Portuguese Style and Luso-African Identity: Precolonial Senegambia, Sixteenth - Nineteenth Centuries 0253215528, 9780253215529

Citation preview

“PORTUGUESE” STYLE AND LUSO-AFRICAN IDENTITY

“PORTUGUESE” STYLE AND LUSO-AFRICAN IDENTITY Precolonial Senegambia, Sixteenth–Nineteenth Centuries

PETER MARK

This book is a publication of Indiana University Press 601 North Morton Street Bloomington, IN 47404-3797 USA http://iupress.indiana.edu Telephone orders 800-842-6796 Fax orders 812-855-7931 Orders by e-mail [email protected] © 2002 by Peter Mark All rights reserved No part of this book may be reproduced or utilized in any form or by any means, electronic or mechanical, including photocopying and recording, or by any information storage and retrieval system, without permission in writing from the publisher. The Association of American University Presses’ Resolution on Permissions constitutes the only exception to this prohibition. The paper used in this publication meets the minimum requirements of American National Standard for Information Sciences—Permanence of Paper for Printed Library Materials, ANSI Z39.48-1984. Manufactured in the United States of America Library of Congress Cataloging-in-Publication Data Mark, Peter, date “Portuguese” style and Luso-African identity : precolonial Senegambia, sixteenth–nineteenth centuries / Peter Mark. p. cm. Includes bibliographical references and index. ISBN 0-253-34155-8 (alk. paper) 1. Architecture, Domestic—Senegambia. 2. Architecture, Portuguese Colonial—Senegambia. 3. Vernacular architecture—Senegambia. 4. Miscegenation—Senegambia. I. Title. NA7467.6.S4 M37 2002 728’.37’08969066—dc21 2002004292 1 2 3 4 5 07 06 05 04 03 02

CONTENTS acknowledgments Introduction

vii 1

ONE.

The Evolution of “Portuguese” Identity: Luso-Africans on the Upper Guinea Coast from the Sixteenth Century to the Early Nineteenth Century TWO. Sixteenth- and Seventeenth-Century Architecture in the Gambia-Geba Region and the Articulation of Luso-African Ethnicity THREE. Reconstructing West African Architectural History: Images of Seventeenth-Century “Portuguese”-Style Houses in Brazil FOUR. “The People There Are Beginning to Take on English Manners”: Mixed Manners in Seventeenthand Early-Eighteenth-Century Gambia FIVE. Senegambia from the Mid-Eighteenth Century to the Mid-Nineteenth Century SIX. Casamance Architecture from 1850 to the Establishment of Colonial Administration Conclusions and Observations

114 144

notes bibliography index

149 189 199

13 33 59 81 97

Acknowledgments To undertake the extended challenge of writing a book is, on one level, a lonely task. But on another level it is a collective undertaking. To gather historical information from archival, published, or oral sources entails a dialogue—with chroniclers long dead, with historians past and present, and with informants. Sharing one’s nascent interpretations with colleagues constitutes another form of dialogue, one quite necessary for refining the work. I have been fortunate to have the assistance and encouragement of many individuals. And I have been blessed—though at the time I might have used another term—with friends and colleagues whose critiques have forced me to rethink and refine the ideas expressed in the pages that follow. Several institutions also offered crucial material support for my research and writing. Wesleyan University is reputed to offer the most generous faculty sabbatical policy this side of paradise. This book is a product of four leaves generously granted by Wesleyan. I am grateful, too, for five Wesleyan Project Grants that have enabled me to return to Senegal on a regular basis. I can only hope that this book begins to justify such generosity on the part of my home institution. The National Endowment for the Humanities supported a semester’s research and travel. I gratefully acknowledge this support. In Senegal, technical assistance was provided by Forutsen and housing by CODESRIA. The West African Research Center (WARC) and the History Department of the University of Dakar both offered opportunities to share my ideas in seminars with Senegalese students and colleagues. In Bamako, Mali, I was welcomed at Point Sud and there benefited from the scintillating intellectual atmosphere. In Frankfurt am Main, the Frobenius-Institut afforded access to their excellent library and kindly allowed me to publish several drawings from their archives. The Basel Mission kindly made available their extensive photographic archives. Challenging debate was provided by my colleagues at the Leiden University African Studies Center, the Université Paris-7, the Centre de Recherches Africaines (Université-Paris I), and the Mande Studies Association. In the Casamance, I have enjoyed the friendship of Dr. Bakary Sane for many years and have greatly benefited from his profound knowledge of Jola (Diola) history and culture. Together, we have traveled across the region. More than once, during the all-too-frequent outbreaks of violence that have tortured this most beautiful land, his reassuring presence has given me the strength to carry on field work. In Fogny, Fiacre Badji has given freely of his hospitality, and in the community of Thionk-Essyl, I am beholden, as ever, to my friend of twenty-five years, chef of Batine

viii

Acknowledgments

ward, Abba Badji. I am also indebted to my hosts, the family of Mamadou “Songo” Djiba. At the University of Dakar, Ibrahima Thioub and Boubacar Barry have offered their historical insight and, more than once, a place to stay. The Archives Nationales du Sénégal have been an essential resource. I offer my sincere thanks to the director, Saliou Mbaye, and to Mamadou Ndiaye. Finally, on Gorée, that magical island that transcends time, Lulu Lapolice has provided friendly advice and a place to call home. Many colleagues offered their time and professional expertise to read and critique parts of the manuscript. I wish to express special thanks to Mamadou Diawara for his painstaking assessment of Chapters 1–3. Tom Spear worked tirelessly to help me improve a version of Chapter 1 that was subsequently published in The Journal of African History. David Robinson also read that chapter with the same care he once took reading my graduate school essays nearly thirty years ago. Catherine CoqueryVidrovitch, the late P. E. H. Hair (the expert on early Portuguese sources), Jan Vansina, and Jean Boulègue all suggested important bibliographical additions and conceptual revisions. I am deeply grateful for their assistance. Long conversations with José da Silva Horta gave me the beginnings of an understanding of the relationship of color and identity in sixteenth- and seventeenth-century Portugal and the Cape Verde Islands. Don Wright, friend since our paths first crossed in Senegal in 1975, read several chapters; his insightful work on oral traditions first inspired me to study the history of the “Portuguese” in the Gambia. Walter Hawthorne and Joshua Forrest kindly made available their work in progress, followed by stimulating discussions that have increased my understanding of Guinea-Bissau’s early history. Michelle Gilbert helped me to a deeper appreciation of the social symbolism of architectural form. Clark Maines and Phil Wagoner, my colleagues at Wesleyan, and Peter Golden have taught me to read and draw architectural plans. Dominique Malaquais, whose brilliant dissertation influenced my own approach to architectural history, invited me to co-teach an architectural study tour of Senegal in 1997. My photographs of Carabane and Mlomp date from that trip. In the spring of 2001, a projected return visit to these communities fell victim to the heightened strife in the Casamance. But if I lost the opportunity to take a few photographs, countless thousands of Casamançais lost far more—their economic well-being, their health and, for several thousand people, their lives. The Casamance has suffered from fighting for almost twenty years. It is long since time for the weapons to fall silent, for the violence to end. Jay Edwards first suggested to me the historical significance of Frans Post’s depictions of Afro-Brazilian architecture. I express my sincere gratitude. David Henige published my first two essays on “Portuguese” architecture—after discreetly suggesting improvements—in History in Africa. George Brooks offered encouragement. Rod McIntosh brought

Acknowledgments

ix

me up to speed on the archaeological evidence for Mande architecture of the first millennium. And my anonymous readers for Indiana University Press and The Journal of African History made crucial suggestions for sharpening my historical arguments. To all of these people I offer deep thanks. The family of my deeply missed friend and mentor Louis-Vincent Thomas kindly offered me his personal archive of photographs. To his daughter, Domitienne, I express particular thanks. Father Henri Goovers of the Holy Ghost Fathers Mission, who took many of these 1950s photographs, was able, in 2001, to identify most of them. I am deeply indebted to him. Permission to reproduce photographs was also kindly provided by Peter Steigerwald (Frobenius-Institut), Alain Hardy, and Paul Pélissier, to each of whom I am most grateful. Jos van der Klei, whom I first met in Buluf in 1974, has been both colleague and friend ever since, as well as host and drinking companion in Amsterdam. Ferdinand de Jong has offered his insightful analyses of Casamance society; I have benefited from his dry wit and spirited critique which, as often as not, has won my agreement. Philip Havik has been most helpful with Portuguese sources. Paul Jenkins introduced me to the critical use of photography as historical source and warmly welcomed me to the Basel Mission Archives. And I have greatly benefited from the friendly critique and the uncritical friendship of Hélène d’Almeida-Topor, Peter Geschiere, Jan Jansen, and Amanda Sackur. In Strasbourg, the legal advice of my friend Michel Weckel magically helped me to obtain a carte de résident when, having overstayed my visa, I risked seeing my research in France interrupted by higher authorities. I also extend warm thanks to Meyer Gross and Tom Mark, as well as to Carla Antonaccio and Roger Haeffele. I owe a special thanks to my editor at Indiana, Dee Mortensen. Her encouragement and guidance through one major revision and subsequent fine-tuning of the manuscript have made the process much easier and, at times, quite enjoyable. Janet Rabinowitch, Editorial Director at Indiana University Press, encouraged this project from an early stage. For this I am most grateful. Kate Babbitt has been a wonderful copy editor and email friend. Amy Rettew was enormously helpful as I struggled to bring logic and clarity to the manuscript. The expertise she applies to writing legal briefs helped to improve my historical exposition. Nancy Allen provided invaluable assistance as copy editor. And my son Chris demonstrated the value of a Columbia history degree. His perceptive critical commentary— and his courage in offering it to his father—convinced me to undertake yet another rewrite and filled me with parental pride. To my parents, Dr. Herbert Mark and Avra Mark, who first nurtured my love for intellectual inquiry and who, even today, encourage me to hold my writing to the highest standard, I owe the longest and the deepest

x

Acknowledgments

debt. To my father I also owe my love of photography, perhaps evident in some of the images that accompany this text. And to Odile Goerg, esteemed colleague, my life’s partner, and in-house critic—and whose support on the home front has made possible several field trips to Africa—I offer this book. One more person has played an important role not only in the gestation of this book, but also throughout my career. When I arrived in Senegal in 1974, slightly overwhelmed by a first taste of Africa, Martin Klein showed me the ropes, from the Archives Nationales to the restaurants of Dakar. My first conference paper was presented at his invitation; so too, I realize, was my most recent one. Marty has been an inspiration as a scholar. His insightful and delicately offered critique has greatly improved the present study. In addition, he has been a professional role model, unselfishly sharing his knowledge and offering encouragement to younger scholars. I dedicate this book to Herbert and Avra Mark and to Martin Klein. Ziguinchor June 18, 2001

“PORTUGUESE” STYLE AND LUSO-AFRICAN IDENTITY

Introduction

In precolonial Senegambia from the sixteenth to the late nineteenth century, the construction of social and cultural identities and the construction of houses constituted two intimately related processes. These processes led to the creation, on the one hand, of abstract and flexible social concepts and categories and, on the other hand, of physical structures. The buildings and the social concepts were, however, closely interrelated. The houses that people build serve not only to provide shelter but also to symbolize or to articulate the owners’ social and cultural identity. Along the Upper Guinea Coast, from the Petite Côte south to Cacheu, domestic architecture constituted an important element in the expression of social status, economic position, and material culture. The geopolitical situation in Senegambia was complex. North of the Gambia River, the medieval state of Jolof extended over a vast area from the Senegal River south to the Gambia. But during the sixteenth century, Jolof ceded autonomy to the component Wolof and Serer (Sereer) states of Kajoor, Baol, Waalo, Siin and Saloum. Islam, which had been present in the region since at least the eleventh century c.e., was widely disseminated, although it was also frequently mixed with local religious practices. The population of the lower Gambia was a mixture of indigenous Serer, Wolofs, and the descendents of Manding traders and warriors who had migrated west from the heartland of the Mande-speaking peoples. Many of these long-distance merchants were Muslims. This Manding migration and the accompanying cultural assimilation of local populations probably had begun by the fourteenth century. The coast from the Gambia south through present-day Guinea-Bissau is a region of low-lying plateaus and tidal swamps bisected by numerous coastal waterways. It is also a region of significant linguistic and cultural diversity. Early Portuguese sources suggest the existence of Bainunk- and Mande-dominated states along the Casamance River. The largest and most powerful of these states was Kaabu, situated southeast of the Casamance between the coastal lowlands and the interior. In the sixteenth century, this Mande-dominated state exerted commercial and limited political influence north to the Gambia. In the late sixteenth and seventeenth centuries, sociopolitical organization in the Casamance probably did not surpass the level of intervillage alliances. Throughout the CasamanceBissau coastal region until the end of the nineteenth century, Islam was

2

“portuguese” style and luso-african identity

1. Map of the Senegambian trading centers and escales from the Petite Côte to Bissau. Map by Jim Hull and the author.

limited to a few long-distance merchants (Dyulas). Each local group had its own particular religious practices. Identity has been defined as the dynamic product of a dialectical interaction between self-definition and ascribed or imposed definitions.1 No community, no social group, can articulate its identity independently of the outside world. Those who are not members of a given group may help to create and revise the identity of the group. Identity thus entails a dia-

Introduction

3

2. Map of major rivers, cities, and towns in the Senegambia. Map by Jim Hull and the author.

logue or discourse between members of the group in question and members of other groups with whom they enter into contact. Architecture, the most visible and durable component of material culture, influences the creation of images pertaining to that culture. These images are articulated not only by those who build and live in the houses but also by outsiders. The images outsiders form of members of a group eventually define the group’s ascribed identity. These images become part

4

“portuguese” style and luso-african identity

of the discourse and, ultimately, they may be integrated into the group’s own sense of who they are. When these external images differ significantly from the group’s sense of being, they may engender a countervailing assertion as group members seek to deny the ascribed identity. Inevitably, however, members of a society have to respond to the manner in which they are perceived and defined. This is particularly true where the outside world exerts local political and economic influence, as was the case with European colonial powers in nineteenth-century Senegambia. THE ETHNOGRAPHIC SETTING Along the northern Upper Guinea Coast and in the immediate hinterland from the Gambia River south to Cacheu (Figure 1, Figure 2), the buildings people lived in and the physical layout of their communities served as important elements in the articulation of their cultural identity. In this region of extensive long-distance trade, architecture reflected contact among the different populations. In the constantly changing constellation of diverse populations, identity was in part a measure of material culture, of language, and of local traditions of origin. At any given moment, a group’s identity was the dynamic product of ongoing cultural interaction and assimilation. Among the groups inhabiting this region were several thousand LusoAfricans, or “Portuguese,” as they were widely known. The local African populations of West-Atlantic speakers—Floups, Bagnuns, Bijogos, Papels, Balantas—were for the most part organized into small-scale, decentralized societies. Mande-speaking peoples inhabited the small states of the lower Gambia and the more important state of Kaabu in the northeastern part of modern-day Guinea-Bissau. By the nineteenth century, Fulbe migrants from the Futa Jalon had also established communities in the Middle and Upper Casamance. From the perspective of the Mande-speaking Juula (dyula) merchants whose caravans connected the Gambia and Casamance Rivers with the regions to the east, the Upper Guinea Coast represented the western limit of the Mande commercial diaspora. The associated Mande cultural zone extended nearly 1,000 miles, to the inland delta of the Niger River in Mali. From a European perspective, on the other hand, the coast was itself the center of commercial activity and the focal point for contact with the cultures of West Africa. H I S TO R I CA L Q U E S T I O N S From an historical perspective there are fundamental questions about precolonial identities in Senegambia. How did these identities change between the time of first contact with Portuguese traders in the fifteenth

Introduction

5

century and the eve of colonial control? What parameters served to define group identity, and how did these parameters change? Did the sixteenthand seventeenth-century presence of European and Cape Verdean merchants, and of a local Luso-African community, change the manner in which identities were articulated in the Gambia-Casamance-Cacheu region? What role did Mande-speaking merchants and Mande culture play in the ongoing evolution of local cultures and identities? A fundamental question regarding the process of identity articulation in precolonial Senegambia is: Does this process accord with older Western anthropological models for understanding identity and identity formation, notably that of Fredrick Barth? Barth’s model, which is central to more recent studies, posits that boundaries are fundamental to the definition of groups.2 Closely associated with this model is the principle that people define themselves by reference to that which they are not: an “other.” In fact, the peoples of precolonial Senegambia do not appear to have exhibited what we could term a standard or classical model of identity formation. One might also ask whether this situation was peculiar to Senegambia. Several additional observations are of paramount importance to a discussion of identity and identity formation. First, I argue that identity is the product of a continuous process or, more precisely, a continuously dynamic process. This fact is especially relevant to the discussion of métissage culturel in seventeenth-century Gambia, the subject of Chapter 4. Métissage culturel may appear to describe interaction and assimilation between two (or more) discrete cultures, the product of which is a new mixed culture. In fact, the component cultures were themselves the result of interaction and assimilation. Rather than constituting fixed cultural entities, the different elements that interacted in seventeenth-century Gambia, as elsewhere in the Senegambia-Cacheu region, were at any given moment the product of ongoing culture contact and change. Métissage culturel thus refers to the continuous process whereby two or more dynamic and continuously evolving cultures interact to create a new constellation of cultural elements that is distinct from its progenitors. This model poses a terminological challenge: How does the historian designate the component cultures or the product of their interaction? Where seventeenth-century sources designate groups by specific names, as with the “Mandingues” or the “Portugais,” I retain these designations. It is, however, apparent that what constituted a “Portuguese” in 1600 was very different from what constituted a “Portuguese” in 1850. One of the aims of this study is to discern the changing parameters by which members of this group identified themselves as “Portuguese” or were so designated by outsiders. In the case of many other seventeenth-century communities, no term existed that corresponds to twentieth-century ethnic designations. In some instances, migrations and the attendant assimilation

6

“portuguese” style and luso-african identity

of earlier inhabitants by more recent arrivals have altered the local cultural map. This is the case with the formerly “Bagnun” populations of the Gambia-Soungrougrou region and of Fogny, north of the Casamance River. In other instances, the labels associated with particular groups have changed, a process that also reflects the gradual coalescing, on the level of group identification, of several local populations (“peuples,” or “races,” as they were sometimes designated in seventeenth- and eighteenth-century French sources) into a geographically more extensive group. This is the situation with the Floups of Fogny, who in the course of the nineteenth century came to be identified as “Yolas” or “Diolas” (English: Jolas). Today the Jolas are described as an ethnic group. Members of this group now widely assert Jola identity. The term Diola is a nineteenth-century concept that is closely tied to the French colonial presence (administrators, missionaries, and anthropologists); before about 1900 the ancestors of the Diolas would not have considered themselves as such. The ancestors of the present-day Diola were formerly often called the Floups. No one calls them that today, least of all themselves. There are as many as ten subgroups of Diolas, distinguished mostly on the basis of language dialects.3 The national dialect of the Diola language is Diola Fogny; the people who live in Fogny, north of the Casamance River and south of The Gambia, are also called Diola-Fogny. Their neighbors to the west, in Buluf, are, not surprisingly, the Diola-Buluf. When I use the term Mlomp-Buluf, I do so to refer to the village in Buluf called Mlomp and to distinguish it from the village in Kasa, south of the Casamance River, also called Mlomp. So, in brief, many or most Diolas have Floup ancestry. One would not distinguish between Floups and Diolas. A few mid-nineteenth-century sources do that, but those writers existed at an historical moment when the appellation “Floup” was just being replaced by the term “Diola.” By the turn of the twentieth century, the term “Diola” had taken on a broader meaning that included all the speakers of Diola dialects, and approximated what we would perhaps call an ethnic group. Given the recent creation of Jola ethnicity, how does one refer to those seventeenth-century ancestors who themselves had no sense of being Jola? It is clearly advisable to avoid the proleptic use of later ethnic labels. However, alternative appellations, such as “the ancestors of the Jolas,” are awkward, and they oversimplify the complex interactions that led to the creation of the present cultural and linguistic group. Nevertheless, the admittedly awkward circumlocution “ancestors of” is preferable because it is less ahistorical than the anachronistic use of contemporary ethnic labels. The history of cultural interaction and evolution of cultural identity evokes a related question: Do contemporary identity concepts have meaning when applied retroactively across four centuries of time? The concept “ethnic group” is a relatively recent creation. The development and application of ethnic labels was often, as is now widely recognized by histori-

Introduction

7

ans and anthropologists, a product of the colonial period. In this limited sense of the term, many contemporary Senegambian ethnic groups did not exist before the nineteenth century. If, on the other hand, one defines “ethnic group” as a population whose members speak a common language or closely related dialects, exhibit a common social organization, share religious rituals (except in the case of differential conversion to Islam or Christianity), and perceive themselves as constituting a common identity, then one could reasonably argue that such groups did exist long before the colonial period. In this sense, sociocultural groupings analogous to contemporary “ethnic groups” have a demonstrable historical depth in Senegambia. A central aim of this book is to provide a detailed view of one specific group in the southern Senegambia-Cacheu region: the Luso-Africans or, as they are referred to in contemporary documents, the “Portuguese.” How did this relatively small group come into existence? What were the parameters of group identity? How, and in response to what external factors, did these criteria evolve from the sixteenth to the nineteenth century? How closely does the dynamic of this precolonial identity formation resemble the process whereby ethnic groups were discovered, defined, or created during the colonial period? In terms of the parameters of “Portuguese” identity, as well as the dynamics of identity formation, it would not be misleading to apply the label “ethnic group” to these precolonial Luso-Africans. In this regard, they constituted an ethnic group avant la lettre. The concept of ethnicity is itself problematic, especially if one attempts to apply it to cultures separated by four centuries. For the concept is not immutable in time and space. As Carola Lentz observes, ethnicity is not a strict analytical category that has a single and precise meaning. It is, instead, a rather vague concept: “I do not employ the concept of ethnicity as a strictly analytical category that lends itself to a single precise definition. Ethnicity is perhaps an imprecise concept.”4 Both precolonial and colonial identity formation entailed a dialectical interaction between the Luso-Africans, their African neighbors, and European merchants whose understanding of who was “Portuguese” differed from the Luso-Africans’ own sense of who they were. Pre-nineteenthcentury identities were not, however, a product of the same power relationships that characterized the colonial period. This factor suggests an important distinction between precolonial and colonial identity formation. The observations of Carola Lentz are again pertinent: “The processes of ethnic identity creation are historically and regionally specific.”5 In the chapters that follow, I seek to elucidate the processes by which “Portuguese” identity was formed and subsequently reformulated from the sixteenth century to the advent of French colonial administration in southern Senegambia in the second half of the nineteenth century. While the focus is on the region comprising the Petite Côte, the lower Gambia,

8

“portuguese” style and luso-african identity

the Casamance, and northwestern Guinea-Bissau (S. Domingos-Cacheu), historical reference is also made to “Portuguese” communities farther south along the Atlantic coast. While I endeavor to define the historical factors that were specific to the “Portuguese” experience, it is nevertheless my hope that this work may prove useful to scholars investigating the formation of precolonial identities elsewhere in West Africa. The richness of pre-nineteenth-century documents covering the region from the Petite Côte to Cacheu makes it possible to trace the history of Luso-African identity with a depth unmatched for any other group in West Africa. Neighboring peoples—including the Manding and Floups— also figure in this historical record, but without the detailed attention accorded the Luso-Africans. Furthermore, only the Luso-Africans left written records. Two important Portuguese sixteenth-century chroniclers, André Alvares de Almada and André Donelha, belonged to the LusoAfrican community in the Cape Verde Islands. The seventeenth-century writer Francisco de Lemos Coelho also belonged to the Luso-African trading community.6 These sources offer a “Portuguese” perspective on local culture. Pre-nineteenth-century sources offer only sporadic glimpses of identity formation and reformulation among other Senegambian peoples. I endeavor to weave these brief, sometimes anecdotal references into the more complete narrative of “Portuguese” identity. This is crucial, because “Portuguese” identity was itself negotiated or articulated in accordance with prevailing Senegambian models of group identity. The nature of LusoAfrican identity can be understood only within the context of these local social and cultural templates. In discussing the evolution of precolonial identities, one needs to consider which specific terms to employ. Do these terms imply an outsider’s definition of the group in question, or were they used by members of the group? It is clearly preferable, where possible, to use indigenous terms that best reflect the underlying concepts that were part of the local discourse on identity.7 Historical sources for Senegambia limit our knowledge of local identity terms and concepts, except in the case of the “Portuguese.” European and Cape Verdean chroniclers refer to the Luso-Africans as “Portuguese” in a context that clearly indicates this was the term used by both the LusoAfricans themselves and their Senegambian neighbors. In referring to this group as “Portuguese,” the chroniclers were conforming to local usage. These sources do not place the term Portuguese in quotation marks. I have chosen to do so in order to differentiate for the reader between LusoAfricans (“Portuguese”) and those Lusitanians who were born and lived most of their lives in Europe (Portuguese). This distinction was admittedly not made, so far as we can know, by members of the Luso-African community. Nor does it correspond to the Senegambian understanding of identity as fluid, contextually defined, and generally based on cultural pa-

Introduction

9

rameters. To have avoided the use of quotation marks would, however, have risked confusing what are today generally considered to be two separate categories: Luso-Africans and western Iberians. The discerning reader will recognize that “Portuguese” does not imply, on my part, either rejection of or skepticism toward the Luso-Africans’ understanding that who they were was Portuguese. Written sources offer a detailed and more or less continuous record of the Luso-African community and of the attendant historical evolution of “Portuguese” identity from the sixteenth to the nineteenth century. Early sources also provide information about the movement of individuals between two or more identities. These instances provide important insight not only into the process of identity change but also into the nature of identity in precolonial Senegambia and Bissau. The result is an historical tableau, unique for its detail and its depth of time, of individual and group identity creation and transformation throughout the Upper Guinea Coast in the nearly half-millennium that preceded European colonial expansion. C O N S T RU C T I N G I D E N T I T Y: A RC H I T E C T U R E The buildings that individuals and groups construct tell much about their lives, their professions, their economic position, and the image of social status they may wish to project. In Senegambia and Bissau, as elsewhere in West Africa, architectural style and building technology show responses to challenging local climate conditions and environments. The existence of similar elements in the architecture of two groups may constitute evidence of contact and borrowing; alternatively, similarities may reflect the independent evolution of parallel solutions to building in that environment. The historian is not always able to determine which of these historical processes was at work. Buildings also have a history or, rather, histories, that are more subjective. The important role architecture plays in the construction of images of a given society derives in part from the obvious fact that architecture is often the most salient feature of material culture. The manner in which members of a culture view and interpret their own buildings, or the symbolic meaning imputed to the structures, reflects the self-image of the group and of its individual members. The same buildings, viewed differently by outsiders, play a significant role in the articulation of external images of that society. Hence, a given architectural style may have multiple histories and multiple meanings. Each of these histories reflects discrete attitudes toward the culture whose members built and lived in these buildings. The history, or the sequence of histories, associated with the architecture of the Luso-Africans and their neighbors on the Upper Guinea Coast

10

“portuguese” style and luso-african identity

reflects the changing status of the Luso-Africans in West African society and economy. Part of a worldwide commercial network and a hemispheric trading diaspora in the sixteenth and seventeenth centuries, LusoAfricans were associated with “Portuguese”-style architecture, or maisons à la portugaise, variants of which were found in communities as far away as Brazil. Descriptions of Luso-African dwellings and seventeenth-century illustrations of Afro-Brazilian architecture offer graphic evidence of the movement of people—European and Euro-African traders as well as African slaves—between the Guinea Coast and Brazil. However, by the second half of the nineteenth century, French colonial observers in Senegambia had redefined “Portuguese” architecture. The buildings were now seen as evidence of both Portuguese and West African cultural decadence. This process of redefinition was part of a widespread colonial reading of African culture and history in a manner congruent with European preconceptions about the superiority of European culture (and the concomitant inferiority of African culture). Ultimately, the history of architecture is about more than the structures themselves. It is about the meanings ascribed to these structures, both by those who built them and by outsiders. In the case of precolonial architecture in southern Senegambia and Bissau, all of the groups concerned—Africans, Euro-Africans, and Europeans—viewed the buildings as symbols of who lived in them and what sort of people these inhabitants were. Architecture served both as an instrument of self-definition and as an instrument of externally imposed definition. Furthermore, architecture was both a symbolic and a subjective marker of identity. The buildings that constitute a focus of this study were, with very few exceptions, constructed of sun-dried clay or earth, generally with thatched roofs. These structures, of course, no longer exist. Many of the villages, too, have either disappeared or changed location. Some have succumbed to warfare, others to changes in the courses of the river channels beside which they were built or to the disappearance of the trade routes that gave them economic life. Perhaps some day archaeologists may study the sites of these former villages. But archaeology has barely been undertaken in southern Senegambia. Until it is, the history of local architecture before 1850 will remain a history without the buildings themselves. Precolonial architectural history is also largely lacking in visual representations of these vanished structures. To reconstruct the history of Senegambian housing forms, one is constrained to use primarily written descriptions. These descriptions exist in travel narratives, as incidental references in Portuguese records from the sixteenth and seventeenth century, and in occasional missionary records. With the exception of the Cape Verdean authors cited above, these sources are almost all written by Europeans. How reliable are they? Travelers’ descriptions of houses, mosques, and fortifications tend to be straightforward. Usually, the raw data of visual description can be separated from ethnocentric bias. The cultural bias

Introduction

11

that characterizes many eighteenth- and nineteenth-century narratives does not often lead to significant distortion in descriptions of the physical form of dwellings of local political leaders or of defensive structures or of the few monumental buildings that existed in Senegambia. Nevertheless, visual images of pre-nineteenth-century architecture from the Gambia-Bissau region are quite rare. At the end of the seventeenth century, Jean Barbot illustrated forts and other buildings, but Barbot did not have firsthand experience of this part of the coast. A small number of late-seventeenth- or early-eighteenth-century maps incorporate schematic images of houses and of village walls. Two maps in particular (see Figure 3 and Figure 4, discussed in Chapter 2) associated with the travel narrative of Michel Jajolet de la Courbe may include images that derive from personal observation. Yet here, too, the images are schematic and tend to be generic. The most accurate depictions of seventeenth-century African domestic architecture may, in fact, come not from Africa, but from Brazil. During and after his seven-year residency in Brazil (1637–1644), Dutch artist Frans Post made numerous sketches and paintings of the houses of Portuguese settlers. Close commercial ties and an active slave trade linked Brazil to the Luso-African traders of the Upper Guinea Coast. Post’s illustrations (see Figures 12–14, discussed more fully in Chapter 3) depict a few settlers’ homes that appear to incorporate elements of the “Portuguese” style, a style likely brought to the New World by Luso-African traders and African slaves. Even more significant, Post’s scenes of slave life depict dwellings that had clearly been built by Africans and that appear to reflect West African prototypes. These paintings provide indirect but detailed visual evidence of architectural forms from early- to midseventeenth-century West Africa. From the middle of the nineteenth century, several French visitors to Senegambia made sketches of the region and of some buildings. The most extensive artistic representations are those by Hyacinthe Hecquard, who traveled from the Casamance to the Futa Jalon in 1850–1851 and who depicted the trading posts of Ziguinchor and Carabane (Figure 18, Chapter 5, and Figure 21, Chapter 6). Thirty years later, the Bayol expedition from Rio Pongo to Futa Jalon was abundantly illustrated by a professional artist named Noirot. In both instances, however, the main visual record is of the Futa Jalon, an area which lies outside the focus of the present study. Only at the very end of the period considered in the present study, beginning in 1889, was southern Senegambia documented by photography. The last two decades of the nineteenth century constitute a watershed in the history of southern Senegambia. In 1886, France acquired the Casamance and its administrative post, Ziguinchor, from Portugal. The Luso-Africans who had inhabited the region since the sixteenth century lost both their political autonomy and the remnants of their formerly dominant economic role as traders. The establishment of French colonial

12

“portuguese” style and luso-african identity

administration marginalized the few remaining Luso-Africans. Drawn into a discourse on identity whose parameters were determined by Europeans, the “Portuguese” were effectively redefined as Africans. By the end of the century, virtually all that remained of their former identity was a vague memory of their reputation as master builders. Yet even this memory was subsumed by an ethnocentric discourse that disparaged African culture and essentially redefined “Portuguese”-style houses as Europeaninspired dwellings.

ONE

The Evolution of “Portuguese” Identity: Luso-Africans on the Upper Guinea Coast from the Sixteenth Century to the Early Nineteenth Century

T H E E S TA B L I S H M E N T O F E A R LY P O RT U G U E S E C O M M U N I T I E S During the late fifteenth and early sixteenth centuries, Portugal established a trading presence along the Upper Guinea Coast, that part of the Atlantic coast extending from Senegal to Sierra Leone. Portuguese ascendancy in the African trade began with fifteenth-century seaborne explorations. By the early 1600s, however, the joint Spanish-Portuguese monarchy’s financial difficulties, combined with the rise of Dutch commerce, had begun to undermine Portuguese supremacy. This process was abetted by the Dutch conquest of northeastern Brazil beginning in 1630 and by the establishment of a Dutch trading post on Gorée Island off the Senegalese coast in 1621. A generation later, the French establishment of St. Louis at the mouth of the Senegal River in 1659, followed by the capture of Gorée in 1678 marked a growing French commercial presence. Britain became an important player at about the same time. The establishment of an English garrison at James Fort in the lower Gambia River, shortly after mid-century, marked the beginning of a military and commercial rivalry in the region between the French and the British that would continue until the end of the Napoleonic Wars in 1815. Emigrants from Portugal (some of whom were Jews seeking to escape religious persecution),1 who were known as lançados,2 settled along the coast, where many of them married women from local communities. By the early sixteenth century, the offspring of these unions, Luso-Africans, or “Portuguese,” as they called themselves, were established at trading centers from the Petite Côte in Senegal to Sierra Leone in the south (Figure 1, Figure 2).3 Descendants of Portuguese emigrants, of Cape Verde Is-

14

“portuguese” style and luso-african identity

landers, and of West Africans, the Luso-Africans developed a culture that was itself a synthesis of African and European elements. Rich historical documentation allows a case study of the changing ways Luso-Africans identified themselves over the course of three centuries. Several “Portuguese” Cape Verdean merchants, including André Alvares de Almada (fl. 1590) and André Donelha (fl. 1570–1625), wrote accounts of the coastal trade; their descriptions present Luso-Africans from the perspective of the Cape Verdean elite. The earliest lançados established themselves along the coast as commercial middlemen between African and European traders4 and as coastal traders between Sierra Leone and Senegambia. In 1623, Dierck Ruiters, a Dutch merchant who had traveled to Guinea in the first decade of the century, described Luso-African trade at Cacheu: “The trade of the Portuguese in Cacheu is of two kinds, first, trade from Portugal, second coastal trade . . . mostly undertaken in small ships, pinnacles, and launches, by Portuguese who live on Santiago Island.”5 The lançados’ commercial activity was formally discouraged by the Portuguese Crown until the second decade of the seventeenth century. They nevertheless played an important role in trade between Portugal and the Cape Verde Islands. Lançado communities were permanently settled on the Petite Côte, while in Sierra Leone and the Rio Nunez of Guinea much early commerce was in the hands of the lançados who sailed there regularly from S. Domingos, in presentday Guinea-Bissau, north of the city of Bissau.6 The offspring of these lançados and African women were called filhos de terra [sons of the land] and were generally considered “Portuguese.”7 Throughout the sixteenth century, the descendents of the lançados maintained close commercial ties with the Cape Verde Islands. Many Cape Verdeans were themselves the offspring of mixed Portuguese and West African marriages. By the late sixteenth century, the island of Santiago, whose population was overwhelmingly of African origin, constituted a Creole society.8 Sharing elements of a common culture and united by marriage and economic ties, the mainland Luso-Africans and the Cape Verdeans represented a socially complex and geographically dispersed community. Cape Verdeans, like mainland Luso-Africans, resolutely maintained that they were “Portuguese.” Both populations used the same criteria, which were essentially cultural, to identify themselves as subgroups. T H E C H A R AC T E R I S T I C S O F L U S O - A F R I CA N I D E N T I T Y Throughout the sixteenth century, membership in the Luso-African community was not associated with physical features. Rather, the “Portuguese” were defined, broadly speaking, by cultural and socioeconomic characteristics.9 The first defining characteristic of “Portuguese” identity

The Evolution of “Portuguese” Identity

15

was occupation: to be “Portuguese” was to be a trader, much as to be Juula in Senegambia implied that one was a long-distance merchant.10 Another important way “Portuguese” identified themselves was through language. Initially they spoke Portuguese, but over time the language they used gradually developed into Creole, or Crioulo. Crioulo, which conjoins vocabulary derived from Portuguese with a grammatical structure derived from West Atlantic languages, is the result of a process of cultural assimilation. “Portuguese” language thus emphasized the hybrid aspect of the community and characterized the assimilative nature of the culture with which it was associated. The development of Crioulo was a long process, which historical sources do not permit us to date precisely. Regional variants may have formed at different times. Written sources clearly attest to the existence of Crioulo in Bissau, but only toward the end of the seventeenth century.11 In 1582, Francisco de Andrade, sergeant-major of Santiago, wrote that African traders on the Petite Côte spoke French and Spanish; he mentions no trading language.12 Twelve years later, “Portuguese” traders reported that Africans had mastered their language; in 1594, André Alvares de Almada observed that the Bainunks of the Cacheu region, “because of the close contact they have always had with our people, speak Portuguese very well.”13 And in 1600, Soares de Albergaria reported that many Africans along the Rio de S. Domingos near Cacheu also spoke Portuguese.14 Likewise, in 1606, the Jesuit Balthasar Barreira wrote of a Sierra Leonean king who, “having been brought up with the Portuguese, understands and speaks our language quite reasonably.”15 For extended conversations, the king could meet with Portuguese who were fluent in his language.16 That conversations were in Portuguese or in the local African language strongly suggests the absence of a common trading language, such as Creole, in Sierra Leone.17 However, it is likely that a form of Crioulo had evolved in the Petite Côte–Gambia region by the first half of the seventeenth century. In 1602, at Portudale on the Petite Côte, Dutch trader Pieter de Marees noted that local people “have their own language, a mixture of many different languages.”18 This reference strongly suggests the existence, at an early date, of a hybrid or trading language along the Petite Côte. Given the preponderance of Luso-Africans and lançados along this part of the coast, such a language would surely have incorporated Portuguese. Further south, in The Gambia in 1661, English traders relied on an African marabout who spoke Portuguese to act as their interpreter.19 By this date, Gambian Portuguese may already have been transformed into a nonstandard version of the language; this is suggested by a 1646 report from an Andalusian Capucine mission. At Juffure in the North Bank kingdom of Niumi, the missionaries reported “many Christians . . . who have never taken Confession and who speak a little Portuguese, by means of which they express themselves as best they can” (italics mine).20 This

16

“portuguese” style and luso-african identity

reference to Luso-Africans, together with de Marees’s earlier report, implies that a form of Crioulo or nonstandard Portuguese was spoken on the Petite Côte and along the lower Gambia River by the middle of the seventeenth century.21 Another characteristic of “Portuguese” identity was religion, as it was for other ethnic groups in the region. They were Catholics. Along the Upper Guinea Coast, it was common for each group to be characterized by its own religious practices. Thus, Floups (Jolas) and Bagnuns (Bainunks) had their own shrines and associated rituals, although a powerful and efficacious shrine might well be acquired by one group from another.22 Luso-African religion was actually an amalgam of Christian, Jewish, and African practices. Among the lançados settled on the Petite Côte in the sixteenth and early seventeenth centuries were Jews and “New Christians” (recently converted Jews).23 Few priests visited the coast, so Christians did not have regular access to the sacraments. Partly as a result of this isolation, some “Portuguese” participated in African religious rituals. However, unorthodox Christian practices are poorly documented. This is hardly surprising, as “Portuguese” who followed syncretic rituals would not have broadcast that fact to mission priests, the major chroniclers of the everyday life of Luso-Africans during this era. In 1601, Portugal briefly granted New Christians the right to settle in Portugal’s overseas possessions.24 Additional New Christians came to the Petite Côte until 1610, when Philip II, king of the recently united Spain and Portugal, revoked the right of former Jews to leave Portugal.25 At Portudale in 1606, Father Barreira noted “a village of 100 Portuguese who follow the laws of Moses.”26 Some of these Jews had lived in Amsterdam after fleeing the Inquisition and had established ties with Dutch merchants.27 In Senegambia, many New Christians profited from the distance from Portuguese secular and religious authority to return to their ancestral religion. They maintained discretion, however, as Portuguese reprisals in response to their commerce with the Dutch were a constant threat.28 Hence, “Portuguese” Jews may well have been content to be perceived by outsiders as Christians. The Jewish presence on the Petite Côte gradually diminished after the first quarter of the seventeenth century; but the number of “Portuguese” Christians whose religion was mixed with African rituals increased over time. References describe both Africanized “Portuguese” and Christianized Africans. In 1607, Father Barreira decried the presence in Sierra Leone of “Christian Blacks who . . . by contact with the heathen had so forgotten the obligations of our holy faith that either they possessed chinas29 themselves or they allowed their slaves to do so and they had dealings with these chinas and made them offerings.”30 There is a centuries-old tendency among non-Muslims in the Casamance and northern Guinea Bissau for each group to have its own religious shrines and rituals.31 The neighbors of the Luso-Africans clearly

The Evolution of “Portuguese” Identity

17

understood “Portuguese” religion from an African perspective of “to each his religion.” This is unwittingly indicated in a 1606 report by Father Barreira that when the Bijogos “see images of Christ or of Our Lady, they call them ‘white man’s china,’ or ‘Christians’ china.’”32 Elsewhere, Barreira observed “Portuguese” “turned wild, whose very way of life is more heathen than Christian, men who go many years without sacraments or mass.”33 These individuals, merchants who lived in coastal West African communities rather than among other Europeans, were referred to as “tangomaos.” Initially, the Luso-Africans’ religious affiliation called attention to family ties and cultural connections to Europe. Yet if Christian identity served to distinguish and differentiate its practitioners, this exclusivist aspect was counterbalanced by the religion’s claims to universality, a stance intimately tied to the missionary ideology that was an important factor in Luso-African and African relations. Ironically, missionary activity served to efface the very distinction the Christian religion created between members of the “Portuguese” community and their African neighbors. When the pitifully few Catholic priests sent to Senegambia converted members of local African communities, the cultural boundaries between Luso-Africans and their neighbors became less clearly defined. A case in point may be the Floup (Jola) village of Bolole, adjacent to S. Domingos, in presentday Guinea-Bissau. By the late seventeenth century, many villagers had Christian names; some may have converted.34 On an individual level, conversion to Christianity, an important factor in cultural assimilation, opened the possibility of further cultural transformations. In such cases, conversion was a first step in the process of becoming “Portuguese.”35 The final parameter that defined membership in the “Portuguese” communities of the Senegambia and Guinea was material culture. The LusoAfricans were identified by the distinctive architecture of their houses.36 These houses had a vestibule at the entrance or, alternatively, were surrounded by a veranda or porch, known as an alpainter.37 The dwellings were rectangular, and their exterior walls were whitened either with a wash of clay or with lime. Along the coast from Senegal to Bissau, this style came to be called architecture à la portugaise. Indeed, the sobriquet was employed in the eighteenth century for buildings as far away as the island of Réunion.38 “Portuguese”-style dwellings were ideally suited both to the climate and to the Luso-Africans’ role as commercial middlemen. The owner could receive traveling merchants in the vestibule. A clear description of the style, together with an illustration of its function, is provided by La Courbe in 1685. As he traveled south from the Gambia River, he was welcomed by a signare: “She received us most civilly in a Portuguese-style house, which is to say having walls of whitened earth and a small vestibule in front of the entry where we were seated upon mats, in the fresh air.”39 Architectural style could, of course, readily be appropriated by neighboring populations. It is significant that the sobriquet “mai-

18

“portuguese” style and luso-african identity

sons à la portugaise” does not necessarily indicate that this building style originated either in Portugal or among the Luso-Africans. Rather, the spread of architectural elements both from “Portuguese” to Africans and from Africans to “Portuguese” was characteristic of public and vernacular architecture in the Gambia-Bissau region from the sixteenth to the eighteenth century. Each of the defining characteristics of the “Portuguese” community could be shared by members of adjacent populations. The more isolated a Luso-African community became from the Cape Verde Islands and from the few centers of Luso-African population on the coast, the greater the likelihood that its members would lose their culturally defined distinctiveness. Consequently, identity transformations in both directions were undoubtedly rather common. Just as “Portuguese” and their descendents could become, for instance, Manding, so too individual Africans could become “Portuguese.” The isolation of smaller Luso-African communities increased after about 1700 as the Portuguese mercantile economy continued to weaken. It is precisely under such conditions of isolation and impending loss of cultural specificity, however, that one might expect the “Portuguese” most emphatically to assert the exclusivity of those few cultural markers they could control. T H E S I X T E E N T H C E N T U RY Luso-African communities along the Upper Guinea Coast were closely linked to the Cape Verde Islands by seaborne commerce. From the early sixteenth century, Islanders carried out extensive trade with the mainland, from the Petite Côte to Sierra Leone. In 1582, Andrade wrote that along the entire coast from Sierra Leone to S. Domingos, “There are many Portuguese who carry out trade with the Negroes of the land and send out the ships and armaments that go to the island [of Santiago].”40 At trading centers along the Rio Grande, as many as twenty or thirty ships might be seen trading for ivory, gold, and slaves from the interior.41 Andrade’s report is confirmed by the more detailed observations of the Cape Verdean trader André Donelha, whose account, written about 1625, was based upon firsthand knowledge of the coast obtained primarily in the decade before 1585.42 Donelha, a member of the mercantile elite of Santiago, was probably Cape Verdean by birth and, according to José da Silva Horta, he may have been mulatto.43 Donelha described lançados—both Christians and Jews—living among the Wolofs of the Petite Côte.44 Both Donelha and Almada present a nuanced picture of the complex relations that prevailed between the Cape Verde Islands and the mainland. This situation was characterized by two-way migration of individuals and by an active process of cultural interaction and borrowing between the various Portuguese groups and their African trading partners,

The Evolution of “Portuguese” Identity

19

as well as a continuing process of intermarriage between Cape Verdeans, Africans, and the descendents of the lançados. Two Sapes45 rulers from Sierra Leone exemplify cultural assimilation and, perhaps, ethnic transformation. Donelha recounts that these rulers had fled the Mane Invasion in about 1550. The daughter of one king became a Christian and moved to Santiago. During the 1583 famine, however, she returned to Guinea. A son of the other ruler also converted, taking the name Ventura. During the 1583 famine, Ventura also returned to the mainland near Cacheu, where he and his fellow Sapes refugees established their own settlement. Almada describes their community: “They live together in a separate section of the town, with a king . . . a Christian called Ventura de Siqueira. He can read and write, as he was brought up in Santiago Island. All the other Blacks of this settlement are Christians and he has all the babies who are born there baptized.”46 The movement of individuals back and forth, both between the physical spaces of Sierra Leone, S. Domingos, and the Cape Verdean island of Santiago, and, more significantly, between the cultural spaces of African, lançado, and Cape Verdean “Portuguese” society suggests a crucially important characteristic of mainland Luso-African society: it was not firmly bounded, nor was it exclusionist with regard to those of African origin. Rather, at the margins, Luso-African culture was open to individual assimilation. The Portuguese maintained an active commercial role in the flourishing Gambia River trade in the 1620s, in spite of growing competition from other European nations. The most important Portuguese trading center in The Gambia was Casão, located sixty leagues upstream from the ocean, on the north bank of the Gambia River. There, Donelha, writing of his own trading experience at the end of the sixteenth century, reported, “I found many well-known tangomaos.”47 The tangomaos traded with Manding merchants, “the best traders in Guinea.”48 These Muslims, or bixiiris,49 themselves sometimes traveled as far as S. Domingos and the Rio Grande.50 The free movement of individuals to Sierra Leone, Rio S. Domingos, and Casão linked the Cape Verde Islands to the coast. At Casão, Donelha found nine ships (likely from Santiago).51 In this trading town, Donelha came upon Gaspar Vaz, a tailor whom he had known in Santiago as the slave of his neighbor.52 At the time they met in Casão, Vaz was wearing Muslim robes. Vaz, observing Donelha’s discomfort, took him aside and, lifting his robe, disclosed a crucifix around his neck. He then assured the Cape Verdean that once he became sandegil, or chief of the local trading community, he would reassert his underlying Christian faith. Thereupon, Donelha and he became trading partners. Vaz may in fact have understood his Cape Verdean counterpart better than Donelha understood him. The two men quickly established a close commercial relationship. Vaz served as Donelha’s host and facilitated his guest’s commercial ventures.53 He helped Donelha purchase merchandise at the prevailing price for ex-

20

“portuguese” style and luso-african identity

changes between Manding traders rather than at the significantly higher prices they charged the tangomaos.54 It is not evident from Donelha’s account precisely how Vaz had come to be a slave in Santiago or how he reestablished his earlier, elevated social status in The Gambia. For in addition to being a tailor and a merchant, Vaz claimed to be the nephew of the sandegil (satigi) of Casão (a term Donelha translates as “duke”).55 It is clear that Gaspar Vaz—like Ventura, the son of the exiled Sierra Leonean king—was able to move back and forth between Cape Verdean “Portuguese” and African societies. When he returned to The Gambia, he brought with him professional skills that he had refined in Santiago. He spoke Portuguese, which enabled him to serve as interpreter for Donelha and, presumably, for other visiting merchants. Vaz, like Ventura, was able to navigate in the social and economic spaces of two cultures.56 In the process, these men undoubtedly served as conduits for the transmission of cultural traits between “Portuguese” and African society. In early-seventeenth-century Sierra Leone, too, European traders relied on individuals who, by their economic and cultural position, served as a bridge between European, “Portuguese,” and African society. Dierck Ruiters mentions two such traders, one named Mathew Fernandos and the other Francisco Mendes.57 Even contemporary descriptions do not always clearly indicate whether such men as these were Luso-African or African. Indeed, the distinction was probably blurred. Material culture in Casão reflected extensive intercultural contact. The remarkable architecture there attracted Donelha’s attention. Most houses were round and made of whitewashed adobe bricks, although a few were rectangular.58 In a subsequent passage Donelha describes the courthouse, or tribunal, which, in addition to having a two-story elevation and whitewashed brick walls, was rectangular. It is noteworthy that in late-sixteenth-century Casão, the tribunal and the houses associated with the sandegil—the preeminent expressions, respectively, of public and private architectural space in this Manding community—incorporated the elements that Europeans would later associate with Luso-African vernacular architecture. Some elements of the “Portuguese” style probably developed independently among both West African and European builders as logical responses to the challenge of constructing houses that would provide a cool interior in a hot climate. But Donelha’s description of Casão suggests that Mande public architecture served as an important source of inspiration for the development of the “Portuguese” style of domestic architecture. Gaspar Vaz personifies the fluidity of ethnic identity on the Upper Guinea Coast in the sixteenth century. He traded with Luso-African merchants. He had lived in the Cape Verde Islands and he spoke Portuguese; in addition, as his name indicates, he was a Christian (at least when not in Casão). As a (sometime) Christian in Mande society, his position was

The Evolution of “Portuguese” Identity

21

highly unusual.59 One has the impression that Vaz switched identities almost as readily as he switched robes. He and others of comparable stature who likewise lived in both African and Portuguese society served as conduits through whom elements of material culture were exchanged between the two societies.60 In The Gambia and Bissau in the late sixteenth and seventeenth centuries, architectural style, like religion and language, reflected the extensive contact that linked the populations of the region. AN INDIGENOUS MODEL O F I D E N T I T Y F O R M AT I O N During the sixteenth and seventeenth centuries, Luso-African identity was consistent with the historically fluid and contextually defined manner in which ethnic identity was established throughout the Gambia-Bissau region before the colonial period.61 The history of ethnic identities in West Africa has attracted significant scholarly attention in recent years. JeanLoup Amselle’s seminal 1990 study, Logiques métisses: Anthropologie de l’Identité en Afrique et ailleurs, articulated the position that fixed ethnic groups or categories—les ethnies—were the invention of colonial administrators and ethnologists.62 The effect of colonial officials’ work was to rigidify and thereby transform preexisting identities.63 Often, as in the case of the “Bamana,” a purely descriptive toponym was transformed into a race, or an ethnic group.64 More recently, historians have turned their attention to the history and the nature of identities elsewhere in the savanna and in the forestland of the Upper Guinea Coast. These studies provide mounting evidence that throughout this geographical space, precolonial identities were complex and fluid. In their introduction to a series of articles in Mande Studies, Mirjam De Bruijn and Han van Dijk argue that “from the perspective of mobility . . . ethnicity becomes something fluid, a creation arising out of the interaction with others and thus something without clear borders.”65 Allen Howard, in an accompanying article, articulates the position that “ethnicity is situational and relational.”66 Drawing on his own work with Mande and Fulbe identities in Sierra Leone, he concludes that “Mande and Fulbe have repeatedly differentiated and merged in various ways in particular contexts.”67 Individuals of Fulbe and other backgrounds could become Mande through intermarriage and by becoming traders.68 In his contribution to the same volume of Mande Studies, Martin Klein cogently argues that “the savannah zone of West Africa is marked by intermixture of ethnic groups.”69 Identity changes were consummated by the adoption of new names and a new genealogy. Klein enumerates multiple instances where Serers and Wolofs interacted, eventually assuming shared identities. In Wolof and Serer society, he argues, “movement was easier between states than across the rather rigid lines between castes

22

“portuguese” style and luso-african identity

and orders.”70 The historical fluidity of identity leads Klein to ask, “Was Saluum a Wolof or a Serer state?”71 The region of Cacheu, too, as Eve Crowley demonstrates, was characterized by a complex mixture of cultural and linguistic traits. There, “an unusually fluid social structure . . . facilitated the incorporation of temporary and permanent guests.”72 Crowley’s observations about Cacheu are supported by Walter Hawthorne’s work among the neighboring Balantas. He observes that some Balantas intermarried with Manding neighbors to create a new identity, one “neither Balanta nor Manding.”73 Writing about nineteenth-century Guinea-Bissau as a whole, Joshua Forrest observes “the relatively malleable, multiple-identity character of ethnic group formation.”74 Throughout the Senegambia and Cacheu region of Guinea, as the work of these scholars shows, movement across ethnic frontiers was a widespread phenomenon. The resultant incorporation of individuals from different linguistic and cultural backgrounds was characteristic both of local communities and of states. Among the African populations of the Gambia-Casamance region, even today, oral traditions reflect a long and complex history of migrations and of identity transformations. In this matrix of local identity, identities have historically been fluid and individuals have often held multiple identities. In sixteenth- and seventeenth-century Senegambia, the permeable nature of “Portuguese” identity reflected the fact that many of the lançados and their descendents had married local African women. Luso-Africans maintained an understanding of group membership based essentially on cultural characteristics, and this facilitated the assimilation into “Portuguese” society both of filhos da terra and, as is suggested by the lives of people like Vaz and Ventura, of individual creolized African men and women. I argue that sixteenth-century “Portuguese” and other Senegambian peoples illustrate a process of identity formation that does not accord with the standard bipolar (or “us/them”) model. I find remarkably few indications that sixteenth-century “Portuguese” had recourse to a sense of boundaries in establishing and maintaining their community or that they discouraged assimilation into their society. Implicit contrasts certainly existed (free men versus slaves; Christians versus Muslims). Yet even here the boundaries were fluid (Vaz was evidently both slave and noble, Christian and Muslim). Boundaries were not fixed indicators of the “otherness” of neighboring populations. In this respect, Luso-Africans appear to represent a model of identity formation that one might term a nonstandard model. Luso-African identity construction stands in contrast to Barth’s bipolar model of identity formation. Where Barth, focusing on “the boundaries that define groups,”75 rightly observes that boundaries persist in spite of the movement of individuals across them, he overstates—or over-

The Evolution of “Portuguese” Identity

23

schematizes—the distinction between group members and others when he writes of “a dichotomization of others as strangers.”76 Not all boundaries are alike; some are more permeable than others. One is led if not to deny that coastal Luso-Africans conceptualized “otherness” in the construction of their own sense of being “Portuguese,” then at least to suggest that this sense of the “other” played a relatively circumscribed role in creating their images of who they were. As the growing body of ethnographic and historical literature on the Upper Guinea Coast demonstrates, Luso-Africans constructed their identity or identities in a manner that conformed with Senegambian culture. This model of identity formation—flexible, malleable, and based on cultural and socioeconomic factors—is characteristic of the manner in which local societies along the Upper Guinea Coast perceived themselves. To posit a simple bipolar model is to overlook the fact that the “Portuguese” were closely related culturally and physically to their African neighbors— at times they were virtually indistinguishable from them—and that individuals could move back and forth between Luso-African and African society. A bipolar model presumes a later, essentially Western, approach to identity, one that is not appropriate for the sixteenth- and early-seventeenth-century Upper Guinea Coast. In this respect, seventeenth-century Senegambian society was similar to Creole society in the New World. Writing about both the French Caribbean islands and Chesapeake society in the first half of the seventeenth century, Ira Berlin notes “the fluidity of colonial society, the ill-defined meaning of slavery, and the ambiguous notions of race.”77 Caribbean Creole communities, where “the lines between free and slave, black and white were porous,”78 did indeed, as Berlin observes, follow a West African model of cultural identity. Berlin further observes that Black people in the Chesapeake did not themselves begin to emphasize their separateness from “those who would deny their birthright” until the imposition of rigid social and racial categories.79 Here, too, one may observe a parallel to the situation of the Luso-Africans in Senegambia, who only began to define themselves in terms of what they were not as a response to the imposition of rigid identity categories. T H E L AT E S E V E N T E E N T H A N D EIGHTEENTH CENTURIES: T H E “ P O RT U G U E S E ” R E D E F I N E D During the late seventeenth and eighteenth centuries, three significant historical developments had a profound impact on Luso-African communities, affecting not only their economic livelihood but also the manner in which they defined their individual and group identity. The first factor was the rapid expansion of commercial competition by Dutch, French, and English trading companies. Portuguese commercial ascendancy along

24

“portuguese” style and luso-african identity

the Petite Côte and the Lower Gambia was only a memory by the time the French captured Senegal’s Gorée Island from the Dutch in 1677. In The Gambia, English commercial dominance may be roughly dated to the capture of James Island in 1661. Second, these Europeans, several of whom left written accounts, brought with them a conception of identity that contrasted with the fluid, culturally based sense of identity characteristic of the “Portuguese” and their West African neighbors. Finally, outside of Bissau, Cacheu, and The Gambia, Luso-Africans, deprived of their economic base and increasingly isolated from the Cape Verde Islands, began to lose the cultural traits that had been the foundation of their identity. These developments had several consequences: many smaller, isolated “Portuguese” trading settlements south of Bissau and along the Petite Côte disappeared as their inhabitants either left or were assimilated into neighboring African cultures. The more remote Luso-African communities that survived seem to have gradually altered the manner in which they defined their identity in order to stress their distinctiveness from their African neighbors. Along the Petite Côte, for example, religion took on an increasingly important role as a marker of cultural difference. Finally, the “Portuguese” in The Gambia and Bissau were brought into the European discourse on identity. “Portuguese” responses to this challenge are not directly recorded by Luso-African chroniclers; the last important Cape Verdean source is Lemos Coelho, dated circa 1684.80 One can nevertheless partially reconstruct their side of the confrontation between contrasting discourses on identity by means of careful comparison and interpretation of European sources. A distinction between European and Luso-African parameters of identity is already evident in some early-seventeenth-century sources. In 1620 in The Gambia, English traveler Richard Jobson observed another people . . . as they call themselves, Portingales, and some few of them seem the same; others of them are Molatoes . . . but the most part as black as the natural inhabitants. . . . They do generally employ themselves in buying . . . commodities . . . still reserving carefully the use of the Portingall tongue and with a kind of affectionate zeal, the name of Christian, taking it a great disdain, be they never [sic] so black, to be called a Negro.81

For Jobson, the salient characteristic of the Luso-Africans was their skin color, which did not coincide with his own conception of what European Portuguese should look like. At the same time, his rather indulgent tone— “affectionate zeal”—betrays no strongly negative reaction to their selfidentification. In fact, Jobson cites three criteria widely accepted by the “Portuguese” themselves as defining their Luso-African identity: their profession, their language, and their religion. This suggests that the Englishman may have been aware of the importance of these parameters in

The Evolution of “Portuguese” Identity

25

defining who was “Portuguese” and even who was “white.” As for the “Portuguese,” the refusal to accept the label “Negro” or “Black” would remain central to their sense of identity for over 200 years. In 1685, La Courbe expressed the conviction that a people who were not white could not, a priori, be Portuguese. He described the LusoAfricans in The Gambia as “certains nègres et mulâtres qui se disent Portugais, par ce qu’ils sont issus de quelques Portugais qui y ont habité autrefois” [certain Negroes and mulattoes who call themselves Portuguese because they are descended from some Portuguese who formerly lived there].82 The Frenchman shows a more judgmental attitude (“who call themselves Portuguese”) than did Jobson. Significantly more pejorative was Jacques Le Maire’s 1695 definition of the Luso-Africans of the Petite Côte: a species of Portuguese, people who refer to themselves this way because they used to serve, and are descended from, those who first lived along this coast. . . . From the Negresses whom they married, were born these mulattoes, from whom in turn came even darker ones.83

Dark skin is here associated with lower status; it reflects, in Le Maire’s view, a presumed heritage of servitude. The association of skin color with social inferiority undoubtedly reflects the growing influence of the Atlantic slave trade, which in turn influenced European attitudes about race.84 Eighteenth-century European descriptions of Luso-Africans betray a progressively more negative judgment, frequently tinged with irony or sarcasm. The English factor in The Gambia, Francis Moore, whose extensive contact with Luso-Africans during his 1732 stay seems not to have engendered much respect for them, described the “Portuguese” as speaking “a sort of bastard Portuguese language.”85 Moore refers presumably to a form of Creole, but in a manner that conceptualizes a cultural equivalent to the physical mixture of Portuguese and Manding, whose result was that “they are now very near as Black as [the Africans].”86 His ironical view of the Luso-Africans is expressed bluntly: “They reckon themselves still as well as if they were actually white.”87 To Moore, light skin color confers elevated standing. In his construction, complexion serves—and this is significant—as the primary marker of identity. Moore’s words were cited almost verbatim by Thomas Astley in a 1745 compilation of travel narratives. Astley’s compendium, which also cites Jobson and Labat (who plagiarized La Courbe), reflects the spread in England of a popular image of Luso-Africans. This image clearly contests both their identity and the criteria on which the “Portuguese” based their sense of self: “Nothing angers them more than to be called Negroes. This proceeds from their not understanding the true meaning of the word, which they use only for slaves.”88 A generation later, John Mathews, traveling among the isolated descendents of Luso-African traders in the Rio Pongo, voiced an even more sarcastic and pejorative image of the “Portuguese”: “The principal

