Oduduwa's Chain: Locations of Culture in the Yoruba-Atlantic 9780226506555

Citation preview

Oduduwa’s Chain

Oduduwa’s Chain Locations of Culture in the Yoruba-Atlantic

ANDREW APTER

The University of Chicago Press Chicago and London

The University of Chicago Press, Chicago 60637 The University of Chicago Press, Ltd., London © 2018 by The University of Chicago All rights reserved. No part of this book may be used or reproduced in any manner whatsoever without written permission, except in the case of brief quotations in critical articles and reviews. For more information, contact the University of Chicago Press, 1427 E. 60th St., Chicago, IL 60637. Published 2018 Printed in the United States of America 27 26 25 24 23 22 21 20 19 18

1 2 3 4 5

ISBN-13: 978-0-226-50638-8 (cloth) ISBN-13: 978-0-226-50641-8 (paper) ISBN-13: 978-0-226-50655-5 (e-book) DOI: 10.7208/chicago/9780226506555.001.0001 Library of Congress Cataloging-in-Publication Data Names: Apter, Andrew H. (Andrew Herman), author. Title: Oduduwa’s chain : locations of culture in the Yoruba-Atlantic / Andrew Apter. Description: Chicago : The University of Chicago Press, 2018. | Includes bibliographical references and index. Identifiers: LCCN 2017018047 | ISBN 9780226506388 (cloth : alk. paper) | ISBN 9780226506418 (pbk. : alk. paper) | ISBN 9780226506555 (e-book) Subjects: LCSH: Yoruba (African people)—Atlantic Ocean Region—Religion. | African diaspora. | Cults—Atlantic Ocean Region. | Orisha religion—Atlantic Ocean Region. | Atlantic Ocean Region—Religion. | Nigeria, Southwest—Religion. Classification: LCC BL2480.Y6 A78 2018 | DDC 299.6/8333—dc23 LC record available at https://lccn.loc.gov/2017018047 ♾ This paper meets the requirements of ANSI/NISO Z39.48-1992 (Permanence of Paper).

CONTENTS

Preface / vii

Introduction / 1 ONE

T WO

/ Herskovits’s Heritage / 17

/ Creolization and Connaissance / 39

THREE

FOUR

FIVE

/ Notes from Ekitiland / 68

/ The Blood of Mothers / 97

/ Ethnogenesis from Within / 122

Afterword: Beyond the Mirror of Narcissus / 152 Notes / 157 References / 171 Index / 189

P R E FAC E

Even the maestro-critic Paul de Man insisted that “however negative it may sound, deconstruction implies the possibility of rebuilding” (1983, 140), and it is in this spirit of critical reconstruction that I have reassembled the five essays of the present volume. Written and published between 1991 and 2013, they represent the working through of a particular problematic in African Diaspora and Black Atlantic studies on the basis of my first extended foray into Yoruba ritual and politics and “the hermeneutics of power” (Apter 1992) that I encountered in Nigeria. Although I focused exclusively on Africa at the time, my subsequent explorations of the Afro-Atlantic world through a self-consciously Yoruba lens have brought previously unacknowledged patterns into view, pertaining to dominant paradigms in the African Diaspora literature, the complex dynamics of Afro-Caribbean religions, and their historical trajectories as sociopolitical communities with variable forms and degrees of self-consciousness. As I mention in the introduction, I did not set out with this broader purview in mind, but after the dust of fieldwork settled and my own understanding of Yoruba critical frameworks coalesced, I couldn’t help “seeing” what I “saw,” in various locations of what I am calling the Yoruba-Atlantic. Which brings me back to Paul de Man and the blindness of whatever insights I may have to offer in these pages. As a Yale undergraduate in the 1970s, I was caught up in the first wave of deconstruction in the United States, when the Yale Critics and their followers were coming into being. And Gayatri Spivak’s “translator’s preface” to Jacques Derrida’s Of Grammatology provided the intellectual looking glass though which a new generation of acolytes ventured, intent on bringing down the Old Order by tracking tropes and demystifying texts as we marched through Frederic Jameson’s prison house of language to the inaugural “Lit-Z” cotaught by Paul de Man and

viii / Preface

Geoffrey Hartman, mingling with visiting luminaries such as Umberto Eco and Jacques Derrida along the way. As a philosophy major, I wrote my senior thesis on Kant’s Critique of Pure Reason, arguing that Kant’s deletion of the subjective deduction in the B edition, and his recasting of transcendental synthesis in the figure of synthetic unity, represented a futile attempt to disguise the rhetorical function of metaphorical substitution at the core of his critical philosophy. The good old days, to be sure, and useful training for the textual turn in the postmodern anthropology of the 1980s and early 1990s that followed. Thus I was, and remain, no stranger to deconstructive criticism, and I have no difficulty appreciating its relevance to diasporic frameworks predicated on rhetorics of displacement from an originary Africa. Indeed, I take those “truths” to be self-evident, part and parcel of the negative dialectic of creative remaking, or “rebuilding,” as de Man would have it. But the negation of Africa as the figural ground of cultural dissemination works best in transition or transformation, as a place from which to proceed rather than permanently dwell. And over the years I have found myself arguing against the “dwellers,” those critics who, in the name of decolonizing knowledge, insist on disavowing Africa as a locus of insight into its Atlantic cultural legacies, relegating it instead to the mystifying margins of the European imagination and its inscription devices. Against this enduring trend with its impressive iterations, I offer a range of counterperspectives that attempt to rebuild an Afrocentric point of departure—not by retreating into precritical scholarship but by sifting through the rubble of its decolonized remains. The rebuilding I have in mind involves three related forms of critical recuperation discussed in the introduction: of Africa as a viable locus of Atlantic historical and cultural interpretation; of culture as a viable conceptual framework for illuminating Atlantic historical trajectories; and of Afrocentric (as opposed to Eurocentric) critical paradigms as powerful shapers of Afro-Atlantic history. The overall thrust of the essays that follow is that Yoruba revisionary and generative schemes—and their deconstructive and reconstructive ritual strategies—illuminate both highly specific pathways of religious syncretism and cultural creolization in the Americas and more general strategies of apprehending Africa throughout the Black Atlantic. This slippage between specific and general goals is principled but may cause some confusion. I am not implying that the Yoruba-Atlantic derived exclusively from a Yorubaland bounded by territory, shared language and beliefs, and common religious forms and kinship norms à la “tribal society” of old-school ethnography, and which came to predominate over other African legacies in the diaspora—although on this point I am bound to be

Preface / ix

mischaracterized. In some cases, specific identifications are indeed interesting and appropriate, such as New World remappings of Yoruba sacred geographies with their diagnostic toponyms. But in many other cases, the Yoruba cultural lens that I deploy highlights significant affinities with a wider West African cultural geography, including Fon-Dahomean legacies in Haiti and Brazil, or related Igbo gender constructions in the North American Lowcountry, precisely because such West African commonalities belong to broader regional similarities and levels of cultural abstraction. Frankly, I am not particularly interested in claiming specifically Yoruba “origins” throughout the Yoruba-Atlantic, but I am interested in using Yoruba cultural frames when they illuminate broader West African regional similarities in the Americas. Nor is such slippage merely methodological but rather belongs part and parcel to the history of Afro-Atlantic ethnogenesis. As I discuss in both the introduction and the afterword, the very category “Yoruba” as proper name and self-conscious ethnicity contains such referential and historic instabilities, but these pose no problem when appropriately located, and illuminate trajectories of re-Africanization as well. My broader goal is not to champion Yoruba contributions as such throughout the Black Atlantic, but to provide a critical method for future argument and research. When I do push the Yoruba connections as far as I can, as in chapters 4 and 5, it is more in the spirit of a hypothesis than in making particular empirical claims: that is, how much can we “see” through a revisionary Yoruba lens that we wouldn’t see otherwise, and what do these insights add to our historical interpretations? I invite others to make similar interventions through comparative Afrocentric critical frames. As for acknowledging ancestors, elders, friends, teachers, colleagues, students, and fellow-travelers who have offered guidance and support throughout these Atlantic crossings, including the spirited interlocutors whom I engage in these pages, you are far too numerous to name, but I am forever in your debt. Two who must stand out, however, are T. David Brent, editor extraordinaire at the University of Chicago Press, who back in the day took a chance on publishing the first book of an unknown aspiring Africanist fresh out of graduate school and has stayed with me ever since; and to Robin Derby, partner extraordinaire, who back in the day took a chance on marrying an aspiring assistant professor, took me to the Caribbean, and has also stayed with me ever since: Jẹ́ k’ọ wọ́ yin á máa lọ síwájú oo!

Earlier versions of chapters 1–5 appeared in the following journals: “Herskovits’s Heritage: Rethinking Syncretism in the African Diaspora,” Diaspora

x / Preface

1, no. 3 (1991): 235–60; “On African Origins: Creolization and Connaissance in Haitian Vodou,” American Ethnologist 29, no. 2 (2002): 233–60; “Notes on Orisha Cults in the Ekiti Yoruba Highlands: A Tribute to Pierre Verger,” Cahiers d’études africaines 35, cahier 138–39 (1995): 369–401; “The Blood of Mothers: Women, Money and Markets in Yoruba-Atlantic Perspective,” Journal of African American History 98, no. 1 (2013): 72–98; and “Yoruba Ethnogenesis from Within,” Comparative Studies in Society and History 55, no. 2 (2013): 356–87. Revisions, particularly of the earlier essays, are mainly limited to relating my arguments to new debates and extensive relevant publications, mostly in the endnotes, and updating spellings (e.g., “Vodou”). Permission to reprint them is gratefully acknowledged.

INTRODUCTION

When I first set out to study Yoruba orisha worship in the autumn of 1982, my focus was exclusively on southwest Nigeria; more specifically that subregion of the Ekiti Yoruba that posed particular problems of variation and change. To this day, the classic Yoruba kingdoms of Ancient Ife and Old Oyo have claimed the lion’s share of scholarly attention: hitting the map when Hugh Clapperton and Richard Lander visited “Katunga” (Old Oyo) in 1826, beginning in earnest with Samuel Ajayi Johnson’s monumental History of the Yorubas, enshrining Oyo as the model of a developing Yoruba identity at the end of the nineteenth century, and with the predatory archeology of Leo Frobenius in the sacred groves of Ile-Ife at the turn of the twentieth—all paving the way for the histories, art histories, and ethnographies that have followed in their wake.1 Indeed, as I would later argue, there are historical reasons why Yoruba oral traditions have coalesced around two ritual fields, one associated with the rise of the Oyo empire in the eighteenth century until its demise circa 1836, the other focused on Ife as an ideological foil to Oyo’s hegemonic claims, and which together sustain two competing corpora of Yoruba origins in dynamic performative tension (Apter 1987a). But I am jumping ahead of myself. In the early 1980s, my goals were more simply and narrowly defined as the placing of orisha worship in its relevant political contexts—the historic or “traditional” kingdoms ruled by sacred kings, and segmented into quarters (àdúgbò) ruled by chiefs and elders. Doing so, my instincts told me (and my hypothesis overconfidently stated), would explain the otherwise bewildering variety of priestly hierarchies and rival pantheons that rendered orisha worship so difficult to pin down. What were the political factors that controlled its mercurial frameworks and forms, both regionally in space and historically over time? I pursued this problem in Ekitiland, the northeast Yoruba periphery of

2 / Introduction

southwest Nigeria, for all the wrong reasons that a more critical anthropology has since brought to light. Because Ekitiland was relatively rural, less “developed,” and thus more “traditional” than the cosmopolitan centers of Lagos, Ibadan, or even Ife itself, the orisha cult system (and, yes, I used the word “cult,” and still do in its nonsectarian sense) would be more active and intact. Furthermore, controlled comparison between its systematic variations would be easier, because of the close proximity of the characteristic ministates (Obayemi 1971) that developed as defensive enclaves against eighteenth-century Oyo expansionism (Law 1977), the nineteenth-century military campaigns of the Ibadans from the west (Ajayi and Smith 1964; Akintoye 1971), and the Nupe slave raiders from the Niger-Benue confluence (Mason 1970). Following the advice of the late Jacob Ade Ajayi, whose hometown of Ikole was just miles away, I set up in the historic kingdom of Ayede, founded in the 1840s as a centralized military autocracy in contrast to the decentralized polities surrounding it. Here I could compare the centralized kingdom and its associated orisha cults with the more decentralized “village-clusters” of Ishan and Itaji, tracking change and variation over time and space. What I concluded was that the orisha not only mediated the negotiation of political authority though the cosmological renewal of its annual festivals, but also destabilized and transformed the polity by mobilizing dynastic factions and political constituencies. This transformative power of ritual was real, I argued, which is why, through its idioms and deep knowledge claims, it was, and remains, such a serious concern. My dissertation changed emphasis when substantially revised to address what I called the hermeneutics of power, accounting for the revisionary strategies of deep ritual knowledge (imọ jinlẹ̀ ) in political and discursive terms (Apter 1992). Gender, colonialism, Christianity, Islam, and the postcolonial state—both civilian and military—also gained greater attention as important dimensions refracted through ritual. Furthermore, anthropology had started to critically question its objects of knowledge, including the invention of Africa by the colonial library (Mudimbe 1988), all of which influenced my subsequent research and thinking. I mention this not just to rehash my past, but to acknowledge that my initial interests and research problematic had nothing to do with “the Yoruba diaspora in the Atlantic world” (Falola and Childs 2004). In fact, my dissertation adviser, M. G. Smith, who had worked extensively in both northern Nigeria and the British West Indies and was no stranger to the orisha in Grenada and Carriacou, was resolutely skeptical of the whole idea of “Africanisms” in the Americas, undermining the cultural trajectories of Herskovits’s “New World Negro” with a methodologically prescribed “battery of conditions” (Smith 1957,

Introduction / 3

45) virtually guaranteeing that no imputed transatlantic continuities would survive the test.2 The last thing on my mind as I struggled to make sense of my Yoruba material was how it played out in the African diaspora. Thus, at the outset I had no explicit stake in the back-to-Africa debates of the New World scholarship, whether endorsing Herskovits’s baseline cultures, deconstructing the tropes of diasporic narratives, or shifting the entire paradigm from precolonial roots to the triangulated routes of Atlantic slavery and the commercial flows that followed (Gilroy 1993). In fact, I was woefully ignorant of the massive plantation society literature that established the historical conditions of the variable Africanities that developed in the black Americas; and, unlike most of the stakeholders in these debates who worked in the “diaspora,” I was first and foremost an Africanist, my blood, sweat, tears, laughter, and intellectual energies—not to mention gifts and cash—primarily invested in Nigeria. To be honest, although I knew the titles of Herskovits’s most famous books and collections of essays, particularly his two-volume Dahomey: An Ancient West African Kingdom, first published in 1938, which contained Yoruba-related material on politics and ritual, and his more famous Myth of the Negro Past ([1941] 1958) and The New World Negro (1966), which did so much to establish Afro-American studies at Northwestern University and in the US academy more broadly, I had not actually read them! Assuming that his studies of Africanisms were deeply compromised by the ideological essentialisms of the colonial era, I was not expecting too much when I finally took the plunge, preparing for a new undergraduate lecture class I developed on rethinking the black diaspora at the University of Chicago. Working my way through Herskovits’s syncretic paradigm for analyzing the African heritage in the Americas, however, I was struck by how it resonated with my revisionary model of Yoruba hermeneutics, subject to major updates of his key terms and concepts. But when I published an article (reprinted here as chapter 1) that critically reformulated his syncretic paradigm along these lines, I nearly lost my job. Immediately branded as a “neo-Herskovitsian,” I caught it from both sides. It was bad enough that the critical vanguard (of which I thought I was a card-carrying member) established my guilt by association with the tribal essentialisms and genealogical fictions that I thought I had taken pains to deconstruct and reconstruct. It appeared that reconstruction was the wrong way to go, whatever the overt qualifiers and interpretive results. But even worse was the reaction of the committee of tyrants that ruled Chicago’s anthropology department and ruled over my fate as an aspiring assistant professor. With the exception of Jim Fernandez, who had been a Herskovits student and liked my piece (and as a result, didn’t really weigh in as a tyrant), I had

4 / Introduction

clearly hit a sensitive nerve. If the antipathy toward Herskovits was primarily intellectual, directed against his impressionistic culturalism and sociological naïveté, I also felt a submerged institutional memory at play, pushing against the imperious personality who had dominated African studies up the road in Evanston. In any case, things hit rock bottom when, in preparation for my third-year review, the chair suggested that it might be prudent to publish an article with at least a footnote disavowing my prior take on “Herskovits’s heritage.” Needless to say, I refused and prevailed, but it was hard out there being a neo-Herskovitsian, particularly since I didn’t self-identify as such. In the quarter century that has followed my baptism by fire into this dynamic field of intellectual production, the field itself has dramatically shifted from diasporic dispersals of the Middle Passage to Deleuzian models of rhizomatic circulation associated with Atlantic slavery, merchant capitalism, and the neoliberal ethnoscapes of globalization.3 Emphasis on African origins had already shifted away from the terra firma of West and Central Africa to the new social frameworks (Bastide [1960] 1978) and the creolizing contexts (Mintz and Price [1976] 1992) of slave ships and New World plantation societies, where neo-African identities and homelands, more produced than recovered, destabilized (if not altogether undermined) transatlantic cultural origins and continuities. Attention also shifted to how Afro-Caribbean cultural genealogies were further appropriated and mythologized by postindependence nationalist projects, as in Papa Doc’s noirisme (Dayan 1998 Nicholls 1979), Brazil’s so-called racial democracy (Afolabi 2009; Skidmore [1974] 1993), or the folkloric soup (ajiaco) of Fernando Ortiz (1940; see also Moore 1994; Palmié 2013, 78–112). Today, the very idea of establishing African origins in other than ideological terms seems quaint and naive if not perniciously misguided (Scott 1991, 1999), with the more exciting work focusing on coterminous Atlantic trajectories (Palmié 2002, 2010), the transatlantic circuits of Brazilian or Cuban returnees to Lagos (Matory 2005; Otero 2010), re-Africanization in diaspora (Capone [1999] 2010; P. Johnson 2007), the intellectual and dialogical production of Afro-Atlantic objects of knowledge and value (Matory 2006; Palmié 2013; Sansi 2007; Yelvington 2006a, 2006b), the metalinguistic framing of “African”-derived ritual languages (Wirtz 2007, 2008, 2014), or the transnational networks of Yoruba community as the orisha go global (Argyriadis and Capone 2011; Clarke 2004; Capone 2005, 2016a; Palmié 2005) and Santería goes electric (Beliso-De Jesús 2015). Why bother with an outdated if not fallacious problematic—a critical reappraisal of Yoruba origins—when Afro-Atlantic studies have clearly moved on in such interesting new directions?4 Why insist that, despite the distortions of baseline genealogies and

Introduction / 5

associated myths of tribal purity, West African cultural frameworks—when critically reformulated—illuminate important Atlantic historical trajectories? The problem has been on my mind for a long time. But to help frame the answer to this question, I begin with a surprising phone call that I received in September 2015.

