson estimated that babassu represented 20% of a total of 386 stems, in a relatively stable composition (Anderson, May and Balick 1991: 86, 189-190). When natural succession follows a disturbance, babassu's dominance in the landscape first declines in the initial 2-3 years, and then increases abruptly, as Anderson's data demonstrated: from a population of 366 trees composed of a total of 31 species in 1 ha of the above mentioned primary forest's succession, babassu's relative dominance increased to more than 80% (Anderson, May, and Balick 1991: 88-90, 191-192). Either in the case of recently cleared forests, or when the sites are converted to pastures, this dominance consists of a two stratum population, the lower one of seedlings and juvenile palms, and a superior layer of adult palms. Considering that babassu palm life-span reaches a maximum of 184 years, in the case of land uses with the predominance of pastures associated with babassu, or shortterm shifting cultivation, an apparent situation of equilibrium will endure only up to the senescence of the upper strata. Conversely, in the case of shifting cultivation, especially with fallow periods greater than eight years, conditions provided for constant recruitment of juvenile palms and the consequent long-term renovation of the population will pertain (Anderson, May, and Balick 1991: 74).

Mearim and Grajadi valleys: rivers, roads, and people
The mid course of the Mearim and Grajafi rivers belongs to the "cocais" ecological zone of the babassu area. Primary forests can still be observed only at the western boundaries of the Grajaui valley, in the locally termed "regido da mata ", the forested region. According to the classification of the Brazilian Institute of Geography and Statistics (IBGE), lands in the mid-course of these two rivers are included in the Pindar6 and Mearim micro-regions, illustrated by Figure 1.6. These two rivers, running in the south-north direction, constituted important commercial channels and ways to access the coast up to the 1950s, when a road connecting Bel6m with the Northeast region (BR-316) was completed, consolidating the integration of most of the area with the state capital, the Northeast, and with the national market. Also, road construction shifted the major focus of urban development and demographic growth in the region, from the area of influence of Pedreiras to the area of influence of Bacabal, the two major cities and urban markets, with a 1991 urban population of around 65,000 (Bacabal), and Pedreiras (40,000).
Introducing the historical perspective to be developed in Chapter 2, Figure 1.6 also provides a visualization of the process of peasant occupation along the Mearim and Grajai valleys. Generally stated, is possible to envision three major phases in the spontaneous peasant colonization of the Mearim and Grajafi valleys. The first one is the oldest occupation, dating from the second half of the 196 century, and associated with the end of the slavery-based economy. This situation predominated in the northern and eastern portions of the Mearim micro-region, and resulted in today's predominant presence of black people in towns such as Bacabal, Pedreiras, and Sio Luis Gonzaga do Maranhilo.

Mearim river
Sao Jos Baca BR-316
Alto- M to Aleare
\Grajau legre ~ MA-122
N + -N -Estrela i
Figure 1.6: Mearim and Pindar6 micrq-regions in Maranhio, and location of the
five peasant villages included in this study.
The second situation, intermediate in occupation, and geographically centralized, is associated with the expansion of Maranhio's agricultural frontier due to the arrival of northeastern immigrants since the 1870s, but mostly in the 1940s and 1950s. This situation includes lands between the two rivers. However it does not reach the western and southern portions of the Grajai valley. Those, constitute the third situation, the "regido da mata" (forested region), effectively occupied since the late 1960s. It is in the south and western parts of the counties (municipios) of Altamira do Maranhlo, Esperantin6polis, Paulo Ramos, Lago da Pedra, and Vitorino Freire, that remnants of the tropical forest still cover considerable extensions of land, a situation no longer seen in the other two zones.

Bacabal and Pedreiras: the city of Pedreiras attained legal status in 1889, whereas Bacabal was officially created only in 1920. Both are located at the margins of the Mearim river, the latter being about 65 km south of the former. The oldest town of the valley, created in 1854, is Sio Luis Gonzaga, previously named Ipixuna, and half-way along the river between those two cities. Since the 1870s, even before its existence as a city, Pedreiras received immigrants from the Northeast occupying lands for agriculture (Lago 1976: 8). Associated with cotton, rice, and babassu production, this immigration increased after the 1910s, and provoked a "demographic jump" in the 1950s, when peasants of the "expansion front" consolidated the Mearim valley occupation. After road construction and expansion of the agricultural frontier, the annual indices of demographic growth of Pedreiras and Bacabal averaged respectively 6.98%, and 5.07%, doubling figures of the previous decade. If restricted to urban areas, these indices attest to Bacabal's supremacy as an urban center: whereas Pedreiras' annual index of urban demographic growth decreased from 8.10% in the 1940/1950 period to 3.27% in the following decade, Bacabal's increased from 1.6% to 12.1% in the same period! (Musumeci 1988: 230-232).
Bacabal's urban perimeter is crossed by highway BR-316, which connects the city to Belim (650 km), and Teresina (280 km). Sao Luis, the coastal capital of the state, is 250 km north of Bacabal, via highway BR-135. Pedreiras is linked to major roadways, to Bacabal and other urban centers by state roads MA-122 and MA-245. Table 1.3 illustrates demographic patterns of this region, through 1991 data from 13 of the 24 municoios of the Pindar6 and Mearim micro-regions. Figure 1.6 shows these municipios, including those containing peasant communities focused in this study, and those adjacent to them.

Table 1.3: Area, population, and demographic density of counties in the Mearim and Pindar6 micro-regions, Maranho.
(municipio) (KM2) (inhab.) (inhab.) % (inhab.) % (Inh I (Inh I
km2) km2)
Altamira do MA 456.2 20,003 2,535 12.7 17,468 87.3 43.9 38.3 Bacabal 1,744.0 98,793 64,783 65.6 34,010 34.4 56.7 19.5
Esperantin6polis 1,383.5 30,567 6,789 22.2 23,778 77.8 22.1 17.2 Igarap6 Grande 643.5 14,810 4,781 32.3 10,029 67.7 23.0 15.6 Lago da Pedra 2,419.4 46,877 17,562 37.5 29,315 62.5 19.4 12.1 Lago do Junco 552.3 19,276 2,841 14.7 16,435 85.3 34.9 29.8
Lima Campos 397.2 12,404 4,943 39.9 7,461 60.2 31.2 18.8
O.D'Agua Cunh&s 577.4 16,249 7,125 43.9 9,124 56.2 28.1 15.8 Paulo Ramos 2,508.9 26,978 5,660 21.0 21,318 79.0 10.8 8.5
Pedreiras 873.2 50,603 39,694 78.4 10,909 21.6 58.0 12.5
Poq o de Pedras 526.0 24,481 7,311 29.9 17,170 70.1 46.5 32.6 S. Luis Gonzaga 1,087.5 26,085 5,266 20.2 20,819 79.8 24.0 19.1 Vitorino Freire 718.1 30,885 12,187 39.5 18,698 60.5 43.0 26.0 Total 13,887.2 418,011 181,477 43.4 236,534 56.6 30.1 17.0
MaranhAo 333,565 4,930,253 1,972,421 40.0 2,957,832 60.0 14.8 8.9
% of State total 4.16 8.48 9.20 8.00
Source: IBGE, 1992.
A glimpse at concentration of land ownership
Figures 1.7 to 1.9 illustrate land ownership and predominant patterns of land use in
these 13 counties, according to 1985 IBGE figures (the latest available). Tables with full
data are in Appendix 2. As in most of the state of Maranhio, there is a large disproportion
between number of rural holdings (estabelecimentos) and their area, according to the
producer category. Figure 1.7 shows that in 1985 less than one fourth of the holdings
(estabelecimentos) were legally owned by a proprietor (landowner), representing almost
95% of the total area. In contrast, renters and squatters cropped three fourths of the

holdings, which however consisted of only 5% of the total area Concentration of land
ownership is more evident through Figure 1 8 It shows that in these 13 counties, in 1985
more than three fourths of land holdings had less than five hectares, representing only 4%
of the total area On the other hand, less than 1% of the farms had more than 500 hectares,
but occupied more than 40% of the area Figure 1 9 demonstrates the association between
concentration of land ownership and pasture conversion. The figure shows that annual
crops occur in about 94% of the holdings, and in less than 9% of the area. In contrast,
planted pastures occurred in one sixth of the holdings, in an area close to half of the total
Distribution by type of producer
1.400,000 25000O
1,000,000 i
percentage of occurrence sqatr(25.73%) g 800,000 1500
* 600,000 -10000
400sharco (14%) number f popet
renter (48.30%) j i 200,000
laondowner Weer Sharecropper squer type ofproducer
number of properties = tota area percentage of occurrence
squatter (2.079%)
sharecropper (0.74%) nber of pr es
renter (48.30%)
landowner (24.37%)
squatter (2.09%)
sharecropper (0.07%) ttlae
renter (2.93%)
landowner (94.92%) (ha)
Figure 1 7 1985 distribution of number and area of rniral holdings in 1 counties in
r'e Meaim andk Graja vaileys' icco rdIng oOuIilcers :ategory "Source IBGE