26

“portuguese” style and luso-african identity

people call themselves Portuguese . . . though they do not retain the smallest trace of European extraction; but having had a white man once in the family is sufficient to give them the appellation. They also profess the Roman Catholic religion . . . yet the most enlightened among them are merely nominal Christians.”89 Not only is Mathews’s conception of identity based on physical characteristics, but these traits determine status in a hierarchical order in which to be European is to be superior. The connection Moore implied between physical appearance and level of cultural development is here made explicit. Furthermore, religion, specifically the Luso-Africans’ departure from European Christian norms, becomes a measure of “Portuguese” decadence, commensurate with the darkening of their complexion over the generations. The origins of the Luso-Africans’ use of “white” and “black” may be traced to the sixteenth century.90 Almada uses the term “black” to refer to Africans and “white” to refer to “Portuguese” and Cape Verdeans. His use of the terms likely reflects attitudes about color prevalent in late-sixteenthcentury Portugal.91 At the same time, however, Almada’s terminology does not follow a simple binary oppositional model, and it transforms the preexisting Iberian terminology. Already in the sixteenth century, Cape Verdean society was widely intermarried; Almada himself was apparently a mulatto.92 Nevertheless, he places himself in the category “white.” That his family belonged to the Island elite and that they owned slaves suggests that “white” and “black” referred not to skin color but, at least in part, to social class. A sense of the Luso-Africans’ early use of color terminology can also be inferred from Jobson’s observation that they objected, “be they never so black,” to being called “Negroes.”93 To the “Portuguese,” many of whom were professional slave traders, “Negro” meant “slave.” As social status remained relatively fluid in southern Senegambia in the seventeenth century, the distinction between free and slave did not imply the adoption of a rigid system of identity categories by the “Portuguese.” A century after Jobson, Francis Moore’s 1732 observation confirms the association of social status with the terms “white” and “black.” And in 1818, at the Luso-African trading community of Geba in present-day Guinea-Bissau, Gaspard Mollien observed “blacks and mulattoes who are nevertheless called white, because all who are free claim this title.”94 From the late sixteenth to the nineteenth century, one perceives a consistent system of ascribed meaning, whereby “white” implies both social status (slave trader, free man) and references to cultural and blood ties to Portugal.95 In this system of meanings, there exist similarities to the language of identity of the Maures of southern Mauritania. There, as James Webb has demonstrated, the term “white”—bidan—implied not skin color but rather cultural identity and social status; they were free.96 Like the “Por-

The Evolution of “Portuguese” Identity

27

tuguese,” the Maures had a diverse heritage, one that incorporated Wolofs and other black Africans. Some Maures, like the Luso-Africans, were traders, although others were herders or warriors. It is perhaps significant that bidan identity developed in the late seventeenth century, not long after “Portuguese” identity. These may be two manifestations of a single identity construct that encompassed long-distance commercial networks from the southern Sahara to the Upper Guinea Coast. Webb’s characterization of “white” and “black” as “regional constructs which refer to the cultural identities . . . rather than to skin color”97 applies to the LusoAfricans as well as to the bidan. L AT E R “ P O RT U G U E S E ” R E L I G I O N John Mathews’s reference to “merely nominal Christians” elicits the question, What was the nature of Luso-African religion in remote areas at the end of the eighteenth century? The syncretic nature of nineteenth-century Luso-African religious rituals in Guinea-Bissau is the subject of George Brooks’s case study of All Souls’ Day.98 Analogous syncretism in day-to-day practices may be observed in late-eighteenth-century Sierra Leone, where Swedish naturalist Adam Afzelius described a characteristic instance of religious syncretism in 1795. Near Freetown, Afzelius witnessed the poison ordeal of a man who had been accused of witchcraft. The principal accuser was an elderly merchant named Domingos. Domingos had a vested interest in the outcome of this trial, for those individuals found guilty were sold into slavery and he, not coincidentally, was the main slave trader in the region. The old man was fluent in Portuguese, and he prayed every day using a Portuguese missal.99 In addition, he always wore a string of prayer beads given to him by a Portuguese priest, and “he [had] expressed great concern that for some years past he had seen no priest. . . . He has left orders that as soon as he dies, two slaves shall be sent to Santiago to a priest there, who may intercede for him.”100 Domingos was one of the few remaining Luso-Africans along this part of the coast. During the poison ordeal, some local Temnes placed their gris-gris on the ground as a public invocation in order to ensure a guilty verdict. Then Domingos, “not being satisfied with the Timany Grisgris, or thinking them not powerful enough, let his own Portuguese Grisgris be brought forward and put near to the former.”101 Notwithstanding this intercession and the prayers of Domingos, the accused survived his ordeal and was found innocent. This incident illustrates several points about the Christianity of Domingos. His use of the prayer beads demonstrates a local West African ritual emphasis on instrumentality: the gris-gris served the specific function of achieving desired ends. The fact that Domingos was actively involved in witchcraft trials also strongly suggests that he had assimilated

28

“portuguese” style and luso-african identity

an important aspect of local metaphysical concepts concerning the interrelationship of the physical and spiritual realms. The diminished ritual and metaphysical distance separating Domingos from his Temne neighbors did not, however, deter him from using his Christianity as an important identity marker. His religion evoked ties (religious, economic, and cultural) to the Cape Verde Islands, where a priest would pray for his soul after his demise. T H E “ P O RT U G U E S E ” R E S P O N S E : IDENTITY DEFENDED ON THE PETITE CÔTE Eighteenth-century Europeans articulated a view of “Portuguese” identity that was based on a priori characteristics, primarily skin color, rather than upon cultural traits. Luso-Africans were forced to respond to this ascribed identity at a moment when their increasing isolation was eroding the cultural markers upon which their own sense of identity had been based. In response to the intrusion of an a priori identity model, they began to articulate a more assertive self-identification; consequently, an important change ensued in their self-conceptualization. Precisely because the dominant European model denied their distinctiveness from other African populations, the “Portuguese” came increasingly to define themselves by reference to who or what they were not. An illustration of this process may be seen among the Luso-Africans of the Petite Côte. There, 200 years of Dutch and French commercial dominance had largely isolated the remaining “Portuguese.” By the nineteenth century, the Jewish community had disappeared, and the remnants of the “Portuguese” at Joal in The Gambia based their identity upon an oppositional model. They distinguished themselves from their Serer (Sereer) neighbors on the basis of religion and “color.” These “Portuguese” no longer spoke either Portuguese or Creole,102 yet they continued to give their children Portuguese names and they resolutely continued to call themselves Christians.103 And as the Abbé Boilat explained, they steadfastly maintained that they were white: Although they are as black as the purest black Africans, they make the modest claim to be pure whites, and it is a great insult to consider them Negroes or Serers. They want to be called the whites of Joal, the Christians of Joal, because they are the direct descendents of the Portuguese. . . . Succinctly stated, to be Christian is to be white; to be white is to be free.104

For these Luso-Africans, to call themselves “blanc” was specifically to differentiate themselves from other local populations. To be white was to be nonslave, recalling that their ancestors had been slave traders. Although their distinctiveness no longer resided in a unique Luso-African culture, the loss of cultural and physical markers of separation did not diminish

The Evolution of “Portuguese” Identity

29

their sense that they constituted a discrete group. Their dual identity— “être chrétien, c’est être blanc”—implicitly required the existence of its opposite number: non-Christian and black, a classical bipolar model of identity formation. The formation in Joal of an exclusivist “Portuguese” identity differed, however, from the later process whereby, under the impetus of colonial administrators and ethnologists, “ethnic” boundaries were established and subsequently accepted by the Africans themselves.105 The redefinition of Luso-African identity took place essentially before the colonial period. Furthermore, the “Portuguese” had borne the brunt of the precolonial European discourse on identity, which largely spared other Senegambian peoples. The reason for this is clear: only the “Portuguese,” whose very existence was testimony to cultural and physical assimilation between Africans and Europeans, posed an ontological challenge to European identity which was based on the premise of a non-white, non-European “other” by the time of the Enlightenment. Groups that do not fit comfortably into any single category, or those whose members may lay claim to one category while the outside society assigns them to a different one, are a focal point for particularly intense contestation. The existence of such ambiguous or liminal groups may call into question the entire categorical structure. To apply the label “Portuguese” to the dark-skinned descendents of the Luso-Africans would have had direct implications for the broader, socially constructed category into which the group was to be situated. By denying the Luso-Africans’ selfdefinition, eighteenth- and early-nineteenth-century observers were defending what subsequently evolved into a racially based definition of peoples and cultures. In order to maintain their sense that they constituted a discrete group, the “Portuguese” were forced to adopt a model that stressed their differences from their African neighbors. With this historical development, religion became the crucial parameter by which Luso-African communities defined themselves. Whereas in the sixteenth century, missionary activity had stressed the universality of the Christian faith, by the early nineteenth century, Christianity had become, for the “Portuguese,” both an exclusivist religion and a marker of the boundary between themselves and other Africans. Amanda Sackur observes that “the creole identity . . . required definition against two competing ‘others’: Europeans and Africans.”106 I would argue that until the eighteenth century, this was not the case. However, for the nineteenth century, Sackur is quite correct. The need to redefine themselves in a manner that emphasized their distance from other African societies led the “Portuguese” to an exclusivist approach to their religion. In 1846, when the Abbé Boilat visited the fishing village of Joal, he found that the villagers defined themselves as Christians, although local religious practices were now an amalgam of Christian prayer (the two elders who led communal prayers knew only the “Ave Maria”)107 and local

30

“portuguese” style and luso-african identity

rituals. One powerful shrine, called baton, attracted clientele from SaintLouis, the Isle of Gorée, and The Gambia, while the spirit of another powerful shrine, named Mamanguey, had formerly appeared in broad daylight in the guise of a uniformed naval officer.108 Despite their Africanized religious practices, the people of Joal were distinguished by their religion from the surrounding populations. NonChristian neighbors who wished to be baptized and whose parents were either Muslims or “idolâtres” had to pay the price of one iron bar.109 The distinction between Christian and non-Christian was religiously maintained. In Joal, no parent would permit their offspring to marry a nonChristian: In practice, none of the inhabitants of Joal would ever . . . allow their children to marry the children of the neighboring village, because the latter are fetishists [sic]. They would never permit a Muslim or a fetishist to be buried on their land.110

Ultimately, it was by means of the creation of ritual prohibitions— against marriage with non-Christians and against the burial of outsiders in the local cemetery—that the people of Joal were able to give concrete expression to their distinctive identity. Christianity was thus central to that identity, which was based upon an oppositional relation between Christians and everybody else, including both Muslims and “idolâtres.” Boilat, himself a missionary, also observed that this attitude led the people of Joal to oppose missionary activity among their non-Christian Serer neighbors.111 The Christians of Joal thus exhibited an exclusivist approach both to religion and to identity, an approach diametrically opposed to that of the early Luso-Africans. L U S O - A F R I CA N S O N T H E E V E O F THE COLONIAL ERA A somewhat different attitude toward religion prevailed among the nineteenth-century “Portuguese” of the Casamance region, just south of The Gambia. Like the inhabitants of Joal, the “Portuguese” of Ziguinchor, a trading center on the south bank of the Casamance River, asserted a Christian identity. As with the “Portuguese” of Joal, the inhabitants of Ziguinchor had assimilated elements of local religious practice. The emphasis placed by their Floup neighbors on instrumentality, on the use of protective charms (called “bubenben” or “bufókub” in the Jola language),112 found a close parallel in the “Portuguese” use of Christian medals: “All the inhabitants call themselves Christians, although they fulfill none of the obligations of this religion. . . . They deeply venerate images, medallions, and [medals of] Christ, to which they ascribe the power to protect them from all accidents.”113

The Evolution of “Portuguese” Identity

31

Unlike the inhabitants of Joal, however, the “Portuguese” of Ziguinchor maintained religious interaction with their non-Christian Floup neighbors. In the early nineteenth century, the “Portuguese” sold some of these Christian charms to the Floups.114 This commerce is reminiscent of the commercial activities of Muslim marabouts. Since the Floups, too, had faith in the power of Christian medals, the Portuguese carried out a thriving trade, exchanging crucifixes for slaves. The “Portuguese” community of Ziguinchor, more numerous and less isolated from other “Portuguese” than the Luso-African settlement of Joal, could maintain such interaction with their neighbors without losing their cultural identity. It is significant that although numerous eighteenth- and nineteenthcentury French and English narrators questioned “Portuguese” identity, members of local African communities continued to accept the LusoAfricans’ self-identification. In their interactions with African traders, nineteenth-century Luso-Africans continued to be identified by profession, religion, and, to a lesser degree, by material culture. Muslim merchants from as far away as the Mande heartland in present-day Mali understood that their “Portuguese” counterparts maintained their own religion.115 To Muslim traders, religion was a primary marker of “Portuguese” identity. Just as to be a Juula was to be a professional trader and a Muslim, so to be “Portuguese” was to be a trader and a Christian. Mungo Park reported that he met a wealthy salt trader along the Niger “who had traveled to Rio Grande and spoke very highly of the Christians.”116 Although Europeans were often skeptical of the Luso-Africans’ religious affiliation, the Muslims with whom they did business identified the “Portuguese” primarily as Christians. To some African trading partners, material culture, too, appears to have remained one of the markers of “Portuguese” identity. In 1850, the French explorer Hyacinthe Hecquard described a “Portuguese” Christian who was a respected merchant and architect in the Futa Jalon.117 The man, named Wolli, had constructed the interior of the mosque at Timbo.118 That this “Portuguese” Christian, who reportedly had lived for many years in the Futa, was entrusted with such a project reflects the respect with which he was regarded by his Muslim peers.119 Eighteenth- and nineteenth-century African trading partners continued to view the Luso-Africans according to long-established parameters of identity that were common to both the Upper Guinea Coast and the Mande diaspora. Historical sources for Senegambia and northern Guinea-Bissau afford the unique opportunity to trace 250 years of changing discourse on “Portuguese” identity. Challenged by Europeans’ denial of their identity, eighteenth- and nineteenth-century Luso-Africans were forced to redefine themselves. Contact with a model of identity based on a priori physical characteristics led the Luso-Africans to move from a local discourse

32

“portuguese” style and luso-african identity

marked by flexible and inclusive identities to a more rigid paradigm based on an oppositional and exclusivist understanding of identity. This historical process constitutes the earliest documented confrontation between West African and European models of identity. It also serves as prologue to the widespread imposition of more rigid ethnic categories during the colonial period.

TWO

Sixteenth- and Seventeenth-Century Architecture in the Gambia-Geba Region and the Articulation of Luso-African Ethnicity

One result of the establishment of Portuguese and Luso-African trading communities along the northern Upper Guinea Coast in the fifteenth and early sixteenth centuries was the development of a distinctive style of domestic architecture. These houses were suited to the climate and they used locally available building materials. The history of Luso-African architecture raises several related questions that are highly significant for the history of cultural interaction between Europeans and Africans. These questions include: What were the respective roles of Africans, Europeans, and Luso-Africans in the development of a distinctive domestic architecture? Is it possible to discern the influence of evolving Luso-African construction on local African architecture or of local building styles on AfroEuropean construction? In other words, to what extent does architecture reflect mutual two-way interaction between European and African societies? Not surprisingly, European written descriptions of the Gambia-Casamance-Geba region focus on commercial centers and trade. Villages that contained communities of lançado or Luso-African traders are disproportionately well represented in accounts written by European and Cape Verdean chroniclers. Discussions of buildings and living space are clearly subordinate to concerns about trade in these sources. The one architectural idiom that is most copiously documented, however, is the form of housing that came to be known as the “Portuguese” style, or maisons à la portugaise. The abundant references to “Portuguese”-style dwellings document interaction between local construction techniques and building forms and materials and techniques brought to West Africa from Europe. Nevertheless, several factors complicate the historian’s task. First, with the excep-

34

“portuguese” style and luso-african identity

tion of buildings in the port of Cacheu, none of the seventeenth-century physical structures survive. Second, contemporary written descriptions of local African buildings are almost all cursory, and they are widely separated both geographically and chronologically. In a region of considerable cultural diversity, one cannot necessarily draw conclusions about construction in one locale on the basis of architecture in another community. Thus, any attempt to describe the early evolution of building styles outside of the trading communities will necessarily be hypothetical. With some exceptions, Luso-African architecture is much more fully documented than local African buildings. Manding architecture in Casão on the Gambia River constitutes one important exception; it is described in detail by André Donelha. In addition, the layout of villages and types of buildings are relatively well documented from the sixteenth century for the people known as Floups, who lived near Rio San Domingos (Cacheu), and, from the late seventeenth century, for the Bagnuns (Bainunks) and northern Floups of Vintang and Fogny (north of the Casamance River). This clustering of sources is significant, as the societies that are relatively well documented are culturally and linguistically related to one another. The Floups of S. Domingos—also referred to as Felupes in Portuguese sources—spoke a form of the Jola language (also Diola) of the Bak subgroup of West Atlantic languages. They were closely related to the Floups of Vintang-Fogny—Floops in English records—referred to by La Courbe and Labat.1 The northern Floups, in turn, have a long history of cultural interaction with autochthonous Bagnuns and with the Manding people of The Gambia. From north to south, the Gambia-Cacheu region may be divided into Gambia River, Vintang Creek, and Fogny, roughly bounded to the east by the Soungrougrou River (or Sangrédegou), the Casamance River, the Kasa region, and Rio San Domingos. In the sixteenth and seventeenth centuries, the Floups were renowned for their refusal to participate in trade with Europeans. One may ask why they, more than other local populations, attracted the attention of foreigners whose primary interest lay in trade. One obvious answer is that Floup architecture so impressed the visitors that those who saw it were moved to write about it. Another possible explanation is that not all Floup groups deserved their reputation for avoiding commercial contact with Europeans. Floup (or Jola) involvement in trade is evidenced, for example, by extensive commercial contact between the nineteenth-century people of Thionk-Essyl, north of the Casamance River and west of Fogny, and the French at Carabane.2 Bertrand-Bocandé (1849) says of the Floups who lived north of Cacheu and S. Domingos, “Almost all of the inhabitants speak Portuguese Creole very well.” That they spoke the local trading language strongly suggests that they were involved in trade.3 The earliest published description of architecture in the Rio Geba– Casamance area is found in Valentim Fernandes’s account from about

16th- and 17th-Century Architecture in Gambia-Geba

35

1506. Although Fernandes, a German living in Portugal, never traveled to Africa, he had informants who were familiar with the Gambia-Geba region. His description of the Casamance River “where a lot of trade is carried out” and where a mélange of peoples including Balantas and Floups lived under a Manding ruler at whose court there were many Portuguese traders,4 gives cursory mention of “the house of the king.” He then writes of the Mansa Falup, the king of the Floups, who lived between the Casamance River and Rio San Domingos (the future site of Cacheu). These Floups were bellicose warriors. The king’s habitation was surrounded by hedges made of stakes in the manner of a barbacane and there are five of these hedges. The inhabitants have to enter in a circle and not directly; in the center of these palisades there is a fortified area and here he lives, and this house is surrounded by an arm of the river and it is very strong.5

The defensive strength of this royal compound, surrounded by five concentric circles of stakes, each with a single entrance facing in a different direction, and with a heavily fortified central enclosure that was also protected by a tidal waterway, is consistent with the warlike image sixteenthcentury chroniclers present of Floup society. The early-sixteenth-century date of this description has implications for the origins of such defensive structures. Significantly, Fernandes’s informants were describing the Gambia-Geba region during the first forty or fifty years of contact with Europeans. While it is possible that the concentric palisades were a recent innovation, meant to protect against slave raids, the Atlantic slave trade had only just begun and was still relatively small in scale. The complexity of the fortifications suggests that they were the result of a long-term process of development. Very likely, the existence of these architectural defenses reflected military threats or instability predating the late-fifteenth-century inception of the Portuguese slave trade. The image emerges of a political situation characterized by the absence of centralized authority and by instability or warfare. In this respect, the southern Floups, with their heavily fortified, compact village settlements, probably differed from their southern neighbors, the Balantes, who adopted similar compactly defended villages partly as a response to the slave trade.6 Similar defensive elements are referred to in much later descriptions of Floup architecture. One is left to guess at the form of the central courtyard and building. The fact that the structure is described as “very strong” makes it likely that the material used in construction was sun-dried clay soil rather than wattle or vegetal fiber, the only other commonly used materials. This is Fernandes’s most detailed description of any local community. He does mention that several villages between the Gambia and the Casamance Rivers are surrounded by palisades.7 These communities, whose populations he gives as five to ten thousand—a gross exaggeration, but

36

“portuguese” style and luso-african identity

characteristic of Mande (and Mande-influenced) peoples in the use of round numbers such as 10,000 to indicate a very large number8—each had their own ruler. Furthermore, Fernandes writes that the compound of the Mandi Mansa, a Manding ruler in the Lower Gambia, was surrounded by six rings of stockades, with the portal of the inner palisade guarded by archers.9 Fernandes’s account also contains what may be the earliest description of the style later referred to as Luso-African architecture. While the specific reference is to Sierra Leone, the building style is common to Portuguese lançados and Afro-Portuguese traders in the Gambia-Geba region. The houses of the rich, Fernandes writes, “are made of adobes and bricks; in the interior they are whitewashed; on the exterior covered with chalk or kaolin. The interior is well finished and these are the best houses in all of Guinea.”10 The use of mud or mud bricks covered on the exterior with chalk or white clay—or perhaps, if Fernandes misunderstood his informant, with lime—recalls the white-limed outer walls of Portuguese forts and trading castles from Elmina (Gold Coast) to Mombasa (East Africa). Elmina, the earliest of these fortresses, dates to 1482, while the East African bastions are roughly contemporary with Fernandes’s account. Portuguese military architecture may have influenced Luso-African housing construction. However, the “Portuguese” were likely also influenced by local African practices. Alternatively, the use of whitewashing could constitute an independently derived solution to the problem, common to all houses in a tropical climate, of how to cool the interior. The most important sixteenth-century historical source for the northern Upper Guinea Coast is the Tratado Breve dos Rios de Guiné do Capo Verde, written about 159411 by André Alvares de Almada, a merchant from the Cape Verde Islands. Almada had lived on the coast during the 1560s and 1570s. His account is based on extensive firsthand experience. The major transformation that had occurred in the three generations since the time of Fernandes was the expansion of the slave trade throughout the Gambia-Geba region. Almada describes the Gambia as “a river with an extensive slave trade”; the local Manding were “very warlike” and had constructed fortresses along the Gambia and its tributaries.12 These strong wooden redoubts, surrounded by palisades of stakes, attested to increasing instability; a result, one may surmise, of the slave raiding which was now widespread.13 Indeed, north of the Casamance River, the Floups and the closely related “Arriatas” had only recently learned to protect themselves by military means against slave-raiders, and they now captured or killed their enemies.14 These groups thus managed to escape the growing depredations of slave raiding. The Bagnuns, however, and their Floup neighbors south of the Casamance River were more deeply involved in the slave trade.15 Furthermore, the Cacheu River, then called the Rio de San Domingos, was the center of an extensive trade in slaves.16

16th- and 17th-Century Architecture in Gambia-Geba

37

Almada’s account depicts a region in which slave raiding and the resulting lack of security had grown significantly since the beginning of the sixteenth century. All of the local peoples either participated in the trade, provided some of its victims or, like the Arriatas, were forced to take strong defensive measures to isolate themselves from that commerce. In addition, at the beginning of the seventeenth century, all of the coastal peoples, including the “Portuguese,” were subject to repeated raids by Bijogos warriors from the adjacent Bissagos archipelago. Almada provides only three descriptions of local African buildings in the Casamance-Geba area. One of these clearly shows that the Buramos, or Papels, were forced to live in dwellings that were more labyrinths than houses, for protection against the slave raids of the Bijogos, the inhabitants of the Bissagos Islands.17 The Bijogos’ depredations continued into the second decade of the seventeenth century. In 1607, these raids induced the Portuguese to reconstruct and strengthen their fort at Cacheu. Their intent was to protect commerce and the inhabitants against the islanders, who were described as having laid waste to twelve “kingdoms.”18 Almada’s second description of local dwellings is of Farim. Located at the head of navigation on the Cacheu (S. Domingos) River, this trading escale was over 100 kilometers from the coast and hence was protected from the Bijogos’ seaborne attacks. Almada writes: “The houses of these people are of the same type as those in Casamance, covered with leaves [?] with large circles of wooden stakes, made into a wall called a ‘tapada,’ behind which circles are arranged the houses, in accord with the means of the owners [moradores].”19 This description resembles housing compounds found today in the adjacent Upper Casamance near Kolda east of Sedhiu, although defensive palisades are no longer built there. Both Farim and Kolda lie beyond the low-lying coastal land crisscrossed by tidal streams, or marigots, that was home to the Floups and Bagnuns. The final passage in which Almada describes local dwellings focuses on the physical layout of the habitations. The Beafadas, unlike other African societies with which the author was familiar, did not construct their houses in compact villages. Rather, their dwellings were built at a distance from each other. In each isolated compound lived an extended family, which was under the authority of the eldest member.20 Casamance peoples today exhibit a diversity of settlement patterns. This range of housing patterns is also documented during the nineteenth century; Almada’s description is evidence that the variety is of considerable age. The diversity of settlement patterns is further complicated by the element of variation over time in response to political and economic factors and to climate change.21 Among factors affecting alteration in settlement patterns and building design, the most immediate was warfare and, more specifically, slave raiding. Virtually every European chronicler from the sixteenth century to the colonial period refers to villages surrounded by stockades, to palisaded housing compounds, or to fortress-like dwellings.

38

“portuguese” style and luso-african identity

Protection against attack was a recurrent theme in the Gambia-Geba region throughout this period. The need to make one’s compound secure from attack limited the options for construction and played an important role in the elaboration of local architectural forms. At the end of the sixteenth century, Almada describes Mande defensive fortifications in The Gambia that were surrounded by wooden palisades and backed by an earthen rampart. Almada’s contemporary André Donelha visited Manding communities in The Gambia that were “entirely surrounded by stakes driven into the earth.”22 A generation later, Richard Jobson reported that the important Gambian trading center of Casão was “circled with posts and pieces of trees set close and fast into the ground, five feet high, so thick that, except in stiles . . . a single man cannot get through and, in like manner, a small distance off the like defense, and this is as they signify to us, to keep off the force of horse.”23 At the beginning of the eighteenth century, in a passage almost certainly plagiarized from a now lost portion of La Courbe’s travel account, Père Labat described similar fortifications surrounding Floup houses: “We traversed some Floup villages that were surrounded by five or six rows of palisades.”24 Labat also describes Geregia as surrounded by “a double palisade of thick stakes.” Inside this perimeter, seven or eight “Portuguese”-style houses were in turn surrounded by another palisade.25 According to Labat (actually La Courbe), the fortifications served to protect against wild animals and human enemies, “who do not dare to attack these redoubts, where ten men can easily hold one hundred at bay.”26 Fifty years after La Courbe, Francis Moore wrote about the Floups of Fogny that “each of their towns [is] fortified with sticks drove all round, and filled up with clay.”27 La Courbe also wrote of the Floup roi de Bolole that “his house is like a labyrinth.”28 In summation, it is clear that both labyrinths and palisades were part of a regional architectural technology that was widely shared by the ancestors of the different peoples of the Gambia-Bissau region. Further information about early architecture in the region is provided by cartography. Upper Guinea Coast architecture is illustrated on two early maps whose authorship has been convincingly attributed by Jean Boulègue to La Courbe and dated to the period 1690–1699.29 The two maps, both of which have two parts, are closely related, one being a finished version of the other. The former bears the legend “Carte de la Coste d’Afrique depuis la rivière de Gambie jusqu’à celle de Cherbe ou Madrebombe . . . présentée à M. de Pontchartrain.”30 The second is a rough draft, entitled “Carte de la Concession de la Compagnie du Sénégal, depuis le Cap Blanc jusqu’à la Rivière de Siera Leone.”31 The finished map includes most of the place-names and peoples visited by La Courbe in his journey from Vintang to Bissau. Several of these labels are accompanied by illustrations of houses. Some Manding villages are represented as groups of round houses surrounded by a circular wall in the form of an undifferentiated barrier (Figure 3). The Portuguese com-

16th- and 17th-Century Architecture in Gambia-Geba

39

munity of Farim, on the Rio San Domingos, is shown surrounded by a rectangular palisade of stakes. Inside the enclosure are eleven rectangular buildings, maisons à la portugaise. One of these structures has a tower and clearly represents a church. Cacheu, too, is illustrated by rectangular houses, one of which has a tower. La Courbe paid particular attention to the Portuguese forts, including those at Ziguinchor and S. Domingos, which he represented by aerial-view diagrams. Kaabu (Cabo), a Mande state centered in northeast Guinea-Bissau, is also depicted as a group of houses surrounded by a circular perimeter of stakes. In contrast with the walls that surround La Courbe’s depictions of other Manding villages, these palisades consist of clearly differentiated stakes. Several other locations on this map are accompanied by images of palisaded houses (Figure 4). These include the “Feloupes sauvages” north of the Casamance River, near present-day Buluf, and the “Feloupes sauvages” south of the Casamance River, in the region now called Kasa. In addition, as Walter Hawthorne has noted, a number of Balanta communities are surrounded by timber fortifications.32 It is possible that this detail was derived from a reading of La Courbe’s text. That text has not survived, but Labat’s version, closely based on La Courbe’s original, mentions Floup villages surrounded by several rings of palisades. On the other hand, as Boulègue has shown, La Courbe was directly involved in the visual representation of the Portuguese forts. He would have seen, if not helped to draw, the sketch version of the map. The images on it must have corresponded at least generally with his recollections. The fact that on this draft several Floup communities are shown surrounded by palisades may indicate that La Courbe himself associated this defensive element specifically with the Floups. The prominent role of fortifications in Floup villages confirms Almada’s earlier observation that they successfully defended themselves against slave raids. Seventeenth-century houses of both African and Luso-African origin are relatively well documented in written sources by French and English, as well as by Portuguese, narrators. The most detailed account of the region south of the Casamance River is by Francisco de Lemos Coelho, a Portuguese trader who spent twenty-three years on the coast between about 1640 and 1665, lived in Cacheu and Bissau, and visited the Bissagos Islands.33 His Description of the Coast of Guinea, written in 1669 and revised in 1684, offers a detailed description of Cacheu and its Portuguese and Luso-African communities. He also offers useful information about settlements along the lower Gambia River. Lemos Coelho’s comments about Vintang (or Bintang), a mixed Manding and Bagnun village that was the northern terminus of an overland trade route from Cacheu to the Gambia River, are limited to commercial considerations. Six leagues south of Vintang, in the Bagnun kingdom of Sangédegu,34 he visited the village of Jamai. “The whites,” he

40

“portuguese” style and luso-african identity

3. Map of the coast of Africa showing Farim, circa 1690. Michel Jajolet de la Courbe. Photo courtesy Bibliothèque nationale de France.

16th- and 17th-Century Architecture in Gambia-Geba

4. Map of the coast of Africa showing “Feloupes sauvages,” circa 1690. Michel Jajolet de la Courbe. Photo courtesy Bibliothèque nationale de France.

41

42

“portuguese” style and luso-african identity

writes, “used to live in this village, which is stockaded around because of war.” Lemos Coelho was himself involved in the slave trade that was a major cause of this conflict; he inadvertently provides information about local crafts when he reports that seventeen of his own slaves had run away from him and that the fugitives included carpenters.35 While these men likely specialized in boat-building, their skills may have also been used to build houses. Lemos Coelho’s descriptions of architecture and village layout focus on the Portuguese settlement at Cacheu. The history of this community has been written by George Brooks; here, I concentrate on the buildings. There were two main streets built along the river, Rua de Diante (Front Street) where the wealthy lived, and Rua de Tabanca, at the foot of the fortification. Cacheu was further divided into Vila Fria, which contained the parish church and the houses belonging to the commandant, and Vila Quente, where most of the free black population lived.36 Cacheu, founded in the 1580s, was an important slave-trading port for captives from the entire coastal area from Vintang south to Beafada country.37 It was officially recognized as a vila, or town, in 1605.38 Sixteen years later, the town contained seventy or eighty houses of Portuguese merchants. The original fort, dating from the time of the founding of the community, was rebuilt after 1615 using slave labor. Brooks estimates the population at that time at about 1,500, of whom one-third were probably Cape Verdeans or Portuguese.39 Of particular interest in Lemos Coelho’s account of Cacheu are his descriptions of the stronghouse, or fort, including the materials of which it was constructed and the manner in which it was built, as well as a detailed account of the houses and storage buildings of the Portuguese community. The passage describing the stronghouse is worth citing at length: Its only resemblance to a fort is that it is made of stone and lime, that it belongs to His Highness and is maintained in his name, and that the Commander resides there with the twenty-five soldiers allocated to him, of whom often . . . not as many as three are actually present. Overlooking the water, the Strong House has a terrace of stone and lime, so small that in the rainy season it is thatched over like the other buildings.40

The stone and lime were locally available, the stone from a nearby reef. The lime was produced by burning quantities of seashells.41 In this respect, the stronghouse at Cacheu differed from the earliest Portuguese fortress in West Africa. El Mina, constructed in 1482, was built of stone imported from Europe. These materials afforded solidity, a measure of security, and relative durability. The use of slave labor in the construction of the fort as well as the connection to the Crown would have magnified the association of the structure with physical and social power. Symbolically, the stone con-

16th- and 17th-Century Architecture in Gambia-Geba

43

struction would have been associated with permanence and with elevated status.42 This symbolism is borne out by the fact that only one other structure in Cacheu was built of stone and lime: the parish church, another symbol of Portuguese authority.43 Whitewashing is an example of a building practice that cannot be traced exclusively to either European or African origins. Within West Africa, some Mande peoples whitewashed their buildings at least as early as the sixteenth century (see below). The walls of the Portuguese settlers’ dwellings along the East African coast in the seventeenth century were also covered with lime. Liming by the Portuguese in East and West Africa may have had a common inspiration in the architecture of the Algarve in southern Portugal.44 But local African models also existed. There were, of course, practical as well as symbolic reasons to use any of the several forms of whitewash. The use of lime may have served to cool the interior of the buildings by increasing the amount of sunlight reflected away from the exterior wall. Furthermore, a sufficiently thick layer of lime applied to a mud wall serves as protection against erosion by rainwater. This in turn reduces the frequency with which the adobe must be reapplied. Contemporary written descriptions do not indicate whether a sufficient thickness of lime was applied to buildings in West Africa to protect against rainfall. It is unlikely that the technology was widely practiced, for it is not specifically mentioned in written sources. Protective liming would perhaps have been more effective in the zones of eastern Senegambia and Mali with lower annual amounts of rainfall.45 By the end of the seventeenth century, whitewashed façades had also taken on a symbolic meaning as indicators of wealth and social standing. The application of lime and other forms of whitewashing that were found in The Gambia is highly labor intensive; only wealthy individuals could add this material to their houses.46 The articulation of an Afro-Portuguese housing style that was symbolically associated with status and wealth is reflected in the writing of another contemporary of Lemos Coelho, the French-born Jean Barbot. Although Barbot never visited Cacheu, as he freely admitted, he did interview two pilots who visited the town about 1680. Barbot writes: “The inhabitants of Cacheu, who are almost all Portuguese mulattoes, are entirely supported by the trade they have with these regions. . . . They all live in Portuguese style and are dependents of the King of Portugal” (italics mine).47 M A I S O N S À L A P O RT U G A I S E The earliest commentator to describe in detail “Portuguese”-style architecture is the Frenchman Michel Jajolet de la Courbe, who journeyed from Albreda in The Gambia overland through the Casamance to Cacheu

44

“portuguese” style and luso-african identity

and Bissau in 1686. Until the twentieth century, his account was known only through Père Labat’s plagiarized version attributed to André Brue. However, La Courbe was the original source for Labat and for all later travelers who quoted Labat.48 His description of the house belonging to a wealthy trader at Albreda, La Belinguère, effectively defines houses à la portugaise: “She received us most civilly in a Portuguese style house, that is to say, with whitewashed earthen walls and a small vestibule before the entry, where we were invited to sit in the open air on woven mats.”49 This passage is echoed by La Courbe’s description of the parish church in Bissau: “It is built almost on the seashore, made of earth like the Portuguese houses, and covered [roofed] with thatch . . . and in front there is a kind of encircling vestibule or gallery.”50 Labat, who borrowed most of his material on this part of the coast from La Courbe (although, as Jean Boulègue has demonstrated, he did have access to at least one other earlyeighteenth-century source), gives a similar description of “Portuguese” houses in Cacheu: “The houses there are built of tamped earth, whitewashed both inside and outside with lime.”51 Significantly however, “Portuguese”-style houses were not the exclusive privilege of Portuguese and Luso-Africans. At Jillifry (Juffure) on the Gambia River, the English built their trading house à la portugaise.52 South of Vintang, La Courbe observed that the king of Fogny, who was a Bagnun, lived in a house that was also “faite à la portugaise.”53 And the most impressive dwelling in Vintang belonged to English traders.54 Throughout the Gambia-Casamance-Bissau region during the lateseventeenth- and early-eighteenth centuries, wherever the presence of Luso-African traders provided a model, houses “in the Portuguese style” were adopted by local rulers and merchants as symbols of social status and wealth.55 The meaning of “à la portugaise” may have varied slightly from one observer to another, but this definition generally entailed construction in dried earth or sun-dried bricks—the local soil has a high clay content, making it well-suited for adobe and banco—covered with clay or lime to give a whitewashed effect, and a vestibule in which to receive visitors. For traders, this vestibule was important. Commerce in the GambiaGeba region was based on personal relations between host and guest; the latter was often a traveling merchant.56 Visiting traders were welcomed and given accommodations by their hosts, who in turn received a percentage of their guests’ commercial gains as well as the prestige that accrued to those who hosted foreign visitors. Among many Manding and Diolas even today, to accord hospitality to strangers is deemed an important social virtue.57 Landlord-stranger relations were based, as Brooks observes, on reciprocity: “hospitality and protection for traders and their goods [in return for] the first opportunity to trade with the strangers.”58 The vestibule served both as the locus for extending the expected hospitality

16th- and 17th-Century Architecture in Gambia-Geba

45

and as a transitional space between exterior public space and interior private space. If the traders’ houses served to welcome visiting merchants and as a place to transact business, one may well ask where the goods were kept. Many of the items exchanged—kola, iron, glass beads, amber among imports; ivory and wax among exports—could have been stored in rooms or storage areas in the house, but what about the slaves who constituted an important part of the commerce? Where were captives kept during halts on the way to the coast? Travelers, even those involved in the slave trade, are not explicit on this point. However, some of the “Portuguese” merchants’ houses were located within fortified compounds. At Geregia, for example, near the mouth of Vintang Creek, although the village itself was surrounded by a double barrier of stakes, the maisons à la portugaise located inside these walls were themselves enclosed within another palisade. The inner wall would not likely have been needed for protection against attack. It probably served to keep captives inside. Labat writes: In the middle we saw a large number of houses of Blacks, enclosed within a double barrier of palisades 10 or 12 feet high, 7 or 8 houses in Portuguese style . . . that were enclosed within another palisaded barrier, with a very small, narrow and low door, that looked rather more like the entry of a prison.59

For the captives who were likely held in this inner courtyard while on their way to the coast, a prison is precisely what the structure was. Luso-African merchants erected their distinctive style of architecture in the riverine trading centers and upstream at the more remote escales. At the trading community of Vintang, La Courbe observed: There are a number of houses built in the Portuguese style. . . . The Portuguese are rather numerous in this place. They seem to be rich; their houses, although covered with fronds, are beautiful, spacious and well furnished for the country. They have a church that is larger and cleaner than the church at Juffure.60

The size of these houses probably permitted the owners to lodge African gourmettes (grumetes) who purchased wax in remote villages and brought it to the escales. The houses served to show off the wealth of their merchant owners while at the same time functioning to receive traveling merchants. La Courbe, who was welcomed by a wealthy senhora, or signare, the widow of a Luso-African merchant and herself a wealthy trader, was entertained in the vestibule.61 There, he was served kola nuts imported, appropriately, from Sierra Leone via Cacheu. In Geregia, seven leagues south of Vintang, La Courbe was likewise welcomed in a maison à la portugaise.62 There were also “Portuguese” at each of his successive stops: at Pasqua; at James, a center for trade in

46

“portuguese” style and luso-african identity

beeswax where “the Portuguese have several attractive houses”;63 and at Guinguin south of the Casamance River. BUILDING STYLE AND “ P O RT U G U E S E ” I D E N T I T Y The whitewashed “Portuguese” houses expressed and embodied the economic role of the Luso-Africans who lived in them. Just as their layout facilitated the commerce that was carried out inside, so too their distinctive white exterior represented the distinctive identity of the members of the local Luso-African community.64 By the late seventeenth century, members of the Luso-African communities, especially in remote areas, were physically indistinguishable from other local African populations. Yet they called themselves “Portuguese” and considered themselves to be a group. Along the coast and in the immediate hinterland, where communities of diverse ethnolinguistic origin have lived in close proximity for centuries, the defining characteristics of identity still include language, profession, religion, and architecture. From a distance, it is often possible to identify who lives in a community by the form of their houses. In the Lower Casamance,65 where dried-earth construction still predominates, the Jolas build with courses of dried mud (as in Figure 22, Chapter 6), while the Manding use dried-mud bricks. Seventeenth-century LusoAfricans’ houses of whitewashed, sun-dried bricks were adapted to their social and economic role as middlemen in long-distance trade. These dwellings constituted one of the defining symbols of their wealth and of their identity as “Portuguese.” Viewed in this context, the repeated references to houses built à la portugaise assume added sociocultural significance. L O CA L A F R I CA N I N F L U E N C E O N “ P O RT U G U E S E ” - S T Y L E B U I L D I N G S Houses built à la portugaise made use of local building materials for the walls and the roof. Only the forts, some churches and, by the nineteenth century, a few of the houses in Bissau66 and possibly Ziguinchor were constructed of stone. Most European and Cape Verdean merchants living in trading centers such as Cacheu used adobe bricks. While adobe was less durable than stone, its use dramatically reduced building costs and the labor required for construction. Interestingly, no local industry tied to stone-cutting, brick-making, or lime production seems to have developed in the Gambia-Geba region before the nineteenth century. By contrast, in Saint-Louis, La Courbe reported that he himself set some of the local “whites” to work making bricks and lime.67 “Portuguese”-style houses were roofed with thatch68 or covered with palm fronds.69 Thatched roofs are characteristic of African architecture in

16th- and 17th-Century Architecture in Gambia-Geba

47

the Gambia-Bissau region. Palm leaves provide a temporary covering; they are installed quickly and are used even today to protect buildings under construction and for temporary structures. Roof tiles were generally used only for churches and forts. According to Lemos Coelho, the parish church in Cacheu was roofed with tiles.70 Another category of building, besides the fort, that did not have a thatched roof was the cumbete [storehouse]. Lemos Coelho’s rather enigmatic description of “a building with earth on top” does not indicate the method of construction. In the rainy coastal climate, flat adobe roofs would not have lasted long. Nevertheless, perhaps adobe was adequate for a small surface, such as a granary roof. As Lemos Coelho explained, the nonflammable material was necessary: “The cumbetes are the storehouses of all residences, in order to counter the danger of fire. Because they have earth on top, even if the town is burned down, as very often happens, what is stored . . . is never burned.”71 It would not be advisable to extrapolate from twentieth-century building techniques to reconstruct these seventeenth-century structures “with earth on top.” Yet it is worth parenthetical mention that thatched houses in Guinea-Bissau are today sometimes protected against fire by constructing an armature over the roof made of long sticks that are covered first with leaves and thatch and then with clay soil.72 The possibility that similar building techniques were relatively widespread in seventeenth-century Bissau and Casamance is suggested by La Courbe’s description of the architecture of the Floup community of Bolole (see below). It would therefore appear that for these small storehouses, the “Portuguese” of Cacheu were following a local model for building durable, fire-resistant dwellings. V E R A N DA S The question of how the veranda developed in West African architecture poses a complex historical problem. Verandas, roofed porches that extend around part or all of a dwelling, are common to many cultures throughout West Africa. They also rapidly became an integral element of European colonial architecture. Verandas are found on plantation houses in eighteenth-century Virginia, they appear in East and West Africa, and they are characteristic of British colonial buildings in India and French colonial structures in Southeast Asia. In short, the veranda is a nearly ubiquitous phenomenon in tropical and subtropical climates. Verandas or galleries were found in Africa from the earliest contact with Europeans (and probably before); they appear in Brazil and, later, throughout the New World. But galleries and enclosed porches were also features of European architecture. By the sixteenth century, they were found in Portugal as well as in Italy and in Burgundian court architecture. Indeed, as Catherine Coquery-Vidrovitch has observed, Portuguese architecture from the Algarve, which may have influenced the early development of maisons à la portugaise, reflects in turn North African “arabo-musulmane” influence. The

48

“portuguese” style and luso-african identity

5. Interior veranda of an impluvium house in Etama. Photo by Louis-Vincent Thomas, 1954.

attempt to establish specific historical origins for the veranda may therefore be pointless.73 Close observation of local architecture in Senegambia and GuineaBissau suggests that vestibules and semi-enclosed roofed porches are indigenous to this region. Round buildings are indigenous to much of West Africa, including the Senegambia and Guinea-Bissau. No one would argue that such structures were introduced here from Europe. The congruence of building materials, design, and function further suggests that in Casamance and in Guinea-Bissau, round houses of dried earth have long incorporated porches. The Papels (Manjaks), the southern Floups, and some of the northern Jolas or Floups all construct round houses. These dwellings often have a vestibule or veranda, as they do among the northern Jolas of Buluf. In the case of impluvium houses (among the Jolas of Enampore, the Jola-Esulalu, the Manjak, and the Nalu), the dwellings are built around a central courtyard (Figure 5), often with an interior porch. A well-documented example of the round house with an encircling veranda is the Nalu dwelling from south of Rio Geba that was documented by Teixeira da Mota in 1948.74 The walls of the Nalu house form two concentric circles of dried earth. These walls are connected at four points by short transverse walls like radii of the circle. The inner circular wall is pierced by two opposing doors. In line with these doors, the outer wall is interrupted both front and back for about one-eighth of its circumference. The roof extends to cover both inner and outer walls. The resulting structure thus includes both an enclosed veranda on either side of the house and an open porch or vestibule in front and in back.

16th- and 17th-Century Architecture in Gambia-Geba

49

The structure of the Nalu building assures maximum strength, since the short transverse walls serve as buttresses that convey lateral thrust from the inner circular wall to the lower outer wall. In addition to their structural role, the double walls provide a shaded porch and a transitional space between exterior and interior; furthermore, the house receives crossventilation through the front and rear vestibules and the two aligned doors. Both structurally and functionally, this is an optimum design. Such a marriage of function and local form suggests that the veranda was not borrowed from a foreign source. Had it been appropriated from European design, the veranda would initially have been part of the design of rectangular dwellings, and it would necessarily have been reinterpreted to fit a round design. In the Nalu construction, however, the veranda appears as an integral part of the design. A F R I CA N O R I G I N S O F T H E “ P O RT U G U E S E ” S T Y L E The “Portuguese”-style house was the product of reciprocal influence between Portuguese and local African architecture. The defining elements of the style may themselves be amalgams of Lusitanian and West African construction techniques, brought together to create a building appropriate to a wet tropical climate. This model of historical interaction and architectural blending suggests that it would not be fruitful to seek the origins of the style exclusively in either Portugal or Africa Nevertheless, most, if not all, of the essential elements that defined the “Portuguese” style of domestic architecture were in fact indigenous to West Africa. While the absence of pre-contact written sources makes it difficult to prove that specific construction materials or forms predate the fifteenth-century arrival of the Portuguese, the presence of these forms is documented by the sixteenth century. The earliest references point to western Mande origins for some of these forms. This, in turn, suggests an historical association with the monumental architecture of the Western Sudan. Such a connection establishes the historical agency of West African, as well as European, architectural influences. The maisons à la portugaise were characterized by four elements: they were built of banco, or sun-dried clay (essentially adobe); they were rectangular in form; the outer walls were often whitewashed; and they had either a continuous veranda around the exterior wall or a smaller vestibule in front of the entrance. The use of dried mud or clay, as opposed to stone or masonry, is clearly a West African characteristic. Early Portuguese settlers, even in important trading centers such as Cacheu, rarely used stone to build. In Bissau, La Courbe observed that not only the houses of the Portuguese but even the parish church were constructed of earth.75 Lemos Co-

50

“portuguese” style and luso-african identity

elho, who had lived in Cacheu, wrote that the house of the commander was built of sun-dried bricks,76 although he observed that the local church was of stone and lime with a tile roof.77 While one may safely assume that construction in banco is an ancient and indigenous practice among the peoples of the Guinea Coast south of the Gambia River, archaeological studies have not been carried out in this geographical area. However, closer to the center of the Mande cultural zone, archaeological excavation at the early urban center of Jenné-Jeno in the inland delta of the Niger River in Mali proves that construction with sun-dried clay dates back well over a millennium. Dried-mud construction is found in the earliest monumental architecture of the Mande heartland. Nineteenth-century photographs of the ruins of the Jenné mosque show that the earlier structure had been built of dried mud. This is confirmed by René Caillié’s 1828 account. Caillié, the first modern European visitor to describe Jenné, observed: “In Jenné there is a large mosque made of earth, dominated by two massive low towers.”78 Whatever changes were made in the building’s form during its 1906 reconstruction, the construction materials remained the same.79 The Friday Mosque was characteristic of mosque construction throughout the inland delta, for which it probably served as a model. Most of the architecture of Jenné, including the Friday Mosque, was made of sun-dried brick. Caillié described in detail the method of house construction that was common to both Jenné and Timbuktu. In Jenné, “The houses are built of bricks that are dried in the sun. The sand of the island of Jenné is mixed with a little clay; they use this to make bricks that are round, but rather solid. The houses are quite as large as those of European villagers. Most of them have a second floor.”80 The walls, however, were not whitewashed, because lime was not readily available so far from the coast. In Timbuktu, Caillié added further observations about the method of making these bricks: “I had the opportunity to observe the manner in which the masons of this country construct. . . . They make round bricks. . . . These bricks are similar to those of Jenné. . . . Each house is in the form of a square.”81 Construction with round, sun-dried bricks is, in fact, an ancient tradition in the region of Jenné. The archaeological excavations at Jenné-Jeno have documented the use of the round Jenné-ferey to the eighth century. While the oldest brick structures at Jenné-Jeno were circular, these were replaced by rectangular buildings by the eleventh century. The foundations of these structures are still visible today (Figure 6), clear proof that rectangular houses made of adobe bricks were part of the Mande building tradition long before the arrival of the Portuguese on the West African coast. Rectangular adobe buildings, both religious and secular, were associated with the period of the ascendancy of the Mali empire. Their geo-

16th- and 17th-Century Architecture in Gambia-Geba

51

6. Foundations of a rectangular house in Jenné-Jeno, Mali, eleventh–fourteenth centuries, showing round sun-dried bricks. Photo by Peter Mark.

graphical distribution, too, followed the expansion of Mande influence. By the sixteenth century, rectangular multistoried buildings made of banco existed as far west as the limits of Mande expansion. In The Gambia, the inspiration for such architecture probably came not from Europe but from the Mande heartland. In his travel narrative written around 1625 but based primarily upon his firsthand knowledge of the coast in the decade up to 1585,82 André Donelha described the Gambian town of Casão, a Manding Muslim community that was also inhabited by numerous tangomaos. Donelha’s description is a crucial source of information about early western Mande architecture. This architecture was remarkable. Donelha offers detailed descriptions of the buildings: The village of Casão is small and round in form. The houses are round, made of sun-dried bricks that are whitened with a white mud that resembles lime. Some of them, like that of the Duke or Sandegil, have a second floor and most of them have banquettes in the interior, made of unfired bricks. . . . All have doors. Their locks and wooden keys were the first I had ever seen.83

52

“portuguese” style and luso-african identity

The public buildings of Casão strongly reflected the influence of Mande Sudanic-style architecture. In a subsequent passage Donelha describes these public buildings, including the courthouse, a two-story building that had whitewashed brick walls and was rectangular in form: “This house was square, made of unfired bricks that were very white, surrounded by a banquette. . . . The house had two doors, one opposite the other.”84 It is not plausible that the low bench surrounding the tribunal was open to the sun in this climate; one may safely assume that the roof extended beyond the walls to provide shade.85 Such a construction would have formed an encircling porch (or piazza) similar to the verandas that later typified the “Portuguese” houses. This would make Donelha’s description among the earliest, perhaps the first, of a porch or veranda in an African building.86 Upon leaving the tribunal, Donelha entered a courtyard surrounded by multistoried buildings, further evidence that the monumental architecture of the Western Sudan provided a model for large-scale public buildings among Manding traders settled in The Gambia: “We left this house and we passed through another doorway and penetrated into a courtyard surrounded by multistoried houses. . . . The wall was covered, as if with a tapestry, by a large mat with designs on it.”87 While the tribunal and associated buildings clearly follow a Mande architectural model, they effectively have all the elements of “Portuguese”style architecture. The tribunal is rectangular, it apparently has a veranda, it is whitewashed, and it is made of banco. In fact, all of the defining characteristics of “Portuguese”-style architecture are to be found in the Mande style of monumental construction, a style that predates the Portuguese presence. Furthermore, the one element that may distinguish the tribunal from later “Portuguese”-style buildings—the fact that it is constructed of bricks rather than simply of mud courses—clearly links this building to the Mande tradition of building with bricks. Casão was primarily inhabited by Manding traders. As Donelha himself observes, “the wealthiest merchants in Guinea are the Manding, especially the bixirins, their priests.”88 They were a local social and political elite, and they used established Mande models for the design of their public buildings. The Manding were themselves immigrants to the Lower Gambia. In 1585, their presence was still rather recent, dating perhaps to as late as the previous century, and it is likely that other groups besides Manding and tangomaos lived in Casão. In such a mixed population, it would be logical for the Mande political hierarchy to have used architectural forms that served to express and confirm their authority. Hence, they constructed buildings, such as the tribunal, whose style referred to the source of Mande political authority: the empire of Mali. Donelha adds: “These bixirins are to be found in all of the coastal trading ports . . . including S. Domingos, Rio Grande.” Clearly, some of

16th- and 17th-Century Architecture in Gambia-Geba

53

the Manding frequented the Portuguese trading centers near the coast. While it is possible that they observed early examples of Luso-African house construction during these visits, the possibility is extremely remote that they borrowed an entire architectural idiom from these briefly visited trading posts, brought the style upriver to Manding villages such as Casão, and then applied the borrowed styles of domestic architecture to public buildings such as the courthouse but not to their own dwellings. It should also be pointed out that Donelha’s description of Casão is based on observations contemporaneous with the establishment of the Portuguese trading center at Cacheu. Donelha gives no indication that the tribunal and associated buildings at Casão were recently constructed. Yet they could not have been even a few years old in 1585 if they were based in part on the newly established community of Cacheu. The likelihood that any of the buildings at Cacheu served as a model at Casão is extremely remote. A Mande prototype for the tribunal of Casão is certainly more plausible than the hypothesis that the Manding appropriated the style of their public buildings—but not of their private houses—from a small community of lançados and tangomaos. This in turn raises the intriguing possibility that in The Gambia, the “Portuguese” style derived, in significant measure, from Mande architecture. If indeed, the maisons à la portugaise were inspired by Manding architecture, by what specific process did Luso-Africans assimilate an African building style? Donelha’s account does not address this issue. An eighteenth-century source from The Gambia, however, indicates that the initiative for architectural borrowings may have come from local African traders. In 1732, Francis Moore described the French buildings at Albreda “like unto the Portuguese houses, with walls about ten feet high, covered with thatch, being supported by strong forkillas, and a space left between the walls and the roof to let in the air.”89 The forkillas [posts] supported a veranda or porch. Moore’s own experience was to teach him the economic and social reasons for building in the “Portuguese” style. When he decided to build his own house, Moore began with six forkillas. He continues: “We next made a porch, by the natives called an ‘Alpainter.’ . . . The natives say they have a right to have an ‘Alpainter’ or porch at every factory to be without doors, where they might have access and shelter.”90 It was thus at the behest of their African trading partners that Europeans in The Gambia constructed trading houses with a porch or vestibule. Unlike the Luso-Africans, who were native to the region and were familiar with local traditions of hospitality, French and British factors were foreigners to whom the local trading and building customs had to be explained. Moore’s porch was built on the advice of local peoples, and it served the same function as the verandas of “Portuguese”-style houses.

54

“portuguese” style and luso-african identity

The role of Africans in the design process illustrates that architecture, like identity, was sometimes the result of a process of negotiation. By the eighteenth century, the porch was considered by both Europeans and Africans to be characteristic of “Portuguese” architecture; this is clearly demonstrated by the fact that Moore’s Gambian informants used the Portuguese word “alpainter” (alpendre). Yet Donelha’s earlier account suggests the possibility of Mande origins. Furthermore, the vestibule was indigenous to architecture along much of the Guinea Coast. It was found throughout the region from Senegal to Sierra Leone that was frequented by the Portuguese in the sixteenth century. It is likely that as early as the sixteenth century, Manding traders and perhaps other African merchants convinced their lançado counterparts to incorporate a porch or veranda into their houses. Once this design element had diffused widely among Luso-Africans, it came to be associated by Europeans and Africans alike with the “Portuguese.” Buildings such as those described by Francis Moore were found throughout the area of Luso-African commercial activity. With their characteristic form, adapted to hosting long-distance traders, these structures reflected the fact that the “Portuguese” who inhabited them were traders and middlemen. In fact, the likely provenance of the “Portuguese” style, deriving at least in part from architecture of the western Mande, suggests that this idiom would be more accurately characterized as “long-distance traders’ style.” Nevertheless, in European eyes, at least from the time of La Courbe, “Portuguese”-style houses were an integral part of Luso-African cultural identity. The maisons à la portugaise appear also to reflect the LusoAfricans’ own sense of their identity. In all Luso-African communities, ranging from Cacheu to remote trading outposts on the Rivers of Guinea, the houses of the Luso-Africans were invariably constructed in the “Portuguese” style. To be a member of the “Portuguese” community implied that one was involved in trade; to live in such a square, whitewashed house with a porch or vestibule was to express one’s identity as a trader and, by extension, as “Portuguese.” L U S O - A F R I CA N I N F L U E N C E O N A F R I CA N A RC H I T E C T U R E Luso-African dwellings served as symbols of social status and wealth wherever they were found, even in the hinterland along the trade route between Vintang and Cacheu. It is therefore not surprising that some local Africans used elements of the “Portuguese” building style for their own habitations. La Courbe visited the king of Fogny, a wealthy Bagnun trader living near Vintang, who received his guest in the vestibule, “car sa case est faite à la portugaise.”91 The characteristic feature of “Portuguese”-style dwellings, an antechamber that served as intermediate space between the

16th- and 17th-Century Architecture in Gambia-Geba

55

public (outside) and the private (interior), must have been as useful to a local Casamance trader as it was to Luso-African and European merchants. In the case of Vintang, a trading town with both Portuguese and Manding inhabitants, La Courbe’s use of the term “à la portugaise” may, of course, be misleading. It is not clear whether this Bagnun ruler had borrowed the style of his house from the former group or from the latter. What to the Frenchman was clearly a “Portuguese” style may in fact have represented a direct instance of cultural borrowing from Manding merchants established at Casão who traded in the Lower Gambia. Since the time of Almada, the Floups had been reputed to avoid contact with the “Portuguese” (not an unwise practice, since their isolation helped to protect them against slave raids).92 But according to La Courbe, the Floups of Bolole, four leagues from the mouth of the Rio San Domingos north of Cacheu, traded with the Portuguese in Cacheu. He describes Bolole as “one of the most beautiful [villages] I have seen in Guinea and one of the most populated.”93 Each family compound had a tower of sunbaked earth that served as granary and protected the rice stores against fire. The custom of constructing fireproof granaries was certainly of long duration among these rice farmers, who lived in a region where riziculture dates to the first millennium. One wonders whether such granaries may have inspired the “Portuguese” at Cacheu to build their cumbetes with earthen roofs. In fact, there is ample evidence of intercultural contact and influence. The inhabitants of Bolole sold rice to the “Portuguese” at Cacheu, and many Floups had taken European names.94 In this context of trade and intercultural borrowing, architectural influence, too, seems to have moved in both directions. The Floup king’s house was built with a vestibule for the royal audiences, much like the residence of the king of Fogny. La Courbe’s extraordinary description of the royal housing compound in Bolole is worth citing in full: The king’s house is one of the prettiest that I have seen in this land; in front there was a large square with an avenue and two rows of rather well-aligned trees. The compound is like a labyrinth, planted around with banana trees like the compound of the king of Bissau, with many small houses placed here and there for his wives and his slaves, and after making several turns and detours, you finally arrive at the middle of his house, in the front of which there was a covered vestibule where he listens to his palabres [audiences; trade talk]. . . . We found him sitting in a chair that the Portuguese had given to him.95

While the vestibule reflects familiarity with Luso-African or Mande housing, the general layout of the compound is quite African. The many small dwellings reserved for the king’s wives and slaves reflect local social organization and especially the king’s elevated status. (Only the wealthiest members of Floup society could afford to have many wives, and slav-

56

“portuguese” style and luso-african identity

ery was not common on a large scale.) The planting of trees to define and order the spatial arrangement of the compound is probably a foreign element introduced by the Portuguese. Even today in Casamance, oral sources attribute the planting of trees for shade and fruit to European influence. While landscaping is not strictly speaking part of the architecture, trees were an important element in the human transformation of the environment and the creation of living space. The labyrinthine layout of the royal compound recalls the facts that the devastating slave raids of the Bijogos were still part of living memory in the 1680s and that the Atlantic slave trade was then at its height. Thus, in its organization of space and constructed form, the late-seventeenthcentury royal compound in Bolole conjoins elements of Floup spatial layout, of innovative responses to the existence of slave raiding, and of selective borrowings from Portuguese and possibly Mande architecture. E I G H T E E N T H - C E N T U RY E U RO P E A N CONCEPTIONS OF “ P O RT U G U E S E ” H O U S E S At the end of the eighteenth century, the French writer Jean Baptiste Durand, who lived in Senegal between 1785 and 1788 but freely admitted to having consulted other sources,96 copied Labat’s plagiarism of La Courbe. Durand writes: “Vintain [Vintang], the usual abode of the emperor of Fogny, was formerly of greater size than it is today. Almost all of the houses are constructed in the Portuguese style. . . . The Portuguese population of Vintain is rather large.”97 Durand, writing a century after La Courbe, was unaware that the “Portuguese” were no longer commercially dominant in The Gambia. By recapitulating Labat’s early-eighteenth-century plagiarism of La Courbe, he was thus unwittingly propagating an image of “Portuguese” commercial presence and architectural influence in The Gambia long after they had dissipated. Similar houses had also spread to the Bissagos Islands. Durand’s tendency to attribute architectural achievement to the “Portuguese” is even more marked when he describes “Portuguese”-style dwellings among the Bijogo people. Here too, he repeats Labat (actually La Courbe). He reports that the Bijogo king of the island of Cazegut lived in a house “built in the Portuguese style, [which] was large, white on the exterior and the interior, with a vestibule at the entry, open on all sides; it was surrounded by large palm trees, and commodiously furnished with chairs and wooden seats.”98 Durand’s description of the house in Cazegut replaces La Courbe’s “bien batie à la portugaise” with a description of the “Portuguese” style. But Durand has introduced one other very important change. Where La Courbe writes that the dwelling is “built in the Portuguese style,” Durand opines that the Bijogos had learned to build such large houses from the

16th- and 17th-Century Architecture in Gambia-Geba

57

Portuguese: “They have learned from the Portuguese how to build large and commodious houses.”99 La Courbe understood the complex nature of cultural relations and marriage between Luso-Africans, Portuguese, and local Africans. For example, he observes of the Bissagos Islands: “There are many black Christians who come from these islands, raised among the Portuguese, and with whose help we are able to trade with [the Bijogos].”100 Sometimes too, he uses the terms “Portugais noir,” “gourmettes portugais” [grumetes], or “Portugais blanc”101 to refer to members of the Luso-African community. The term gourmette derives from grumete, which in Portuguese denotes an apprentice seaman. Initially, grumetes were coastal Africans hired by Portuguese and Luso-Africans.102 Occasionally in travel narratives and archival reports, gourmette implies primarily “trader.” The meaning of “gourmette” would change in the nineteenth century. Consistent with this nuanced usage, when La Courbe applies the term “à la portugaise” to architecture, he is not referring specifically to Portugal. A century later, when Durand translates this expression as “they learned from the Portuguese,” he is imputing a meaning that was foreign to La Courbe’s text. Durand perceives “Portuguese” style as European architecture that was brought to Africa and taught to the local populations. For Durand, the large and commodious “Portuguese” houses, rather than signifying a hybrid Afro-European culture, served instead as evidence of the civilizing impact of European influence. This transformation of meaning is related to the broader process whereby Europeans imposed rigid identity categories upon all Senegambians during the eighteenth and nineteenth centuries. Late-eighteenth-century European observers were blind to the possibility that African architectural traditions could have had an impact on European traders or their Afro-European descendents. Confronted with impressive examples of local architecture, many European observers assumed that the Portuguese had taught the Senegambians how to build. Another, slightly earlier, manifestation of this attitude was M. Adanson’s 1759 description of Manding houses in the Lower Gambia: Their huts are better built and perhaps they are obliged for their taste of architecture to the Portuguese, who were formerly settled in those parts. The walls are made of a fat binding clay which soon hardens. They are all thatched with straw, which hangs down to another little wall breast high, and this makes a small gallery around the hut, where they are sheltered from the rays of the sun.103

Adanson’s description of houses built of sun-dried adobe surrounded by a gallery or veranda contains nothing foreign to Mande building traditions. As André Donelha’s description of the tribunal in Casão demonstrates, the Manding of The Gambia had been constructing rectangular multistoried houses with porches since at least the sixteenth century. Yet

58

“portuguese” style and luso-african identity

two centuries after Donelha, Adanson postulated Portuguese influence rather than credit the African peoples of the Senegambia with the capacity to build durable, impressive buildings. Adanson’s and Durand’s comments on Senegambian architecture reflect an attitude toward West African culture that is distinctly more negative than that expressed in the writings of many earlier observers, from Fernandes through La Courbe. Here, during the Enlightenment, one finds the roots of a later, more consistently pejorative image of African material culture. Clearly, theories of foreign architectural influence in Africa had roots long before the colonial period. The distinctive elements of “Portuguese”-style houses were found not only in sixteenth-century Portugal but also in the Inland Delta of the Niger River in Mali and in the Senegambia. This suggests the impossibility of definitively attributing the origins of maisons à la portugaise exclusively to either Iberia or West Africa. The rapid spread of this architectural style through coastal Senegambia may have been possible because all parties in the coastal commerce had some familiarity with both rectangular structures and verandas. Once the style had developed, however, it became subject to the imposition of symbolic associations, including theories about its origins. These associations evolved independently of the buildings themselves, or of the architects, thereby gradually changing the cultural meaning of these “Portuguese”-style houses. In her study of the history of African cities, Catherine CoqueryVidrovitch suggests that the roots of colonial architecture are to be found in the period before colonization.104 The origins of an architecture adapted to the social needs, the climate, and the materials of sub-Saharan Africa predate the fifteenth-century arrival of Europeans on the Atlantic coast of Africa. Inspiration for the architectural style later favored by longdistance traders—Africans, Europeans, and “Portuguese”—derived in part from Mande culture of the early second millennium c.e. While it is during the period of the earliest Luso-African presence on the West African coast, particularly in Senegambia and Bissau, that a “Portuguese” style was elaborated, the first centuries of contact were characterized no less by African influence on “Portuguese” construction than by European influence on the architecture of local African elites. At the same time, the elaboration of a distinctive “Portuguese” style that was actually EuroAfrican architecture, more African than European, coincided with the establishment of a Luso-African identity. Through the seventeenth century, the existence of maisons à la portugaise helped to define that group.