An Unusual Request When Olu Ibitoye called from his home in Chicago, I hadn’t heard from him in several years, but that doesn’t matter with an ọ mọ ìyá (child of the same mother), whose “actual” mother—the commanding Shango priestess in Ayede—had taken me under her wing and backed my efforts to gain some level of access to the major orisha cults when others had viewed me with suspicion and distrust. In an instant, Olu and I were back together, laughing at the feeling of space-time compression as our lives collapsed into the intimacy of our collaborative selves. After the formulaic pleasantries, Olu announced that he had a very big favor to ask, and would I please oblige. “Of course, certainly, anything, no problem,” I replied, hoping that I had not committed myself to a herculean labor of exorbitant costs and proportions. It turned out that with a new ọ ba (king) recently installed in Ayede, less hostile to “paganism” than his Pentecostal predecessor, the orisha cults were getting a new lease on life, enjoying something of renaissance as a valued form of cultural patrimony.5 But there was a problem. The master drummer Ajayi was getting old and could no longer perform up to speed. Young initiates were being called, but questions of procedure and precedent were coming up. Could I please digitize all the videotapes and audio recordings that I had made during fieldwork, representing festival cycles in 1983, 1984, 1990, and 1993? Could I please burn them onto DVDs and Olu would take them back, providing a record of past festival performances and settling some issues of correct ritual protocols? Or even better, would I please go back with him and receive a chieftaincy title that the new ọ ba was proffering? The implications and ramifications of such a request are frankly astonishing, for many reasons ranging from the personal and ethical to the political and methodological. Most immediate was the ironic reversal of positions that had structured my apprenticeship in the field, captured so succinctly by one local chief who had stated in no uncertain terms, “No matter how many years you work with the priestesses, they will never give you anything; they will never leak their secrets!” Indeed, the social organization of secrecy became one of the guiding threads of my research precisely because I had such limited access to its labile contents, and although the chief had a point, his

6 / Introduction

was certainly not the last word on the subject. Nonetheless, I could not help being struck by the momentous shift in relations of access, as the priestesses were coming to me for material, even if they still knew better than I what to do with it. More important, I was deeply honored to oblige, not only on the personal grounds of giving something meaningful back to the those who helped me launch my career, but also professionally, as the nightmarish weight of dead colonial officers was slightly alleviated from my anthropological brain. Most of my generation of Africanists has struggled with the politics of the colonial library (Mudimbe 1988), and with a discipline that emerged with European empire and overrule, and even if I have worked out certain positions that take these conditions of genesis into account without sacrificing the discipline tout court (Apter 1999; 2007, 1–14), one never rests easy with anthropology’s heart of darkness. Thus the fact that the priestesses (and a few priests too) were coming to me for assistance in their work felt politically liberating, showing that our anthropological projects can be valuable to those whom we study, on terms that remain theirs. Neither were the “postmodern” lessons lost on me, as I risked becoming a Vargas-Llosaian storyteller, rescripting through my camerawork the very ritual performances I have sought to represent, exemplifying the anthropologist’s invention of culture and tradition if not the more dialectically nuanced ethnographic interface (Palmié 2013, 49–54), replete with looping effects, converging agendas, and objectifying correlates. Many of us working in Nigeria or Cuba have had similar experiences of interviewing diviners and priests who, at the end of a session, reveal the sacred document that tells it all, whether a dog-eared photocopy of Bascom’s Ifa Divination or the tattered text of Samuel Ajayi Johnson’s History of the Yorubas, tucked away in their shrines. Was the priestesses’ request not just a more contemporary iteration of this feedback loop, similar to the circulating videotapes “possessing” the past so brilliantly analyzed in Beliso-De Jesús’s Electric Santería? Am I now, more than ever, deeply implicated in the construction of a sacred tradition? Contra the prevailing winds of current scholarship, but without recommending a great leap backward, I want to pursue a different course, one that acknowledges the social networks, dialogical pathways, textual relations, working conditions, political agendas, and historic inequalities that have produced an “original” Yoruba culture both in Nigeria and in diaspora but which still endorses a relatively objective “something else” that exists over and above its multiply mediated actors and formats, at different levels of scale and abstraction, that we can call, invoking Kant’s critical philosophy, the anthropological object (culture) = x.6 Just as Kant sought the limits of

Introduction / 7

pure reason in his first Critique by building a prior framework of objectconstitution to gain experience and objective knowledge, he also posited an ontological ground, beyond his framework, where things existed in themselves. I mention this distinction (between phenomena and noumena) because, when relativized in historical and cultural (and thus Weberian) terms, it helps us see how epistemology and ontology can both improve and impair our vision.7 Briefly stated, if our (hegemonic and ideological) anthropological apparatus of theories, concepts, discourses, notebooks, networks, professional organizations, laptops, tape recorders, and yes, video cameras—all part of the inscription devices (Latour and Woolgar 1979, 51) that generate ethnographic facts—can be taken as the prior framework of anthropological object-constitution, including the objectification of other cultures, there nonetheless remains (as Jean Comaroff used to say) a “‘there,’ there” to which this apparatus applies, providing some kind of access to some kind of reality beyond the apparatus itself. If we can only “know” that reality through our ever-evolving frameworks, it does not follow that we create that reality; or, put another way, that the facts that we construct are 100 percent reducible to the frameworks through which they are apprehended and interpreted. If they were, we would be forced into one of the two epistemological extremes that continue to define the limits of our discipline; namely, the deconstructive assimilation of all referential claims to endless chains of tropological displacement (the mise en abîme of postmodern anthropology) or the utopian myth of scientific objectivism, in which the correct theory finally gets reality right—an approach that has had a checkered anthropological career to say the least, whether in yesteryear’s pseudoscientific ruminations on race or today’s scary forays into evolutionary psychology. So what is to be done? How do we build a viable bridge between the facts that we create and the world we wish to know? For Kant, the answer lay in synthetic a priori propositions that gave objective determination to the world (making experience possible) while somehow letting part of the world slip in. Whether such propositions actually provide the synthetic unity that Kant sought is a question best left to Kant scholars. More useful to me is the critically framed relationship between epistemology and ontology, a relationship that has gained new currency in anthropology’s recent ontological turn.8 If I appeal to an earlier conception of ontology that ultimately reflects “on what there is” (Quine 1948), however, it is as the destination of interpretation and analysis rather than the jumping-off point for its current congeners—a destination, moreover,

8 / Introduction

that requires epistemology to get there. All of which is preparing the ground for the following compound question: Does Yoruba culture exist, and if so, where and when is it located in West Africa and in the Americas? Most critical anthropologists concerned with such issues answer in the negative for a variety of reasons. First, there is no “Yoruba culture” in the hypostasized sense of an objective entity because such fictions belong, together with “race” and “tribe,” with the naturalizing categories of European empire and administrative overrule, with prior roots in the stereotyped African “nations” generated by Atlantic slavery. The culture concept for many is a “thing” of the past, no longer an issue worthy of debate, although even when deconstructed, it still seems to slip back in sous rature. Second, and more central to current research, is a shift of emphasis away from Yoruba culture qua culture toward the historical construction of Yoruba identity (ethnogenesis) by intersecting categories of motivated actors in West Africa and throughout the Atlantic, including anthropologists, missionaries, elites, entrepreneurs, and ritual specialists (the latter primarily in Cuba and Brazil, but now throughout Latin America and Europe) who have coproduced a body of knowledge that is mistakenly retrojected into a mythic African past. As Stephan Palmié provocatively argues, what is “Yoruba” in Afro-Cuban religions is not some cultural heritage out of Africa, but rather (to oversimplify his prodigious explications) the historical relations of textual production that generated an Afro-Atlantic space catalogued under BL.2532.S3 in the Library of Congress. In effect, he replaces the colonial “discourses” of an invented Africa with the “metadiscursive frameworks” (2013, 49) and historical trajectories of an “ethnographic interface,” one that gave rise to powerful texts and temples, but one that should never be confounded with Yoruba origins. This latter shift to language about language, or to the metadiscursive construction of a “boundary object” called “Yoruba tradition” (Palmié 2013, 54), has developed in dialogue with Kristina Wirtz (2007, 2008, 2014) on the vernacular discourses framing Cuban Santería and the Lucumí language of its songs, invocations, and sacred texts. Reassessing the paradigmatic domain of “Yoruba” ritual language in Santería, with its clearly recognizable Yoruba morphemes and systematic phonological shifts, Wirtz deemphasizes the Yoruba origin of Lucumí in favor of the sociolinguistic strategies (indexical, pragmatic, metapragmatic) deployed by ritual specialists to register and recuperate it as Yoruba speech, thereby giving rise not to prima facie evidence of its linguistic origins, but rather to a discursive chronotope that “imbues it with sacred and historical value” (2008, 143). Evidence for this more indexically grounded semiotic reading stems, in part, from her observation that the meanings of Lucumí terms were provided by practitioners not in the form

Introduction / 9

of semantic glosses but according to their performative contexts of use. As with Palmié, Wirtz’s rich explorations and explications of these performative domains far exceed synoptic summary and illuminate complex histories and dialogical mediations that give the lie to any simplistic attribution of Lucumí’s self-evident Yoruba origins. Nonetheless, reduced to essentials, I would argue that by replacing the linguistic problem of Yoruba origins with the linguistic ideology of Yoruba origins, constituted through Lucumí’s semiotic enregisterment as a sacred ritual language, Wirtz throws the Yoruba baby out with the bathwater, sacrificing the epistemological ground on which to establish a richer sense of linguistic derivation.9 For example, as I have shown in my own research (Apter 1992, 117–48; 1998), Yoruba ritual language in Nigeria also is ideologically framed as dangerous and deep (ìjinlẹ̀ ), possessing àṣ ẹ . It also is metapragmatically motivated, marking off sacred and secular space, shifting participation frameworks (accounting for similar semantic instabilities), and, perhaps most important, recuperating sacred histories and centers through chronotopic transpositions. How else do the priests and priestesses reestablish Ile-Ife, the original locus of sacred kingship, within their rival centers and shrines? How else do they activate the “time immemorial” (láíláí) of mythic origins within their ritual repossession of the past? My broader point is not that Palmié and Wirtz are wrong in what they bring to light. They have opened up highly productive pathways for pursuing the construction of a cultural tradition, revealing important mediations like Supreme Court rulings and Cuban “brutology” (minstrelsy, blackface) that take us by surprise. Rather, it is the negative implication that they variably develop, that anything like “Yoruba culture” with an earlier history in West Africa cannot be meaningfully engaged to illuminate its Afro-Caribbean manifestations, an approach that impairs our vision. Bringing a relativized Kant back into the discussion, with a nod (or should I say a wink?) to Gilbert Ryle (1949, 6–12), such a negative implication is based on a category mistake; a confounding, I would argue, of epistemology with ontology. In this case, by reducing the reality of Yoruba culture to its historically emergent interfaces and frames (intellectual, discursive, textual, political, etc.), the representational apparatus is ultimately equated with the culture and tradition of which it speaks. In the end, there is no ontological difference between Yoruba culture and unicorns (Palmié 2013, 255).10 The best we can pursue is the historical anthropology of their representations and of the emergent networks and apparatuses in which they have been entextualized, contested, objectified, and catalogued.11 While I applaud the daring of this provocative move, and appreciate the critical balance it restores to the current proliferation of Afro-Atlantic

10 / Introduction

databases, I won’t give up on the ontological claim that there exists, in time and space, a dynamic if semibounded x that has come to be known as “Yoruba culture” that can be critically reformulated to illuminate significant patterns and trajectories in West Africa and the Americas.12 Perhaps it has something to do with my requested videos, which somehow get beyond the zooms and pans of the camera, my relationship with the priestesses, and even the Western audiovisual registers of representing Africa, to a “there” that remains there, with or without my physical copresence. But how do we get “there,” beyond the history of representations to the reality they represent, without reifying a fixed tablet (Bhabha 2004, 2) of culture and tradition?

Oduduwa’s Chain If Yoruba culture is a slippery object, let us begin with its myths of origin, which fall into two genres: creation myths and myths of migration. I do so not to equate myth with culture in a Lévi-Straussian algebraic reduction, but to highlight a central Yoruba schematism that refabricates origins wherever it goes, thus illustrating the kind of critical reformulation of the culture concept that I have in mind. Creation myths generally state that in the beginning, Olodumare, the Yoruba high god, had a son, Oduduwa, who climbed down a chain (ẹ̀ wọ̀ n) from heaven (ọ̀ run, also “sky”) to an uninhabited world that was covered with water. Oduduwa pulled out a handful of earth and a cock from a bag that he had brought with him and placed them on the water. As the cock began to scratch the earth about, land spread out over the water. According to this myth, Ile-Ife is the sacred locus of Oduduwa’s original descent, where he became the first Yoruba king (ọ ba) and fathered future generations of Yoruba monarchs through sixteen sons. This basic narrative, with elaborated variations, is generally thought to express the origin of Yoruba monarchy at Ife, or at last Ife’s importance as an early center of ritual prestige and political power. It is, moreover, the mythic charter of sacred kingship throughout Yorubaland, as any ọ ba of whatever rank must establish a connection with Ile-Ife to justify ownership of a sacred crown, either directly through one of Oduduwa’s sons, or cognatically through one his daughters. Migration myths, on the other hand, tell how Oduduwa, progenitor of the Yoruba people, came from somewhere in the east. Some myths are vague about his point of departure; others cite Nupeland, Egypt, or Medina, although Johnson’s History brings Oduduwa from Mecca (Law 1973). In this latter, Oyo, version, Oduduwa, son of a Meccan king, rebelled against his (unnamed) father and Islam and fled with his children to Ife, where he

Introduction / 11

founded Yoruba kingship. Of his seven sons in this Oyo account, the first six left Ife to found the historical kingdoms of Owu, Ketu, Benin, Ila, Sabe, and Popo (Dahomey), and the last-born, Oranyan, succeeded his father at Ife. Once settled and sufficiently strong, Oranyan set out to overthrow the Meccan dynasty of his forefathers. Delegating one of his father’s slaves to worship the orisha in his absence, he traveled northeast with an army through Nupeland until he reached the banks of the Niger River. Here his way was blocked, but rather than suffer the humiliation of defeat, he settled in the bush and founded the town of Oyo Ile, which later became the capital of the Oyo empire. As I have argued elsewhere (Apter 1987a), this Oyo-centric migration myth displaces the authority of Ife as the locus of sacred kingship by minimizing its association with Oduduwa’s cosmic descent from the sky and by establishing a new, post-Islamic beginning associated with Oyo’s expansion and political consolidation through the imperial cult of Shango. Oyo’s mythic revisionism had its limits, however, and could never quite sunder Oduduwa’s heavenly chain. Despite Oyo’s best attempts to downplay Ife origins, Ife remained enshrined in the rituals of its conquered territories and in the highly coded verses (odù) of Ifa divination. Ife’s importance as a sacred center of kingship, moreover, intensified as a foil to Oyo expansionism, generating an oppositional ritual field to Oyo’s hegemonic claims. My point is not to retrace the historical topography of myth and ritual associated with shifting centers of sacred kingship, but to emphasize the historicity and flexibility of their originary frames. If the world began at Ile-Ife, its beginnings were eminently unstable, predating humanity or postdating Islam, with Oduduwa descending from heaven or arriving from the east, but always bringing the outside in. In the creation myth, Oduduwa’s chain is a complex figure that fulfills a number of poetic functions: as a figure of chiasmus, it crosses the thresholds of heaven (ọ̀ run) and earth (ilẹ̀ ), human and divine, quite literally preparing the sacred ground of kingly dissemination from Ile-Ife. It also establishes (to invoke Jakobsonian poetics) a syntagmatic axis, a genealogical chain of signifiers linking all descendants from a primary progenitor throughout Yorubaland and beyond; to the kingdoms of Dahomey, Benin, and—as should become clear from my expansive emphasis—the reconstituted ritual kingdoms of the Yoruba diaspora.13 This latter association with ritual genealogies overseas evokes the historical density of Oduduwa’s chain—in some mythic variants made out of gold—as a key symbol of Atlantic slavery associated with its wealth, its ships, and the visceral materiality of human bondage. As a creation myth, Oduduwa’s descent is always already historical. As for the migration genre of Yoruba origins, it crosscuts Oduduwa’s

12 / Introduction

syntagmatic chain with a paradigmatic axis of substitution, replacing Oduduwa’s heavenly father (Olodumare) with a Meccan king—identified in some variants as Lamurudu—against whom Oduduwa rebelled. Much ink has been spilled pondering the historicity of these migration myths, leading to speculations about invading conquerors bringing kingship from afar (Beier 1956), the impact of Islam on Yoruba cultural idioms, or, among nineteenth- and twentieth-century Yoruba elites, whether Lamurudu was a mistaken corruption of Nimrod, anchoring Yoruba origins in the Old Testament. In one of the most complex variations of this theme, Samuel Johnson recounts a migration myth in which Oduduwa seizes a copy of the Qur’an while escaping Mecca en route to Ife and, after reaching his destination, relocates it in Ife’s central shrines where it “was not only venerated by succeeding generations as a sacred relic, but is even worshipped to this day under the name of Ìdì, signifying Something tied up” (S. Johnson 1921, 4). Thus appropriated as a sacred relic where it remains a closed book, the Qur’an is textually refetishized within the mythical origins of Yoruba kingship. But the plot thickens. Johnson is bothered by “traces of error” in this account, and in his speculative rereading “cannot resist” concluding “that the book was not the Koran at all, but a copy of the Holy Scriptures in rolls, the form in which ancient manuscripts were preserved” (7). Johnson thus effects a doubledisplacement of the origins of Yoruba sacred kingship; first from paganism to Islam, then from Islam to Christianity. At the beginnings of Yoruba founding myths we find what Spivak (1988, 308) has called “catachresis at the origin,” a space not just of multiple substitutions, but of infinite substitutability. What are we to make of Oduduwa’s chain; its multiple variations, dynamic instabilities, and divergent historicities? The current trend of a more historical anthropology is to identify what Peel (1989, 198) has called the “cultural work of Yoruba ethnogenesis” a project deeply imbricated in the rise of a Christianized literate elite and its late nineteenth-century cultural nationalism (Peel, 2000). It is no coincidence that the creation and migration myths of Johnson’s History were eventually published, in English, by the Church Missionary Society and reflect the concerns of an emerging native elite with their social location within the developing colonial order; reconciling Christianity with tradition, ethnicity with race, and, not surprisingly, mythic origins with racist theories of Hamitic diffusion purveyed by the Europeans of Lagos Colony. Moreover, as Matory (2005, 38–72) has shown, the cosmopolitan community was further influenced by Atlantic vectors of Brazilian returnees combining diasporic commerce with ideas of cultural purity. Historicizing the cultural production of myth rather than treating myth as a powerful cultural framework belongs to the epistemological shift,