Distribution by classes of area
600000 2500
-4000n 5E
32000OfW 10,000 Figure 1.8: 1985
100.0woo concentration of land
0 1a 111. ae ownership in 13
cka, of me(ha) counties at the
-~ wo WWMM..w Mearim and Grajau
percentage of ocrrence valleys, according to
00 haMA2%)size of the holding. 00 5Wha lS.MQ) (Source: IB3GE. 1985
1-10 ha (40.2M) Prowerie Censo Agropecuario)
-41 ha (3121%
3. 50 ha (40.85%)
100 600o ha (23.0%) toti area
10-400 he(21.14%) (a
1-40 ha (4016%)
I1 ha (0.79%) 2>
Distribution by land use classes
GK0.00 l40=00
3W4 O .00,OI
-10.000 -JT 10
El .fl Figure 19: number of
PWm ann. fasw n. past pi.pwit natfor. I hodnsan oa
cls"of land use hlig n oa
-nmM*. o Pr.,~t =] to w". area for the 1985
distribution of land
afo PV percentage of occuIrlnce use cl asses in 13
pl~pL (1XM41counties at the fanw P%) umb of prp Mearim and Grajau
am. t&TM)valleys, according to Per. KIM0% the predominant
fo~fr. (A4Veconomic activity. n. posL(MA1%) towa (Source: IBGE. 1985
MM IO(1RS) -Censo Agropecuario)

2.1 Introduction
The purpose of this chapter is a critical review of the literature in the field of
peasant economics and peasant social organization, and of the studies about the dynamic of the political economy associated with the formation of a free peasantry in the Brazilian Eastern Amazon, particularly in Maranhio. The first section provides a retrospective of the major theories of peasant economic behavior and debates concerning factors affecting peasant allocation of resources. The second section examines the impact of impediments in access to land and other resources at the level of the social group. It addresses processes of peasant collective action leading to rural conflicts, and the institutionalization of new types of peasant social movements in recent years. The final section focuses on a historical perspective of the political economy in Maranhio, through the major facts associated with the existence of a peasantry: its formation following slavery, the intensification of Northeastern migration upon changes in the plantation economy, their expropriation through land commoditization, and the processes of reaction of contemporary peasants.

2.3 Revisitingt Economic Strategies of Peasant Households and Communities
Definitions of peasants as a social category, and interpretations of peasant
economics have been formulated since the end of the 19th century, and still are the motive of scholarly debate, having strong policy implications in developing countries. Anthropology, political science, sociology, and economics are the disciplines converging and contributing to the generation of an extensive literature of peasant studies. This literature contributes to the understanding of themes of contemporary relevance such as environmental conservation,. sustainable agriculture, rural-urban integration, and economic globalization.
A recent work by Michael Kearney (1996), focusing on the anthropological
perspective but not limited to it, is one of the best references for a substantive review of a number of these contributions, and how the "intellectual warfare!' evolved among them. Kearney points out two major dimensions in the theoretical and political debate of peasant societies. These are of the left-wing vis-a-vis right-wing disposition, and the modem versus romantic dimension. Whereas the left-wing debate incurs between the proletarianism (modem) and peasantism (romantic) schools of thought, debates within the right-wing are expressed by the confrontations among modernization theorists and those that Kearney calls "romantic populists" (Chayanov, Redfield, James Scott as the most significant ones).

Asserting that higher stakes are present in those debates that cross both of the
dimensions1, Kearney also recognizes transformations in this framework resulting from the self-deconstruction of the left-right confrontation and from the evolving development of articulation theory, the sustainability discourse, and post-development images.
Prior to a more comprehensive discussion of Kearney's reconceptualization,
moving away from dualistic thoughts about peasant societies, this section starts with an overview of the major theories in peasant studies, and more specifically, in peasant economics, focusing on the investigation of decision making among economic alternatives. It departs from definitions of peasant economy provided by the early works of Wolf (1955, 1957, 1959, 1966, 1969), who described peasant survival strategies as mechanisms for sharing resources in response to selective pressures. One of Wolf s major contributions to peasant economics is the notion of the provision of multiple funds (replacement, ceremonial, rent, and then, profit) guiding peasant decision-making, and the existence of outside forces setting "asymmetrical ratios of exchange" for the food and other goods produced by peasants vis-a-vis items (and cash) they obtain from the market. According to Wolt "the appearance of the state marks the threshold of transition between food cultivator in general and peasants.... Only when the cultivator becomes subject to the demands and sanctions of power-holders outside his social stratum that we can speak of peasantry" (Wolf 1966: 6-11). Based on Wolf's perspective, Stavenhagen describes peasant economics with the following productionist statement:
1 Keamey (1996: 108-109) mentions as example of intense intellectual warfare the debate between classical Marxists (proletarianists) and Chayanovians.

"the peasant economy can be defined quite simply as aform ofproduction (and
associated activities) in which the producer and his family till the land
themselves, generally utilizing their own means ofproduction (tools and
instruments) with the objective of directly satisfying their basic need, although
for a number of reasons they may find themselves required to sell part of their
produce on the market in order to obtain goody which they do not produce. The
peasant economy occurs on small production units, non wage labor
predominates, possibilities for accumulation are limited or absent, and the
principal purpose of economic activity is not to obtain or maximize profits but to
guarantee subsistence" (Stavenhagen 1978: 31) Allocation of resources:
Sahlins introduced the concept of the "domestic mode of production" (DMP) to characterize how households in primitive societies support themselves. DMP's basic elements are the sexual division of labor, the relationship between human beings and tools, and finite production objectives. For Sablins, household relations are the dominant relations of production in primitive societies: "the normal activities of an adult man, taken in conjunction with the normal activities of an adult woman, practically exhaust the customary works of society" (Sahlins 1972: 79). Human labor is what drives the domestic mode of production, and the tools are subordinated to the human being, contrary to capitalist relations, in which technology drives the relations of production and social relations. According to Sahlins, economies organized by domestic groups and characterized by kinship relations run below their capacity, are "underproductive" (Sahlins 1972: 39- 41). Referring specifically to agricultural societies practicing slash-and-burn, shifting cultivation, he concludes that although sufficiency is the target of domestic production for use, and a considerable portion of the available labor force is unused, heterogeneity within primitive societies caused a substantial degree of "domestic economic

failure" (Sahlins 1972: 68-69). The significance of the DWi for peasant societies is observed by Sablins himself who sees peasant articulation with the market as favoring the manifestation of the under-use of labor-force, setting as a pattern for the group the attainment of a level of household effectiveness within the reach of the majority, although resulting inefficiencies among the most capable (Sahlins 1972: 89-91).
In sumn, the allocation of factors of production by peasant households results from a number of factors. First, it varies according to specific features of the domestic mode of production, the division of labor, and attributions within the household. Second, it is influenced by mechanisms existing at the community level, such as the intensity and forms of reciprocity and redistribution, as well as the existing degree of internal differentiation within the community. Third, it will be constrained or favored by endowments of the natural environment. Finally, aspects of the broader society, and the political economy resulting from state policies and market intervention play an essential role in peasant decision making2.
Understanding economic behavior in the so-called non-western, pre-industrial societies motivated during the 1960s and 1970s a great deal of academic controversy in economic anthropology. Debates between the substantivist and formalist schools were centered on the applicability of microeconomic theory in the analysis of the allocation of resources by those societies. To some extent this debate also reached the study of peasant economics, through different assessments given to rationality. 2Leslie Anderson, in her theory of peasant political ecology, considers four components of the peasant world: the individual, the village, the natural environment, and the national society. (Anderson, L.E. 1994: 7-8).

Rationality is considered by a number of authors as the major guideline for the allocation of resources by peasants. This hypothesis, however, assumed distinct explanations, starting with Schultz's theory of the profit maximizing peasant,' which emphasized that peasants operate in a way that no alternative allocation of inputs nor adjustments in outputs would give higher monetary or non-monetary net income to the household (Ellis 1988: 64). Schultz's assumption that each farmer, by maximizing utility, prevents any major inefficiency in the allocation of traditional factors (Schultz 1964: 39), is derived from neoclassical economics perfect competition model, which indicates that maximum profit occurs when marginal cost equals marginal revenue (what Shultz called marginal value-product equalization). Despite the recognition of the knowledge acquired by peasants by simply repeatedly executing their traditional tasks, the obtainability of profit maximization is criticized by Lipton and Adams, who enumerated as major constraints factors such as environmental instabilities, market and institutional imperfections (Lipton 1968: 346), and the lack of information (Adams 1986: 277). In addition, both condemned Shultz's misuse of evidence, through the adoption of selective data (Lipton 1968: 346; Adams 1986: 277).
In 1965, one year after the publication of Schultz's work, George Foster
introduced the concept of "the image of limited good" as dominant cognitive orientation4
' Schultz's theory is based on two ethnographic studies: Sol Tax's "Penny Capitalism", and W.D. Hooper's "The Economic Organization of a Village in North-Central India". From the latter, he quotes "each man comes close to doing the best he can with his knowledge and cultural background". (Schultz 1964: 45-46)
4' Foster defines cognitive orientation as the "unverbalized, implicit expression of a society's understandings of the "rules of the games" of living, imposed upon them by their social, natural,

of classic peasant societies. Based on his field work in Michoacin, Mexico, Foster suggests that "peasants view their social, economic, and natural universes their total environment as one in which all of the desired things in life ... exist infinite quantity and are always in short supply, as far as the peasant is concerned" (Foster 1967: 304). He sees a number of features in peasant behavior adversely influencing economic growth, but highly rational and indispensable for the maintenance of the peasant society in its classic form. According to Foster, unexpandable goods in a closed system would allow individual or household economic progress only at the expense of others, and consequently, "apparent relative improvement in someone 's position with respect to any good is viewed as a threat to the entire community' (Foster 1967: 305). Thus, changes in the status quo of the relationships are discouraged, through a preferred behavior that, as Sahlins' analysis of the DMP corroborates, will maximize peasant security, preserving his relative position in the traditional order of things. Advocating greater peasant participation in national development, Foster suggests that only creating economic opportunities (changing the economic rules of the game), the peasant would be "encouraged to abandon his traditional and increasingly unrealistic cognitive orientation for a new one that reflects the realities of the modern world" (Foster 1967: 304).
Another way to address how rationality is expressed in peasant decision-making is the debate between political scientists James Scott and Samuel Popkin, through their respective theories (Scott 1976; Popkin 1979). The debate, one of the most preeminent in the field of peasant studies, confronts Scott's social security mechanisms and peasant and supernatural universes. Foster, G. M. (1967:300)