THREE

Reconstructing West African Architectural History: Images of Seventeenth-Century “Portuguese”-Style Houses in Brazil

The basic form of “Portuguese”-style architecture was not unique to West Africa. In the eighteenth century, similar structures were built as far away as the Indian Ocean. An early-eighteenth-century report from the French island of Réunion reflects the extraordinary geographical range of this style, or at least of the appellation “Portuguese,” since the parameters of that term varied, united by the common feature of a veranda: “The distinguishing elements . . . among others, are the veranda, a sort of portico that forms the entry to the house, in the Portuguese manner.”1 Even within the more limited geographical area of the Upper Guinea Coast, both the form and the materials used for “Portuguese”-style houses varied over the three centuries during which their existence is documented. Certain features, however, were fundamental to the idiom. The houses were square or rectangular, they were usually whitewashed on the exterior, and they always had either a porch or veranda or a smaller, semienclosed vestibule by the entrance. Construction materials varied according to the wealth of the owner and the use of either local or imported material. Thus, roofs might be thatched2 or made of tile. In Senegambia, walls were generally of banco (dried clay earth or mud) or dried bricks or, occasionally, of wattle and daub or even wood, but rarely, at least until the nineteenth century, of stone. Another variant of the “Portuguese” house is detailed by Thomas Astley, whose 1745 description derives from Moore and Labat: Their houses have nothing but the ground for a floor, which however is raised 2 or 3 feet to keep them dry. They build them pretty long, dividing them into several rooms, with small windows, on account of the heat. They have always a porch open on every side, where they receive visits, eat . . . and do all their business.3

60

“portuguese” style and luso-african identity

7. House in Elana showing suspended-roof construction, circa 1955. Courtesy of Louis-Vincent Thomas.

The common characteristic that defined “Portuguese”-style houses both in Réunion and in West Africa was a porch. Yet even this feature varied from a small vestibule to a veranda running the length of the building. In the West African version, porches, where present, were supported by forkillas [forkhilas]. In The Gambia in 1732, Francis Moore observed: [The French] have two or three very handsome houses built of clay, like unto the Portuguese houses, with walls about 10 feet high, covered with thatch, being supported by strong forkillas and a space left between the walls and the roof to let in the air.4

The space Moore refers to enabled night air to circulate through and cool the dwelling. Similar construction techniques are used even today by the southern Jolas (whose seventeenth-century ancestors were called Floups). The roof is raised up on stakes that are anchored in the top of the banco walls. A photograph taken in the Buluf (northern Jola) community of Elana about 1955 shows how the system effectively suspends the roof above the structure of the house itself. This permits circulation of air around the top of the house (Figure 7).5 The “Portuguese” style was not monolithic; rather, it varied with time and in different geographical locations. The wide diversity of features associated with “Portuguese”-style houses suggests that the historical study

Reconstructing West African Architectural History

61

8. View of Rufisque (near present-day Dakar), 1680–1682. Jean Barbot. Courtesy of the Hakluyt Society.

of these buildings should specify both the geographical region and the period under consideration. References to “Portuguese”-style houses in different regions that are unsupported by detailed descriptions or by other evidence do not necessarily refer to the same architectural style. Where similar buildings were constructed in areas geographically distant from one another, only the existence of direct historical connections between the two regions establishes the probability that the buildings were in fact related to each other. The history of maisons à la portugaise is further complicated by the fact that no detailed seventeenth- or early-eighteenth-century illustrations of these structures in West Africa are known. None of the early Portuguese or Cape Verdean chroniclers illustrated their writings. The first traveler who illustrated his work with reasonably well-executed drawings was Jean Barbot.6 Several of his drawings depict coastal settlements. A few of these, although they are not sketched with an eye for architectural detail, show buildings that may be in the “Portuguese” style. Barbot’s illustration of Rufisque, near present-day Dakar (Figure 8), depicts two rectangular buildings near the center of the village, which is otherwise composed of small round dwellings with tall conical roofs. Many Luso-Africans still lived in Rufisque in 1681. The two rectangular structures, contrasting as they do with the other buildings in the picture, may represent “Portuguese”-style houses. They stand side by side and are the largest structures portrayed. Ultimately, however, Barbot’s summary treatment shows nothing of the materials used in construction or, indeed, of any details.7

62

“portuguese” style and luso-african identity

Barbot’s detailed drawings of European forts along the Gold Coast also represent rectangular houses and commercial buildings. A few of these structures have rooflines that hint at the presence of a veranda or covered porch. Particularly noteworthy in his illustration of the Dutch Fort Batensteen at Butre is the second house from the left, at the foot of the hill.8 In Barbot’s illustration of Accra, one house, situated to the left of the Portuguese fort at São Fr. Xavier, has a double-pitched roof, suggesting the presence of an exterior porch. Here, in a Portuguese trading post on the Gold Coast, is a likely depiction of a “Portuguese”-style house. Yet, given the miniscule scale of the drawing, this representation is of primarily anecdotal interest.9 A few other seventeenth-century sources contain illustrations. These are not, however, clearly drawn eyewitness representations of houses. Although Olfert Dapper’s illustrations are frequently cited as historical documents,10 Dapper himself never visited Africa; few, if any, of the illustrations to the first edition of Dapper’s work are based on firsthand sketches.11 Even Dapper’s representations of coastal forts are more a reflection of European fantasy than of African reality; they lack the accuracy of Barbot’s drawings. Similarly, Dapper’s African houses are also based on European models. These illustrations are not particularly useful for the historian of West African vernacular architecture. I L L U S T R AT I O N S O F CA P E V E R D E I S L A N D H O U S E S More than seventy years after the posthumous publication of Barbot’s illustrated text, Joseph Corry published his Observations upon the Windward Coast of Africa (1807). This work is illustrated with engravings based on the author’s own sketches.12 Corry’s drawings include views of the island of Gorée, of the newly founded British colony at Sierra Leone (actually Freetown), and of Porto Praya, Santiago, in the Cape Verde Islands. These panoramic landscapes are reasonably accurate, yet they too lack architectural detail. Their relatively late date raises another problem: architectural style in early-nineteenth-century Gorée, Freetown, and Santiago likely diverged from the model of earlier maisons à la portugaise. Corry’s panorama of Santiago (Figure 9), the most important of the Cape Verde Islands, is significant for its depiction of houses. The scene includes three buildings that stand near the shore to the left of a promontory in the middle of the picture. These rectangular single-story structures have low-pitched, four-sided roofs.13 Small windows in the walls suggest a floor plan with several rooms horizontally aligned; this is reminiscent of Astley’s 1745 description of Luso-African cazas in The Gambia: “They build them pretty long, dividing them into several rooms, with small win-

Reconstructing West African Architectural History

63

9. Porto Praya, Santiago Island. From Joseph Corry, Observations upon the Windward Coast of Africa, 1807. Courtesy of the Yale Center for British Art, Paul Mellon Collection.

dows, on account of the heat.”14 But in Corry’s distant view, it is not clear whether these houses have a veranda or vestibule. While the overall form is consistent with “Portuguese”-style houses, the small scale of the representations provides only a tantalizing generalized outline of the structure. From the sixteenth century on, the population of the Cape Verde Islands maintained close contact with the adjacent mainland. Commercial ties were complemented by migration back and forth, establishing what George Brooks has called an economic, social, and cultural nexus.15 Indeed, the Luso-African communities in Senegambia, in the Rivers of Guinea region, and on the Cape Verde Islands shared a common “Portuguese” cultural identity. This Creole society constituted, as José da Silva Horta suggests, “not only a Luso-African world but, more specifically, a Cape Verdean–Luso-African world.”16 Material culture constituted one element of this Luso-African identity. It is possible that “Portuguese”-style houses existed at an early date in the Cape Verde Islands. Conceivably, the islands may have served as a diffusion point for Iberian architectural techniques, including masonry construction. Accordingly, illustrations of the Islands could provide evidence of “Portuguese”-style architecture. Unfortunately, pre-nineteenth-century images are almost as rare for the Islands as they are for the mainland. The illustrations to Joseph Corry’s narrative are among the earliest that exist. Mid-nineteenth-century pictures have survived. These later architectural images do not constitute historical evidence for the seventeenth century. Nevertheless, these illus-

64

“portuguese” style and luso-african identity

10. Houses in Calheta, Santiago, Cape Verde Islands. From Ilidio Amaral, Santiago de Cabo Verde, a Terra e os Homens, 1964.

trations, which are primarily from the island of Santiago, deserve brief consideration, if only because they suggest how earlier Cape Verdean buildings might have looked. The sixteenth-century cidade [town] of Ribeira Grande, which was abandoned by the nineteenth century, contained impressive buildings of masonry. These structures included a cathedral (remnants of which still exist) and numerous smaller, one-story dwellings. Some of these houses are clearly visible in a mid-nineteenth-century lithograph by an anonymous artist.17 The houses are rectangular and the façades are whitewashed (or limed) and have two small windows, one on either side of a central doorway. A noteworthy feature of some of the dwellings is a four-sided roof. However, neither verandas nor porches are visible. Similar, though more rustic houses without the four-sided sloping roofs appear in other lithographs in the same series.18 Some of these simple dwellings may be found even today, for example in the small community of Calheta. There, both two-sided and four-sided roofs may be seen side by side (Figure 10).19 Whether any of these simple dwellings approximates seventeenth-century construction on Santiago is uncertain. It is possible that this housing style did not change extensively in the century or two before the 1864 illustrations. On the other hand, as the absence of a porch or veranda so clearly indicates, these island houses are not identical with “Portuguese”style houses in Senegambia.

Reconstructing West African Architectural History

65

S E V E N T E E N T H - C E N T U RY DUTCH BRAZIL For the Senegambian mainland, as for the Cape Verde Islands, published travel narratives provide precious little visual documentation of LusoAfrican architecture before the nineteenth century. In fact, Hyacinthe Hecquard’s detailed drawings and watercolors from 1850 (most of which remain unpublished) are among the earliest detailed representations of African dwellings.20 For early images of maisons à la portugaise, one must look to another part of the Portuguese mercantile empire. Illustrations of “Portuguese”-style dwellings do exist, but they are from seventeenth-century Brazil. The Upper Guinea Coast was part of Europe’s first global commercial system, a worldwide network that linked West Africa to Europe, Asia, and the New World. In Brazil, as early as the 1540s, Portuguese settlers established a plantation economy based on the use of slaves to grow sugar. By the early seventeenth century, this system was heavily dependent upon the importation of slave labor from western Africa. The slave trade, which was formally established in Brazil in 1518 with the granting of the first assiento [royal contract], ensured not only a fresh labor supply but also regular and direct contact between northeastern Brazil, the Upper Guinea Coast, and the Cape Verde Islands. In the sixteenth century, relatively few slaves were sent to Brazil. The number of Africans in Bahia by 1600 has been estimated at only 7,000. The origins of these early slaves were diverse, yet the Guinea Coast was most heavily represented during this early period.21 By the 1620s, the bulk of the traffic came from Angola, although slaves shipped from Cacheu and Cape Verde may still have represented a significant portion of the trade in the early 1640s.22 In Dutch Brazil (Pernambuco, northeast of Bahia), imports from Angola surpassed imports from the Guinea Coast only after 1643, with the Dutch occupation of Angola. Total imports of slaves into Dutch Brazil from 1636 to 1651 included 14,073 from Angola and 11,437 from the entire Guinea Coast.23 By the late 1640s, the Angolan portion of the trade had expanded. During this period, the Cape Verde–Rivers of Guinea region was commercially linked primarily to Spanish America.24 Nevertheless, some captives, especially those transported on English and French vessels, were still sent from the Upper Guinea Coast.25 Following the Dutch occupation of Luanda in 1641, the Portuguese shipped increased numbers of captives from the Cape Verde Islands to Brazil. It was only in the 1660s that the percentage of captives who originated in Senegambia fell to 5 percent of total imports.26 Clearly, the African origins of slaves in Bahia and Pernambuco were diverse through the early seventeenth century. The fact that the Upper

66

“portuguese” style and luso-african identity

Guinea Coast was disproportionately represented among the earliest Brazilian slaves may have had implications for Afro-Brazilian material culture. In other words, while significant aspects of material culture, including housing styles, must have represented an amalgam of Guinea Coast and Angolan influence, it would not be unreasonable to ascribe primary importance, at least during the initial elaboration of domestic housing forms among Brazilian slaves, to the Upper Guinea Coast. Not only captive Africans but also Portuguese followed the trade route from Africa to the New World. In Brazil, Portuguese colonists were faced with a climate, similar to that of the Guinea Coast, that made comparable demands on architectural design. Portuguese settlers in Brazil constructed a wide range of dwellings. Wealthy plantation owners enjoyed a level of prosperity unknown to the Luso-Africans of Senegambia. Their two-storied habitations were larger and more imposing than any seventeenth-century Luso-African houses. Brazilian plantation houses bear little relation to early “Portuguese” dwellings on the Guinea Coast. If Lusitanians in diaspora contributed to the spread of domestic architectural styles from Upper Guinea to Brazil, then it is to the more modest dwellings of less wealthy Brazilian settlers that one might look for similarities to the “Portuguese”-style houses of West Africa. An historical connection between domestic architecture of the Guinea Coast and sixteenth- and seventeenth-century settler buildings in the Caribbean and Brazil has been proposed by Jay Edwards, an anthropologist of New World material culture. Edwards observes that European New World vernacular architecture may have derived, not primarily from Renaissance European prototypes, as had previously been hypothesized, but rather from the Antilles. Edwards further posits that the “Antillian Creole house,” which was characterized by a porch or veranda and was sometimes raised above ground level and by a whitewashed façade, may in turn have been inspired by West African vernacular traditions.27 Edwards’s insight revises the widely accepted theory of exclusively European origins for the porch. His proposal is impressive, as he was working backward chronologically in an effort to establish the origins of French Creole architecture in the New World. Edwards’s theory has the added advantage of encompassing the possibility that specific architectural elements, present in several disparate building traditions, may in fact have derived from more than one source. In 1630, the Dutch captured Recife and Olinda, in northeast Brazil, from Portugal. For the next twenty-four years, until the colony was returned to Portugal in 1654, Dutch Brazil was colonized by the Dutch West India Company. In 1637, Johan Maurits of Nassau was appointed governor-general of Dutch Brazil, a post he would hold for seven years. Maurits was an important patron of the arts and of the study of natural science, two branches of intellectual endeavor that were closely interrelated in seventeenth-century Europe.28 As governor, he sponsored the

Reconstructing West African Architectural History

67

magisterial Historia Naturalis Brazilae (1648), a compendium of geographical and natural scientific knowledge about the colony. In order to record the landscape, the diverse populations living in the colony, and the military and domestic architecture of Brazil, Maurits brought two professional artists from the Netherlands: Albert Eckhout (1610–1664) and Frans Post (1612–1680). From 1637 to 1644, the two painters created a pictorial record of Dutch Brazil that is comprehensive, detailed, and of great ethnographic and historical significance. “ P O RT U G U E S E ” - S T Y L E H O U S E S I N F R A N S P O S T ’ S WO R K Frans Post is known primarily for the artistic products of his seven years spent in Brazil. His colonial oeuvre is phenomenal; it includes 800 drawings that depict animals, plants, fortifications, and houses.29 While in Brazil, Post painted at least seven panoramic landscapes.30 After his return to the Netherlands, he painted more than 100 additional landscapes. Post was particularly adept at rendering architecture. This fact has occasioned historical speculation that Johan Maurits employed him specifically to make detailed plans and illustrations of Portuguese fortifications, which would have been of military concern to the newly installed Dutch government. The artist conveys the form and clearly describes the location of forts, plantations, houses, sugar mills, and chapels, as well as the smaller houses of Native Americans and African slaves. His attention to architectural detail is so precise that both the materials and the techniques of construction are frequently evident.31 The drawings and paintings of Frans Post provide a record of architectural form and style in early-seventeenth-century northeast Brazil at the moment of transfer from Portuguese to Dutch control. Most of the structures Post represented dated from the first Portuguese colonization. The buildings that he depicted, from plantation houses to simple onestory dwellings, encompass the entire range of habitations built during the Portuguese occupation. A few of the smaller settler houses appear to incorporate elements of the “Portuguese” style. Post’s work offers visual evidence suggesting that some Portuguese—or Luso-African—settlers in Brazil adopted architectural elements that had originated in sixteenth-century Guinea. And some of the slave dwellings painted by Post are built in the “Portuguese” style. This visual evidence strongly supports Edwards’s hypothesis that New World Creole architecture had African roots. Paradoxically, the earliest accurate visual representations of “Portuguese”-style houses come not from Africa but from seventeenth-century Brazil. Comparison of Frans Post’s representations of modestly scaled Brazilian houses to seventeenth-century written descriptions of maisons à la portugaise in Senegambia suggests that Post’s paintings and drawings

68

“portuguese” style and luso-african identity

document a Brazilian variant of the “Portuguese” style. While the landscapes painted after the artist’s return to Europe rework the Brazilian material to accord with artistic conventions for landscape painting,32 they do not substantially transform the architecture. Scholars of Post’s work disagree about whether the drawings, most of which carry the inscription “1645,” represent his original Brazilian output, finished and dated after his return to Europe, or whether they, like many of his paintings, were made in the Netherlands after earlier sketches that have now been lost. It is likely that the drawings are Post’s original Brazilian work.33 In the period from 1647 to 1669, Post made over 100 paintings that carefully portray building styles in Brazil. In creating these works, as Whitehead and Boeseman argue, he must have relied on sketches made in Brazil.34 Even if the buildings in Post’s landscape paintings were slightly reworked from the initial drawings (and indeed they appear to have been given a romanticized interpretation with an added veneer of dilapidation similar to that in Hobbema’s slightly later depictions of Dutch farmhouses), Post nevertheless maintains scrupulous attention to architectural detail. Hence, these finished works are important documents for the history of Brazilian (and Afro-Brazilian) architecture. T H E M A RC G R A F M A P In addition to the landscape paintings, another series of illustrations of Brazilian scenes and architecture appears to be closely based on Post’s drawings. In 1643, Georg Marcgraf created a map of Brazil, also under the sponsorship of Johan Maurits. This map remained the most accurate cartographic representation of the country until the nineteenth century. The Marcgraf Map—actually a series of maps—is illustrated with engravings that offer panoramas of Brazilian life. In composition and style, these vignettes evoke Post’s work; they were probably based directly on his drawings.35 The map was first published by Johan Blaeu in 1646. A year later, the original copperplates were republished in Amsterdam by Caspar Barlaeus in his Rerum per octennium in Brasilia. In their detail and precision, the Marcgraf engravings strongly suggest a direct connection to Post’s work. This may be seen in a comparison of plantation houses. The map incorporates an engraving that portrays a plantation house behind a sugar mill (Figure 11). The dwelling is rectangular. The upper floor, containing the living quarters, is raised on pillars above a windowless ground floor. The structure is covered by a low-pitched, four-sided tile roof that extends to cover a semi-enclosed gallery. The porch extends nearly the entire length of the upper floor. This building provides a classic example of the Portuguese casa grande in Brazil. The representation is quite similar to the plantation house in a 1652 painting by Post now in the Rijksmuseum. The

Reconstructing West African Architectural History

69

11. A Brazilian plantation house showing the influence of “Portuguese” style. Illustrated in the Marcgraf Map, from Caspar Barlaeus, Rerum per octennium in Brasilia, 1647, rep. 1657. Courtesy of the Library of the University of Leiden.

Marcgraf engraving is almost certainly based on architectural sketches by Post, whose paintings of plantation houses give similar attention to detail. If the Marcgraf engravings derive from architectural drawings by Post, they provide evidence of the artist’s own lost, or hitherto uncataloged, works. The Marcgraf Map also offers evidence that suggests Luso-African influence on Brazilian vernacular architecture36 in a manner that Post’s finished landscapes do not. One illustration from the Marcgraf Map depicts a river scene with a mill in the foreground.37 Beside the mill stands a small dwelling with a four-sided roof with eaves that extend low over the façade and project forward to create a sheltered space. This is a humble structure; it is no plantation house. The dwelling is rectangular in form, with small windows and with walls rendered in the engraving as unshaded surfaces, suggesting that the artist sought to represent whitewashing. This unpretentious structure shows features that typify West African “Portuguese”-style houses. On the right side of the façade, the artist appears to have sketched two posts. These posts support the extension of the roof, creating a small porch. They are similar to the forkillas described by Francis Moore (1738) in The Gambia, a characteristic feature of “Portu-

70

“portuguese” style and luso-african identity

guese”-style houses there. It is not possible, on the basis of the rather worn impression from Allart’s 1657 engraving, to determine the materials of which the house is constructed. In other respects, however, this dwelling resembles the descriptions of Senegambian houses except for the roof tiles, which are rarely found in West Africa before the nineteenth century. This otherwise unremarkable dwelling, far removed both in its scale and its simplicity from the two-story houses of wealthy Luso-Brazilian sugar planters, embodies trans-Atlantic architectural continuity. The picture suggests that the distinctive domestic architectural style that evolved through the interaction of Portuguese, West Africans, and Luso-Africans on the Upper Guinea Coast also spread to Portuguese Brazil by the first third of the seventeenth century. But whereas in West Africa the maisons à la portugaise were associated with wealthy traders, ultimately becoming a symbol of social status among both “Portuguese” and Africans, in seventeenth-century Brazil, less wealthy members of the Luso-Brazilian community inhabited buildings that were built in this style.38 The style may have been brought to the New World by Luso-African slave traders, who played an important role in the early export of slaves from Upper Guinea. It is conceivable that the houses were built by Portuguese immigrants who had sojourned first in West Africa. Alternatively, this architectural style could have been created independently in Brazil, the product of interaction between Portuguese immigrants and slaves brought from the Guinea coast. The similarities may reflect both the interaction of the same two cultural groups to create Brazilian and West African “Portuguese”-style houses and the parallel response by people with limited material means to the challenge of developing housing suited to comparable climates. The end result was the creation of a vernacular idiom that was so close to the “Portuguese” style that, except for the tiled roofs, surviving illustrations from seventeenth-century northeast Brazil could be illustrations of West African maisons à la portugaise. It is not clear how many Portuguese settlers in Brazil had previously spent time in Upper Guinea or the Cape Verde Islands. Ten years after Bahia was founded in 1549, the Portuguese Crown authorized a direct slave trade from Africa to Brazil. By 1570, the rapid expansion of sugar plantations had created a greatly increased demand for African labor.39 During this early period, the overwhelming majority of captives came from Upper Guinea.40 Cape Verdean and Luso-African merchants were deeply involved in this trade. It is likely that some of them eventually settled in Brazil, especially as the expansion of plantation agriculture and rising sugar prices brought rapidly increasing prosperity to the Brazilian Portuguese at the same time that drought and famine afflicted the Cape Verde Islands. Frans Post’s paintings focus on fortresses, plantation houses, mission stations, and sugar mills, the visual embodiment of economic, social, and military power. Representations of the dwellings of the dispossessed—In-

Reconstructing West African Architectural History

71

12. Distant Landscape with Church Ruins, Chapel, and Farms. Frans Post, oil on canvas. São Paulo, Museum of Art.

dians and African slaves—and of poor settlers are frequently consigned to the background. Yet Post’s observation of architectural detail is so precise that even these anecdotal or background features provide an historical record of the architectural landscape in seventeenth-century Brazil. The more distant images offer information about vernacular building styles. As Whitehead and Boeseman observe, “It seems very probable that the scenes depicted were actually observed and although redrawn and recombined afterwards, they can be trusted as authentic records.”41 One of Post’s paintings, entitled simply Distant Landscape with Church Ruins, Chapel, and Farms42 (Wide Landscape with Buildings, according to Joaquim de Sousa-Leao), shows a landscape bisected in the middle distance by a river (Figure 12). Beyond the water, at the left side of the picture, is a small village. The buildings are modest, except for the ruins of a stucco church. Beside the church stand a small chapel and four or five austere rectangular houses. Immediately in front of the chapel, almost hidden by a copse of trees, is a small square structure that resembles the “Portuguese”-style house in the Marcgraf illustration. The plastered walls gleam in the sunlight, suggestive of whitewash. Beneath a four-sided, lowpitched tiled roof, two walls are turned toward the viewer. Both walls are interrupted by long rectangular apertures. These openings seem to be part of a gallery or veranda. The same architectural elements that dominate the

72

“portuguese” style and luso-african identity

upper floors of the plantation houses painted by Post appear here on a much reduced scale. Rather than representing a scaled-down version of the casa grande, however, this small structure, its outer walls partially open to breezes from the river, may reflect a common prototype. This small rectangular white-washed dwelling, much like the house in the Marcgraf engraving, is similar in form to West African maisons à la portugaise. The building is located on a rise, just above the riverbank. Its veranda faces the water, open to the breezes that would blow in the evening and early morning. The location and orientation maximize the cooling effects of the immediate environment. In West Africa, historical sources rarely specify the orientation of “Portuguese”-style houses, but they were invariably built near waterways, if primarily for proximity to trading routes (cf. S. Domingos, the settlements along the Gambia River, in Vintang, and in Ziguinchor). In Vintang, the maisons à la portugaise were built on a hillside, overlooking Vintang Creek.43 In the seventeenth century, the Floup king of Bolole constructed his “Portuguese”-style house next to a stream, partly for defensive purposes44 but also, conceivably, to catch the cooling breezes. In Brazil, as in Senegambia, it would not have taken the Portuguese long to learn the advantages of selecting locations adjacent to—and, topography permitting, overlooking—bodies of water. In Senegambia, they would have learned this from their neighbors. Today, among the Jolas of the Casamance (descendents of the Floups) a word for evening (gÌrÌssÌ) derives from “erus” [breeze],45 the cooling air currents that are the result of diurnal temperature variations. Such local knowledge was widespread in the days before ceiling fans or air-conditioning.46 In one important respect, the small house in Frans Post’s Distant Landscape with Church Ruins, Chapel and Farms differs markedly from seventeenth- and eighteenth-century descriptions of West African maisons à la portugaise. Almost invisible behind the copse of trees that partly screen this dwelling is a series of thin posts. Like the larger plantation houses in Post’s other paintings, this modest house, too, is elevated. The accident of composition that masks the ground floor calls attention to the likely West African prototype. The Brazilian house may have evolved through the addition of an arcaded ground floor or a raised basement to the basic “Portuguese”-style house. Elevated houses with storage space on the ground floor and living space above were also characteristic of French Louisiana and colonial South Carolina. The North American models, however, date to the eighteenth century.47 Elevated structures became common in nineteenth-century West African colonial architecture, too.48 Second-story living quarters had lower humidity and were open to cooling breezes. By the last decades of the century, they were found throughout the French colonies and, with certain stylistic variations, in British and German possessions as well.49

Reconstructing West African Architectural History

73

H I S TO R I CA L C RO S S - C U R R E N T S I N N E W WO R L D A RC H I T E C T U R E The historical sources that influenced Luso-Brazilian buildings were extraordinarily complex. Similar architectural elements existed in many diverse cultural traditions. The presence of verandas or vestibules or the larger enclosed galleries can be traced to more than one geographical region. Nevertheless, trade between West Africa and Brazil and the presence of Cape Verdean or Luso-African settlers in Brazil suggest a direct historical connection to the “Portuguese”-style houses of the Guinea Coast. Upper Guinea and northeast Brazil were both part of the Portuguese cultural and economic sphere. If the descendents of Portuguese settlers arrived at similar solutions to the technical problems of building in both regions, those solutions did not evolve entirely independently. In view of the commercial ties and the flow of human beings, seventeenth-century West African influence upon Brazilian architecture is not surprising. In fact, it prefigures and provides a mirror image of the spread of Afro-Brazilian architectural style from Brazil back to the coastal trading ports of Nigeria, Dahomey, and Togo, a process which occurred when Afro-Brazilian merchants migrated back to West Africa two centuries later.50 Numerous societies in Europe and Africa—not to mention Asia—had developed some form of veranda or gallery. This fact, together with the complex patterns of direct and indirect interaction among many of these cultures, makes it quite impossible to trace stylistic influence from one culture to another simply on the basis of the presence or absence of this architectural feature.51 The subsequent historical development of Brazilian vernacular architecture is increasingly complex. By the mid-seventeenth century, most slaves were imported not from the Upper Guinea Coast but from elsewhere in West Africa. By the second half of the century, direct ties between Brazil and Senegambia had become attenuated and the influence of Senegambia on Afro-Brazilian culture had diminished. Derived in significant measure from indigenous West African architecture, the distinctive building style which Europeans in West Africa termed maisons à la portugaise influenced domestic architecture in late-sixteenth- and early-seventeenth-century Brazil. Over the ensuing two centuries, it is likely that Afro-Brazilian style, characterized by the veranda and raised living quarters, influenced architecture throughout the Caribbean basin. Jay Edwards suggests that these elements spread to North America through the intermediary of Dutch refugees who settled in the Antilles after Brazil was returned to Portugal in 1654.52 He acknowledges the complex influences that could have inspired what he terms New World Creole architecture. Brazilian influence

74

“portuguese” style and luso-african identity

is one of seven possible sources he enumerates. Nevertheless, I believe that he overestimates the role of the Dutch as intermediaries between Portuguese colonists and French and English settlers. All four of these European nations were actively engaged in the Atlantic slave trade. In West Africa, all of them were exposed to both African and Luso-African architecture. (Francis Moore’s experience in The Gambia where, in 1732, he was constrained to build his trading factory in the “Portuguese” style, is a case in point.) It is likely that the similarities between “Portuguese,” French, and English Creole New World building styles were due to a combination of factors. All European colonists faced similar challenges in constructing dwellings suited to a tropical or semi-tropical climate, all were familiar with the solution to this problem that was first articulated in West Africa, and all relied heavily on West African slaves for labor and for technical expertise in a tropical or subtropical environment. B R A Z I L I A N S L AV E A RC H I T E C T U R E African slaves brought their architectural knowledge to Brazil, and they influenced early Brazilian vernacular building styles. Slaves undoubtedly constructed the smaller settlers’ houses such as the one illustrated in the Marcgraf Map. Some West African captives, including Mande slaves, would have had experience in Africa constructing two-story buildings with verandas. Their technical expertise cannot be discounted as a possible factor in the development of Brazilian and New World plantation architecture. Moreover, it is likely that the slaves’ own houses in late-sixteenth- and early-seventeenth-century Brazil incorporated architectural forms from Senegambia and the Upper Guinea Coast. Influence from this region would have been particularly strong during the early period, when a high proportion of captives came from Upper Guinea. Brazilian slave architecture is clearly documented by Frans Post. In a work formerly in the National Gallery in Prague, entitled Brazilian Landscape with Farms, Ruins and Figurative Group, he depicts a group of Africans; behind them, in the middle distance, are three small dwellings.53 The nearest of these is close enough to show structural detail. The building is small and rectangular, with only one or two rooms and a low, red (tile?) roof. The front of this cabin, partially obscured by foreground foliage, consists of a whitewashed wall and an entry porch or vestibule whose roof is supported by a wooden post. The presence of African figures indicates clearly that this is a slave dwelling. This painting, with its ethnographic descriptive detail, suggests that some African slaves in Brazil constructed houses similar to, although simpler in form than, the maisons à la portugaise. Post’s 1656 painting, Brazilian Landscape, now in the Wadsworth Atheneum in Hartford, Connecticut (Figure 13) depicts a group of fes-

Reconstructing West African Architectural History

75

13. Brazilian Landscape, 1656. Frans Post, oil on canvas. Wadsworth Atheneum, Hartford.

tively attired dancing slaves in the foreground.54 To the left and behind them stands a rectangular dwelling, with thatched roof and earthen walls. The doorway lies at the narrow end of the house and is shaded by a veranda. Four unadorned veranda posts support a thatched roof that sits on a framework of branches. This building, too, sits on a hill overlooking a body of water. Post’s Brazilian Landscape with Huts and Casa Grande (1656) (Figure 14), which is in the Carmen Thyssen-Bornemisza Collection in Madrid,55 depicts a plantation with the casa grande in the background, dancing slaves in the foreground and, in the middle distance, five or six slave houses. These rustic dwellings show a variety of forms; all are basically rectangular or square, though somewhat irregular in shape. The nearest and largest of them shows traces of whitewashing; it also has a large porch in front of the entrance. This veranda is shaded by a sloping roof that is in turn supported by four posts, reminiscent of the veranda and forkillas described by Moore in The Gambia. Partially hidden behind this slave house is a dwelling with a four-sided roof whose eaves extend well beyond the outer wall, forming a shaded area. This design is similar to the form of veranda still found in some twentieth-century dwellings

76

“portuguese” style and luso-african identity

14. Brazilian Landscape with Huts and Casa Grande, 1656. Frans Post. Carmen Thyssen-Bornemisza Collection, Madrid.

among Jola and Manjak populations in the Casamance and Guinea-Bissau. It also resembles the shaded façade of the “Portuguese”-style house depicted in the Marcgraf Map. Parenthetically, the simple rectangular form capped by a four-sided tile roof resembles nineteenth- and twentieth-century farmhouses from Santiago in the Cape Verde Islands. If one assumes that seventeenth-century Cape Verdean houses resembled these more recent structures, then the Islands may have played a role in the seventeenth-century articulation of a common style on both sides of the Atlantic. Cape Verdean merchants may have brought their building style to the New World. A third landscape, which Post painted after his return to the Netherlands, depicts slave dwellings that closely resemble seventeenth- and eighteenth-century descriptions of Luso-African architecture. Brazilian Landscape with Natives [sic] on a Road Approaching a Village, dated 1665,56 depicts six Africans in the foreground and, to their right, a two-story farmhouse and a chapel. Lower down the hillside, overlooking a distant body of water, are three simple rectangular structures with thatched roofs and overhanging eaves and rough walls that are not whitewashed and that appear to be made of dried earth.57 One building has a small porch attached to the end wall. These humble structures of simple building materials were clearly built by slaves for their own use. Essentially West African houses transposed to the New World, these dwellings display some of the

Reconstructing West African Architectural History

77

characteristics of “Portuguese”-style architecture. They are not, however, whitewashed. Here, one faces a fundamental problem with the identification of images of “Portuguese”-style houses in the New World. The style itself is basic and unadorned. This makes it difficult to differentiate a generic representation of a simple adobe house with a porch from a specific “Portuguese”-style building. The historical context of mid-seventeenth-century northeast Brazil, with its African population of mixed Upper Guinea Coast, Guinea Coast, and Angolan heritage, further complicates the issue. By 1650, slave dwellings actually constituted an amalgam of styles from the different regions from which the captives had come. The Upper Guinea Coast initially played a determinant role in the articulation of Brazilian slave architecture. Nevertheless, the “Portuguese” vernacular style in Senegambia would have been modified by analogous forms that were imported in the mid-seventeenth century by numerically dominant Angolan slaves. Hence, Frans Post’s images of slave houses in Pernambuco should be seen either as generic representations of Afro-Brazilian vernacular architecture or as representations of a generic variant of the “Portuguese” style. POSTSCRIPT Do any surviving Senegambian buildings offer a concrete illustration of earlier “Portuguese” houses? Our knowledge of sixteenth- and seventeenth-century styles remains general and somewhat conjectural. Even Post’s detailed Brazilian images constitute only indirect evidence for the West African “Portuguese” style. A few colonial-era buildings that appear to reflect earlier descriptions and illustrations do, however, exist in Casamance. Until 1886, when it became a French possession, Ziguinchor was the northernmost Portuguese trading community in the Casamance. The oldest part of the city contains a few two-story buildings that resemble the “Portuguese” houses painted by Post. One fine late-nineteenth-century example of the “Portuguese” style (Figure 15) may, in fact, date to the era of Portuguese control.58 This two-story rectangular structure is surrounded on three sides by an open veranda. The roof flares to cover the end porches. Except for the absence of a hipped roof, this building is very similar to the small raised plantation houses in several of Post’s illustrations. In the old French colonial quarter of Ziguinchor, a rather dilapidated turn-of-the-century, single-family dwelling offers a beautiful example of the single-story “Portuguese”-style house (Figure 16, Figure 17). This rectangular whitewashed building with surrounding veranda59 retains much of its classical elegance in spite of its sorry condition. Nearly identical houses were constructed by the French elsewhere in West Africa around 1910.60

78

“portuguese” style and luso-african identity

15. Two-story “Portuguese”-style house in Ziguinchor from the late nineteenth century. Photo by Peter Mark.

These buildings may be the architectural descendents of the “Portuguese” style, directly modeled on earlier houses. But it is also possible that this relatively uncomplicated style was independently reinvented in the nineteenth century. Although they do not necessarily represent a continuous architectural tradition, these houses do illustrate the physical form of at least some of their eighteenth-century antecedents. The important corpus of architectural sketches and oil paintings created by the Dutch artist Frans Post during and after his extended stay in Dutch Brazil from 1637 to 1644 provides visual documentation of three kinds of houses, each of which evokes Luso-African architecture. These structures include the large two-story plantation house with enclosed porches or verandas (casa grande) and, closer to a West African prototype, smaller dwellings of less wealthy Portuguese Brazilians. The third group of dwellings that resemble seventeenth-century descriptions of the “Portuguese”-style houses of the Upper Guinea Coast are the rough cottages built by slaves in Brazil for their own use. The architectural form referred to on the Upper Guinea Coast as maisons à la portugaise was a product of cultural interaction and assimilation among Africans, Europeans, and Euro-Africans. The spread of this

Reconstructing West African Architectural History

16. French colonial house in the “Portuguese” style built in Ziguinchor circa 1910. Photo by Peter Mark.

17. Side view of the French colonial house shown in Figure 16. Photo by Peter Mark.

79

80

“portuguese” style and luso-african identity

domestic architectural style from West Africa to Brazil and, ultimately, back to West Africa reflected complex patterns of cultural interaction. The resulting Afro-Brazilian architecture, which was closely associated with the expansion of the Portuguese seaborne empire in the sixteenth and early seventeenth centuries, may be seen as the earliest instance of a transAtlantic architectural idiom.

FOUR

“The People There Are Beginning to Take on English Manners”: Mixed Manners in Seventeenth- and Early-Eighteenth-Century Gambia

Precolonial West African trading centers were characterized by complex and often intimate relations between Africans, Europeans, and Afro-Europeans. These relations were reflected in the manners and customs of local populations. Since etiquette and manners directly concerned merchants who visited these communities, customs often figure prominently in accounts written by visiting European traders. These detailed descriptions of local manners portray a microcosm of social relations in coastal communities as early as the seventeenth century. On the basis of these accounts, it is possible to reconstruct and analyze intergroup relations and patterns of cultural assimilation between members of different groups. One region of early and sustained commercial contact between local African populations, European merchants, and locally settled Luso-Africans was the lower Gambia River and the adjacent mouths of the Casamance River, Rio Cacheu, and Rio Geba to the south. By 1700, these waterways had been part of the Atlantic trade for more than two centuries. Trading centers along each of these rivers were home to Luso-African communities and were frequented by European merchants. Local society was characterized by extensive commercial and cultural interaction; intermarriage further facilitated the movement of individuals back and forth among African communities and between African and Luso-African society.1 Indeed, the many different groups that lived in this region, which included Manding, Bagnuns, Papels, Manjaks, Floups, and Luso-Africans, were not clearly bounded. Free movement of cultural traits and of identities typified this zone of extensive long-distance commerce. Individual identities were extremely fluid, and cultural characteristics were extensively shared among the different peoples. The result was an extraordinary degree of cultural admixture and assimilation.

82

“portuguese” style and luso-african identity

In the late seventeenth century, three trading centers along and adjacent to the lower Gambia River—Albreda, Geregia, and Vintang—exhibited diverse forms of cultural mixing to a degree that may have been unusual even for the Upper Guinea Coast. Two specific aspects of local culture in these communities—religion and social etiquette—illuminate the open nature of the societies as well as their extraordinary diversity and extensive intercultural borrowings. These Lower Gambian communities are documented in narratives written by, or based upon the reports of, contemporary European and Cape Verdean merchants. Additional information referring to the nearby trading communities along the Rio de San Domingos and the Rio Grande complements the documentation of the Gambian communities, providing a more complete image of these multicultural trading centers. The most detailed information about late-seventeenth-century trading communities in The Gambia, Casamance and, to a lesser degree, Cacheu is provided by Michel Jajolet de la Courbe’s account of an overland journey from Albreda to Cacheu that he made during his visit to Senegal from July 1686 to February 1687. La Courbe was inspector and, from 1689 to 1693, director of the French concessionary company in Senegal. The part of La Courbe’s manuscript describing the Casamance has been lost. This passage, however, formed the basis for Père Jean-Baptiste Labat’s travel narrative, which he plagiarized from La Courbe and falsely attributed to André Brue.2 Labat’s version, as Jean Boulègue has demonstrated, appears to provide a faithful copy of La Courbe’s description of Casamance peoples and villages.3 I refer to the cultures and society of seventeenth-century Gambian trading centers as “mixed.” The term may be read as roughly equivalent to the French métis. “Mixed,” however, has the advantage of being free of the complex connotations that have grown up in French ethnography around the concept of métis. Recent French anthropological theory uses métissage to describe the formation of cultures that are the product of interaction and assimilation between two or more societies. This usage does describe the historical situation in seventeenth-century Lower Gambia. A problem with the concept of cultural métissage, however, is its lack of a consistent and clear definition.4 Métissage can connote many different processes. To group what may be disparate historical processes under the rubric of métissage culturel is to risk presenting a false sense that these cases are the same. For example, in his highly stimulating, superbly documented, and detailed study of sixteenth- and seventeenth-century Spanish-Amerindian-African interaction in Mexico, La Pensée Métisse, Serge Gruzinski refers to “les métissages” as if they constituted an objectifiable cultural process, implying by extension that “culture métisse” is likewise a category of culture that can be objectively defined.5 Gruzinski refines his use of the term “métissage” by means of synonyms or comparable terms, including “les mélanges,” “les brassages,” “adaptation,” “des mécanisms

“The People There Are Beginning to Take on English Manners”

83

d’association et d’imbrication,” and finally, “ces emprunts et ces appropriations” [these borrowings and appropriations] that occur because they are “dans l’air du temps” [in the air at the moment].6 Ultimately, however, there is no specific and clearly defined definition of métissage. By contrast, he uses “hybridation” to apply to mixtures that develop within a single civilization.7 But since human civilizations are all the product of prior and ongoing intercultural contact—that is, métissage—one fails to see a fundamental distinction between “hybridation” and métissage. In brief, métissage is a useful descriptive term but is of only limited use as an analytical category. The vagueness of the concept limits its usefulness. Hence, in this chapter, I use the term only as a descriptive adjective, akin to “mixed.” Neither “métis” nor “mixed” is to be understood, in the pages that follow, as an analytical category. In 1686, when Michel Jajolet de la Courbe visited the trading villages of the Lower Gambia and then journeyed overland through the Casamance to Cacheu, a century had passed since André Donelha’s visit.8 The cultural contact and assimilation described by Donelha and exemplified in the person of his trading partner, the Manding–Cape Verdean Gaspar Vaz,9 continued to characterize the elite in Gambian society. Men and women like Vaz moved between African and Luso-African society, bringing with them cultural traits that soon spread among the different groups living in those trading centers. The effects of a century of continuous interaction are clearly apparent in La Courbe’s account. The Lower Gambian trading communities that he visited in 1686 represented more thorough cultural assimilation than Casão had in 1585. Part of the difference may be due to the fact that Geregia and Vintang had mixed Manding-Bagnun populations under Bagnun political authority, whereas Casão in 1585 was a Manding community under Manding authority. The more centralized and hierarchical nature of Manding society very likely limited intercultural assimilation. However, the passage of a century of trade, intercultural contact, and resultant assimilation (e.g., Gaspar Vaz) had surely left its impact on the communities La Courbe visited. Vintang Creek, which today is of virtually no economic significance, was then an important commercial route. It was the northern terminus of a north-south trading corridor that connected The Gambia to the fertile Casamance and, south of that, to the slave-trading center of Cacheu. Vintang Creek lay a league upstream from the French trading post at Albreda and the nearby English fort at James Island. The English regularly sailed up the creek as far as Vintang village. La Courbe describes his arrival by boat: This entire country, along the river, appeared to us quite beautiful; on the right was the heavily forested kingdom of Fogny. . . . Bentam [Vintang] ap-

84

“portuguese” style and luso-african identity

peared from a distance to be a rather large town; it is located on the slope of a hill, with many trees, and several Portuguese-style houses, which surpass those of the Blacks, looked like churches; but as we drew near, we recognized that all that was only made of thatch, except for those of the Portuguese, which were built of earth and covered with palm fronds as large as tablecloths.10

A physical description of Vintang and Geregia is also provided by Francis Moore who, in 1732, was the Chief Factor for the English trading company in The Gambia: The river Vintain [Vintang], whose mouth is about a mile over . . . is navigable a great many leagues. About three leagues from the mouth is the town of Vintain, situated in Fonia [Fogny], and above that, on the same side of the river, is Geregia, at each of which places the Company have a factory, chiefly for elephants’ teeth, bees-wax and other dry goods.11

Moore further describes Geregia as “a small town . . . inhabited by Portuguese and Banyoons [Bagnuns].”12 The town had already diminished in size and importance since La Courbe’s visit fifty years before. The size and wealth of Vintang and Geregia derived from commerce, the same factor that attracted their diverse populations. In addition to the “Portuguese” communities of Luso-African merchants, there was a large Manding population. Among the Manding were long-distance traders who provided commercial ties to the Upper Gambia13 and, very likely, to Kaabu, located to the southeast. In 1682, La Courbe’s contemporary, the Portuguese merchant Francisco de Lemos Coelho, wrote that Vintang was “the best village on the river, having much trade in hides as well as in wax, ivory and blacks.”14 The wax came from nearby Fogny in the Casamance, whereas the hides and ivory were available both near the coast and in the interior and were likely the product of local as well as long-distance trade. Slaves, too, were obtained in Fogny.15 La Courbe writes, for example, that the king of Fogny was frequently fighting the Floups and that this warfare provided slaves to Vintang. But captives were also brought downriver from the Upper Gambia16 and perhaps also from Kaabu. For Juula merchants coming from the interior, Vintang would have been a convenient alternative market to Albreda. At the end of the seventeenth century, Vintang was a thriving trading town where diverse groups and individuals encountered one another and intermingled. There, as in other Gambian centers of commerce, African, European, and Afro-European customs were mixed together, sometimes in a single family. In addition, all the major religious traditions of West Africa coexisted peacefully at Vintang. One might say that here, in the seventeenth century, was an early instance of mixed cultures. At Vintang, La Courbe encountered a Spaniard who had settled and had married a daughter of the local ruler, the king of Geregia. This king also ruled over the populations of northern Fogny. He was a Bagnun and

“The People There Are Beginning to Take on English Manners”

85

is described as an “idolâtre.” That is, he followed Bagnun rituals although, according to his son-in-law, he believed in God and might have become a Christian if not for political considerations.17 The king’s daughter, however, was a Christian. Most of the “Portuguese” in the Vintang area, on the other hand, were Jewish.18 The date of 1686 is late for the persistence of the Jewish faith among Senegambian “Portuguese”; nevertheless, corroboration appears in a contemporary account by Jacques Le Maire (1682), who asserted that the Luso-Africans of the Petite Côte were “partie Juifs, partie Catholiques.”19 La Courbe did not mention a Manding presence at either Vintang or Geregia. This is a difficult lapse to explain, especially as Lemos Coelho wrote in 1682 that “the majority of the people in the village are Mandingoes and they are Mohammedans.”20 Fifty years later, Francis Moore wrote that Vintang “belongs to one of the emperors of Fonia . . . inhabited both by Portuguese and Mahometans, the latter having for their devotion a handsome mosque with an ostrich’s egg at the top.”21 There were also many non-Muslim Africans, including Floups and Bagnuns. In this borderland area between the largely Manding Gambia and the Casamance with its culturally heterogeneous, politically decentralized population, there were frequent and ongoing identity changes between Bagnuns, Floups, and Manding. Local African identities were so fluid that it is misleading to speak of these three groups as entirely separate entities. Cultural traits, including religious rituals, were shared among Floups, Bagnuns, and, to a lesser extent, Manding. A significant part of the northern Jola population of Buluf and Fogny today traces its ancestry, in part, to the Manding. At the end of the nineteenth century, some Manding migrated west and south into the Casamance, while some Jolas migrated as wage laborers or agricultural workers to The Gambia. Jola oral traditions, however, also tell of much earlier waves of Manding migration from Kaabu. While these traditions cannot be precisely dated, they clearly refer to a period before the seventeenth-century migration of Floups from Kasa north into Buluf.22 Throughout Buluf, local traditions recount an ancient migration pattern from the region of Pakao in the east into Kasa, the Floup heartland south of the Casamance River. This early population movement is sometimes identified with an initial immigration westward from the Mande state of Kaabu, a process that occurred before the arrival in Buluf of the first Jola-speaking immigrants from Kasa.23 The latter migration occurred at least 350 years ago. As the Floups were well established in Kasa long before some of them migrated north across the Casamance River, the historical memory of Mande origins must refer to a period well before the seventeenth century.24 For example, informants in Thionk-Essyl, the largest community in Buluf, maintain that the ancestors of those immigrants who long ago arrived from Kasa had in fact originated farther east at Kaour, in present-

86

“portuguese” style and luso-african identity

day Pakao. Pakao was historically populated by Manding and Bagnuns.25 The oeyi, the religious leader who was priest of the rain shrine, also originated in the east.26 The first oeyi is said to have been a Manding of the Mané family from Kaabu. Thus, this important Jola religious institution is associated with Manding origins. Diola communities in southern Buluf, too, claim their origins in the east, specifically in Kaabu27 or in Pakao.28 Yet it would be misleading to posit that waves of Floup migrants arrived from the east and swept away or assimilated Bagnun inhabitants. Oral traditions tend to represent gradual processes of cultural transformation as abrupt movements of discrete groups. A more nuanced reading of these same traditions would suggest that what is actually represented is a continuous process of interaction among Mande-, Bagnun-, and Jola-speakers. Implicit in this reading is a history of ongoing identity transformations. An awareness of this process is perhaps reflected in the interpretation offered by a local historian, Cherif Chamsedine Haidara, son of the Mauritanian cleric who first brought the Muslim religion to the Jolas. In his words, “The Jolas are said to come from the Bagnuns, who are said to come from the Manding.”29 As he discussed the early history of the Jolas, Haidara further observed that when the Bagnuns arrived in Buluf, “they began to become Jola.” This felicitous phrasing is, in fact, more accurate than an image of migration and conquest.30 RELIGIOUS MIXING Religion was a prominent aspect of mixed cultures at trading centers from The Gambia to Bissau. European merchants in the Vintang region included Portuguese and French Catholics and English Protestants. Indeed, there were representatives of all the religions that existed on the Upper Guinea Coast: Islam, Judaism, Catholic and Protestant Christianity, and African religions (Jola, Bagnun, Soninké). Significantly, however, seventeenthcentury sources mention neither strife nor active proselytizing or competition between followers of the various sects. Peaceful coexistence seems to have characterized the extraordinary religious diversity of these trading communities in the Lower Gambia region. There is a certain irony to the Portuguese name “Geregia,” which may derive from “Heregia” [heretics];31 one might say that the town’s population was so diverse that no matter what one’s religious orientation, it was full of heretics.32 In the Gambia-Casamance-Cacheu region, religious practices were characterized by the widespread sharing of shrines and their attendant rituals.33 In time of illness, even today, both Muslims and non-Muslims offer sacrifices at reputedly powerful Jola shrines.34 Elements of Muslim religion, too, were incorporated into local practices. A graphic example of such syncretism is the existence of an initiation mask covered with Arabic writing that invokes the name of God. This mask dates to the mid-eigh-

“The People There Are Beginning to Take on English Manners”

87

teenth century, long before the northern Jolas became Muslims.35 Likewise, in the early nineteenth century, non-Christian Floups sought Catholic amulets for protection against gunshots. In the seventeenth century, some southern Floups at Bolole had Christian names. It is conceivable that not only Christian charms but also some of the religion’s metaphysical concepts had already influenced individual Floups and Bagnuns. The Bagnun king of Geregia may have been influenced by such religious interaction. Although seventeenth-century written sources do not afford a direct view of the impact of Christianity on local African ritual, there is strong indirect evidence that Christianity profoundly influenced some Africans who, nevertheless, did not convert. Following his overland journey from Vintang to Cacheu, La Courbe visited the Bissagos Islands. On the island of Cazegut, he met a wealthy local merchant. Not only had this man built for himself a “Portuguese”-style house but he had also constructed a chapel: “He invited us to enter a little church that was nearby; he rang the bell and told us that he himself had built the chapel, even though he was not a Christian, because he liked Christians.”36 Perhaps this man was simply a brilliant businessman who knew how to make his European counterparts feel at home. But the chapel, complete with a bell to call to prayer, was probably not just a business investment. That this trader did not call himself a Christian means he was not baptized, perhaps because he was unwilling to give up that perk of wealthy Senegambians: many wives.37 He nevertheless had constructed a shrine where the most potent of Christian rituals, the Mass, could take place. In Guinea-Bissau today, powerful shrines from one society may be recreated or reproduced by members of other societies, thereby establishing a form of intercultural religious borrowing. It is conceivable that in like manner, this seventeenth-century trader, implicitly recognizing the spiritual power of Christian ritual, was attempting to gain access to the source of that power by building himself a Christian shrine. Christianity would thereby have exerted a profound impact on him, though he continued to follow his local religion. Nor was this merchant an isolated example. Other inhabitants of the Bissagos Islands were influenced by Christianity. For example, there were several chrétiens noirs who had been raised among the “Portuguese.”38 For those who followed both Christian and non-Christian rituals, religious identity was quite likely flexible and contextually defined. Many Christians continued to offer ritual sacrifices at local shrines. Similarly, as Amanda Sackur notes, eighteenth-century Christians in Saint-Louis and on Gorée Island had recourse to Muslim and local rituals.39 If some individuals followed both Christian and local rituals, others, like Gaspar Vaz had a century earlier, performed both Christian and Muslim prayers, depending on the context. La Courbe says of “certain Blacks and mulattos who call themselves Portuguese” that “most neither

88

“portuguese” style and luso-african identity

pray nor offer sala while others do both; for when they are among the Negroes they perform sala, and when they see whites, they take their rosary and make like them.”40 The implication is that for individuals who followed ritual practices from more than one religious tradition, their choice of ritual tended to be contextually determined.41 With regard to the assimilation of Christian rituals, one may ask who defined a “Christian.” There is evidence that some Africans who considered themselves Christian were not so viewed by the Portuguese priests at Bissau: “These good Fathers told me that it was almost impossible to convert these people; because when they were with Christians they attended Mass and they subsequently sacrificed to their idols or at least performed several other pagan ceremonies.”42 Just as Africans could be more or less Christian while retaining more or less of their local religion, so too “Portuguese” Christians could become progressively Africanized. It is sometimes impossible to distinguish in the historical record between Africanized Christians and Christianized Africans. Upriver from Cacheu, for example, La Courbe learned that “all along the river from Cacheu to Farim were located, at intervals, villages inhabited by Portuguese gourmettes who traveled deep into the country to seek wax.”43 The term “gourmettes” generally refers in French sources to “black Christians.”44 In this passage, such a reading gives the meaning “black Christian Portuguese.” But La Courbe contrasts these “gourmettes portugais” to “les Portugais de Cacheu,” thereby implying that they are different. If the “gourmettes portugais” are not “Portuguese” but are nevertheless Christian, then they would seem to constitute an intermediate category in terms of culture and religious orientation. Are they Christianized Papels or Africanized lançados? The point is moot. In the seventeenth-century Gambia-Bissau region, individuals could not always be situated at a distinct point along the continuum describing religious orientation. R E L I G I O N , P RO F E S S I O N , A N D I D E N T I T Y The complexity and intimacy of cultural interactions in seventeenthcentury Gambia, Casamance, and Guinea-Bissau are reflected in the existence of fluid and dynamic sociocultural identities that challenge any clearly delimited classification. Thus, for example, at the trading center of Guinala on the Rio Grande, La Courbe observed “a village consisting of numerous Portuguese houses.” “Then we went ashore at the house of Signor Patrisio-Parese, a white Portuguese, the son of a Dutchman and a mulatto woman.”45 Parese had evidently taken his name and his “Portuguese” identity from his mother.46 The descriptive term “Portugais blanc” is intriguing, since his mother was mulatto and neither parent is described as Portuguese. That the son of a mulatto could be “white” did not accord with usual French or English categories. “White” may reflect the local

“The People There Are Beginning to Take on English Manners”

89

perception at Guinala of Parese’s mother’s identity; she was surely AfroEuropean and hence probably Luso-African. If she was also a trader, she may have been referred to as “white.” But if she was, then it is difficult to understand why La Courbe refers to her simply as “une mulâtresse.” Most likely “blanc” refers to Signor Parese’s socioeconomic position; he was the first person La Courbe visited in Guinala and therefore was undoubtedly an important merchant. His social stature and his profession were likely also determinant elements in his “Portuguese” identity. Social status and occupation, rather than biological heritage or physical appearance, were the primary markers of being “white” and “Portuguese.”47 Ultimately, physical characteristics were irrelevant to one’s social and economic position and hence had little or no bearing on whether one was considered “Portuguese.” This is made abundantly clear by Labat, citing Brue, who visited Bissau fifteen years after La Courbe and who described Guinala in these terms: “It is a large village of Portuguese who have long been established there from father [mother?] to son . . . significant both for the number of its houses and for its inhabitants: white Portuguese, black Portuguese, mixed Portuguese, tanned, and mulatto Portuguese.”48 What these Portuguese had in common was certainly not their physical appearance. Rather, it was their profession and their culture. From a Luso-African perspective, these sociocultural characteristics defined them not only as “Portuguese,” but also as “whites.” As La Courbe observes, “Although they are black, they nevertheless maintain that they are white, by which they mean to signify that they are Christians like the whites.”49 Even so careful an observer as La Courbe is, however, inconsistent in his application of these labels. He refers to Manuel Alves, the “Portuguese” governor of Geba, a man who had received the prestigious Order of Christ from Lisbon, as “nègre chrétien” rather than by the expected— and appropriate—“Portugais.” This inconsistency quite possibly reflects the fact that La Courbe is himself caught between two perspectives or two discourses on identity. In consequence, sometimes he uses “white” in the European sense of physical description (by this measure Alves is “nègre chrétien”), while at other times he adopts the Luso-African definition (by which Parese is a “white Portuguese”). M A R I TA L A L L I A N C E S Among wealthy traders, the intercultural assimilation characteristic of Vintang society is clearly expressed in marriage patterns. These unions incorporated European and Euro-African merchants into a local kinshipbased society. At Albreda, Geregia, and Vintang, “Portuguese” women, many of whom were either important traders or, like the wife of the Spaniard at Vintang, the daughters of local rulers, contracted alliances with European men.50 At Albreda, the famous trader and courtesan La Belinguère51 established a series of alliances with several European mer-

90

“portuguese” style and luso-african identity

chants of different nationalities. Such unions served to secure the commercial interests of both partners.52 La Courbe, in his tongue-in-cheek, rather prissy description, alludes to the intersection of La Belinguère’s personal and professional life: “She had a noble manner and a refined tongue and she spoke good Portuguese, French, and English, a certain indication of the extensive commerce she had carried on with all these nations.”53 La Belinguère, like the wife of the Spanish trader at Vintang, was the daughter of a local ruler. La Courbe depicts her as having taken on the cultural markers of “Portuguese” identity. She was a successful trader, she spoke Portuguese and lived in a “Portuguese”-style house, and she dressed in cloth imported from the Cape Verde Islands. No mention is made of her religion, but in other respects, she is someone who has made the transition to a “Portuguese” identity. Also in Vintang, a Luso-African woman, a mulatto, played a vital role in the economic integration into the community of European traders through her alliances with them. La Courbe paid her the expected social visit, too. As at Albreda, the meeting took place in the vestibule of a “Portuguese”-style house. The meal was served “à la manière des Portugais,” complete with imported red glassware.54 The hostess was married to an English merchant named Agis(?), who had obtained commercial benefits from this union. In fact, economic self-interest evidently led Agis to remain with his wife in spite of her public affair with the alquier [mayor] of Geregia.55 Such unions, at least in the perception of other European traders, brought economic advantages to both parties. E T I Q U E T T E : M I X E D M A N N E RS Seventeenth-century society in the Vintang area may be described as mixed not only in terms of intermarriage but also with respect to the widespread assimilation and merging of material culture and social practice. While economic factors played a role in the proliferation of marriages between European merchants and women from the local social elite, a direct outgrowth of these unions was the development of a social etiquette, part of an elite material culture that merged elements of diverse societies. Housing styles, clothing, table manners, and comportment were all transformed under the influence of close contact, marriage, and mutual borrowings between individuals and groups. The Bagnun king of Geregia shows how the most disparate elements could come together to create a royal etiquette. These syncretic manners served above all to welcome foreign guests, who were frequently traders. To host strangers well constitutes an important aspect of Gambian and Casamance social values.56 It is widely considered an honor, even today, to entertain visitors, and to perform this task well is a source of pride. In the seventeenth century and, indeed, even early in the twentieth century, trade