Introduction / 13

shared by Palmié and Wirtz, from explanans to explanandum—from a “culture” that explains to one that is explained. But if we reverse the arrows, not dogmatically but dialectically, to work both ways in opposition and synthesis, we can restore the explanatory dimensions of Yoruba culture, whether explicitly claimed as an ethnic identity or implicitly shared by its localized communities in West Africa (Aku, Egba, Ijebu, Ekiti, etc.) and the Americas (Lucumí, Nagô). The question of where and when such a culture is located will be explored in the chapters that follow. For now, I would like to refocus attention on an alternative reading of Oduduwa’s chain and the cultural work of its poetic displacements. It is not surprising that it is clear that the syntagmatic axis is a powerful genealogical idiom linking people, kings, and kingdoms—and thus the flow of àṣ ẹ —to a deified progenitor; and like any genealogical charter, it can be stretched or contracted to fit conditions on the ground (Bohannan 1952). Less obvious, but equally significant, is its relationship to the ground, which Oduduwa not only produces with his dirt-kicking cock but also decenters and recenters through his princely progeny. Oduduwa’s chain is what Ivy (1995, 13) calls a “mobile sign, detachable from locale but dependent on perpetually evoking it,” and in fact reconstitutes origins wherever one finds a palace, a beaded crown, or an orisha shrine, whether in Ibadan, Ijebu, Bahia, or Union City.14 In fact, many orisha shrines in Benin and Nigeria have chains in the ground marking the precise spot where the ancestral hero, usually standing on one foot, disappeared into the earth to become a god. All such shrines are, furthermore, ritual kingdoms unto themselves, replete with titled officers and junior wives and slaves who take possession of the political kingdom and make it their own during their annual festivals, reconstituting the sovereignty of their former towns of origin through the invocation and praises (oríkì) of their gods (Apter 1992). In this fashion, Oduduwa’s chain attaches itself to multiple terrains, generating cosmic centers, ritual communities, and sovereign spaces wherever it goes. Finally, such syntagmatic site selection at the end of the chain (effected wherever an orisha mounts or possesses a devotee) is matched by paradigmatic substitution at its source, whether in Olodumare’s heaven or somewhere in Arabia. Like a geometric Klein bottle, with its empty center outside itself, the father of the Yoruba people is infinitely replaceable by bringing the outside in, be it Islam or Christianity from the Middle East, or, as I shall argue in the final chapter, patriarchal whiteness from the plantation master’s house. Indeed, such substitutability at the source extends throughout the genealogical chain, rerouting the flow of social and initiation bloodlines at any potential link when conditions are right. As we shall see, the same cultural matrix mediating

14 / Introduction

political fission, dynastic usurpation, and contests for power between civil chiefs in Yorubaland governs similar dynamics within the ritual houses and plantation societies of the Americas. Indeed, what Palmié (2013, 46–49), referencing Brandon (1983, 89), calls the “great transformation” in late nineteenth-century Cuba, when ritual initiation (rebirth) superseded birth into a kin group as a mode of becoming Lucumí, in no way problematizes the principles of Yoruba cultural transmission, as Palmié suggests, but actually exemplifies them, illustrating how strategies of incorporating “strangers” within Yoruba houses (ilé), kingdoms (ìlú), and shrines in West Africa were redeployed in the Creole societies of the Americas (see also Brandon 1993, 83). Ironically, the very separation of social and ritual kinship is not really a problem for most Yoruba in West Africa because both spheres emerge from the same generative category. Nor would the initiation of a white Cuban babalao such as Palmié’s example of Bonifacio Valdés raise many eyebrows. Rather, such ethnic displacements of genealogical transmission—what Roach (1996, 2–3, 28–29) calls performative or ritual “surrogation”—are already enshrined in Oduduwa’s chain, not as a timeless tablet of tradition, but as a generative framework of refabrication that not merely is produced by historical actors, but also has shaped New World societies and trajectories: in what cultural registers, on what levels of abstraction, and in what Yoruba-Atlantic locations will be explored in the chapters that follow.

The first chapter, “Herskovits’s Heritage,” returns to the classical problem of religious syncretism in Haitian Vodou, Cuban Santería, and Brazilian Candomblé. Reformulating Herskovits’s syncretic paradigm in more dynamic and critical terms, based on the Yoruba hermeneutics of political revision, I argue that the ritual association of African gods with Catholic saints developed less as a psychological mechanism of acculturation, as Herskovits maintained, and more as a strategy of collective empowerment and agency among slaves and free blacks in the Americas. Reanalyzed through the lens of Yoruba ritual organization, I argue that the cultural continuities between West African and New World religious “cults” are much closer than the standard narrative of deracination allows, a theme developed in subsequent chapters in different spheres of social practice. Thus, in chapter 2, I apply Yoruba revisionary strategies to the New World dynamics of creolization. Focusing on the Petwo paradox in Haitian Vodou, which opposes Creole powers of money and magic to the venerated,

Introduction / 15

if enervated, authority of Ginen (Africa), I address a fairly narrow debate on the division of Petwo and Rada deities and their imputed Creole versus African origins. Against the ideology of Haitian Vodou, and its misleading influence on various scholars, a Yoruba–Dahomean cultural hermeneutic reveals the African origins and revisionary principles of the Petwo and Rada opposition, as it emerged before the Haitian revolution and realigned with class relations under François Duvalier. In chapter 3, building and reflecting on the pioneering work of Pierre Verger, who undertook the first intensive comparison of Yoruba orisha worship in West Africa and Brazil (and documented nineteenth-century Brazilian returnees), I examine the sociopolitical dimension of orisha cult organization and change in two Ekiti Yoruba kingdoms. Comparison of the politico-ritual configurations of decentralized Ishan kingdom with those of centralized Ayede kingdom, and their very different historical transformations from circa 1845 to the present, reveals political segmentation, not family or lineage, as the dominant principle of cult organization, even if it is cast within lineage ideology. The orisha cult clustering that thus occurs in the Ekiti Yoruba highlands, a ritual characteristic that Verger attributes to innovation in Brazilian Candomblé, suggests that West African orisha worship is closer to its New World manifestations than has generally been acknowledged. Chapter 4 explores historic locations of Yoruba gender ideologies and their characteristic idioms of fertility and witchcraft within a range of social practices and economic institutions throughout the Black Atlantic world. These include ideas about childbirth, lactation, infanticide, and sexuality associated with women’s economic activities in the marketplace and home. Relating the blood of mothers—the irreducible “secret” of human reproduction—to the circulation of money and commodities in the body social, I identify emergent vectors of black female agency in a broad range of plantation societies in the Americas. Chapter 5 provides a polemical counterpoint to externalist accounts of Yoruba ethnogenesis, in which the development of Yoruba identity in the late nineteenth century is attributed to Fulani perspectives on their Oyo neighbors, Christian missionaries, and the politics of conversion (Peel 2000), as well as to Afro-Brazilian merchants in diaspora reconnecting with their homeland (Matory 2005). In chapter 5 I both complement and destabilize these externalist perspectives by focusing on the Yoruba logic of home and house (ilé), relating residence, genealogy, and regional identities to their reconstituted ritual frameworks in Cuba and Brazil. Following

16 / Introduction

Barber’s (1991) analysis of Yoruba praise-poetry (oríkì) and Verran’s (2001) work on Yoruba quantification, I reexamine the semantics of the category ilé in the emergence of Lucumí and Nagô houses to illuminate the racialized dialectics of ritual purity and provide an internal counterperspective on Yoruba ethnogenesis.

ONE

Herskovits’s Heritage

It is customary, if not mandatory, in contemporary African diaspora studies to invoke the pioneering spirit of Melville J. Herskovits (1895–1963), whose lifework was dedicated to the repossession of Africa’s heritage in the New World.1 Not that Herskovits was the first to engage in such research. Others before him included W. E. B. Du Bois and Carter G. Woodson, as well as Jean Price-Mars of Haiti, Fernando Ortiz of Cuba, and—more his contemporaries than predecessors—Zora Neale Hurston and the Brazilian ethnologists René Ribeiro, Arthur Ramos, and Gilberto Freyre.2 But Herskovits more than any other scholar posed the African American connection as a theoretical problem that, in the service of a progressive if intellectually circumscribed political agenda, demanded systematic research into an unprecedented range of West African and New World cultures. It is not my aim to praise a great ancestor, whose flaws and limitations are as legendary as his virtues, but to assess the relevance of his theoretical program to contemporary African American research. In particular, I will focus on his syncretic paradigm, which continues—even among those who disavow it as crudely essentialist or unwittingly racist—to inform the current renaissance in studies of the African diaspora. There is much that seems wrong, misconceived, and simply outdated in Herskovits’s syncretic paradigm when it is evaluated against the current standards of a more critical anthropology. It is not difficult to see how Herskovits essentialized tribal origins in Africa, perpetuated myths of cultural purity in the New World, overlooked class formation, and developed passive notions of acculturation and cultural resistance, all of which distorted the ethnographic record under the guise of an imputed scientific objectivity. But there also is something elusively tenacious about the concept of syncretism. Even when critically deconstructed, it somehow creeps back into

18 / Chapter One

any meaningful discussion of Africanity in the New World. And even as we recognize that “Africa” has been ideologically constructed to create imagined communities in the black Americas—as Guinée in Haitian Vodou, Lucumí in Cuban Santería, or the Nagô nation of Brazilian Candomblé—such invented identities cannot be totally severed from their cultural analogues (dare we say origins?) in West and West Central Africa. The goal of this chapter is to rethink syncretism in a way that does justice to both sides of this methodological divide: equally to the inventedness (and inventiveness) of New World African identities and to their cultural and historical associations with African peoples. This resolution requires greater clarity about just what it is we are comparing, contextualizing, and historicizing on both sides of the Atlantic, an exercise that gives a new twist to Herskovits’s ethnohistorical method and owes its revisionary strategy to some critical lessons that I learned from the Yoruba in Nigeria.3 It is perhaps noteworthy that recent studies of “Africanisms” in the Americas, focusing on particular deities like Ogun, on religions like Santería and Vodou, or even on transatlantic aesthetic and philosophical complexes, have effectively erased syncretism from their lexicons.4 These excellent studies indeed reveal that there is much more to such New World cultural forms than a blending of two distinct traditions into a hybrid form. Scholarly emphasis has shifted to disclose the nuanced complexities of the historical conditions in which African identities are remembered and forgotten, fractured and fused, invoked, possessed, repossessed, transposed, and reconfigured within rural peasantries, urban centers, immigrant communities, national arenas, and even at transnational conferences on, for example, the Yoruba-based Orisha tradition.5 Consistent with this move is a shift away from cultural form toward cultural performance and practice, to traditions in the making rather than those already made, preserved, or retained. These developments are welcome as part of the positive trend in cultural studies, but it is equally clear that nothing as powerful as the syncretic paradigm has arisen from its ashes, resulting in a compelling crisis of representation. In brief, what is Africa’s place in the New World? Indeed, what is “Africa”?6 Rather than address this rapidly growing literature, I will return to the essentials of Herskovits’s syncretic paradigm in order to extract the interpretive kernel from its scientistic shell. This involves a pilgrimage to the classic shrines of New World syncretism, where, in Brazilian Candomblé, Cuban Santería, and Haitian Vodou, African gods embrace Catholic saints to promote new religious empires. I privilege these sites not only because, for Herskovits, they represented the clearest cases of syncretism as such, but

Herskovits’s Heritage / 19

because for us, they provide the clearest examples of how African critical practices in the Americas can inform our own research.

The Syncretic Paradigm Herskovits ([1941] 1958, xxii) credited Arthur Ramos as one of the first to employ the concept of syncretism to account for the identification of African deities with Catholic saints in Brazilian Candomblé. Syncretism suggested, for Herskovits, “a pattern of first importance” in the study of Afro-American culture contact and change, the dynamics of which he continued to document and theorize while refining his method over the years. If the scientificity of this method appears contrived today, we should appreciate that Herskovits posed a radical challenge to the sociological interpretations of American and New World “Negro” institutions and practices which, according to E. Franklin Frazier and Robert E. Park (among others), represented functional adaptations to socioeconomic conditions rather than African cultural holdovers or survivals (see Frazier 1939; Park 1919, and Smith 1957). Herskovits called for greater sensitivity to history and culture in the acculturative process, arguing effectively that synchronic sociological reductionism not only violated the ethnographic record, but worse, supported the racist myth that the Negro had no meaningful African history or heritage. It is with this spirit and strategy in mind that his syncretic paradigm must be understood. Under the rubric of “ethnohistorical method,” Herskovits meant simply that ethnology and history should be combined “to recover the predominant regional and tribal origins of the New World Negroes” and “to establish the cultural baselines from which the processes of change began” (1966, 49). This baseline, he argued, was restricted to the West African coastal and rainforest belts, running from Senegal to Angola, because the smaller and later shipments of slaves from East Africa, Mozambique, and Madagascar had minimal cultural impact on previously established West African patterns in the New World. This claim is debatable, although probably true, but it is not my aim to evaluate its empirical plausibility in light of new evidence (e.g., Curtin 1969), but to clarify its underlying logic. In this view, West African cultures are figured as discrete, coherent wholes, which, with various degrees of purity and in different spheres of social life, left their impress on the black Americas. To assess the relative purity of African retentions and to specify their social domains, Herskovits developed a number of related concepts that together can be glossed as the syncretic paradigm.7 In addition to the ethnohistorical method, these concepts are (1) scale of intensity, (2) cultural

20 / Chapter One

focus, (3) syncretism proper, (4) reinterpretation, and (5) cultural imponderables. I will review these concepts not only to point out their profound limitations, but also to draw out their theoretical relevance to contemporary diaspora studies. If Herskovits regarded his scale of intensities as one of his greatest methodological achievements, with hindsight it seems to parody the epistemology of liberal social science. In an effort to quantify New World Africanisms, albeit for heuristic rather than statistical purposes, Herskovits developed a logical continuum from most to least African, segmented into (a) very African, (b) quite African, (c) somewhat African, (d) a little African, (e) trace of African customs or absent, and (?) no report. These relative values were placed in a two-dimensional array, with New World regions and communities like Guiana (bush and Paramaribo), Haiti (peasant and urban), and the United States (Gullah Islands, rural South, urban North) along a vertical axis, and with specific sociocultural domains (technology, economics, social organization, religion, art, music, etc.) segmented along a horizontal axis to represent variable degrees of African intensity within each region or community. Despite internal variations, the resulting table (table 1.1) reveals that “the progression of Guiana, Haiti, Brazil, Jamaica, Trinidad, Cuba, Virgin Islands, the Gullah Islands, and southern and northern United States comprise a series wherein a decreasing intensity of Africanisms is manifest” (Herskovits 1966, 54). Today, the empirical conclusions can be revised. For example, we know from the Prices’ work in Surinam that Guiana is much more creolized than Herskovits ever imagined (see R. Price 1983; S. Price 1984; Price and Price 1980). But the conceptual problems of the schema are more serious. Clearly the intensities themselves (such as “very,” “quite,” and “somewhat African”) are highly relative and subjective. Also, the sociocultural domains (social organization, religion, art) are in no way discrete and ignore class divisions, and the regional and community designations are inconsistent with each other. For example, only Haiti is divided into “urban” and “peasant,” even though this is a distinction that Bastide ([1960] 1978) has shown to be highly salient in Brazil. As a form of knowledge, the scale of intensities resembles the anthropometric measures of physical anthropology, which Herskovits deployed in his postdoctoral research on the phenotypical effects of miscegenation in North America (see Herskovits 1928). Although he was always explicit—following his mentor, Franz Boas—about separating race from culture and language, his scale of intensities echoes a blood-based logic by transposing notions of purity and dilution from racial stocks to cultural genealogies. Thus he could claim that “the Bush Negroes of the Guiana forests manifest African

Herskovits’s Heritage / 21 Table 1.1. Scale of intensities of New World Africanisms (from Herskovits 1966, 53).

culture in purer form than is to be encountered anywhere else outside Africa” (Herskovits [1941] 1958, 124) and that “rural and urban Negro cultures took on somewhat different shadings” (135) with the darker peasants more African than their lighter, more acculturated urban brothers and sisters. I do not mean to suggest that Herskovits was a closet racist—which would be a rather cheap shot against a scholar whose progressive views were so ahead of his time.8 But it should remain clear how easily the language of race entered into the discourse of syncretism in the New World, particularly when the rhetoric of science was wedded to essentialized concepts of African culture in comparative studies funded by the Carnegie Corporation.9

22 / Chapter One

If the scale of intensities represented variable degrees and domains of African retention—high for religion, low for economics and art—it offered no explanations. To understand why some practices thrived when and where they did while others went underground or disappeared, Herskovits developed his general theory of syncretism, supplemented by concepts of reinterpretation, cultural focus, and what he called (no doubt echoing Malinowski [(1922) 1961, 20–21]) “cultural imponderables.” Narrowly defined, syncretism is produced in situations of contact between cultures from “the tendency to identify those elements in the new culture with similar elements in the old one, enabling the persons experiencing the contact to move from one to the other, and back again, with psychological ease” (Herskovits 1966, 57). Thus, he notes, in the Catholic New World, African gods are identified with saints of the Catholic Church, whereas in Protestant areas the religious associations are subtler, and African cultural retentions (e.g., mourning and shouting) less intense. We can perceive in this notion a strong psychological emphasis on the individual as syncretic agent, on identification as the syncretic process, and on adaptation and integration (“psychological ease”) as syncretic functions, which extend secondarily to groups in their new cultural contexts. The same process occurs, according to Herskovits, “in substance rather than form, in psychological value rather than in name” (1966, 57) when the resemblance between cultural elements is too weak to afford a fully syncretic relationship but is strong enough to allow a reinterpretation of the new by the old. Herskovits’s favorite example of reinterpretation is his claim that African polygyny was retained under monogamous constraints, with his arguing that polygyny was reinterpreted in diachronic terms by the practice of serial unions. This example reveals Herskovits’s sociological naïveté in downplaying contemporaneous social conditions in favor of imputed cultural continuities, but the principle itself, I will argue, is extremely salient when recast as a revisionary strategy. If syncretism and reinterpretation are mainly psychological concepts that explain how the new culture is adopted within the framework of the old, the concept of cultural focus shifts the analysis to culture as sui generis. To explain why African religious beliefs and practices in the New World are retained with greater clarity and vigor than, for example, kinship, economic, and political institutions, Herskovits argues that religion itself constitutes the cultural focus of African peoples—their “particular emphasis,” “distinguishing flavor,” and “essential orientation” (1966, 59). That which is given highest cultural priority by a people will offer the greatest “resistance” to change, he argues, and will thus rank high on the scale of New World Africanisms. Therefore culture, by way of its distinguishing focus, plays a determinative