communities' harmony, with Popkin's concept of the rational peasant guided by selfinterest, and cooperation in societies based on task-specific incentives and calculations. Whereas Popkin repeatedly stressed peasants' individual calculations in order to raise their standard of living, Scott depicts peasant rationality as shaped by economic, political, and social constraints. These books, published in the mid to late 1970s, linked peasant decision-making at the individual and household level, with processes occurring in their communities, and with the adoption of collective action. As Adams points out, whereas Scott emphasizes harmony between individual and collective interests, Popkin stresses the sources of conflict between them.
On the other hand, when examined at the microeconomic level, Scott's theory of peasant behavior confronts capitalist enterprises's economic rationality. According to him, peasants, threatened by a variety of social claims and impersonal forces, have as their central economic preoccupation to feed their households reliably, despite the minimal marginal return of the extended labor allocated.5 The strategy of avoiding falling below what he calls a "subsistence danger level" combines with the risk avoidance behavior pointed out by many economists studying peasant agriculture in developing countries (Roumasset et al. 1979; Binswanger and Sillers, 1983). Following this principle, cultivators would prefer to minimize the probability of a disaster rather than to maximize
s Scott lists rent, taxes, and debts as social claims. As impersonal forces, he mentions uncertainties of weather, soil quality, level of techniques, the risk of illness, and the availability of arable land. (Scott 1975: 505-507).
' As "subsistence danger level" Scott means the threshold below which the qualitative decline in subsistence security, family cohesion, and social status is massive and painful. (Scott 1975: 507).

their average return. In so doing, they will provoke a sub-optimal use of variable inputs, and outputs below the level expected when profit is maximized (Ellis 1988: 100).
Risk aversion behavior, as well as the theory of profit maximization, fails in
considering only the pursuit of a single goal as determining the allocation of resources by a peasant household (Ellis 1988: 102). A different interpretation of peasant economic practices is the analysis presented by A. V. Chayanov. His "Theory of Peasant Economy" based on the concept of the peasant farm as "a family labor farm in which the family as a result of its labor receives a single labor income and weights its efforts against the material results obtained' (Chayanov 1976: 41). The basic concepts of Chayanov's theory are self-exploitation of the peasant labor force7 and demographic differentiation'. According to this theory, given a socially accepted minimum standard of living, the achievement of a point of natural equilibrium in the allocation of resources occurs when the drudgery of marginal labor expenditures will be equal to the subjective evaluation of the marginal utility of the sum obtained by this labor. In a separate essay from his "The Theory of Peasant Economy", Chayanov identifies peasant economy as having singularities regarding productive and reproductive strategies constituting a mode of production distinct from the capitalist and the socialist.9 Chayanov (1976: 76) points out that the level of labor intensity is determined by internal and external factors. The internal factors are those related to pressures of family consumption demands. The external ones are those production conditions which determine the level of productivity. sChayanov (1976: 60) explains demographic differentiation as: "every family, depending on its age, is in its different phases of development a completely distinct labor machine, as regards labor force, intensity of demand, consumer-worker ratio, and the possibilities of applying the principles of complex cooperation"
9 "On the Theory of Non-Capitalist Economic System", included in pp 1-28 of the op. cited.

The supposed absence of a market for wage labor in 1920s Russia was the motive of Chayanov's analysis considering the peasant family as a unity in abstraction from external relations (the utilitarian individualism). Critiques of his theory, although recognizing the consistency of his postulates, and moreover its scientific advance, allude to the existence of a growing network of social relations between households, mediated through product and factor markets. As Heynig points out, Chayanov's theory, besides ignoring social differentiation, considers the peasant family farm as independent of the larger social formation (Heynig 1982: 129). Kearney points out that Chayanov had little or no concern for peasants as social and cultural beings, limiting his analysis only to production and consumption aspects of the peasant household. In addition, Chayanov's vision of the peasant household as a unity is contested by the fact that real family units are internally differentiated by age and sex (Harrison 1979: 89; 1976: 329-334). Harrison's critique of Chayanov's view of the peasant family opposes his analytical aggregation of family members into a single economic agent possessing a unified, inherent rationality and consciousness, encompassing a conceptual dissolution of internal social relations. The individual and the household in the reconceptualization of the peasant:
Deere's approach to household and class relations addresses both of the
shortcomings attributed to Chayanov's theory. First, she looks at the class processes that characterize economic activity. She departs from Resnick and Wolff's interpretation of Marxist class analysis'0 "by employing the concept of class relation to refer to a '0 Resnick and Wolff distinguish between fundamental and subsumed class processes. The former refers to performance and to extraction of surplus labor, whereas the latter consists of its distribution, through specific social functions. Subsumed classes, according to them, maintain

fundamental class process and the conditions of existence necessary to distinguish different class processes from one another (Deere 1990: 13). Deere attributes to peasants' participation in a number (multiple) of fundamental, or in both fundamental and subsumed class relations, the basic condition for the reproduction of the peasant household as a unit of production and reproduction. In other words, multiple income-generating activities help explain the persistence of the peasantry. Second, she introduces the concept of household relations to describe the economic, cultural, and political practices ruling the formation of households and the relationship between the unit of production and the unit of reproduction. As the multiple class positions of peasant households were often based on the gender division of labor, her analysis brings together the potential tensions in household and gender relations in order to understand their interaction with class relations and the potential for social change. Within her framework, household members may engage in feudal, petty, communal, and capitalist class relations at the same time (Deere 1990: 13-15) ". Rather than consider petty production as the primary relation of peasants (as Warman did), Deere asserts that specific circumstances and opportunities characterizing each household will dictate peasants' primary relation of production. That
themselves by means of shares of extracted surplus labor distributed by the fimidamental extracting classes. (Resnick and Wolff 1982: 3).
"Deere characterizes feudal class relations by the existence of a class of landlords and a class of direct producers vwho must pay rent in the form of labor services in order to obtain access to the means of production. In capitalist class relations, a class of dispossessed workers sell both their necessary and surplus labor for a wage, to the owners of the means of production. Communal class relations are characterized by collective possession of the means of production by direct producers, all of whom perform labor. Finally, the class relations embodied in petty production are characterized by the petty producer's ownership of the means of production and control over the labor process, allowing the individual appropriation of surplus labor.

is conditioned by the class relations in which household members participate and the outcome of the economic activities, which interact and are mediated by household and gender relations, and by extra-household relations among kin. Household and class relations vary according to the family life cycle (Chayanov's demographic differentiation), as well as in response to the broader political economy. Stressing interactions among household and class relations, she considers that
"a focus on peasant household reproduction from a micro point of view provides
a different understanding of how peasants may survive in the face of
impoverishment and proletarization, opening up space for the analysis of the
subjective forces of history, the actions of men and women in daily life as well as
the relations between them (Deere 1990: 326)
As was already pointed out, Kearney's working assumption of peasants'
reconceptualization advocates a "dismantling of spatial and temporal binary oppositions in the present historical moment, characterized as transnational, post-developmental, and global" (Kearney 1996: 43). Influenced by articulation theory assertions of complex interrelationships between capitalism and other modes of production, Kearney emphasizes the "dissolution of much of the oppositions between rural and urban, developed and underdeveloped and between peasant and non-peasanf' (Kearney 1996: 120). Contrasting dependency theory's emphasis on global regions and nations as primary units of analysis, he borrows from articulation theory the analytical focus on community, household, and individual levels. Working with the concept of social identities affected by internal differentiation resulting from migration, the penetration of agro-industry, and capitalist agriculture dominance, Kearney attributes to peasants rather the complex transforming entities who defy social bounds than Wolf's definition by productionist

criteria as "polybians": "creatures moving in and out of multiple niches, ... adapting their being to different modes of existence opportunistically occupying different hlife spaces"(Kearney 1996: 141). Hence, although mechanisms of the capitalist economy in several cases resulted in the disappearance of peasant societies, it is also true that peasant economy can coexist side by side with capitalism. Shanin, for instance, points out that a number of authors
"have documented for different environments the tendency of agribusiness to
withdraw from the process ofproduction in agriculture, focusing their profit
making activities on credit, supply of inputs, contracting and selling, while
leaving farming to the small holders and skimming them rather than replacing
them". (Shanin. 1976: 4)
The importance of Kearney's reconceptualization of the peasant extends also to the field of decision-making research, which he considers largely contained to the sphere of agricultural production and predicated by rationality and maximization of economic value. Kearney advocates a refocus of decision-making theory in order to recognize the totality of spaces occupied by the individual and elements of the identity of which he or she is composed: "ethnographic research on decision-making concerning agricultural production is inseparable from the ethnographies of the other socialfields in which the small part-time producer lives and that are often more important to biological and social reproduction than is agricultural production "(Kearney 1996: 159).
Definitively, Kearney refuses the adoption of productionist issues, decision-making models and rational actor theory when referring to his polybian-type of peasant (Kearney 1996: 148). His theory envisions these peasants as part of a dynamic process of transformation of value and power, shaping the political economy, "redefining agrarian

issues and establishing the local politics of human rights and ecology within international contexts, eventually favoring subaltern groups in the balance offorces with the nation-states" (Kearney 1996: 169).
Whereas Kearney expresses a great deal of human agency through the sequential engagement in various economic opportunities, his characterization does not sufficiently addresses the role of household in the establishment and maintenance of the polybian-type of social identity. My interpretation is that his theory fits suitably for situations in which peasants experience massive penetration of cultural and economic value characteristics of capitalist society, especially in transnational contexts. It has less application, however, for scenarios of peasant communities that, in spite of ongoing changes, have not yet undergone a complete revolution in the process of production. To these scenarios, Deere's theory seems to be of more analytical validation. Taken together, Deere's and Kearney's approaches substantially contribute to the understanding of the economic calculations and practices of the contemporary peasant.
2.3 Social Organization and Collective Action as Peasant Survival Strategies:
Socioeconomic transformation restricting access to the resource bases modify
prior arrangements and attributions not only at individual and household levels, as seen in the second section of Chapter 1. They provoke changes internally in the entire peasant social group. When the availability of resources for the productive process reaches a certain minimum, the status quo of the relationships within the community, and previous mechanisms regulating the access to labor and to common resources begin to deteriorate.