“The People There Are Beginning to Take on English Manners”

91

was carried on in the context of personal relations between client and patron, or host and guest. Hence, locally prescribed etiquette for welcoming strangers was imbued with both social and economic significance. Following his 1682 visit, Jacques Le Maire wrote that the Senegambians “are hospitable and never allow any foreigner to pass through without giving them something to drink and eat, even for several days.”57 This passage is particularly striking in view of Le Maire’s generally pejorative view of Africans. As a foreign visitor and a merchant, La Courbe was expected to pay courtesy calls on political leaders or officials in each community he visited. They, in turn, were obligated to receive him graciously and generously. This was the context in which he visited the king of Geregia: The alquier led us to see the king of Geregia: we found him at the entry to his house; he was a small and robust man; he had on a Portuguese cap and an African robe; in his hand he carried a Spanish style sword upon which he leaned. After greeting us, he bade us enter his house and had us sit on wooden seats; I told him the reason for my visit. . . . I made him a present of a little iron and some schnapps and, shortly after, he invited us to enter another room, for his house is built in the Portuguese style; we found lunch waiting. . . . He joined us to eat, as did his wife, which showed me that, in this place, they have begun to take on English manners.58 (emphasis mine)

The king’s clothes combined local African robes with a Portuguese hat. Such hats were frequently given to local notables by Lusitanian officials as an emblem of authority, a symbolic function that they continued to serve as recently as the beginning of the twentieth century. The sword, too, was part of the regalia that indicated his rank; one can only guess as to its provenance, which perhaps was not unrelated to the place of origin of his son-in-law. Manding rulers in the nearby state of Kaabu possessed iron staffs of office made by Mande blacksmiths; it may be that this local Bagnun ruler was inspired by Manding tradition to use his sword as an emblem of authority and power.59 The European origin of his sword would also have emphasized the king’s association with long-distance traders. The purpose of La Courbe’s visit—to discuss the establishment of a trading post—has its place in the complex routine of social etiquette. The two men discuss business before the Frenchman makes the requisite gifts; this conversation takes place in the first room of the house. La Courbe does not specify the location but, as this was a “Portuguese”-style house, their initial meeting almost certainly took place in the vestibule or on the veranda. After business comes hospitality; the king invites his guest into an interior room for dinner. This is an unusual practice, one unheard of north of The Gambia, where Europeans did not have access to the private interior space of their host’s house, which was reserved for family. And the

92

“portuguese” style and luso-african identity

most unusual practice of all, as La Courbe observes, is the fact that the king and his wife both join the Frenchman to eat. Men and women do not eat together in either Bagnun or Floup society. The king of Geregia may indeed have been following a European example. The social customs accompanying this international business meeting on the banks of Vintang Creek represented a microcosm of the society itself. Late-seventeenth-century trading culture, with its mixture of Spanish, Portuguese, French, and English goods and customs, not to mention the ongoing cultural assimilation between Bagnuns, Floups and Manding, illustrates the degree to which Gambian society was characterized by cultural assimilation and appropriation. The same process that led to the creation of a distinctive social etiquette among the local elite also molded religious practices. This process also defines the intercultural adaptation and appropriation that led to the development of the distinctive architectural style called “Portuguese,” but which, in reality, incorporates African techniques and building materials along with (possibly) some Portuguese stylistic elements. And, while our knowledge of the Creole language before the eighteenth century is extremely limited, this aspect, too, of coastal trading societies reflects a dynamic of assimilation and adaptation. The end result was indeed a mixed culture. A C O N C E P T UA L M O D E L O F I N T E R AC T I O N I N M I X E D C U LT U R E S Sixteenth- and seventeenth-century trading centers from The Gambia to Bissau were characterized by extensive cultural interaction that fostered fluid and sometimes multiple identities. A study of the social manners reflected in La Courbe’s account of a visit to the king of Geregia illuminates the microprocesses of cultural interaction and assimilation that were the building blocks of these flexible and fluid identities and that led to the creation of new, hybrid cultural patterns. For the Lower Gambia, Casamance, and coastal Bissau, a model of essentially fluid cultural patterns that dynamically interact and mutually transform one another more closely accords with the region’s history than does the image of discrete ethnic groups. In this model, as with J.-L. Amselle’s concept of cultures that form “un ensemble mouvant” [a changing ensemble],60 dynamic process rather than fixed sociocultural categories serves as the foundation for both social interaction and the formation of group identities. A similar approach to seventeenth-century cultural interaction is proposed by Gruzinski. In his study of Spanish-Aztec interaction in Mexico, he writes: “An understanding of métissage must confront intellectual habits that tend to prefer monolithic ensembles to intermediate spaces. It is easier, in effect, to identify solid blocks rather than the spaces between them that have no name.”61 This is a telling point. Gruzinski further associates the métisse cultures of sixteenth- and seventeenth-century Mexico

“The People There Are Beginning to Take on English Manners”

93

with Iberian expansion and the creation of a world economy under Spanish domination in the sixteenth century. Throughout part of the sixteenth century, while Portugal was under the Spanish monarchy, the Guinea Coast was part of the same commercial empire as Mexico. Indeed, Portugal’s seaborne empire was part of not only an economic but also a cultural world system. Yet in precolonial West Africa, unlike in Mexico, European culture was ultimately subsumed by African culture. The reasons for this were complex. A relatively small number of Portuguese settled in Senegambia and, by the eighteenth century, their Luso-African descendents were isolated as the Portuguese colonial empire declined. Thus, although these two cases followed a similar model of cultural interaction, the historical circumstances differed widely and the results were hardly comparable. Another significant difference may have been the fact that African culture, particularly the religion of the Muslim traders with whom the Portuguese and Luso-Africans had extensive contact, was more familiar to the Europeans than was pre-conquest Aztec religion. Conversely, the more extensive efforts to convert the indigenous residents of Mexico after the Spanish conquest of a centralized state had a more lasting, although mitigated, Christianizing impact. In Senegambia, the strong Muslim presence, especially among traders, afforded the Portuguese a familiar reference by means of which they might incorporate Africans into their accustomed intellectual universe. Iberia had experienced centuries of Muslim occupation. No such reference existed for the Portuguese or Spanish who were confronted with Native American cultures in the New World. On the Guinea Coast, European familiarity with an important aspect of local culture and beliefs—the Muslim religion—undoubtedly facilitated early interaction. It may also have played a role in the rapidity with which two-way cultural assimilation between Africans and Portuguese manifested itself. Cultural assimilation in sixteenth- and seventeenth-century Senegambia is in certain respects similar to the cultures métisses analyzed by Gruzinski in late-sixteenth-century Spanish Mexico. Both regions, as well as the Cape Verde Islands and the sugar-producing regions of Portuguese Brazil, were important components of the Iberian commercial empire. Additional research may enable scholars to further refine the familiar hypothesis that Iberian society, with its history of coexistence and interaction between diverse cultures, was more open than northern European cultures to intercultural assimilation in both Africa and the New World. Nevertheless, and in spite of broad similarities between cultural processes in Senegambia and Mexico, the two regions should also be viewed in their specific social and historical contexts. A more narrowly focused approach underlines those areas where each situation was unique. It is important to seek particular characteristics of Upper Guinea Coast society that may have facilitated the growth in the Gambia-Cacheu region of

94

“portuguese” style and luso-african identity

the mixed cultures that are the subject of the present study. Indeed, intercultural assimilation and fluid identities were characteristic of the entire Senegambia-Bissau region. Cultural mixing was a characteristic of societies throughout the Gambia-Bissau region before the arrival of Portuguese traders in the mid-fifteenth century. Evidence for this history, while circumstantial, is broad and convincing. First, the interactive and dynamic nature of local society has long outlasted the Portuguese presence in the Gambia and Casamance regions. Casamance society is defined, even today, by its assimilative and transformative qualities. Second, although the nature of pre-contact Upper Guinea Coast culture can only be known in its broadest outline—ironically, Portuguese written sources constitute the earliest historical documentation for this area—oral traditions in the Casamance suggest that cultural mixing and the resultant transformations in identity are a longestablished dynamic that predates the Portuguese commercial presence. Local traditions north of the Casamance River attest to the transformation of an earlier Bagnun-speaking population into the ancestors of the present Jola-speaking community. These oral traditions cannot be precisely dated, but comparison of the traditions to early Portuguese sources suggests that the cultural assimilation of the Bagnuns by the Jolas north of the Casamance River began no later than the sixteenth century.62 The history of precolonial cultural assimilation in the Lower Casamance is a central theme in my previous work63 and it is not my intention to cover the same material here. A brief outline may, nevertheless, help to date more accurately early cultural transformations in the region. In northern Fogny, the seventeenth-century warfare between Floups and Bagnuns mentioned by La Courbe was part of a progressive expansion of Floups into the formerly Bagnun territories north of the Casamance River. This expansion, resulting in the assimilation of Bagnun communities by Floup newcomers, is attested to by oral traditions among the Buluf-Diola, descendents of this amalgamation. Buluf traditions date the cultural transformation to a period more than eighteen generations ago, a chronology that is consistent with Almada’s late-sixteenth-century description (ca. 1594).64 South of the Casamance River, the Bagnun or their ancestors also constituted an autochthonous population, at least in the sense that no historical memory exists of any earlier group.65 Here, too, they were gradually assimilated by other groups, including Floups, Balantas, and probably Mande-speakers in Kaabu.66 Cultural assimilation thus characterizes the entire region of Casamance and northern Guinea-Bissau from the earliest accessible oral record through the earliest written sources and throughout the precolonial period. Historical memory among the northern Diola also asserts that they and the Manding have intermarried so much that in this borderland region individuals cannot be clearly identified as belonging to one or the other group. An earlier historical layer of Mande immigration into the

“The People There Are Beginning to Take on English Manners”

95

Lower Casamance is reflected in the tradition, preserved in Thionk-Essyl, that the institution of kingship (the oeyi, or priest-king) was originally brought to the Lower Casamance from Kaabu by a Mande named Malamanka Kaabu.67 The fact that the most important institution of local spiritual authority is recognized by some Jolas as being of Mande origin speaks eloquently of the long history of assimilation between Jola-speakers and Mande-speakers. In addition, the limited information afforded by historical linguistics and by the study of place-names is consistent with the thesis that cultural assimilation and transformation have been an integral part of regional history since the first half of the second millennium c.e. Stephan Bühnen suggests, for example, that the names “Ziguinchor” and “Geregia” both derive from Bagnun terms.68 Taken together, this oral historical and linguistic evidence suggests that the Portuguese in Senegambia and their LusoAfrican successors adapted themselves to a preexisting Gambian-Casamance pattern of cultural interaction. European acceptance of local cultural patterns is all the more likely in view of the small numbers of “Portuguese” living along the coast, even as late as the seventeenth century.69 La Courbe and Lemos Coelho visited Vintang at the height of its lateseventeenth-century commercial prosperity. Shortly afterward the town experienced a decline, which both Francis Moore and Père Labat chronicled. It has long been acknowledged that Labat extensively plagiarized La Courbe’s manuscript in his description of the journey from Geregia and Vintang to Cacheu.70 While much of the corresponding portion of La Courbe’s narrative is lost, comparison to Labat of the surviving manuscript has enabled Boulègue to demonstrate that Labat complemented his unacknowledged use of La Courbe with information that can be dated to about 1705.71 Labat’s description of Vintang clearly derives from La Courbe but is prefaced by the assertion that “the town or village of Bintam was formerly larger than it is at present.”72 In comparing his 1705 source to La Courbe’s detailed description of the town, Labat must have concluded that the town had diminished in size. Moore’s subsequent description of Geregia as “a small town” and his cursory treatment of Vintang suggest that the commercial fortunes of the town continued to decline over the next two decades. The decline continued; by the time of Demanet’s visit to The Gambia in 1764, the French apparently no longer maintained a trading presence at Geregia.73 The diminishing importance of commerce at Vintang and Geregia in the eighteenth century led to the decline of the local merchant class. This professional elite was identified with the distinctive mixed culture that had developed during the sixteenth century and that, by the late seventeenth century, was a noteworthy feature of trading centers along the lower Gambia River. When Demanet traveled upriver in 1764, the de-

96

“portuguese” style and luso-african identity

scendents of the Luso-African traders still retained the memory of their Portuguese ancestors, but their family traditions were virtually all that remained of their earlier prosperity. Even these traditions were already only a dim recollection of “the time when the Portuguese were the masters of the Gambia River and maintained trading centers all along the African coast.”74 The period of extensive cultural mixing was already past its zenith. With the decline in their commercial fortunes and the loss of their distinctive professional role, the Gambian “Portuguese” (whatever their European ancestry) became increasingly African. Yet, even today, two centuries later, an understanding of this history and of its dynamics is crucial if one is to appreciate the distinctive cultures of the region from The Gambia to Bissau.

FIVE

Senegambia from the Mid-Eighteenth Century to the Mid-Nineteenth Century

Through the late eighteenth and the early nineteenth centuries, the French imposed increasingly rigid identity categories on Senegambians. This served both to deny the Luso-Africans their identity as “Portuguese” and to contest the widespread Senegambian usage of the concept of “white” as a socioeconomic label rather than as an a priori category based on skin color. As the French extended their influence south into the Casamance in the nineteenth century, political factors played an increasing role in their identity discourse. At the same time, in this discourse, local domestic architecture became a powerful index reflecting the presumed level of civilization of different Senegambian societies. During the first half of the nineteenth century, commerce from The Gambia south through the Casamance region became increasingly competitive as French and English traders competed with the Luso-Africans for control of the coastal export trade. Manding merchants continued to dominate the trade from the interior to the coast. In 1836, the commandant at Gorée, Dagorne, who was leading a French mission to Casamance, observed that “the Manding . . . are outstanding merchants and . . . they are the exclusive agents of commerce; it is through their intermediary role that all the products from the interior arrive at the coast and that European goods, on the other hand, circulate in the interior.”1 The commercial role of the Manding was particularly important in the Gambia-Soungrougrou region, near the former commercial center of Vintang. Further south, a few Wolof traders were beginning to establish themselves on the island of Carabane on the south bank of the Casamance River and in Jola villages in the Lower Casamance.2 The establishment of a French fort at Sedhiu in 1837, along with the development of a trading post at Carabane beginning in 1836, significantly increased commercial competition with the Luso-Africans throughout the region. Nevertheless, at mid-century, “Portuguese” traders, some based in Ziguinchor on the south bank of the

98

“portuguese” style and luso-african identity

Casamance River and others based in Bissau, still played a significant role in commerce throughout the Lower and Middle Casamance.3 The Portuguese commercial and administrative post at Ziguinchor was the only significant Luso-African community north of Cacheu. From this settlement, “Portuguese” merchants carried out a limited trade with Manding traders along the left bank of the river and, on the right bank, in the Middle Casamance. On the occasion of his exploratory visit to the Casamance, Dagorne wrote of the “Portuguese,” “They send pirogues upriver to trade a small quantity of . . . products in villages along the river.”4 In return for salt, the Luso-Africans obtained cotton; they also purchased rice, beeswax, and ivory. The “Portuguese” acquired trading goods of European manufacture from the English in The Gambia. However, high import and export duties in the Portuguese territory prevented this commerce from remaining competitive once other European nations had established trade in the region.5 In 1836, Ziguinchor itself was an unimpressive village of sun-dried earthen houses, surrounded by a barbican and four earthen bastions that housed a few old cannon: The walls of the houses consist of dried mud. The town, if one dares use the term for such a pitiable grouping of miserable huts, is defended by a tata in the form of interlaced branches with, at each of the four corners, a sort of bastion, walled in dried clay with embrasures through which pass the volleys of a few old cannons without gun carriages. . . . It is a miracle that these walls don’t simply tumble down.6

Dagorne’s view of the “Portuguese” inhabitants of Ziguinchor was that they were as decadent as the town in which they lived. He attributed their willingness to live in precarious and decrepit conditions to a natural apathy that derived from their African heritage: “They are resigned to this by habit and by the natural apathy that derives from their origin, for they are all black, even the Povedor [local governor].”7 In a sense, architecture reflected the absence of motivation or initiative; indeed, it embodied the character of the townspeople. Since in Dagorne’s estimation these qualities were inborn, he viewed the buildings as a reflection of the natural apathy of the black “Portuguese.” The earliest illustration of Ziguinchor is a watercolor sketch of the town made thirteen years after Dagorne’s visit by Hyacinthe Hecquard (Figure 18). Hecquard, an officer in the spahis sénégalais, precursor to the military force called the tirailleurs sénégalais, arrived in the Casamance in 1850 to begin an overland journey to the Futa Jalon.8 He stopped first at Carabane, where he visited with Emmanuel Bertrand-Bocandé, the French résident and the most knowledgeable nineteenth-century ethnographer of the Casamance. After leaving Carabane, Hecquard stopped briefly at Ziguinchor. The watercolor that he sketched, depicting the town as seen from the river, was never published by the artist. This work is one of a se-

Senegambia from the Mid-18th to the Mid-19th Century

99

18. Zéguinchor—Comptoir portugais (Cazamance), 1851. Hyacinthe Hecquard, watercolor. Courtesy of Frobenius-Institut, Universität Frankfurt am Main, Bilder-Archiv.

ries of thirty-one watercolor sketches and pencil drawings that are now preserved in the archives of the Frobenius-Institut in Frankfurt am Main.9 Hecquard’s drawing, labeled Zéguinchor—Comptoir portugais (Cazamance),10 depicts, in the foreground, half a dozen small boats with masts, plying the river. Snuggled behind a low wooden palisade that hugs the shoreline are a dozen tightly clustered rectangular houses with thatched roofs. Scattered among these buildings are six smaller round structures. The round houses are surrounded by posts, representing the forkillas that support encircling verandas. A flag flies from the roof of one building. At the left (eastern) edge of the village, one rectangular structure is crowned, not by thatch, but rather by a flat sloping roof. This flat roof calls to mind Lemos Coelho’s enigmatic seventeenth-century description of the cumbetes at Cacheu, which he describes as having mud roofs. Perhaps this is a depiction of just such a structure. Besides these few architectural features, Hecquard has provided little detail; one cannot tell, for example, whether the buildings are whitewashed. The few visible architectural elements are consistent with written descriptions of “Portuguese”-style dwellings. This sketch is complemented by Hecquard’s published description of Ziguinchor, which indicates that he drew only a portion of the village. The community consisted of about 100 rectangular buildings. Their thatched roofs could be removed during the dry season to reduce the risk of fire, a feature that Mollien also observed in Bissau in 1818.11 Hecquard observed no hierarchy of scale among the buildings. The governor’s house was not

100

“portuguese” style and luso-african identity

significantly different from any other dwelling, and the lone chapel had long since fallen into disrepair.12 The chapel was apparently often in ruins; fourteen years after Hecquard’s visit, in 1864, Père Lacombe of the Holy Ghost Fathers passed through Ziguinchor. He observed: “For a year there has been no church; the downpours of the rainy season caused it to collapse, because here one builds only of clay-soil. Local resources do not permit the people to reconstruct it.”13 The fortunes of Ziguinchor continued to decline during the final decades of Portuguese control. In 1874, Father Sène, also of the Holy Ghost Fathers, found the village to be impoverished and run down and evidently also totally unkempt: “Nothing could be sadder than the appearance of Ziguinchor. . . . It is only an agglomeration of huts, roughly constructed and covered with thatching. Along the quais, in the streets, everywhere, is complete chaos. . . . The least village of the Blacks does not show such impoverished scenes.”14 In 1837, however, the situation was not yet so desperate for the “Portuguese” in Ziguinchor or the surrounding territory. Along the south bank of the Casamance River, trade was still largely in “Portuguese” hands. Returning from Sedhiu in that year, Dagorne observed pirogues belonging to Luso-Africans who were evidently trading with the Balantas.15 In 1857, when Bertrand-Bocandé traveled upriver to the Middle Casamance, he found this area to be frequented by “Portuguese traders who arrive there by land.”16 Luso-Africans from Bissau, Cacheu, and even as far afield as Sierra Leone brought kola and palm oil to Carabane to sell to the French. Ziguinchor was the base from which “Portuguese” traders acquired Jola rice in exchange for cattle and cotton cloth. Much of this rice was then traded to the French at Carabane, leading Bocandé to write, “The Portuguese of Ziguinchor have become the representatives of French commercial interests in Casamance.”17 These mid-nineteenth-century descendents of Luso-African traders had adjusted their activity to changing patterns of European involvement along the northern Upper Guinea Coast. Nevertheless, they continued to play the long-established commercial role, developed by their Portuguese and lançado ancestors, of intermediaries between Africans and Europeans. North of the Casamance River, along the Soungrougrou River, a Manding population that was expanding westward from the Middle Casamance and south from The Gambia had begun to impose political control over local Jola-speaking populations (referred to as “Yola” by French visitors). Traveling along the Soungrougrou in 1861, a French administrator from Sedhiu accurately prophesied: “The Manding will soon take over this country.”18 Attesting to the climate of instability that prevailed along this waterway, Jola villages were generally fortified against attack: “The Yolas take pains to surround their clusters of houses with a palisade [illegible] a strong resistance.”19 This area, virtually unknown to French officials de-

Senegambia from the Mid-18th to the Mid-19th Century

101

19. Pirogues on the shore of the Casamance River. Photo by Peter Mark.

spite its proximity to Sedhiu, was still visited by “Portuguese” traders, who continued to use pirogues to navigate the shallow marigots that provided access to remote Soungrougrou villages. In 1857, Bertrand-Bocandé described the “Portuguese” pirogues, “hollowed from a single tree trunk . . . the gunwales raised by thick boards hewn with an axe.”20 Even today, the Jolas of Buluf north of the Casamance River continue to make pirogues from the trunks of hollowed-out busaanab (silk-cotton) trees. Similar vessels still ply the lower Casamance River (Figure 19). The Soungrougrou River runs north from the Casamance practically to The Gambia; its northern reaches are close to Vintang-Geregia. These trading centers were largely abandoned by the nineteenth century. The influence of the earlier commerce was nevertheless still reflected in a greater openness among local peoples toward foreign traders. Trading routes followed by Manding merchants from The Gambia and from the Upper Casamance region joined near the Soungrougrou, stimulating the local production of cotton cloth for export.21 In 1861, the French administrator

102

“portuguese” style and luso-african identity

20. Ruins of the nineteenth-century French trading post at Albreda on the lower Gambia River. Photo by Peter Mark.

at Sedhiu wrote: “This part of the Songrougrou, being close to the routes to The Gambia and the Upper Casamance, and having been often visited by [courtiers? / illegible] who carried out quite a successful trade before our merchants even had the idea of sending traitants [traders] there, is far more civilized than the lower Songrougrou.”22 Along the Soungrougrou, the “Portuguese” did their best to hinder the establishment of French commerce. In one instance, a French administrator who had traveled up the river to meet and sign treaties with local village leaders was abandoned by a “Portuguese” trader who was to have served as his guide and interpreter. Unable to communicate with the local Jola-speakers, the Frenchman was forced to abort his mission.23 Ultimately, however, France’s increasing commercial and military superiority led to the extension of colonial control based at Saint-Louis and Gorée Island. The Luso-Africans who lived and traded in the Lower and Middle Casamance soon became pawns in this colonial encounter. Farther north, along the lower Gambia River, possession of the trading escales at Albreda and James Fort switched back and forth between

Senegambia from the Mid-18th to the Mid-19th Century

103

France and England throughout the second half of the eighteenth century, reflecting the shifting fortunes of war between the two European nations.24 Access to The Gambia was barred to France several times before 1815. As a result of the continuing conflict, the fortunes of Albreda (Figure 20) and the trading centers along Vintang Creek declined. Even after the restitution of its Senegalese possessions in 1815, France was never able to reestablish its previous commercial importance in this region. Finally, in 1857, France ceded Albreda to England.25 F R E N C H I M AG E S O F T H E “ P O RT U G U E S E ” AT B I S SAU South of the Casamance, Portuguese administrative authority remained secure throughout the nineteenth century. At Bissau, a Portuguese fort built of stone and masonry dominated a town of about 100 houses that was inhabited by a population of Luso-Africans. Dagorne wrote: The town, as it is called, consists of about 100 square houses of sun-dried earth, covered with an earthen platform in the dry season and, in the rainy season, topped with a sharply canted thatched roof. These houses are inhabited by a totally black population who are supposedly Catholic—goodness knows why—but who, with the exception of their houses, have the same customs and the same manners as the surrounding Papels, from whom they stem.26

Along the river stood three or four stone houses belonging to Europeans. The description of the domestic architecture adds a bizarre twist to Mollien’s 1818 report that the roofs of “Portuguese”-style houses were removed during the dry season. An anonymous contemporary of Dagorne also described these temporary rainy-season roofs; they were so immense that they blocked the cannons of the fortress.27 But the most striking aspect of Dagorne’s own report is the fact that he perceives the population of Bissau as consisting of three categories—Papels, black “Portuguese,” and Europeans—each group with its characteristic architecture. In this description, it is only the form of their houses that differentiates the “Portuguese” from the local Africans. Physically, the two groups are indistinguishable, and even if the “Portuguese” claim to be Catholics, their religious practices are the same as those of the Papels, “from whom they stem.” Dagorne does not mention language. The only marker of “Portuguese” identity that he accepts is their architecture. In every respect except their houses [au logement près] they have the same customs and traditions as the Papels. Yet ultimately Dagorne does question the Luso-Africans’ identity; in concluding his discussion of Bissau, he refers to the Luso-Africans not as “Portuguese,” but as the “noirs gourmettes qui demeurent dans la ville” [the black gourmettes who live in the town].28

104

“portuguese” style and luso-african identity

A few years later, in 1845, a report published in the Revue Coloniale develops this line of argumentation one step further. The author divides the population of the “Portuguese settlement” [établissement portugais] at Bissau into two components, the Papels and the gourmettes. The gourmettes are individuals of diverse nations; they are more or less Christian, and they are traders: A people of diverse nations, all more or less Christian . . . these are the Gourmettes. They never do any agricultural labor, for their natural element is the sea. They outfit all of the numerous pirogues that ply the Géba River to acquire the products that abound in this region . . . wax, ivory, skins, palm oil. . . . Some of the Gourmettes sail with ships that travel as far as Europe; others are workmen of all sorts and they often serve as pilots for the boatmen of diverse nations.29

In this report, the “Portuguese,” or Luso-Africans, have largely disappeared, subsumed into the category of gourmettes. While they are “more or less Christian” and although by profession they are sailors and traders, the members of this group are not recognized as “Portuguese” or even as “black Portuguese” by their French observer. In the nineteenth century, Europeans progressively contested the Luso-Africans’ “Portuguese” identity until, ultimately, as exemplified by this French report, the “Portuguese” began to be defined out of existence. In a precisely contemporary account, the Abbé David Boilat, himself a Senegalese from the Petite Côte, presents a less pejorative interpretation of the term “gourmette.” In his Esquisses Sénégalaises, Boilat imputes to the French at Gorée and Saint-Louis an openness regarding skin color that reflects his own attitude: “Color prejudice is unknown in Senegal.”30 Boilat proceeds to enumerate the groups that, together, constitute “les habitants.” First are “les mulâtres,” followed by “les gourmets ou noirs baptisés. . . . sont aussi traitants: leurs femmes sont appelées signares.”31 Boilat further refers to the gourmettes of Gorée and Saint-Louis. The gourmettes are traders, they are Christians, and at least some etymological connection to the Creole language is implied by the term “signare.” The term is thus close in meaning to the contemporary appellation “Creole.” In its broadest sense, when applied to the Luso-Africans of Bissau, the term clearly encompasses the descendents of the “Portuguese.” In southern Senegambia, the growing political rivalry between France and Portugal played a role in the rejection by French observers of the Luso-Africans’ self-definition as “Portuguese.” By dismissing these agents of Portugal’s commercial empire as “Africans” or “gourmets,” French officials challenged the Portuguese claim to the territory. If Lisbon’s representatives in Ziguinchor were not really Portuguese, then Portugal’s claim to territorial sovereignty over the Lower Casamance could be challenged. Increasingly in the second half of the century, political considerations assume a central role in the determination of “Portuguese” identity.

Senegambia from the Mid-18th to the Mid-19th Century

105

In at least one respect, the anonymous report of 1845, like Dagorne’s perception of the absence of any distinction between Luso-Africans and their African neighbors, contains an element of truth: there were no clear boundaries—physical or cultural—between the two groups. Indeed, through much of the nineteenth century, as at the time of Donelha’s sixteenth-century visit to Casão, the movement of individuals back and forth between “Portuguese” and other African identities was undoubtedly quite common. Even though the “Portuguese” make only brief appearances in nineteenth-century French sources, historical evidence of such transformations of identity does exist. I D E N T I T Y T R A N S F O R M AT I O N S I N P O L I T I CA L LY C O N T E S T E D S PAC E S In 1849, a trading vessel accompanied by a French trader from Gorée was attacked by a group of marauders, evidently Bijogos from Cazegut in the Bissagos Islands. The official French report referred to the attacking party as “the people of Baje-coot [elsewhere ‘Cazegut’] who had chosen Manuel Gara as their chief.” The reasons why the Bijogos selected this man as their leader are not stated. In the report, he is referred to as “this pirate chief” and as “a Portuguese from Bissau.”32 Manuel Gara (or Sara) had become the leader of this group of Bijogos, a role that was aborted by the naval vessel that investigated the incident. Had the French not intervened, he would effectively have become a member of the Cazegut community where he was also the chief. Ultimately, the fact that he was from Bissau, where he was a “Portuguese,” would probably have been forgotten. The process of identity transmutation, or of Gara adding Cazegut [Bijogo] identity to his “Portuguese” identity, was here in process but perhaps not yet completed. This was not a unique example of a Luso-African becoming the chief or leader of a local non-“Portuguese” community. In 1796, the naturalist Adam Afzelius, who was visiting the Rio Pongo, reported several instances of “Portuguese” who had become local chiefs. One of the headmen was a “white Portuguese” named Gomez, who had been educated in England. According to Afzelius, there were other local chiefs whose “Portuguese” identity was far more attenuated. Afzelius wrote disapprovingly of these men, “The Native Chiefs of Portuguese extraction are more cruel and, if they can, they catch [an adulterer] and sell him for a slave.”33 Although these individuals were descended from the Portuguese, they had become local chiefs. Their involvement in the slave trade recalls the fact that one of the defining characteristics of “Portuguese” identity was to be a merchant. Perhaps they remained “Portuguese” for as long as they maintained their profession as slave traders. At the end of the eighteenth and into the nineteenth century, interaction between “Portuguese” and other local Africans continued to engender a blurring of distinctions

106

“portuguese” style and luso-african identity

throughout the area of the Luso-Africans’ former commercial dominance. Generally, clearly defined borders did not exist between “Portuguese” and non-“Portuguese.” In mid-nineteenth-century Casamance, the gradual establishment of French authority marginalized the “Portuguese” and reduced their capacity to exercise a political role in rural areas. The implantation of French authority also served to underscore the fact that identity, at the dawn of colonial administration, was determined in some measure by political factors. Thus, whether a given individual was identified as “French,” “Portuguese,” or African (Balanta, Felupe, etc.) depended in part upon that person’s affiliation with competing colonial authorities. An episode from 1879 at Adeane, south of the Casamance River, illustrates the subtle and complex manner in which political considerations helped to determine “ethnic” identity.34 In that year, a French administrator visited the local chief, one Manuel Verda, to inform him that Adeane was henceforth under French jurisdiction. Verda responded that he, unlike the chiefs of surrounding villages, had not signed a treaty with France and that, like his father Fabrice Verda, who had purchased the land from local rulers, he, Manuel, was Portuguese, and so was his village.35 The administrator countered that the other local rulers had accepted French sovereignty and that therefore Verda and his entire village were also now on French territory. Verda responded that he would continue to consider his village to be Portuguese until he was informed to the contrary by Cacheu, Bissau, or Lisbon. Verda is here claiming both Portuguese citizenship and “Portuguese” identity, whereas the administrator, of course, is arguing territorial sovereignty. But the latter claim has direct implications for the question of identity. The established cultural and professional parameters of “Portuguese” identity are being challenged and, gradually, supplanted by a sense of identity more closely in accord with European notions of nationality. Ironically, this development is exemplified in the person of the official French interpreter on the expedition to Adeane. The interpreter was one M. Bocandé, an employee of the trading house of Maurel & Prom and a son of Emmanuel Bertrand-Bocandé. The elder Bocandé had been the French résident at Carabane from 1849 to 1867; he had married a Luso-African woman “à la mode du pays.”36 The younger Bocandé was a French citizen through his father. But through his mother he was “Portuguese.” Verda, the chief of Adeane, was “Portuguese” through his father, but he too was almost certainly a Luso-African. Thus, in this colonial encounter, some Luso-Africans are Portuguese while others are French. Increasingly, European identities based on nationality or citizenship and associated with the territorial claims of the colonizers have begun to intrude, even in Casamance. Verda in particular finds his room to maneuver limited by the expansion of French control of the Middle Casamance. By his own claim he is Portuguese and also chief of a local village;

Senegambia from the Mid-18th to the Mid-19th Century

107

he now finds himself redefined, against his strong protest, as “un simple particulier” [just an individual] who happens to own property located on French territory. Adeane lay in contested territory. Long inhabited by a Bagnun-speaking population, the south bank of the Casamance River had come under attack shortly before 1850 by Balantas arriving from the south. The neighboring chiefs who had sold the village to Manuel Verda’s father may have been offering him land that had been depopulated by the Balanta invasion. But the chiefs may also have sought to gain protection by attracting a “Portuguese” as a local ruler, who, they hoped, would bring Adeane and the surrounding territory under the aegis of the garrison at Ziguinchor. The Verdas seem to constitute yet another instance of Luso-Africans who became chiefs of a community of non-“Portuguese” Africans. As at Cazegut in 1849, where the local population had chosen as their chief the Luso-African “pirate” Manuel Gara (or Sara), it appears that the local populations near Adeane consciously selected an individual of Afro-Portuguese heritage as a leader. The French report of 1879 gives no direct information about African motives for having welcomed the Verdas. However, it is likely that the transaction was related to the political upheavals that disrupted the south bank of the Casamance toward mid-century. IDENTITY AND COLOR One significant aspect of individual identity that did not change in the Casamance-Bissau region during the first half of the nineteenth century was the fact that whether or not an individual was considered “white” had nothing to do with their skin color. The color terminology that had begun to develop as early as the sixteenth century whereby “black” and “white” reflected social status, including profession, still existed on the eve of the colonial period. Gaspard Mollien’s 1818 observation that among Rio Géba traders “all who are free” were called “white”37 also applied to Casamance societies. In 1861, the French administrator at Sedhiu reported that some of the more isolated groups he had visited along the Soungrougrou were astonished to see a pale-skinned human being. He was in country inhabited primarily by Yolas (Jolas) at the time: “These good people had been under the impression until now that the ‘whites’ were our black traitants who visited them, and they had absolutely no idea that there were also ‘whites’ of another color.”38 To these Casamançais, to be “white” was to be a trader. Just as long-distance merchants who brought captives, ivory, and other goods from the east were Juulas, so too merchants who came upriver bringing cotton cloth were “whites.” Parenthetically, one should note that more than a century later, the northern Jolas frequently use the term “Juula” as synonym for “trader.” One can become a Juula by becoming a professional long-distance trader.

108

“portuguese” style and luso-african identity

Not once, however, during two decades of fieldwork in Buluf, Fogny, and, briefly, along the Soungrougrou, have I heard the term elúlúm (“white”) used to refer to a trader. A century of contact with Manding traders and with whites who were truly “d’une autre couleur” [of another color] has apparently effaced the memory of the earlier usage of the term. In nineteenth-century Casamance, however, the categories of “black” and “white” were not based upon skin color. Rather, the terms were associated with profession and social status. There is evidence that this usage was not limited to the region of LusoAfrican commercial activity. In 1856, Anne Raffenel reported on his travels through eastern Senegal along the Falémé and upper Gambia Rivers. He found that traitants from Saint-Louis were known as “whites.” He wrote: “By a most annoying confusion, the term white is given, in those lands frequented by our traitants, to any individual who lives in SaintLouis, without consideration of their color. Hence, for this reason, my Negroes were ‘whites.’”39 The use of “white” to connote professional traders was entrenched in the Senegambia region long before the colonial period. In the Senegal River valley, as James Webb has demonstrated, the term “bidan” was often associated with traders, regardless of their complexion. In view of the existence of similar color terminology from the Senegal River valley to the coast south of The Gambia, this may have been a regional phenomenon. There is some evidence that Africans who traded with both the Portuguese and the Maures referred to both groups by the same word. In 1828, during a prolonged stay at Timé in upper Guinea, a village inhabited by Manding and Bamana, René Caillié observed that his hosts “have a strong image of the wealth of the ‘whites,’ and even of the Arabs, whom they place in the same category.”40 The context for Caillié’s comment, a discussion by Manding merchants about French and English traders at Albreda, clearly indicates that both the “blancs” and the “Arabes” in question were merchants. Professional traders, whether Arabs, Europeans, or Luso-Africans, were perceived as comprising a single category. This confirms the observation that the term “white” was not a physical description but rather a professional category. And traders were generally associated with riches. Hence, to be a trader (“white”) was to be associated with wealth. Recently, Bogumil Jewsiewicki has characterized “white” as a synonym for “bourgeois” in urban Africa rather than as a physical description. He writes: “For many years now, in popular urban culture in many Black African societies, the term ‘white’ has served as a synonym for bourgeois, without particular regard for skin color.”41 In the greater Senegambia region, the history of the use of the term “blanc” and the closely associated words “bidan” and “Portuguese” strongly suggests that here, at least, the phenomenon to which Jewsie-

Senegambia from the Mid-18th to the Mid-19th Century

109

wicki refers is not limited to urban settings. Indeed, it is likely that the particular usage to which he refers reflects a recent modification. The dissociation of “white” from simple description of skin color dates not to many years ago but rather to many centuries ago. F R E N C H I M AG E S O F CA SA M A N C E P E O P L E S Southern Senegambia and Bissau were less frequented by French merchants in 1800 than they had been in 1700. Consequently, the local populations were less well known to French writers. Late-eighteenth-century French observers sojourned in the administrative and commercial centers of Saint-Louis and the Isle of Gorée; occasionally, too, they briefly visited the French escale at Albreda on the lower Gambia River. Men such as Durand, Adanson, Pommegorge, and Geoffroy de Villeneuve were generally familiar with northern Senegal but were largely ignorant of the peoples and cultures south of the Gambia River. For example, in his fourvolume account of the peoples of Senegal, Villeneuve, who spent a total of four years in Senegal in the 1780s and served as aide de camp to the governor, Chevalier de Boufflers,42 has little to say about the populations south of The Gambia. He imagines these peoples as living in a perpetual state of warfare with each other, and he presents a brief and particularly pejorative view of the Floups of the Casamance. However, Villeneuve does have the distinction of being possibly the first European to record and publish the oral tradition according to which the Serers and the Jolas are descended from a common ancestor. He also is the earliest writer to use the term “Jola” to refer to the peoples who had hitherto appeared in the historical record as Floups. Villeneuve was reasonably well informed about the Serers and he undoubtedly had this oral tradition from them, though this tradition is indeed recounted by members of both groups. He writes: “It is thought that the Serrers [sic] are colonies of a nation called Guiola, who still exist along the banks of the Gambia. Similarities between these two peoples in language, customs, and common practices confirm this opinion.”43 This passage accurately records the Serer migration tradition that is otherwise found only in much later sources. Such early versions of oral traditions that survive to the present are rare. Not only does this passage demonstrate the continuity over two centuries of the tradition in question, but it affords the earliest known usage of the term “Guiola” (Jola, Diola). Furthermore, it suggests that the name Guiola itself originated among either the Serers or, conceivably, among their relatives, the Jolas themselves. This is highly significant, since most sources from the nineteenth century to the present have suggested a foreign etymology, ascribing the term either to Wolof traders or to the Manding.44

110

“portuguese” style and luso-african identity

Nevertheless, when it comes to describing the people called “Guiola,” Villeneuve not only reverts to the more common terminology, he presents a disparaging picture: One sometimes encounters, in the land of Barra, a kind of Negro who have retained all the crudeness and customs of savagery. They inhabit a small country on the banks of the Casamance River, close to The Gambia, and they occasionally cross the latter river to sell their cattle. They are called Floups. The Floups are almost naked, aside from a small skirt that passes between their thighs. They tightly bind their upper arms and wrists, their upper thighs and above the knee, their upper and lower legs with leather thongs that are dyed red. In consequence, the intervening parts of their tightly bound limbs are much thicker than normal. They scarify their faces and body, inscribing all manner of figures and weird and bizarre designs. These Blacks have thickly curling hair that they gather on the crown of their head above the forehead, in a feather-like projection, 5 or 6 inches long. They let their beard grow to a point. They are covered with amulets or gris-gris; their weapons include bows, arrows, shields hanging from their back. In their left hand they carry several azagayes, or javelins, of various sizes, which they throw with great accuracy. Their language is crude, their pronunciation lively, loud, and guttural; though they communicate little, they are not at all fierce, and they live in peace with the neighboring hordes.45

The physical description of these people, emphasizing their nudity and the transformation of young men’s bodies by means of cicatrized and bound limbs, serves to convey their supposedly rude and savage nature. Villeneuve is clearly not using the term “la vie sauvage” as a reference to some ideal state of nature gleaned from a reading of Rousseau. This image is complemented by the description of their hair, which they wear long and tied into locks, as well as of the amulets, an implicit mark of supposedly primitive religious beliefs, with which they cover their bodies. Likewise, their weapons are both quaint and primitive. To Villeneuve, their rude and guttural language gives additional proof of their uncivilized state. Of course what Villeneuve is saying is that he does not understand Jola, hence it sounds strange and incomprehensible to him.46 Practically the only positive observation he makes is that the Floups are not warlike. This image of the Floups as peaceable would soon be modified as the French tried to impose their territorial sovereignty over Casamance villages. By the end of the nineteenth century, many of the Jola subgroups had earned reputations as ferocious warriors. A RC H I T E C T U R E A N D “ C I V I L I Z AT I O N ” In 1837, Dagorne, who was commanding a reconnaissance expedition that led to the establishment of the French post at Sedhiu, found the Jolas—or Yolas, as he called them—to be warlike and feudal in their political organization:

Senegambia from the Mid-18th to the Mid-19th Century

111

The Yolas, on the other hand, who inhabit the islands at the mouth of the Casamance and who have no ongoing relations with the continent, are rude and have no political constitution. Each village forms a separate state which is often at war with neighboring villages, and each village is a form of republic where each family head acts in an utterly independent manner.47

Furthermore, to this nineteenth-century observer, architecture reflected a society’s political organization: “Along the coast one might almost recognize the specific form of government by the form of the dwellings of the inhabitants.”48 Accordingly, Dagorne argued, the Wolofs constructed flimsy straw huts because their system of government gave total control over the individual’s property and life to the ruler. The Manding, “more advanced by reason of contact with civilization,” constructed solid and enduring houses of sun-dried clay.49 The Jolas provided a more complicated study. Their dwellings were massive and enduring, yet they did not show evidence of a more advanced level of civilization. Dagorne, who had obviously visited Jola villages, described their large farmhouses, which were characterized by solid walls of sun-dried clay, many rooms, and twisting corridors separated by narrow doorways. He stressed—not without reason—the defensive nature of these massive dwellings. However, rather than be impressed by their solidity or their scale, Dagorne interpreted these buildings as evidence of the anarchic and warlike nature of Jola society. Ironically, he overlooked the slave raiding that had necessitated such defensive architecture. The Yola, without laws or rulers, always ready to defend his property or to invade that of his neighbor, builds a vast house of sun-dried clay, with high and thick walls, multiple narrow compartments, tortuous corridors divided by numerous narrow doors. Clearly, he seeks to defend himself to the last. In a sense, his home is—except for the construction materials and the scale and with, relatively speaking, quite as much strength—the equivalent of the castles of our warlike and frequently unscrupulous medieval lords.50

Dagorne here elaborates a theory whereby each African society has its characteristic form of architecture: one people, one style. In Dagorne’s idiosyncratic interpretation, each style reflects the form of social organization peculiar to that group. And the social organization, in turn, corresponds with the level of civilization that Dagorne attributes to the group. The logic is circular—or at least self-confirming. The dominant image of the Floups at the dawn of the colonial period in Senegal was that they were a backward, primitive people. Had observers such as Dagorne not adhered to this view, they might well have been impressed by Floup farmhouses. Instead, Jola domestic architecture, no matter how sturdy, durable, and massive, was not interpreted as evidence of a high level of cultural achievement but as indication of primitive social and political organization. A generation after Dagorne, Vallon, who was later governor of Senegal, expressed a similar disparaging attitude toward a Casamance people.

112

“portuguese” style and luso-african identity

However, this image does not coincide with the description he gives of the architecture of the Baiot or Bayottes, who are today classed as a subgroup of the Jolas. He refers to “Baiot” as a village: This village has an impoverished appearance. Nevertheless, the houses are built of earth, as elsewhere, and [are] quite as well-constructed; they are square rather than round; beside each house is a small plot for cotton, to meet the family’s needs. But the people are so ugly, so repulsive, that I was left with a bad impression.51

Vallon’s description of the houses shows them to be well constructed and rather impressive. Nevertheless, he is unimpressed by the inhabitants, if only because he finds them unattractive. Here the negative image of the Bayottes does not coincide with the description of their architecture. The people of Casamance find themselves in something of a catch-22 situation: their architecture, if impressive, is ignored in favor of less positive characteristics of their material culture. However, if the buildings are less than impressive, this fact is taken as demonstration of a lower level of civilization. By associating each specific architectural style with a corresponding level of civilization, Dagorne was following the lead of earlier French observers of Senegalese culture. At the end of the eighteenth century, Villeneuve had reserved his highest praise for those Senegalese who built using the most permanent materials. These Senegalese just happened to be the most Europeanized, at least in terms of their houses: “The inhabitants of Dakar are the most civilized of the entire coast; they very much like the French and have begun to take on European manners and customs. In fact, several, when I was in Africa, had already constructed houses of stone.”52 This passage, with terminology that echoes La Courbe’s seventeenth-century description of the courtly (and partially European-influenced) manners of the king of Geregia, but from a more Eurocentric perspective, is founded on the assumption that lasting structures are a mark of civilization. For, in another passage, Villeneuve writes: “Nowhere does one encounter any building, nor any monument, constructed with the intention of transmitting an awareness of great deeds to posterity.”53 Monumentality and permanence are precisely the characteristics that Villeneuve finds lacking in most Senegalese architecture. These qualities are approached only in the stone constructions of the more civilized inhabitants of Dakar. North of the Cape Verde peninsula, in the region of Kajoor (Cayor), the Wolof population also did not construct buildings in permanent materials, a fact that Mamadou Diouf attributes to long-standing policy on the part of the Damel: “The ruling class in Kajoor always prohibited the construction of any building in permanent materials on their territory. According to tradition, the Damels . . . always asserted that houses or trading

Senegambia from the Mid-18th to the Mid-19th Century

113

entrepôts built [of stone] were susceptible to be used as a base for attack or defense.”54 According to Diouf, the absence of more lasting building materials— even of banco—in Cayor constituted a conscious political and military policy on the part of local rulers. This was quite different from the situation that existed from The Gambia to Bissau where, in the absence of any centralized political authority, local African communities as well as European merchants and Luso-African middlemen all built houses of banco that were solid enough to withstand military attack.55 Ironically, when Dagorne argues that the flimsy construction materials of Wolof houses reflect the control exercised over the population by their rulers, he is, strictly speaking, correct. His ethnocentric bias intrudes, however, when he asserts that people who build houses of sun-dried clay or stone demonstrate by this architecture that they are more advanced on a relative scale of civilization. And, of course, Villeneuve’s association of stone houses with “the most civilized people found along the entire coast” imposes European architectural norms on cultures that have radically different parameters for constructing buildings. Villeneuve thereby extends a French norm into a universal standard. It is a short step from Villeneuve’s argument to the articulation of a full-fledged French “civilizing mission.”

SIX

Casamance Architecture from 1850 to the Establishment of Colonial Administration

On April 8, 1850, Emmanuel Bertrand-Bocandé, the French résident on the island of Carabane in the Casamance, arrived by boat at the community of Thionk-Essyl or, as he called it, “Kion,” located on the Jigouches plateau north of the Casamance River. The people of Jigouches were a subgroup of the Floups; their descendents consider themselves to belong to the Jougoutes, or Buluf, group of the Diolas. The inhabitants of “Kion” were well known to Bocandé as rice farmers, fishermen, and wage laborers, but above all as redoubtable marauders or pirates. Bocandé had lived at Carabane since 1837; he spoke Crioulo and Manding and he had at least a rudimentary familiarity with the Floup language, now called Diola (Jola).1 His knowledge of local history, including Floup and Bagnun migrations, is unsurpassed to the present day,2 based as it was on his long and generally mutually respectful relationship with prominent inhabitants of Thionk-Essyl and other Floup villages. On this April day, however, Bocandé’s friendship was mitigated by his mission: he had traveled to Kion to obtain the release of a woman and child from Carabane. These two residents of French territory (Carabane had been French by treaty since 1837)3 had been kidnapped by “pirates” from Kion, who were about to sell them to some Manding traders. Bocandé’s rescue mission, including the details of his successful negotiations for the captives’ freedom, is the subject of a long report, fourteen manuscript pages, that he wrote upon his return to Carabane on April 10th. This manuscript, preserved in the Archives Nationales in Dakar,4 offers a view of Bocandé’s relations with the Floups a generation before the colonial conquest of the Casamance. It also provides detailed information about Thionk-Essyl, including the community’s social organization and physical layout. Architecture figures prominently in Bocandé’s description of the village, which affords a view of the constructed environment, including

Casamance Architecture from 1850 to the Colonial Administration 115

building style, some methods of construction, and the layout of family compounds. The report predates the maraboutic wars of the 1870s and 1880s and the subsequent urban migration that followed the reestablishment of peace, two factors that had contributed by the early twentieth century to the profound transformation of local social organization and architecture. In addition, Bocandé’s written description is a generation earlier than the birth of the oldest Buluf informants whom it was my privilege to interview when I first lived in Thionk-Essyl, in 1974. Diola houses, as Louis-Vincent Thomas, the first modern scholar to study the Diola peoples, observed, are strong enough to endure for over half a century.5 Hence, some of the buildings remembered by my oldest informants, notably Cheikh Suleman San∑ (b. ca. 1885) and Kepi Diatta (b. ca. 1884), probably already existed at the time of Bertrand-Bocandé’s visit. And a few of the structures that drew Bocandé’s attention in turn dated from the first decades of the nineteenth century. Thus, it is possible to trace the history of domestic architecture in the largest Diola community north of the Casamance River6 through the entire nineteenth century. In 1850, as today, Thionk-Essyl consisted of a cluster of geographically dispersed wards (kalol; pl. ulolau), each effectively a semi-autonomous village. Even within these smaller units, the population was not densely compacted. Rather, as Bocandé observed: “Each inhabitant builds his home in isolation, for himself and his family, in the midst of his field, on land elevated above the rice fields that border the tidal streams.”7 The physical separation of these extended family compounds (fankaf ) and their location adjacent to the rice polders reflected the fact that the family group also constituted an economic unit.8 The architecture emphasized this autonomy. Each compound was closed to the outside world and served as a fortification for those within. The fankaf consisted of numerous earthen houses, each in the form of a small tower covered with a thatched roof. The individual structures were linked together by earthen walls. This image closely parallels the recollections of elders born a generation after Bocandé’s visit. For even in the last decades of the nineteenth century, all of the members of the extended family—the family patriarch and, in this generally patrilocal society, his adult married sons, each with their wives and children—constructed their houses together, partly for defensive purposes. Bocandé’s important description deserves to be cited at length: A dwelling consists of diverse earthen houses in the form of small towers, without any windows on the outside, more or less numerous depending on the wealth of the family; these small towers, covered with thatched roofs, are joined together by earthen walls protected from the rain by hollow trunks of palm trees, in the form of a slab that collects and funnels off the rainwater that falls from the roof. These various structures are in turn at least partially surrounded by a palisade; from outside they have the appearance of small fortresses with their towers and courtyards, while inside their asymmetrical

116

“portuguese” style and luso-african identity

form creates a small labyrinth, with two openings, in opposite directions, to permit entry or to facilitate flight in time of danger. At irregular distances there are small but dense woods.9

This passage, which is remarkable for its detailed description of domestic architecture, provides evidence of long-term continuity of certain elements of construction. The labyrinthine form is reminiscent of La Courbe’s description of Floup residential compounds at the end of the seventeenth century.10 The use of exterior palisades is documented even earlier (see below). Protection against armed attack remained a central consideration for Casamance architects from the sixteenth century to the end of the nineteenth century. This is not surprising, for these four centuries coincided with the growth of the Atlantic slave trade, followed in the nineteenth century by a burgeoning domestic slave trade. Ironically, in Casamance, the abolition of the overseas trade in slaves and the growth of legitimate commerce served to increase the demand for captives to grow peanuts in the Gambia and the Middle Casamance regions. Indeed, strong demand for captive labor in the Manding-dominated groundnut-producing areas to the north and east most probably induced the men of “Kion” to take up seaborne slave raids. But the people of Thionk were themselves vulnerable to similar attacks. Oral traditions in Thionk recount the story of children from one kalol being captured by men from other wards.11 Hence, defensive dwellings were a necessity, even after Bocandé’s visit. The period from 1877 to 1893 was a time of heightened danger as Manding and Wolof marauders launched regular slave raids against villages between the Gambia and Casamance Rivers. The most notorious of these warriors, Fodé Kaba12 and Fodé Sylla (or Combo Sylla, as the Diolas called him), led raids throughout Fogny, located to the west of Djougoutes. They burned villages and slaughtered or captured the inhabitants. In 1886, Combo Sylla tried to invade Djougoutes from the north by crossing the marigot that separates the region from Fogny. But the villages of Djougoutes had formed a defensive alliance. When the people of Kartiak, northeast of Thionk, heard Sylla’s advance guard approaching, they gave the alarm. The men of Kartiak, Thionk-Essyl, and Thiobon attacked Sylla’s cavalry while they were still stuck in the swampy estuary. The invaders were routed. Never again did any Manding slave-raider dare to attack the large villages of western Djougoutes. Thionk captured Sylla’s war drum, which is still carefully preserved along with the historical memory of the victory. Although the danger of slave raids was significantly diminished in Buluf after 1886, defensive considerations nevertheless remained an important aspect of architectural design until the end of the century. Armed attacks remained a threat until 1893, when Fodé Kaba was forced by combined French and British colonial forces to sign a peace treaty and curtail his attacks in Fogny and adjacent British Kombo.

Casamance Architecture from 1850 to the Colonial Administration 117

Bertrand-Bocandé’s description shows that the fankaf in Thionk were designed with a triple system of protection against attack. The earthen rampart that connected the windowless outer walls of the individual houses formed a solid and virtually indestructible primary line of defense.13 Outside the wall was a palisade. Finally, inside the walls, the structures formed a confusing labyrinth. Bocandé does not specify the construction material of the palisade, but it was almost certainly made of wooden stakes driven into the ground. In French accounts, the term “palissade” generally implies such wooden construction. A decade after Bertrand-Bocandé’s report, the French administrator at Sedhiu visited the Soungrougrou River, which separates Fogny from the Middle Casamance. This region was largely isolated from European contact. The administrator observed that the Diola villages along the river were surrounded by palisades. Within the villages, individual housing compounds, too, were palisaded: “The Jolas take care to surround their housing compounds with a palisade that provides solid protection.”14 In 1877, the construction method for these fences was described by Père Spiesser of the Holy Ghost Fathers’ Mission (Pères du St. Esprit): “At Carabane one day I entered a courtyard containing several houses, each with its small barrier made of sticks bound together with bark.”15 Carabane was a community of immigrants from both north and south of the Casamance River. Hecquard’s 1851 sketch (Figure 21) depicts the town from the river. It shows a mixture of rectangular and round dwellings and one or two two-story structures. Many of the houses there were constructed by seasonal migrants from Thionk-Essyl and the neighboring community of Mlomp.16 Bertrand-Bocandé observed that the men of Thionk traveled throughout the Lower Casamance and that their many occupations included the construction of earthen houses.17 Adobe house construction is still practiced in Thionk-Essyl today. Although the architectural forms have changed significantly, construction techniques for walls and for the wooden structures that support the roof are probably reminiscent of techniques used for nineteenth-century buildings (Figure 22, Figure 23). Palisades such as these were found throughout the Lower Casamance. This architectural feature was not limited to the Diolas. However, it is noteworthy that the Floups and Diolas, who were expert in the art of constructing houses out of banco, do not appear to have taken the additional defensive precaution of surrounding their villages or their compounds with mud walls. The dispersed layout of their compounds made Diola villages too large to surround by walls of banco. This situation was in marked contrast to the mud-walled Manding villages of nineteenthcentury Pakao, east of the Soungrougrou.18 The third line of defense described by Bocandé was a product of the compound’s design. The asymmetrical arrangement of the multiple interior courtyards created a labyrinth. Any enemy who succeeded in breach-

118

“portuguese” style and luso-african identity

21. Carabane-Comptoir français (Cazamance), 1851. Hyacinthe Hecquard, watercolor. Courtesy of Frobenius-Institut, Universität Frankfurt am Main, Bilder-Archiv.

22. Thionk-Essyl, house under construction, 1975. Photo by Peter Mark.

Casamance Architecture from 1850 to the Colonial Administration 119

23. Thionk-Essyl, roof under construction, 1975. Photo by Peter Mark.

ing the two outer lines of defense would become lost in the confusion of interior spaces. This could allow a few individuals to hold off far more numerous attackers. There were only two doorways to the compound. In an emergency, family members could exit through one of the doorways, leaving the raiders to wander blindly through the maze. Since at least the sixteenth century, Casamance housing compounds had been surrounded by palisades of wooden stakes; labyrinthine construction is documented by the seventeenth century. The earliest description of Floup architecture—Valentim Fernandes’s portrayal of the compound of the Mansa Falup, or king of the Floups—affirms that his houses were surrounded by wooden stakes that formed concentric circles. Entry to the compound could be gained only by walking around the perimeter of each circle, as the entries were not aligned.19 Fernandes specifically mentions that similar palisades protected the Mandi Mansa’s compound in The Gambia. So this feature appears to have been a regional phenomenon. Found among the Floups and the western Mande, it predates the expansion of the Atlantic slave trade.