Herskovits’s Heritage / 23

role in the selective process of what is and is not retained, in what form, with what degree of intensity, and in what sphere of social life. Thus, for Herskovits, [M]ore elements which lie in the area of focus of a receiving culture will be retained than those appertaining to other aspects of the culture, acceptance being greater in those phases of culture further removed from the focal area. When a culture is under pressure by a dominant group who seek to induce acceptance of its traditions, elements lying in the focal area will be retained longer than those outside it. (1966, 59)

We have no clearer commitment to (a relativized) cultural determinism than in this passage, which provides a rather strange take on the initial socioeconomic context of African culture contact—that of slavery in the New World. Elsewhere Herskovits ([1941] 1958, 86–109) was clearly aware of the different forms of plantation slavery and the modes of passive and active resistance that the slaves deployed, including foot-dragging, suicide, escape, marronage, and organized revolt. But here we are led to believe that religion—more than kinship, politics, or economics—persisted in a world that turned Africans into laboring chattel and destroyed their families because it served as the dominant cultural focus. Such a position seems to defy rational argument if not common political sense, displaying what Jackson identifies, in another context, as Herskovits’s “curious naïveté about the relationship between culture and power” (Jackson 1986b, 114). But even here, I will argue, lies the germ of an idea that helps to explain the power of syncretic practices in real and effective terms. I will argue that the hermeneutical principles of West African religions—particularly Yoruba religion, which has thrived in various New World guises—have provided salient forms of popular resistance in a variety of oppressive conditions. The final concept in Herskovits’s syncretic paradigm—that of cultural imponderables—introduces the variable dimension of consciousness in African diaspora research and prefigures the study of practical and embodied knowledge in current anthropology. For in addition to consciously retained Africanisms emanating from the cultural core, Herskovits discerned a range of “retentions” that “are carried below the level of consciousness” (1966, 59) and persist in everyday practices. These include the linguistic patterns of accents, dialects, and creoles; the musical styles of, for instance, son, rumba, mambo, and the blues; the “motor habits” of expressive gestures and dance; and the “codes of etiquette” that inform greetings and politeness formulae. At a time when such phenomena were often explained in racist terms, as

24 / Chapter One

transmitted through blood, Herskovits took great pains to emphasize their cultural character, as acquired by successive generations. More interesting for us, however, was his understanding that such culturally embodied imponderables persisted as retentions because they resisted change. Clearly, this notion of resistance is passive, attributed elsewhere to “the force of cultural conservatism” (Herskovits 1966, 57), and is conceived negatively, as the absence of assimilation. But as we shall see, when “updated,” this notion foreshadows theories of active resistance that identify bodily practices as contested sites of symbolic and ideological struggle (Comaroff and Comaroff 1991, 1:24–25). Thus reduced to its essentials, Herskovits’s syncretic paradigm highlights the limitations of American liberal scholarship. In retrospect, its major features—the racist overtones of the scale of intensities; the psychologistic orientation of syncretism and reinterpretation, which privilege adaptation and accommodation over opposition and contradiction; the absence of any sustained class analysis; the emphasis on an inertial cultural focus over and above the dialectics of power and identity construction; and finally, the essentially conservative vision of cultural retention as that which resists change—all seem to relegate Herskovits to a dubious past. Having dissected the syncretic paradigm to criticize its component parts, highlighting their limitations while flagging their redeeming features for later discussion, I will now turn to the paradigm as a whole in its more substantive applications, in order to interrogate the notion of cultural origins that informs the syncretic process.

Deconstructing Origins In Herskovits’s “African Gods and Catholic Saints in New World Negro Belief,” first published in American Anthropologist (1937), we find the official liberal blueprint of the science of New World syncretisms, based on fieldwork in Haiti as well as on published Cuban and Brazilian data. To be sure, Herskovits was one of the first North American anthropologists to recognize the validity—indeed the privileged status—of Caribbean societies as objects of ethnographic study, at a time when the discipline preferred “pristine” primitives in “natural” habitats to cultures reconstituted in “diluted” forms through slavery or other coercive dislocations. Such settings, he argued, provided “laboratory situations” (Herskovits 1966, 46) for investigating the dynamics of acculturation—of what is lost, modified, or retained through culture contact, and the mechanisms that govern the process. In this respect, Herskovits perceived the universal significance of New

Herskovits’s Heritage / 25

World syncretisms for the “science of man,” in that they provided the clearest cases, assuming adequate data, of more general principles of culture contact and change the world over. In this view, New World cultures moved from a marginal ethnographic status to center stage, and if the assumptions that guided his study appear simplistic and naive, distorting (as I shall now illustrate) the very data themselves, the larger conceptual revolution that he inspired linked empirical studies of the African diaspora to a general theory of culture. According to Herskovits, the syncretic identification of African gods with Catholic saints was shaped by two primary factors in the New World: by slavery as the dominant institution of social life (or, for Orlando Patterson [1982b], “social death”) and by Catholicism as the official religion of the masters. These two factors together account for the distinctive patterns of syncretism found in Haitian Vodou, Cuban Santería, and Brazilian Candomblé, in that the slaves were summarily baptized as they came off the ships and were thrust into sugar mills and plantations, where they secretly continued to worship their gods under the cloak of official Catholicism. Banned by the authorities, the African cults were forced underground, where they provided a focus for sporadic slave revolts, were uneasily tolerated during Catholic holidays, and fragmented into local groups that were mainly shaped and dominated by the personalities of their leaders (Herskovits 1966, 322). I do not have space in this limited review to discuss the complex social, political, and religious variations that historically unfolded, except to mention the most general cultural consequence identified by Herskovits. This is the profound fragmentation of aboriginal African unities, a forced fusion of different African cultures and the dismemberment of religious cult hierarchies into shattered splinter groups, “reflected in a resulting confusion of theological concept” (1966, 322–23). It is this model of cultural and theological fragmentation that I will challenge, focusing first on Herskovits’s tropes of aboriginal unity. Despite his call for rigor in identifying the numerous cultural origins of New World slaves, Herskovits reduces African influences to two principal sources—Fon and Yoruba.10 The deities identified with Catholic saints are limited—at least upon first inspection—to the pantheons of these two great West African cultures, such that Fon gods like the trickster Legba, the rainbow-serpent Damballa, and the Marassa twins syncretized with Saint Anthony, Saint Patrick, and the twin saints Cosmas and Damian, while Yoruba gods like the trickster Eshu, the thunder-god Shango, and the water goddess Yemoja (Yemanja) were similarly identified with the Devil, Santa Barbara, and the Virgen de Regla. The correspondences are based, we may recall, on the similarities of

26 / Chapter One

religious elements, such that the saintly icons depicted on Catholic chromolithographs exhibit symbolic features of African counterparts; thus, the serpents on Saint Patrick’s image invoke the Fon serpent-deity Damballa, whereas the twin saints Cosmas and Damian resemble the Fon Marassa or Yoruba Ibeji twins. Under repressive conditions of official opprobrium, the slaves—so the argument goes—were able to worship in two worlds at once; outwardly Catholic, inwardly they honored their African gods. Following this syncretic principle of identification, Herskovits constructed a table of correspondences between African gods and Catholic saints in Brazil, Cuba, and Haiti, listing African deities in a vertical left-hand column with their saintly counterparts to the right (table 1.2). Again, it is not the specific correspondences that I will challenge (because, as Herskovits perceived, they vary both regionally and over time), but the table itself as a form of knowledge and the assumptions embedded within it. For they amount to a specific discursive modality, a way of constructing African identities and differences, of figuring (or as Mudimbe might say, conjugating) Africanity in the New World as an acculturative process. And it is this discourse that can be fruitfully deconstructed, not abstractly from the lofty heights of postmodern criticism, but concretely, on the basis of internal evidence supplemented by empirical data from Nigeria. First, I would call attention to a subtle but powerful slippage subsumed by the category of the left-hand column—that of “African deities,” wherein several significant contrasts are neutralized. Following the logic of the syncretic paradigm, “Africa” refers to a West African baseline, an ethnohistorical reality circumscribed by space and time and identified as the source of African influence in the New World. Here the Fon and Yoruba figure as dominant cultural origins (with Congo receiving a passing reference associated with Haitian Simbi deities), in that either Fon or Yoruba deities can be identified with Catholic saints. But here also is where the cultural waters get muddied, for apart from a few central deities like the Fon Damballa and the Yoruba Ibeji, it is impossible to distinguish the two religious pantheons as culturally discrete. I will not recount the complex history of DahomeanYoruba political relations, except to mention that from at least the sixteenth century (and probably earlier) through the mid-nineteenth century, warfare, slave raiding, migration, and ritual reciprocity between the kingdoms of Dahomey, Ketu, and Old Oyo persisted, with much cultural mixing of religious deities and institutions (see Akinjogbin 1967; Law 1977; and Parrinder 1967). There is a general historiographic tendency to see Yoruba gods like Ogun (of war and iron) and Ifa (of divination) recoded in Dahomey as Gu and Fa, suggesting a regional Yoruba diaspora to the west. But one

Table 1.2. Correspondence between African gods and Catholic saints in Brazil, Cuba, and Haiti (from Herskovits 1966, 327–28) African deities as found in

Brazil

Obatala

Obatala; Orisala; Orixala (Oxala) Grand Mambo Batala Shango Elegbara; Elegua, Alegua Legba Exu Ogun Ogun Balandjo Ogun Ferraille Oxun Yemanja Maitresse Erzulie; Erzilie; Erzilie Freda Dahomey Saponam Osa-Ose (Oxossi) Ololu; Omolu Agomme Tonnere Ibeji (Brazil and Cuba); Marassa (Haiti) Father of the Marassa Orunbila (Orunmila) Loco Babayu Ayi Ifa Yansan (wife of Shango) Damballa Father of Damballa Pierre d’Ambala loa St. Pierre Agwe Roi d’Agoueseau Daguy Bologuay La Sirene loa Christalline Adamisu Wedo loa Kpanyol Aizan Simbi Simbi en Deux Eaux Azaka Meda ’Ti Jean Petro

Cuba

Haiti

Virgen de las Mercedes; the Most Sacred Sacrament; Christ on the Cross “Nosso Senhor do Bomfim” at Bahia; St. Anne; “Senhor do Bomfim” at Rio (because of influence of Bahia) St. Barbara at Bahia; St. Michael the Archangel at Rio; St. Jerome (the husband of St. Barbara) at Bahia

St. Anne St. Barbara “Animas benditas del Purgatorio”; “Anima Sola” St. Anthony; St. Peter

The Devil St. George at Rio St. Jerome; St. Anthony at Bahia

St. Peter St. James the Elder; St. Joseph St. James

Virgin Mary; N.D. de Candeias Virgin Mary; N.S. de Rosario (at Bahia); N.D. de Conceicao (Rio)

Virgen de la Caridad del Cobre Virgen de Regla Holy Virgin (of the Nativity); St. Barbara; Mater Dolorosa

The Sacred Sacrament St. George at Bahia; St. Sebastian at Rio St. Bento

St. Alberto; St. Hubert St. John the Baptist St. John the Baptist St. Cosmas and Damien

St. Cosmas and Damien

St. Nicholas St. Francisco St. Francisco St. Lazarus The Most Sacred Sacrament St. Barbara (wife of St. Jerome) St. Patrick Moses St. Peter St. Peter St. Expeditus St. Louis (King of France) St. Joseph The Assumption; N.D. de Grace St. Philomena St. Anne N.D. de Alta Gracia Christ (?) St. Andrew St. Andrew the Hermit St. Andrew (?) St. Anthony the Hermit

28 / Chapter One

equally finds the Yoruba trickster Eshu referred to as Eshu-Elegba as far east as the Ekiti region of Yorubaland, suggesting a complementary infusion of Fon deities into Yoruba pantheons. My point is not to argue which gods came from where—a possible and quite valuable regional exercise within limited terms—but to challenge the aboriginal purity of Herskovits’s tribal baseline. In brief, Fon and Yoruba are not pure cultural categories. Indeed, the very notion of a singular Yoruba people was a missionary invention of the mid-nineteenth century, subsuming Egba, Egbado, Ijebu, Ijesha, and Ekiti peoples, among others, under a standardized Oyo-Yoruba linguistic and cultural model. Whether wittingly or not, Herskovits avoids this problem of interposed origins by lumping them under the general category of African. His table of correspondences erases the difference between Fon and Yoruba, which remains implicit in the names of certain deities, but which also remains highly ambiguous. The baseline is thus occluded by the trope of aboriginal Africa, grounding a primordial cultural genealogy that quickly vanishes into an unknown past. Nor does systematic slippage stop here. If we examine the left-hand column further, we find “African” deities such as la Sirène, loa Christalline, loa St. Pierre, loa Kpanyol (the Spaniard), Maitresse Erzulie, and ’Ti Jean Petro (little John Petro), deities that never did or could exist in precontact West Africa because they represent European mythic and historical allusions and social stereotypes. Part of this problem lies with Herskovits’s failure to distinguish what he as an external observer and trained anthropologist calls African from what Haitians, Brazilians, and Cubans call African. The table of correspondences confuses both perspectives under “Africa,” merging the etic with the emic. Thus, Herskovits notes The Haitian . . . does not merely stop at identifying the saints with African gods, for saints are occasionally themselves conceived as loa . . . thus St. Louis, the patron of the town of Mirebalais where this field work was carried on, is a loa in his own right. Similarly two of the kings who figure in the image that depicts the Adoration of the Christ Child, Balthazar and Gaspar, are also held to be vodun deities. (1966, 325)

Before proposing a more critical solution to this problem, it is enough to point out that Herskovits has failed, in the terms established by his syncretic paradigm and table of correspondences, to clearly distinguish cultural origins and African deities from their reconstructed and syncretic forms. The ethnohistorical baseline remains a myth of African origins, not a

Herskovits’s Heritage / 29

documented or even documentable point of empirical departure. This myth is significant not only as a foundational fiction, but because it was elaborated by Herskovits in substantive claims that have continued to misguide much New World research. This brings me to the second tropic function of aboriginal unity, that of coherent, unified, and standardized theologies and pantheons in West Africa that were uprooted and fragmented through the slave trade to be reconstituted in locally variable and confused forms in the Catholic New World. I call this elaboration tropic because it establishes a nostalgic topos of a theological Golden Age that never existed in West Africa and should be abandoned as a comparative standard for studying New World religions. The general idea derives from Herskovits’s discovery that Vodou theology is highly inconsistent. He found differences of opinion not only from region to region, but within a given region even between members of the same group concerning such details of cult belief and practice as the names of deities, modes of ritual procedure, or the genealogies of the gods, to say nothing of concepts regarding the powers and attributes of the African spirits in relation to one another and to the total pantheon. (1966, 323)

Eliciting lists of deities from a single Haitian valley, he discovered that “the differences between these lists were much greater than the resemblances; and . . . in identifying deities with Catholic saints, an even greater divergence of opinion was found” (323). The same indeterminacies and patterns of variation apply, on a larger scale, to the African pantheons and their syncretic manifestations in Brazilian Candomblé and Cuban Santería, summarized by the table of correspondences. It is not the indeterminate character of Herskovits’s data that I would challenge but his inference that a greater uniformity ever existed in West Africa. It is in fact the very idea of listing African deities as discrete mystical entities, in fixed relations within pantheons associated with stable sociological correlates on the ground, that my own research on Yoruba òrìṣ à worship undermines. Yoruba deities are not only vested in lineages (what Bascom, following Herskovits, called “sib-based” cults), but articulate with more inclusive corporate groups (Bascom’s “multi-sib cults”), such as quarters (àdúgbò), each ruled by a town chief, or in the case of royal cults, the town or kingdom as a whole (ìlú). Thus, in a crude sense, the ritual configuration of òrìṣ à cults within any kingdom represents its dominant relations of political segmentation, the patterns of which vary spatially, from one kingdom to

30 / Chapter One

another, and within kingdoms over time, accommodating (and on important occasions, precipitating) political fission, fusion, or the reranking of civil chiefs (see Apter 1992; Barber 1981; and Bascom 1944). The methodological implication of this politico-ritual complementarity—a very gross reduction of a complex dialectic—is that no two Yoruba kingdoms arrange their pantheon of òrìṣ à in the same way. The situation is further complicated by the fact that within kingdoms, the òrìṣ à cults of different town quarters organize their pantheons around their own principal deities, so that if, officially, a civil chief pays ritual obeisance to the superior òrìṣ à of the town king, secretly, within the confines of his own quarter’s cult, the chief and his followers recognize the hidden paramountcy of their òrìṣ à, around which their pantheon revolves (Apter 1992, 149–61). The cosmological principles that render such polyvocalities possible and intelligible are grounded in Yoruba notions of “deep knowledge” (imọ jinlẹ̀ ), referring to the privileged access of powerful priests and priestesses to hidden truths and secrets. I will return to the power of such knowledge in due course. For now, it is enough to point out that within this ritually safeguarded space of interpretive possibilities, official dynasties and genealogies are revised, deities are repositioned to express rival political claims, and the deities themselves are fragmented and fused into multiple and singular identities. Small wonder that Herskovits had trouble with his lists, for even in Nigeria, no òrìṣ à cult, community, or Yoruba kingdom (let alone two individuals) would produce the same list or pantheon of òrìṣ à. From this stems a second dominant misconception—that òrìṣ à cults in West Africa represent discrete deities, with one cult worshipping Shango, another Yemoja, a third Obatala, and so on. In fact, all òrìṣ à cults house clusters of deities that are represented by specific priests and priestesses, altars, and sacrifices and are grafted onto an apical deity.11 Within these microarenas, the configuration of the clustered deities also shifts with changes in the status of their associated lineages and titled representatives and according to contesting claims from within. Under these conditions, no definitive list of deities is possible. More significant for syncretic models is the mistaken claim— and here Bastide and Verger keep company with Herskovits—that formerly discrete cults in Africa were restructured in the New World to house a multiplicity of African deities and Catholic saints. The ethnohistorical record reveals that Yoruba òrìṣ à cults were never discrete in the first place. Zora Neale Hurston was one of the first to grasp the elusive polymorphism of the Haitian loa in her more personal (and in many ways protoexperimental) ethnography of Vodou, understanding that “No one knows the name of every loa because every major section of Haiti has its own

Herskovits’s Heritage / 31

variation” ([1938] 1990, 114), and that the loa themselves are both multiple and singular. My point has been to extend this indeterminacy back to southwest Nigeria, where Yoruba religion reveals much greater continuity with its syncretic manifestations than Herskovits ever imagined. The theological confusion in New World cults and pantheons that for Herskovits resulted from the upheavals of slavery is actually endemic to òrìṣ à worship (and I suspect to Fon religion as well) and resolves into a critical hermeneutics of power, once its relevant dimensions are grasped. In thus reformulating Herskovits’s ideas, we will see that he was onto something very important. He concluded his seminal comparison of Brazilian, Cuban, and Haitian syncretisms with the observation that, despite the confusion surrounding African deities and their Catholic correspondences, a general syncretic process was at work: Considered as a whole . . . the data show quite clearly to what extent the inner logic of the aboriginal African cultures of the Negroes, when brought in contact with foreign traditions, worked out to achieve an end that, despite the handicaps of slavery, has been relatively the same wherever the forces for making change have been comparable. (Herskovits 1966, 328, emphasis added)

It is this inner logic, recast as a cultural hermeneutics, that provides the key to understanding the West African contribution to New World syncretic forms and does so without recourse to foundational fictions of essentialized aboriginal unity.