They deteriorate up to a point in which individual initiatives often result in conflicts with the general perception of the social group. As Almeida and Mour~o pointed out, although reciprocity and redistribution principles strengthen community cooperation and protect the group from external interferences, they are affected by limitations in material conditions, of which access to land is the primary one (Almeida and Mourao 1976: 19). Thus, the reduction of resources increases individualism (Ostrom's "free-riders"), and Hardin's "tragedy-of-commons"' (Hardin 1968: 1243-1248) is manifested, deteriorating patterns of social interaction ". The peasant community becomes fragile, and parallel to the adoption of distinct economic practices, households are stimulated to rearrange their broader social conduct.
One of these rearrangements consists of internal mechanisms of regulation through incentives for out-migration, reducing pressures upon resources. That is represented by the departure of some members of the household, or of the entire social unit. The first case usually involves youths' seasonal leaves to work in activities such as mining, migration to urban centers or to poles of capitalist expansion in need of labor in rural areas. Migration of the entire household, however, is more often associated with the existence of an "expansion front". This alternative, therefore, depends upon the availability of "free" lands in the frontier, where peasants reacquire an ideal stock of resources for the practice of their system of cultivation. It will last, however, up to the point in which the capitalist mode production subsequently reach the newly occupied area and the peasantry.
12 According to Hardin, "each man is locked into a system that compels him to increase his herd without limit in a world that is limited. [and] ... freedom to breed will bring ruin to all"

Proletarization of members of the peasant household, on the other hand, generate remittances that contribute to income composition and for the social reproduction of the remaining group. The adoption of out-migration, although partially dismantling community linkages, is thus a regulatory strategy, extending the conditions for the prevalence of petty class relations for peasant subsistence.
The lack of the "expansion-front" alternative, or the return of those who outmigrated and assimilated different perceptions of reality in urban centers or regions undergoing agricultural modernization, are some of the factors favoring the intensification of another process, peasant collective action, as a mechanism adopted by communities to deal with the subordination derived from concentration of land ownership. Eric Wolf sees peasant movements of protest centered upon a myth of a more egalitarian social order, sufficient however only to provide a common vision, to unite peasants, and not to organize them (Wolf 1966: 106-108). Peasant reaction to exploitation, in fact, assumes heterogeneous forms, from rebellions and revolutions to nonviolent protest and everyday resistance3. In proposing a theory of rural class conflict, Paige attributes the prevalence of major forms of rural social movements to particular combinations of economic and political behaviors of those who cultivate and those who do not cultivate the land. He asserts that agrarian revolts, described as "short, intense movements aimed at seizing land but lacking long-run political objectives" characterizes the situation combining a landed upper class and a landed starved peasantry. According to Paige, a reformist process
'3A bibliography on peasant revolution includes the already mentioned works of Scott and Popkin; Wolf 1969; Paige 1975; Migdal 1974. Works on nonviolent peasant protest are Sharpe 1977; Zamosc 1986. For everyday resistance, see Scott 1985; Hyden 1980, and Gaventa 1980.

prevails in two cases, both of them characterized by a rural upper class having commercial and/or industrial interests rather than depending on land revenues. The first instance occurs in small-holding systems, where conflict focuses. on the control over the commodity market. The second is the reform labor movement, in which political conflict targets working conditions and salaries. Finally, Paige associates agrarian revolutions to "the conflict between a radical, well organized class conscious work-force and a landed upper class unable and unwilling to grant any political or economic concession" (Paige 1975: 39-60). Incorporating a vision of interdependence among individual, village, ecosystem, and society, L. Anderson, attributes peasants' survival to their flexible range of political actions and consequent readjustments to an ever-changing world (Anderson, L.E. 1994: 170). According to her, as the wider political world becomes part of peasants' perspective, collective action incorporates the ecological interdependence and reciprocity that characterizes the relationship between peasants and their broad environment4.
The enunciation of Anderson's political ecology emphasizes the incorporation of a distinct strategy, the establishment of new forms of peasant organizations. Their origin is flexible, either resulting from a process of revolt, or emerging from the sequential stages of spontaneous disruptive actions. In the context of fewer closed corporate peasant communities'", the establishment of dynamic, non-perennial forms of socioeconomic organizations has been increasingly adopted to reduce peasants' subordination and overexploitation resulting from the subsumption of their labor and products. The creation 14 For an analysis of the political ecology of the peasant, see Anderson, L. E. (1994: 14-18) 15 For the characterization of closed, corporate communities, see Wolf, 1966.

of these institutions, in fact, follows Wolf's assertion about peasants' social organization, which emphasizes a" ... strong bias towards autonomy, and an equally strong tendency to form coalitions on a more or less unstable basis for short-term endc'(Wolf 1966: 91). These new social movements articulate interests that are other than class-based. They consist of institutions targeting the reproduction of peasant social and cultural identity, through the adoption of economically-focused coalitions, initially within the community, and subsequently with external groups or entities, in which peasants encounter appropriate allies. In a certain sense, new social movements within the peasantry act as a contemporary substitute for the patron/client relationship. When constrained by a reduced resource base and facing transformations in the dominant mode of production, these organizations deal with the narrow limits represented by the maintenance of peasant identity simultaneous to the everyday provision of material needs for the household.
Given processes of socioeconomic transformation occurring in their specific area of influence, peasants' contemporary organizations attempt to channel their members' economic logic and capacity to survive in adverse conditions, to a more active market integration, and in the formulation and claim for adequate public policies. This process represents an internally assimilated response to external pressures, and allocates strategies and instruments of collective action toward a new set of determinants. They are essentially operational and localized in their interests, but are characterized by a strong capacity of regional aggregation and by the envisionment of broader aspects affecting the existence of the interest groups represented (Almeida 1994: 522-523). Their unstable and flexible

condition reduces restrictions common to bureaucratic corporative institutions, allowing the expression of values characteristics of peasant societies in the contemporary world.
Socioeconomic transformations push the peasantry to organized strategies, and the creation of institutions such as associations or cooperatives can be a tool for the social reproduction of the peasantry and improvement in their living conditions within capitalism. In effect, peasants, due to the need to incorporate longer term production planning, influenced by a broader vision achieved through coalitions, will have the awareness and strategies to avoid Meillassoux's remark that,
" ... the shiftfrom production for self-sustenance and self-perpetuation to
production for an external market must necessarily brings a radical
transformation, if not the social destruction of the communities, as indeed we
witness the process nowadays. The attempts to superimpose productive and
mercantile institutions, such as the cooperative, on the lineage or village
community, is bound to fail or to transform the society into an eventual class
system". (Meillassoux 1980: 10)
On the contrary, the activity of peasants associations, cooperatives, and similar
organizations will enhance their ability to attract allies in other sectors of the civil society, and through vertical coalitions, counterbalance the exploitation imposed by the hegemonic class. By acting over the three major constraints for peasant social reproduction, such organizations attempt the reconstitution of funds that have been often suppressed in most contemporary peasant societies. First, their activity targets the recovery and/or effective maintenance of tenure rights. Second, they emphasize the definition of patterns for land use and conservation of resources given ever increasing pressures. Third, they attempt to raise peasant awareness of the exchange value of their labor and of their products. In this respect, the formalization of interest groups for commercial purposes assumes a buffering

function, attenuating negative effects of market exchange, by absorbing them at the institutional level. In doing so, they revert these commercial transactions to forms of integration reflecting mechanical solidarity among households. On this, I agree with Popkin when he states that" ... throughout the wor peasants have fought for access to market not as a last gasp when all else has failed, but when they were secure enough to want to raise their economic level and redefine cultural standards" (Popkin 1979:80).
The extension of interactions brought up by contemporary forms of peasant
organizations and the coalitions they form also change the cultural distance between the peasant and the outside world. Acting as brokers in bringing an extended universe to the social group, leaders and representatives of these organizations, as well as members of allied groups in constant connection with the peasant community, have the extreme responsibility of counterbalancing the often early negative impacts of the industrial society.
Nevertheless, important internal restrictions also affect the activity of these peasant organizations. Their positioning should observe the social context. Economic initiatives must be shaped by social experiences of the group, by "everyday experiences peasants have in their class situations and in their various social roles in institutions such as the church, the family, and external kinship groups" (Sharpe 1977: 121). In effect, in order to achieve their goals, these new institutions have to consider the adoption and/or gradual modification of practices present in traditional social relations, avoiding sharp discontinuities. Particularly important are tensions between the household and the social group, or community. Far from Banfield's "immoral familism" (Banfield 110-113), there are certain attributions traditionally pertaining to the former, and that cannot be simply

substituted through the initiative of a broader organization. For a certain peasant group, the undertaking of a collective agricultural field, for instance, can be of extreme importance during certain phases of the struggle for land. However, once this phase is terminated, it may not be accepted as a normal pattern. Existing differences among households in a peasant community also have to be observed during the implementation of economic initiatives such as settlement schemes and development projects. Political negotiation and deadlines for implementation, however, have been forcing the execution of such actions disregarding these aspects and jeopardizing not only the long run feasibility of these initiatives, but also, and primarily, the internal organization of the group. This is exactly one of the most significant challenges for the contemporary peasantry in Maranhio. The next section will thus introduce the literature referring specifically to that peasantry.
2.4 Formation and Consolidation of a "Shifting-Cultivator Peasantry"
Society in the babassu zone, and in Maranhao, is predominantly rural. According to the 1991 Brazilian demographic census, sixty percent of the state population was classified as "rural", the lowest urbanization index in the country. The origin of this society resides in descendants of native Indian people6, descendants of former Afican
6According to Van Damme (1990), the indigenous population that inhabited the coast of the state upon the arrival of the Europeans was estimated at I million people. Forced into slave labor up to 1754, the indigenous groups progressively moved to the interior, usually along the rivers. For more details about native Indian groups in Maranhio, see Gayoso 1970. Compendio Historico-Politico dos Principios da Lavoura no Maranhio.