120

“portuguese” style and luso-african identity

At the end of the sixteenth century, Almada described Mande defensive fortifications in The Gambia that were also surrounded by wooden palisades and backed by an earthen rampart.20 Almada’s contemporary André Donelha visited Manding communities in The Gambia that were “entirely surrounded by stakes driven into the earth.”21 A generation later, Jobson reported that the important Gambian trading center of Casão was “circled with posts and pieces of trees set close and fast into the ground, five feet high, so thick that, except in stiles . . . a single man cannot get through and, in like manner, a small distance off the like defense, and this is, as they signify to us, to keep off the force of horse.”22 At the beginning of the eighteenth century, in a passage almost certainly plagiarized from a now lost portion of La Courbe’s travel account, Père Labat described similar fortifications surrounding Floup houses: “We traversed some Floup villages that were surrounded by five or six rows of palisades.”23 Labat also described Geregia as surrounded by “a double enclosure of thick stakes.”24 Inside this perimeter, seven or eight “Portuguese”-style houses were in turn surrounded by another palisade. According to Labat (actually La Courbe), the fortifications served to protect against wild animals and human enemies. Fifty years after La Courbe, Francis Moore wrote about the Floups of Fogny that “each of their towns [is] fortified with sticks drove all round and filled up with clay.”25 Almada also mentioned that the Papels, southern neighbors of the Floups, constructed buildings in the form of labyrinths.26 And La Courbe wrote of the Floup roi de Bolole that “his house is like a labyrinthe.”27 Clearly, both defensive design features mentioned by Bocandé—labyrinths and palisades—were part of a regional architectural technology that was widely shared by the ancestors of the different peoples of the Gambia-Bissau region. This technology was at least several centuries old at the time of Bocandé’s visit to Thionk. Such long-established features of defensive architectural design continued to serve the Diolas—as the Floups’ descendents were generally known by the late 1800s28—in the face of extensive late-nineteenth-century slave raids. The large communities of Buluf also formed inter-village alliances that protected them against the late-nineteenth-century Muslim warriors. Further north and east, however, the devastation caused by Fodé Kaba and other marauders led to the depopulation of wide areas on the Fogny plateau and along the Soungrougrou River. The magnitude of this destruction is suggested by a photograph taken during the French bordersurveying expedition of 1889–1890. The image of an unidentified Fogny village, which was taken by Varenhorst, shows several bodies sprawled amidst the ruins of four burned-out houses (Figure 24).29 The remains of the three houses that are fully visible30 clearly show that these structures were round in form. The remnants of porches and veranda posts are prominent in the ruins. At ground level, the low outer veranda wall of one house is embellished with a series of round-

Casamance Architecture from 1850 to the Colonial Administration 121

24. A northern Jola village immediately after a slave raid, 1890. Photo by Varenhorst. Courtesy Société de Géographie, Paris, France.

ed adobe arcs. There is poignancy to this wantonly destroyed decoration. Amidst the charred ruins lie seven bodies. Clearly the photograph was taken very soon after the slave raid, since the corpses are not yet bloated in the heat. The dwellings had not been constructed in the fortress-like style of Thionk; there are no protective outer walls, nor is there any evidence in the picture of an outer palisade. Varenhorst’s horrifying picture is probably the only photograph of a nineteenth-century slave raid in Senegambia. Why did the people of mid-nineteenth-century Djougoutes live in heavily fortified compounds while the population of Fogny initially did not? The answer may lie partly in Kion’s own slave-raiding history, which made its residents acutely aware of the potential threat of kidnapping. The Essylians’ proclivity to capture children both from their Floup neighbors south of the Casamance River31 and from neighbors closer to hand in the other wards (ulolau) of Thionk taught them the importance of dwellings that could easily be defended. In addition, there was a long history of intervillage warfare in Buluf. Local traditions still recall fighting with the nearby community of Tendouk. These conflicts produced few casualties but probably some captives. The endemic nature of intervillage fighting was a strong incentive to maintain fortified compounds. The villages of eastern Buluf, which were vulnerable to attacks from the Fogny plateau, also developed defensive housing construction. Informants in the community of Mandégane recalled that before the establish-

122

“portuguese” style and luso-african identity

ment of French administration, extended-family housing compounds were encircled by a barrier that had only a single entry. At night, animals were brought into the central courtyard for protection against theft. This architecture around a courtyard, termed fankaf, gave way to the present patterns of dispersed housing after the establishment of peace with the arrival of the French and conversion to Islam.32 Although the warrior-marabouts sought to legitimize their actions by couching them in the discourse of jihad, in the eyes of their potential victims, they were simply slave-raiders. The only religious conversion effected by their attacks was of those captives who became Muslims as part of a process of mandingization after they were sold into slavery in the Manding villages of The Gambia and in Pakao in the Middle Casamance.33 Those inhabitants of Fogny and Buluf who were not captured resisted not only the military assaults but also all efforts at Islamic proselytizing. Half a century after Bertrand-Bocandé described the architecture of Thionk-Essyl, another French visitor, Dr. Lasnet, published an account of the Casamance that clearly shows the impact of the intervening generation of warfare on domestic architecture in Fogny. He found that houses in Fogny were now designed to maximize protection against attack by mounted warriors. The Diolas of Fogny, long vulnerable to cavalry raids by Fodé Kaba and Combo Sylla, had been forced to sacrifice comfort for defense: Fogny houses are completely different, also made of compacted earth [adobe] but small and circular, without any interior subdivisions; they are partially buried under the ground and are neither aerated nor lighted [i.e., no windows] but have narrow and low openings, covered with palm fronds; sometimes they are located in the middle of several circular enclosures made of closely spaced stakes whose openings are diametrically opposite one another. . . . The people of Fogny, once they return from their rice fields or from the forest, go to bed in these houses and they do not like to leave them; often they take bamboo straws and pass them through the walls to serve as urinals.34

These dark, windowless, half-subterranean structures, surrounded by concentric palisades of stakes, constituted an architecture of desperation for a population that, for a generation, had suffered from attacks by mounted Manding warriors. Lasnet comments with distaste on the dark and insalubrious state of these houses, “where the people of Fogny are quite content.” The phrase “ni aérées, ni éclairées” suggests that the French observers were describing ekumbane houses. Characteristic of Fogny, these structures contain an interior room with especially thick walls and no windows. The ekumbane room was home to the family patriarch; here the old man would sit by a fire to warm himself during the cool evenings of December and January.35 But the general living conditions were merely the

Casamance Architecture from 1850 to the Colonial Administration 123

result of a state of virtual siege rather than any innate attraction to filthy conditions.36 The fact that these people did not even dare to leave their fortress dwellings to relieve themselves at night illustrates the danger and insecurity that had come to characterize the region at the end of the nineteenth century. In fact, Lasnet himself makes the association between this particular form of architecture and the Manding depredations: “These low, halfburied houses, well-suited to defense, are peculiar to the parts of Fogny that were most exposed to Manding incursions; it may be that defensive factors led, in these regions, to the modification of Diola buildings and led the inhabitants to burrow into the earth and to live in these filthy redoubts.”37 Lasnet’s explanation is, in fact, precisely accurate. The defensive measures taken by the people of Fogny, like those taken in Thionk-Essyl a generation earlier, incorporated elements that are documented in descriptions of housing in the region as early as the seventeenth century. However, the description of houses partially buried underground is unique in the literature of the Gambia-Casamance region, although similar subterranean dwellings are documented elsewhere in West Africa. The small, round, dark houses of late-nineteenth-century Fogny, “half buried in the earth,” were intended to afford protection and shelter. Unseen by attacking cavalry, those within could use the narrow and low openings, which were inaccessible to mounted warriors, to fire at the bellies of the horses while remaining virtually invulnerable to the cavalry. The vulnerability of the people of Fogny to Manding raiders was initially due in part to their lack of familiarity with cavalry. In Buluf, the memory of the 1886 victory over Combo Sylla incorporates the recollection that until one man from Thionk captured a horse, the Diolas thought horse and rider were one terrifying animal.38 It did not take the Djougoutes Diolas long, however, to learn to defend themselves against these attacks. The densely populated Djougoutes plateau was protected from attack from the west by the Diouloulou marigot, from the south by the Casamance River, and from the north by the Baila marigot. It was while attempting to ford the latter waterway that Sylla was ambushed.39 The two dozen communities of Djougoutes could only be invaded from the east. In 1886, another Muslim warrior, Birahim Ndiaye, attacked from that direction; his assault carried only as far as eastern Buluf.40 In the wake of Sylla’s defeat, Djougoutes was spared further attacks. In Fogny, the population was geographically vulnerable to cavalry attack from both The Gambia and Pakao. Most of Fogny consists of a low plateau unbroken by the waterways and nearly impenetrable mangrove thickets that protect Djougoutes. This lightly forested plateau is readily traversed by horse; nineteenth-century Djougoutes, on the other hand, was heavily forested. In addition, Fogny villages were smaller than those of Djougoutes; there were fewer men to mount a defense. Furthermore, the Diola-Fogny, at least in the nineteenth century, lacked experience in

124

“portuguese” style and luso-african identity

warfare. In 1880, Père Lacomb of the Sedhiu mission of the Holy Ghost Fathers described the Diola-Fogny as “defenseless and unfamiliar with warfare,”41 unlike the redoubtable men of Buluf. Why should the inhabitants of Fogny have been far more pacifist than their Buluf neighbors? One reason may have been the extensive commercial contact between Thionk-Essyl and Carabane. Even before 1850, labor migration to Carabane afforded the Buluf-Diola access to firearms. Perhaps the difference was also due to the fact that Fogny was formerly Bagnun territory. The Bagnun showed little proclivity for fighting, even in their sixteenth-century empire of Kasa. The Diola-Fogny had assimilated parts of the earlier Bagnun population, and it is likely that the pacific orientation of nineteenth-century Fogny society derived partly from their ancestors. The Bagnun states of Kasa (sixteenth century) and Geregia-Fogny (seventeenth century) were based on long-distance trade. Warfare was antithetical to such commercial enterprise. The Bagnuns had little incentive to develop the arts of war. For seventeenth-century Bagnuns, their nonviolent disposition42 was a fundamental aspect of the economic role of Geregia and Vintang as interethnic and, indeed, international trading centers. The absence of a military tradition, however, did not serve their descendents well in the unstable and violent conditions of late-nineteenth-century Senegambia. Fodé Kaba’s attacks on Fogny began in 1877–1878. Over the following decade, his cavalry transformed the region into a wasteland and its population into refugees, captives, or besieged farmers like those whose living conditions were documented by Lasnet. By 1880, entire villages were abandoned by the Diola, who fled south and east before the onslaught until their way was blocked by the Soungrougrou or Casamance Rivers. One eyewitness, missionary Père Lacombe, observed: At his approach, the terrified Diolas flee, driving their cattle before them. . . . From Adeane all the way to Ziguinchor, which is a distance of ten leagues, it is not unusual to come across numerous families of Diolas spread out across the villages of the coast [riverbank] with nothing to live on but forest plants or palm kernels.43

These refugees were perhaps the lucky ones. Their compatriots who did not flee in time became war booty to Fodé Kaba, whose express ambition was “to make of the province of Fogny a Muslim empire placed under his domination.”44 Although slave raiding had long been practiced by the sea-warriors of Thionk, it was on a small scale and took the form of kidnapping. Captives were ransomed by their families, sold to the Manding, or forcibly incorporated into their captor’s family. Among the northern Diolas, and probably their Bagnun ancestors, domestic slavery was a relatively mild institution. The captor was responsible for the well-being of his slaves, who gradually were assimilated into the family. The distinction between

Casamance Architecture from 1850 to the Colonial Administration 125

descendents of an amigel [slave] and free persons disappeared over a few generations. Ironically, twenty years after the French forcibly put an end to Djougoutes slave raiding,45 the northern Diola were themselves victims of a vastly larger slave trade; indeed, it was quite likely larger than during the era of the Atlantic slave trade. The ranks of Fodé Kaba’s captives, swelling to a continuous stream during the dry season of 1880, fed a growing slave market at Marsassoum, on the east bank of the Soungrougrou River. This market was visited by Père Lacombe, who hoped to purchase the freedom of some children. The missionary was, however, unprepared for the magnitude of human misery he encountered there: Marsassoum is the rendezvous of the merchants of human flesh. Every day there arrive here captives from the right bank, driven in bands either by the Manding or by the soldiers of Fodé Kaba. These poor unfortunates are the object of commercial transactions of a very considerable magnitude. Playthings at the mercy of their masters’ caprice. . . . To say it all in a word, they must serve every imaginable function.46

At Marsassoum, men, women, and children were sold into bondage. Father Lacombe reported the price of a child to be 200 to 250 francs.47 He added, “I was able to purchase the freedom of a few youths from Fogny.” Eight of the newly liberated youths were to be sent to the mission of the Holy Ghost Fathers in Dakar to receive religious instruction. Ultimately, the priest hoped that these Diolas would serve as auxiliaries to help spread the Catholic faith. It is worth noting that the most remote parts of Fogny are only about fifty miles from Marsassoum. These Diola captives were within a few days’ march of their homes. One might expect this proximity to have increased the likelihood of flight and lowered their value as slaves. It is true that from Marsassoum, many were driven further east into Pakao; in addition, the crocodile-infested Soungrougrou was a barrier to flight. Father Lacombe observed Diola refugees crowded along the banks of the river, imploring passersby to transport them across to Pakao: “I watched in pain as men, women, and children, clumped along the shore under the mangrove bushes, squatted in the mud at the mercy of the intemperate weather, and I listened without, alas, being able to help them, as they called to travelers navigating on the river, imploring to be carried across to the other bank.”48 The fact that these captives were enslaved so close to their homes further suggests that a high proportion of them were children who would have been unlikely to escape.49 It also suggests a level of devastation and social dislocation in the wake of the slave raids so severe that the captives may literally have had no home community to which they could return. In the places Fodé Kaba attacked, Fogny society had been essentially destroyed. Between the slaves and the refugees, many communities no longer existed.

126

“portuguese” style and luso-african identity

The warfare and the slave raids gradually diminished after 1886, when Fodé Sylla’s defeat diminished one source of danger. More important to security over the long run was the establishment of colonial administration throughout the Gambia and Casamance regions. Borders were fixed between French territory and British territory; this facilitated cooperation between military forces on both sides of the Gambian border and led to a coordinated campaign against Fodé Kaba and Combo Sylla.50 In 1893, the Jahanké warrior was forced to sign a treaty recognizing French sovereignty over Fogny; this effectively ended fifteen years of warfare in the region. The establishment of a French military and administrative post at Bignona in central Fogny in 1894 further marked the transition to colonial rule and facilitated the imposition of peace. What happened to the refugees or to those Diolas who had been enslaved? No census was taken of refugees in the first years of French presence in Fogny. Oral testimony collected in the Marsassoum-Sedhiu region in 1975 indicates that some individuals living in that region are the mandingized descendents of Diola slaves.51 Many of the refugees who had fled to villages along the Soungrougrou at the time of Père Lacombe’s visit did return to their home villages following the treaty of 1893. In 1899, the résident at Bignona reported that in Fogny, returned former slaves were fighting to regain their rice fields from the followers of the warrior-chiefs.52 The rapid expansion of the wild rubber trade in Fogny and French Combo also attests to the repopulation of those regions as early as 1897.53 By 1909, Fogny was again densely settled. French estimates of population were as high as twenty persons per square kilometer.54 The land from which one comes is vitally important to the Diolas. Not only is it a source of sustenance and a possession inherited from the ancestors, the earth itself is associated with spiritual forces (bekíín; pl. ukíín) who are responsible for life and health. Ritual offerings are made at shrines (also called ukíín) affiliated with the extended family, the lineage, or the village. Diola religion has been characterized as “a religion of the land.”55 The connection to the land of one’s origin is as much religious as economic. Refugees would have had very strong incentive to return home after the defeat of Fodé Kaba. By the early years of the twentieth century, Fogny was no longer a deserted region, although, ironically, northern Fogny villages were again partially depopulated during World War I as young men fled to The Gambia to escape conscription.56 Lasnet’s description of underground dwellings in Fogny, which was published in 1900, was undoubtedly based on observations made a few years earlier. The houses had been constructed during the period of disruption and danger, probably before 1893. Lasnet’s account of the lifestyle of the Diola-Fogny depicts a people still under siege. Indeed, while security in the region improved after the French post was established at Bignona, Fodé Kaba did not completely cease his incursions. Using the Gambian border as protection, he continued sporadic raids against the

Casamance Architecture from 1850 to the Colonial Administration 127

25. Impluvium house at Kajinol, 1976. Photo by Peter Mark with thanks to Robert Baum.

Diola. Finally, in 1901, a concerted French-British military expedition cornered him at Médina in The Gambia and killed him. With his death, the distinctive defensive dwellings of the Diola-Fogny were no longer needed. K A SA FA R M H O U S E S : C O L U M N S , CA P I TA L S , A N D T WO - S TO RY S T RU C T U R E S Just as the fortress-like farms of Thionk-Essyl had elicited Bertrand-Bocandé’s praise in 1850, so too the large farmhouses of the southern Diolas, which were safe from the distant raids of Fodé Kaba, deeply impressed colonial observers at the end of the century. In 1899, Lasnet extolled these large, rectangular adobe structures with their rounded corners and seven or eight rooms. Each of these “true Norman farmhouses” lodged a single family, “the manpower that cultivates the same rice fields.”57 These massive dwellings were also durable, lasting through many rainy seasons. Lasnet was particularly impressed by the arcaded colonnades that embellished some of the structures: “These buildings show an oriental decoration with superimposed colonnades and arcades of pointed arches that can hardly be attributed to so primitive a population; most are roofed in clay so as to avoid any risk from fire and some have a second story accessible by an interior circular staircase.”58

128

“portuguese” style and luso-african identity

26. Impluvium house at Elana showing the interior veranda, circa 1954. Photo by LouisVincent Thomas and Henri Goovers.

A decade later, Anfreville de la Salle adds details about another variant of southern Jola farmhouses. These massive structures, less common today than they were a century ago, are built around a courtyard where cattle are kept at night to guard against theft (Figure 25). Furthermore, the columns are topped by capitals that are sometimes adorned with shells. The houses, rectangular and solidly built of thick adobe walls, are pierced by narrow doorways and windows covered by wooden bars. In the middle of the crescent-shaped structure is a courtyard . . . where the cattle are kept. Each house is inhabited by a single family. . . . Often the houses are constructed with columns whose rudimentary capitals are decorated with mosaics of shells.59

A few such impluvium houses survive today. They often incorporate an interior veranda (Figure 26). There, in previous centuries, family members could safely sleep in the outside air without fear of attack by slave-raiding neighbors.60 This impressive domestic architecture does not coincide with either Lasnet’s or Anfreville’s preconceived notions of the Diolas. The latter author adds that these houses, though beautiful, are dirty and dark. Lasnet, for his part, has difficulty accepting that the Floups were capable of such impressive architectural design and construction; he postulates foreign influence (see below). Nevertheless, in spite of his assessment of the Diolas as “primitive,” Lasnet is a careful observer of their material culture. His may be the earliest description of their distinctive two-story buildings that were supported by columns and arcades. A century later, similar houses are still being built in the Kasa community of Mlomp.

Casamance Architecture from 1850 to the Colonial Administration 129

27. Two-story house in Mlomp, 1974. Photo by Alain Hardy.

28. Two-story house in Mlomp under construction, 1974. Photo by Alain Hardy.

130

“portuguese” style and luso-african identity

These houses, constructed in adobe61 and durable enough to last for decades, are unique to Mlomp (Figure 27, Figure 28). In recent decades, they have become a tourist attraction. The origins of this distinctive style have not, however, been traced. Lasnet does not specify the community whose buildings he describes. While he probably refers to Mlomp, it is possible that these two-story houses were formerly found in neighboring villages, too. Ironically, Lasnet’s account, by its date, effectively disproves interpretations that are recounted by some present-day inhabitants of Mlomp of the history of the buildings.62 Informants asserted that the earliest two-story dwellings in Mlomp were built shortly after World War I by an ancien combattant who had been inspired by the tall buildings he saw in Europe during the war. Since Lasnet’s account predates the war by fifteen years, this explanation cannot literally be true. The Buluf-Diola are intensely conscious of their history and frequently seek to attribute the greatest possible age (or the earliest possible date) to events or traditions they deem important. The Diolas of Esulalu, the region that includes Mlomp in Kasa, also have a strong historical consciousness and they, too, frequently ascribe importance to events and traditions thought to have occurred at an early date.63 Yet here is an instance of a recent oral account that, rather than exaggerating, actually truncates the age of an important element of local culture. Why should this be? Perhaps it is because France is still perceived by some, forty years after Independence, as the epitome of an advanced culture.64 Hence the imputation of French origins may be seen as more prestigious than chronologically earlier or geographically closer origins. The history of this distinctive architecture raises several questions. When was the style introduced in Mlomp? Were two-story houses formerly more widespread in the Casamance? Is there an historical link to colonial architecture or to the architecture of another West African culture? Independent local invention of this distinctive style cannot be ruled out a priori. However, the incorporation of columns and capitals, rounded arches and arcades, combined with a second story all in the same structure when none of these elements is documented elsewhere among the Floups or their Diola descendents is suggestive of outside influence. When did the people of Mlomp develop their distinctive building style? The history of migrations from south of the Casamance River to Buluf may provide a clue to the dating. The inhabitants of Thionk-Essyl trace their ancestors to Essyl south of the river, while the adjacent Buluf community of Mlomp traces its origins to immigrants from the southern Mlomp. These two migrations are said to have occurred at about the same time. The settlement of Thionk can be dated to about 400 years ago,65 which suggests that Mlomp-Buluf, too, was settled by Diola-speakers around 1600.66

Casamance Architecture from 1850 to the Colonial Administration 131

29. Etablissement anglais à Georges Town (Gambie), 1851. Hyacinthe Hecquard, watercolor sketch. Courtesy of Frobenius-Institut, Universität Frankfurt am Main, Bilder-Archiv.

Neither Buluf community has any two-story buildings, nor is the local use of columns, round arches, or arcades documented in either written or oral sources. Bertrand-Bocandé makes no mention of these features in his 1850 report, although his references to “petites tourettes” and to “petites forteresses avec leurs tours” may be construed to imply that part of the extended family compound was multistoried. Nor did older informants in Thionk or Mlomp make reference to houses with a second floor when queried in 1975 about their recollections of precolonial architecture. Such buildings are not part of local historical memory. If the architectural style of Mlomp had developed prior to the migration northward, evidence of that style would likely have existed in Buluf. No such evidence exists. “Portuguese” prototypes existed for two-story buildings with roundarched arcades. There were such buildings at Santiago in the Cape Verde Islands as early as the sixteenth century. Some of these architectural elements were to be found much closer to the Casamance after the construction of a Portuguese fort and church at Cacheu in the 1580s. Some southern Floups engaged in trade with the Luso-Africans at S. Domingos as early as the sixteenth century. “Portuguese”-style architecture was adopted by some Floups at Bolole, and it may have influenced their neighbors in Mlomp, which also lay within the commercial sphere of the “Portuguese” at S. Domingos. The sixteenth- or early-seventeenth-century establishment

132

“portuguese” style and luso-african identity

30. Former trading station of Maurel & Prom built in Ziguinchor in the late nineteenth century. South façade. Note the recently walled-off second-floor gallery. Photo by Peter Mark.

of a trading post at Ziguinchor on the Casamance River67 brought the “Portuguese” even closer to Mlomp. Luso-African architectural influence is thus possible. But the absence of a similar style in Mlomp-Buluf and Thionk-Essyl argues against “Portuguese” influence much before the midseventeenth century. In the nineteenth century, arcaded two-story buildings existed at Bathurst (Banjul), at Albreda (as in Figure 20), and at Georgetown in The Gambia, which is depicted by Hecquard in an 1851 sketch (Figure 29). These structures could have served as inspiration for the houses at Mlomp. It is improbable, however, that Gambian influence should appear in Mlomp (which is two days distant from The Gambia by sea) and nowhere else in the Lower Casamance. Although Mlomp traditions incorrectly represent the chronology of the first two-story houses, the association with French architecture may contain a grain of truth. The inspiration for these buildings may have come not from Metropolitan France but from the colonies. Models for two-story verandas in the form of superimposed galleries with rounded arches existed in French colonial buildings in Senegal before 1850. By

Casamance Architecture from 1850 to the Colonial Administration 133

31. Gorée, Hôpital militaire, 1820–1845. View from the northeast. The bay that is farthest to the east was added shortly after 1863. Photo by Peter Mark.

1890, they were found in Ziguinchor, a relatively short trip from Mlomp by pirogue. The trading house of Maurel & Prom, which was constructed shortly before that date, located on the waterfront in Ziguinchor, incorporated a two-story veranda with superimposed galleries. This structure survives, although it has been extensively remodeled (Figure 30). The upper porch has been walled off and the original brick and basalt walls covered over with plaster.68 If the legendary Mlomp architect lived in the mid-nineteenth century, two other sources of inspiration may have been available to him, one on Gorée Island adjacent to Dakar and the other at Sedhiu in the Middle Casamance. The French established a settlement at Sedhiu, which is less than 100 kilometers to the east of Ziguinchor, in 1836–1837. It soon became the center for their commercial activity in Casamance. A fort, begun in 1838, features superimposed galleries with rounded arches. Sedhiu could have been reached in a day’s sail from Carabane. Gorée, a French naval base, lay further afield, a few days’ sail or one day by steamer north along the coast. The Hôpital militaire of Gorée (Figure 31, Figure 32), which is built on the edge of a bluff overlooking the water, is a majestic two-story edifice. Each story is 4.2 meters high. The north façade is a classic example of the two-story gallery with superimposed arcades; both stories incorporate eight rounded arches. The present structure, which was built before 1850—historians have posited widely di-

134

“portuguese” style and luso-african identity

32. Gorée, Hôpital militaire. North façade. Photo by Peter Mark.

vergent dates—was enlarged after 1863.69 This monumental structure, which dominates the center of Gorée, would very likely have impressed any architecturally minded southern Floup who arrived there as a gourmette or a tirailleur.70 Furthermore, the most impressive elements of this building besides its scale are the superimposed arcades and the rounded arches—precisely the elements that distinguish the Mlomp houses. Any of these colonial buildings, or several of them cumulatively, could have served as a model for the houses. An historical association with Gorée, which is certainly plausible, would further suggest a mid-nineteenth-century date for the Mlomp architecture. A RC H I T E C T U R E A N D C O L O N I A L D I S C O U RS E As early as the beginning of the sixteenth century, Floup architecture had impressed European visitors to the Casamance with its scale, its solidity of construction, and its complexity of form. By the late nineteenth century,

Casamance Architecture from 1850 to the Colonial Administration 135

33. Rip, Senegal, Wolof, or Serer housing compound. Photo by Peter Mark.

however, this positive regard was incorporated into and modified by an increasingly pejorative colonial discourse about West African culture. During the early decades of the colonial period, French observers echoed the positive image of Casamance architecture that had characterized descriptions from Fernandes and La Courbe through Bertrand-Bocandé. Indeed, Floup houses were often singled out for praise, while they at the same time were contrasted to buildings of other Senegambian peoples. A late example of this juxtaposition occurs in 1909, in the writings of missionary Père Wintz: “The traveler crossing Diola country is struck with admiration at the sight of the beautiful houses that the natives build for their families. This is not the small straw hut, dark and narrow, of some parts of Africa, but a large and impressive adobe structure.”71 The contrast in construction material and form would have been especially vivid for visitors accustomed to the small, round structures of the Wolofs and Serers that were characteristic of northern Senegal (Figure 33). In fact, Jola farmhouses are not that different from the architecture of some of their southern neighbors in the Casamance and Guinea-Bissau, including the Papels and the Balantas. For French colonial officials and missionaries whose experience was primarily of the region north of the Gambia, however, the contrast tended to be more striking. After the southern Basse Casamance became part of Senegal in 1886,72 the impressive architectural accomplishments of the Diolas were incorpo-

136

“portuguese” style and luso-african identity

rated into a discourse affected by both colonial power relations and a disparaging French view of African culture. The Casamance was the last region of Senegal to be pacified. Armed opposition to the imposition of taxation and to the establishment of French military posts continued both north and south of the Casamance River until World War I. Opposition to taxation and to the exercise of colonial jurisdiction earned some Diolas a reputation that they were living in a state of complete anarchy.73 In 1902, the résident at Bignona described the Diolas of Fogny as comprising “two clearly separate peoples” that were distinguished from each other by whether or not they accepted French authority! Those who did, he characterized as “hardworking, peaceful in character.” Those who did not, particularly those villages that refused to pay the impôt, were designated as “the Flouk [sic] peoples, mean in demeanor and warlike in character” [la race Flouk à l’aspect farouche et au caractère belliqueux].74 The latter group faced the threat of military action like the French attack that had burned parts of Thionk Essyl and Mlomp in 1895.75 That colonial attack demonstrated the contradiction inherent in French attitudes toward Diola culture. In order to force these warlike primitives to accept the mission civilisatrice—i.e., to get them to pay the head tax—it was deemed necessary to burn down the very farmhouses that were among the most impressive architectural landmarks of French West Africa. The irony of this position was lost on the colonial administration. As opponents of colonial rule, the peoples of Casamance were not, of course, viewed as évolués [educated people]. The first chefs de canton named as intermediaries between the French administration and the local population were non-Diolas. They were chosen partly because of the linguistic barrier, but partly too because the French tended to view the Diolas as representing a lower level of cultural evolution than the Muslim inhabitants of the north. Given this political and military context, the earlier highly positive assessment of Diola architecture was modified. These impressive dwellings came to embody a double contradiction. On a general level, they were seen as an exception to the presumed West African norm of insubstantial and ephemeral houses. At the same time, the buildings appeared to constitute something of an aberration with respect to the Diolas’ own supposedly lower level of cultural evolution within the hierarchy of Senegalese societies. This double contradiction led European commentators to mitigate their praise of Casamance architecture in two ways: either they contrasted it with a presumed less evolved African norm or they depicted the buildings as an admittedly impressive anomaly within Diola culture. The former approach is taken by Père Wintz when he writes, “This is not the little straw hut of some parts of Africa.” The second approach is taken by those observers who qualify their praise for Diola farmhouses by calling attention to less impressive structures built elsewhere in the Casa-

Casamance Architecture from 1850 to the Colonial Administration 137

mance or to less impressive aspects of Diola culture. Lasnet’s extensive discussion of the underground filthy dwellings of the Diola-Fogny, for example, tempers his description of the commodious two-story farmhouses of Kasa by making the latter appear as an exception, even among the Diolas. Lasnet could not deny the incontrovertible evidence of Diola architectural achievement. Instead, he and other observers took pains to elaborate theories of cultural diffusion that would enable them to attribute to other, presumably more “advanced,” cultures the inspiration and requisite technological expertise for these buildings. Lasnet enumerates three hypotheses. The first, “that the Diolas had been in contact with the Sudanese peoples of the Middle Niger from whom they learned [to build] that architecture one still finds at Jenné and Timbuktu,”76 he rejects out of hand; the two styles are simply, in his judgment, too different. The second hypothesis, which he cites without comment and implicitly rejects, is that of distant Phoenician influence. Here, one perceives an echo of the theory, which was elaborated at approximately the same time, of Carthaginian influence on the architecture of Great Zimbabwe. In the context of earlytwentieth-century attitudes toward race and culture, the efforts of European observers to attribute Diola architecture to Mediterranean influence could, ironically, be seen as the highest compliment! Lasnet’s third hypothesis, which was accepted by “le plus grand nombre” and hence, in his view, presumably the most plausible, is that the Diolas had been influenced in their housing style by the Portuguese: “Most observers, including Reclus, attribute these buildings to the Portuguese influence which made itself felt along the coast beginning in the fifteenth century.”77 This theory, first proposed in the eighteenth century,78 elides the distinction between European Portuguese and Luso-African “Portuguese.” It also ignores the cultural complexity of the Portuguese and Luso-African communities. Most important, it overlooks the fact that “Portuguese” communities were heavily influenced in their own architecture by local African cultures, including the Floups/Diolas. In other words, it is likely that Lasnet reverses the historical process whereby local building techniques and materials and, probably, design played a crucial role in the development of the building style that came to be known as maisons à la portugaise. Lasnet’s theory of Portuguese influence on the Floups is characteristic of early-twentieth-century French assessments of African architecture. It is pertinent to ask why Lasnet dismisses the theory of Sudanese (Mande) influence out of hand in favor of Portuguese origins. One explanation is the existence, since the eighteenth century, of a literature that developed the theory of Portuguese influence on domestic architecture in the Casamance-Bissau region. A second, closely related factor is that the Portuguese hypothesis imputed European origins to these buildings, whereas the “Jenné hypothesis” implied African origins. At the moment when

138

“portuguese” style and luso-african identity

34. Kankétéfa, Résidence du Roi Bakar, Paquési, 1851. Hyacinthe Hecquard, watercolor. Courtesy of Frobenius-Institut, Universität Frankfurt am Main, Bilder-Archiv.

colonial authority was being imposed, the former explanation was consistent with the dominant French image of the Diolas as uncivilized. In this context, it is also relevant to note that Lasnet’s book was published in conjunction with the Paris Exposition Universelle of 1900. The tendency to ascribe Portuguese origins to local African architecture that was particularly impressive or monumental affected even nineteenth-century observers who were aware of African technical skills. Hecquard recounts an episode that illustrates the process by which he came to recognize Mande engineering expertise. Traveling through the eastern Gambia region and Paquési in 1851, he arrived at Kankétéfa, a town of 2,000 inhabitants. There he observed that the king’s compound was surrounded by a double rampart twelve meters high and one meter thick. The fortress walls traced a series of acute angles; the walls were pierced with slits to enable the defenders to pour a double enfilade of fire upon any attacking force. The projecting salients were in turn flanked by two-story towers. Hecquard sketched the fortress at Kankétéfa (Figure 34).79 He then asked several local rulers how they had come to build in this manner: Seeing these defenses, which truly have something of European science and which one finds in Bundu and in Bambuk, several travelers have voiced the opinion that these people had undoubtedly constructed them after having seen our trading comptoirs [posts]. A few have even suggested that the Por-

Casamance Architecture from 1850 to the Colonial Administration 139 tuguese must have taught them the rudiments of [building] science. I too was struck by the regularity of form and the power of these tatas [forts] . . . and I asked several chiefs why they built their walls with acute angles. . . . They were unanimous in answering me by pointing out a rather obvious thing that I had not thought of: That is, that it would be impossible for them to do otherwise because a simple [flat] wall of this height and length, constructed of earth, would have absolutely no solidity.80

A direct and forthright conversation with local leaders thus disabused Hecquard of the notion that the people of Bundu and Bambuk had acquired the necessary engineering skill from Europeans. His was a rare, but not unique, case. Early- and mid-nineteenth-century observers from René Caillié to Bertrand-Bocandé avoided theories of foreign architectural influence. Nevertheless, as the colonial conquest progressed, Europeans— most of whom were military officers or administrators—were rarely prepared to credit Africans with the technical knowledge necessary to have constructed solid and lasting buildings. The French effort to incorporate Diola architecture into a broader colonial discourse and, more specifically, to deny indigenous origins to local architectural achievements may be likened to the process whereby ethnographers and colonial officials, notably Maurice Delafosse, sought to attribute the creation of the ancient empire of Ghana to a white Semitic dynasty. As Jean-Louis Triaud has recently demonstrated, Delafosse’s theory, not unlike the well-known “Hamitic myth” with which it was roughly contemporaneous, ascribed the origins of Ghana to a white (in this instance Judeo-Syrian) immigrant population.81 Delafosse subsequently refined this theory to suggest that the progressively darker-skinned descendents of Semitic immigrants and local African women ultimately engendered the Soninké rulers of Ghana.82 He writes, “It was they, undoubtedly, who built the stone structures of Koumbi and who took over power.”83 In this articulation of a theory of foreign cultural influence, the establishment of political power and the manifestation of cultural superiority through architectural accomplishment are perfectly equated. To build impressive and monumental structures is to give concrete evidence of one’s superiority. Not coincidentally, elevated cultural levels are also associated with a heritage of light-complexioned ancestors. Colonial administrators and scholars thus elaborated an historical model of African material culture and cultural history in which architectural achievement was tied to notions of an evolutionary hierarchy. Architecture, in turn, according to this model, served as concrete evidence of the a priori cultural superiority of light-skinned peoples and their descendents. A similar assessment of Diola domestic architecture is expressed by the French chronicler Anfreville de la Salle. In 1909, Anfreville praised Casamance houses, but he tempered his praise by making disparaging

140

“portuguese” style and luso-african identity

comments about Diola living habits. Here, he follows Lasnet: Casamance architecture is presented as truly impressive, but it is also portrayed as an anomaly within Diola culture. Anfreville, too, then hypothesizes Portuguese origins for the houses. In effect, he argues that if these dwellings differ from the expected African norm, it is because the architecture must have originated elsewhere. The similar organization of the two texts suggests that Anfreville may have modeled his account on Lasnet’s earlier study. Anfreville’s description of Casamance culture is tempered, or, rather, distorted, by his racial theories. He views the Casamance as inhabited by “différentes races”; these races, each distinctive, are in turn composed of “tribus.” Several of the races, presumably including the Diolas, he situated “on the lowest rung of human groups.” Nevertheless, the Diolas construct “veritable houses that one is astonished to find built by these people.”84 Anfreville was enthralled by the houses of the southern Floups. Not surprisingly, however, he was unable to attribute such evidence of civilization to a people who were presumably so low on the scale of human cultural development. Accordingly, he called attention to less impressive aspects of their material culture. Floup houses may be large and impressively built but, he also observed, they were dirty and dark. (Here, too, Anfreville follows Lasnet’s description of the dwellings in Fogny.) He added, “If the Floups are remarkable for the relative beauty of their homes, they nevertheless dress in the same rude fashion as all the other Diolas.”85 Hence, they were not so civilized as their architecture might suggest. Just as three years later Delafosse would attribute the material culture and political organization of ancient Ghana to Judeo-Syrian immigrants, so too Anfreville and Lasnet attributed Diola architecture to Portuguese influence: “It is said that they [the Portuguese] were . . . the architectural teachers of the Diolas. This is perhaps going a bit far, but the dwellings of these lusitanized Blacks appear to be well built after a European model that has hardly been changed at all.”86 The term “noirs lusitanisés” is an oblique reference to the Luso-Africans of Ziguinchor; the town had only recently become a French possession. This is the only context in which French colonial observers consider that group, whom they generally disparage, to be in any way “Portuguese.” And of course the sobriquet of Blacks with only a veneer of Portuguese heritage (“noirs lusitanisés”) serves to create a rather disparaging image. In Anfreville’s eyes, these aren’t really Portuguese after all. Nevertheless, in the context of early-twentieth-century European attitudes toward race, the fact that Diola buildings should be attributed to people of even remote European origin constitutes backhanded recognition of the beauty and refinement of Floup domestic architecture. French colonial observers were not inclined to praise the culture of a people who were still

Casamance Architecture from 1850 to the Colonial Administration 141

actively resisting the imposition of colonial authority . . . especially when it was occasionally deemed necessary to burn down the houses the colonizers might otherwise have praised. In the writings of these three contemporaries—Delafosse, Lasnet, and Anfreville—the concept of race is inextricably interwoven with the perception and interpretation of African material culture. Anfreville’s use of the phrase “noirs lusitanisés” to refer to the Luso-Africans and perhaps also to the Floups (the antecedent of his term is rather vague) parallels Delafosse’s characterization of the Soninké as “des descendants de plus en plus métissés” [increasingly dark-skinned descendents] of the founders of ancient Ghana.87 Indeed, while Anfreville’s term “lusitanisés” refers ostensibly to the architecture and material culture of the Casamançais, the implicit reference is quite clearly also to their mixed genetic heritage. Delafosse’s description of the Soninkés’ presumed Semitic ancestry focuses on intermarriage (“by taking wives among their subjects who belonged to the black race”).88 The ensuing assumption of political power which Delafosse postulates thus takes on an aura of cultural and political decadence. To a degree with Delafosse, but to an even greater extent in Anfreville’s description of the people of Casamance, the distinction between culture and race is elided. Anfreville fuses the two concepts. In these descriptions of both ancient Ghana and turn-of-the-century Casamance, the perceived level of African material culture serves to confirm the European observers’ notions of their own superiority. More specifically it is architecture, whether it be the farmhouses of the Diola or the stone buildings of Kumbi Saleh, that embodies the cultural superiority of Europeans and their descendents. In the system of immutable categories that rigidly situates each group within the colonial classification of races, material culture embodies and confirms genetic heritage. E N VO I Early European settlers on the Upper Guinea Coast were assimilated into a world comprising cultures whose membership was fluid and often contextually determined. Membership in these groups was largely defined by parameters that were cultural; this magnified the flexible nature of precolonial identities. Consequently, individuals could and often did change identities, depending upon the specific context in which they found themselves. Among the cultural characteristics that served to define one’s social identity, both in terms of profession and social status, a significant and dual role was played by material culture and specifically by domestic architecture. The buildings people constructed conveyed a concrete sense of their wealth and prestige. They also served as markers that helped to identify the inhabitants as members of a particular social class; for example, long-distance traders. But since one’s profession, particularly in the case

142

“portuguese” style and luso-african identity

of traders, also constituted an important aspect of one’s belonging to a particular group, domestic architecture could serve as a marker of group identity. From the sixteenth to the eighteenth century, European observers, themselves generally merchants, readily associated one housing type that had been adapted by long-distance traders as “Portuguese” style. These foreign observers recognized the rectangular whitewashed buildings that had a vestibule or veranda to accommodate visitors as a characteristic of the “Portuguese,” or Luso-African, community. Even when these structures were built by members of other local African societies, they were widely identified as maisons à la portugaise. To construct and live in such a building was to assert one’s status as (relatively) wealthy and as a trader. Provided that the individual showed the other cultural characteristics of “Portuguese” identity—language and religion—these houses also served to establish one’s “Portuguese” identity to outsiders. Hence, in a society where identity was not based on a priori characteristics such as skin color, and in which a person might readily become a member of a different cultural or “ethnic group” than either of his or her parents, architecture was among several criteria that permitted one to establish a given identity. In a world where identity was flexible, domestic architecture was an important means of maintaining or, presumably, of changing one’s identity. By the late nineteenth century in Senegambia, the establishment of European economic predominance, followed by the imposition of colonial military and political control, was accompanied by the spread of notions of identity based on physical characteristics. Competition between France and Portugal for political control of the Casamance accentuated French efforts to redefine the “Portuguese” of Ziguinchor as black Africans. Identity was now perceived to be inborn and inherited and hence, for the individual, immutable. Furthermore, one’s membership in a particular group was thought to be associated with intellectual capacity and, by extension, with cultural achievement. The result of these developments was the articulation of a rigid categorization of humanity. Each individual now fit into a fixed group, to which he or she was born. Culture, according to this system, rather than being an expression of malleable and even multiple identities, became instead the almost inevitable result of a people’s heritage. The imposition of rigid a priori categories not only transformed the discourse on identity, it also fundamentally changed the role of architecture in this discourse. Writers such as Delafosse and Anfreville essentially reversed the role of architecture as a marker of West African identity. Instead of being a flexible and freely chosen means of establishing one’s social and cultural identity, the kind of dwelling one lived in became a reflection of one’s inherent nature. Thanks to some ex post facto historical hypothesizing, architecture now also served as an indication of the presence or absence of European or

Casamance Architecture from 1850 to the Colonial Administration 143

Semitic blood in one’s ancestry. At the very least, one’s housing style reflected the influence of European culture upon one’s African ancestors. For colonial officials and administrators such as Delafosse and Lasnet, architecture assumed a central place in the establishment of a rigid system of categories by means of which African cultures were interpreted and redefined.

Conclusions and Observations

The history of the structures that people build and live in constitutes an important element of the broader history of culture, and it is appropriately addressed from an historical perspective. This perspective illuminates the evolving social importance and symbolic significance of architecture within the communities that built the houses. But architecture is also situated within a wider social setting: the contact among different groups that helps to determine the exchange of technologies and of styles. Like any other aspect of culture, building forms and styles do not evolve in isolation. The history of Upper Guinea Coast architecture demonstrates the central and continuous role played by contact among cultures in the elaboration and evolution of both forms of housing and the social meanings ascribed to these forms. The history of architecture is, therefore, situated within and subsumed by broader issues of culture contact. Along the West African coast from the fifteenth to the nineteenth century, local societies were characterized by extensive cultural interaction. This process created a distinctive mixed culture, particularly among a social elite composed largely of traders. The cultural métissage that defined this group is itself a proper object of study. Material culture illuminates the process of métissage. Architecture, particularly the houses of wealthy African and Euro-African traders, was central to the formation of seventeenth-century mixed culture; architecture also constituted an integral part of the local identity discourse. Material culture, specifically the houses built by wealthy and powerful individuals, reflects the nature of cultural interaction and assimilation. An architectural history of seventeenth-century southern Senegambia reflects broader processes of interaction between cultures. For the GambiaCasamance-Bissau region, particularly in the sixteenth and seventeenth centuries, this wider process was distinctive. These cultures did not define themselves by an oppositional model of identity in which members of a specific group conceived of themselves primarily by reference to an “other” or to people or groups whom they perceived as different. Rather, local models of identity were characterized by fluidity and by the fact that different identities were not perceived to be mutually exclusive. Instead, individuals could and often did assume more than one identity, sometimes sequentially but often contextually contingent. Categories of identity were not fixed. Rather, they were flexible, allowing individuals and sometimes

Conclusions and Observations

145

communities to assume different identities at different times. Furthermore, these identities were characterized by their assimilative nature; they developed through appropriation and mutual transformation, or again, through what we might term a process of cultural métissage. This process was not characterized by boundaries that delimited and separated, at least not in the Western sense of the term. T H E O R E T I CA L A N D M E T H O D O L O G I CA L I M P L I CAT I O N S A study of the construction of architecture and identity in precolonial Senegambia may afford insights into processes whereby material culture and cultural identity are created and reformulated elsewhere in West Africa. This study may also afford a broader theoretical and methodological perspective. The history of material culture may itself yield a wealth of information about the intellectual processes whereby members of particular societies conceptualized themselves and understood their relationship to others. Specifically, architecture plays an important role in the discourse about group and individual identities. During the eighteenth and nineteenth centuries, this discourse was of critical importance as part of the interaction between West Africans and Europeans. After 1850, the growing economic and political authority of Europeans enabled them to impose a particular identity discourse, one that contested and ultimately transformed local identity models. Methodologically, too, it is my hope that the present study may provide a useful point of departure for other cultural and intellectual historians. To study the history of indigenous discourses on identity, given the paucity of written sources that address the issue and the dearth of sources written from a local perspective, the historian needs to find other, nonwritten sources. Such sources exist. They include both manners and etiquette. Seventeenth-century Gambian trading society provides a case study that illustrates the importance of manners as historical documents reflecting cultural interaction. To reconstruct a history of discourses on identity, to observe and interpret local etiquette or manners, and to interpret the social meanings of local architecture, the historian must rely almost exclusively on written descriptions. Precolonial written sources are abundant for the SenegambiaBissau region in comparison to what is available for other areas of West Africa. Because most of these sources were written by Europeans, however, an additional cultural filter is interposed between the African communities and the historian. Fortunately, several early sources were written by Cape Verdeans, members of the “Portuguese” community. Nevertheless, caution is warranted in the interpretation of these written sources; the historian is here using written sources to document areas of human activity that would normally be accessible through direct observation.

146

“portuguese” style and luso-african identity

In architecture as well as in comportment and interpersonal relations, style reflects social discourse and, by extension, attitudes. Social etiquette, or manners, and domestic architecture both serve as indicators for a history of ideas. Style, in both of these domains, reflects the intimate interaction of different groups and ultimately the establishment of common ground. This common ground constitutes the crucible for the formation of new social and cultural identities. R E F L E C T I O N S O N CAT E G O R I E S In conclusion, a short commentary may reflect my personal perspective regarding social categories and identities as well as the wider context of late-twentieth-century American society in which these chapters have been written. It seems to me that the crucial feature of precolonial social and cultural history along the northern Upper Guinea Coast is precisely the flexibility or malleability of individual identity and the apparent ease with which people could assume multiple, or at least a sequence of, different identities. Historical sources strongly suggest that these characteristics were common to the many African and Euro-African groups throughout Senegambia from the sixteenth to the nineteenth century. In all likelihood, a similar situation prevailed before the earliest fifteenth-century written records. Written sources are clear on this point. Nevertheless, my interpretation of these sources is certainly influenced by my experiences living in and carrying out field research among the populations of the Casamance region, primarily among peoples who today identify themselves as Jola or Manding. Flexibility of individual identity has long been a prominent characteristic of local society. My awareness of this flexibility and fluidity was certainly conditioned by my experiences in Casamance. I recall, for example, the elderly Jola in Ziguinchor who, when he learned that I was studying Bagnun religion, informed me that he was himself Bagnun. There was also a Jola informant in Thionk Essyl who succinctly observed: “Where you see a Jola, you are actually speaking with a Manding, and where you see a Manding, you are actually speaking with a Jola.” But above all, there was my first visit to Guinea-Bissau. In S. Domingos, a young African was introduced to me as “Portuguese,” with the implication that “he is like you.” My immediate reaction was incredulity. He is only calling himself “Portuguese,” I thought, to ingratiate himself with me. Today, whenever I reread Le Maire or other seventeenth- and eighteenth-century narrators who rejected Luso-African claims to be “Portuguese,” I recall this incident and I cringe. Over the years, these and similar encounters have led me to reflect on what it means to be Jola, Manding, or Portuguese. Eventually, I realized that there was an underlying thread connecting these experiences. Identity is a product of one’s heritage but it is also something one can manipulate. This manipulation occurs, it seems to me,

Conclusions and Observations

147

not in a cynical sense. Rather, one can form one’s self, within limits, in response to one’s social environment, much as an architect works with adobe or other materials to construct his house. This realization, in turn, made sense of disparate experiences, including the above-mentioned anecdotes. After twenty-five years of fieldwork, I finally felt that I had begun to understand the dynamics of group formation and of individual identities in the Casamance. Undoubtedly, this insight, like all revelations that one experiences during fieldwork, has its limitations. But I have yet to come upon a situation in the Senegambian historical record that was not better illuminated when recast in the light of this understanding of the fluidity and multiplicity of local identities. My interpretation of precolonial Senegambian discourses on identity is informed by my experiences in contemporary Casamance, but my perspective is no less affected by my experience of living in American society through the second half of the twentieth century. Questions of identity, particularly “ethnic” identity, and the role of “ethnic minorities” in a multicultural society are central to American social and political discourse. It seemed to me, as I developed the present study, that my observations and conceivably my conclusions would be of direct relevance to contemporary society. Ethnic identity and “color” remain as central to American social and political discourse today as they were a century ago, when the issue was succinctly articulated by W. E. B. DuBois. Elsewhere in the world, the significance of ethnic identity has certainly not diminished either, as events in Eastern Europe, Rwanda or, sadly, in Casamance itself bear witness. I should like to think that in the history of seventeenth-century southern Senegambia there is a message that might benefit contemporary societies. One part of the message is that plural societies that do not rigidly and methodically categorize each individual may thereby encourage sharing or métissage. As a result of this sharing or mixture, the richness of all the component cultures becomes part of a common heritage. Another part of the message is the realization that the boundaries of the categories by means of which we classify and divide individuals and communities are, ultimately, arbitrary. Change the boundaries and you create new communities. Flexible boundaries or multiple identities may lessen distinctions and create new bonds, thereby establishing a broader human community. When I began the study, I initially intended to focus on the history of precolonial Senegambian domestic architecture. The story of the “Portuguese” seemed at first an incidental corollary. However, I soon became enthralled by the history of this community whose members, in direct opposition to my own culture’s understanding of identity, defined themselves almost exclusively by cultural and social parameters and who, moreover, steadfastly asserted their identity as “whites,” regardless of the color of their skin. I found myself rooting for them to maintain their iden-

148

“portuguese” style and luso-african identity

tity in the face of European assertions that they were neither Portuguese nor white. I was even tempted to end this historical study before I arrived at the nineteenth-century denouement. My discomfort was all the greater as I recalled the episode in S. Domingos and realized my own small role in denying Portuguese identity. Perhaps, in some small measure, this book is my way of making amends for that experience.

Notes INTRODUCTION 1. Wallerstein theorized identity formation as a dialectical process as early as 1969 in “Ethnicity and National Integration in West Africa,” Cahiers d’Etudes Africaines, no. 3 (October 1969): 129–138. For a more recent discussion of the subject, see P. Poutignat and J. Streiff-Fenart, eds., Théories de l’Ethnicité (Paris: PUF, 1995), 155. 2. Fredrik Barth, “Ethnic Groups and Boundaries,” in Process and Form in Social Life: Selected Essays of Fredrik Barth (London: Routledge & Kegan Paul, 1981), 198–227. 3. Louis-Vincent Thomas, Les Diola: Essai d’analyse fonctionelle sur une population de Basse-Casamance (Dakar: IFAN, 1959), 1: 13. 4. “Ich verwende den Begriff Ethnicität nicht als streng analytische Kategorie, die sich eindeutig definieren liesse. Ethnicität ist vielleicht ein Schillerndes Konzept.” Carola Lentz, Die Konstruktion von Ethnizität: Eine politische Geschichte Nord-West Ghanas, 1870–1990 (Köln: Rüdiger Koppe Verlag, 1998), 31. 5. “Die Prozesse der Schöpfung ethnischer Identität sind historisch und regional spezifisch.” Ibid., 34. 6. Lemos Coelho spent twenty-three years on the West African coast as a trader (circa 1640–1665), and he lived for several years in Bissau (1660s) and Cacheu. See P. E. H. Hair, “Introduction,” in Francisco de Lemos Coelho, Description of the Coast of Guinea (1684), translated and with an introduction by P. E. H. Hair (University of Liverpool, Department of History: Privately issued by Prof. Hair, 1985), iii. 7. In a similar historical context, Lentz uses the local term “yir” [patrilineage, family] in preference to ethnic labels in a discussion of identity. Die Konstruktion von Ethnizität, 45. 1. THE EVOLUTION OF “PORTUGUESE” IDENTITY 1. The Inquisition was established in Portugal in 1536. 2. See Maria Emilia Madeira Santos, “Origem e desenvolvimento da colonização: Os primeiros ‘lançados’ na costa da Guiné; aventureiros e comerciantes,” in Portugal no Mundo, edited by Luis de Albuquerque (Lisboa: Publicações Alfa, 1989), 2: 125– 136. 3. The Petite Côte extends from Dakar south to Joal. 4. On the economic history of Luso-Africans, see George Brooks, Landlords and Strangers: Ecology, Society and Trade in Western Africa, 1000–1630 (Boulder, Colo.: Westview Press, 1993). See also Brooks, “Perspectives on Luso-African Commerce and Settlement in the Gambia and Guinea-Bissau Regions” (Boston University African Studies Center, Working Papers, 1980); and Jean Boulègue, Les Luso-Africains de Sénégambie, XVIe-XIXe siècle (Paris: Université de Paris I, Centre de Recherches Africaines, 1989). For the seventeenth-century Petite Côte, see Nize Isabel de Moraes, A la découverte de la Petite Côte au XVIIe siècle, 3 vols. (Dakar: Université Cheikh Anta DiopIFAN, 1993–). For Serra Leoa, the southeastern part of “Guinea of Cabo Verde,” the history of the Luso-African community in the sixteenth and seventeenth centuries is covered by the richly annotated translations of Portuguese, Dutch, and English documents published by P. E. H. Hair in Africana Research Bulletin between 1974 and 1981. 5. Dierck Ruiters, Toortse der Zeevaert [The Torch of Sea Travel], translated by P. E. H. Hair in “Sources for Early Sierra Leone (2): Andrade (1582), Ruiters (1623),

150

Notes to Pages 14–15

Carvalho (1632),” Africana Research Bulletin 5, no. 1 (October 1974): 47–56. Dierck Ruiters was a Dutch sea captain who made several trips to West Africa and participated in Dutch military campaigns to capture Brazil from Portugal, all between about 1602 and 1634. His The Torch of Sea Travel (1623) provides information about the African coast from Senegal to the Congo. 6. For references to this presence, see Francisco de Andrade, “Relação de Francisco de Andrade sobre as ilhas de Cabo Verde (1582),” in Monumenta Missionaria Africana, edited by António Brásio. Series 2, Volume 3: Africa Ocidental, 1570–1600 (Lisboa: Agência Geral do Ultramar, Divisão de Publicacões e Biblioteca, 1964), 102. The entire text of “Relação de Francisco de Andrade” is on pages 97–107. Donelha exists in both English and French translation. The English translation as well as the annotations in both editions are by P. E. H. Hair. Except where otherwise noted, I refer to the French edition: Déscription de la Serra Leoa et des Rios de Guiné du Cabo Verde (1625), edited by A. Teixeira da Mota, translated by Léon Bourdon (Lisbon: Junta de Investigações Científicas do Ultramar, 1977). Donelha mentions a tangomao named Luis Lopes Rabelo who had spent years at Rio Nunez (163). He also writes that south of the Rio Grande and as far as Sierra Leone, “there is no large river to which ships might travel to trade, but the barques of the tangomaos do come there to purchase a few Blacks, wax, and ivory, by traveling from one port to another” (179). 7. See Madeira Santos, “Origem e desenvolvimento da colonizaçao.” 8. See José da Silva Horta, “Evidence for a Luso-African Identity in ‘Portuguese’ Accounts on ‘Guinea of Cape Verde’ (Sixteenth–Seventeenth Centuries),” History in Africa 27 (2000): 99–130. I wish to express my gratitude to the author, who kindly provided me a manuscript copy of this article. 9. Elsewhere, I have analyzed the material culture upon which “Portuguese” identity was based. See Peter Mark, “Constructing Identity: Sixteenth and Seventeenth Century Architecture in the Gambia-Geba Region and the Articulation of Luso-African Ethnicity,” History in Africa 22 (1995): 307–327; see also Mark, “‘Portuguese’ Architecture and Luso-African Identity in Senegambia and Guinea, 1730–1890,” History in Africa 23 (1996): 179–196. 10. In contemporary Mande society, one’s profession is presumed to reflect one’s ancestry. In the case of those belonging to the nyamakalaw groups, individuals or groups may, however, change profession, a change sometimes recognized by ascription of a new social identity to the individuals in question. See R. Launay, “The Dieli of Korhogo,” in Status and Identity in West Africa, edited by David Conrad and Barbara Frank (Bloomington: Indiana University Press, 1995), 153–169. 11. In 1694, the Portuguese priest Portuense observed that Bacampolo Co, king of Bissau, was able to speak Creole: “Entende muito bem a lingua portuguesa e podera falar o crioulo se quisera; porem, entre todos aqueles reis gentios estea introduzido por gravidade o fatarem por inteprete ou chalona.” [He understands the Portuguese language very well and could speak Creole if he wanted to; however, among all those heathen kings he might introduce an interpreter for the sake of gravity.] A. Teixiera da Mota, ed., As viagens do Bispo D. Frei Vitoriano Portuense a Guiné e a Cristianizacao dos Reis de Bissau (Lisbon: Junta de Investigações Científicas do Ultramar, 1974), 70. Mota adds that only Lemos Coelho gives an earlier clear reference to the existence of Crioulo. 12. Andrade, “Relação de Francisco de Andrade sobre as ilhas de Cabo Verde,” 102. 13. P. E. H. Hair, trans., “An Interim and Makeshift Edition of André Alvares de Almada’s Brief Treatise on the Rivers of Guinea . . . ,” translation, introduction, and notes on Chapters 13–19 by P. E. H. Hair, organized by A. Teixeira da

Notes to Pages 15–16

151

Mota (University of Liverpool, Department of History: Privately issued by Prof. Hair, 1984), Chapter 9, page 87. Mota edited Almada’s Brief Treatise but died before completing the project, whereupon Hair wrote the introductory notes. 14. Lopo Soares de Albergaria, “Relação de Lopo Soares de Albergaria sobre a Guiné do Cabo Verde [ca. 1600],” in Monumenta Missionaria Africana, edited by António Brásio. Series 2, Volume 4: Africa Ocidental, 1600–1622 (Lisboa: Agência Geral do Ultramar, Divisão de Publicacões e Biblioteca, 1968), 3–5. Albergaria mentions that Banhus, Casangas, and Buramos all spoke Portuguese. 15. See P. E. H. Hair, “Sources on Early Sierra Leone (5): Barreira,” Africana Research Bulletin 5, no. 4 (July 1975): 90. 16. Ibid., 97. 17. Ibid. Hair observes that many African ruling groups in Sierra Leone, including both men and women, “had been in extensive contact with the Portuguese in childhood and had acquired some European customs and a knowledge of the Portuguese language.” 18. Quoted in de Moraes, A la découverte de la Petite Côte au XVIIe siècle, 1: 55. 19. Anonymous, cited in ibid., 416. Marabouts are Muslim holy men and religious leaders. 20. Ibid., 361. Juffure was close to the Luso-African trading community of San Domingos (not to be confused with S. Domingos). 21. “C’est au cours du XVIIe siècle que le portugais parlé sur la côte de Guinée évolua vers le créole.” Boulègue, Les Luso-Africains de Sénégambie, 51. 22. For an excellent study of the interethnic and even international following attracted to a Manjak shrine, see Eve Crowley, “Contracts with the Spirits: Religion, Asylum and Ethnic Identity in the Cacheu Region of Guinea-Bissau” (Ph.D. diss., Yale University, 1990). 23. Almada describes a Jew named Ferreira who traveled from The Gambia to the Grand Fulo, one of whose daughters he married; Hair, trans., “An Interim and Makeshift Edition of André Alvares de Almada’s Brief Treatise on the Rivers of Guinea,” Chapters 2, 23. 24. See de Moraes, A la découverte de la Petite Côte au XVIIe siècle, 1: 32. 25. The records of the Jesuit Mission to Cape Verde (1604–1617) are an important source for primary sources documenting the presence of Jews in Guinea during the seventeenth century. See Barreira, Letter of February 20, 1606, in Monumenta Missionaria Africana, edited by António Brásio. Series 2, Volume 4: Africa Ocidental, 1600–1622 (Lisboa: Agência Geral do Ultramar, Divisão de Publicacões e Biblioteca, 1968), 97–113 and 159–174. See also Hair, “Sources on Early Sierra Leone (7): Barreira, Letter of 9.3.1607,” Africana Research Bulletin VI, no. 2 (January 1976): 45–70; and Hair, “Sources on Early Sierra Leone (17): Alvares at Mitombo, 1611,” Africana Research Bulletin XI, no. 1–2 (March 1981): 92–140. In 1497, the Portuguese Crown decreed the forcible baptism of all Jews. After the establishment of the Inquisition in 1536, “New Christians” faced increasing oppression. But by 1600, Portugal’s mounting financial difficulties led the Double Kingdom (Spain/Portugal) to grant converted Jews the right to leave the kingdom in return for a cash payment. In 1610, Philip II (Philip III of Spain) revoked this agreement. However, after 1609, the beginning of the Twelve Years’ Truce opened Dutch commerce with Portugal and led to increased migration of Portuguese Jews and “New Christians” to Amsterdam. On the origins of the Jewish community in Amsterdam see Miriam Bodian, Hebrews of the Portuguese Nation: Conversos and Community in Early Modern Amsterdam (Bloomington: Indiana University Press, 1997). On the Portuguese crown’s treatment of its Jews, see de Moraes, A la découverte de la Petite Côte au XVIIe siècle, 1: 25–33.

152

Notes to Pages 16–19

26. Letter of Balthasar Barreira to Father J. Alvares, in de Moraes, A la découverte de la Petite Côte au XVIIe siècle, 1: 116. 27. In 1608, P. van den Broecke visited in Portodale (Portudale) with one Simon Rodrigos, who had lived in Amsterdam. That Rodrigos is characterized as “an excellent trader” indicates that this was not simply a social visit; de Moraes, A la découverte de la Petite Côte au XVIIe siècle, 1: 138. Some of these Portuguese-Dutch commercial connections were undoubtedly based on family ties. 28. On the efforts of one governor of Cape Verde to extirpate Jewish merchants from the Petite Côte, see de Moraes, A la découverte de la Petite Côte au XVIIe siècle, 2: 305. For the original document, see Francesco de Moura to the King, “Lembranças de D. Francisco de Moura para sua Majestade ver [1622],” in Monumenta Missionaria Africana, edited by António Brásio. Series 2, Volume 4: Africa Ocidental, 1600–1622 (Lisboa: Agência Geral do Ultramar, Divisão de Publicacões e Biblioteca, 1968), 698. 29. Chinas were local shrines. The term may be etymologically related to the word “bekin,” which in the Jola language refers to a spirit force or to the shrine that such forces inhabit. 30. Hair, “Sources on Early Sierra Leone (7): Barreira Letter of 9.3.1607,” 63. 31. Religious rituals showed significant local variation. Nevertheless, there is also a long tradition in southern Senegambia of cultural interaction and sharing (see below, chapter 4). On this point see Crowley, “Contracts with the Spirits.” For the historian, the dynamic nature of culture makes it hard to attribute rituals to specific groups. 32. See the 1606 report of Father Balthasar Barreira of the Jesuit mission, “Das coisas do Cabo Verde e Costa da Guiné,” in Monumenta Missionaria Africana, edited by António Brásio. Series 2, Volume 4: Africa Ocidental, 1600–1622 (Lisboa: Agência Geral do Ultramar, Divisão de Publicacões e Biblioteca, 1968), 203: “Quando vêem nossas imagens de Christo ou de Nossa Senhora lhe chamam China do branco ou China do christao.” 33. Hair, “Sources on Early Sierra Leone (5): Barreira,” 88. 34. Jajolet de la Courbe reports that the people of Bolole were ydolâtres. But the names do suggest religious syncretism, or at least nominal acceptance of Christianity. Premier voyage du Sieur de la Courbe faite à la coste de l’Afrique en 1685, edited by Prosper Cultru (Paris: E. Champion, 1913), 260. 35. See below for the case of Ventura, son of a Sierra Leonean king. 36. On the nexus between Luso-African architecture and “Portuguese” identity, see Chapter 2. 37. “Alpendre” is the Portuguese term for porch or veranda (on the ground floor). 38. See Yves Goasguen Leven, “Architecture coloniale à l’ile de la Réunion” (Thèse de doctorat, Université Lyon II, 1997), 23. Leven cites Père Houbert, circa 1720. 39. “Elle nous receut [sic] fort civilement, dans une case à la portugaise, c’est à dire ayant des murailles de terre blanchie et un petit vestibule devant la porte où l’on nous fit asseoir à l’air sur des nattes.” La Courbe, Premier voyage du Sieur de la Courbe, 196. 40. Andrade, “Relação de Francisco de Andrade,” 102. 41. Ibid., 105. 42. Donelha, Déscription de la Serra Leoa, 15ff. 43. See José da Silva Horta, “Evidence for a Luso-African Identity,” 114. 44. Donelha, Déscription de la Serra Leoa, 129. 45. “Sapes” identity is problematic. Almada, who is inconsistent in his use of the term, uses “Sapes” to refer to several different groups. He writes, “In this kingdom of the Sapes are the following nations of people: Bagas, Tagunchos [?],

Notes to Pages 19–21

153

Sapes . . . Temenes, Limbas . . . and all these understand each other”; Hair, trans., “An Interim and Makeshift Edition of André Alvares de Almada’s Brief Treatise on the Rivers of Guinea,” Chapter 14, page 17. Elsewhere, however (Chapter 13, page 6), Almada appears to distinguish between Bagas and Sapes when he writes that “the Baga blacks extend as far as Cape Verga where the Sapes begin.” 46. Ibid., Chapter 9, page 10. 47. Donelha, Déscription de la Serra Leoa, 147. 48. Ibid., 161. On Gambian bixirins, see also Hair, “An Interim and Makeshift Edition of André Alvares de Almada’s Brief Treatise on the Rivers of Guinea,” Chapter 5, page 46. 49. “Bixiiris” may be read as a transliteration for “bi-serinyi.” Hair suggests that the term derives from “serigne” and ultimately from the Arabic “mubecherin” [one who spreads the faith]; Hair, in André Donelha, An Account of Sierra Leone and the Rivers of Cape Verde (1625), Portuguese introduction and notes by A. Teixeira da Mota, English translation and notes by P. E. H. Hair (Lisbon: Junta de Investigações Científicas do Ultramar, 1977), 303. 50. Ibid., 161. 51. The author does not specify the origins of these vessels. Elsewhere, however, he is specific when vessels he mentions come from European ports. 52. Donelha, Déscription de la Serra Leoa, 147ff. 53. Trade in Gambian Manding society was based upon such individual relationships between hosts and guests. The guest would be expected to share some of his profits with the host. 54. Donelha, Déscription de la Serra Leoa, 147ff. Clearly, European traders were not the only ones who established artificially high prices when trading on the African coast. 55. In Manding, the term “satigi” or “saltigi” may be translated as maître de la route [chief trader]; it designates rulers (or chief traders) in both Manding and Fulbe communities. See Jean-Loup Amselle, Logiques métisses: Anthropologie de l’Identité en Afrique et ailleurs (Paris: Payot, 1990), 73; see also Hair’s note in Donelha, Déscription de la Serra Leoa, 301. 56. José Horta cites Gaspar Vaz as “an excellent example of a ‘grumete’” (“Evidence for a Luso-African Identity,” 113). While the term is apposite, I would argue that, from Vaz’s perspective, he was alternately a Muslim Manding or a Christian Cape Verdean (arguably “Portuguese”). In Landlords and Strangers, Brooks also discusses Gaspar Vaz, whom he calls “a Mandinka youth who had formerly been a slave on Sao Tiago” (250). 57. See Hair, “Sources for Early Sierra Leone (4): Ruiters (1623),” Africana Research Bulletin V, no. 3 (April 1975): 61–62 and 70n. For a more extensive discussion of Mateus Fernandes, see Hair, “Hamlet in an Afro-Portuguese Setting: New Perspectives on Sierra Leone in 1607,” History in Africa 5 (1978): 36; reprinted in Hair, Africa Encountered: European Contacts and Evidence, 1450– 1700 (Hampshire, Great Britain and Brookfield, Vt.: Variorum, 1997). Hair presents a detailed treatment of Sierra Leonean Africans “who had themselves come to accept a measure of European culture.” 58. Donelha, Déscription de la Serra Leoa, 151. 59. That Vaz’s position as a Christian in a predominantly Muslim society was politically delicate is suggested by Donelha’s description of him; ibid., 149. 60. Later, during the eighteenth century, the idea that members of the Manding elite living along the lower Gambia had assimilated elements of Luso-African architectural style was expressed in popularized European travel accounts. In his 1745 compendium of travel narratives, Thomas Astley argued from his reading of Labat that cazas, or Portuguese houses, had spread to the kingdom of Barra,

154

Notes to Pages 21–23

where “the king and his lords have of them.” Astley, A New General Collection of Voyages and Travels, 4 vols. (London: Printed for T. Astley, 1745), 2: 246. This of course does not prove the primacy of the Luso-African model. Influences may just as well have moved in the other direction. 61. On changing identity in precolonial Casamance, see Peter Mark, The Wild Bull and the Sacred Forest: Form, Meaning, and Change in Senegambian Initiation Masks (New York: Cambridge University Press, 1992); for the Lower Gambia, see Donald Wright, The World and a Very Small Place in Africa (Armonk, N.Y.: M. E. Sharpe, 1997). 62. Amselle, Logiques métisses, 22. 63. “Il serait préferable de postuler une situation pluriethnique originaire qui aurait donné naissance aux différentes entités ethniques telle qu’elles ont été figées par la pensée coloniale.” [It might be preferable to posit a primordial pluriethnic situation; this would have led in turn to the development of different ethnic entities much like those which were rigidified by colonial thought.] Ibid., 74. 64. Ibid., 80. 65. Mirjam de Bruijn and Han van Dijk, “Introduction: Changing Frontiers,” Mande Studies 1 (1999): 10. See also “Introduction: Dialectique des Constructions Identitaires,” in Peuls et Mandingues: dialectique des constructions identitaires, edited by Mirjam de Bruijn and Han van Dijk (Paris: Karthala, 1997). 66. Allen Howard, “Mande and Fulbe Interaction and Identity in Northwestern Sierra Leone, Late 18th through Early 19th Centuries,” Mande Studies 1 (1999): 13. 67. Ibid., 16. 68. Ibid., 27. 69. Martin Klein, “Ethnic Pluralism and Homogeneity in the Western Sudan: Saluum, Segu, Wasulu,” Mande Studies 1 (1999): 119. 70. Ibid., 112. 71. Ibid. 72. Eve Crowley, “Institutions, Identities and the Incorporation of Immigrants within Local Frontiers of the Upper Guinea Coast,” in Migrations anciennes et peuplement actuel des côtes guinéennes, edited by Gérald Gaillard (Paris: L’Harmattan, 2000), 128. 73. Walter Hawthorne, “The Interior Past of an Acephalous Society: Institutional Change among the Balanta of Guinea-Bissau, c. 1400–1950” (Ph.D. diss., Stanford University, 1998). 74. Joshua Forrest, “The Lineages of a Soft State in Africa: State and Civil Society in Guinea-Bissau,” unpublished book ms., 8 (in press, Ohio University Press). Forrest attributes this flexibility of identity formation in part to the heritage of the Kaabu (Gaabu) empire: “A great deal of mutual acculturation occurred. . . . Non-Mandinka among them more readily assumed a Mandinka identity. Still, the constructive, inter-ethnic ties that in part defined Gaabu’s social relations would represent an enduring tradition within the associated regions of the north-central areas of what are now Guinea-Bissau and the Senegambia” (24). 75. Fredrik Barth, “Ethnic Groups and Boundaries [1969],” in Process and Form in Social Life: Selected Essays of Fredrik Barth (reprint, London, Routledge & Kegan Paul, 1981), 198–227. 76. Ibid., 204. 77. Ira Berlin, Many Thousands Gone: The First Two Centuries of Slavery in North America (Cambridge, Mass.: Harvard University Press, 1998), 41 and Chapter 2. 78. Ibid., 27. 79. Ibid., 46.