From Syncretic to Critical Practice Thus far I have deconstructed Herskovits’s myth of African origins—its figures of cultural purity, theological unity, lineage-based cult organization, and cult singularity (that is, one deity per ritual collectivity)—in order to extract the interpretive kernel from its ideological shell. If Herskovits distorted the West African baseline with misconceptions that can be scrapped or readjusted, he also established the ground of a cultural argument that can be further developed. The goal of this argument, as I mentioned earlier, is to determine what is African in the African diaspora, focusing on religious syncretism as a clear case of Africanity in order to theorize its subtler forms of influence in the New World. First, however, we must acknowledge that the trope of “Africa” has served as a dominant ideological category in the service of empire, a category that has naturalized, as Mudimbe (1988) so cogently demonstrates,

32 / Chapter One

the normative and territorial dominions of Europe’s “civilizing mission” (see also Comaroff and Comaroff 1991, 1:86–125). But it also must be emphasized that a deconstruction of this master trope, as Mudimbe’s radical project demands, does not do justice to the other side, to that which lies beyond the ideological limits of an Africa produced by missionary-colonial discourse. This other side of Africa must not be taken to refer to the fiction of pristine cultures stipulated by Herskovits, but neither is it reducible to some unknowable Other in a move that annihilates the very histories of peoples who have come to define themselves as Africans with specific national and ethnic identities. If African worlds are as much the constructs of Africanist discourses as the objects of their inquiry, then it is within this dialectic of invention and observation that the concept of syncretism performs a double synthesis. In brief, Africa is assimilated to the New World through culture contact if and only if an invented Africa is assimilated to an Africa observed. We are caught in a double bind. Either we essentialize Africa or renounce it. One way out of this ideological dilemma is to focus on the inner logic of syncretic practices as strategies of appropriation and empowerment. What Haitian Vodou, Brazilian Candomblé, and Cuban Santería have in common is a history of accommodation and resistance, not merely in the cultural terms of allying uneasily with Catholicism but also in the political contexts of class division and the state. It is precisely this relation between implicit social knowledge and political economy—what in my Yoruba research has emerged as a hermeneutics of power—that defines the horizon of Africanity in the New World: not as core values or cultural templates but as dynamic and critical practices. Nowhere is this critical relation between forms of knowledge and relations of domination more evident than in the history of Haitian Vodou. The conventional historiography traces a grand development from the late eighteenth century to the regime of François Duvalier (“Papa Doc”), over the course of which Vodou’s original revolutionary impulse was gradually co-opted by the state. Thus, in 1791 rites conducted by the famous houngan (priest) Boukman inspired the first organized blow against the French plantocracy (see Courlander 1966). According to Haitian historical memory, it was in a clearing of the Bois Caïman that, “under a raging tropical downpour accompanied by lightning and the cracking of giant trees, [Boukman] performed a Petro ceremony” in which “a pig was sacrificed and its blood, mixed with gunpowder, was distributed among participants to strengthen their will to win” (Bastien 1966, 42).12 The revolutionary triumvirate of Toussaint L’Ouverture, Jean-Jacques Dessalines, and Henri Christophe were quick to capitalize on Vodou’s popular appeal and its secret channels of

Herskovits’s Heritage / 33

communication in mobilizing the masses. Hence it is generally accepted that Vodou played a strategic role in achieving Haitian independence. After the revolution’s success in 1804, however, Vodou played into the hands of both the center and the opposition. In his efforts to stabilize the new government, Dessalines allied with the mulatto elite and tried to foster Catholicism. He was murdered in 1806 and then resurrected as a loa, thereby incorporated into the very pantheon of deities that as official leader he had grown to oppose. His successor, Christophe, suffered a similar reversal. As king of northern Haiti until his suicide in 1820, he encouraged Catholicism as the official religion of state administration and alienated himself from the peasantry. Subsequent Haitian leaders, such as the self-proclaimed emperor Faustin Soulouque (1847–59), cultivated alliances and reputations among the houngan and their followers while paying official lip service to Catholicism. Under Antoine Simon (1906–11), for example, the National Palace was recognized as a sacred site of Vodou activities. Thus, in the first century of Haitian independence, Vodou articulated with a complex set of emerging political and class divisions. Centralized government favored Catholicism as the official religion of administration, based in the urban centers of Cap Haïtien, the northern seat of black elite power, and in Port-au-Prince, the southern seat of the mulattos. Vodou remained a predominantly rural religion of the black peasantry, providing a powerful political resource for black leaders like Soulouque, who could play the peasantry against the mulatto elite by invoking the popular religion of the people. In this postcolonial context, the religious syncretism of Vodou and Catholicism played into the dialectics of class and color stratification. As Catholicism spread into the countryside and merged with Vodou, the latter seeped, as it were, into the palace, providing unofficial access to leadership and state power. This trend was not constant, since Vodou was periodically attacked by mulatto leaders who allied with the Vatican after the concordat of 1860 and waged antisuperstition campaigns against the houngan (Ramsey 2011). But these campaigns only reinforced the black opposition, which eventually united under Duvalier into a deadly combination of Vodou, noirisme, nationalism, and state power. The stage for Duvalier’s consolidation was prepared by the US occupation of Haiti (1915–34), which inspired Haitian intellectuals like Jean Price-Mars to rediscover within Vodou the “genius” of the Haitian people in his celebration of négritude. Formerly disdained by the educated elites, Vodou was suddenly elevated in the respectable language of poetry and folklore to the status of national heritage and identity (Dayan 1998). Leftist intellectuals allied with the peasantry to demand an authentic and truly autonomous

34 / Chapter One

Haiti. It was in such an atmosphere that Papa Doc appropriated Vodou to consolidate dictatorial control. As a peasant religion, Vodou mobilized enough popular support to counterbalance mulatto and church opposition; as a symbol of Haitian nationalism, it appealed to leftist intellectuals and noiristes; and with its underground network of secret Bizongo and Secte Rouge societies (W. Davis 1988, 241–84), it provided an ideal channel for administering state power and terror while effectively dividing all organized opposition to the self-proclaimed president-for-life. It appears that the history of Haitian Vodou is a history of popular resistance and state appropriation—of the high and official appropriating the low and popular—in that the religion that originally inspired revolution came to uphold a dictatorial state. In a basic sense this reading is true, but what is lost in such an instrumental interpretation is how Vodou and its associated notions of Africanity (Guinée) have mediated the complex dialectics of political competition and class division (a theme I pursue further in chapter 2). To be sure, Vodou remained a powerful resource in the hands of revolutionary leaders and shrewd politicians and has clearly made a difference in Haitian history. But what has made Vodou so powerful? Strategic explanations that Vodou mobilized collective action still beg the question of why it was so politically effective in the first place. At this point, I would like to tie the various threads of this discussion together, rethinking syncretism in the African diaspora as a critical and revisionary practice, one that reconfigures dominant discourses with variable, and at times quite significant, consequences. Haiti provides the clearest illustration that resistance waged through syncretic struggle—through the appropriation by African powers of Catholic saints, postrevolutionary kings, and nationalist rhetoric—was more than symbolic wish fulfillment. But it also illustrates the other side of syncretism, in that the dominant categories that were semantically revised also were, in more formal terms, reproduced and perpetuated.13 If Vodou took possession of Catholic hierarchies through the very gods that possessed their devotees, it also reproduced the authoritative structure of God the Father and his saintly messengers, disseminating popular Catholicism throughout the countryside. It was this double aspect of syncretism that Herskovits identified as an acculturative process, as the uneasy adaptation of cultures in collision. What Herskovits missed was the critical relation between cultural form and hegemony, although he intuited the variable modalities that this relation could take. Returning to the classic syncretism of African gods and Catholic saints, we can recast its historical genesis as a grand counterhegemonic strategy. What Herskovits perceived as a psychological mechanism of cultural integration,

Herskovits’s Heritage / 35

allowing blacks to move between African and colonial orders with relative conceptual and emotional ease, was in fact a much more powerful process of discursive appropriation. If in Haiti, as in Cuba and Brazil, the dominant discourse of Catholicism baptized Africans into slavery, it also was Africanized through syncretic associations to establish black nations, identities, and idioms of resistance. The role of Vodou in the Haitian revolution may stand out for its remarkable impact, but parallel developments occurred throughout the New World. Thus, in Brazil the quilombos and mocambos, or black republics of escaped slaves, began as religious protest movements that Africanized Portuguese Catholicism along various ethnic lines. Palmares, the largest and most famous of the quilombos, recreated Bantu models of social organization and government, combining African effigies with Catholic icons in its shrines as early as 1645 (Bastide [1960] 1978, 83–90). And in Cuba, the 1844 slave revolt called La Escalera—so named to commemorate the ladders to which the vanquished slaves were tied and tortured—grew out of “an elaborate conspiracy in Matanzas, organized through the cabildos and drum dances of the sugar estates, the ‘king’ and ‘queen’ of the weekly dance being the agents of conspiracy” (Thomas 1971, 205). I mention these famous uprisings not merely to illustrate a strategic relationship between slave religion and organized revolt but to argue that the power of syncretic revision was real and that when conditions were right, the African communities thus imagined and organized asserted themselves with considerable impact. The syncretic revision of dominant discourses sought to transform the authority that these discourses upheld. To be sure, radical ruptures were exceptional and stand out in Caribbean history as memorable flashpoints in the perduring black struggle. But the general point I wish to emphasize is that the power and violence mobilized by slave revolts and revolution were built into the logic of New World syncretism itself. The Catholicism of Vodou, Candomblé, and Santería was not merely an ecumenical screen, hiding the worship of African deities from official persecution. It was the religion of the masters, revised, transformed, and appropriated by slaves to harness its power within their universes of discourse. In this way the enslaved took possession of Catholicism and thereby repossessed themselves as active spiritual subjects. Nor was this revisionary strategy specific to slavery; it developed also under subsequent conditions of class and color stratification, among black rural peasantries and urban proletariats (see Bastide [1960] 1978). The political dimensions of such syncretic revision began not with social protest and calls to arms but with the unmaking of hegemony itself. As Carnival and possession rituals so clearly illustrate, hegemony is unmade by reversing high and low categories with blacks above whites in the Africanized

36 / Chapter One

streets and shrines, by recentering Catholic hierarchies around African gods, by reinscribing ritual space with palm fronds, crossroads, and kingly thrones, by marking time to different drum rhythms and ritual calendars, and by liberating the body from its disciplined constraints (see Alonso 1990; DaMatta [1979] 1991; Parker 1991). Possession by spirits—which include Catholic saints as well as African deities, for even these two orders ritually collapse— involves sexual transgression and gender crossing because it transcends and transforms the most fundamental categories of the natural and social worlds. But if hegemony is unmade through syncretic ritual, it is also remade, and it would be wrong to equate its religious impulse with protorevolutionary struggle pure and simple. As noted earlier, the ritual revision of dominant discourses also reproduces their grammar and syntax, which it reconstructs from below. In Vodou, this unmaking and remaking of hegemony corresponds to two sets of spiritual powers: the cool Rada deities of the right hand (often traced back to Allada in Dahomey), who sanction authority, and the hot Petro (Petwo) deities of the left hand (identified as chthonic), who lampoon and decenter the status quo. In Brazilian Candomblé and Cuban Santería, as in Yoruba òrìṣ à worship, both types of power inhabit one general pantheon and associate with cool and hot deities such as Obatala and Shango. One can trace the permutations of this basic opposition through innumerable examples, but the point I wish to highlight is that syncretism necessarily involves both the unmaking and remaking of hegemony and thus is intrinsically political. Returning to Herskovits’s syncretic paradigm and locating it within a context of cultural hegemony, we can reduce its basic concepts to a more general dialectic of revision and reproduction. Those things Herskovits reified into categorical distinctions—between syncretism proper, reinterpretation, cultural focus, and embodied forms of expressive culture—reflect variable modalities of cultural resistance, not in his passive sense of resisting change but actively, as counterhegemonic strategy. By appropriating the categories of the dominant classes, ranging from official Catholicism to more nuanced markers of social status and cultural style and by resisting the dominant disciplines of bodily reform through the “hysterical fits” (Larose 1977, 86) of spiritual possession, New World blacks empowered their bodies and souls to remake their place within Caribbean societies. As we have seen, the material consequences of these revisionary strategies range from negligible to revolutionary: the spiritual nationalism asserted by Vodou sword-flags brandished before the National Palace; the self-conscious nationalism of Jean Price-Mars, reclaiming Vodou as the model of négritude; the Haitian revolution itself, for a few examples. And as we also have noted, the power

Herskovits’s Heritage / 37

of revisionary challenges from below could be reappropriated by the elites, in the academic folklore of Fernando Ortiz and Gilberto Freyre, or in the Machiavellian statecraft of François Duvalier. There is no single trajectory of exalted class struggle built into syncretic forms of revision and resistance, or vice versa (as has been suggested for Haiti). What concerns us is the hermeneutics of revision as such and the interpretive conditions of its possibility.

This final concern brings us back to our initial inquiry into what is properly African in the African diaspora. I have deconstructed Herskovits’s essentialized cultural baseline, its trope of an aboriginal Golden Age, and its attendant reifications of cultural purity and dilution, without renouncing the logic of cultural genealogies. I will conclude by making my position explicit, by establishing the historically critical relationship between West Africa and the New World. Boldly stated, the revisionary power of the syncretic religions derives from West African hermeneutical traditions that disseminated through the slave trade and took shape in black communities to remake the New World in the idioms of the old. It is not the elements of Old and New World cultures that should be meaningfully juxtaposed in the concept of syncretism—as Herskovits maintained—but the orthodox and heterodox discourses in which such elements have been deployed and the tropic operations that they have performed. I have dwelt perhaps excessively on refiguration and revision because these are the strategies that have made, and continue to make, a difference—rhetorical, pragmatic, and, in key moments, political—among blacks, mulattos, and whites in the Americas (see Gates 1988). These also are the discursive strategies that characterize West African religions, particularly Yoruba religion, which has had a long history of reconfiguring hegemony, documentable from the rise and fall of the Old Oyo empire (1600–1836), through the nineteenth century Yoruba wars, to the appropriation of Christian and colonial rhetoric in Nigeria’s long march to independence. Thus West Africa’s contribution to the African diaspora lies not merely in specific ritual symbols and forms, but also in the interpretive practices that generate their meanings. In Yoruba cosmology, for example, deep knowledge (imọ jinlẹ̀ ) has no determinate content but rather safeguards a space for opposing hegemony. Sanctioned by ritual and safeguarded by secrecy, deep knowledge claims are invoked to revise dynastic genealogies, the rankings of civil chiefs, and even the relative positions of deities within official pantheons. Deep knowledge by definition opposes public discourse, and the authoritative taxonomies that it upholds—whatever they may be. If this is what

38 / Chapter One

has made West African religions powerful in relation to local, colonial, and postcolonial hegemonies, it also has informed syncretic revisions of dominant hierarchies in the New World, incorporating them within more popular pantheons and cosmological fields of command. The concepts of syncretism, reinterpretation, and cultural imponderables, which for Herskovits distinguished types of African retentions, are recast in my argument as modalities of revision and resistance. I have traced them back not to a pristine cultural baseline but to a dynamic variety of West African interpretive strategies, thereby revising Herskovits’s concept of cultural focus into a more critical concept of cultural hermeneutics. If I seem to have succumbed to the indeed substantial hegemony of Yoruba chauvinism in black diaspora debates, it is not to assert that Yoruba cosmology has had the greatest impact in the New World, although its impact has been and remains profound, but because its hermeneutical principles of refiguration and revision are so clearly at work in the classic syncretic religions and illuminate their power. I have restricted my discussion to Herskovits’s ethnohistorical project (which, if groundbreaking in its time, appears narrow next to current research on colonial mimesis, public culture, and transnational identity) because within this more global set of issues, it reminds us that even after they are deconstructed, the Old World origins of the African diaspora can be recovered and their heritage explored in endless depth.

T WO

Creolization and Connaissance

The genealogy of morals in African diaspora research encounters the politics of race at nearly every turn. Implicated in racial ideologies, nationalist movements, and academic battles since its formal emergence in the 1930s, the field of Afro-American studies was politically charged from the start.1 In the United States, Carter G. Woodson (1936) and W. E. B. Du Bois (1939) produced the first programmatic studies of African cultural contributions to the Americas, enlisting the recovery of this history against Jim Crow and the color bar. Following their lead, Melville J. Herskovits recast Afro-American research in the more clinical language of social science, posing African acculturation in the New World as a central anthropological problem while exposing the myth of an “absent Negro past” as a symptom of racist denial (Herskovits [1941] 1958, 1966). Parallel developments in the Caribbean took various forms as African cultural practices were uneasily domesticated into inclusive idioms of national culture, sanitized as folklore in the work of Fernando Ortiz (1916), Jean Price-Mars (1928), and Gilberto Freyre ([1933] 1986), legitimated by ideologies of mestizaje and créolité (Creole identity), and animated by the more militant movements of noirisme and négritude. Despite their different contexts, moments, and commitments to empirical research, the scholars of this first generation of Afro-Americanists shared a basic notion of African origins that, uprooted and fragmented by the Middle Passage and plantation slavery, could be recovered and in some sense traced back to the motherland. It is not my aim to recount the history and politics of this scholarship in any detail—a history that includes French luminaries like Pierre Verger (1982) and Roger Bastide ([1960] 1978, 1971), négritude poets like Aimé Césaire and Léopold Senghor, and their intellectual followers, as well as the rise of black studies in the 1960s—but rather to return to the problematic

40 / Chapter Two

notion of origins that continues to haunt African diaspora research and motivate its debates. Unless critically deconstructed, the idea of African origins is decidedly out of favor. On methodological grounds alone, criteria for establishing African provenance have remained controversial since the Herskovits-Frazier debate, demanding strict functional correspondences that can never realistically be found (Smith 1957), involving essentialized tribal designations that should be abandoned (Mintz and Price [1976] 1992), or invoking a play of tropes within a historically situated discursive field that can never be transcended (Scott 1991, 1997; see also Palmié 2013). According to these methodological strictures, African cultural practices may well exist in the Americas, but they cannot be known with any specificity. They lie beyond the limits of anthropological reason. In this chapter, I return to the locus classicus of origins in African diaspora research to defend the taboo position that they illuminate the dynamics of creolization. I am not arguing for a return to the first-generation notion of origins, but for a critical reappraisal. African origins are indeed ideologically motivated, rhetorically structured, and historically situated within genealogical narratives and discursive fields. Historically, they have evoked essentialized tribal and ethnic designations that are today untenable. But when appropriately theorized and located, they continue to illuminate African diaspora research. To show how such an invocation of origins can work, I address a fairly narrow debate among scholars specializing in studies of Haitian Vodou. The debate concerns the category of Petwo (petro) deities or spirits, their Creole status within the Vodou pantheon, and the sense in which they have come to oppose “African” gods. My analysis deepens the previous argument of chapter 1 about syncretism in the African diaspora by applying it to the complexities of a specific case. First, I will briefly review the revisionary logic of Yoruba deep knowledge, locating its power and indeterminacy within the dialectics of kingship in Yorubaland. Second, I will apply this deep-knowledge paradigm to the division between Petwo and Rada deities within Haitian Vodou to see how the Yoruba model informs the genesis and structure of this important opposition. From the standpoint of deep Yoruba ritual strategies, the radically revisionary Petwo powers—which one might expect to be associated with African origins—are paradoxically associated by Haitians with Creole deities. In the third section, I account for this paradoxical association of the most powerful with the least African mystical agencies and resolve the Petwo paradox. Against the ideology of Haitian Vodou and its misleading influence on various scholars, a Yoruba–Dahomean cultural hermeneutic reveals the West African origins and revisionary principles of the Petwo and Rada opposition

Creolization and Connaissance / 41

itself, as it emerged before the Haitian revolution and realigned with class relations under François Duvalier.