slaves7, and Portuguese colonizers. Its present configuration resulted from the arrival of migrants coming since the end of the 19* century from the arid Northeast, mainly from the states of Ceari and Piaui". Net migration to Maranhao, and to the babassu zone, ceased in the 1970s. According to Sawyer, average indices of both annual population growth, and annual population natural growth have decreased in the 1980s. The former dropped from 2.9 in the 1970-1980 period, to 1.9 in the following decade, while the latter dropped from
3.3 to 2.5 in the same period (Sawyer 1993).
The dynamic resulting in the configuration of Maranhio's rural society reflects the processes associated with the formation of a peasantry in the Eastern Amazon. Consolidating the occupation of the area, agro-extractive, "shifting-cultivator peasant", relying essentially on human labor and nature (no tractors, no animal power in agricultural operations, no fertilizers and only limited use of pesticides), since the 1950s placed Maranhao as one of the top Brazilian producers of rice and manioc"9. In addition, the largely female-and-child-based extraction of products derived from the babassu palm, and the nature of the palm forest's secondary growth itselfW suitably integrates extractive
1 The prohibition of indigenous slavery in the 1750s, motivated the trade of African slaves, in exchange of primary goods acquired by the English. For details about slavery in Maranhao, see Viveiros 1992. "Historia do Com~rcio no Maranhao 1612-1895". 18 A succession of severe droughts since the beginning of the century, and the fragmentation of properties are considered the main reasons for this movement westward. For more details see Andrade 1984. Ensaios sobre a Realidade Maranhense. According to the IBGE, the 1992 area destined in Maranhao to rice production (740,000 ha) and manioc production (240,000 ha) corresponds respectively to 17% and 13% of the total area destined to these two crops in Brazil. The 650,00 tons of rice and 2 million tons of manioc harvested in Maranhao, corresponded to 6.5% and 9% of the national production. (IBGE 1994 Anuirio Estatistico do Brasil).

activity with the mostly male-based agricultural practices. And although babassu kernels are no longer the most important product of Maranh~o's primary economy as they were up to the 1970s, secondary forests of babassu decisively contribute to the livelihood of the larger agro-extractive segment in Brazil'. Maranhio produces more than seventy five percent of the country's babassu kernels, through an activity that occupies more than 300,000 families. As Shiraishi's recent research attests, an increasing portion of babassu extraction is executed by peasants expelled from their land, currently living in the periphery ("pontas de rud') of urban areas (Shiraishi Neto 1997).
Paradoxically, since the late 1960s, trends of the political economy and regional development resulted in the negation of these peasants' tenure rights. According to IBGE's figures, there were 425,000 landless families (more than 2.1 million people) classified either as squatters, renters, or sharecroppers in 1985. Land ownership has in fact been an issue of controversy and dispute in this region. As Valverde pointed out since 1957: ... no other Brazilian region, except ,Mo Paulo and Parant 's plateau faced so intense and abusive a process of illegal land appropriation (grilagem)" (Valverde 1957: 17, my translation). This section presents an historical perspective of processes resulting in the formation and consolidation of the "agro-extractive, shifting cultivator peasantry" in Maranblo, and the dynamic of the political economy associated with its existence. Although assuming the risk of placing boundaries where they do not fit, I will present
20 According to the IBGE, in 1986 there were 202,000 families (188,000 of those in Maranhio) involved in the extraction of babassu. In the same year, population estimations for rubber and Brazil-nut were respectively 95,000 and 40,000 families. (IBGE. 1986. Censo da Produq~o Agropecuiria).

these processes according to a simplified classification consisting of four major periods: the consolidation of a peasantry following the end of slavery, the intensification of Northeastern migration due to the disintegration of its plantation economy, the process of expropriation due to land incorporation to the market, and the recent processes of reaction of contemporary peasants.
From slavery to peasantry: land occupation and common resource use:
The creation of the General Commercial Company of Gr~o Pari and Maranhao (GCCGM), in 1756, is considered the threshold for the economic formation of the state (Tribuzi 1981: 13). Until Brazilian Independence in 1822, Portuguese colonizers were granted large extensions of land (sesmarias), measuring 18 x 6 km. The exploitation of these areas, however, became effective only upon intensification of the African slave trade through the GCCGM monopoly. Coinciding with the aftermath of the Independence of the United States, this period opened the European cotton market for Brazilian production, and Maranhao's cotton plantations received a significant impulse, although until 1815 all profits were channeled to Portugal. In 1822, more than half of the 175,000 population of the state wa composed of African slaves, working mainly on cotton plantations in the Itapecuru valley, and on sugar-cane plantations in the Pindar6 valley (Tribuzi 1981:14-16; Andrade 1980:184). The prohibition of slave trade by the British crown in 1850 initiates a process of decline of the plantation economy in Maranhao, culminating thirty eight years later with the abolition of slavery.
The decline of a system based on the capturing of the labor-force through slavery occurs parallel to the formation of a peasantry adopting a "subsistence" economy based on

the domestic unit of production. Emphasizing the establishment of these units outside the plantation, Andrade describes the period as one in which
"... the Negroes left the fazendas and sugar mills. They hid away in the babassu
groves andforests, easily withdrawing their subsistence. There was abundant unoccupied government land, from which the Indian was already removed but which the cohnists had not yet occupiedfor agriculture, and they were able to
live in a closed, non monetary subsistence economy."(Andrade 1980:184)
However, even though unoccupied lands were incorporated, this was not the
predominant feature associated with the formation of the "black-peasantry" in Maranhio. As a matter of fact, two major trends were observed. On the one hand, ex-slaves remained within the former plantation, especially in areas of older occupation at the Itapecuru valley (Almeida and Mourio 1976:10). Landowners, adopting extensive livestock grazing on natural pastures, captured this labor-force through mechanisms of payment of rent for land use, and commercial exploitation. On the other hand, in areas of more recent occupation, the disintegration of the plantation farm led to its transformation into a constellation of autonomous domestic units of production, based on inter-household cooperation and the common use of resources (Almeida 1996: 19). Tenure rights were transferred from failed landowners to families of ex-slaves, either through donation, sale, or by mediation of municipal governments, as Esterci mentions for the Pedreiras case (Esterci 1977: 73). In addition, Almeida includes in this latter category, so-called "lands of the black" (terras de preto), state concessions of land to runaway slaves as a payment for their warrior services, mainly in the Balaiada war (1838-1841), well before abolition (Almeida 1989: 174-175).
Indeed, instead of the natural economy targeting subsistence as Andrade's description signalizes, what prevails in both situations following the collapse of the

plantation system in Maranhio, as Musumeci emphasizes, is capitalist accumulation through the commercial exploitation of rural cultivators. Cotton, and later rice and babassu, heavily integrated into market mechanisms, remained as the major products of this newly formed peasantry (Musumeci 1988: 173-174). Contrary to a natural economy, there was in fact
"an intention to be integrated into the market and overrule the agrarian situation stipulated by the 1850 Law of the Land [regulating the measurement, demarcation,
and sale of the designated Imperial lands], as large landowners, upon losing their
power and being unable to feed and sustain their slaves, proposed agreements including freedom (alforria) and land." (Almeida 1996: 18-19, my translation)
Market integration through the commoditization of their production seemed to be incorporated into the economic horizon of that society. But if we refer to peasant segments entitled to access to land, and "terra de preto" was only one of a multiplicity of situations, the same logic did not apply regarding their perception of land. Opposed to individual appropriation of territories and land commoditization was the notion of common resource use, defined by Almneida as a category where
"resource control is not freely and individually executed by the specific domestic
unit ofproduction or by one of its members. Resource control, instead, occurs
through rules that go beyond the legal framework, being adopted by consemsus
throughout inter-household social relatiorLns.... Territoriality functions as an
element of identification, defense, and empowerment. Solidarity liaisons feed a set of regulations over a common physical basis, essential and inalienable, in spite of
eventual successional dispositions" (Almneida, 1989: 163, my translation)
Common resource use represented a strategy for their social reproduction in the face of capitalist accumulation. Common resource use usually does not include common appropriation of the results of the labor (agricultural fields are individually appropriated), even though exchange labor and other forms of cooperation are observed. Also, common

resource use entails differential notions of entitlements among the social group. Maintained in a number of present situations, although the object of threats since its pristine appearance", common land and resource use are viewed as an impediment for the incorporation into the market of these resources, and a consequent defense against the dispossession of the peasant unit. In a distinct context, it was indeed associated with the other primary process in the formation of Maranhao's peasantry, the northeastern "expansion front".
Northeasterners in the moving frontier: market integration or commercial exploitation'?
Gradual and non revolutionary transformation in Northeastern plantations, from a system that immobilizes the labor-force through direct hierarchical relations, to authoritarian capitalism, is indicated by Velho as the key factor for the appearance of a marginal peasantry in Northeast Brazil (Velho 1972: 120, 187). According to him, "secular decadence of Northeastern sugar plantations ... made it more convenient for Northeastern planters to resort to forms of immobilization of the working force in which labor itself took care of its subsistence ... both outside and inside plantation physical domains" (Velho 1972: 118, 123). Parallel to the capturing of the necessary labor-force for plantation needs, Velho sees the formation of a peasantry marginal to the plantation, acting as a reservoir of labor-force for seasonal operations. Such transformations, however, as Velho himself attests, occurred with greater intensity since the 1950s, with the introduction of a more typically capitalist system of free wage labor (Velho 1972:
21 Alneida (1989: 171) mentions the annihilation of the Canudos and Contestado movements, and the repression against units absorbing runaway slaves (quilombos).