Notes to Pages 24–26

155

80. On the life of Francisco de Lemos Coelho and the dating of his 1669 manuscript, which was expanded in about 1684, see P. E. H. Hair, “Introduction,” in Francisco de Lemos Coelho, Description of the Coast of Guinea (1684), translated and with an introduction by P. E. H. Hair (University of Liverpool, Department of History: Privately issued by Prof. Hair, 1985). 81. Richard Jobson, The Golden Trade, or a Discovery of the River Gambra (London: Nicholas Okes, 1623), 28–30. 82. La Courbe, Premier voyage du Sieur La Courbe, 193 (my translation). 83. Jacques Le Maire, Les Voyages du Sieur Le Maire aux Isles Canaries, Cap-Verd, Sénégal et Gambie (Paris: J. Callombat, 1695), 38 (my translation). Original text: “Une espèce de Portugais, gens qui se nomment ainsi parce qu’ils ont autrefois servis, et qu’ils descendent de ceux qui habitèrent les premiers cette côte, après l’avoir découverte. Des Négresses qu’ils épousèrent nâquirent ces Mulâtres, de qui [sic] viennent des plus noirs qu’eux.” 84. On European attitudes toward Africans before the development of the Atlantic slave trade, see Peter Mark, Africans in European Eyes: The Portrayal of Black Africans in Fourteenth and Fifteenth Century Europe (Syracuse: Maxwell School of Citizenship and Public Affairs, Syracuse University, 1974). For a contrasting view that focuses on fifteenth-century Portuguese conceptions of Africans, see José da Silva Horta, “A represantação do Africano no literatura de viagens do Senegal à Serra Leoa (1453–1508),” Mare Liberum, no. 2 (1991): 209–338. 85. Francis Moore, Travels into the Inland Parts of Africa (London: Printed by E. Cave for the author, 1738), 29. 86. Ibid. 87. Ibid. 88. Astley, A New General Collection of Voyages, 2: 245. 89. John Mathews, A Voyage to the River Sierra Leone (London: B. White, 1788), 13–14. 90. On the use of “brancos” or “negros” in the writings of Almada and Donelha, see José da Silva Horta, “Portuguese Accounts of the Guiné do Cabo Verde and Luso-African Identity (Sixteenth–Seventeenth Centuries)” (paper presented at the Mande Studies Association Triennial Meeting, Banjul, The Gambia, June 1998). For a revised version of this paper, see “Evidence for a Luso-African Identity.” I wish to express my gratitude to José da Silva Horta for sharing his profound knowledge of early Portuguese sources with me. My understanding of Luso-African self-identification and of the relationship of these concepts to earlier Portuguese models has been deeply influenced by our discussions. 91. Almada had traveled to Iberia, and he may have written the Tratado breve while in Lisbon. See A. Teixeira da Mota, introductory notes, revised by Hair, in Hair, trans., “An Interim and Makeshift Edition of André Alvares de Almada’s Brief Treatise on the Rivers of Guinea,” translation, introduction, and notes on Chapters 13–19 by P. E. H. Hair, organized by A. Teixeira da Mota (University of Liverpool, Department of History: Privately issued by Prof. Hair, 1984), Introduction, 2–3. 92. Ibid. 93. Jobson, The Golden Trade, 30. 94. “Noirs et mulâtres qu’on appele cependant blanc, parce que tout ce qui est libre prétend à ce titre.” Gaspard Mollien, Voyage à l’Intérieur de l’Afrique aux Sources du Sénégal et de la Gambie, 2 vols. (Paris: Imprimerie de Mme Veuve Courcier, 1820), 2: 218. See also Mark, “Constructing Identity,” 307–327, esp. 314. In Saint-Louis in the mid-nineteenth century, free Africans were also called “whites”; see Anne Raffenel, Nouveau Voyage dans le pays des Nègres (Paris:

156

Notes to Pages 26–29

Impr. de. N. Chaix et cie, 1856), 57. The reader should not be confused about Raffenel’s identity; he was a man. 95. In this latter sense, Mathews’s sarcastic comments about European ancestry may contain a grain of truth. 96. James Webb, Desert Frontier: Ecological and Economic Change along the Western Sahel (1600–1850) (Madison: University of Wisconsin Press, 1995). 97. Ibid., xxv. 98. George Brooks, “The Observance of All Souls’ Day in the Guinea-Bissau Region: A Christian Holy Day, an African Harvest Festival, or an African New Year’s Celebration?” (paper presented at Africa Symposium, Ohio State University, 1982). 99. As a merchant, Domingos was well known to several English visitors, including Clarkson and Macaulay. Their observations are cited by Alexander Kup in his extensive annotations of Afzelius’s manuscript; see Afzelius, Sierra Leone Journal 1795–1796, edited by Alexander Kup (Uppsala: Studia Ethnographica Upsaliensis, 1967), 89n. 100. Ibid. 101. Ibid. 102. Abbé P.-D. Boilat, Esquisses Sénégalaises (Paris: P. Bertrand, 1853; reprint, Paris: Karthala, 1984), 110. At the beginning of the nineteenth century, however, Geoffroy de Villeneuve remarked on the existence of a Creole language. He observed: “Joal . . . fut habité autrefois par des Portugais qui, par leurs alliances avec des négresses du pays, y ont laissés des descendans. Ce sont eux qui forment la population de Joal. . . . Ils se disent cependant encore Portugais; mais ils n’ont conservé de leurs pères qu’un jargon presqu’inintelligible et un reste du culte du christianisme fort alteré.” [Joal . . . was formerly inhabited by Portuguese who, through their alliances with the Negro women of the country, left descendents there. It is they who constitute the population of Joal. . . . Nevertheless they still call themselves Portuguese; but they have preserved nothing from their ancestors (literally: fathers) other than a practically unintelligible jargon and the remnants of a Christian cult that has been radically changed.] Villeneuve, L’Afrique: ou histoire, moeurs, usages et coutumes des africains: le Sénégal, 4 vols. (Paris: Nepveu, 1814), 1: 82. 103. A mission was reestablished at Joal in 1850. On the Portuguese community in Joal, see Martin Klein, Islam and Imperialism in Senegal, Sine-Saloum, 1847–1914 (Stanford: Stanford University for the Hoover Institute on War, Revolution, and Peace, 1968), 49. In 1847, only two men could make the sign of the cross, and all the Joal Christians wore gris-gris. 104. “Bien que noirs, comme les plus noirs Africains, ils ont la modeste prétention d’être blancs tout purs, et c’est leur faire une grande injure que de les prendre pour des nègres ou pour des Sérères. Ils veulent qu’on les appelle les blancs de Joal, les chrétiens de Joal, parce qu’ils descendent des Portugais en ligne directe. . . . Pour tout dire, être chrétien, c’est être blanc; être blanc, c’est être libre”; Boilat, Esquisses Sénégalaises, 110. It should be noted that the culturally assimilated Boilat was himself from Joal. 105. I refer, inter alia, to Amselle, Logiques métisses. See also Jean-Loup Amselle and Elikia M’Bokolo, Au coeur de l’ethnie: Ethnies, tribalisme et Etat en Afrique (Paris: La Découverte, 1985). 106. This excellent essay is, unfortunately, unpublished; see Amanda Sackur, “Religion and the Construction of Creole Identity in Precolonial Senegambia” (paper presented at the African Studies Association of the United Kingdom, September 1992), 2. See also (Karen) Amanda Sackur, “The Development of Creole Society and Culture in Saint-Louis and Gorée, 1719–1817” (Ph.D. thesis, The School of Oriental and African Studies, 1999).

Notes to Pages 29–34

157

107. Boilat’s description is preserved in manuscript form as Boilat, “Voyage à Joal, 1846,” SG Carton Bo-Bon, Département des Cartes et Plans, Bibliothèque Nationale (hereafter BN). For a later and somewhat different account, see Boilat’s Esquisses Sénégalaises, especially 20–117. 108. Boilat, “Voyage à Joal, 1846.” 109. Ibid. 110. “En effet, tous les habitants de Joal sans exception ne veulent jamais . . . permettre à leurs enfants de contracter d’alliance avec les enfants du village voison, parce que ces derniers sont fétichistes, soit enterré sur leur territoire.” Ibid. 111. “As for the Serers, the priest must never maintain relations with them, and he most assuredly must not offer them [religious] instruction” (my translation); Boilat, Esquisses Sénégalaises, 117. See also Sackur, “Religion and the Construction of Creole Identity,” 6. I wish to thank Amanda Sackur for calling these passages to my attention. In Les Luso-Africains en Sénégambie, Jean Boulègue makes the same point when he writes, “The weakening of their cultural identity did not prevent the Luso-Africans of Joal from vigorously asserting their uniqueness with regard to the neighboring Serer population. They based this distinctiveness on religion.” 112. “Bubenben” glosses as “medicine,” while “bufókub” means literally “buried thing.” 113. Hyacinthe Hecquard, Voyage sur la Côte et dans l’Intérieur de l’Afrique Occidentale (Paris: Bénard et compagnie, 1855), 110. Hecquard derives much of his information from Bertrand-Bocandé, who for twenty years was the French résident at Carabane. 114. By the early twentieth century, colonial administrators had classified the Floups and several other groups who spoke dialects (which were not always mutually comprehensible) of the same language under the category of “Diola” (Jola). This is a clear instance of the “creation” of ethnicity by French colonial administrators, at least one of whom, Dr. Maclaud, was also an ethnologist. 115. Juula merchants may have perceived that their “Portuguese” counterparts derived professional advantage from their religion. Christianity, as Sackur observes, provided Luso-African traders with similar advantages to those that Islam provided to the Juula. For instance, shared religious orientation may have provided a sense of common values and a consequent foundation for trust and cooperation, which was important for merchants living in diaspora. See Sackur, “Religion and the Construction of Creole Identity,” 1, 8, 13–14. 116. Mungo Park, Travels in the Interior Districts of Africa (London: W. Bulmer, 1799), 215ff. In view of the similar discourse on identity and color among the Maures and the “Portuguese,” it is worth noting that the salt trader cited by Park was himself a Maure. 117. The Futa Jalon is in the highlands of interior Guinea. In the nineteenth century, it took about two weeks to walk there from the coast. 118. Hecquard, Voyage sur la Côte, 283. 119. Ibid. 2. SIXTEENTH- AND SEVENTEENTH-CENTURY ARCHITECTURE IN THE GAMBIA-GEBA REGION AND THE ARTICULATION OF LUSO-AFRICAN ETHNICITY 1. Jola (or Diola) ethnic identity is largely a product of the colonial period. See Peter Mark, The Wild Bull and the Sacred Forest: Form, Meaning and Change in Senegambian Initiation Masks (New York: Cambridge University Press, 1992). Before the late nineteenth century, the component peoples lacked a sense of com-

158

Notes to Pages 34–36

mon identity. Nevertheless, cultural and linguistic similarities existed then, as they do today, and were partly the product of migration from Kasa north across the Casamance River into Buluf. This population movement and the gradual assimilation of Bagnun-speakers by Jola-speakers is reflected in late-seventeenth-century sources, particularly La Courbe. By 1849, the Jola-speakers had reduced the Bagnun-speaking population to a few villages in northern Fogny and along the Soungrougrou River. See Emmanuel Bertrand-Bocandé, “Notes sur la Guinée portugaise ou Sénégambie méridionale,” Bulletin de la Société de Géographie XII, no. 67 (1849): 68. 2. Carabane is an island near the mouth of the Casamance River. 3. Bertrand-Bocandé, “Notes sur la Guinée-Portugaise,” 327. Note that these are the same group described as traders by La Courbe in 1686: “Presque tous les habitants parlent très bien le créole portugais.” [Almost all the inhabitants speak excellent Creole.] Jajolet de la Courbe, Premier voyage du Sieur de la Courbe fait à la coste d’Afrique en 1685, edited by Prosper Cultru (Paris: E. Champion, 1913), 56. 4. Valentim Fernandes, Déscription de la Côte occidentale d’Afrique (Sénégal au Cap de Monte, Archipels), translated by R. Mauny, Th. Monod, and A. Teixera da Mota (Bissau: Centro de Estudos da Guiné Portuguesa, 1951), 59. 5. “Cercada da estacas da paos e maneyra de barbacaã e cinco destas cercas tem em volta e assy ha dentrar e no direyto e dentro de todo tem sua praça onde elle esta assentado e esta casa tem derrador de sy huum esteyro dagoa e he muy forte casa.” Ibid., 63. 6. Walter Hawthorne, “The Interior Past of an Acephalous Society: Institutional Change among the Balanta of Guinea-Bissau, c. 1400–c. 1950” (Ph.D. diss., Stanford University, 1998), 137. Until the early sixteenth century, probably only a few hundred slaves were taken annually by the Portuguese from the entire Guinea Coast. Many early slaves went to the Cape Verde Islands. 7. Fernandes, Déscription de la Côte occidentale d’Afrique, 57. 8. More accurate late-nineteenth-century estimates placed the population of the largest villages in the Casamance at between 2,000 and 3,000; see 96 ter, Sénégal et Dépendences I, Archives Nationales du Sénégal (hereafter ANS). The use of figures such as 10,000 to symbolize a very large sum is common in the Mande world. In 1500, The Gambia formed the western extremity of this cultural area. Some of Fernandes’s informants likely had Mande informants themselves. Thus, 10,000 inhabitants should not be understood literally. See Mamadou Diawara, “Contribution to the Study of Social Differentiation in the Jaara Kingdom,” History in Africa 22 (1994): 132. 9. Fernandes, Déscription de la Côte occidentale d’Afrique, 37. 10. “Som fectas de adobes e de tijollo pera coser e dentro bem cayadas e defora co cree ou barro braco e de dentro muy bem lavradas e so as melhores cases de toda Guynee.” Ibid., 93. 11. The Tratado Breve was not published until 1733. However, later seventeenth-century chroniclers had either direct or indirect access to Almada’s account. Pierre Davity closely follows the narrative in his description of the Casamance region. Later authors such as Dapper rely on Davity (Adam Jones, personal communication). Dapper, in turn, was used by Barbot. Thus, as late as 1732, Almada’s writing helped to establish European images of the Guinea Coast; see Paul Hair, Adam Jones, and Robin Law, eds., Barbot on Guinea: The Writings of Jean Barbot on West Africa, 1678–1712, 2 vols. (London: Hakluyt Society, 1992). 12. “E Rio de grande trato de escravos,” Almada continues. “Ha algumas fortelazzas de guerra chamadas por eles Cao-sans, ao longo do Rio e esteiros, fortes de madeira muito forte, fincada toda a pique e terra-plenada, com suas guaritas.” [There are some fortresses which are called Cao-sans, situated along the river and estuaries, forts of very strong wood, surrounded entirely with stakes

Notes to Pages 36–37

159

and tamped earth.] André Alvares d’Almada, Tratado Breve dos Rios de Guiné do Capo Verde [1594],” in Monumenta Missionaria Africana, edited by António Brásio. Series 2, Volume 3: Africa Ocidental, 1570–1600 (Lisbon: Agência Geral do Ultramar, Divisão de Publicacões e Biblioteca, 1964), 272. The entire text of the Tratado Breve is on pages 229–378. 13. See ibid., 274: “The slaves which they obtain and sell, they acquire in wars.” [Os escravos que hao e vendem cativam em guerras.] 14. “A outra não fugia nem se defendia; o uso disto os fez ja terem melhor conhecimento, porque pelejam e se defendem e matam e cativam aos imigos.” [The other (enemy) does not flee nor defend himself; this fact means they already will have better knowledge, because they do fight and defend themselves and kill and enslave their enemies.] Ibid., 288. On the identification of the “Arriatas,” see Peter Mark, A Cultural, Economic and Religious History of the Basse Casamance Since 1500 (Stuttgart: Steiner Verlag/Frobenius-Institut, 1985), 20ff. 15. “A terra de Iziguchor que são Banhus qual ha trata de cero e escravos” [The land of Iziguchor who are Banhus (Bagnuns/Bainunks), which carries out trade of wax and slaves]. Almada, Tratado Breve, 289. On slave trading among the ancestors of the Diola-Esulalu (southern Floups), see Robert Baum, Shrines of the Slave Trade: Diola Religion and Society in Precolonial Senegambia (Oxford: Oxford University Press, 1999). 16. For a description of the Portuguese and Luso-African trading networks, see George Brooks, Landlords and Strangers: Ecology, Society and Trade in Western Africa, 1000–1630 (Boulder, Colo.: Westview Press, 1993), esp. 79–113 and 260ff. Lançados renamed the lower part of the Cacheu River the Rio de San Domingos in the sixteenth century; they called the upper portion of the river the Rio de Farim. “Portuguese” focused their mercantile efforts on the Rio de San Domingos. See ibid., 229. See Almada, Tratado Breve, 304: “O porte de Cacheu . . . por causa do muito trato que havia nesta terra de escravos, mantimentos, muita cera” [The port of Cacheu . . . because of the extensive trade that this region had in slaves, food, and lots of wax]. 17. “(As casas) são muito boas, e são mais labirinthos que casas. E fastem-nos desta maniera por causa de um nação de negros chamados Bijagos . . . os quais têm continuademente guerra con estes.” [(The houses) are very good, and they are more labyrinths than houses. They are secured in this manner because of a nation of Blacks called Bijagos . . . who are constantly at war with them.] Almada, Tratado Breve, 307. 18. “Os Bijogozos, os quais têm destruido doze Reinos, que ora estão despovoados. . . .” [The Bijagos who have destroyed twelve kingdoms, which are now depopulated. . . . ] Sebastião Frenandes Cação, “Letter to King Philip II, April 20, 1607,” in Monumenta Missionaria Africana, edited by António Brásio. Series 2, Volume 4: Africa Ocidental, 1600–1622 (Lisboa: Agência Geral do Ultramar, Divisão de Publicacões e Biblioteca, 1968), 248. See also “Letter of King Bamalá of Guinala to King Philip II, May 1, 1607,” in ibid., 255. See also Brooks, Landlords and Strangers, 263. 19. “As casas da dita povoacão são da taipa como as de Casamanca, delas sobradados, cobertas de olas, com grandes cercas de pau fincados a pique, feito um muro de palha a que chamam tapadas, e por dentro destas cercas vão as casas por dentro, segundo as posses dos moradores.” Almada, Tratado Breve, 299. See also the depiction of Farim, which was surrounded by a palisade, in the late-seventeenth-century map attributed by Boulègue to La Courbe. Jean Boulège, “Contribution des Sources françaises à la connaissance de l’actuelle Guinée-Bissau à la fin du XVIIème siècle,” History in Africa 28 (2001): 48 (see Figure 3). 20. “Estes Beafaras não têm as suas casas aldeadas, como as outros naçoes, senão afastadas algum tanto umas das outras, e as fazem segundo a posse de cada

160

Notes to Pages 37–42

um. E no lugar donde as fazem vivem ali os parentes todos juntos, reconhecendo ao mais velho, a quem dão obediencia.” [These Beafadas do not have their houses in villages like the other nations but separate each as well as the other, and they do this in accord with every individual’s capability. And in the place where they construct them, all the relatives live together, recognizing and offering obedience to the eldest.] Almada, Tratado Breve, 332. 21. Climate change on a regional scale is a central theme of George Brooks’s Landlords and Strangers. 22. P. E. H. Hair, trans. “An Interim and Makeshift Edition of André Alvares de Almada’s Brief Treatise on the Rivers of Guinea. . . . ,” translated and introduction by P. E. H. Hair, organized by A. Teixeira da Mota (University of Liverpool, Department of History: Privately issued by Prof. Hair, 1984), Chapter 5, page 44; André Donelha, Déscription de la Serra Leoa et des Rios de Guiné de Cabo Verde (1625), edited by A. Teixeira da Mota, translated by Léon Bourdon (Lisbon: Junta de Investigações Científicas do Ultramar, 1977), 141. 23. Richard Jobson, The Golden Trade, or a Discovery of the River Gambra (London: Nicholas Okes, 1620), 56. 24. “On passa au travers de quelques villages de Floupes qui étaient environnez [sic] de cinq ou six rangs de palisades.” Jean-Baptiste Labat, Nouvelle Relation de l’Afrique Occidentale, 5 vols. (Paris: Cavelier, 1728), 5: 33. 25. Ibid., 14. 26. “Qui n’osent pas les attaquer dans ces reduits, où dix hommes sont capables de tenir tête à cent.” Ibid., 32. 27. Francis Moore, Travels into the Inland Parts of Africa (London; Printed by E. Cave for the author, 1738), 36. 28. La Courbe, Premier Voyage du Sieur de la Courbe, 238. 29. Jean Boulègue’s “Contribution des sources françaises” discusses the accuracy of these representations and their precise dating. I wish to express my deepest appreciation to Professor Boulègue both for sharing this piece and for making available to me the photographic copies of the two maps, which are preserved in the library of the Centre de Recherches Africaines, Université Paris I. 30. S. H. M., portefeuille 111-2, piece 5/1, Département de Cartes et Plans, BN, fig. 3, 4. 31. Rés. Ge B 1964, Département Cartes et Plans, BN. This map, in two parts, has been published by Teixeira da Mota, As Viagens do Bispo D. Frei Vitoriano Portuense à Guiné e a cristianização dos reis de Bissau (Lisbon: Junta de Investigações Científicas do Ultramar, 1974), Figure 8. Hawthorne comments on the timber fortifications in “The Interior Past of an Acephalous Society,” 183– 184. He accurately observes that not only Floup communities but also Balanta villages are depicted as being surrounded by palisades. 32. Hawthorne, “The Interior Past of an Acephalous Society,” 183. 33. P. E. H. Hair, “Introduction,” in Francisco de Lemos Coelho, Description of the Coast of Guinea (1684), translated and with an introduction by P. E. H. Hair (University of Liverpool, Department of History: Privately issued by Prof. Hair, 1985). 34. Ibid., 11. The name probably derives from Soungrougrou, the north-south waterway followed by seventeenth-century merchants. 35. Ibid., Chapter 3, paragraph 7. 36. On Cacheu, see ibid., 8–18. Many of the free blacks worked as grumetes (or grumettes or gourmettes); see also Brooks, Landlords and Strangers, 243. 37. For a statistical analysis of the ethnic origins of slaves from Cacheu, see Stephan Bühnen, “Ethnic Origins of Peruvian Slaves (1584–1650)” Paideuma: Mitteilungen zur Kulturkunde 39 (1993): 57–110.

Notes to Pages 42–44

161

38. For documentation of the founding of Cacheu and the establishment of its fortifications, see royal decree of November 15, 1605, in Monumenta Missionaria Africana, edited by António Brásio. Series 2, Volume 3: Africa Ocidental, 1570– 1600 (Lisboa: Agência Geral do Ultramar, Divisão de Publicacões e Biblioteca, 1964), 88–89; and royal ordinance dated April 4, 1615, in Monumenta Missionaria Africana, edited by António Brásio. Series 2, Volume 4: Africa Ocidental, 1600–1622 (Lisboa: Agência Geral do Ultramar, Divisão de Publicacões e Biblioteca, 1968), 573. See also G. Thilmans and Nize Isabel de Moraes, “Le routier de la côte de Guinée de Francisco Pirez de Carvalho (1635),” Bulletin de l’Institut Fondamental d’Afrique Noire 32, no. 2 (1970): 343–369. 39. Brooks, Landlords and Strangers, 243; see also royal ordinance dated April 4, 1615, in Monumenta Missionaria Africana, 4: 573. 40. Lemos Coelho, Description of the Coast of Guinea, 20. The original fort, built in the 1580s and replaced shortly after 1610, was again rebuilt in the 1660s; see Hair, Jones, and Law, eds., Barbot on Guinea, 1: 166. By the nineteenth century, the fort was reduced to “rotten and indefensible wooden palissades and the bastions [to] no more than mounds of dirt”; George Brooks, “A Nhara of the Guinea-Bissau Region: Mae Aurelia Correia,” in Women and Slavery in Africa, edited by Claire Robertson and Martin Klein (Madison: University of Wisconsin Press, 1983), 305. 41. Lemos Coelho, Description of the Coast of Guinea, 24. The ready availability of seashells to make lime certainly facilitated the whitewashing of buildings in Portuguese trading towns. Much later, in the mid-nineteenth century, at least one French escale developed a local industry that produced lime. Carabane, on the Casamance River, exported locally produced lime to the island of Gorée and to The Gambia; see Hyacinthe Hecquard, Voyage sur la côte et dans l’intérieur de l’Afrique Occidentale (Paris Impr. de Bénard et compagnie, 1855), 109. 42. For an analogous example of the symbolic association of stone construction with permanence and social standing, see John Middleton’s discussion of Swahili architecture in The World of the Swahili (New Haven, Conn.: Yale University Press, 1993), 5, 62ff. 43. Lemos Coelho, Description of the Coast of Guinea, 21. 44. Manuel Teixeira, “Portuguese Traditional Settlements: A Result of Cultural Miscegenation,” Traditional Dwellings and Settlements Review 1, no. 2 (Spring 1990): 29. 45. In Yemen, liming of the exterior walls of mud buildings protects monumental ten-story dwellings against erosion by rainfall. Some of these structures are at least 350 years old. I wish to thank Caterina Borelli for explaining the application procedure and the functions of lime in Yemeni architecture. See Pamela Jerome, Giacomo Chiari, and Caterina Borelli, “The Architecture of Mud: Construction and Repair Technology in the Hadhramaut Region of Yemen,” APT Bulletin: The Journal of Preservation Technology XXX, no. 2–3 (1999): 39–48. Lime may extend the life of adobe walls even under conditions of high humidity such as are common in coastal Guinea (Jan Vansina, personal communication, September 13, 1994). 46. Villeneuve mentions Gambian houses that had been whitewashed with a solution made by combining powder made from burned bones and gum water; see Geoffroy de Villeneuve, L’Afrique, ou histoire, moeurs, usages et coutumes des Africains: le Sénégal, 4 vols. (Paris: Nepveu, 1814), 4: 139. 47. On Barbot’s sources, see Hair, Jones, and Law, eds., Barbot on Guinea, 1: introduction. See also 1: 160. 48. Boulègue observes that the missing portion of La Courbe’s manuscript is undoubtedly the source of Labat’s description. The beginning and end of the

162

Notes to Pages 44–45

overland voyage from Vintang to Cacheu appear in both La Courbe and Labat; Labat clearly plagiarized. One may assume that the plagiarism extends to the lost portion of La Courbe’s manuscript and that, consequently, the missing part is preserved in Labat. Boulègue, “Contribution des sources françaises,” 43–51. 49. “Elle nous receut fort civilement, dans une case à la portugaise, c’est à dire ayant des murailles de terre blanchies, et un petit vestibule devant la porte où l’on nous fit asseoir à l’air sur des nattes.” Ibid., 196. 50. “Elle est presque sur le bord de la mer, batie de terre comme les cases des Portugais et couverte de paille . . . et au devant il y a comme un vestibule, une galerie, qui tourne à l’entour.” La Courbe, Premier Voyage du Sieur de la Courbe, 212. 51. “Les maisons n’y sont que de terre battue, blanchies dehors et dedans avec de la chaux.” Labat, Nouvelle relation de l’Afrique occidentale, 5: 68. 52. La Courbe, Premier Voyage du Sieur de la Courbe, 198. The village of Juffure moved many times in its history, as Donald Wright observes in The World and a Very Small Place in Africa. One of its eighteenth-century incarnations was the trading village from which Alex Haley’s putative ancestor, Kunte Kinte, may or may not have begun his voyage to New World captivity. 53. La Courbe, Premier Voyage du Sieur de la Courbe, 203–204. This king was himself a Bagnun who often launched slave raids against the neighboring Floups; see 207. 54. Labat, Nouvelle Relation de l’Afrique Occidentale, 5: 5. 55. La Courbe remarked on the existence of several “Portuguese”-style houses in the trading village of Guinala south of Bissau on the Rio Grande, too. Premier Voyage du Sieur de la Courbe, 228. 56. On the importance of rapport between traders and their hosts, see Mark, A Cultural, Economic and Religious History of the Basse Casamance Since 1500, 61ff; see also Brooks, Landlords and Strangers. 57. In the nineteenth century, Diola men who traveled to The Gambia to gather and sell palm produce owed a percentage of the proceeds to their Manding hosts; see Mark, A Cultural, Economic and Religious History of the Basse Casamance Since 1500, 98. The Diola-Fogny language has a word for such hosts: ajoeti. 58. Brooks, “A Nhara of the Guinea-Bissau Region,” 296. 59. “On voyait au milieu un grand nombre de cases de Nègres renfermées dans une double enceinte de gros pieux de 10 à 12 pieds de hauteur, 7 ou 8 maisons à la portugaise . . . qui étaient encore enfermées dans une seconde enceinte de palissades, avec une très petite porte étroite et basse, qui semblait plutôt être le guichet d’une prison.” Labat, Nouvelle Relation de l’Afrique Occidentale, 5: 12. 60. “Il y a un numbre de maisons bâties à la portugaise. . . . Les Portugais sont en assez grand nombre dans ce lieu; il y paraissent riches, leurs maisons quoique couvertes de feuilles de latanier, sont belles, grandes et bien meublées pour le payis [sic]: Ils ont une Eglise plus grande et plus propre que celle de Gilfroy.” Ibid., 5: 5. Compare to La Courbe, “Bentam semblait être quelque bourg considérable de loing; il est situé sur le penchant d’une colline remplie d’arbres et plusieurs maisons portugaises, qui surpassent celles des nègres parraissaient comme des églises; mais, nous étant approchés de plus près, nous reconnûmes que tout cela n’était fait que de paille, hormis celles des Portugais, qui étaient de terre et couvert de feuilles de lataniers.” [From a distance Bentam [Vintang] appeared to be a rather large town; it sits on the slope of a hill, covered with trees and several Portuguese houses, which surpass those of the Blacks, appear like churches; however, once we had drawn near, we recognized that this was all only built of straw, except for those of the Portuguese, which were constructed of

Notes to Pages 45–50

163

earth and covered with the leaves of latanier trees.] Premier Voyage du Sieur de la Courbe, 201. 61. Ibid., 203; also Labat, Nouvelle Relation de l’Afrique Occidentale, 5: 7. 62. La Courbe, Premier Voyage du Sieur de la Courbe, 206. 63. Labat, Nouvelle Relation de l’Afrique Occidentale, 5: 43 and 50. Elsewhere I identify James with the Kujaamatay, a region now inhabited largely by Jola-speakers and extending from the Soungrougrou west to Fogny (A Cultural, Economic and Religious History of the Basse Casamance Since 1500, 31). For an alternative identification of James, see Bühnen, “Place Names as an Historical Source: An Introduction with Examples from Southern Senegambia and Germany,” History in Africa 20 (1992): 76. 64. I do not mean to imply that the color of the walls was intended to symbolize the color, blanc, of the Luso-Africans. No evidence exists to suggest such an association. 65. The Lower Casamance is, roughly, that part of the Casamance located east of the Soungrougrou River, both north and south of the Casamance River. 66. See Mark, A Cultural, Economic and Religious History of the Basse Casamance Since 1500, Chapter 4. 67. La Courbe, Premier Voyage du Sieur de la Courbe, 39. 68. Lemos Coelho, Description of the Coast of Guinea, 21. 69. La Courbe, Premier Voyage du Sieur de la Courbe, 212; Mollien, Travels in the Interior of Africa, 320. 70. Lemos Coelho, Description of the Coast of Guinea, 21. 71. Ibid. 72. See Mário G. Ventim Neves, “Técnica e características,” in A Habitacão Indigena na Guiné Portuguesa, edited by A. Teixeira da Mota and Mário G. Ventim Neves (Bissau: Centro de Estudos de Guiné Portuguesa, 1948), 148. 73. This argument in favor of the independent development of verandas in different cultures closely follows the reasoning articulated by Odile Goerg in Pouvoir Coloniale, Municipalités et Espaces Urbaines: Conakry-Freetown des années 1880–1914, 2 vols. (Paris: L’Harmattan, 1997), 2: 213ff: “La véranda, ou une de ses multiples variantes, puisait à des influences variées, accumulation d’expériences coloniales . . . ou non. Ce lieu couvert pouvait prendre plusieurs formes selon sa fonction et sa localisation: porche s’il ne couvrait que l’entrée de l’édifice, galerie de plain-pied pour la plupart des constructions administratives ou commerciales, véranda fermée en partie par un muret au niveau d’un rez-de-chaussée surélévé ou close par une balustrade à l’étage, balcon. . . . Cet élément . . . connut également des mutations successives.” [The veranda, or one of its many variants, derived from diverse influences, an accumulation of colonial experiences . . . or otherwise. This covered place could take several forms depending on its function and locale: a porch if it only covered the entryway, a gallery for most administrative or commercial structures, a veranda partially enclosed by a small wall on the level of an elevated ground floor or closed by a balustrade on the first floor, a balcony. . . . This element . . . also went through successive mutations.] See also Catherine Coquery-Vidrovitch, “Luso-Africains et Afro-Bréziliens du 16e au 19e siècle” (paper presented to the conference Le Portugal et L’Atlantique, Paris, May 18–19, 2000), 12. 74. See Teixeira da Mota and Neves, eds., A Habitacão Indigena na Guiné Portuguesa, 443. 75. “L’église paroissale . . . bâtie de terre comme les cases des Portugais et couverte de paille” [The parish church . . . built of earth like the Portuguese and covered with thatching]. La Courbe, Premier Voyage du Sieur de la Courbe, 212. 76. P. E. H. Hair, “Introduction,” in Francisco de Lemos Coelho, Description of the Coast of Guinea (1684), translated and with an introduction by P. E. H.

164

Notes to Pages 50 –52

Hair (University of Liverpool, Department of History: Privately issued by Prof. Hair, 1985), v and 21. 77. The discrepancy between these two contemporary eyewitness accounts is difficult to reconcile. Could they have been describing different churches? Another contemporary, Jean Barbot, also gives a physical description of Cacheu. He writes that the houses “are made of clapboards in the Portuguese style.” Barbot, however, freely admitted that he had never visited Cacheu; he clearly must have misunderstood his informants. See Hair, Jones, and Law, eds., Barbot on Guinea, vol. 1. 78. “Il y a à Jenné une grande mosquée en terre, dominée par deux tours massives et peu élevées.” René Caillié, Journal d’un voyage à Temboctou et à Jenné, dans l’Afrique Centrale, 3 vols. (1830; reprint: Paris: La Découverte, 1996), 2: 147. 79. On the history of the Jenné mosque, see J.-L. Bourgeois, Spectacular Vernacular (1983; reprint, New York: Aperture Books, 1999). See also Labelle Prussin, Hatumere: Islamic Design in West Africa (Berkeley: University of California Press, 1986). Bourgeois cogently argues that the nineteenth-century building was in fact the original building. This contradicts the idea, propagated by Pularspeaking invaders in the nineteenth century, that the early-nineteenth-century mosque was not located on the site of the original fourteenth-century structure. 80. “Les maisons sont construites en briques cuites au soleil. Le sable de l’isle de Jenné est mêlé d’un peu d’argile; ils l’emploient à faire des briques d’une forme ronde, mais assez solide. Les maisons sont aussi grandes que celles des villageois en Europe. La plupart ont un étage.” Caillié, Voyage à Tembouctou, 2: 147. 81. “J’eus l’occasion d’observer la manière de construire des maçons du pays . . . on fait des briques de forme ronde . . . ces briques sont semblables à celles de Jenné . . . Chaque maison forme un carré.” Ibid., 240. 82. Donelha, Déscription de la Serra Leoa et des Rios de Guiné du Cabo Verde. See Teixeira da Mota’s introduction, 15ff. 83. “Le village de Casao est petit, de forme ronde. Les maisons sont rondes, en briques crues, blanchies avec une argile blanche qui ressemble à de la chaux. Il en est quelques-unes comme celle du duc ou ‘sandegil’ qui ont un étage, et la plupart ont des banquettes à l’intérieur, en briques crues. . . . Tous ont des portes. Leurs serrures et leurs clefs de bois furent les premières que j’ai vues.” Donelha, Déscription de la Serra Leoa et des Rios de Guiné du Cabo Verde, 149 (in Bourdon’s French translation of the Portuguese). “A aldea é piquena, redonda, as casas redondas, de adobes caiados com um barro branco que parece cal. Há alguas sobradadas, como as do duque ou Sandeguil, as mais delas com poiais por dentro, de adobes, pera se sentarem, todas com portas, as fechaduras e chaves de pao foram as primeiras que vi” (in the original Portuguese). 84. “Cette maison (était celle du tribunal). . . . Cette maison était carrée, faite de briques crues, très blanches, avec une banquette tout autour . . . La maison avait deux portes, une en face à l’autre.” Donelha, Déscription de la Serra Leoa et des Rios de Guiné du Cabo Verde, 151 (in Bourdon’s French translation of the Portuguese). “Achei a casa quadrada, feita de adobes, mui alva, com um poial ao redor. . . . A cassa tinha duas portas” (in the original Portuguese). 85. Donelha’s original text is ambiguous on this point. Bourdon correctly retains that ambiguity in his French translation. Hair’s English translation implies that the benches were in the interior of the courthouse building, an interpretation I find dubious. 86. This early reference from a trading center frequented by the “Portuguese” lends circumstantial support to the thesis proposed by Jay Edwards that the origins of the veranda in New World Creole architecture can be traced—by way of

Notes to Pages 52–56

165

Brazil—to West Africa. See Edwards, “The Complex Origins of the American Domestic Piazza-Veranda-Gallery,” Material Culture 21, no. 2 (Summer 1989): 23ff. 87. “Nous sortîmes de cette maison . . . nous franchimes une autre porte, et nous pénétrâmes dans une cour entourée de maisons à étage. . . . La paroi . . . avait, en guise de tapisserie, une grande natte ouvragée.” Donelha, Déscription de la Serra Leoa, 153 (in Bourdon’s French translation of the Portuguese). “Saímos dessa casa pela porta dos vencidos, a u“a rua estreita, entrámos por outra porta, demos em um páteo cecado de casas sobradadas. Tinha a parede em que estava a caixa, por goademisim, u“a esteira grande lavrada” (in the original Portuguese). 88. Donelha, Déscription de la Serra Leoa et des Rios de Guiné du Cabo Verde, 161. 89. Moore, Travels into the Inland Parts of Africa, 57. It is also significant that the Englishman uses the Portuguese term for stake or pole, forkilha or forkilla; this, too, points to Lusitanian influence in local architecture. 90. Ibid., 177–178. This passage is also analyzed by Catherine CoqueryVidrovitch in “Villes africaines anciennes: une civilisation mercantile pré-négrière dans l’ouest africain, XVIe et XVIIe siècles,” Annales: Économies, Sociétés, Civilisations, no. 6 (novembre–décembre 1991): 1389–1410. The author points out that “alpainter” derives from the Portuguese term for veranda, “alpendre.” 91. “Because his house was built in the Portuguese style.” La Courbe, Premier Voyage du Sieur de la Courbe, 206. 92. Bühnen’s statistics for the Peruvian slave trade partially bear out this point. For the period 1548–1650, Floups constituted 8.6 percent of the 3,167 African slaves from Upper Guinea identified in Peru. By contrast, those identified as Bran made up 27.4 percent, Biafada 17.3 percent, Manding 9 percent, and Bagnun 10.7 percent of the total. See Stephan Bühnen, “Ethnic Origins,” 57–110. We do not know total population figures for the region for this period, however. 93. “Un des plus beaux que j’aye vu en Guinée et un des mieux peuplés.” La Courbe, Premier Voyage du Sieur de la Courbe, 259. 94. “Ils sont tous ydolastres et néantmoins à l’imitation des Portugais, ils se nomment presque tous d’un nom de saint.” [They are all idolaters and nevertheless, imitating the Portuguese, they nearly all have [Christian] names of saints.] Ibid., 260. It is possible that individual Floups actually had two names, one Portuguese (Christian) and the other Floup, just as many of their Jola descendents today have both a Jola name and a Christian or a Muslim name. It would also be interesting to know whether the Floups’ use of Portuguese names of saints may reflect more religious borrowing than is implied by La Courbe’s use of the term “idolâtres.” In other contexts, La Courbe uses the term to imply a mixture of local and Christian religious practices. 95. “La case du roi est une des plus jolies que j’aye vû dans ce pays; il y avait au devant une grande place avec une avenue de deux rangées d’arbres assez bien plantés. La case est comme un labyrinthe, toute plantée de bananiers comme celle du roy de Bissau avec quantité de cases d’espace en espace pour ses femmes et ses esclaves, et après plusieurs tours et detours vous arrivez vers le milieu de la sienne, devant laquelle il y avait un vestibule couvert où il escoute les palabres . . . nous le trouvâmes assis dans un fauteuil que les Portugais luy avaient donné.” Ibid., 238. 96. “Je n’ai pas tout vu; . . . Sur les parties qui me sont inconnues, j’ai consulté les habitans du pays et les auteurs. . . . Labat.” [I have not seen everything (that I describe). . . . For those regions unknown to me, I have consulted the inhabitants of the country and other writers. . . . (including) Labat.] Jean Baptiste Durand, Voyage au Sénégal (Paris: Henri Agasse, 1802), xxviii.

166

Notes to Pages 56–59

97. “Vintain, demeure ordinaire de l’empereur de Foigni, était autrefois plus considérable qu’il n’est à présent. Presque toutes les maisons sont bâties ‘à la portugaise.’ . . . La population des Portugais à Vintain est assez considérable.” Ibid., 86. 98. Ibid., 101. “La maison, bâtie à la portugaise, était grande, blanche dehors et dedans, avec un vestibule à l’entrée, ouverte de tous côtés; elle était environnée de grands palmiers, garnie de meubles commodes, de chaises et de selles de bois.” 99. “Ils ont appris des Portugais la manière de batir des maisons grandes et commodes.” Durand, Voyage au Sénégal, 101. 100. “Il y a plusieurs chrétiens noirs qui sont de ces isles, élevés parmi les Portugais par le moyen desquels on va negocier avec eux.” La Courbe, Premier Voyage du Sieur de la Courbe, 220. 101. Ibid., 232 and 228. 102. See José da Silva Horta, “Evidence for a Luso-African Identity in ‘Portuguese’ Accounts on ‘Guinea of Cape Verde’ (Sixteenth–Seventeenth Centuries),” History in Africa 27 (2000): 99–130. The author opines (correctly, in my judgment) that grumetes could and did choose to be seen as either Africans or Luso-Africans, depending on the context. See also Jean Boulègue, Les LusoAfricains, XVIe–XIXe siècle (Paris: Université de Paris I, Centre de Recherches Africaines, 1989), 15: “Ces serviteurs libres, plus ou moins lusitanisés, constituèrent aussi un élément de la population luso-africaine. Les serviteurs libres, qui avaient un rôle important dans le commerce, étaient appelés grumetes.” [These free servants, more or less lusitanized, also constituted a part of the Luso-African population. The free servants, who played an important role in trade, were called grumetes.] See also P. E. H. Hair, “Hamlet in an Afro-Portuguese Setting: New Perspectives on Sierra Leone in 1607,” in his Africa Encountered: European Contacts and Evidence 1450–1700 (Hampshire, Great Britain and Brookfield, Vt.: Variorum, 1997). 103. M. Adanson, A Voyage to Senegal, the Isle of Gorée, and the River Gambia (London: J. Nourse, 1759), 162. 104. Catherine Coquery-Vidrovitch, Histoire des Villes d’Afrique Noire: des Origines à la colonisation (Paris: Albin Michel, 1993), 151. 3. RECONSTRUCTING WEST AFRICAN ARCHITECTURAL HISTORY 1. Yves Goasgen Leven, “Architecture coloniale à l’île de Réunion” (dissertation, Université Lyon-II, 1997), 23, citing Père Houbert (1722) as quoted in A. Lougnon, Relation du R.P.Houbert (Madagascar: Imprimerie de l’Imerina, 1938). 2. In late-seventeenth-century Cacheu and early-nineteenth-century Bissau and Ziguinchor, some Luso-Africans lived in houses with roofs of palm fronds that were removed in the dry season to reduce the risk of fire. See J.-B. Labat, Nouvelle Relation de l’Afrique Occidentale, 5 vols. (Paris: Cavelier, 1728), 5: 65. See also G. Mollien, Travels in the Interior of Africa to the Sources of the Senegal and Gambia (London: H. Colburn & Co., 1820), 384; and Hyacinthe Hecquard, Voyage sur la Côte et dans l’Intérieur de l’Afrique Occidentale (Paris: Impr. de Bénard et compagnie, 1854). 3. Thomas Astley, A New General Collection of Voyages and Travels, 4 vols. (London: Printed for T. Astley, 1745), 2: 246. Astley adds that some “Portuguese”-style houses had walls of wattle and daub: “Their walls are 7 or 8 feet high made of reeds or hurdles, covered inside and out with stiff clay mixed with cut straw.”

Notes to Pages 60–64

167

4. Francis Moore, Travels into the Inland Parts of Africa (London: Printed by E. Cave for the author, 1738), 57. 5. Photograph from the archives formerly belonging to Professor LouisVincent Thomas; Album II, 23. I wish to express my profound gratitude to the heirs of Prof. Thomas, especially to his daughter, Domitienne, for kindly providing access to his entire archive of photographs. Some of these photographs were taken by Prof. Thomas’s friend and longtime acquaintance, Catholic missionary Henri Goovers. 6. See P. E. H. Hair, “An Accomplished Traveller Will Take Draughts: Barbot’s Illustrations of Guinea,” Hakluyt Society Annual Report for 1991 (London: Hakluyt Society, 1991), 14. 7. Adam Jones argues, along with Hair, that only Barbot, among African travelers, indisputably made on-the-spot drawings that served as the basis for later published illustrations. Although I have elsewhere argued that Froger’s 1697 illustration of a Gambian initiation masquerade is ethnographically accurate, I would not make the same claim for Froger’s depictions of architecture. 8. See too the building just to the left of the tower in the illustration of the English fort at Anomabu; illustrated in P. E. H. Hair, Adam Jones, and Robin Law, eds., Barbot on Guinea: The Writings of Jean Barbot on West Africa, 1678– 1712, 2 vols. (London: Hakluyt Society, 1992), 2: 415. 9. Ibid., 2: 435. 10. See, for example, Frank Willett, African Art (London: Thames & Hudson, 1993), 101–102. Willett describes Dapper in the context of “travellers’ illustrations and descriptions” without, apparently, being aware that Dapper never went to Benin. “The accuracy of the drawings,” Willett writes, “has been confirmed in a number of ways,” and he considers them to be a significant late-seventeenthcentury historical source. The assessment that they represent eyewitness views is unsubstantiated. 11. See Adam Jones, “Olfert Dapper et sa déscription de l’Afrique,” in Objets Interdits (Paris: Fondation Dapper, 1989). 12. Published in London in 1807 and based on Corry’s travels in 1805 and 1806. Each illustration is signed “R. Cocking delint. from a sketch by I. Corry . . . I. C. Stadler sculpt.” 13. Low roofs that are pitched on all four sides. Because the artist has not shaded the walls, the three houses stand out against the darker background. This artistic device may represent whitewashing. 14. Astley, A New General Collection of Voyages and Travels, 2: 246. 15. See George Brooks, Landlords and Strangers: Ecology, Society and Trade in Western Africa, 1000–1630 (Boulder, Colo.: Westview Press, 1993), 143. 16. José da Silva Horta, “Evidence for a Luso-African Identity in ‘Portuguese’ Accounts on ‘Guinea of Cape Verde’ (Sixteenth–Seventeenth Centuries),” History in Africa 27 (2000): 104. 17. This work is one of a collection of nine lithographs created in about 1864. Initially published in Francisco Travassos Valdez, Africa Ocidental (Lisbon: Imprensa Nacional, 1864), these illustrations are reprinted in Maria Emilia Madeira Santos, ed., História Geral de Cabo Verde (Lisbon: Centro de Estudos de História e Cartografia Antiga, Instituto de Investigação Científica Tropical; and Praia [Cape Verde]: Instituto Nacional da Cultura de Cabo Verde, 1995), vol. 2. 18. See the view of Monte de Galianca reproduced as Figure 17 in Madeira Santos, ed., História Geral de Cabo Verde, volume 2. 19. Ilidio do Amaral, Santiago de Cabo Verde, a Terra e os Homens (Lisbon: Junta de Investigações Científicas do Ultramar, 1964), Plate LXX: Formas de povoamento. A povoação de Calheta e a baia.

168

Notes to Pages 65–67

20. Hecquard’s drawings are preserved in the Bilder-Archiv of the FrobeniusInstitut in Frankfurt am Main. Several of these illustrations, which are based on direct observation, served as the basis for lithographs first published in the German edition of Hecquard’s travel account, Reise an die Küste und in das Inneren von West Afrika (1854). For a discussion of some of the thirty extant drawings, see P. Mark, “Hyacinthe Hecquard’s Drawings and Watercolors from Grand Bassam, the Futa Jalon and the Casamance: A Source for Mid-Nineteenth Century West African History,” Paideuma: Mitteilungen zur Kulturkunde 36 (1990): 173– 184. 21. Pierre Verger, Trade Relations between the Bight of Benin and Bahia from the 17th to the 19th Century, translated by Evelyn Crawford (Ibadan, Nigeria: Ibadan University Press, 1976), 2. Verger terms the sixteenth century “the Guinea cycle.” See also Boris Fausto, Historia do Brasil (São Paulo: Editora da Universidade de São Paulo, 1994), 51. 22. During the seventeenth century many, perhaps most, slaves came from Angola and Sao Tomé. For a recent study of the Cape Verdean slave trade, see Antonio Correia e Silva, “Cabo Verde e a Geopolitica do Atlantico,” in Madeira Santos, ed., Historia Geral de Cabo Verde, 2. In a comment made regarding the period 1648–1668, Verger writes: “The slave trade between the Mina Coast and Bahia was still very insignificant, the main traffic being with Cacheu and Cabo Verde in the North and mostly Gabon and Angola in the South” (Trade Relations between the Bight of Benin and Bahia, 46). 23. Johannes Postma, The Dutch in the Atlantic Slave Trade 1600–1815 (Cambridge: Cambridge University Press, 1990), 21ff. 24. Madeira Santos, ed., Historia Geral de Cabo Verde, 2: 26. Some of this trade left directly from San Domingos; ibid., 2: 35ff. Correia e Silva cites contemporary sources suggesting that in the 1630s, Portuguese ships transported almost their entire human cargo to “the Indies of Castille”; ibid., 2: 58. 25. Ibid., 2: 21. 26. See Postma, The Dutch in the Atlantic Slave Trade 1600–1815, Table 5.2: Slave Origins under the Old WIC, 1658–1674. This table represents figures for the Dutch West India Company. When one takes into account Portuguese, French, and English vessels, the percentage of slaves acquired from Senegambia during this period may have been higher. 27. Jay Edwards, “The Complex Origins of the American Piazza-VerandaGallery,” Material Culture 21, no. 2 (Summer 1989): 3–58. I am most grateful to Jay Edwards for freely sharing his research, both published and unpublished, with me. I wish to acknowledge my methodological debt to Edwards, who first suggested the use of Post’s paintings as a source for the historical reconstruction of New World Creole architecture. 28. Maurits is best known by art historians for the fact that he constructed the Mauritshuis as his personal residence. On Maurits as patron, see P. J. P. Whitehead and M. Boeseman, A Portrait of Dutch 17th Century Brazil: Animals, Plants and People by the Artists of Johan Maurits of Nassau (Amsterdam: North Holland Publishing Company, 1989). 29. Whitehead and Boeseman, A Portrait of Dutch 17th Century Brazil, 22. These drawings disappeared from a Berlin collection during World War II. In 1977 some of them “reappeared” in Cracow. 30. One of the scenes from his Brazilian period came on the art market in February 1997. It was sold at auction for over $4 million. Presumably, the purchaser was more impressed by the work’s rarity than by the accuracy of its ethnographic description. 31. Frans Post’s brother Pieter was an architect whom Maurits employed to design the Mauritshuis. While it is not known whether Frans had been trained as

Notes to Pages 68–71

169

an architect, his drawings and paintings betray familiarity with the technical aspects of construction. 32. See Uwe Fleckner, “Die Erfahrung der Fremde: Albert Eckhouts und Frans Posts Brasilienreise (1636–1644) und ihre Gestaltung in Porträt und Landschaft,” in Reisen des Barock: Selbst- und Fremdenerfahrungen und ihre Darstellung, edited by Regina Pleithner (Bonn: Romanistischer Verlag, 1991), 37: “Ethnographische oder topographische Beschreibung muss mit den besonderen ästhetischen Erfordernissen die eine Bildconvention an die Gestaltung des Gegenstandes richtet, im Kunstwerk vermittelt werden.” [Ethnographic or topographical description must be informed by the specific aesthetic demands that direct a painterly convention, in the formulation of the particular themes, in a work of art.] 33. Unfortunately, the drawings are unpublished. Fleckner suggests that the drawings were made in Brazil and that Post added finishing touches and dated them in 1645. Whitehead and Boeseman offer a different interpretation. They believe that only four of the drawings were made in Brazil (A Portrait of Dutch 17th Century Brazil, 181). On the dating of Post’s drawings, Fleckner writes, “Obschon erst 1645 datiert und signiert, dürfen die Landschaftsskizzen Posts als in Brazilien entstanden betrachtet werden: es ist wahrscheinlich, dass Post seine Zeichnungen zu Druckvorlagen umgearbeitet hat, als Caspar van Baerle [Baerlaeus] ein Jahr nach der Wiederkehr des Grafen Maurits begann, sein Geschichtswerk zu verfassen.” [While Post’s landscape sketches are only dated and signed 1645, they may nevertheless be considered as the product of his time in Brazil. Post probably reworked his drawings in preparation for engraving when Caspar Baerle began to assemble his historical work a year after Maurits’s return.] “Die Erfahrung der Fremde,” 31–32. Baerle’s work is the Rerum per octennium in Brasilia (1647), with copperplate engravings based on Post’s drawings. 34. Whitehead and Boeseman, A Portrait of Dutch 17th Century Brazil, 193. 35. On the Marcgraf Map, see ibid., 151ff. and plates 81 and 82. 36. By “vernacular architecture” I mean dwellings obviously constructed by individuals of limited means who would certainly not have had recourse to architects; churches and fortifications would be excluded from this category. 37. A third edition of the map, published in 1664 by Clement de Jonghe, incorporates Blaeu’s original copperplates as copied by Allart in 1657. There is much less clarity in Allart’s version than in Blaeu’s originals. 38. The specific appellations “Portuguese” style and maisons à la portugaise evolved in European travel literature from West Africa. I do not mean to imply that these labels were also used in seventeenth-century Brazil. 39. On sugar production and the growth of the slave trade, see L. McAlister, Spain and Portugal in the New World (Minneapolis: University of Minnesota Press, 1984), 276ff. See also A. Correia e Silva, “Cabo Verde e a Geopolitica do Atlantico,” in Madeira Santos, ed., Historia Geral de Cabo Verde, vol. 2. 40. The chief West African ports along that part of the coast were Guinala and, from its founding in the 1580s, Cacheu. 41. Whitehead and Boeseman, A Portrait of Dutch 17th Century Brazil, 159. 42. This title was used in the Frans Post exhibition at the Basel Kunsthalle in 1990; see Thomas Kellein and Urs-Beat Frei, eds., Frans Post, 1612–1680 (Basel: Kunsthalle Basel, 1990), 97. Since 1959, the painting has been in the collection of the Museu de Arte, São Paulo. However, the fact that it was in a Danish collection in 1773 (ibid.) suggests that this was another of the landscapes painted after Post returned to Europe. This interpretation is corroborated by the fact that the foreground represents an idealized and highly fanciful botanical composition dominated by palm trees and exotic fruits. The romanticized vision contrasts

170

Notes to Pages 72–73

dramatically with the detailed realism of the more distant village scene. On Post’s romanticized reworking of his sketches, see Fleckner, “Die Erfahrung der Fremde.” Most of Post’s landscapes have purely descriptive titles. Sometimes a given work has titles in different languages that do not correspond. See also Sousa-Leao’s Catalog raisonné: Frans Post, 1612–1680 (Amsterdam: A. L. van Gendt, 1973), no. 91. Sousa-Leao suggests that the work represents Santo Amaro, near Recife. 43. Michel Jajolet de la Courbe, Premier voyage du Sieur de la Courbe fait à la coste d’Afrique en 1685, edited by Prosper Cultru (Paris: E. Champion, 1913), 201. 44. Ibid., 238. 45. “GÌrÌssÌ” is in the Gussilay dialect, which is spoken in Thionk-Essyl. On the seventeenth-century history of this region, see Peter Mark, A Cultural, Religious and Economic History of the Basse Casamance Since 1500 (Stuttgart: Steiner Verlag/Frobenius-Institut, 1985). 46. The siting of houses was a critical component of Brazilian Creole and, later in North America, colonial architecture. In the antebellum southern states, plantation houses as far north as the James River in Virginia were built on promontories overlooking waterways, often with verandas open to the cooling river breezes. 47. See Edwards, “The Complex Origins of the American Piazza-VerandaGallery,”15. 48. See Odile Goerg, Pouvoir Colonial, Municipalités et Espaces Urbains: Conakry-Freetown des Années 1880 à 1914 (Paris: L’Harmattan, 1997) 2: 197ff. See also Goerg, “La Guinée Conakry,” in Rives Coloniales: Architectures, de Saint-Louis à Douala, edited by Jacques Soulillou (Paris: ORSTOM/Parenthèses, 1993), 82. Elevating the living quarters was a means of minimizing humidity in the house. See also Sylvie Kandé, Terres, Urbanisme et Architecture “Créoles” en Sierra Leone, XVIIIe–XIXe siècles (Paris: L’Harmattan, 1998), 205. Kandé discusses Maroon architecture in the Caribbean and its subsequent impact on Sierra Leone, where it was (re)introduced by Jamaican immigrants. Kandé suggests a connection between the elaboration of Creole architecture and the establishment of a distinctive Creole identity in nineteenth-century Sierra Leone (220ff.). 49. For excellent discussions of colonial architecture in all three West African colonial empires, see Soulillou, ed., Rives Coloniales. 50. Excellent discussions of the nineteenth-century Afro-Brazilian style are to be found in Soulillou, ed., Rives coloniales. See especially Thierry Lulle, “Le Togo,” 170–205; Luc Gnacadja, “Le Benin,” 209–242; and Alan VaughanRichards, “Le Nigeria,” 243–288. On Brazilian migration back to West Africa, see also Catherine Coquery-Vidrovitch, “Luso-Africains et Afro-Bréziliens du 16e au 19e siècle” (paper presented to the conference Le Portugal et L’Atlantique, Paris, May 18–19, 2000). 51. A personal anecdote may illustrate the futility of such an exercise. I recently visited Dijon, site of the fifteenth-century Burgundian court. In the courtyard of a late-fifteenth- or sixteenth-century hôtel was a stunning second-story wooden gallery. “Clearly the inspiration for the galleries on Brazilian plantation houses,” I joked. Later, I realized that in 1434 the Duke of Burgundy had married a member of the Portuguese royal family, thereby linking the Burgundian and Portuguese courts. Already at this early date, Burgundian artists, including painters from the Low Countries, then part of Burgundy, were traveling to Lisbon. In other words, cultural and artistic ties between Portugal and Burgundy and the (future) Netherlands were established two centuries before the Dutch took possession of Brazil. Architectural forms may well have been shared even before the era of Portuguese seaborne exploration of Africa began.