Deep Knowledge My revisionary approach to New World syncretism developed out of my previous research on Yoruba ritual and politics in Nigeria. Two general questions motivated my initial Yoruba project. The first concerned the extraordinary variation in orisha cult organization and hierarchy throughout Yorubaland—including relations between deities, rival pantheons, and their corresponding sociopolitical bases. As the extensive Yoruba literature attests, the orisha are associated with historic kingdoms that have local, regional, and Pan-Yoruba deities only partially rooted in lineages.2 The deities themselves are similarly mercurial, defying conventional classifications into ancestors, culture heroes, or so-called nature spirits by variably combining the attributes of each (Barber 1981; Horton 1983). Through a comparison of three Ekiti Yoruba kingdoms—two loosely federated as “village clusters” and one highly centralized as a former military autocracy—I sought the political factors that controlled such variations. Complementing this structural attention to cult organization and context, I focused on festival cycles and performances in order to account for the power of ritual practices. The festivals of the orisha were clearly associated with kingship, through ritual investment and mimetic appropriation, but because of the risky thresholds and danger zones that were navigated, and the collective anxiety that such passages generated, more than mechanical legitimation was involved. Reduced to essentials, my research revealed that political segmentation, not lineage organization as such, governed the variable bases of orisha cults, which generally conformed to three levels: that of the kingdom (ìlú) with its royal cult; the ward or quarter (àdúgbò), each governed by a civil chief forming a dominant segment within the town; and the lineage (ìdílé), each governed by an elder (baálé) from within the quarter, constituting the lowest level of political segmentation (fig. 2.1). Such a sketch provides an ideal type of a political logic that is more complex on the ground because political organization also includes subquarters, age sets, and lineage segments, as well as subordinate towns and hamlets. It is this underlying political logic that accounts for orisha cult variation and change. Briefly stated, the relative status of an orisha within a kingdom corresponds to its political location within this scheme; Ogun as a lineage orisha in one town would be a minor force in comparison with Ogun associated with the kingship in a neighboring kingdom. More significantly, as the social histories of the orisha cults

42 / Chapter Two

Figure 2.1. Power and authority in Yoruba government.

reveal, minor orisha can rise up with political factions that in some cases effect dynastic coups, pushing the exiled ruling line and its associated orisha to the sociopolitical margins. Royal cults are inclusive and so can absorb civil and lineage deities within their ritual fields, incorporating them into the rooms and altars of the town shrines. Town shrines thus house clusters of deities.3 Only the lineage orisha of particular compounds (agbo ilé) maintain a discrete relation between deity and shrine. Expressed in more dynamic terms, the political organization of Yoruba kingdoms highlights competitive fields of power and authority that pit chiefs against each other in promoting their respective jurisdictions or unite them against the king, whose power can be checked only by collective veto. Power competition operates within the transformative limits of fission, through which a royal prince or ambitious chief breaks away to establish an independent kingdom, and deposition, through which the excesses of the ọ ba (king) can be constitutionally checked. Within these limits, power transforms authority structures by usurping the kingship, revising dynastic genealogies, reranking hierarchies of civil chiefs and quarters, dividing kingdoms, and violating the sociopolitical order with variable degrees of impunity. Power thus unleashed is dangerous and labile, promising unlimited agency to the ambitious and challenging those in control. Power sui generis is transgressive and transformative, exceeding boundaries, subverting structures, even turning hierarchies upside down; it must be harnessed and domesticated,

Creolization and Connaissance / 43

contained by authority structures and channeled for the collective good. Power must be cooled, centered, and properly oriented. Here lies the work of the orisha, whose rituals negotiate the margins of power and authority to mediate competition and regenerate the kingdom. This characterization is itself a hybrid, merging the political frameworks of M. G. Smith (1956; 1975, 29) and Peter C. Lloyd (1968) with Yoruba models and representations of government to illuminate the dominant idioms and challenges of Yoruba ritual. As a mechanism of conversion, Yoruba ritual domesticates power, bringing it from the bush into the center of the town where the king is recrowned (Apter 1992, 97–116). In orisha festivals owned by chiefs, ritual empowerment deploys icons of kingship to signal the potential king that lies within every person, and who—when conditions are right—can rise up and take control. Ritual is risky because if its mechanisms fail or fall to rivals, unexpected transformations of authority can take place. Its outcomes are uncertain. It provides a space of mediation and maneuvering in which submerged factions can seize control. If in formal terms such a concept of power is transgressive, transformative, and pitted against the rule-governed hierarchies of administrative authority, then in Yoruba terms it is hot, polluted, and dangerous, a pure potency that must be purified, cooled, and contained. Thus sketched, the power of Yoruba ritual should be understood concretely, as transformative collective action. To be sure, ritual empowerment is generally reproductive, restoring the body politic to the status quo ante, but it is not always so. It is the ruptures, rebellions, and reconfigurations of these historic exceptions that illuminate the levels of realpolitik at play. In the most basic Yoruba ritual terms, transformative power versus reproductive authority are manifest in two basic categories or families of orisha, corresponding to “cool” and “hot.” The cool òrìṣ à funfun (white deities) such as Yemoja, Oshun, Olokun, and Obatala, associated with water, cool rhythms, fertility, and integration, are opposed to the hot, or warrior, deities like Shango and Ogun, whose staccato rhythms and explosive choreographies invoke legendary associations with lightning, fire, war, demolition, differentiation, death, and even immolation. As I have argued (Apter 1992, 154–56), any orisha can serve in both capacities, as hot and cool, reproductive and transformative. It is not the deity as such but the categorical opposition of their agentive attributes that is most important for my purposes in this chapter. To deepen understanding of the power of ritual as mobilized by hot and cool deities alike, I approach a closely guarded realm of secrecy safeguarding the imọ jinlẹ̀ (deep knowledge) of the cults. Two aspects of secrets are especially relevant. The first concerns their

44 / Chapter Two

subversive character as icons and indices of sociopolitical revolt. The dominant symbol of the royal Yemoja festival in Ayede, for example, is the calabash (igbá Yemoja) of concentrated ritual potency (àṣ ẹ ) that is carried—balanced on the Yeyeolokun’s (high priestess’s, lit. “grandmother of Olokun’s”) head—from the bush to the palace, where it empowers the king’s person and revitalizes the body politic. Like any dominant symbol, it embraces a span of meanings ranging from explicit normative community blessings (“it brings children and wealth, keeps the king healthy”) to implicit, forbidden themes of division and bloodshed, and it is this latter pole that is powerful and deep. Yemoja’s fructifying calabash represents the womb of motherhood, the head of good destiny, the crown of the king, the integrity of the town, even the cosmological closure of sky and earth. But its surfaces are decorated with signs of a deadlier power within, indicated by red parrot feathers—signs of ritual negation. Evoking the witchcraft of the priestesses and their mechanism for deposing the king, red parrot feathers on the calabash simultaneously assert a broken womb, miscarried delivery, bad destiny, and a headless (and crownless) king, as well as political fission and a cosmos out of control. Such negative themes are rarely voiced in public, but they nonetheless constitute a repertoire of potential interpretations that under certain conditions can be invoked so as to mobilize opposition against the status quo. The deep knowledge of royal ritual actually involves the king’s sacrifice and rebirth, whereby his icons of personal power and royal authority are literally taken apart and reassembled by authorized priests and priestesses, culminating in the crowning moment of Yemoja’s calabash. In the case of nonroyal festivals, the orisha’s calabash serves as a potential crown to remind the king that his chiefs can always rise up and take over. Such themes are enhanced by the various genres of ritual speech, which invoke repressed histories and veiled warnings of former kings and warriors who can prevail again (Apter 1992, 117–48; 1998). The dominant visual and verbal tropes that express these themes include those of inversion (e.g., in the image of a capsized canoe), reversal (e.g., from right to left), and mimetic appropriation (e.g., of European crowns). The latter symbolic function is particularly relevant to bringing outside icons of power within local fields of ritual command, absorbing symbols of foreign value and authority through metaphoric and metonymic associations (see endnote 29). But if this aspect of deep knowledge invokes fission, usurpation, and militant dismemberment, it does so through mechanisms of formal opposition to received historical and genealogical charters. The deeper one goes, in a sense, the less fixed and determinate the character of the secret, and the more formal the mechanisms of reversal and inversion. Ultimately, the secret

Creolization and Connaissance / 45

behind the secret is that deep knowledge has no content at all but derives its power from context-specific opposition to the authoritative discourses that it implicitly challenges.4 Like Griaule’s (1965, x) discussion of Dogon esoterica, a salient distinction between exoteric (“paroles de face”) and esoteric (“la parole claire”) knowledge is here at play (see also Griaule and Dieterlen 1991, 55), but unlike Griaule, I maintain that the deep dogma has no fixed content into which all ritual elders are eventually initiated.5 If the ideology of deep knowledge asserts a fixed corpus of secrets, then it should not be taken at face value, but as a screen that allows its pragmatic functions to masquerade as sanctified wisdom and learning. As such, deep knowledge is powerful because it is revisionary, sustaining possibilities of political transformation through the revaluation and reversal of established orders. In the sociocultural contexts of historic Yoruba kingdoms, the political lines of contestation and division are formed around segmentary opposition and hierarchical inclusion. To be sure, the considerable impact of colonial administration, commercialized agriculture, Islam, Christianity, national politics, and so forth on such local ritual fields has been negotiated by the orisha cults, which have rearticulated with emerging elites and patterns of stratification. As I have argued, it is precisely the revisionary principles of deep knowledge that have enabled the orisha cults to extend their interpretive horizons and embrace colonialism, class formation, and the postcolonial state while simultaneously bringing the outside in. But as a generative template or structuring structure, the revisionary model of Yoruba hermeneutics and its associated hot and cool valences are grounded in the dialectics of kingship as has been sketched above. Applied to local pantheons, even the coolest orisha can become hot or dangerous, with hidden deadly powers that are protected by secrecy and activated by its praises (oríkì). Extending these principles to the logic of syncretism in chapter 1, I argued that the reinterpretation of Catholic saints as African gods in Haitian Vodou, Cuban Santería, and Brazilian Candomblé—a process that Herskovits characterized in psychological terms—was recast by the model of Yoruba deep knowledge as a mode of political revaluation and revision. Less a screen for maintaining African traditions than a form of collective appropriation, the saints were Africanized by New World blacks as double agents in their religious sanctuaries and societies. If the public identity of a saint was European Catholic, then its secret, deeper, and more powerful African manifestation could be invoked and manipulated by initiates. The very spatial organization of altars seems to support this vision, with chromolithographs of popular saints displayed on top and above, and icons—including fundamentals—of African deities hidden below or locked within cabinets.6 Historically, there

46 / Chapter Two

is no question that African deities and their ritual arenas are associated with major revolts and resistance movements, including the Haitian revolution. Boukman’s legendary Vodou ceremony of Bois Caïman, which triggered the revolution’s call to arms, is the most celebrated exemplification of the power of the gods.7 Palmares, the most famous of the Brazilian quilombos (maroon communities) of escaped slaves, and the Cuban slave revolt known as La Escalera, commemorating the ladders to which the vanquished slaves were tied and tortured, are among the significant variations on this political theme. Thus proposed as a hypothesis, the deep-knowledge model could explain how New World syncretism mobilized resistance and opposition. To be sure, in that the sociopolitical contexts of plantation societies and the peasantries that followed were radically different, the structural bases of political action were more about race and class than kings and chiefs, but it was precisely the hermeneutical mapping of deep-knowledge claims within these oppressive contexts that provided possibilities of collective empowerment. With this hypothesis up front, I turn to Haiti and the Petwo paradox.

The Petwo Paradox It is practically a catechism of Haitian Vodou that its extraordinary range of deities and powers, including mainly lwa (loa, “gods” and “goddesses”) but also spiritual agents like zombies, pwen (points or manufactured deities), and djab (spirit devils), fall into two general divisions—Rada and Petwo. In official publications and popular accounts, the Rada deities, associated with the pure unadulterated tradition from Ginen (Guinea or Africa), generally trace back to the town of Arada (Allada) in the historic Dahomean empire; Rada rituals involve cool drum rhythms, choreographies, and spiritual demeanors appropriate to revered authority. The Petwo deities, by contrast, are hot and transgressive, riding their mounts (possessed devotees) during ritual performances with the explosive fury and self-abandonment of faster tempos and accentuated off-beats. According to Métraux, The word petro inescapably conjures up visions of implacable force, of roughness and even ferocity—qualities which are not a priori associations of the word rada. Epithets such as “unyielding” “bitter” and even “salty” are applied to the petro while the rada are “gentle.” The petro loa are, moreover, specialists in magic. All charms come under their control. . . . Everything which has to do with petro is shadowed with doubt and inspires fear. ([1959] 1972, 88)

Creolization and Connaissance / 47

Implicit in this contrast between Rada and Petwo divisions is the latter’s dubious association with money and magic, linking the more immediate efficacy of its lwa with moral compromise and even wickedness. Unlike the righteous Rada deities, the Petwo operate like invisible loan sharks, breaking moral, social, and even physical boundaries to confer quick results at high— and literally binding—rates of return. As Hurston explains, Before we go into a description of the outdoor altar to Petro, let me give you some idea of the differences between a Rada god and a Petro divinity. As has been said before, Damballa and his suite are high and pure. They do only good things for people, but they are slow and lacking in power. The Petro gods on the other hand are terrible and wicked, but they are more powerful and quick. They can be made to do good things, however, as well as evil. They give big doses of medicine and effect quick cures. So these Petro gods are resorted to by a vast number of people who wish to gain something but fear them at the same time. The Rada spirits demand nothing more than chickens and pigeons, and there are no consequences or hereafter to what they do for you, while the Petros demand hogs, goats, sheep, cows, dogs, and in some instances they have been known to take dead bodies from the tombs. The Petros work for you only if you make a promise to service them. . . . [T]he promise must be kept or the spirits begin to take revenge. ([1938] 1990, 164–67)

It is clear from these and other classic comparisons (e.g., Bastien 1966, 42–43) that despite their purity, authority, and Dahomean ancestry, the Rada deities are less effective and ritually potent than the more powerful, if somewhat polluted, Petwo spirits who work outside official channels.8 Indeed, the power of the Petwo deity innovates and transforms, requiring the blood of the four-legged sacrifice to mobilize personal and collective agency. Quoting a leader of a Bizongo society, one of the extreme institutional branches of Petwo sorcery, Beauvoir-Dominique (1995) shows how the arts of self-transformation into ordinary animals and predators of the night enable nothing less than the transformation of the world. In a remarkable ratification of the politics of revision, the houngan (priest) relates concrete intervention to the ritual power of transmutation: Bizongo is to prove that man can learn to change. That’s what Bizongo means; learn to change. We’re in the world, and we can transform the world. In the world; just as people watching us can see our members metamorphose,

48 / Chapter Two become pigs, chickens, donkeys, any kind of animal . . . so we change. And that is what is called Bizongo—changes, that’s what it’s all about, changing form. (Beauvoir-Dominique 1995, 166, emphasis added)

The political implications of Petwo transformations are clearly evident in Vodou history and practice. The celebrated sacrifice in Bois Caïman that triggered the Haitian revolution in nationalist accounts was specifically a Petwo ceremony, so identified by the killing of a pig. The use of whips, firearms, and stylized fragments of military insurgency in its ceremonies has long associated the power and iconography of its services with the Haitian revolution, a connection conveyed by the langaj (ritual language, possessed speech) of officiating houngan in invocations of Jean Petro and other revolutionary leaders of the first black republic, such as Christophe, Dessalines, Pétion, Rigaud, and Toussaint L’Ouverture (Mars 1953, 222). Indeed, the revolutionary power of the Petwo division works as a form of condensed historical memory, which includes the experiences of the slave’s social death and transfiguration within plantation society (Dayan 1998). In this capacity, Vodou with its Petwo powers exemplifies Taussig’s (1987, 366–92) explorations of history as sorcery. But I am less concerned with repressed histories and forms of historical consciousness for now—this issue returns in another exposition (Apter and Derby 2010)—than with the explicitly Creole status attributed by Haitians to the Petwo spirits themselves. This attribution, and its association with sorcery and magic, initially challenges the deepknowledge paradigm. If the Creole genealogy of the Petwo spirits remains an unsolved mystery in Vodou historiography, its founding ancestor is generally identified with the semimythic figure of Don Pedro—a popular sorcerer in colonial Saint Dominique. In a passage from Moreau de Saint-Méry (1797), an innovative dance associated with violent possession and insurrection marks the birth of the Petwo nation: In 1768 a Negro of Petit-Goave, a Spaniard by birth, abused the credulity of the Negroes with superstitious tricks and gave them the idea of a dance, similar to the Voodoo dance, but more hectic in its movements. To give them extra filip [sic] they added well crushed gunpowder to the rum which they drank while dancing. Sometimes this dance, called the Danse à Don Pèdre or The Don Pedro, inflicted fatal casualties on the Negroes; and sometimes the very spectators, electrified by the convulsive movements, shared the madness of the dancers, and drove them on, with their chanting and hurrying rhythm to a crisis which, to a certain extent, they shared. The Don

Creolization and Connaissance / 49 Pedro was forbidden under threat of direst penalty—sometimes without avail. (44; see also Courlander [1960] 1973, 129; Métraux [1959] 1972, 39)