196). Thus, if transformations on the plantation economy are coherent with the appearance of a peasantry in the Northeastern sugar-belt, other explanations should be added for the Northeastern influence in the constitution of a peasantry in Maranhao.
In fact, since the third quarter of the 19' century, the already debilitated plantation economy, affected by the consequences of periodic droughts in the arid interior regions, was not able to maintain immobilized their reserves of labor, and families began to move either to cities in the Northeast, to cities and coffee plantations in the Center-South, or to the Amazon. An estimated half million Northeasterners arrived in the Amazon during the rubber boom, until 1912, when the region fell into a new period of great stagnation. Thus, although observed since the end of the 191 century (Lago 1976: 8), the movement of surplus Northeastern population towards Maranhao was intensified in the 1920s, following the end of the rubber extraction alternative. As Veiho states, this migration "at first tended to cross already settled although decadent areas close to the coast. But with time it spread towards the unsettledjorest regions of the river valleys further West." (Veiho 1972: 213).
The "Northeastern front" introduces distinct cultural patterns in Maranhao' s
Indian and African heritage. Patterns that did not always favor integration, and in certain circumstances, as will be described in the next chapter, are still generating internal conflicts. Although deeper interpretation of such a cultural confrontation goes beyond the objectives of this thesis, the following narrative made in 1957 by geographer Orlando Valverde gives an idea of the usual representations of these patterns, attributing a positive image to the Northeasterner in the face of Maranh~o's "native" cultural habits:

"The cearense works hard. He has initiative, has disposition to open doors for a
better future to his family, whatever the cost. It is a patriarchy, but everybody
works: men, women, and children... Maranhdo 's caboclo is a loser. Economic
dependency, the slavery legacy, and indolence defeated him. His socialfinction is almost limited to mere reproduction of the species. Women, however, start to
work early. At 6 or 7 years old they begin to crack babassu to buy clothes, but at
16 or 17, most of them are prostitutes".(Valverde 1957: 32, my translation)
Northeasterners in fact, through the so-called "spontaneous" migration -- that Almeida appropriately portrays as an actual process of forced migration (Almeida 1974: 46) -- initially arrived at the already occupied and pauperized lands of the Itapecuru valley, characterized by large holdings and the extraction of rent from peasants. Such an influx of cultivators resulted in dismantling the previous equilibrium, increasing competition over limited land. Raising prices of the rent, it stimulated the joining of a peasantry previously established in those areas of older occupation (for instance, segments of the "black peasantry" mentioned before), to the westward progression of Northeasterner migrants (Almeida and Mourio 1976: 11). For these, as Valverde emphasizes, the search for social and economic independency was the major motivation to immigrate. A reinsertion in a system that would immobilize them was not at all desirable (Valverde 1957: 17).
The opportunity to occupy unsettled, free lands was viewed as favorable by those peasants, given the exacerbation of restrictions imposed by rental agreements. Not usually associated with the "expansion front" into the Mearim, Graja6 and Pindar6 valleys, these arrangements later became part of the subsumption of peasant labor even in these areas, after land incorporation into the market, began in the late 1960s. Still today, this mechanism is largely used in Maranhio.

The usual agreement between renter and landowner stipulates payments using part of the harvest, often rice. Frequently through informal contracts, the concession of land during the agricultural season is usually the only contribution of the landowner. The estimated obligation of such contracts consisted in a payment of an average of two "alqueires per linha', (180 kg of rice per hectare), corresponding to 5 to 20 percent of the production, according to the yield 2. Although severe restrictions for the cultivation of perennial crops (except in a small area surrounding the household), the inter-cropped production of maize, beans, and manioc is usually free of rent, and rent was flexibly charged in years of bad weather or when unexpected problems affected a household. Landowners wealth to a large extent depended upon the relations of production with peasants. In many cases, living in the village, landowners appropriated portions of their funds through the land rent, by requiring unpaid labor obligations for their own benefit, and in some circumstances even through processes of transformation and commercialization of agricultural and extractive production.
Increased rental and other peasant obligations in areas of older colonization contributed to the occupation, by the "expansion front", of free lands west of the 440 meridian, entering into a territory that later would be termed as "Legal Amazon". By installing themselves through their characteristic system of occupation2, peasants in the
" The "alqueires per linha" (one "alqueire"is equal to 30 kg of rice, and one "linha" equivalent to
0.33 hectares) still is the unit of measure for the rent paid by sharecroppers. According to my personal observation, peasants in Maranhio consider one "alqueire" as a fair payment. Two "alqueires" are accepted as normal. Three "alqueires" are considered a violation of social rules. 23 For the description of the peasant system of occupation in the "expansion front", whose basic element is the center-edge opposition, see Velho (1972: 224-225); Almeida (1974: 16-21)

frontier seldom asked for rental payments for those in their social group. As land was not perceived as a commodity4, rental obligations were not mandatory for newcomers in a recently settled area (centro) (Santos 1981: 19). Convenient for both the squatter in need of labor force for the expansion of his agricultural area, and for those newly arrived in need of start-up support, such cases involved a great deal of labor exchange, and eventual payments of an attenuated rent functioned for the recognition of early settler's tenure rights in that site. Constituting a pioneer agricultural frontier, these peasants heavily depended on rice cultivation, that joined babassu as the most important product of Maranhio's economy in the 1950s, the decade signaling the intensification of this frontier.
By specializing in rice cultivation, generating a product of inferior quality (compared to the rice produced in the southern and central Brazilian states)25 and therefore important for the supply of the lower strata of the urban population, this peasantry integrates itself into a commercial circuit of capitalist accumulation. In areas of older occupation the predominant pattern is the landowner control of agricultural commercialization. In the frontier, conversely, the commercial circuit has the middleman as its key element, who also often own rice mills (Droulers and Maury 1981: 1041). Velho stresses road construction in the late 1950s as the cause of a process of vertical integration of this peasantry with the national market, through the supply of rice. Although peasant
' Hypothesis refused by Musumeci. She condemns the categorization of a peasant mode of production in the "expansion front" based on common land use, opposed to capitalist private property. See Musumeci (1988: 30-42)
2 "rice from Maranhao competes at the SMo Paulo and Rio markets due to its low price. At the fairs and groceries stores, it is always the cheapest one. It has an irregular aspect, is yellowish, has a lot of impurities and broken grains. (Valverde 1957: 15; my translation)

production was effectively reaching further markets, Velho's statement that through these commercial transactions "this peasantry seems to be gradually demarginalizing itself" (Velho 1972: 216-218) denies market integration already existent since at least the end of the 19" century, as other authors have demonstrated26.
Beyond demonstrating market integration, Musumeci contests theories attributing to peasants in the frontier a survival strategy based on self-subsistence, only secondarily targeting the market. Her work focuses on the Mearim valley, and mentions Velho's apparent reconsideration of the subject. She initially establishes her point by documenting peasant engagement in cotton cultivation during the 1930s, a decade of record production and exports of cotton in Maranhao, when the Mearim valley, recently occupied by Northeasterners, was the biggest cotton producer zone of the state, predominantly through small peasant units. Musumeci questions the preferential adoption of rice cultivation by migrant peasants from the Northeast, as they had no tradition of rice cultivation nor was rice their staple food. She emphasizes that if receiving market signals and stimuli, peasant units of production and small scale agriculture do not impede an export-oriented production, neither are they an obstacle to mercantile accumulation. Musumeci highlights peasant subordination and capital accumulation through commercial and speculative mechanisms, a system that does not require land monopoly nor direct control of the labor-force. According to her, these peasants "are mercantile producers, integrated into a specific chain of commercial transactions which dictates predominant
26 Valverde (1957: 15) mentions a Caio Prado Junior work in which the latter attests that "since colonial times Maranhao's rice was an important product for local consumption and for exports to the Northeast". See also Almeida (1996: 19)

market trends and heavily influences choices of the unit ofproduction." (Musumeci 1988: 181-196).27
The interpretation of Manoel da Conceiggo, a peasant leader who personifies in his life the progression of the "expansion front" through the Itapecuru, Mearim, Pindar6 and Turi valleys, realistically summarizes peasant commercial subordination. A union and political activist, who was mutilated, tortured, imprisoned, and exiled by the authoritarian regime, he states:
"There are books and theories about a certain subsistence economy. They explain
that workers sell their surplus production. This is not true because the worker
does not sell his surplus. He himself gets the 'surplus', if something remains after
he pays all his debts to the middleman. If nothing remains it is because the
middleman got the last grain of rice. Thus there is no surplus, the worker sold all
his produce. "(Conceioo 1980: 59; my translation)
Hence, economic and political control and the accumulation of wealth through commercial operations are based in a series of practices, including the provision of advanced credit, to be converted in rice payments"3, lower prices paid at the harvest, and higher prices for basic staple and consumer goods for the household. In addition, the middleman imposes exclusive rights to purchase babassu production, and process rice, manioc, and maize. Most landowners installed a commercial store and became middlemen.
27 Musumeci's research in the Mearim was coordinated by Velho, as part of the project: The process of colonization of Maranhio's humid valleys and its follow-ups. In Maranhio, January to April is the hardest period for peasants. Due to previous obligations to the middleman, their supply of rice is usually exhausted. Peasant engagement in wage activities conflicts with labor requirements in their own fields. In addition, this period coincides with the year's lowest production of babassu. Middlemen advance rice at a 2 or 3:1 ratio, being paid after the harvest. A lower value is stipulated for peasant's rice when middlemen lend cash, forcing a significant portion of the harvest to be used for this payment. For a categorization of the economic transactions pertaining to the rice commercial circuit in Maranhio, see Musumeci (1988: 264-286)