Notes to Pages 73–82

171

52. Edwards, “The Complex Origins of the American Piazza-VerandaGallery,” 45. 53. Sousa-Leao, Frans Post, 1612–1680, no. 94, offers the English title Landscape with Buildings and Ruins. The painting, acquired by Prague’s National Gallery in 1948, is reproduced in Kellein and Frei, eds., Frans Post, 1612–1680, 77. 54. The Ella Gallup Sumner and Mary Catlin Sumner Collection Fund, 1949.460, Wadsworth Atheneum, Hartford, Connecticut. 55. Reproduced in Kellein and Frei, eds., Frans Post, 45. Sousa-Leao, Frans Post, 1612–1680, no. 23, offers the title Landscape with Houses. According to the National Gallery, this work was returned to private ownership in 1968. Unfortunately, the Gallery is legally enjoined from divulging the identity of that owner. Therefore, it is not possible to provide an illustration in the present volume. 56. Lot 37, oil on panel, 10⫻16 inches, signed and dated, Sotheby’s, New York, sale of January 28, 2000; see catalogue Old Masters 2000 (New York: Sotheby’s, 2000). 57. Both the large farmhouse and the chapel are roofed with tile. 58. Appropriately enough, this building currently houses the local consulate of Guinea-Bissau. 59. The veranda was undoubtedly enclosed with walls at a later date, much like some of the Brazilian houses painted by Post. 60. See, for example, the Villa les Iris, illustrated in Goerg, “Guinée Conakry,” in Soulillou, ed., Rives Coloniales, Architectures de Saint-Louis à Douala, 98. Except for later additions, the two buildings are nearly identical. Both have a central raised roof, surrounded by a lower and separately pitched roof to cover the encircling veranda. The Conakry villa is roofed with metal sheets, while the Ziguinchor house has a tile roof. It is possible that the basic plan for this structure was widely available to French colonial officials at the beginning of the twentieth century. 4. “THE PEOPLE THERE ARE BEGINNING TO TAKE ON ENGLISH MANNERS” 1. Cultural interaction, intermarriage, and the absence of emphatic demarcation based on color were all characteristics that spread to Caribbean Creole society before the development of a slave-based plantation economy. See Ira Berlin, Many Thousands Gone: The First Two Centuries of Slavery in North America (Cambridge, Mass.: Harvard University Press, 1998), introduction, especially 27. 2. See Sieur Michel Jajolet de la Courbe, Premier voyage du Sieur de la Courbe fait à la coste d’Afrique en 1685, edited by Prosper Cultru (Paris: E. Champion, 1913). Cultru’s introduction is important, as it documents Labat’s appropriation without attribution of La Courbe’s description of the Casamance. 3. See Jean Boulègue, “Contribution des Sources françaises à la connaissance de l’actuelle Guinée-Bissau à la fin du XVIIe siècle,” History in Africa 28 (2001): 43–51. By comparing portions of Labat’s account which overlap the preserved portion of La Courbe’s description with the original passages by La Courbe, Boulègue convincingly shows that Labat, in his plagiarism, is faithful to the original. 4. Richard Fardon makes a similar point in his critique of the English translation of Amselle’s Logiques métisses; see Fardon, “Review Article: ‘Métissage’ or Curate’s Egg,” Africa 70, no. 1 (2000): 149. 5. Serge Gruzinski, La Pensée Métisse (Paris: Fayard, 1999). It should be pointed out that in his seminal 1990 study, Logiques métisses: Anthropologie de l’identité en Afrique et ailleurs (Paris: Payot, 1990), Amselle presents a nuanced

172

Notes to Pages 83–84

conceptual framework for analyzing the historical processes of culture and identity formation: “Les cultures . . . prennent place dans un ensemble mouvant qui est lui-même un champ structuré de relations.” [Cultures . . . are positioned in a moving ensemble that is itself a structured field of interrelationships] (55). In this conceptual model, he argues for “une ‘logique métisse’ c’est-à-dire une approche continuiste qui . . . mettrait l’accent sur l’indistinction ou le syncretisme originaire” [a “métisse logic,” which is to say a continuous approach that would emphasize indistinctness or a state of original syncretism]. 6. Gruzinski, La Pensée Métisse, 68, 94, 204. The precise nuances that distinguish these terms, roughly “melanges, conjoining, adaptation, associative and overlapping mechanisms, borrowings and appropriations,” are not rendered in English translation. My point is that the term métissage does not have a precise meaning, but rather a constellation of approximate meanings. 7. Ibid., 57. 8. See André Donelha, Déscription de la Serra Leoa et des Rios de Guiné da Cabo Verde (1625), edited by A. Teixeira da Mota, translated by Léon Bourdon, notes by P. E. H. Hair (Lisbon: Junta de Investigações Científicas do Ultramar, 1977); see also Donelha, An Account of Sierra Leone and the Rivers of Cape Verde (1625), Portuguese introduction and notes by A. Teixeira da Mota, English translation and notes by P. E. H. Hair (Lisbon: Junta de Investigações Científicas do Ultramar, 1977). La Courbe did not travel upriver beyond Geregia; thus, he never visited Casão. 9. On Gaspar Vaz and cultural assimilation at Casão, see above, Chapter 1. For an earlier version of this discussion, see Peter Mark, “The Evolution of ‘Portuguese’ Identity: Luso-Africans on the Upper Guinea Coast from the Sixteenth to the Early Nineteenth Century,” The Journal of African History 40 (June 1999): 180–182. 10. “Tout ce pays, le long de cette rivière, nous parut parfaitement beau; celui de nôtre droit était du royaume de Faugrit (Fogny) et paraissait fort boisé. . . . Bentam (Vintain) semblait être quelque bourg considérable de loin; il est situé sur le penchant d’une colline remplie d’arbres et plusieurs maisons portugaises, qui surpassent celles des nègres, paraissaient comme des églises; mais, nous étant approchés de plus près, nous reconnûmes que tout cela n’était fait que de paille, hormis celles des Portugais, qui étaient de terre et couvertes de feuilles de lataniers, grandes comme des napes [sic].” La Courbe, Premier Voyage du Sieur de la Courbe, 201–202. 11. Francis Moore, Travels into the Inland Parts of Africa (London: Printed by E. Cave for the author, 1738), 24. 12. Ibid., 52. 13. That many of the Manding in the Lower Gambia were traders is reflected in the description of trade at Albreda, one league from Vintang Creek, offered by the Abbé Demanet after his 1764 visit: “Tous les marchands Mandingues qui descendent et remontent fréquement cette rivière, tant par eau que par terre.” [All of the Manding merchants who frequently descend and ascend this river, by water and by land.] Abbé Demanet, Nouvelle histoire de l’Afrique française (Paris, 1767), 129. 14. Francisco de Lemos Coelho, Description of the Coast of Guinea (1684), translated and with an introduction by P. E. H. Hair (University of Liverpool, Department of History: Privately issued by Prof. Hair, 1984),Chapter 22, para. 9. 15. La Courbe, Premier Voyage du Sieur de la Courbe, 207. 16. On long-distance trade in slaves along the lower Gambia, see, inter alia, Demanet, Nouvelle histoire de l’Afrique française, 148, 157, and 161: “Janvier, qui est le temps le plus ordinaire du passage des marchands Mandingues qui amènent toutes les années au moins douze cents captifs à la traite, et une quantité

Notes to Pages 85–86

173

prodigieuse d’or et de morphil.” [January, which is the month when, most commonly, Manding merchants come by and who annually bring at least twelve hundred captives to trade and a prodigious amount of gold and ivory.] 17. La Courbe, Premier Voyage du Sieur de la Courbe, 207. 18. Ibid., 208. 19. Jacques Le Maire, Les voyages du Sieur Le Maire aux Iles Canaries, Cap Vert, Sénégal et Gambie (Paris: J. Callombat, 1695). 20. Lemos Coelho, Description of the Coast of Guinea (1684), Chapter 22, para. 9. It is, of course, possible that La Courbe did mention Manding traders and that the reference was deleted in Labat’s plagiarized account. 21. Moore, Travels into the Inland Parts of Africa, 74. 22. Buluf is that part of the Lower Casamance that lies north of the Casamance River and east and south of the Dioloulou marigot. It is a heavily populated rice-growing region containing twenty-two villages. 23. Interview with the elders of Kamanar ward, Thionk-Essyl, Senegal, February 12, 1975. On early Jola population movements, see Mark, A Cultural, Economic, and Religious History of the Basse Casamance Since 1500 (Stuttgart: Steiner Verlag/Frobenius-Institut, 1985), Chapter 2. 24. An initial migration from the Manding region of Badibu to the east was also recalled by the elders of Niaganan ward, Thionk-Essyl, Senegal, February 12, 1975. 25. Elders of Kamanar ward, Thionk-Essyl, Senegal, February 12, 1975. These informants localized the initial immigration from the village of Kaour in Pakao. From Pakao, the ancestors are said to have moved to Burofai near Ziguinchor and then on to Essyl in Kasa. Informants in Niaganan ward (Bassiru Djiba, Boubacar Diatta, Ousman Djiba, Bacary Diatta, Malan Djiba) asserted that the initial migration came from Badibu, also in the Manding region to the east; February 12, 1975. 26. Elders of Niaganan ward (themselves members of the family which formerly provided the oeyi), Thionk-Essyl, Senegal, February 12, 1975. Information on the first oeyi was offered by Suleyman San∑ (b. ca. 1886), Niaganan ward, Thionk-Essyl, Senegal, April 1984. 27. Interview with Ernest Sagna (b. 1919), village chief of Elana, Senegal, April 2, 1975. 28. Interview with Suleyman Sagna of Elana, Senegal, April 3, 1975. 29. “Les Jolas seraient venus des Bagnuns, qui seraient venus des Mandings.” Interview with Cherif Chamsedine Haidara at his home in Dar Ould Khair, Casamance, May 25, 1975. Haidara himself published a history of the Islamization of Casamance. Haidara, a non-Jola who had settled in the Casamance, was literate in Arabic, and spoke Manding but not Jola, represented an interesting perspective, one partly informed by local traditions. (Interview conducted in Manding through an interpreter.) 30. It is important to note that Islam spread into Buluf from the Manding in the region of Pakao. The Middle Casamance is thus associated, in the minds of many Jola Muslims, with the prestige of Islam. Hence, traditions of Pakao origins may contain an element of association with a prestigious geographical location. 31. P. E. H. Hair offers this etymology; see Lemos Coelho, Description of the Coast of Guinea (1684), Chapter 22, para. 9. An alternative etymology is proposed by Stephan Bühnen, who suggests that “Geregia” derives from the Bagnun term “jireji” [crocodile place]. See “Place Names as an Historical Source: An Introduction with Examples from Southern Senegambia and Germany,” History in Africa 19 (1992): 63. 32. Of course, such an orthodox view, which was appropriate in Portugal with its Inquisition, was wholly inappropriate in the Vintang area.

174

Notes to Pages 86–88

33. See Eve Crowley, “Contracts with the Spirits: Religion, Asylum and Ethnic Identity in the Cacheu region of Guinea-Bissau” (Ph.D. dissertation, Yale University, 1990). 34. Interview with Aneta Diatta, owner of the shrine called Murehei, Tendouk, Senegal, February 8, 1975. See Mark, A Cultural, Economic, and Religious History of the Basse Casamance, 81. 35. See Peter Mark, The Wild Bull and the Sacred Forest: Form, Meaning and Change in Senegambian Initiation Masks (New York: Cambridge University Press, 1992), appendix. This mask is in the Musée de l’Homme, Paris. It was collected before the French Revolution. On the Islamization of the Jola north of the Casamance River, see Mark, A Cultural, Economic and Religious History of the Basse Casamance. The Jola of Fogny were first attracted to Islam at the end of the nineteenth century. However, at least fifty years earlier, Muslim marabouts were selling amulets with Arabic writing to non-Muslim Jolas. 36. “Il nous fit entrer dans une petite église qui en était proche dont il fit sonner la cloche et nous dit que c’était lui qui l’avait fait bâtir quoi qu’il ne fut pas chrétien mais parce qu’il aimait les chrétiens.” La Courbe, Premier Voyage du Sieur de la Courbe, 218. 37. The Portuguese priests at Bissau told La Courbe that “la defense de notre religion d’avoir plusieurs femmes est un grand obstacle à leur conversion” [our religion’s prohibition against having several wives is a serious obstacle to their conversion]. Ibid., 212. 38. La Courbe writes that these African Christians (grumettes) were commercial intermediaries between the Portuguese and the local population. Ibid., 220. In this instance, the supposed moral enlightenment afforded by Christianity was mitigated by the fact that the trade they facilitated was in human beings. Four hundred captives annually were purchased on the islands by Europeans. 39. (Karen) Amanda Sackur, “The Development of Creole Society and Culture in Saint-Louis and Gorée, 1719–1817” (Ph.D. diss., School of Oriental and African Studies, University of London, 1999), 237ff. 40. “Certains noirs et mulâtres qui se disent Portugais . . . la plupart de ceux ne font ni prières, ni sala, et d’autres font tous les deux; car étant avec les nègres ils font sala, et quand ils voient des blancs, ils prennent leur chapelet et font comme eux.” La Courbe, Premier Voyage du Sieur de la Courbe, 193. 41. Sackur emphasizes the contextual nature of religious choice among Creoles in “The Development of Creole Society and Culture in Saint-Louis and Gorée.” 42. La Courbe, Premier Voyage du Sieur de la Courbe, 212. This is a common complaint among missionaries in the Casamance-Bissau region, even in the nineteenth and twentieth centuries. 43. “Depuis Cacheu jusqu’à (Farim) il y avait le long de la rivière d’espace en espace des villages habités par des gourmettes portugais qui allaient chercher du cire dans le milieu du pays.” Ibid., 232. 44. See, for example, Mamadou Diouf, “Assimilation coloniale et identité religieuse de la civilité des originaires des Quatre Communes (Sénégal),” in AOF, Réalités et héritages: Sociétés ouest-africaines et ordre colonial, 1895–1960, edited by Charles Becker, Saliou Mbaye, and Ibrahima Thioub, 2 vols. (Dakar: Direction des Archives du Sénégal, 1997), 2: 839. 45. “Un village composé de plusieurs maisons portugaises. . . . Ensuite nous allâmes à terre à la maison du signor Patrisio-Parese, Portugais blanc, fils d’un Hollandais et d’une mulâtresse.” La Courbe, Premier Voyage du Sieur de la Courbe, 228. 46. One can say with some confidence that Parese is not a Dutch name. Sr. Parese could not have received this name as a patronym unless his father was a Dutch Jew of Portuguese heritage.

Notes to Pages 89–90

175

47. As Jean Boulègue has pointed out, one could probably become “Portuguese” by becoming a prominent trader; seminar, Centre de Recherches Africaines, Paris, February 18, 1999. Generally, however, to become “Portuguese,” one needed also to demonstrate the other cultural traits—language, religion, material culture—associated with this identity. 48. “C’est un gros village tout composé de Portugais qui y sont établis de père [sic] en fils depuis longtemps . . . considérable par le nombre de ses maisons et de ses habitants Portugais blancs, noirs, bazanez et mulâtres.” Jean-Baptiste Labat, Nouvelle Relation de l’Afrique Occidentale, 5 vols. (Paris: Cavelier, 1728), 5: 160. 49. “Quoi qu’ils soient noirs, ils assurent néanmoins qu’ils sont blancs, voulant signifier par là qu’ils sont chrétiens comme les blancs.” La Courbe, Premier Voyage du Sieur de la Courbe, 192. 50. Marriages also served to link the Creoles of Saint-Louis and the Isle of Gorée to nearby African societies. In her detailed study of Creole society, Sackur shows that such unions were relatively rare and increasingly uncommon after the mid-eighteenth century. Sackur, “The Development of Creole Society and Culture in Saint-Louis and Gorée,” 155ff. 51. The name is from the Wolof “samba linguèere” [a very wealthy individual]. 52. Europeans and Africans brought from their respective cultures similar practices of cementing alliances through marriage. An earlier example of a marriage that cemented economic ties between European and African families was the sixteenth-century union, mentioned by Alvarez de Almada, of a Portuguese Jewish merchant and the daughter of the Grand Fulo; P. E. H. Hair, trans., “An Interim and Makeshift Edition of André Alvares de Almada’s Brief Treatise on the Rivers of Guinea . . . ,” translation, introduction, and notes on Chapters 13–19 by P. E. H. Hair, organized by A. Teixeira da Mota (University of Liverpool, Department of History: Privately issued by Prof. Hair, 1984), Chapter 2, page 23. Family connections also helped to secure commercial ties among European traders; see the example of Amsterdam-based Jews who traded with New Christian counterparts on the Petite Côte in Mark, “The Evolution of ‘Portuguese’ Identity: Luso-Africans on the Upper Guinea Coast from the 16th to the Early 19th Century,” Journal of African History 40, no. 2 (June 1999): 173– 191. 53. “Elle avait l’air noble et la langue bien affilée et parlait bon Portugais, Français et Anglais, marque très assurée du grand commerce qu’elle avait eu avec toutes ces nations.” La Courbe, Premier Voyage du Sieur de la Courbe, 198. The Frenchman’s description of her forceful and seductive personality continues: “Elle est belle, grande, et bien faite, quoiqu’un peu sur l’age, et est l’écüeil où quantité de blancs de plusieurs nations ont fait naufrage . . . Après lui avoir fait un present de corail et d’ambre, je me retirai de ce lieu comme Ulisse fit de la maison de Circé.” [She is beautiful, large, and well-built, though a bit past her prime, and she is the shoal upon which a number of whites from several nations have been shipwrecked. . . . After having made her a present of some coral and amber, I withdrew from this place like Ulysses from the house of Circe] (196, 198). 54. La Courbe, Premier Voyage du Sieur de la Courbe, 203. 55. La Courbe reports that the alquier had fathered one of her children and that the Englishman had had the child put to death but that he nevertheless remained with her “parce qu’elle lui avait apporté beaucoup de bien” [because she had brought him significant benefits]. Ibid., 204. 56. On client-host relations and trade in the Casamance see Mark, A Cultural, Economic and Religious History of the Basse Casamance. In 1850, BertrandBocandé, the French résident at Carabane in the Casamance, wrote, “Les Floups . . . attachent le plus grand honneur à reçevoir des traitants chez eux, ils en sont

176

Notes to Pages 91–94

fiers comme s’ils possédaient eux-mêmes les marchandises, qu’ils aident à vendre.” [The Floups . . . deem it a great honor to welcome traders in their homes, and they are as proud as if they themselves owned the merchandise that they help to sell.] E. Bertrand-Bocandé, “Voyage au Pays de Kion” (1850), sous-série 1G 23, ANS. 57. “Sont hospitaliers et ne laissent passer aucun étranger de leur nation, sans leur donner à boire et à manger de ce qu’ils ont, et même durant quelques jours.” Jacques Le Maire, Les Voyages du Sieur Le Maire aux Iles Canaries, Cap-Verd, Sénégal et Gambie (Paris: J. Callombat, 1695), 119. 58. “L’alquier . . . nous mena voir le roi de Guerègue: nous le trouvâmes à l’entrée de sa case; c’était un petit homme trapu, il avait un bonnet à la portugaise et un habit de nègre, et tenait une épée à l’espagnole à la main sur laquelle il s’appuyait. Après avoir salué, il nous fit entrer dans sa case et nous fit asseoir sur des selles de bois; je lui dis le sujet de ma venue [they discuss commerce]; je luy fis present d’un peu de fer et d’eau de vie et, quelque temps après, il nous pria d’entrer dans une autre chambre; car sa case est faite à la portugaise; nous trouvâmes le dejeuner tout prêt . . . il se mit avec nous et en mangea aussi bien que sa femme, ce qui me fit voir qu’ils commèncent en cet endroit là à prendre les manières des Anglais” (emphasis mine). La Courbe, Premier Voyage du Sieur de la Courbe, 205–206. 59. In “Territoires et droits en Afrique noire, Essai d’anthropologie du droit foncier en milieu rural” (Thèse pour le doctorat, Université de Paris X, 1999), Charles de Lespinay argues convincingly that Bagnun-speakers constituted an autochthonous population in the territory later controlled by the Mande rulers of Kaabu. He further suggests that many political and religious customs throughout the Bagnun region derived from these autochthones. It is possible that the staff of authority constituted such a custom. However, the facts that iron staffs are associated with political (and spiritual) power throughout the Mande culture zone and that this emblem is historically associated with the founding ruler, Sunjata, strongly suggest that the iron staff derives from the Mande heartland. 60. See Amselle, Logiques métisses, 55. 61. “La compréhension du métissage se heurte à des habitudes intellectuelles qui portent à préferer les ensembles monolithiques aux espaces intermédiaires. Il est plus facile en effet d’identifier les blocs solides que des interstices sans nom.” Serge Gruzinski, La Pensée métisse (Paris: Fayard, 1999), 42. 62. See Mark, A Cultural, Economic and Religious History of the Basse Casamance, 17ff. 63. See ibid.; see also Mark, The Wild Bull and the Sacred Forest. 64. Cheikh Suleyman San∑ of Thionk-Essyl, an informant with both an extraordinary memory and a profound historical knowledge unusual even for one of his age, listed eighteen generations of initiations in Thionk-Essyl since the migration of Floup ancestors to Thionk from Essyl in Bandial, south of the Casamance River. The migration, however, is said to have occurred several generations before the people of Thionk first celebrated their bukut initiation in their new home. On the use of initiation lists as a chronological measuring tool, see Mark, A Cultural, Economic, and Religious History. Cheikh Suleyman was about 99 years old when he recounted this bukut list in April 1984. 65. On the early history of the Bagnuns, see Lespinay, “Territoires et droits en Afrique noire,” 239: “The Bagnun appear to have been the first inhabitants (or the inheritors from them) in the entire region of southern Senegal, including the Gambia, in Guiné-Bissau, and in the eastern part of Guinea.” Lespinay argues that the Bagnun established a model for social relations and territorial arrangements of ownership and land use that was later followed by other inhabitants of

Notes to Pages 94–97

177

the region. Oral traditions of the Esulalu-Jola also record an early Bagnun presence in the Esulalu region south of the Casamance River. See Robert Baum, Shrines of the Slave Trade: Diola Religion and Society in Precolonial Senegambia (Oxford: Oxford University Press, 1999), 64. 66. The Esulalu-Jola south of the Casamance River also have oral traditions asserting that they are in part descended from Bagnun (or Bainunk); ibid. 67. Interview with Cheikh Suleyman San∑ of Thionk-Essyl, April 3, 1984. Malamanka Kaabu was the first king, or oeyi, of Essyl, even before the Floups migrated north to Thionk-Essyl. It is unlikely that the migrants would have brought with them the institution of oeyi unless it was already firmly established in Essyl-Bandial. Thus, it must have existed for some time before their migration. Strictly on the basis of this tradition, Malamanka’s arrival must be dated to considerably more than twenty generations ago. The name “Malamanka,” or “Mala Manga,” is equivalent to the patronym “Mané.” Malamanka Kaabu (or Mala Manga) was known as “Afaleyo,” or “I have found it.” His descendent, Djimanga (Mané), was, according to this tradition, the first Floup migrant to Thionk Essyl from Essyl-Bandial. 68. See Stephan Bühnen, “Place Names as an Historical Source: An Introduction with Examples from Southern Senegambia and Germany,” History in Africa 19 (1992): 45–101. This interpretation is quite consistent with La Courbe’s description of Geregia. 69. Boulègue estimates the seventeenth-century Luso-African population of the Petite Côte at about 500. At the end of the eighteenth century, the LusoAfrican inhabitants of Joal may have numbered more than 1,000. Les LusoAfricains de Sénégambie, 29, 80. 70. See Prosper Cultru’s introduction to the 1913 edition of La Courbe, Premier voyage du Sieur de la Courbe. 71. Jean Boulègue, “Contribution des sources françaises à la connaissance de l’actuelle Guinée-Bissau à la fin du XVIIe siècle,” History in Africa 28 (2001): 43–51. Boulègue shows that Labat had access to information from as late as 1705 that he used in his description of the Manding state of Kaabu. If one makes the not-unreasonable assumption that Labat’s most recent information on Vintang dated to about the same time, then the differences between his text and La Courbe’s, specifically the decline of the town’s population, can be dated with relative precision to the period between La Courbe’s 1686 visit and 1705. 72. “Le bourg ou village de Bintam était autrefois plus considérable qu’il n’est à présent.” Labat, Nouvelle Relation de l’Afrique Occidentale, 5: 5. 73. “Les Français ont toujours eu, depuis de longues années, des établissements à Bintan . . . ainsi qu’à Gérèges . . . établissement qu’ils retabliront dès qu’ils le jugeront à propos.” Demanet, Nouvelle Histoire de l’Afrique occidentale, 140. 74. “Le temps auquel les Portugais avaient été les maîtres de la rivière de Gambie et avaient eu des établissements dans la côte d’Afrique.” Ibid., 142. 5. SENEGAMBIA FROM THE MID-EIGHTEENTH CENTURY TO THE MID-NINETEENTH CENTURY 1. “Les Mandingues . . . sont marchands par excellence et . . . ils sont les seuls facteurs du commerce; c’est par leur intermédiaire que viennent à la côte tous les produits de l’intérieur de l’Afrique et que les marchandises européennes circulent en revanche dans l’intérieur.” “Rapport de la commission d’exploration du bas de la côte” (1837) (hereafter “Mission Dagorne,” sous-série F 12 7208, Centre des Archives Nationales, Paris (hereafter CARAN).

178

Notes to Pages 97–100

2. “Voyage dans le Sougrougrou,” January 23, 1861, sous-série 1 G 34 7, ANS. 3. See René Pélissier, Naissance de la Guiné: Portugais et Africains en Sénégambie (1841–1936) (Orgeval, France: Pélissier, 1989). 4. “Ils envoyent dans le haut de la rivière des pirogues qui traitent dans les villages du bord du fleuve une petite quantité des mêmes produits.” “Mission Dagorne,” 16, CARAN. 5. “Mission Dagorne,” 17. Dagorne calculated the entry duties at 24 percent and export duties at an additional 15 percent. Pélissier estimates that Portuguese commerce in 1842 was equal to only 5 percent of combined French and British imports into the Casamance (Naissance de la Guiné, 40). 6. “L’argile sechée au soleil compose les murailles des maisons. La ville, si l’on peut donner ce nom à ce pitoyable assemblage de quelques chetives cases, est defendue par un tata en branches brutes entrecroisés, aux quatre angles du quel sont quatre sortes de bastion, murées en argile seche avec des embrassures qui donnent passage à la volée de quelques vieux canons sans affuts. . . . Il est miraculeux que ces murailles ne tombent.” “Mission Dagorne,” 18–19. 7. “Ils sont résignés par l’habitude et par l’apathée naturelle due à leur origine, car ils sont tous noirs même le Provedor.” Ibid., 19. 8. Hecquard’s journey to Timbo, where he established a friendship with the Almami Oumar, is recounted in detail in Hyacinthe Hecquard, Voyage sur la côte et dans l’Intérieur de l’Afrique Occidentale (Paris: Impr. de Bénard et compagnie, 1855). See also his “Rapport sur un Voyage dans la Casamance,” Revue Coloniale, May 1852. 9. Hecquard Sammlung, Bilder-Archiv, Frobenius-Institut, Goethe Universität, Frankfurt am Main. It is not known how these illustrations come to be at the Frobenius-Institut. According to the former director of the Institut, Dr. Eike Haberland, all records pertaining to this collection were destroyed during World War II. 10. Bilder-Archiv, number 2456. For a discussion of some of the drawings in this collection, see Peter Mark, “Hyacinthe Hecquard’s Drawings and Watercolors from Grand Bassam, the Futa Jalon, and the Casamance: A Source for MidNineteenth Century West African History,” Paideuma: Mitteilungen zur Kulturkunde 36 (1990): 173–184. 11. Gaspard Mollien, Voyage dans l’Intérieur de l’Afrique aux Sources du Sénégal et de la Gambie, 2 vols. (Paris: Imprimerie de Mme Veuve Courcier, 1820), 2: 244. 12. “Zéguichor [sic], établissement portugais formé en 1645. . . . C’est un village composé d’une centaine de cases carrées. . . . Ces cases ont une toiture mobile, en paille, et qu’on enlève pendant la sécheresse, dans la crainte d’incendie et pour avoir plus de fraîcheur. Le gouverneur . . . son logement ne diffère pas beaucoup de celui des autre habitants. . . . Il y avait jadis une chapelle desservie par deux prêtres, mais il n’en reste plus de traces, les ornements ont été conservés et placés dans une case.” [Zéguichor, Portuguese post established in 1645. . . . It is a village of about one hundred square houses. . . . These houses have a mobile roof of thatch, which is removed during the dry season for fear of fire and to have added coolness. The governor . . . his lodging is little different from those of the other inhabitants. . . . Formerly there was a chapel served by two priests, but nothing remains, the (ritual) ornaments have been preserved in a house.] Ibid., 109. 13. “Depuis une année, l’église n’existe plus; les averses de la saison pluvieuse l’avait fait fondre; car ici on ne batit qu’avec de la terre glaise. Les ressources de l’endroit ne permettent pas de la relever.” Père Lucien Lacombe, 1864, “Annales réligieuses de la Casamance,” 6, sous-série Z 2Z2, Microfilm 14, Mi 1610, Centre d’Archives d’Outre-Mer, Aix-en-Provence, France (hereafter CAOM).

Notes to Pages 100–104

179

14. “Zikinchor . . . rien de plus triste que l’aspect de Zikinchor. . . . Ce n’est qu’une agglomération de cases, grossièrement travaillées et recouvertes de paille. Sur les quais, dans les rues, partout, c’est un désordre. . . . Le dernier village de noirs n’offre pas des choses si misérables.” Père Gabriel Sène, “Tournée en Casamance,” 1874. Ibid., 11ff. 15. “Mission Dagorne,” 46. 16. “Les traitants portugais du Géba qui y viennent par terre.” “Rapport de Bertrand Bocandé (1857),” sous-série 11D 1 282, ANS. 17. “Les Portugais de Ziguinchor sont devenus des facteurs du commerce français en Casamance.” Ibid. 18. “Les Mandingues . . . ne tarderont pas à s’emparer du pays.” “Voyage dans le Sougrougrou,” January 23, 1861, Exploration Haute Casamance (1860– 1866), Chemise 7, sous-série 1G 34 7, ANS. 19. “Les Yolas ont soin d’entourer leur groupe de cases d’une palissade [offrante?] une solide résistance.” Ibid. 20. “Les Portugais de Ziguinchor out seulement des pirogues . . . faites d’un seul tronc d’arbre . . . exhaussées de bordages en planches épaisses taillées à la hâche.” “Rapport de Bertrand-Bocandé (1857)” (Extrait du Moniteur Officiel du Sénégal, 1857), sous-série 11D 1 282, ANS. 21. “Voyage dans le Sougrougrou.” See also “Rapport de M. Faliu, Lt. Cdt. Sédhiou,” sous-série 1G 34 12, ANS. 22. “Cette partie du Songrougrou se trouvant voisine des routes de la Gambie et de la Haute Casamance, ayant été fréquentée par des [courtiers?] qui faisaient très bien leurs affaires avant que notre commerce eut l’idée d’y envoyer des traitants, est beaucoup plus civilisée que le bas Songrougrou.” “Voyage dans le Sougrougrou,” January 23, 1861. 23. “Le départ du Portugais me laissant sans interprète j’ai été obligé de quitter Kanbanba, sans avoir obtenu le moindre détail.” [The departure of the Portuguese leaving me without an interpreter, I was obliged to leave Kanbanba without having achieved the least part.] Ibid. 24. See Donald Wright, The World and a Very Small Place in Africa (Armonk, N.Y.: M. E. Sharpe, 1997). 25. Christian Schefer, ed., Instructions Générales données de 1763 à 1870 aux gouverneurs et ordonnateurs des établissements français en Afrique, 2 vols. (Paris: E. Champion, 1927), 2: 299. 26. “La ville [villa] comme ils l’appellent est composée d’une centaine de cases carrées bâties en terre sêchée au soleil, recouvertes d’un platforme en terre pour le beau temps, surmontées pour la saison des pluies d’un toit aigu en paille. Ces cases sont habitées par une population toute noire, qui est supposée catholique— on ne sait pourquoi mais qui au logement près, a les mêmes usages et les mêmes moeurs que les Papels environnants, dont elle est une émanation.” “Mission Dagorne,” 60. 27. “Report of November 14, 1844,” Afrique IV 25 a, CAOM. 28. “Mission Dagorne,” 63. 29. “Gens de divers nations, tous plus ou moins chrétiens, . . . ce sont les Gourmettes. Ils ne s’occupent jamais d’aucun travail de la terre, leur élément est la mer. Ils arment toutes les nombreuses pirogues qui remontent la rivière de Géba, pour prendre les produits qui abondent dans cette partie, cire, ivoire, peaux, huile de palme. . . . Une partie des Gourmettes embarque même sur les navires qui font les voyages d’Europe; d’autres sont ouvriers de toute sorte et servent souvent de pilotes aux cabotiers de diverses nations.” Revue Coloniale 2 (1845): 135ff., Afrique IV 25 a, CAOM. 30. “Les préjugés de couleur sont inconnus au Sénégal.” David Boilat, Esquisses Sénégalaises (1853; reprint, Paris: Karthala, 1984), 209.

180

Notes to Pages 104–109

31. “The gourmets or baptized Blacks . . . are also traders; their wives are known as ‘signares.’” Ibid., 211. 32. “L’acte de piraterie commie par ces barbares qui y étaient conduits par un nommé Manuel Gara portugais de Bissau.” “Relations avec la Gambie” (January 1, 1849), sous-série 1F5 34, ANS. 33. Adam Afzelius, Sierra Leone Journal 1795–1796, edited by Alexander Kup (Uppsala: Studia Ethnographica Upsaliensia, xxvii, 1967), 105. 34. Political power relationships affected representations of Africa and Africans throughout the precolonial period. José da Silva Horta observes that “the ‘Guinea of Cabo Verde’ concept implies a representation of power: an European or Euro-African power projected onto the African regions.” “Evidence for a LusoAfrican Identity.” 35. “Relations avec la Guinée-Portugaise,” dossier 13 (1879), sous-série 2F 4, ANS. 36. E. Bertrand-Bocandé had two families; after his return to Paris, he married and raised a second family. For a biographical sketch of this fascinating individual, see G. F. Joachim-Arnaud Bocandé, “Biographie de la Famille Bocandé” (unpublished seminar paper, Université Paris I-Sorbonne, Centre de Recherches Africaines, 1991). See also Jean Bertrand-Bocandé, G. Debien, and Y. Saint-Martin, “Emmanuel Bertrand-Bocandé (1812–1881), un Nantais en Casamance,” Bulletin de l’Institut Fondamental d’Afrique Noire XXX, no. 1 (1969). 37. Mollien, Voyage dans l’Intérieur de l’Afrique, 2: 218. 38. “Ces braves gens avaient cru jusqu’à présent que les blancs étaient nos traitants noirs qui allaient chez eux, et ils étaient loin de se douter qu’il y en avait d’une autre couleur.” The administrator reported that he had met a few traitants working for a French trading house, as well as some former traitants, themselves Wolofs, who had settled in Fogny. “Voyage dans le Sougrougrou,” January 23, 1861. 39. “Par une confusion très facheuse, la qualification de blanc est donnée, dans les pays où paraissent nos traitants, à tout individu qui habite Saint-Louis, sans distinction de couleur. Ainsi, à ce compte, mes nègres étaient des blancs.” Anne Raffenel, Nouveau Voyage dans le Pays des Nègres (Paris: Impr. de. N. Chaix et cie, 1856), 57. 40. “Ont une grande idée de la richesse des blancs, et même de celle des Arabes, qu’ils mettent dans la même catégorie.” René Caillié, Journal d’un voyage à Temboctou et à Jenné, dans l’Afrique Centrale, 3 vols. (1830; reprint, Paris, La Découverte, 1996), 2: 20. 41. “Depuis de nombreuses années déjà, dans les cultures populaires urbaines de plusieurs sociétés d’Afrique noire, le qualificatif de ‘blanc’ est un synonyme de bourgeois, sans égard particulier à la couleur de la peau.” B. Jewsiewicki, “Race et ethnie, acteurs politiques virtuels de cette fin de siècle,” in Les convergences culturelles dans les sociétés pluriethniques, edited by K. Fall, R. Hadj-Moussa, and D. Simeoni (Sainte Foy: Presses Universitaires de Québec, 1996), 329. 42. Geoffroy de Villeneuve, L’Afrique, ou Histoire, moeurs, usages et coutumes des Africains: le Senegal, 4 vols. (Paris: Nepveu, 1814). In volume 1, Villeneuve writes that he made two trips to Africa, one from 1785–1788 with de Boufflers and the other, lasting two years, to Gorée, where he learned to speak Wolof. 43. “On croit que les Sereres sont des colonies d’une nation nommée Guiola, qui existe encore sur les bords de la Gambie. Des rapports dans la langue des deux peuples, dans les moeurs, dans les usages, confirment cette opinion.” Ibid., 4: 3. 44. See, for example, Emmanuel Bertrand-Bocandé, “Notes sur la Guinée portugaise ou Sénégambie méridionale,” Bulletin de la Société de Géographie XII, no. 67 (1849): 68; Peter Mark, A Cultural, Economic and Religious History

Notes to Pages 110–112

181

of the Basse Casamance Since 1500 (Stuttgart: Steiner Verlag/Frobenius-Institut, 1985). The one scholar who accurately attributes the word “Diola” to the Jola themselves is Louis-Vincent Thomas; see his foreword in Mark, The Wild Bull and the Sacred Forest: Form, Meaning, and Change in Senegambian Initiation Masks (New York: Cambridge University Press, 1992), ix. 45. “On rencontre quelquefois, dans le pays de Barra [in the Lower Gambia] une espèce de nègres qui a conservé toute la rudesse et les habitudes de la vie sauvage. Ils habitent une petite contrée sur les bords de la rivière de Casamance, près de la Gambie, et passent quelquefois ce dernier fleuve pour venir vendre des bestiaux. On les nomme Félups [sic]. Les Félups vont presque nus, hors un très-petit tablier qui passe entre les cuisses. Ils se serrent le haut des bras et près des poignets, et le haut des cuisses et au-dessus du genou, le haut et le bas des jambes, avec des lacets de cuir teints en rouge; de sorte que les intervalles de leurs membres, ainsi serrés, sont beaucoup plus gros que dans l’état naturel. Ils se cicatrisent le visage et le corps, y gravent toutes sortes de figures et de dessins informes et bizarres. Ces nègres ont les cheveux fort laineux et crépus. Ils les rassemblent sur le sommet de la tête, au-dessus du front, et en forment une sorte d’aigrette, qui a 5 à 6 pouces de longueur; ils laissent croître leur barbe et la taillent en pointe. Ils sont couverts d’amulettes ou gris-gris; leurs armes sont des arcs, des fleches, des carquois suspendus derrière le dos. Ils portent de la main gauche plusieurs zagayes ou javelots de diverses grandeurs, qu’ils lancent avec beaucoup de dexterité. Leur langage est rude, leur prononciation vive, sourde et gutturale; quoique peu communicatifs, ils ne sont point féroces, et vivent bien avec les hordes voisins.” Villeneuve, L’Afrique, ou Histoire, moeurs, usages et coutumes des Africains, 4: 174–175. 46. The Fogny dialect, which these northern Jolas would probably have spoken, has always struck me as poetically alliterative, clearly articulated, and somewhat mellifluous. But then, it is the one Senegalese language which I do speak. 47. “Les Yolas au contraire qui habitent les Iles de l’embouchure de la Casamance et qui n’ont pas de relations suivies avec le continent, sont grossiers et sans constitution politique. Chaque village forme un état séparé qui fait souvent la guerre aux villages voisins, tandis que chaque village, même, est une sorte de république où chaque chef de famille agit d’une manière tout à fait indépendante.” “Mission Dagorne,” 8. 48. “On pourrait presque sur la côte, reconnaître l’état gouvernemental à la vue des habitations des indigènes.” Ibid., 9. 49. “Plus avancé sous le rapport de la civilisation.” Ibid., 9–10. 50. “Le Yola, sans lois et sans chef, toujours pret à défendre sa propriété comme à envahir celle d’autrui se construit une vaste maison en argile séchée au soleil, avec des murailles hautes et épaisses, des compartimens étroits et multipliés, des corridors tortueux que coupent de nombreuses et étroites portes. On voit qu’il veut prolonger sa défense jusqu’à la dernière extremité. Sa demeure est, en quelque sorte, à cela près de la différence des matériaux et de l’étendue, et avec autant de force relative ce qu’étaient les chateaux de nos belliqueux et souvent peu scrupuleux seigneurs du moyen age.” Ibid., 10. 51. “Ce village . . . a un aspect misérable. Les cases sont pourtant en terre, comme partout ailleurs et aussi bien construites, carrées au lieu d’être rondes; à coté de chaque case se trouve même un petit enclos de coton pour les besoins de la famille. Mais les habitants sont si laids, si repoussants, que cela m’a laissé une mauvaise impression.” “Voyage de M. Vallon, lieutenant de vaisseau, délégué du Gouverneur en Casamance,” February 1862, sous-série 1G 34 19, ANS. 52. “Les habitans de Dacar sont les plus civilisés de toute la côte; ils aiment beaucoup les Français et commencent à prendre les manières et les habitudes des Européens. Plusieurs même, lorsque j’étais en Afrique, avaient déjà construit des

182

Notes to Pages 112–116

maisons en pierres.” Villeneuve, L’Afrique, ou Histoire, moeurs, usages et coutumes des Africains, 1: 124–125. 53. “On ne rencontre nulle part aucune construction, aucun monument par lequel on ait voulu transmettre des hauts faits à la posterité.” Ibid., 4: 140. 54. “La classe dirigeante ajoor a toujours refusé [aux européens] toute construction en dur sur son territoire. La tradition rapporte que les damel . . . ont toujours affirmé que les demeures ou entrepôts en dur étaient susceptibles de servir de base d’agression ou de défense.” Mamadou Diouf, Le Kajoor au XIXe siècle, pouvoir ceddo et conquête coloniale (Paris: Karthala, 1990), 119. 55. In the Middle and Upper Casamance at mid-century, Manding Muslims destroyed the palisades that surrounded Peulh and non-Muslim Manding villages. This was a means of subjugating the communities. See “Voyage dans le Sougrougrou.” 6. CASAMANCE ARCHITECTURE FROM 1850 TO THE ESTABLISHMENT OF COLONIAL ADMINISTRATION 1. Bocandé spoke some Jola. In several official reports, now in the Archives Nationales in Dakar, he glosses Diola terms. He may have spoken the language well. 2. The ethnographic study of Casamance that he published in 1849 is the most important nineteenth-century source about the social structure and history of the Lower Casamance. See Emmanuel Bertrand-Bocandé, “Notes sur la GuinéPortugaise ou Sénégambie Méridionale,” Bulletin de la Société de Géographie XII, no. 67/68 (1849). 3. Ibid., 305. 4. E. Bertrand-Bocandé, “Voyage au pays de Kion” (1850), sous-série 1G 23, ANS. This report has been published; see Jean Bertrand-Bocandé, G. Debien, and Y. Saint-Martin, “Emmanuel Bertrand-Bocandé (1812–1881), un Nantais en Casamance,” Bulletin de l’Institut Fondamental d’Afrique Noire XXXI, no. 1 (1969). 5. Louis-Vincent Thomas, Les Diola: Essai d’analyse fonctionelle sur une population de Basse-Casamance (Dakar: Institut Fondamental d’Afrique Noire, 1959), 128. 6. Thionk is the largest community, with the exception of the regional administrative center of Bignona. Bignona, however, was founded only in 1893– 1894. 7. “Chaque habitant bâtit sa demeure isolément pour lui et sa famille, au milieu de son champ, sur des terrains élevés au dessus des terres cultivés en riz au bord des marigots.” “Voyage au Pays de Kion.” 8. See also Dr. Lasnet, et al., Une Mission au Sénégal (Paris: A. Challamel, 1900), 157. 9. “Une habitation se compose de diverses cases en terre en forme de petite tourettes [tourelles?], sans fenêtres au dehors, plus ou moins nombreuses suivant la richesse et la famille qui l’habite; ces tourettes couvertes d’un toit de chaume sont réunies entr’elles par des murs également en terre garantis des pluies par des troncs de palmier creusés en forme de dalle qui reçoivent aussi pour les conduire au dehors les eaux qui s’écoulent des toits. Ces diverses constructions sont encore ceintes au moins en partie d’une palissade; elles paraissent à l’extérieur de petites forteresses avec leurs tours et leurs courtines, à l’intérieur elles composent par leur réunion sans symmétrie un petit labyrinthe, où deux issues sont pratiquées dans deux côtés opposés, pour permettre l’entrée, ou faciliter la fuite dans le danger. Des petits bosquets épais et fourrés sont encore ménagés de distance en distance.” “Voyage au pays de Kion.”

Notes to Pages 116–120

183

10. Sieur Michel Jajolet de la Courbe, Premier Voyage du Sieur de la Courbe faite à la coste d’Afrique en 1685, edited by Prosper Cultru (Paris: E. Champion, 1913), 238. 11. Interview with the elders of Batine ward, including Bakari Badji, imam of Batine’s mosque, and chef de quartier Cheikh Abba Badji, Thionk-Essyl, February 13, 1975. For archival confirmation of the sale of children to Dioula traders, see “Rapport de la situation politique du Fogny, décembre 1899,” sous-série 2G1 69, ANS. 12. Fodé Kaba Doumbouya was actually a Jahanké. 13. Indestructible, that is, until the French arrived with artillery in 1895. 14. “Les Yolas ont soin d’entourer leur groupe de cases d’une palissade offrante une solide résistance.” “Voyage dans le Sougrougrou,” January 23, 1861, sous-série 1G 34 7, ANS. 15. “A Carabane j’étais un jour entré dans une cour où il y avait plusieurs cases ayant chacune sa petite enceinte, faite de bâtons liés avec l’écorce.” Père Spiesser, cited in “Annales Religieuses,” 52, sous-série 14 Mi 1610, CAOM. 16. “Bignona, Rapport politique, 2e trimestre 1902,” sous-série 2G 2 20, ANS. 17. “Ils voyagent dans toute la basse Casamanace et se répandent dans tous les villages; . . . ils aident à la construction des cases en terre.” [They travel throughout the Lower Casamance and are found in all the villages . . . they help to construct the earthen houses.] “Rapport de Bertrand Bocandé (1857),” soussérie 11D 1 282, ANS. 18. The administrateur at Sedhiu describes Diannah (population 3,000) as the largest village in Pakao: “Un mur d’enceinte en terre aujourd’hui presque entièrement demoli, les servait jadis de défense.” [A protective wall of earth, today almost entirely demolished, formerly served for defense.] “Exploration Haute Casamance (1860–1866),” sous-série 1G 34, ANS. 19. Valentim Fernandes, Description de la Côte occidentale d’Afrique (Sénégal au Cap de Monte, Archipels), edited and translated by R. Mauny, Th. Monod, and A. Teixera da Mota (Bissau: Centro de Estudos da Guiné Portuguesa, 1951), 59. 20. P. E. H. Hair, trans., “An Interim and Makeshift Edition of André Alvares de Almada’s Brief Treatise on the Rivers of Guinea . . . ,” translation, introduction, and notes on Chapters 13–19 by P. E. H. Hair, organized by A. Teixeira da Mota (University of Liverpool, Department of History: Privately issued by Prof. Hair, 1984), Chapter 5, page 44. 21. André Donelha, Description de la Serra Leoa et des Rios de Guiné du Cabo Verde (1625), edited by A. Teixeira da Mota, translated into French by Léon Bourdon (Lisbon: Junta de Investigações Científicas do Ultramar, 1977), 141. 22. Richard Jobson, The Golden Trade, or a Discovery of the River Gambra (London: Nicholas Okes, 1623), 56. 23. “On passa au travers de quelques villages de Floupes qui étaient environnez [sic] de cinq ou six rangs de palissiades.” Jean-Baptiste Labat, Nouvelle Relation de l’Afrique Occidentale, 5 vols. (Paris: Cavelier, 1728), 5: 33. 24. “Une double enceinte de gros pieux.” Ibid., 4: 414. 25. Francis Moore, Travels into the Inland Parts of Africa (London: Printed by E. Cave for the author, 1738), 36. 26. Hair, trans., “An Interim and Makeshift Edition of André Alvares de Almada’s Brief Treatise,” Chapter 9, page 90. 27. La Courbe, Premier Voyage du Sieur de la Courbe, 238. 28. In this chronological context, the use of the term “Diola” is slightly proleptic. Among French officials and missionaries, “Diola”—originally used interchangeably with “Yola”—came into use gradually during the second half of the

184

Notes to Pages 120–123

nineteenth century. By about 1900, “Diola” had gained generic connotation. It was subsequently applied to all of the groups—“Floups,” “Djougoutes,” “Bandials,” and others—living north and south of the lower Casamance River who spoke related dialects of a common language, now also called Diola. 29. Varenhorst was the photographer for the expedition commanded by Captain Brosselard-Faidherbe in 1889–1890. See collection formerly housed in the Société de Géographie, Série We, Département des Cartes et Plans, BN. 30. The remains of the veranda of a fourth house are just visible in the right corner of the photograph. 31. Thionk-Essyl’s participation in the nineteenth-century slave trade is well documented both in archival sources and in local oral traditions. See, for example, Bertrand-Bocandé: “Les Jigouches de Kion pillent partout, se volent euxmêmes et vendent aux Mandingues de la Gambie les captifs qu’ils ont enlevés.” “Rapport de Bertrand-Bocandé (1857),” sous-série 11D 1 282, ANS. See also Mark, The Wild Bull and the Sacred Forest. 32. Interview with Sekou Coudiaby (b. ca. 1924, son of the first Muslim in Mandégane) and Ansoumana Badji (assistant chief), Mandégane, March 24, 1975. These informants at first attributed the end of nineteenth-century warfare to Islamization. Subsequent and more detailed discussion, however, led them to the assertion that peace was a consequence of the establishment of colonial administration: “Nya kululumek kujalo nya kankan kasumay.” [When the Europeans arrived here, then they created peace.] Interview, Mandégane, March 26, 1975. In Kasa, south of the Casamance River, the term for this central courtyard, applied also to the extended family compound as a whole, is “hank.” For information about large, extended family compounds in Thionk-Essyl, see interview with Kepi Diatta (b. ca. 1884), Thionk-Essyl, December 30, 1974. 33. One still encounters people in Sedhiu and Pakao who identify themselves as Manding and speak Mandinka but have retained Diola patronyms. Some of them are descended from Diola slaves captured during the so-called maraboutic wars. On Diola resistance to conversion by force, see Mark, A Cultural, Religious and Economic History of the Basse Casamance Since 1500 (Stuttgart: Steiner Verlag/Frobenius-Institut, 1985), 70. 34. “Les cases du Fogny sont complètement différentes (de celles de la Basse Casamance), également en pisé mais petites, circulaires, sans cloisonnement intérieur, à moitié enterrées sous le sol, ni aérées, ni éclairées, avec des ouvertures étroites et basses, couvertes de feuilles de latanier; quelquefois elles sont placées au milieu de plusieurs enceintes circulaires faites de pieux juxtaposés et dont les ouvertures sont diamétralement opposées . . . les Fogny, quand ils rentrent de leurs rizières ou de la forêt, ils se couchent dans leurs cases et n’aiment pas en sortir; souvent ils passent à travers le paroi un tube de bambou qui leur sert d’urinoir.” Lasnet et al., Une Mission au Sénégal, 158–159. 35. Information on the function of the ekumbane room was kindly provided by Dr. Bakary Sane, Ziguinchor, June 18, 2001. 36. One need only compare Lasnet’s description of these people, who were living like rats, to the description written by Père Wintz a decade later of life south of the river: “Amis de leur liberté et de leur vie champêtre. Après le travail de leur journée, ils aiment s’égayer, le soir, sur la place du village ou autour d’un feu; on y écoute le récit des aventures de quelques héros d’autrefois.” [Lovers of liberty and of their pastoral life. After their day’s work they like to relax in the evening on the village square around a fire; there they listen to the recitation of the adventures of olden-day heroes.] Père Wintz, 1909, “Annales réligieuses,” 37, sous-série Z 2Z2, microfilm 14, Mi 1610, CAOM. This image of a people who enjoy socializing outside at the end of the day’s work, often recounting stories

Notes to Pages 123–125

185

about family and local history, accords perfectly with my own experience in Buluf during the dry season in 1974–1975, 1976, 1984, and 1986. 37. “Ces cases basses disparaissant à moitié sous le sol et bien faites en somme pour la défense se trouvent surtout dans les régions du Fogny qui étaient le plus exposées aux incursions des Mandingues; il est possible que ce soient des considérations de securité qui aient fait dans ces régions modifier le type des constructions diolas et donné aux habitants l’habitude de se terrer et de vivre dans des réduits malpropres.” Lasnet et al., Mission au Sénégal, 159. 38. Interviews with Bakari Badji (b. ca. 1895), late imam, and the other elders of Batine ward, Thionk-Essyl, February 13, 1975. The episode is remembered in song. 39. A French military officer reported on May 20, 1886, that “Sillah a envahi le pays de Thionk situé à côté de Djougoutes, mais sa témérité lui a été fatale car il a essuyé une sanglante defaite et a été grièvement blessé.” [Sillah invaded the country of Thionk, located at the edge of Djougoutes, but his bravery was fatal to him, for he suffered a bloody defeat and was seriously wounded.] Sous-série 1D 50, ANS; cited in Mark, A Cultural, Economic and Religious History, 70. 40. See “Annales Religieuses,” 57. See also Mark, A Cultural, Economic and Religious History, 69. 41. “Sans défense et peu habitué à la guerre,” 31, sous-série Z 2Z2, Microfilm 14, Mi 1610, CAOM. 42. I do not mean to imply that the Bagnun were nonviolent by nature. 43. “A son approche les Diolas épouvantés, s’enfuient, poussant devant eux leurs troupeaux. . . . Depuis Adéane jusqu’à Ziguinchor c’est-à-dire sur une distance d’environ 10 lieues, il n’est pas rare de rencontrer de nombreuses familles de Diolas échelonnées dans les villages de la côte, n’ayant pour subvenir à leur existence que l’herbe de la forêt ou des amandes de palme.” Père Lacombe, Letter of 1880, “Annales réligieuses,” 31. 44. Ibid., 32. 45. In 1860, a punitive military expedition led by Pinet-Laprade defeated the men of Thionk, thereby ending that community’s slave raiding. 46. “Marsassoum est le rendez-vous des marchands de chair humaine. Il y arrive tous les jours des captifs de la rive droite, conduits en bande, soit par les Mandings, soit par les soldats de Fodé Kaba. Ces pauvres malheureux sont la matière des transactions commerciales les plus importants. Jouets de tous les caprices de leurs maîtres . . . pour tout dire, en un mot, ils servent à tous les usages.” Père Lacombe, Letter, Easter 1880, “Annales réligieuses,” 22. 47. This price is slightly higher than, but generally in line with, those cited by Martin Klein in Slavery and Colonial Rule in French West Africa (Cambridge: Cambridge University Press, 1998), 259. He gives the 1879 price of an adult slave at Saint-Louis at 250 to 300 francs and the price of a child at 150 to 200 francs. The preponderance of children among those newly enslaved was consistent with the preference for women and children throughout West Africa (41). 48. “J’ai eu la douleur de les voir, agglomérés sous les palétuviers de la rive, hommes, femmes, enfants, accroupis dans la boue à la merci de toutes les intempéries de l’air et de les entendre, sans pouvoir hélàs leur porter secours, appeler les voyageurs qui naviguent sur la rivière et les conjurer de les transporter sur l’autre rive.” Père Lacombe, Letter of 1880, 31. Lacombe had no trouble obtaining his eight boys and girls. His description of parents being ripped away from children further suggests that many of the enslaved were youths. 49. In 1899 the résident at Bignona wrote that Juula traders in Fogny continued an illicit trade in slaves, primarily children. “La traite des esclaves n’a pas cessé dans le Fogny. Les dioulas l’entretiennent; ils poussent les diolas à vendre

186

Notes to Pages 126–133

leurs enfants et, au besoin, les enfants volés à d’autres villages.” [The slave trade has not ended in Fogny. The Dioulas keep it going; they push the Diolas to sell their own children and, if need be, children stolen from other villages.] “Rapport de la situation politique du Fogny, décembre 1899.” 50. See Christian Roche, Conquête et Résistences des Peuples de Casamance, 1850–1920 (Dakar: Nouvelles Éditions Africaines, 1976); see also Boubacar Barry, La Sénégambie du XVe au XIXe siècle (Paris: L’Harmattan, 1988), 328. 51. Interviews conducted by the author in February 1975. 52. “Le passage de Fodé Kaba, Brahima N’Diaye et autres chefs pirates dans le Fogny, la distribution du sol faite par eux à certains villages, le retour des esclaves qu’ils avaient enlevés et dont les rizières étaient cultivées par de nouveaux propriétaires ont embrouillés de belle façon la question des propriétés.” [The raids of Fode Kaba, Birahim Ndiaye, and other pirate leaders into Fogny, their distributing lands in some of these villages, and the return of those whom they carried off as slaves and whose rice fields were subsequently cultivated by new owners have truly confused the question of land ownership.] “Rapport sur la situation politique du Fogny, décembre 1899.” 53. On the growth of the rubber trade, see Mark, A Cultural, Economic and Religious History, 94ff. 54. Anfreville de la Salle, Notre vieux Sénégal (Paris: Challamel, 1909), 254. 55. “Une religion du terroir.” Thomas, Les Diola. “Terroir” should be understood in the sense of “land,” or “earth.” 56. See sous-série 1F 13 and sous-série 2G 20 36, ANS. 57. Lasnet et al., Une Mission au Sénégal, 157. 58. “Ces constructions . . . présentent une ornamentation tout orientale avec des colonnades superposées, des arcades en ogive difficilement attribuables à une population aussi primitive; la plupart sont plafonnées en argile de façon à ne pas redouter l’incendie, quelques-unes sont à l’étage avec escalier tournant en terre à l’intérieur.” Ibid. 59. “Les maisons, rectangulaires et solidement faites d’épaisses murailles en pisé, sont percées de portes étroites et de fenêtres munies de barreaux en bois. Au milieu du construction dont l’ensemble forme une sorte de croissant, se trouve une cour . . . où l’on garde des troupeaux. Les membres d’une seule famille habitent chacune de ces maisons . . . souvent composées de colonnes dont les chapiteaux rudimentaires s’adornent de mosaîques en coquillages.” de la Salle, Notre vieux Sénégal, 253. 60. For an excellent discussion of the eighteenth-century slave trade among the southern Floups, see Robert Baum, Shrines of the Slave Trade: Diola Religion and Society in Precolonial Senegambia (Oxford: Oxford University Press, 1999), Chapter 5. 61. The door frame, window frames, and roof frame are made from wood. 62. Interviews conducted in Mlomp, July 1, 1997. A similar account was told to Father Henri Goovers in the early 1950s; personal communication, April 25, 2001. 63. On Kasa historical traditions, see Baum, Shrines of the Slave Trade. 64. My interlocutor may also have believed me to be French. 65. See Mark, A Cultural, Economic and Religious History, 17–22. 66. The first Diola-speaking settlers in Buluf were preceded by Bagnunspeakers. 67. The founding of Ziguinchor is frequently dated to circa 1645. However, Bertrand-Bocandé, writing in 1849, convincingly argues for an earlier date. 68. Unfortunately, permission to photograph the well-preserved galleries on the north side could not be obtained from the building’s owner.

Notes to Pages 134–137

187

69. The Hôpital militaire has been dated variously to the early nineteenth century and to as late as 1870. Ousman Sane, “Urbanisation, Urbanisme et Architecture dans L’île de Gorée au XVIIIe et XIXe siècle” (Mémoire de maîtrise, Université de Dakar, 1972), 91, gives the latter date. However, the rounded arches would have been outmoded by 1870. Similar arcades are frequent in colonial military buildings as early as 1820. A. Grégoire’s assessment (“Monuments Historiques de L’île de Gorée et de St. Louis du Sénégal,” published by UNESCO in 1974, 21), “vraisemblablement du début du 19e siècle, avec double galeries ouvertes à arcades” [apparently from the beginning of the nineteenth century, with double rows of open and arcaded galleries], is more plausible on stylistic grounds. Jean Delcourt dates the hospital to the mid-1840s in La Turbulante Histoire de Gorée (Dakar: Claire Afrique, n.d.), 74. He may be correct. In 1863, the government undertook the enlargement of the hospital; see Batiments de Gorée, “Rapport sur le plan d’agrandissement de l’hôpital militaire,” sous-série P 112, ANS. It was decided “d’agrandir immédiatement l’hôpital de Gorée” [to immediately enlarge the hospital on Gorée]. An architectural plan accompanying this report shows the present structure, but with seven bays rather than the eight existing today. The extra bay was undoubtedly added as part of the ensuing expansion. Indeed, close inspection of the lower gallery reveals that the easternmost bay has a different floor from the rest of the veranda and that the level of the floor of this bay is raised by several inches. Any expansion would necessarily have occurred here, since the western exterior wall is flush with the seaside cliff. The expansion was necessitated by the rapid growth of Dakar. The original seven-bay structure probably dates from circa 1820–1845. 70. Buluf oral traditions record the presence of Jola visitors to Gorée in about 1860. In that year, a French military expedition against Thionk-Essyl took two young men hostage and brought them to Gorée. Interviews with elders of Kamanar ward, Thionk-Essyl, February 12, 1975. 71. “Quand le voyageur traverse le pays Diola, il est frappé d’admiration à la vue des belles cases, que l’indigène se construit pour y abriter sa famille. Ce n’est pas la petite case en paille, sombre et étroite, de certaines parties de l’Afrique; mais c’est un grand et large batisse en argile.” Père Wintz, Congrégation du St. Esprit, 1909, quoted in “Annales réligieuses,” 35, sous-série 2Z2, Microfilm 14 Mi 1610, CAOM. 72. Ziguinchor passed from Portuguese to French control, and the border between the two colonies was fixed close to its final configuration by means of the Franco-Portuguese Convention of May 12, 1886. The boundary between the colonies was to be set at half the distance between the Casamance and San Domingos Rivers. See “Relations avec la Guinée-Portugaise,” sous-série 2 F 4, ANS. 73. See, for example, “Rapport du ‘resident’ de Bignona,” sous-série 2 G 1 69, ANS. 74. In early-twentieth-century French colonial discourse, “race” does not have the connotation of the English word “race.” “Rapport politique 1e trimèstre,” 1902, sous-série 2 G 2 20, ANS. 75. Ibid. 76. “Que les Diolas avaient été en rapport avec les races soudaniennes du moyen Niger et avaient appris auprès d’elles l’architecture que l’on retrouve encore à Dienné et à Tombouctou.” Lasnet et al., Une Mission au Sénégal, 158. 77. “Le plus grand nombre, avec Reclus, attribuent ces constructions à l’influence portugaise qui se faisait sentir sur la côte des le XVe siècle.” Ibid. 78. See above, Chapter 2; and M. Adanson, A Voyage to Senegal, the Isle of Gorée, and the River Gambia (London: J. Nourse, 1759), 162. See also J.-B.-L. Durand, Voyage au Sénégal (Paris: Henri Agasse, 1802), 101.