If or exactly how the Petwo deities derived from the Don Pedro dance is less important than the status this early account has acquired as a veritable myth of origins.9 Elevated to the deity Jean Petro after his death, Don Pedro was both “a Negro of Petit-Goave” and “a Spaniard by birth.” His quintessentially Creole identity—most likely a fugitive Dominican slave who escaped into Haiti (Larose 1977, 111)—establishes a non-African patriline for the new category of hot and raging deities that he inaugurated, exhibiting the “violence and delirium that threw off the shackles of slavery” (W. Davis 1988, 274). Homegrown in Haiti rather than inherited from Africa, the Petwo line has been characterized as a hybrid of Creole and Kongo attributes and powers in marked contrast to the Rada division of Ginen (Courlander [1960] 1973, 328–31).10 It should be clear from this summary sketch that Petwo and Rada resemble the hot and cool valences of Yoruba pantheons and the dialectics of power and authority that they mediate. Unlike the conservative Rada, who preserve the secrets from Africa, Petwo are innovators, deploying money and magic to change human and social forms (Derby 1994). Born in the spirit of the Haitian revolution and enlisted in its struggle, the Petwo are transgressive and subversive, “riding their horses” (i.e., possessing their devotees) brutally and without warning, darting “de ci de là, agité, convulsif, rageur” (Mars 1953, 222). In their extreme form of Bizongo, or Petwo sauvage, their opposition to Rada assumes the symbolism of evil: “Bizongo is in some sense the opposite of Rada. If the latter represents the benign aspect of the faith, Bizongo symbolizes unforgiveness. Rada stands for light and the normal affairs of humanity; Bizongo occurs by night, in the darkness that is the province of the djab, the devil” (W. Davis 1988, 274). Hurston’s characterization of Rada as “high and pure” yet “lacking in power” ([1938] 1990, 164), reveals in full form the distinction between power and authority. As hierarchy, orthodoxy, and rule-governed code, the Rada carry the authority of Africa to their hounfor (temple) and devotees, enshrining the reproduction of the status quo. Outside such authority structures and defying limitation, the Petwo manifest pure power and efficacy, uncontrolled, dangerous, devious, and above all transformative. Like the hot and deep dimensions of the Yoruba orisha, their invocation empowers but also kills. Their force can be controlled but at considerable risk of uncertain outcomes. Against the moral high ground and orthodox tradition of the Rada division, the Petwo gods or spirits pit magic and sorcery. Using medicines and

50 / Chapter Two

charms such as paket kongo and wangole (fetishes that concentrate the nonDahomean secrets of Kongo and Angola in their materials and nomenclatures), the Petwo tread the nefarious shadowlands of points and zombies. Within the strictures of Ginen, they are outsiders and lack legitimacy. The knowledge, or connaissance, of their practitioners is false by the standards of Ginen tradition. Here lies the paradox. If the transformative agency of Petwo lwa occupy the sanctified space of deep knowledge in the Yoruba paradigm, then why is the knowledge associated with their powers normatively devalued as sorcery and magic? If the cultural hermeneutic of deep knowledge was indeed transplanted from West Africa to the Creole Caribbean, then why are Haitian ideas of connaissance linked to the authority of Ginen rather than the power of Petwo? That Haitians do in fact possess notions of cult secrecy, and that they are expressed in idioms of depth, is brought out clearly in the concept of connaissance. As Métraux explains: Voodoo adepts use the word “knowledge” (connaissance) to describe what we would define as “supernatural insight and the power which is derived therefrom.” It is in degrees of “knowledge” that various hungan (priests) and mambo (priestesses) differ from each other. ([1959] 1972, 63–64)

The acquisition of this knowledge, however, occurs in two potentially contradictory ways. The first is through graded induction and initiation. Much like any technical training, the devotee ascends the ranks from novice to priest, serving successively as hunsi (initiate), hungenikon (choirmaster), la-place (master of ceremonies), and confiance (trusted adept) over the course of several years, culminating in la prise des yeux (the gripping of the eyes), the highest stage of initiation, conferring clairvoyance (Métraux [1959] 1972, 68).11 The supreme knowledge of a houngan or mambo represents the true tradition from Ginen, carried to the Americas by the first African slaves. Weakened over time, however, by the death of great priests and their reluctance to divulge all their secrets fully, connaissance becomes understood as a diminished reflection of an original power and presence (Larose 1977, 87– 89). To counteract this entropic trend and recover what was otherwise lost, a second method of instruction through direct spiritual intervention supplements and in some cases overrides the first, whereby the initiate receives knowledge directly from the lwa. Although priests relying exclusively on such claims of supernatural patronage invite suspicion or disdain, their inspired learning, when combined with formal initiation and training, marks

Creolization and Connaissance / 51

the deepest form of recollected knowledge.12 Here is where images of the deep figure prominently. Because the lwa themselves live in riverbeds and the depths of the sea, alternatively defined as the realm of Ginen beneath the waters, priests and priestesses go there to enhance their power through further instruction. Métraux ([1959] 1972) describes testimonies in which “hungan and mambo with great ‘knowledge’ go and visit [the lwa] in their watery homes and stay with them for long periods,” returning with “new powers and sometimes . . . shells—the concrete proof of their exploit” (92). Madame Tisma, a mambo from Marbial, claimed to have “spent three years at the bottom of a river where she had received instruction from water spirits” (Métraux [1959] 1972, 63). Figures of the deep predominate in Yoruba rituals and discourses of initiation, manifested in neophytes “seized” by spirits of the water, myths of revelation in ibú (the deep), and in such praise-names as Yemoja’s Omi, Aríbúṣ ọ lá (Water, she-who-makes-the-deep-a-place-of-honor) (Apter 1992, 129). But if Yoruba deep knowledge establishes a sanctified space of revision and refiguration, such interpretive strategies in Haiti appear pushed to the margins of sorcery and magic, of the dishonorable Petwo against the righteous Rada. Deep knowledge in this formulation is opposed by the powerful Petwo line. I can deepen the problem and prepare its solution by locating the Petwo paradox in its relevant social and political contexts: one rural and generational, based on the developmental cycles of the landed peasantry; the other urban and based on class, focusing on the political machinations of François Duvalier. As I show, these contexts are not isolated but represent poles of a single highly stratified sociopolitical system.

The Dynamics of Creolization The complex and variegated character of Haitian land tenure notwithstanding (Murray 1977; D. Woodson 1990), Larose (1977, 97) has argued that the key to the Guinea–Magic opposition lies in the démembré (lit., “dismembered”) and lakou (family estate)—those cognatic descent groups that hold land collectively and establish “the basic cult units” of rural Vodou. In a vivid account of a generational conflict between a Vodou priest and his aspiring son, Larose (89–92) relates the dynamics of segmentation and fission to accusations of Satanism and sorcery against the authority of Ginen. Thomas, the recognized head of a démembré and, thus, houngan of its associated Vodou cult-house, is challenged by his son Leon, who barges into the inner sanctum, where he is not authorized to enter, in order to learn the secrets of “tying up the wanzins.” In the heated exchange that follows, the

52 / Chapter Two

father accuses his son of following the wrong path, wandering the land with the Sans Poils (hairless society) or Bizongo group and worshipping demons and dark forces for personal gain. The son also was a makout—that is, he served in the Duvalier regime’s dreaded secret police. As far as the father was concerned, however, the son’s knowledge was unfit for the true Vodou priesthood. The son, on his part, demanded that his father retire, and that he himself be appointed successor. Veiled threats of “testing” and poisoning lead to a cautionary tale, wherein Thomas relates how his own father had been similarly “sounded” by an unspecified relation—most likely his junior brother—who consequently died. As Larose (91) succinctly summarizes the argument, “Leon confronting his father, is magic confronting Guinea.” If generational succession lies at the core of the conflict, the broader contexts of démembré and lakou illuminate its positive and negative valuations. Like any cognatic lineage organization (Fortes 1969, 287), démembrés select from multiple lines of descent and filiation to form corporate groups limited by residence “on the ground.” To make matters more complicated, although corporately held land cannot be sold on the open market, family members can sell plots to each other. Inheritance furthermore appears to be per stirpes, such that children of the same father but different mothers receive land in equivalent blocks. If many démembrés united into the larger general estate of the lakou, the per stirpes subgroups within each démembré established lines of future segmentation and fission.13 As Larose explains, “The reluctance of the father to appoint a successor while he is still living, does not ease the finding of a solution. Quarrels over succession often lead the contenders to build up their own installation on another piece of land they may have bought elsewhere or in their own share of the inheritance. Each of the new establishments defines a new démembré for the children of its founder” (1977, 98). In the most basic structural terms, Ginen, with the authority of Africa and the ancestors, upholds the integrity and reproduction of the démembré (or where intact, the lakou), whereas Petwo is associated with its segmentation and fission. As an underlying developmental cycle (Fortes 1971), the reproduction, breakup, and reconstitution of the landed démembré is thus mediated by the relative opposition between Ginen and Petwo.14 What is magic at one phase of the cycle will become Ginen at another. As Larose (1977, 86) has noted, “Everyone is Guinea in his own way and everyone denigrates his neighbor for having added to and thus diluted the inheritance.” In addition to land and lineage as such, money and wage labor enter the picture to further illuminate the magic of Petwo. Women working in rural

Creolization and Connaissance / 53

markets accumulate profits from trade that they use to purchase land from family members. The land thus acquired by a successful trader can hive off at her death to establish an autonomous démembré for her children.15 The same is true for land accumulated by the earnings of wage labor. A junior son with limited prospects of succession can leave the land, earn money, and return to purchase family land, eventually establishing his own démembré. What is initially purchased by money is gradually incorporated into the moral order of the ancestors. In the idioms of Vodou, the reincorporation of purchased land into hallowed lineage relations is like buying a pwen (point)—a personal medicine that over time becomes a respected lwa. In more frequent patterns of inheritance, the lwa of the démembré are inherited with its land, and forgotten lwa of former owners are known to reappear with recriminating demands and grievances. I have ruthlessly simplified Larose’s rich material in order to abstract the dynamic principles of rural Vodou onto which the Yoruba deep-knowledge paradigm can be mapped. Clearly, the power and authority relations of the Yoruba kingdom, with its associated hot and cool ritual idioms, are transposed, in some sense, to the Haitian démembré, where the micropolitics of succession and fission revise authoritative idioms of corporate solidarity. Here magic confronts Ginen not only as the impatient son confronts his unyielding father, but as segmentation and fission divide corporate groups. I do not mean to suggest, however, that such a peasant system operates in rural isolation, for relations of trade and wage labor enter the picture in crucial ways. It is highly significant that the Bizongo or Sans Poil as paradigmatically “Petwo sauvage” are uprooted from the land, wandering satanically in transmogrified shapes and guises. Nor is Leon’s association with the urban capital unusual. Does the magic and sorcery of the Sans Poil express the landless peasantry and urban poor, the commodification of their labor and the fetishism of its alienated forms?16 What I am suggesting is a historical articulation between the developmental cycle of the rural démembré and the polarization of class relations, an articulation that is further framed by the Ginen–Petwo divide.17 As the Petwo represent the coarse self-abandonment of the festive lower body in violence, sexuality, and laughter (Bakhtin 1984, 303–67), the Rada acquire the refined tastes and habits of the elite. Taking on the racialized connotations of Haitian social class totally lacking in the funfun, or white, category of cool Yoruba orisha, the Rada nation of gods represents the crème of Vodou society. Hurston ([1938] 1990) observed that devotees of Erzulie Frieda, the classy mulatta of luxurious love, “powder their faces with talcum” (1990, 122). Larose notes that

54 / Chapter Two the “White spirits,” often identified with the Rada “nation” . . . all share a common fondness for town and foreign goods; French perfumes, soft drinks, white bread and sugar, silver cutlery and porcelain plates. (1977, 100)

Métraux casts the difference more boldly. Between the two classes of lwa, he maintains, there is such a huge gulf that one is tempted to talk in terms of an aristocracy and a proletariat of gods. To the former would belong the loa of African origin venerated in all sanctuaries, and to the second the greater part of the “Creole” gods, called “Creole” because aboriginal and of recent “birth.” (Métraux [1959] 1972, 84)

Here is the curious opposition of creole revolutionaries against an African ancien régime! And here, too, in this inverted revolutionary image, lies the solution to the Petwo paradox: As Vodou developed under plantation slavery, the revisionary and, indeed, revolutionary principles of deep knowledge were mapped onto polarized class relations, with the heterodox Petwo opposing the hegemonic Rada. I call this mapping “the lowness thesis.” Unlike the Yoruba and Dahomean kingdoms where the structural bases of power competition were rooted in segmentary opposition between political factions, in Haiti such bases took on the vertical character of class relations. If, before the revolution, the colonial regime of French sugar and coffee planters established the cultural markers and distinctions of this brutally oppressive class, then after the revolution and the abolition of slavery, the new mulatto elite in many ways occupied its place, replicating its cultural codes and controlling the economy. These polarized relations of class division and conflict established the material conditions of social protest and power against the elite, as generated among landless peasants, exploited laborers, criminals, hustlers, and vagabonds—in effect, the Haitian underworld. As the Rada line of Ginen became high and hegemonic within Vodou, identified symbolically with the Haitian elite, the revisionary principles of deep knowledge were emically located within the low, non-Ginen, Petwo line, associated with the hybrid character and transgressive power of revolutionary Creoles. Thus proposed, the lowness thesis grossly simplifies the complex history of Haitian class relations and racial ideologies, a history marked by the lasting rivalry between northern black and southern mulatto elites (with colonial roots) that fully emerged after the revolution (Garrigus 2006; King 2001;

Creolization and Connaissance / 55

Nicholls 1985, 24–28; Trouillot 1990, 47–50, 109–36). It fixes the model of color stratification neither at the moment of revolution nor in time as a static and unyielding framework, for as Haitian history fully demonstrates, major shifts in economic and political organization from the early schism between Christophe and Pétion, through the US occupation and the rise of noirisme, to the struggles of Aristide’s Lavalas movement, mark swings and turns in the politics of color. What the lowness thesis does emphasize is that the segmentary character of political competition at the national level in Haiti extended its roots into the urban and rural underclass to galvanize political power and popular support. Nowhere is this racialized politics, and its dangerous dealings with the Petwo line, more clearly demonstrated than in the notorious career of François Duvalier.

Fetishes of State François Duvalier, or Papa Doc, was by no means the first Haitian head of state to inspire strong associations with Vodou. Many Haitians believe that Dessalines and Christophe were resurrected into lwa after their deaths, and the self-proclaimed emperor Faustin Soulouque (1847–59) developed the first Vodou political machine, replete with secret police and zinglins (paramilitary thugs) who carried out political massacres and campaigns of terror on his orders (Rotberg 1976, 356). Antoine Simon, president from 1908 to 1911, is reputed to have risen from peasant origins through the hidden powers of his daughter Celestina, a respected mambo and force within the palace, where secret ceremonies fueled the fears and anxieties of the aspiring court society. Assembled at palace functions, Simon’s sycophants regarded the food and wine with the dread suspicion of unspeakable cannibalistic substitutions. As elite society mingled with forbidden powers of the deep, the palace itself became iconic of high and low. Thus, Hurston ([1938] 1990) recounts how during President Simon’s brief and controversial rule, “often it was said that a Voodoo ceremony was going on in the basement chambers while the state function was glittering its farcical way in the salon” (96). Under François Duvalier, similar fears of a basement of blood circulated among peasants and cabinet ministers alike, who attributed clairvoyance, self-protection, and the powers of zombification to the man who modeled himself on Baron Samedi, a Vodou deity and Lord of the Cemetery (Rotberg 1976, 362–65).18 To be sure, Duvalier’s role in elevating Vodou to the respected status of a national doctrine was already established by his involvement with noirisme and the founding of Les Griots, its literary organ, together with Lorimer Denis,

56 / Chapter Two

Carl Brousard, and Clement Magloire in 1938. There Duvalier called for the rehabilitation of Vodou, which he saw as the soul of the Haitian people and called “the supreme factor of Haitian unity” (Dayan 1998, 126). But beneath such ideological representations of authentic tradition, and operating behind the scenes, was a growing network of loyalties and alliances between Bizongo societies, the dreaded secret police known as tonton-makout, and the civil militia or Volontaires de la Sécurité Nationale (VSN), which served as the official face of the security apparatus. As Trouillot (1990, 190) explains, the system operated at many levels to intimidate critics and undermine organized opposition to Duvalier’s dictatorship, but two aspects of this strategy pertain directly to my argument. The first is that by developing and extending this network within Port-au-Prince and throughout the countryside, Duvalier deepened his social base from the middle classes into the peasantry and lumpen proletariat. Recruiting “new urban parasites” (Trouillot 1990, 153), ranging from petty shopkeepers to “thugs and proven criminals” (190), Duvalier appropriated those lower orders whose opposition to the elite could otherwise be mobilized against the state. The black peasantry and Haitian masses celebrated by noirisme were thus strategically recruited to bolster state domination. The second aspect of this strategy concerns the coercive patronage by which this recruitment was accomplished. Redistributing state funds from coffee taxes to the lower orders of clients who, in turn, became so-called barons of their villages and neighborhoods, Duvalier generated a system of “practical consent” based on popular figures and idioms (Trouillot 1990, 191). As Trouillot explains, “It was the individual fidelity of these men and women that provided the famous tonton-makout (literally, ‘Uncle with a basket,’ so-named after a bogeyman in a folk tale) with its critical mass” (1990, 154). In addition to the recruitment of major Vodou priests, Duvalier transformed ritual symbols on makout uniforms into national icons of terror and intimidation: This uniform—blue denim shirt, pants, and hat, with a red kerchief— evoked the traditional costume of the Vodoun god Zaka (the peasant god of agriculture), the colors of the Haitian national flag before Duvalier, and the peasant armies of nineteenth century Haiti, crushed by the Marines during the U.S. occupation. And while the middle-class members of the secret police arrested and tortured opponents of the regime, the peasant-dressed members of the militia, peasants themselves or members of the urban lower classes, marched to the sound of military music in the streets of the capital city, intimidating by their very presence the bourgeoisie and the middle classes. (Trouillot 1990, 190)