In recently occupied areas where there was no land ownership, commercial subordination was initially referred directly to a larger village or town businessman. Gradual differentiation among squatters resulted in the appearance of local middlemen, linked to those in the immediate superior level in the trading chain, which extends to two or three additional components. Through debt peonage, price mechanisms, and processing machines, the middlemen left no room for peasant accumulation of eventual surpluses". While commercial extraction by the landowner in lands of older occupation contributed to the maintenance of the agrarian structure and to the social subordination of the peasant, the economic control of the middlemen was one of the main sources of social differentiation and concentration of land ownership in the "expansion front". Having no means to pay their debts, and attracted by a still open westward frontier, the alternative to negotiate their tenure rights with the middleman was convenient to many families.
In fact, thousands of peasants who had settled in the Mearim Valley by the 1960s moved to the valleys of the Graja6i, Pindari, Turi, and Tocantins rivers, in the 1970s and 1980s. Juxtaposing characteristics of peasant economic strategy based on shiftingcultivation, the availability of unoccupied lands, and pressures resulting from commercial exploitation, Andrade describes this period of "spontaneous" settlement as one in which
"peasants from elsewhere in the Northeast migrated to Maranhdo in search of virgin forest land Upon finding it, they built a small hut, cleared afield in the
forest, burned the brush and trees, and prepared the fieldfor planting.... The
Manioc flour is produced in the village through manual or diesel powered "casas-de-faiinha", under a 5 to 10 percent fee. Rice is processed in mills "piladores de arroz" charging 1 kg per 30 kg, plus by-products (used as swine feed), and intentional scale errors in weighting: 30 kg of paddy are usually converted to 21 kg of rice, although producing up to 24 kg.

following year, because the soil was less fertile, the peasants planted manioc in the same field and cleared a new plot in the forest on which to plant rice. They
movedforward each year, making new clearings and carrying out what might be
called migratory agriculture. The lIandv left behind are being occupied by
merchants, rice millers, or government officials, who fence the landfor cattle raising... The poor peasant, thus, clears areas for the rich livestock raiser to
occupy. (Andrade 1980: 188)
Circuits of commercial and speculative accumulation were more benefitted by the lack of social and infra structural policies targeting the needs of peasants at the "expansion front". Instead of recognizing this peasantry's tenure rights in the land they occupied, state conduct favored the proliferation of illegal land appropriation (grilagem), promoted internal social differentiation, and stimulated the conversion to pastures of lands recently deforested and potentially productive for agriculture. Moreover, it generated an unprecedented wave of violence and expropriation. Land incorporation into the market, pasture conversion and agrarian conflicts
The Superintendency for the development of the Northeast (SUDENE
Superintendencia para o Desenvolvimento do Nordeste), was created in December 1959, having CEPAL's economist Celso Furtado as its first president. One of the bases of SUDENE's policy in its initial operation under the Furtado administration was the colonization of Maranhilo's forested areas, targeting the demarginalization of peasants through their transformation in "efficient food producers" for the Northeast, and the assimilation of population surpluses. (Velho 1972: 214)
"Furtado considered Maranhio a favored area because it experienced
spontaneous migration from the semi-arid zone of the Northeast, and had an
abundance of unoccupied government land.... S(JDENE planned to obtain
donations of extensive areas in Maranhilo: in the West, on the Pindard river, in
the upper Mearim, and later in the Gurupi valley, in the forested zone. These

lands were to be divided into parcels and given to families brought from the semiarid zone. To guarantee the success of this undertaking, Ambavsador Merwin
Bohen, special representative of President Kennedy, went to the region in order to
channelfunds from the Alliance for Progress andfoodstuffs from the Foodfor
Peace program. "(Andrade, M. C. 1980: 195)
Sudene's original "Projeto de Povoamento do Maranhio" (PPM), the Maranhio's colonization project, was created in 1962 with the ambitious goal of settling and coordinating the "spontaneous" settlement of 40,000 peasant families in five years, over three million hectares donated by the state government. Results were meager. It took ten years for the effective operation of the project, under the authoritarian regime. The implementation was restricted to 940,000 hectares at the Pindar6 and Turi valleys in the Northwest portion of the state, for the settlement of 10,700 families in its first two phases (490,000 ha), and the later expansion to a third area with 450,000'o. As in the COMARCO project3, and at the PIC-Barra do Corda32, the two other official colonization initiatives in Maranhilo, multiple agrarian problems surrounded the execution of the schemes. On one hand, the "spontaneous" colonization exceeded expectations, and
3 "PCAT Projeto de Colonizaio do Alto Turi", since 1972 managed by the Northeast Colonization Company (COLONE Companhia de Colonizaqio do Nordeste) and installed in an area crossed by BR-316, the road that connects Bel6mn to the Northeast. Funds for this project originated from a $6.7 million World Bank loan to the Brazilian government. For more details about COLONE, see Droulers. and Maury (1981:1041-1049); Trovio (1989: 137-141). 31"COMARCO Companhia Maranhense de Colonizago", the Maranhbio Colonization Company, created in 1971 by the state government to settle 10,000 families in 30 ha plots at the Grajaii and Pindar6 valleys. In addition, COMARCO would sell 1.4 million ha to corporations, at lower prices and no interest. Whereas the latter action was (illegally) conducted, 1035 peasant families were settled until 1974, and 600 remained four years later. (Troviio 1989: 132-137). 32"PIC Projeto Integrado de Colonizagio de Barra do Corda", the INCRA (Agrarian Reform Agency) led Integrated Colonization Project at the upper Mearim region, targeting settlement of 3,000 peasant families on 340,000 ha. (Almeida and Mourio 1976: 7).

approximately 20,000 families occupied the area even before the conclusion of its first phase, which envisioned the installation of 5,200 families on 240,000 ha. Droulers and Maury describe that the application of the recognized "economic model" was possible only in 1977, and to a subset of only 500 families (Droulers and Maury 1981:1042), favoring the formation of a restricted middle class within the peasantry. On the other hand, and moreover, this area witnessed pressures of ranchers and corporations illegally purchasing properties and initiating land conflicts. As Almeida and Mourao stated, "the magnitude of the conflicts, which rational colonization of available lands should supposedly reduce, became instead even greater, bringing land struggles to the expansion front" (Almeida and Mourio 1976: 9).
Outside the colonization schemes, state action intensified the implementation of a concentration-based development model, with no regard to its social impact. The 1969 state government "Law of the Land" (Lei de Terras), and fiscal incentives and subsidies generated by the authoritarian regime transformed social relations of production in rural Maranhao. A process of property legalization was initiated. Nevertheless it was limited to a circle of better off and more entrepreneurial producers, more often Northeasterners, combining agriculture with small scale livestock, and acting as local middlemen. In many situations, although originally members of the community, these newly recognized landowners excluded and harmed their former friends. With the possession of deeds, they became entitled to claim rural credit and benefit from official programs". 31 See Velho (1972: 228) fbr his interpretation of social stratification in Eastern Amazon's peasantry. See also Santos (1981: 28-31), and Musumeci (1988:118-128) for characterizations of the "inside grileiros", usually referring to better-off peasants taking advantage of their "insider"

In other areas, as happened at the COMARCO scheme, untitled state lands (terras devolutas) were appropriated through "grilagem", the illegal instrument of documentation of large tracts of land and posterior speculative sale of these lands, usually with peasant communities in their interior. The so-called grileiro "is the most important element in the transition of an untitled piece of land in private property, either benefitting corporations, ranchers, or himself" (Trov~o 1989: 91; my translation). Attracted by the profitability of the deal, these corporations and wealthy ranchers became the certified landowners of extensive properties. Most of the peasant communities had neither the information nor the opportunity to demonstrate their tenure rights, and after being expropriated through labor (rent), prohibited access to natural resources (babassu, wood, game, fish), having their house and tools vandalized, they were finally expelled from the land (Almeida 1981: 8Y
Extraction of surpluses through rent and commercialization were no longer the
only sources of capital accumulation for landowners. Indeed, the hegemonic classes were successful in influencing policies in their benefit. The state, with generous offers of subsidized capital and fiscal incentives, determined that "... especially in the 1970s, an active process of pasture conversion has been replaced rice fields, ... through financial support of the two regional development authorities, Maranhio had 77 government subsidized livestock projects in an area of more than one million hectares" (Amaral Filho 1990: 233; my translation).
condition as a disguise for illegal land appropriation.
For an actual description of the stages and elements of a process of"grilagem", see Asselim (1982: 60-61).