188

Notes to Pages 138–141

79. The watercolor is entitled Kankétéfa, Résidence du Roi Bakar, Paquési. B22 2458, Bilder-Archiv, Frobenius-Institut, Goethe Universität, Frankfurt am Main. 80. “En voyant ces travaux de défense, qui ont reéllement quelque chose de la science européenne et que l’on retrouve dans le Bondou et le Bambouck, plusieurs voyageurs ont émi l’opinion que ces peuples les avaient sans doute construits d’après ce qu’ils avaient vu dans nos comptoirs. Quelques uns ont même prétendu que les Portugais avaient dû leur enseigner ces rudiments de la science. Frappé moi aussi par la regularité et la force de ces tatas . . . je demandais à plusieurs chefs pourquoi ils batissaient leurs murs à angles aigus . . . Tous furent unanimes à me répondre une chose bien naturelle et à laquelle je n’avais pas songé: c’est qu’il leur était impossible de faire autrement parce qu’un mur simple de cette hauteur et de cette longueur, construit en terre, ne présenterait aucune solidité.” Hyacinthe Hecquard, Voyage sur la Côte et dans l’Intérieur de l’Afrique Occidentale (Paris: Impr. de Bénard et compagnie, 1855), 205. 81. For a detailed study of this subject see J.-L. Triaud, “Le nom de Ghana: mémoire en exil, mémoire importée, mémoire appropriée,” in Histoire d’Afrique, les Enjeux de Mémoire, edited by Jean-Pierre Chrétien and Jean-Louis Triaud (Paris: Karthala, 1999), 240ff. See also Triaud, “‘Haut-Sénégal-Niger,’ un modèle ‘positiviste’? De la coutume à l’histoire: Maurice Delafosse et l’invention de l’histoire africaine,” in Maurice Delafosse, entre orientalisme et ethnographie, l’itinéraire d’un africaniste (1870–1926), edited by J.-L. Amselle and E. Sibeud (Paris: Maisonneuve et Larose, 1999), 210–232. 82. Triaud, “Le nom de Ghana,” 253–254. 83. Delafosse, Haut-Sénégal Niger, 494, cited in Triaud, “Le nom de Ghana,” 254; “Ce furent eux, sans doute, qui édifièrent les constructions en pierre de Koumbi et s’emparèrent du pouvoir.” [It was undoubtedly they who built the stone structures of Koumbi and who took over (political) power.] 84. “De véritables maisons qu’on s’étonne de voir édifier par elles.” de la Salle, Notre vieux Sénégal, 251–252. 85. “Si les Foups se font surtout remarquer par la relative beauté de leurs demeures, ils s’habillent par contre de la même façon sommaire que les autres Diolas.” Ibid., 253. 86. “On prétend qu’ils [les Portugais] furent . . . les maîtres en architecture des diolas. C’est peut-etre trop dire, mais les habitations des noirs lusitanisés semblent bien baties sur un modèle européen à peine transformé.” Ibid., 256. 87. See Triaud, “Le nom de Ghana,” 253–254. 88. “En prenant femme parmi leurs sujets de race noire.” Jean-Louis Triaud, “Le nom de Ghana,” 235–280.

Bibliography ARCHIVES Centre d’Archives d’Outre-Mer, Aix-en-Provence, France (CAOM) Afrique IV 25a Série Z 2Z2, Mi 1610, Microfilm 14, “Annales Religieuses de la Casamance” Archives Nationales du Sénégal, Dakar, Senegal (ANS) Sénégal et Dépendences Sous-série 1D: Affaires militaires 1D 50 Sous-série 1F: Affaires étrangères, Gambie 1F 5: “Relations avec la Gambie” (1778–1858) 1F 13 Sous-série 1G: Missions, Notices et Monographies 1G 23: E. Bertrand-Bocandé, “Voyage au pays de Kion” (1850) 1G 34 7: “Voyage dans le Sougrougrou” (January 23, 1861) 1 G 34 12: “Rapport de M. Faliu, Lt. Cdt. Sédhiou” 1G 34 19: “Voyage de M. Vallon, Lieutenant de vaisseau, délégué du Gouverneur en Casamance” (February 1862) 1G 154: “Mission Brosselard-Faidherbe, Guinée” (1890–1891) 1G 445: Ousmane Sane, “Urbanisation, Urbanisme et Architecture dans l’Isle de Gorée, aux XVIIIe et XIXe siècle” (mémoire de maîtrise, Université de Dakar, 1972) Sous-série 2F: Affaires étrangères, Guinée portugaise 2F 4: “Relations avec la Guinée-portugaise” (1879) 2F 4 13: Letter by officer on the French naval vessel Castor (February 2, 1879) Sous-série 2G: Rapports Périodiques 2G 1 69: “Rapport de la situation politique du Fogny, décembre 1899” 2G 2 20: “Bignona, Rapport politique, agricole et commercial, 2e trimestre 1902” 2G 20 36: “Rapports périodiques” Sous-série 11D: “Affaires militaires” 11D 1 282: “Rapport de Bertrand-Bocandé” (1857) Série P: Travaux Publiques P112: “Bâtiments de Gorée” (1833–1882) Bibliothèque Nationale, Paris, France (BN) Département des Cartes et Plans, SG Carton Bo-Bon: “Voyage à Joal, 1846” (Boilat) Série We: Photographs from collection of the Société de Géographie Côte S. H. M., portefeuille 111-2, piece 5/1 Côte Rés., Ge B 1964 (cartes 1 et 2) Centre des Archives Nationales, Paris (CARAN) Côte F 12 7208: “Rapport de la commission d’exploration du bas de la Côte” (1837) (Mission Dagorne) Frobenius-Institut, Goethe-Universität, Frankfurt am Main. Bilder-Archiv, Fonds Hecquard Thomas, Louis-Vincent. Personal papers and photographic collection. In author’s possession (available for consultation by researchers)

190

Bibliography PRIMARY AND SECONDARY SOURCES

Adanson, M. A Voyage to Senegal, the Isle of Gorée, and the River Gambia. London: J Nourse, 1759. Albergaria, Lopo Soares de. “Relação de Lopo Soares de Albergaria.” In Monumenta Missionaria Africana, edited by Antonio Brásio. Series 2. Vol. 4, Africa Ocidental, 1600–1622. Lisboa: Agência Geral do Ultramar, Divisão de Publicações e Biblioteca, 1968. Almada, André Alvares de. Tratado breve dos rios de Guine [Brief Treatise on the Rivers of Guinea]. Lisboa: Officina de Miguel Rodrigues, Impressor do Senhor Patriarca, 1733. ———. Tratado Breve des Rios de Guiné do Capo Verde [1594]. In Monumenta Missionaria Africana. Edited by Antonio Brásio. Series 2. Vol. 3: Africa Ocidental, 1570–1600, 229–378. Lisbon: Agência Geral do Ultramar, Divisão de Publicações e Biblioteca, 1964. Amaral, Ilidio. Santiago de Cabo Verde, a Terra e os Homens. Lisbon: Junta de Investigações Científicas do Ultramar, 1964. Amselle, Jean-Loup. Logiques métisses: Anthropologie de l’Identité en Afrique et ailleurs. Paris: Payot, 1990. Amselle, Jean-Loup, and Elikia M’Bokolo. Au coeur de l’ethnie: Ethnies, tribalisme et Etat en Afrique. Paris: La Découverte, Paris, 1985. Andrade, Francisco de. “Relação de Francisco de Andrade sobre as ilhas de Cabo Verde (1582).” In Monumenta Missionaria Africana. Edited by Antonio Brásio. Series 2. Vol. 3, Africa Ocidental, 1570–1600, 97–107. Lisbon: Agência Geral do Ultramar, Divisão de Publicações e Biblioteca, 1964. Anfreville de la Salle. Notre Vieux Sénégal. Paris: Challamel, 1913. Astley, Thomas. A New General Collection of Voyages and Travels. 4 vols. London: Printed for T. Astley, 1745. Afzelius, Adam, Sierra Leone Journal 1795–1796. Edited by Alexander Kup. Uppsala: Studia Ethnographica Upsaliensis, 1967. Barlaei, Casparis [Caspar Berlaeus]. Rerum per octennium in Brasilia et alibi nuper gestarum. Amsterdam: Blaeu, 1647. Barry, Boubacar. La Sénégambie du XVe au XIXe Siècle. Paris: L’Harmattan, 1988. Barth, Fredrik. “Ethnic Groups and Boundaries.” In Process and Form in Social Life: Selected Essays of Fredrik Barth. Reprint, London: Routledge & Kegan Paul, 1981. Baum, Robert. Shrines of the Slave Trade: Diola Religion and Society in Precolonial Senegambia. Oxford: Oxford University Press, 1999. Berlin, Ira. Many Thousands Gone: The First Two Centuries of Slavery in North America. Cambridge, Mass.: Harvard University Press, 1998. Bertrand-Bocandé, Emmanuel. “Notes sur la Guinée portugaise ou Sénégambie méridionale.” Bulletin de la Société de Géographie XII, no. 67 (1849): 68. Bertrand-Bocandé, Jean, G. Debien, and Y. Saint-Martin. “Emmanuel BertrandBocandé (1812–1881), un Nantais en Casamance.” Bulletin de l’Institut Fondamental d’Afrique Noire XXXI, no. 1 (1969). Blier, Suzanne. “Imaging Otherness in Ivory: African Portrayals of the Portuguese ca. 1492.” Art Bulletin (September 1993): 375–396. Bocandé, G. F. Joachim-Arnaud. “Biographie de la famille Bocandé.” Unpublished seminar paper. Paris: Université Paris I, Centre de Recherches Africaines, 1991.

Bibliography

191

Boilat, Abbé P.-D. Esquisses Sénégalaises. Paris: P. Bertrand, 1853. Reprint, Paris: Karthala, 1984. Boulègue, Jean. “Contribution des Sources françaises à la connaissance de l’actuelle Guinée-Bissau à la fin du xviie siécle.” History in Africa 28 (2001): 43–51. ———. Les Luso-Africains de Sénégambie, XVIe–XIXe siècle. Paris: Université de Paris I, Centre de Recherches Africaines, 1989. ———. “Présence européenne, Royaumes sénégalais et sociétés métisses du XVIème au XVIIIème Siècle.” Paper presented at conference on Connaissance des Marines Occidentales, Rochefort, 26–31 mai, 1986. ———. “Réorientation atlantique et politique extérieure des royaumes sénégambiens (fin XVe–XVIe siècles).” II RIHA (1996): 89–94. Bourgeois, J.-L. Spectacular Vernacular. 1983. Reprint, New York: Aperture Books, 1999. Boxer, Charles R. The Church Militant and Iberian Expansion 1440–1770. Baltimore: Johns Hopkins University Press, 1978. Brooks, George. “A Nhara of the Guinea-Bissau Region: Mae Aurelia Correia.” In Women and Slavery in Africa, edited by Claire Robertson and Martin Klein. Madison: University of Wisconsin Press, 1983. ———. Landlords and Strangers: Ecology, Society and Trade in Western Africa, 1000–1630. Boulder, Colo.: Westview Press, 1993. ———. “The Observance of All Souls’ Day in the Guinea-Bissau Region: A Christian Holiday, an African Harvest Festival, or an African New Year’s Celebration?” Africa Symposium, Ohio State University, 1982. ———. “Perspectives on Luso-African Commerce and Settlement in the Gambia and Guinea-Bissau Regions.” Boston University African Studies Center, Working Papers, 1980. Bühnen, Stephan, “Ethnic Origins of Peruvian Slaves (1584–1650).” Paideuma: Mitteilungen für Kulturkunde 39 (1993): 57–110. ———. “Place Names as an Historical Source: An Introduction with Examples from Southern Senegambia and Germany.” History in Africa 19 (1992): 45– 101. Caillié, René. Journal d’un voyage à Temboctou et à Jenné, dans l’Afrique Centrale. 3 vols. 1830. Reprint, Paris: La Découverte, 1996. Carreira, Antonio, ed. Descrições oitocentistas das Ilhas de Cabo Verde. Lisbon: Grafica Europam, 1987. Chrétien, J.-P., and Jean-Louis Triaud, eds. Histoire de l’Afrique: Les Enjeux de Mémoire. Paris: Karthala, 1999. Coquery-Vidrovitch, Catherine. Histoire des Villes d’Afrique Noire, des Origines à la colonisation. Paris. Albin Michel, 1993. ———. “Villes africaines anciennes: une civilisation mercantile prénègrière dans l’ouest africain, XVI et XVII siècles.” Annales Économies, Sociétés, Civilisations, no. 6 (novembre–decembre 1991): 1389–1410. Corry, Joseph. Observations upon the Windward Coast of Africa. London: G. and W. Nicol, 1807. Crowley, Eve. “Contracts with the Spirits: Religion, Asylum and Ethnic Identity in the Cacheu Region of Guinea-Bissau.” Ph.D. dissertation, Yale University, 1990. ———. “Institutions, Identities and the Incorporation of Immigrants within Local Frontiers of the Upper Guinea Coast.” In Migrations anciennes et peuplement actuel des Côtes guinéennes, edited by Gérald Gaillard. Paris: L’Harmattan, 2000.

192

Bibliography

De Bruijn, Mirjam, and Han van Dijk, eds. “Introduction: Changing Frontiers.” Mande Studies 1 (1999): 9–12. ———. Peuls et Mandingues: dialectique des constructions identitaires. Paris: Karthala, 1997. Demanet, Abbé. Nouvelle histoire de l’Afrique française. Paris, 1767. De Jong, Ferdinand. Modern Secrets: The Performance of Locality in Casamance, Senegal. Ph.D. diss., University of Amsterdam, 2001. Diawara, Mamadou. “Contribution to the Study of Social Differentiation in the Jaara Kingdom.” History in Africa 22 (1994): 123–139. Diouf, Mamadou. “Assimilation coloniale et identité religieuse de la civilité des originaires des Quatre Communes (Sénégal).” In AOF, Réalités et héritages: Sociétés ouest-africaines et ordre colonial, 1895–1960, edited by Charles Becker, Saliou Mbaye, and Ibrahima Thioub. 2 vols. Dakar: Direction des Archives du Sénégal, 1997, 2: 837–850. ———. Le Kajoor au XIXe siècle, pouvoir ceddo et conquête coloniale. Paris: Karthala, 1990. Donelha, André. An Account of Sierra Leone and the Rivers of Cape Verde (1625). Portuguese introduction and notes by A. Teixeira da Mota. English translation and notes by P. E. H. Hair. Lisbon: Junta de Investigações Científicas do Ultramar, 1977. ———. Déscription de la Serra Leoa et des Rios de Guiné du Cabo Verde (1625). Edited by A. Teixeira da Mota. French translation by Léon Bourdon. Lisbon: Junta de Investigações Científicas do Ultramar, 1977. Durand, Jean Baptiste. Voyage au Sénégal. Paris: Henri Agasse, 1802. Edwards, Jay. “The Complex Origins of the American Domestic Piazza-VerandaGallery.” Material Culture 21, no. 2 (Summer 1989): 3–56. ———. “The Origins of Creole Architecture.” Winterthur Portfolio 29, no. 2–3 (1994): 155–189. Fardon, Richard. “Review Article: ‘Métissage’ or Curate’s Egg.” (Review of JeanLoup Amselle, Mestizo Logics) Africa 70, no. 1 (2000): 144–151. Fausto, Boris. História do Brasil. São Paulo: Editora da Universidade de São Paulo, 1994. Fernandes, Valentim. Déscription de la Côte occidentale d’Afrique (Sénégal au Cap de Monte, Archipels). Translated by R. Mauny, Th. Monod, and A. Teixera da Mota. Bissau: Centro de Estudos da Guiné Portuguesa, 1951. Fleckner, Uwe. “Die Erfahrung der Fremde: Albert Eckhouts und Frans Posts Brasilienreise (1636–1644) und ihre Gestaltung in Porträt und Landschaft.” In Reisen des Barock: Selbst- und Fremdenerfahrungen und ihre Darstellung, edited by Regina Pleithner. Bonn: Romanistischer Verlag, 1991. Fyfe, Christopher, ed. Anna Maria Falconbridge: Narrative of Two Voyages to the River Sierra Leone. Liverpool: Liverpool University Press, 2000. Gaillard, Gérald, ed. Migrations anciennes et peuplement actuel des côtes guinéennes. Paris: L’Harmattan, 2000. Goerg, Odile. Pouvoir Colonial, Municipalités et Espaces Urbaines: ConakryFreetown des années 1880–1914. Paris: L’Harmattan, 1997. Gonsalves de Mello, José Antonio, ed. Fontes para a História do Brasil Holandes. Recife: Parque Historico Nacional dos Guararapes, 1981. Gruzinski, Serge. La Pensée Métisse. Paris: Fayard, 1999. Hair, P. E. H. Africa Encountered: European Contacts and Evidence, 1450–1700. Hampshire, Great Britain and Brookfield, Vt.: Variorum, 1997.

Bibliography

193

———. “An Accomplished Traveller Will Take Draughts: Barbot’s Illustrations of Guinea.” In Hakluyt Society Annual Report for 1991, 12–19. London: Hakluyt Society, 1991. ———. “Some References to Bolama, 1450–1870.” In Bolama entre a Generosidade da Natureza e a Cobiça dos Homens, edited by Carlos Cardoso. Bissau: INEP, 1996. ———. “Sources on Early Sierra Leone (4): Ruiters (1623).” Africana Research Bulletin V, no. 3 (April 1975): 52–70. ———. “Sources on Early Sierra Leone (5): Barreira.” Africana Research Bulletin V, no. 4 (July 1975): 81–118. ———. “Sources on Early Sierra Leone (7): Barreira, Letter of 9.3.1607.” Africana Research Bulletin VI, no. 2 (January 1976): 45–70. ———. “Sources on Early Sierra Leone (17): Alvares at Mitombo, 1611.” Africana Research Bulletin XI, no. 1–2 (March 1981): 92–140. ———, trans. “An Interim and Makeshift Edition of André Alvares de Almada’s Brief Treatise on the Rivers of Guinea. . . .” Translated and introduction by P. E. H. Hair. Organized by A. Teixeira da Mota. University of Liverpool, Department of History: Privately issued by Prof. Hair, 1984. Hair, P. E. H., Adam Jones, and Robin Law, eds. Barbot on Guinea: The Writings of Jean Barbot on West Africa, 1678–1712. 2 vols. London: Hakluyt Society, 1992. Hawthorne, Walter. “The Interior Past of an Acephalous Society: Institutional Change among the Balanta of Guinea-Bissau, c. 1400–1950.” Ph.D. dissertation, Stanford University, 1998. Hecquard, Hyacinthe. “Rapport sur un Voyage dans la Casamance.” Revue Coloniale (May 1852). ———. Voyage sur la Côte et dans l’Intérieur de l’Afrique Occidentale. Paris: Impr. de Bénard et compagnie, 1855. Horta, José da Silva. “Entre história europeia e história africana, um objecto de charneira: As representações.” In Actas do Colóquio, Construção e ensino da história de Africa. Lisbon: Gulbenkian Foundation, 1994, 189–200. ———. “Evidence for a Luso-African Identity in ‘Portuguese’ Accounts on ‘Guinea of Cape Verde’ (Sixteenth–Seventeenth Centuries).” History in Africa 27 (2000): 99–130. ———. “Portuguese Accounts of the Guiné do Cabo Verde and Luso-African Identity (16th–17th Century).” Paper presented at the Mande Studies Association Triennial Meeting, Banjul, The Gambia, June 1998. ———. “A represantação do Africano no literatura de viagens do Senegal à Serra Leoa (1453–1508).” Mare Liberum no. 2 (1991): 209–338. Howard, Allen. “Mande and Fulbe Interaction and Identity in Northwestern Sierra Leone, Late 18th through Early 19th Centuries.” Mande Studies 1 (1999): 13–39. Jewsiewicki, Bogumil. “Race et éthnie, acteurs politiques virtuels de cette fin de siècle.” In Les convergences culturelles dans les sociétés pluriéthniques, edited by K. Fall, R. Hadj-Moussa, and D. Simeoni. Sainte Foy: Presses Universitaires de Québec, 1996. Jobson, Richard. The Golden Trade, or a Discovery of the River Gambra. London: Nicholas Okes, 1623. Jones, Adam. “Olfert Dapper et sa déscription de l’Afrique.” In Objets Interdits. Paris: Fondation Dapper, 1989.

194

Bibliography

Kandé, Sylvie. Terres, Urbanisme et Architecture “Créoles” en Sierra Leone XVIIIe– XIXe Siècles. Paris: L’Harmattan, 1998. Kellein, Thomas, and Urs-Beat Frei, eds. Frans Post, 1612–1680. Basel: Kunsthalle Basel, 1990. Klein, Martin. “Ethnic Pluralism and Homogeneity in the Western Sudan: Saalum, Segu, Wasulu.” Mande Studies 1 (1999): 109–124. ———. Islam and Imperialism in Senegal, Sine-Saloum, 1847–1914. Stanford: Stanford University for the Hoover Institute on War, Revolution, and Peace, 1968. ———. Slavery and Colonial Rule in French West Africa. Cambridge: Cambridge University Press, 1998. Labat, Jean-Baptiste. Nouvelle Relation de l’Afrique Occidentale. 5 vols. Paris: Cavelier, 1728. La Courbe, Sieur Michel Jajolet de. Premier voyage du Sieur de la Courbe faite à la coste de l’Afrique en 1685. Edited by Prosper Cultru. Paris: E. Champion, 1913. Lasnet, E., et al. Une Mission au Sénégal. Paris: Challamel, 1900. Launay, R. “The Dieli of Korhogo.” In Status and Identity in West Africa: Nyamakalaw of Mande, edited by David Conrad and Barbara Frank. Bloomington: Indiana University Press, 1995. Le Maire, Jacques. Les Voyages du Sieur Le Maire aux Isles Canaries, Cap-Verd, Sénégal et Gambie. Paris: J. Callombat, 1695. Lemos Coelho, Francisco de. Description of the Coast of Guinea (1684). Introduction and English translation by P. E. H. Hair. University of Liverpool, Department of History: Privately issued by Prof. Hair, 1985. Lemmens, Gerard. “Die Schenkung an Ludwig XIV und die Auflösung der brazilianische Sammlung des Johann Moritz 1652–1679.” In Soweit der Erdkreis Reicht, Johann Moritz von Nassau-Siegen, 1604–1679. Kleve: Städtisches Museum Haus Koekkoek, 1980. Lentz, Carola. Die Konstruktion von Ethnizität: Eine politische Geschichte Nord-West Ghanas, 1870–1990. Köln: Rüdiger Koppe Verlag, 1998. Lespinay, Charles de. “Baynunk et Beafad dans l’ensemble guinéen: parentés et alliances anciennes (15e–20e siécle).” In Bolama entre a generosidade da natureza e a cobiça dos homens, edited by Carlos Cardoso. Bissau: Instituto Nacional de Estudos e Pesquisa, 1990. ———. “Territoires et droits en Afrique noire: Essai d’anthropologie du droit foncier en milieu rural.” Thèse de doctorat. Université de Paris-X, 1999. Leven, Yves Goasguen. “Architecture coloniale à l’ile de la Réunion.” Thèse de doctorat, Université Lyon II, 1997. Lobban, Richard. Historical Dictionary of the Republics of Guinea-Bissau and Cape Verde. London: Scarecrow Press, 1979. Madeira Santos, Maria Emilia. “Origem e desenvolvimento da colonização: Os primeiros ‘lançados’ na costa da Guiné: Aventureiros e comerciantes.” In Portugal no Mundo, edited by Luis de Albuquerque. Vol. 2. Lisboa: Publicações Alfa, 1988. ———, ed. História Geral de Cabo Verde. Vol. 2. Lisbon: Centro de Estudos de História e Cartografia Antiga, Instituto de Investigação Científica Tropical; and Praia [Cape Verde]: Instituto Nacional da Cultura de Cabo Verde, 1995. Malaquais, Dominique. “Burning Down the House: Architecture and Resistance in the Bamileke Highlands, 1957–1971.” English trans. from Architecture, pouvoir et dissidence au Cameroun. Paris: Karthala, 2002.

Bibliography

195

———. “Constructing Power: Architecture, Landscape and the Design of Space in a Chieftaincy of the Bamileke Highlands.” Ph.D. dissertation, Columbia University, 1997. ———. “You Are What You Build: Architecture as Identity among the Bamileke of West Cameroon.” Traditional Dwellings and Settlements Review 5, no. 2 (Spring 1994): 21–35. Marcgraf, Georg. Brasiliae Geographica & Hydrographica Tabula nova . . . 1643. Amsterdam: Johan Blaeu, 1646. Mark, Peter. Africans in European Eyes: The Portrayal of Black Africans in Fourteenth and Fifteenth Century Europe. Syracuse: Maxwell School of Citizenship and Public Affairs, Syracuse University, 1974. ———. “Constructing Identity: Sixteenth and Seventeenth Century Architecture in the Gambia-Geba Region and the Articulation of Luso-African Ethnicity.” History in Africa 22 (1995): 307–327. ———. A Cultural, Economic and Religious History of the Basse Casamance Since 1500. Stuttgart: Steiner Verlag/Frobenius-Institut, 1985. ———. “The Evolution of ‘Portuguese’ Identity: Luso-Africans on the Upper Guinea Coast from the 16th to the Early 19th Century.” Journal of African History 40, no. 2 (June 1999): 173–191. ———. “Hyacinthe Hecquard’s Drawings and Watercolors from Grand Bassam . . . A Source for Mid-Nineteenth Century West African History.” Paideuma: Mitteilungen zur Kulturkunde 36 (1990): 173–180. ———. “‘Portuguese’ Architecture and Luso-African Identity in Senegambia and Guinea, 1730–1890.” History in Africa 23 (1996): 179–196. ———. The Wild Bull and the Sacred Forest: Form, Meaning, and Change in Senegambian Initiation Masks. New York: Cambridge University Press, 1992. Mathews, John. A Voyage to the River Sierra Leone. London: B. White, 1788. McAlister, L. Spain and Portugal in the New World. Minneapolis: University of Minneapolis Press, 1984. Middleton, John. The World of the Swahili. New Haven, Conn.: Yale University Press, 1993. Mollien, Gaspard. Travels in the Interior of Africa to the Sources of the Senegal and Gambia. English translation. Edited by T. E. Bowditch. London: H. Colburn & Co., 1820. ———. Voyage à l’Intérieur de l’Afrique aux Sources du Sénégal et de la Gambie. Paris: Imprimerie de Mme Veuve Courcier, 1820. Moore, Francis. Travels into the Inland Parts of Africa. London: Printed by E. Cave for the author, 1738. Moraes, Nize Isabel de. A la découverte de la Petite Côte au XVIIe siècle: Sénégal et Gambie. 3 vols. Dakar: Université Cheikh Anta Diop-IFAN, 1993–. Park, Mungo. Travels in the Interior Districts of Africa. London: W. Bulmer, 1799. Passfield, Oliver, ed. The Voyages of Sieur D. B. to the Islands Dauphine or Madagascar and Bourbon or Mascarene in the Years 1669, 70, 71, and 72. London: Printed for D. Nutt by T. and A. Constable, 1897. Pélissier, René. Naissance de la Guiné: Portugais et Africains en Sénégambie (1841–1936). Orgival, France: Pélissier, 1989. Postma, Johannes. The Dutch in the Atlantic Slave Trade 1600–1815. Cambridge: Cambridge University Press, 1990. Prussin, Labelle. Hatumere: Islamic Design in West Africa. Berkeley: University of California Press, 1986.

196

Bibliography

Raffenel, Anne. Nouveau Voyage dans le pays des Nègres. Paris: Impr. de. N. Chaix et cie, 1856. Roberts, George. The Four Years Voyages of Captain George Roberts; Being a Series of Events which Befell Him in a Voyage to the Islands of the Canaries, Cape de Verde, and Barbadoes. . . London: A. Bettesworth, 1726. Roche, Christian. Conquête et Résistances des Peuples de Casamance, 1850– 1920. Dakar: Nouvelles Éditions Africaines, 1976. Ruiters, Dierck. Toortse der Zeevaert [The Torch of Sea Travel]. Translated by P. E. H. Hair in “Sources for Early Sierra Leone: (2) Andrade (1582), Ruiters (1623), Carvalho (1632).” Africana Research Bulletin 5, no. 1 (October 1974): 47–56. Sackur, Amanda. “Religion and the Construction of Creole Identity in Precolonial Senegambia.” Paper presented at the African Studies Association of the United Kingdom, September 1992. Sackur, (Karen) Amanda. “The Development of Creole Society and Culture in Saint-Louis and Gorée, 1719–1817.” Ph.D. thesis, School of Oriental and African Studies, London, 1999. Sane, Ousman. “Urbanisation, urbanisme et architecture dans l’Isle de Gorée aux XVIIIe et XIXe siècle.” Mémoire de maîtrise, Université de Dakar, 1972. Schefer, Christian, ed. Instructions Générales données de 1763 à 1870 aux gouverneurs et ordonnateurs des établissements français en Afrique. Paris: E. Champion, 1927. Senna, Manuel Roiz Lucas. Dissertação sobre as Ilhas de Cabo Verde 1818. Edited by A. Carreira. Reprint; Lisbon: Gráfica Europam, Mems Martin, 1987. Silva, Ernani Bruno. História do Brasil, geral e regional. Vol. 2. São Paulo: Editora Cultrix, 1967. Soulillou, Jacques, ed. Rives Coloniales, Architectures de Saint-Louis à Douala. Paris: ORSTOM/Parenthèses, 1993. Sousa-Leao, Joaquim de. Frans Post, 1612–1680. Amsterdam: A. L. van Gendt, 1973. Southeby’s. Old Masters 2000 (sale of January 28, 2000). New York: Sotheby’s, 2000. Teixeira, Manuel. “Portuguese Traditional Settlements, a Result of Cultural Miscegenation.” Traditional Dwellings and Settlements Review 1, no. 2 (1990). Teixeira da Mota, Avelino, ed. As Viagens do Bispo D. Frei Vitoriano Portuense à Guiné e a cristianização dos reis de Bissau. Lisboa: Junta de Investigações Científicas do Ultramar, 1974. Teixeira da Mota, Avelino, and Mário G. Ventim Neves, eds. A Habitacão Indigena na Guiné Portuguesa. Bissau: Centro de Estudos de Guiné Portuguesa, 1948. Teixeira da Mota, A., and R. Mauny. “Le voyage du navire ‘Santiago’ à la Sierra Leone et à la Rivière de S. Domingos (1526).” Bulletin de l’Institut Fondamental d’Afrique Noire 37, no. 3 (1975). Thilmans, G., and Nize Isabel de Moraes. “La déscription de la côte de Guinée du Père Baltasar Barreira (1606).” Bulletin de l’Institut Fondamental d’Afrique Noire 34, no. 1 (1972): 1–50. ———. “Le routier de la côte de Guinée de Francisco Pirez de Carvalho (1635).” Bulletin de l’Institut Fondamental d’Afrique Noire 32, no. 2 (1970): 343– 369. Thomas, Louis-Vincent. Les Diola: Essai d’analyse fonctionelle sur une population de Basse-Casamance. Dakar: IFAN, 1959.

Bibliography

197

Triaud, Jean-Louis. “‘Haut-Sénégal-Niger,’ un modèle positiviste? De la coutume à l’histoire: Maurice Delafosse et l’invention de l’histoire africaine.” In Maurice Delafosse, entre orientalisme et ethnographie, l’intinéraire d’un africaniste (1870–1926), edited by J.-L. Amselle and E. Sibeud. Paris: Maisonneuve et Larose, 1999. van der Klei, Jos. Trekarbeid en de Roep van het Heilige Bos. Nijmegen: IKEN, 1989. Verger, Pierre. Trade Relations between the Bight of Benin and Bahia from the 17th to the 19th Century. Translated by Evelyn Crawford. Ibadan, Nigeria: Ibadan University Press, 1976. Villeneuve, Geoffroy de. L’Afrique ou Histoire, moeurs, usages et coutumes des Africains: Le Sénégal. 4 vols. Paris: Nepveu, 1814. Webb, James. Desert Frontier: Ecological and Economic Change along the Western Sahel (1600–1850). Madison: University of Wisconsin Press, 1995. Whitehead, P. J. P., and M. Boeseman. A Portrait of Dutch 17th Century Brazil: Animals, Plants and People by the Artists of Johan Maurits of Nassau. Amsterdam: North Holland Publishing Company, 1989. Wright, Donald. The World and a Very Small Place in Africa. Armonk, N.Y.: M. E. Sharpe, 1997.

Index Page numbers in italics refer to illustrations. Accra, 62 Adanson, M., 57–58 Adeane, 106 adobe (banco), 127–28, 130, 135 in African architecture, 47–48, 50, 51, 52, 57, 60, 111, 113, 115, 117, 118, 121, 122, 127–29, 130, 135 in colonial architecture, 46, 50–52, 60, 133–34 in Floup architecture, 35 and mixed cultures, 20 and “Portuguese”-style architecture, 36, 44, 46, 49, 59, 78–79, 98 African architecture ekumbane houses, 122–23 French views of, 98, 100–13, 123, 134–43 forts, 37, 45, 138, 139, 158–59n12, 161n40 labyrinths, 37, 55–56, 111, 116–17, 119–20, 159n17 multistory architecture, 51, 57, 74, 138, 127–34, 137 palisades, 35–39, 45, 99–100, 115–22, 159n19, 160n31, 182n55, 183n18 ramparts, 117, 120, 138, 183n18 round, 20, 38, 48–51, 61, 116–17, 120, 123, 135 in Sudan, 52 African housing compounds, 112, 135, 138 dispersed, 37, 115, 120–21 extended-family (fankaf), 115–17, 122, 184n32 fortified, 35–38, 45, 55–56, 98, 100, 103, 111, 115–17, 119–22, 128, 138–39 African kings, 177n67 and Creole language, 15, 150n11 and cultural mixing, 19, 84–85, 87, 90–92, 112 and defensive architecture, 35, 38, 55, 119–20, 138 and interethnic warfare, 162n53 and “Portuguese”-style architecture, 44, 54–56, 72, 91 African religious practices of Floups, 16–7, 30–31 and impact of Christianity, 87 of Manding, 86, 173n30, 182n55 of Serers, 157n111

See also amulets; bubenben (bufókub); charms; chinas (shrines); Christianity; gris-gris; idolâtres; Islam; masks, religious; Muslims; shrines; witchcraft trials African resistance to European colonization, 34, 55, 136, 141 Afzelius, Adam, 27, 105 Albergaria, Soares de, 15 Albreda, 83, 102, 109, 132 center of slave trade, 84, 172n13 French lose control of, 103 mixed cultures at, 82, 89 women traders at, 89 Almada, André Alvares de, 8, 14–5, 26, 36–39, 120, 155n91 alpainters, 17, 53–54, 152n37 Alves, Manuel, 89 Amaral, Ilidio, 64 amber, 45 amulets, 30, 87, 110, 174n35. See also bubenben (bufókub); charms; gris-gris Andrade, Francisco de, 15, 18 Angola, 65, 77, 168n22 arches, rounded, 127, 130, 132–34 Arriatas, 36–37 Astley, Thomas, 25, 59 Atlantic slave trade, 35, 56, 74, 116, 119, 165n92 Badibu, 173n25 Bagas, 153n45 Bagnuns (Bainunks) architecture of, 44 and domestic slavery, 124 ethnic identity of, 6, 86 internal migration of, 114 political power of, 1, 83–84, 124 religious rituals of, 16, 84–85 territory of, 37, 81, 86, 94, 176–77n65 as slave traders, 36, 162n53 social organization of, 4, 34, 176–77n65 as traders, 15, 159n15 as victims of slave raids, 165n92 Bahia (Brazil), 65, 70 Bainunks. See Bagnuns Baiot (Bayottes), 112

200

Index

Balantas architecture of, 35, 135, 160n31 intercultural marriage and, 22 social organization of, 4 trade with Luso-Africans, 100 as warriors, 107 Bamana, 108 Bambuk, 138–39 banco. See adobe Baol, state of, 1 Barbot, Jean, 11, 43, 61 Barlaeus, Caspar, 68 Barreira, Father Balthasar, 15–7 Barth, Fredrick, 5, 22–23 Bathurst (Banjul), 3, 132 Bayottes. See Baiot Beafadas, 37, 42, 160n20, 165n92 beeswax, 45–46, 84, 88, 98, 150n6, 159n16 bekin (shrine), 152n29. See also chinas (shrines); shrines Bertrand-Bocandé, Emmanuel marriage to Luso-African woman, 106 observer of Floup culture, 34, 114–15 observer of Luso-African culture, 100–101 résident of Carabane, 98, 114–17, 120, 122, 127, 157n113, 180n36, 182n1 bidan identity, 26–27, 108. See also white identity Bignona, 126, 136, 182n6 Bijogos, 4, 37, 56, 105, 159nn17,18 Bissagos Islands, 2–3, 37, 39, 56–57, 87, 105 Bissau, 2–3, 9 architecture at, 21, 39, 49, 166n2 French at, 109 king of, 55 Luso-Africans at, 24, 89, 97–98, 100, 104 missionaries at, 174nn37,42 mixed cultures at, 21, 88, 94 Portuguese fort at, 103 bixiiris (bixirins), 52, 153n49 black identity architecture and, 140–41 and religion, 57, 88, 104 occupation and, 89, 103–104, 107 “Portuguese” and, 24–25, 27–28 Blaeu, Johan, 68 Bocandé. See Bertrand-Bocandé, Emmanuel Bocandé, M., 106

Boilat, Abbé, 28–30, 104 Bolole (Floup village), 17, 55, 131, 152n34 Boufflers, Chevalier de, 109 Brazil and “Portuguese”-style architecture, 67–73 and slave architecture, 66, 69–70, 74–78 slavery in, 65, 70, 74 bubenben (bufókub), 30. See also amulets; charms; gris-gris Buluf (Diola region), 6, 173n22 architecture of, 60, 115–16 intervillage alliances of, 120 intervillage warfare in, 121 and migration, 85–86, 124, 158n1 oral traditions of, 85–86, 130, 187nn69,70 and slave raiding, 122–23 and spread of Islam, 173n30 Buluf-Diola, 94, 130 Bundu, 138–39 Burofai, 173n25 Cacheu, 2–3, 4, 8, 88, 162n48, 169n40 Luso-African trade at, 14, 24, 42, 98, 100 and mixed cultures, 22, 55, 82–83 Portuguese fort at, 55, 131 “Portuguese”-style architecture at, 34, 39, 42, 49, 166n2 Sapes at, 19 slave trade at, 65, 168n22 Caillié, René, 50, 108 Calheta, 64 Cape Verde Islands architecture of, 62–64, 73, 76, 131 and cultural mixing, 18–20, 24, 26, 28 Jesuit missionaries at, 151n25 Luso-Africans at, 8, 14, 18, 24, 36 and slave trade, 65, 70, 158n6, 168n22 and trade with Portugal, 14, 93 Carabane, 118, 158n2, 161n41 cultural mixing at, 117 French trading post, 11, 34, 97–98, 100, 106, 114, 124, 157n113 Caribbean, 23, 170n48, 171n1 Casão, 172n9 architecture at, 20, 34, 38, 52, 55, 103 Manding community, 51–52, 83, 120 Portuguese trading center, 3, 19–20

Index Catholicism, 16–17, 25–26, 86–87, 103, 125. See also missionaries cattle, 100 Cazegut, Island of, 56, 87, 105, 107 chapels, 67, 71, 76, 87, 100, 171n57, 178n12. See also churches charms, 30–31, 87. See also amulets; bubenben (bufókub); gris-gris Chesapeake, 23 children products of informal unions, 175n55 as victims of slave raids, 125, 183n11, 185nn47,48, 185–86n49 chinas (shrines), 16–17, 152n29. See also bekin (shrine); shrines Christianity conversion to by Africans, 15, 18, 20, 26, 57, 85, 87–89, 104, 156n103 and cultural identity, 7, 17, 22, 31, 87, 104, 153nn56,59, 174n38 Europeans in African communities, 18. See also tangomaos and names of Africans, 17, 24, 87, 165n94 and religious diversity in Senegambia, 86, 153nn56,59, 156nn102,103, 157n115, 165n94 churches, 39, 42–47, 49, 71, 87, 100, 131, 163n75, 164n77. See also chapels colonnades, 127, 130 colonizers in Africa. See Dutch colonizers in Africa; English colonizers in Africa; French colonizers in Africa; Portuguese colonizers in Africa Combo Sylla (Fodé Sylla), 116, 122–23, 125–26, 185n39 Corry, Joseph, 62–63 cotton, 98 cotton cloth, 100–101 Creole architecture, 66–67, 164n86, 170nn46,48 language (Crioulo), 15–6, 25, 28, 34, 104, 114, 150n11, 156n102, 158n3 society, 14, 23, 29, 63, 81–82, 104, 171n1, 175n50 See also mixed cultures cumbetes (storehouses), 47, 55, 99 Dagorne, 97–98, 100, 103, 110–11, 113 Dahomey, 73

201

Dakar, 3, 112, 125, 133 Dappert, Olfert, 62, 167n10 Delafosse, Maurice, 139, 140–43 Diatta, Kepi, 115 Diolas architecture of, 115, 117, 120, 122–23, 135, 137, 139–41 described, 122, 184n36 and domestic slavery, 124 etymology of Diola name, 183–84n28 oral traditions of, 86, 94 reactions to French colonization, 136 religious beliefs of, 126 as victims of slave raids, 125–26, 184n33 Diola-Fogny, 6, 123–27, 137, 162n57 Distant Landscape with Church Ruins, Chapel, and Farms (Frans Post), 71, 71–72 Djougoutes, 116, 121, 123–25, 185n39 Domingos (religious syncretism), 27 Donelha, André describes Luso-African life, 8, 18 describes Senegambian architecture, 20, 34, 51–53, 57, 120 describes Senegambian trade, 14, 150n6 and Gaspar Vaz, 19 Durand, Jean Baptiste, 56 Dutch colonizers in Africa Dutch West India Company, 66, 168n26 on the Gold Coast, 62 at Gorée Island, 13, 23–24 on the Petite Côte, 28 Dutch colonizers in Brazil, 65–67. See also Ruiters, Dierck dyula merchants. See Juula (dyula) merchants ekumbane houses, 122–23 Elana, 128 elephant teeth, 84 Elmina (Portuguese fort), 36–37, 42 English colonizers in Africa, 44, 62 borrow “Portuguese”-style architecture, 44 competition with other colonizers, 23, 97, 178n5 and end of slave raids in Fogny, 116, 126–27 at Freetown, 62

202

Index

at James Island, 13, 24, 83, 102 participation in slave trade, 65, 168n26 See also Astley, Thomas; Corry, Joseph Essyl (Esulalu), 130, 173n25, 177n67. See also Thionk-Essyl (Thionk) Esulalu-Jola, 177nn65,66 etiquette and African values, 90 as marker of ethnic identity, 81–82, 91–92 See also hospitality fankaf. See African housing compounds Farim, 2, 40, 88, 159n19 Fernandes, Valentim, 34–35, 119 filhos de terra, 14, 22 Floup village, map of, 41 Floups (Jolas), 8, 16–17, 37 architecture of, 38–39, 48, 116–17, 119–20, 134–35, 137, 140, 160n31 fluid identity of, 81, 85, 165n94 French views of, 109–11 internal migration of, 114, 177n67 religious practices of, 16–17, 30–31 and slave trade, 36, 84, 121, 159n15, 162n53, 165n92 social organization of, 4, 6 trade with Europeans, 34, 55 as warriors, 35–36, 39, 94, 110 Fodé Kaba, 116, 120, 122, 124–27, 183n12, 186n52 Fodé Sylla. See Combo Sylla Fogny described by La Courbe, 83 home of Floups, 6, 34 interethnic warfare in, 94, 116, 120–22, 124–26, 158n1, 186n52 Manding migration to, 85 and slave trade, 84, 120–22, 124–26, 186n52 social organization in, 38 Wolof traders in, 180n38 Fogny language, 181n46 Fogny region, depopulation of, 126 forkillas, 53, 60, 69, 99 forts in Senegambia African, 37, 45, 138, 158–59n12, 161n40 European, 13, 36–37, 42, 62, 102–103, 131, 133, 161n40 Franco-Portuguese Convention (1886), 187n72

French colonizers in Africa, 53, 83, 120 at Carabane, 114 competition with other colonizers, 23, 142 and destruction of African property, 140–41, 183n13 and end of slave raids in Fogny, 116, 125–27, 184n32, 185n45 at Gorée Island, 24, 102 participation in slave trade, 65, 168n26 resistance to by Africans, 100, 136, 179n23 and rigid identity categories, 6–7, 10–12, 25–26, 28, 97, 106, 108, 111, 144 at Saint-Louis, 102, 104 at Sedhiu, 110, 133 and trade with Luso-Africans, 100 at Ziguinchor, 11, 140 See also Barreira, Father Balthasar; Bertrand-Bocandé, Emmanuel; Boilat, Abbé; Caillié, René; Dagorne; Delafosse, Maurice; Hecquard, Hyacinthe; Labat, Father Jean-Baptiste; Mollien, Gaspard; Raffenel, Anne; Sène, Father; Villeneuve, Geoffroy de Fulbe, 4, 21, 153n55 Futa Jalon, the, 4, 11, 31, 98, 157n117 Gabon, 168n22 Gara (Sara), Manuel, 105, 107 Geba, 2 Georgetown, 132 Geregia decline of, 95, 101 and mixed cultures, 83–84, 89, 124 palisades at, 45 “Portuguese”-style housing at, 45 trading center, 2, 84, 124 Geregia-Fogny, 124 Ghana, 139–41 glass beads, 45 gold, 173n16 Gorée Island, 3 architecture at, 133–34, 187n69 attitudes toward race at, 104 controlled by British, 62 controlled by Dutch, 13 controlled by French, 24, 102, 109, 133 Creole society of, 175n50 lime exported to, 161n41 religious syncretism at, 87

Index shrines at, 30 Villeneuve visits, 180n42 gourmettes (grumetes), 45, 57, 88, 103–104, 134, 166n102, 174n38 gris-gris, 27, 110, 156n103 grummetes. See gourmettes Guinala, 88, 162n55, 169n40 Hecquard, Hyacinthe, 11, 31, 65, 98–100, 117, 138–39, 168n20 Holy Ghost Fathers, 15, 100, 117, 124–25 Hôpital militaire, Gorée Island, 133, 134, 187n69 hospitality, 44, 53, 81, 90–92, 176n56 identity formation nonstandard model of, 22–23 oppositional model of, 22–23, 26, 28 identity in Senegambia architecture and, 9–10, 17–18, 46, 54, 58, 103, 111, 141–43, 144, 170n48 and cultural mixing, 63 etiquette and, 90–92 fluidity of, 19–23, 85–87, 105, 141, 145–48, 154n74 Luso-African, 14–8, 23–32, 105, 175n47 marriage and, 22, 30, 89–90 political factors and, 97, 104, 106–107 profession and, 96, 105 religion and, 24, 27, 29–31, 88–89, 157n111 rigidification of categories of, 6–7, 10–12, 21, 24–26, 28, 31–32, 57–58, 97, 104, 106, 108, 111, 141, 144 See also black identity; mixed cultures; white identity idolâtres, 30, 165n94 impluvium houses, 48, 127–28 intercultural marriage Balantas and, 33 Cape Verdeans and, 19 Creoles and, 175n50 forbidden in Joal, 30 Luso-Africans and, 19, 81, 89–90, 106, 175n52 Manding and, 22 intercultural unions, 90, 175n55 internal migration in Africa of Bagnuns, 114 of Bulufs, 130

203

of Floups, 94, 114, 177n67 and fluid cultural identities, 22 of Mandes, 52, 85, 95 of Mandings, 173nn24,25 and maraboutic wars, 115 and migrant labor, 117, 124 of Serers, 109 and slave trade, 126, 186n52 and trade routes, 10, 18–9 intervillage alliances, 116, 120 iron, 30, 45. See also staffs of office, iron Islam presence in Senegambia, 1–2, 173n30, 174n35, 184n32 and religious conversion, 7, 122 and religious syncretism, 86, 157n115 and tangomaos, 19 See also bixiiris (bixirins); marabouts; mosques; Muslims ivory, 45, 84, 98, 150n6, 173n16 James Island English fort at, 13, 24, 83, 102 “Portuguese”-style housing at, 45 Jenné, 137, 164n79 Jenné-Jeno, 50, 51 Jesuits, 151n25 Jews, 16, 28, 85, 151n25. See also New Christians Joal, 2–3, 28–31, 156nn102,103,104, 157n111 Jobson, Richard, 24–25, 38, 120 Jolas architecture of, 46, 48, 60, 76, 111–12, 117, 121, 127–28, 135 ethnic identity of, 6, 30, 157n114, 157n1 language of, 72, 110, 114, 181n46, 182n1 migration of, 85, 187n70 naming practices of, 17, 165n94 oral traditions of, 94, 109, 177nn65,66 political organization of, 110–11 religious practices of, 16–7, 30, 86–87, 95, 152n29, 173n30, 174n35 as traders, 34, 97, 100–101, 107 See also Diolas; Floups (Jolas) Jolof, state of, 1 Jougoutes (Buluf), 114 Juffure, 2, 15, 151n20 Juula (dyula) merchants, 2, 4, 31, 84, 107, 157n115

204

Index

Kaabu (Cabo or Gaabu) and fluid cultural identity, 154n17 Mande state, 1, 4, 39, 84, 86, 91, 95, 176n59 Kaabu, Malamanka, 95, 177n67 Kaba, Fodé. See Fodé Kaba Kajoor (Cayor or Kaour), 1, 85, 112–13 kalol (geographically dispersed ward), 115–16 Kanbanba, 179n23 Kankétéfa, 138, 139 Kartiak, 116 Kasa, 85, 124, 128, 130, 137, 158n1, 173n25 kidnapping, 114, 124 King Philip II, 16, 151n25 kola products, 45, 100 Koumbi, 139, 188n83 Kumbi Saleh, 141 La Belinguere, 44, 89, 175n53 La Courbe, Michel Jajolet de maps by, 11, 38, 40–41, 159n19 observer of African architecture, 44 observer of cultural mixing, 83, 85, 90–92, 152n34, 158nn1,3, 165n94, 173n20, 174nn37,38, 175n55 observer of interethnic warfare, 94 observer of Luso-African architecture, 17, 39, 43, 54–55, 87–88, 162n55 observer of Portuguese architecture, 39 plagiarized by Father Labat, 44, 56, 82, 95, 161–62n48, 171nn2,3 racial beliefs of, 25, 87–89 La Salle, Anfreville de, 128, 139–42 Labat, Father Jean-Baptiste observer of African architecture, 38–39 observer of cultural mixing, 89 observer of trade, 95 plagiarization of La Courbe, 44, 56, 82, 95, 161–62n48, 171nn2,3 labyrinths. See African architecture, labyrinths Lacombe, Père, 100, 124–26 lançados, 13–16, 18–19, 22, 33, 36, 53–54, 159n16 Lasnet, Dr., 122–24, 126–28, 130, 137, 143 Le Maire, Jacques, 25, 85, 91

Lemos Coelho, Francisco de observer of African society, 85 observer of Luso-African society, 8, 24, 39 as slave owner, 42 as trader, 84, 149n6 liming, 161nn41,45. See also whitewashing maisons à la portugaise. See “Portuguese”-style architecture Mali, 4, 50 Mande architecture of, 20, 38, 52, 57–58, 119–20, 137 diaspora, extent of, 4, 31 as engineers, 138 fluid identity of, 5, 21 internal migration of, 94–95 heartland of, 31, 51, 176n59 political power of, 1, 176n59 Manding architecture of, 34, 46, 57, 111, 117, 120 and intercultural marriage, 22, 25, 81, 83 political power of, 83, 91, 100 religious practices of, 86, 173n30, 182n55 as slave owners, 122 as slave raiders, 36, 116, 122–25 settlement patterns, 38 as traders, 1, 84, 108, 153n53 as victims of slave raids, 165n92 Mané family, 86, 177n67 Mane Invasion (ca. 1550), 19 Manjaks, 75–76, 81 Mansa Falup, 35, 119 maraboutic wars, 115, 122, 184n33 marabouts, 15, 31, 115, 151n19, 174n35. See also Islam; Muslims Marcgraf, Georg, 68 Marcgraf Map, 68, 69, 74, 76 Marees, Pieter de, 15 Maroon architecture, 170n48. See also Caribbean Marsassoum (slave market), 3, 125–26. See also slave trade in Senegambia masks, religious, 86–87, 174n34 Mathews, John, 25, 27 Maurel & Prom, 106, 132, 133 Maures, 27, 108 Maurits, Johan, 66–67, 168n31 Médina, 127 métissage, 5, 82–83, 144–45, 147, 172n6

Index missionaries, 15–17, 28–29, 30, 124–25, 135–36, 174n42. See also Boilat, Abbé; Barreira, Father Balthasar; Labat, Father Jean-Baptiste; Sène, Father; Spiesser, Father; Wintz, Father mixed cultures architecture and, 3, 20, 33, 55, 63, 80, 141–42, 147 etiquette and, 82, 90–92 and identity, 2–3, 20–21 intercultural marriage, 18–19, 22, 81, 89–90, 94–95, 171n1, 175nn50,52,55 internal migration and, 154n74, 158n1 race and, 23–26, 28, 43, 88–89, 105, 107–108, 171n1 religion and, 19, 24, 27–28, 82, 85–88, 93, 95, 152n34, 165n94 slave status and, 23–24, 26, 28, 89, 166n102 trade and, 18, 20, 55, 81, 83–84, 96, 105, 107–108, 124, 151n17, 158n3 See also Creole; identity in Senegambia; métissage; tangomaos; Vaz, Gaspar Mlomp, 3, 117, 128–30, 134, 136 Mlomp-Buluf, 6, 130 Mollien, Gaspard, 26, 107 Moore, Francis, 25–26, 53, 60, 69, 74, 84–85, 95, 120 mosques, 31, 50, 85, 164nn79,80 multistory architecture in Brazil, 69, 70, 78 colonial, 20, 52, 72, 131–34 Diola, 127–28, 129, 130, 137 Luso-African, 77, 78, 79 Mande, 51, 74, 138 Manding, 57 Muslims architecture of, 51 and cultural assimilation, 87, 93, 165n94 and fluid identity, 19, 22 and religious conversion, 30 and religious syncretism, 86 and spread of Islam, 86–87, 173n30, 174n35 and trade, 31 and war, 120, 122–24, 182n55 See also bixiiris (bixirins); Islam; Juula (dyula) merchants; Manding; marabouts; mosques; Vaz, Gaspar

205

Nalu, architecture of, 48–49 Ndiaye, Birahim, 123 New Christians, 16, 151n25, 175n52 Nigeria, 73 Niumi, 15 oeyi (religious leader), 86, 95, 177n67 Olinda, Brazil, 66 Pakao, 183n18 and Diola origins, 86, 173n25, 184n33 Manding at, 117, 173n30 and slave trade, 123, 125 palisades at Farim, 37, 159n19 and Floups, 35, 116, 119–20, 160n31 at Geregia, 45 and Jolas (Diolas), 100, 115, 117, 122 and Mandes, 38 and Manding, 36, 39, 182n55 in Pakao, 183n18 at Ziguinchor, 99 palm leaf, as building material, 46–47 palm products, 100, 162n57 Papels architecture of, 37, 48, 103–104, 120, 135 fluid cultural boundaries of, 81, 88 social organization, 4 Paquési, 138 Pasqua, 45 patrilocal societies, 115, 122–23 Patrisio-Parese, Signor, 88, 174n46 peanuts, 116 Pernambuco, 65, 77 Petite Côte, 7, 18, 24–25, 28, 149n3 Peulh, 182n55 piracy, 105, 186n52 pirates, 114 pirogues, 101 polygyny, 55, 87, 174n37 porches in colonial architecture, 62, 68, 133 cultural importance of, 53 historical origins of, 47–49, 54, 57, 66, 163n73 in “Portuguese”-style architecture, 17, 59–60 precolonial, 52 See also alpainters; verandas; vestibules Porto Praya, 62, 63 Portudale, 3, 16, 152n27

206

Index

Portuguese colonizers in Africa architecture of, 36–37, 39 at Bissau, 103 at Cacheu, 37, 42 at Casão, 19 competition with other colonizers, 23, 98, 104, 106, 142, 178n5 on the Gold Coast, 62 lose holdings to France, 11, 24, 106 participation in slave trade, 35 role of Portuguese Crown, 14, 16 as traders, 4, 13–14, 18, 94, 98 at Ziguinchor, 77, 100 Portuguese colonizers in Brazil, 65–67, 70, 73, 93 Portuguese Inquisition, 16, 149n1, 173n32 “Portuguese” (Luso-Africans), 4, 7, 13–14, 31 as African chiefs, 105–106 cultural identity of, 5, 11, 14–15, 28, 46, 88, 104, 175n47 religion of, 16–17, 24, 26–31, 87–89 and slave trade, 26–27, 31, 45, 55, 105, 116, 150n6 See also identity in Senegambia; Jews; tangomaos; traders in Senegambia “Portuguese”-style architecture, 78–79 in African communities, 38, 44–45, 56–57, 131 borrowed by English, 44, 53 in Brazil, 59–80 and cultural identity, 9–10, 12, 142, 146 and cultural mixing, 18, 33, 80 described, 17, 44, 52, 56, 59, 61, 166n3 in European trading centers, 45, 162nn55,60 features of, 44, 46, 49, 52, 54, 59–60 influenced by African architecture, 137 origins of, 49, 57–58, 78–80. See also adobe (banco); identity in Senegambia; porches; roofs; verandas; vestibules; whitewashing Post, Frans, 11, 67–70, 74, 77–78, 169nn33,34, 169–70n42 Post, Pieter, 168–69n31 prayer beads, 27 Raffenel, Anne, 108 ramparts. See African architecture, ramparts

Recife, Brazil, 66 rectangular architecture in Africa, 17, 39, 49–50, 59, 61–62, 77, 99 in Brazil, 68–69, 71, 75–76 religious syncretism, 16–17, 19, 27–30, 86–88, 174n35 Réunion, 17, 59–60 rice culture, 55, 98, 100, 114, 126–27, 173n22, 186n52 roofs clay or earth, 47, 55, 59, 99, 127 four-sided, 62, 64, 68–69, 71, 75, 167n13 removable, 99, 103, 166n2, 178n12 suspended-construction, 60, 171n60 thatched, 10, 44, 46, 59–61, 72, 75–76, 99, 103, 115, 119, 166n2, 178n12 tile, 47, 50, 68, 70, 71, 75, 76, 171nn57,58,60 See also forkillas round architecture. See African architecture, round rubber trade, 126, 186n53 Rufisque, 3, 61 Ruiters, Dierck, 14, 20, 150n5 S. Domingos, 2, 8, 17–18, 39, 72, 131 Saint-Louis (St. Louis) architecture at, 46, 87 French colonize, 13, 102, 109 intercultural marriage at, 175n50 religious practices at, 30 slave trade at, 185n47 views about skin color at, 104, 108, 155n94 Saloum, 1 salt, 98 sandegil, 19–20 San∑, Cheikh Suleman, 115, 173n26, 176n64 Sangédegu (Bagnun kingdom), 39 Santiago, 19, 76, 131 Santiago Island, 14, 19–20 Sao Tomé, 168n22 Sapes, 19, 152–53n45 Sedhiu, 3, 97, 110, 133, 184n33 semitic dynasty, 139–40 Sène, Father, 100 Senegal, 132 Serers (Sereers), 1, 21, 28 architecture of, 135 oral tradition of, 109

Index religious practices of, 157n111 social organization of, 37, 51 shrines, 16, 30, 86–87, 126, 151n22, 174n34. See also bekin (shrine); chinas (shrines) Sierra Leone, 62, 100 skin color and identity, 24–26, 104, 139–40, 142. See also black identity; white identity slave trade in Senegambia and Brazil, 11, 65, 70 at Cacheu, 36, 42, 159n16, 168n22 children as slaves, 121, 185nn47,48, 185–86n49 fluid social status of slaves, 19, 22 at Geregia, 45 growth of in nineteenth century, 116, 125 and Portugal, 35 “Portuguese” as slave traders, 26–27, 31, 45, 55, 105, 116, 150n6 price of slaves, 125, 185n47 skin color and slave status, 26 slave raiding in Senegambia Bagnun involvement in, 36, 159n15, 162n53 Bijogos involvement in, 37, 56, 159nn17,18 in Buluf, 120–21 Floup involvement in, 36 in Fogny, 84, 120, 121, 122–26, 185–86n49 and Juula merchants, 84, 185–86n49 Manding involvement in, 159nn13,14, 173n16, 184n33 See also Combo Sylla (Fodé Sylla); Fodé Kaba Soninké, 86, 139, 141 Soungrougrou River, 100–101 Spiesser, Father, 117 St. Louis. See Saint-Louis staffs of office, iron, 91, 176n59 stockades in African villages, 37 stone buildings in Africa, 42–43, 113, 188n83 Sudan, 137 sugar plantations, in Brazil, 68–70 tangomaos, 17, 19–20, 51–52, 150n6 Temnes, 27–28 Thiobon, 3, 116 Thionk-Essyl (Thionk), 3, 182n6 architecture of, 117, 118–19, 127

207

colonization of, 136, 185n45, 187n70 economic organization of, 114 language of, 170n45 oral traditions of, 85, 123, 130, 176n64, 177n67 political organization in, 95 and slave trade, 116, 121, 184n31, 185n39 social organization of, 115 trade in, 34, 124 Timbuktu, 50, 137 Togo, 73 trade commodities in Senegambia amber, 45 beeswax, 45–46, 84, 88, 98, 150n6, 159n16 cattle, 100 cotton, 98, 100–101 elephant teeth, 84 glass beads, 45 gold, 173n16 iron, 30, 45 ivory, 45, 84, 98, 150n6, 173n16 kola products, 45, 100 palm products, 100, 162n57 peanuts, 116 rice, 55, 98, 100, 114, 126–27, 173n22, 186n52 rubber, 126, 186n53 traders in Senegambia, 1, 31 Cape Verdean, 5, 14, 18, 36 Diola, 162n57 Dutch, 15, 23 economic decline of, 95 English, 15, 23 European, 5, 7, 10, 81, 86 Floup, 131 French, 23 Juula, 84, 185n49 Luso-African, 11, 15, 17, 81, 84, 97–98, 100, 102, 131 Mande-speaking, 5 Manding, 1, 20, 52–53, 84, 97–98, 101, 108–109, 162n57, 172n13, 173n16 Portuguese, 30, 39 Wolof, 97, 109, 180n38 women as, 17, 44–45, 89–90 Vaz, Gaspar, 19–20, 22, 83, 87, 153nn56,59, 172n9 Ventura, 19–20, 22 verandas and African architecture, 48, 52, 57, 128

208

Index

and Brazilian architecture, 71, 75 in colonial architecture, 132, 163n73, 187n69 cultural significance of, 55 historical origins of, 47, 54, 164–65n86, 170n46 and “Portuguese”-style architecture, 17, 49, 59, 78, 142 See also alpainters; porches; vestibules Verda, Manuel, 106–107 vestibules, 17, 44–45, 49, 55, 59, 73, 142 Villeneuve, Geoffroy de, 109–10, 112, 180n42 Vintang, 3 architecture at, 44–45, 55–56, 72, 89–90, 162–63n60 as center of trade, 54–55, 83–84, 90, 97, 124 cultural mixing at, 39, 82–83, 86, 124, 173n31 economic decline at, 95, 97, 103 and slave trade, 42, 84

whitewashing in African architecture, 161n46 in Brazilian architecture, 64 in colonial architecture, 36, 42, 59 cultural significance of, 43 function of, 36, 43 in “Portuguese”-style architecture, 17, 44, 46, 49, 142, 161n41 Wintz, Father, 135–36 witchcraft trials, 27 Wolofs architecture of, 111–13, 135 Europeans live among, 18 language, 180n42 mixed cultures and, 21, 27 as slave raiders, 116 women as traders, 17, 44–45, 89–90 as victims of slave raids, 125 See also intercultural marriage; intercultural unions World War I, 126, 130

Waalo, 1 war drums, 116 warrior-chiefs, 126 white identity and cultural identity, 25, 27, 88–89, 97 and freedom, 25–26, 28, 107–108, 155n94 and occupation, 107–109 and religion, 28–29 See also bidan identity

Ziguinchor, 2–3 architecture of, 72, 98 colonization of, 187n72 cultural mixing at, 3 decline of, 100 founding of, 186n67 French possession of, 11, 77 Portuguese trading post, 39, 97–98, 107, 132, 142, 178n12 trading center, 30 Zimbabwe, 137

Yemen, 161n45 Yola. See Jolas

Peter Mark is Professor of Art History at Wesleyan University, Middletown, Connecticut.