State policy resulted in the shift from rice to cattle. In the period between 1975
and 1985, IBGE figures show that the area converted to pastures in Maranho more than doubled, from 12,182 km2 to 27,903 kin2, whereas the area for annual agriculture presented a much lower increase, from 10, 140 km2 to 12,181 km2. In the same period, while the state's cattle herd increased from 1.78 million to 3.25 million, rice production decreased from 894,000 tons to 779,000 tons (LBGE. 1976, 1986).
Pasture conversion led to increasing concentration of landownership and agrarian violence, by the removal of squatters and sharecroppers, and the consequent demographic pressure on the remaining agricultural land. Besides the elimination of hundreds of villages (whose existence is witnessed by clusters of mango trees within extensive pastures and on the edges of the roads), peasants were restricted accessing to babassu stands enclosed in ranchers' and corporations' properties. Babassu palm eradication, due to both a false idea of palm competition with pastures, and the ideological motivation to avoid peasant economic activity, became an effective way to accelerate this strategy.
Cooperation and flexibility formerly provided by landowners and middlemen in
critical instances ceased to exist. Land left for agriculture was soon exhausted, resulting in increasing risks and decreasing yields, reducing safety margins of middlemen's commercial gain. Extensive livestock production required only a tiny portion of the peasant labor force, working as waged "cowboys". Westward migration of families to further areas in Pari was still an alternative, but since the mid 1980s even this option has become limited as the same restrictions were progressively applied to the frontier. Peasants were thus left with two major alternatives: migration to urban centers, and seasonal migration, either

working for wages on the installation of agricultural, livestock or infra-structural projects, or to the "garimpos", the mining areas. Urban migration inflated the unbalanced urbanized peripheral zones of Sao Luis, Imperatriz, Caxias, or BacabaP5. Through seasonal migration, peasants from Maranhio constituted an impressive portion of the "garimpeiros", the independent mining prospectors working in the Amazon, and even outside the Brazilian territory '. Both alternatives represented a profound contrast for those who experienced them, in comparison with their original peasant values and culture. However, the existence of these options and the remittances they provided, was crucial for the maintenance of a still numerous peasantry in the state of Maranhio, and particularly in the Mearim and Grajaii valleys.
Changing realities for an aero-extractive peasantry: globalization and reaction
Besides the alleviation of demographic pressures due to out-migration and the
support of eventual cash transfers from urban and mining activities to the original peasant unit, another important factor contributing to the livelihood of the peasantry in large areas of Maranhio is the income provided by babassu. Even with strong restrictions to access the resource base since the 1970s, babassu's unique provision of both monetary and nonmonetary income has effectively been a "subsidy from nature" to the peasant. Nevertheless, although representing one of the bases for the regional economy, babassu
3 According to IBGE census, the 1991 population of the three largest urban centers of Maranhio was 700,000 (Sao Luis), 210,000 (Imperatriz), and 85,000 (Caxias).
3 Maranhio is the state that contributes the largest number of "garimpeiros" in the Amazon. Almeida (1989), for instance, mentions that in 1986, there were 400,000 "garimpeiros" in the Brazilian Amazon, 120,000 of them being from Maranhio.

extractive activity has been progressively losing even the reduced government support that it received until the early 1980s37. Land commoditization relocates the basis of capitalist accumulation from a system based in commercial exploitation to another based on landed ranchers and agro-industrial enterprises. Parallel to pasture conversion and land speculation, occurs the polarization of interests regarding land uses. Extractive activity based on the domestic unit of production was increasingly viewed as an obstacle for rural development of Maranhio. The official discourse, in the name of the interests of these hegemonic classes, attributes to babassu extractivism a generalized backward feature that reproduces the impressionistic narratives of the geographers who visited the "babassu zone" during the 1950s and 1960s38, but without the social concern of those narratives. In addition to the ongoing transformation on entitlements for land use, were added initiatives targeting the transformation of the productive process in the babassu economy. Foreseeing a feasible integration of babassu within the "capitalist livestock enterprise", and matching the interests of ranchers aiming to eliminate continuous peasant entrance in private properties, the reduced group of politicians, researchers, and entrepreneurs sympathetic to babassu addressed technological change as the redemption for the sector.
Different from the traditional babassu oil industry, the proposed babassu agroindustrial complexes considered a process of horizontal integration, through which the 37 During the oil shortage of World War II, fbr instance, a report from the US Vegetable Oil mission evaluated babassu's potential value as "... five times the value of Brazilian coffee crop" (Brazilian Government Trade Bureau, 1950). The last period in which babassu deserved the attention of government authorities was during the mid 1980s oil crisis, when the federal based "Babassu's Institute", was created. After a few years, when petroleum supplies were no longer an acute problem, it ceased to exist.
38 see Azevedo and Matos 1951; Valverde 1957; Andrade 1964.

industry would either own land or establish contracts with ranchers who would "supply" the babassu. Technological development, therefore, rather than inspiring changes in babassu zone's agrarian structure, intended to reinforce it, as seen in Almeida's quote from a report presented in 1991 by the Brazilian Association of Babassu Industries (ABIBA Assoiago Brasileira das Indfistrias do Baba9u), to Maranhao's governor. One of ABIBA's proposals is "to stimulate landowners to invest in the [babassul sector as a complementary source of income in their properties, incorporating a permanent laborforce and avoiding possible invasions of babassu gatherers" (Almeida 1995: 28). The adoption of this technology, the industrial cracking, separation, and integral utilization of babassu's component parts, was attempted in the late 1970s and early 1980s through some plants in the Itapecuru and Mearim valleys. All these projects failed, allegedly due to an inconsistency in babassu supply and refusal of the peasants to being a mere supplier of raw material ". As a matter of fact, the widespread adoption of industrial babassu cracking without agrarian reform and social policies targeting peasant needs would have profound negative impacts, and peasants seem to be aware of that. As May summarizes, consequences include the loss of opportunities for the allocation of peasant labor (as fewer people will be needed to collect whole fruits), imbalances at the gender division of labor within the household, and a substantial reduction of total income due to the loss of other products, especially charcoal and starch (May 1994:13-15). Even when the technology was installed in an area where peasants were reassured of their tenure rights through an
3 For the description of the technology and an analysis of its application, see May (1990:242-252)

INCRA-managed settlement.project, its operation did not last two years, and from the beginning was viewed, especially by women, as worsening household living conditions0.
Parallel to the attempts to develop and industrialize babassu cracking, a series of much stronger technological advances seriously affected the babassu economy. From the 1970s, developments in the industrial and agricultural sectors resulted in the dislocation of babassu oil in both the food and hygiene industries4'. The collapse of most of the babassu oil processing units did not represent the collapse of the industrialists, as most of the involved capital was also subsidized by governmental programs. Industries whose scale was large enough had simply to readapt their operation to the comparatively advantageous raw materials, especially soybeans, whose cultivation in the south of the state was being increasingly supported by governmental programs. Peasants, however, suffering direct consequences of market restrictions, had neither the material conditions nor the choice to adopt any short-term adaptation, as the impact of this substitution was the reduction of the already low prices and a decreasing demand for babassu kernels, affecting the bargaining power of the suppliers in the local and regional economies.
Furthermore, in the early 1990s, still other factor ultimately threatened the
domestic mode of production in the babassu region. Following trends of global trade
'0 An industrial plant was installed in 1989 by the Maranhio Enterprise for Agricultural and Livestock Research (EMAPA Empresa Maranhense de Pesquisa Agropecuiria) at the Monte Cristo settlement project, in Slo Luis Gonzaga. Union and peasant leaders initially evaluated as positive the installation of such a unit, as a way to test whether or not the technology itself would work in a situation of assured tenure rights. The operation of this unit and peasant reactions were observed directly by the author.
" Improved varieties of soybeans replaced babassu oil in the food industry. Indeed, babassu and cottonseed oils were the most consumed in Northeast Brazil until the 1970s. Babassu oil maintains only a small portion of this market, progressively being replaced by soy oil. Synthetic raw materials, on the other hand, are replacing babassu oil in the cosmetic and hygienic industry. See May 1993. "Babassu Palm Product Markets".

liberalization, and acknowledging the interests of the hygiene industry oligopoly (Unilever and Colgate-Palmolive), a policy based on the reduction of vegetable oil import tariffs was initiated. The previous 1990 import tariff of 18% was reduced to 2% in 1995 (Almeida 1995: 45). As a result, increasing quantities of plantation-originated lauric oils (palm, palm-kernel, and copra oil) have been imported since 1991, mainly from Southeast Asia, and especially from Malaysia, replacing extractive based, peasant production of babassu oil.
Coinciding with the intensification of socioeconomic pressures, the last fifteen years have witnessed a process of change in peasant social organization and collective action in Maranh~o. In having their survival threatened, and no longer relying on the "expansion front", sectors of this peasantry began to undertake organizational processes. Following the gap resulting from the authoritarian regime's repression of previous initiatives (Almeida 1981; Conceico 1980), the intensification of more politicized activities within religious communities was usually the first step of this process. Through the participation in the "CEB's Comunidades Fxlesiais de Base", and other social movements linked to or supported by the Catholic Church42, peasants were motivated for organized reactions against socioeconomic oppression. Subsequently, many religious leaders entered in the union movement, which after being coopted by the military regime, was struggling to recover peasants' trust and political participation. Since the mid 1980s, with the spiritual guidance obtained through the CEB's and the effective support of the rural unions, a number of agrarian conflicts resulted in peasant victories. Particularly after
42The "ACR Animation of the Rural Christians", "MEB Movement for the Basis' Evangelization", and "MIER Movement for Rural Evangelization" were the more frequent religiously-linked organizations among peasants in the Mearim valley.

Full Text