Creolization and Connaissance / 57

Thus we see how the peasant sack typically carried by the Petwo deity Azaka doubled as a Vodou medicine pouch and material receptacle for Papa Doc’s coercive patronage, tapping into the forbidden depths of sorcery and social class to oppress the masses, neutralize the opposition, and empower the state. Beyond such symbolism and iconography, to what extent were the Bizongo societies materially involved with Papa Doc’s political project? It is clear from Larose’s material on structural conflict in the rural démembré that Leon, the aspiring son, opposed his father Thomas not only as a Bizongo sorcerer, but also as a tonton-makout, bringing the formidable threat of the secret police to bear on his domestic agenda. More generally, Duvalier’s preference for the Bizongo societies, and their regional networks of houngans doubling as chefs-de-section (rural section chiefs), are readily apparent in Davis’s observations of Bizongo ritual and political organization.19 In his sensationalized search for the secrets of the Haitian zombie, Davis worked through Bizongo networks made available by his “principal informant—a man both deeply religious and deeply patriotic—[who] became the effective head of the Tonton Macoute for a full one-fifth of the country” (W. Davis 1988, 270). As Davis explains, Duvalier was the first Haitian president to be personally involved in the appointment of each chef-de-section throughout Haiti, an interstitial position between the state and the peasantry that combined the formal administrative office with the shadow roles of houngan or Bizongo president and tonton-makout. That the chefs-de-section were recruited from the peasantry further deepens the lowness thesis in that the power of deep knowledge expressed as segmentary opposition in the Yoruba context was transformed by the dialectics of Haitian class and color stratification into the revolutionary potential of the oppressed. Again, it cannot be overemphasized that Bizongo represented the wild side of Petwo in particular and Ginen at large, not as tradition upheld but as the hot iron that strikes. Elevated by François Duvalier into positions of inviolate power, the Bizongo presidents-cum-tonton-makouts were well versed in the official doctrines of the state. As one Bizongo leader proclaimed at a nocturnal séance, Bizongo is the culture of the people, a culture attached to our past, just as letters and science have their place in the civilization of the elite. Just as all peoples and all races have a history, Bizongo has an image of the past, an image taken from an epoch that came before. It is an aspect of our national soul. (W. Davis 1988, 266)

58 / Chapter Two

Such rote renderings of Papa Doc’s noirisme were visually echoed by the preponderance of Haitian flags and presidential portraits decorating the Bizongo peristyle, which by 1984 celebrated the image of his son and successor, Baby Doc. But beneath the smiling face of the society, extending its nocturnal networks throughout the nation, was the nefarious hand of a shadow government that converted the revolutionary potential of the dispossessed into the dirty work of the regime. François Duvalier’s appropriation of the Bizongo societies to consolidate control illustrates that the power they wielded posed enough of a threat from below to be neutralized and redirected from above. By incorporating Petwo powers into his political arsenal and thus deepening his social base, Duvalier co-opted effective opposition and channeled it against the growing tide of bourgeois rebellion. Socially, politically, and economically deep in terms of class relations and racial ideologies, however, it remains to be shown how such sorcery and magic were culturally deep in terms of the deep-knowledge paradigm itself. I have shown that the Petwo line corresponds to the revisionary and revolutionary powers of deep knowledge through its invocations of revolutionary leaders, such as Boukman and Makandal, and the incendiary ceremony at Bois Caïman; its association with fire, fission, transgression, and subversion; and, as a science of transformation, by its capacity to change the world. Riding their “horses” in violent possessions, the Petwo deities demand freedom from social, political, even bodily boundaries, opposing the authoritative order as upheld by Rada. To fully resolve the Petwo paradox according to a West African–derived hermeneutical model, however, I must specify those strategies of revision and revaluation that characterize deep knowledge in more formal terms, as opposed to Rada and within Ginen at large.

Revisions and Reversals If, from the official Rada perspective, the Petwo cults and Bizongo societies adulterate the true tradition with money and magic, such matters appear differently to insiders. Like the orisha cults in Yorubaland that serve alternative pantheons and hierarchies within the inner sanctums of their shrines, what is shallow and heterodox from the outside is deep and orthodox from within. As Larose points out, each society sees itself as the guardian of the true way, whatever its relative denomination, and the Petwo cults are no exception. For their followers and practitioners, the efficacy of such sorcery and magic offers tangible evidence of true Ginen. Moreover, available evidence of their deep-knowledge forms suggests that the most magical and

Creolization and Connaissance / 59

creolized line of deities shares much with West African secret societies. Davis records that—just like the Poro secret society in West Africa—“central to Bizongo signs and signals of recognition is a notion of opposites,” such that in a variety of ritual greetings, entrances, and exits, “heaven becomes earth, the mouth becomes the anus, front is back, up is down, the eyes are ears, the knee is the elbow, the hand is the foot and the foot the hand” (W. Davis 1988, 253). Although the Yoruba Ogboni Society deploys similar codes, such reversals and inversions of the normal are typical of most societies with secret passwords and languages and do not by themselves establish West African provenance. Many reversals play on Masonic symbols and even handshakes, suggesting an appropriation of European or Creole signs of secret power and value through secondary coding. But in relation to the Rada line that Petro opposes, such strategies of refiguration and appropriation are paradigmatically deep.20 Among the Petwo lwa, for example, are found manifestations of Rada deities that have been transformed into hot and dangerous counterparts, indicated by various forms of relexification. Thus the exalted Dambala Wedo (Ouida), often identified as the leader of the Rada gods, appears as Damballah-flangbo among the Petwo, where he has a “bad reputation” (Métraux [1959] 1972, 89) along with other appropriated counterparts such as Oguyansan, Ezili-mapyang, Ezile-Dantor, and all other root names modified by the “Ge-rouge” cognomen of the notorious Secte Rouge.21 Not all Petwo deities mark Rada names with Petwo tags. Others appear to derive from Kongo as well as Creole sources, giving rise to the rather literal-minded debate over the Creole versus Congo character of the Petwo line. The problem is characterized by at least two dimensions of semantic contrast; one between Creole and African names, the other between West African (Dahomean, Yoruba) and Kongo names. Thus a Creole like Don Pedro became a lwa, as opposed to Ogun-Badagary, who came directly from Ginen. But even within the so-called Creole category ambiguities abound, in that Petro deities such as ’Ti Jean Petro represent African spirits who, like the slaves from the old country, were baptized or rechristened with European names (Métraux [1959] 1972, 88). The strategies of relexification have produced a bewildering variety of Haitian lwa, muddying the philological waters beyond possible genealogical reconstruction (with proponents of the Dahomean, Kongo, and even Creole camps pushing their eponymous ancestors). I would argue that such attempts at reconstruction should be abandoned in favor of a fresh consideration of the strategies themselves, so as to view them in a different light. It is not the names of the deities, or even their attributes, that provide the

60 / Chapter Two

key to the origin of the Petwo line but the rhetorical devices of marking and relexification themselves, so clearly rooted in West African principles of hermeneutical revision and revaluation. On a concrete level, the Petwo gods are a mixed group. They are indeed Creole, Kongo, and in some cases Dahomean (not to mention Ibo, Senegalese, etc.), merging together and breaking apart according to the dialectics of ritual empowerment and renewal. But such origins themselves are insignificant. What is significant, I would argue, are the more general principles by which they have coalesced into a powerful Petwo line contra the authority of Rada. This shift from elements to the broader ensemble solves a problem for those pushing the Kongo connection. In his efforts to establish an unequivocally Kongo pedigree for the Petwo gods, de Heusch notes a glaring incongruity between their African ritual association with water and their connection with fire in parts of Haiti: Although there is no doubt about the Kongo origin of simbi, kita, and bumba, their incorporation in voodoo raises a major problem. In central Africa, the aquatic aspect of these nature spirits is clearly marked whereas, in Haiti, they are associated with fire in the petro ritual. Among the Mpangu, the favorite abode of both the simbi and nkita is water. . . . What is the explanation for almost all the Haitian petro loa and, in particular, the so-called Congo Savannah gods, being said to be hot and associated with fire? (1995, 110)

De Heusch appeals to “the structural nature of syncretism” for the answer, providing what is essentially an ad hoc rule of binary contrast that effectively puts the cart before the horse. Because Petwo came to oppose Rada, and Rada were predominantly water deities, the water of Petwo was symbolically transformed into fire. He goes on to note an ambiguous or intermediary class of Nago Ogun deities that mediate between hot and cool in another ad hoc attempt to explain why even within the Rada line, some deities are hot while some petro are friendly and cool (1995, 113). These issues resolve without recourse to ad hoc rules and categories, however, by locating the Petwo gods more firmly within a Dahomean-Yoruba cosmological schema. The Yoruba material sketched earlier indicates how ritual valences of hot and cool mediated the dialectics of power and authority within historic kingdoms and took on the character of class relations in Haiti in the Rada–Petwo opposition. This does not imply, however, that all Rada are cool and all Petwo hot. Rather it implies that the terms of this opposition were already present in the Dahomean-Yoruba system, between categories of deities as distinguished by context. Just as the same Yoruba deities can be hot and cool with respect

Creolization and Connaissance / 61

to authority relations, so the Haitian lwa can slide between camps, such that a Rada lwa in one démembré, town, or region serves the Petwo line in another. Within the corporate organization of every Vodou society or hounfor, I would argue, a relative hot and cool distinction will emerge along the lines of generational succession and structural fission as outlined on the démembré. With these formal properties appropriately distinguished, de Heusch’s problem requires no external structural rules or assumptions. Within the Ginen framework of Haitian Vodou, which evokes the historic Guinea coast of West Africa, devotees distinguish the cool, high Rada from the hot, low Petwo in an initial marked contrast between categories of gods. Within each category, the distinction is reproduced, such that Rada breaks into hot and cool, as does Petwo, extending to whatever relevant subcategories the sociopolitical context supports. Cutting across this segmentary model is a diagonal slash, skewing the model by class to distinguish true Ginen from adulterated Petwo. In this diagonal opposition, Ginen is marked as a restricted or elite set of gods apart from the rest of the deities, which include Congo spirits, Creole ancestors, and those Rada gods marked with Petwo tags (fig. 2.2). Thus incorporated into a West African framework, the contradictions, incongruities, and ambiguities of Vodou begin to make sense. Ginen is at once the entire system, “a general concept that covers the full range of religious practices in Haiti” (W. Davis 1988, 273; see also Beauvoir-Dominique 1995, 167), and a marked subset distinguishing Rada from Petwo and other manifestations of magic, money, and sorcery (Larose 1977, 85–86, 112) such as pwen (points), zombies, djab (devils), even paket Kongo (Thompson 1995).22 Within Rada and Petwo lines, cool and hot valences can further be found, giving rise not to de Heusch’s intermediary categories, but to fluid subdivisions. In this general paradigm, Ginen is authoritative and cool; Petwo powerful and hot. Petwo gods of Kongo origin thus shifted from water to fire not according to an abstract structural transformation as such, but because they were revalued—incorporated into a Dahomean-Yoruba scheme wherein powerful opposition to authoritative hierarchy is ritually potent and hot. Here is where deep knowledge illuminates an important process of creolization. It concerns the historical question of how the Rada line became hegemonic within Vodou in the first place, and how the Petwo line came to oppose it. The data on Vodou in colonial Saint Domingue are scarce and fragmentary. De Heusch cites evidence that the Kongo ritual enclaves were established by maroons who escaped from sugar plantations in the north and set up a base in Nansourkry, “the high place of the Kongo cult” (1995, 116). Increasing imports of Congo slaves in the latter half of the eighteenth

62 / Chapter Two

Figure 2.2. Power and authority in Haitian Vodou.

century would have enhanced the Kongo elements of the Petwo line. Thus positioned against Rada and bolstered by imports of Congolese slaves, Petwo remains a resolutely Kongo cultural complex for de Heusch. Like the Petwo powers of the ritual domain, he has taken on the struggle against Dahomean hegemony in Haitian Vodou, arguing, along with Thompson (1995), for greater appreciation of Central African contributions. What he cannot explain, however, is why Petwo became a residual category for all non-Rada families of gods, including Creole deities and those of other African nations such as Ibo, Zandor, Wangoles, Kanga, Bambara, even Siningal (i.e., Senegal, see Courlander [1960] 1973, 331). To take such lumping into account in the making of a Creole religion and society requires a final if brief excursion into the slippery notion of the nanchon (nation). Underlying much of my

Creolization and Connaissance / 63

discussion of Rada and Petwo families or lines, the nation holds a critical place in the dynamics of creolization in Haiti. Historically, the notion of nation referred to categories of slaves in colonial Saint Domingue, combining European ideas about different African “races” with behavioral stereotypes, food preferences, and ports of embarkation. As in Cuba and Brazil, some nations—like the so-called true Congos— were cast as better house slaves and others, like Ibos, were seen as superior laborers in the field. The actual stereotypes associated with different nations varied within as well as throughout the plantation colonies, reflecting the prejudices of slave owners and traders in addition to segmented productive relations on the ground. As Larose (1977, 101) points out, however, French planters generally avoided organizing work crews by ethnicity, preferring to mix and mingle different nations in a variation of divide and rule.23 Owners evaluated slaves by their capacities to assimilate, and the salient distinction between house and field slaves developed less between imputed nations as such and more between Creoles and newly arrived Africans (bossales). Creolization among African slaves was thus central to the stratification of the slave community, associating higher status with greater degrees of cultural and racial assimilation. On the one hand, nations were ranked within the colonial ideology according to their supposedly innate capacities of work and assimilation. On the other hand, their social significance was structurally skewed by the division of labor between bossales and Creoles. Culturally, many of the status markers of the French plantocracy distinguishing European civilization from African barbarism were replicated within the exploited class, differentiating mulattos and black Creoles from the ill-fated survivors of the Middle Passage. It is thus significant and initially perplexing to find, in Moreau de SaintMéry’s classic account of colonial Saint Domingue, that the dog-eating Radas (Aradas), known for their bloody customs and ferocity, were deemed least capable of all Africans to learn French or Creole, and that their women were rarely employed as domestics because of their quarrelsome character as ceaseless talkers (1797, 31). Using the same pernicious tokenism, Moreau de Saint-Méry writes that the banana-eating Congos (excepting the cannibal Mondongues) were sweet and gay, “loving song, dance and ornaments they make excellent domestics, while their intelligence, their facility in speaking a pure Creole . . . makes them preferred for house service” (1797, 32). In this account, the Rada nation appears at the bottom of the colonial status hierarchy, closest to African barbarism and thus sent to work the fields, whereas the Congo nation, closer to Europeans, is portrayed as intelligent

64 / Chapter Two

and attractive, speaking a pure Creole with a reassuring joie-de-vivre.24 I call this characterization perplexing because within Vodou their positions were already reversed. The same text, published before the Haitian revolution, identifies the Aradas (of the Rada nation) as “the true practitioners of Vaudoux in the colony, maintaining its principles and rules,” adding that “Vaudoux signifies a supernatural and all-powerful being, upon which all the events on the globe depend” (1797, 46, my translations). There is much rich detail on the dances and services associated with Rada Vodou and its alternative forms. These details deserve close examination, but the relevant point for my general argument is that—if Moreau de Saint-Méry’s depictions are in any way typical of colonial ideology—the Rada–Petwo distinction within Vodou was already a reversal of status relations on the ground, with the hegemonic Rada appropriating high cultural markers in contrast to the low valences of Congo and Creole deities of the nascent Petwo line. Within Vodou, the social and ritual nations, and the growing opposition between African and Creole that subsumed them, were reversed. This moment of radical revaluation put Africans on top and their Kongo-Creole brothers and sisters below. After this turning point, and the political revolution that followed, the nations—realigned with the lakou and démembré—would never be the same. Thus, through a characteristic reversal, the Nago and Rada nations inverted the social hierarchy of Africans and Creoles to become hegemonic within Vodou.25 What was socially low became ritually high, endowed with the authority of Ginen and the weight of tradition. And what was socially high within the slave community became ritually low, embodying the power of sorcery and its dubious means. This ritual reversal, unacknowledged in the historiography of Haitian Vodou, is essential to the solution of the Petwo paradox because it explains how Rada became hegemonic in the first place, appropriating elite markers of social class within Vodou in contrast to the socially low icons and idioms of Petwo, and why the latter came to assimilate Kongo and Creole elements alike. Against the grain of Vodou ideology, the deep-knowledge paradigm reveals not the African Rada against the creolized Petwo, but the equal creolization of both categories of powers according to a deeper dialectic of ritual reversal and mediation.

Conclusion If the deep-knowledge paradigm derived from West Africa helps resolve the Petwo paradox, what does such an exercise prove? If I am not simply backtracking to Herskovits’s African baseline, then what am I saying about African

Creolization and Connaissance / 65

origins? And as Mudimbe (1990, 1994) might ask, “which idea of Africa” is guiding my investigation? By deploying a Yoruba hermeneutical model, I am not implying that Haitian Vodou came from Yorubaland. This is decidedly not the point. I have used the ethnic label loosely, as inspired by my research in Nigeria, to represent a regionally coherent set of political forms and ritual practices underlying Dahomean (Fon) and Yoruba variations. These variations are in some sense echoed by the Rada and Nago nations of Haitian Vodou even as these were transformed and distorted by colonial conditions and ideologies.26 The idea is not to identify Fon or Yoruba elements or gods, but to locate a general interpretive framework that informed the invention of Vodou in Haiti and its political advances and retreats. Clearly, the West African model I have in mind is as dynamic and variable as its Caribbean counterpart. By identifying the revisionary power of the deep-knowledge paradigm in relation to hierarchical authority structures, I have isolated the critical connection between West African and Haitian ritual distinctions of hot versus cool, low versus high, and political transformation versus authoritative reproduction.27 Moreover, this connection is important for what it reveals—in this case the emergence of the Haitian Rada–Petwo divide in relation to class formation, color stratification, and the developmental cycle of the rural démembré. Haitian Vodou is above all a creolized religion. My appeal to a West African hermeneutics of power in solving the Petwo paradox is not directed at recovering the African heritage as such, but in illuminating the dynamics of cultural invention in the Creole Caribbean. As Vodou appropriated European icons and idioms into Rada codes and choreographies, associating the authority of Africa with Haiti’s privileged elites, Petwo became a structurally hot category for incorporating Kongo, Creole, and other African elements into a ritually low-oppositional phalanx. What the deep-knowledge framework reveals are the revisionary mechanisms by which this basic opposition took shape; how, for instance, Rada as against Congo and Creole slaves reversed status locations; how Kongolese water spirits were turned into fire; how Rada root lwa were revalued and given Petwo tags; how Petwo gods came to embody the spirit of a revolutionary class; and, thus, why Duvalier appropriated Bizongo networks to neutralize rebellion and deepen his social base. Nor is the Rada–Petwo opposition purely parochial. In Cuba, a similar divide between the ocha (orisha) of Santería and the Kongo-based practices of Palo-Monte and Mayombe recapitulates the same tension between a pure tradition and a deeper, if more polluted, sorcery (Argyriadis 2000), as do the Yoruba gods of Candomblé against the Bantu and caboclo spirits of Macumba in Brazil (Bastide [1960] 1978, 285–303).

66 / Chapter Two

In each of these cases, African, Creole, and even Amerindian spirits populate an ideological field of stereotypes, and this field brings me back to the idea of Africa. What the Rada nation protects as genuinely Ginen against the powerful, if polluting, innovations of Petwo occludes the critical transatlantic connection underlying the opposition itself, misleading scholars into fruitless debates over what is Creole and what is Kongo. The ideology of Africa within Vodou is of course important for what it signifies about status and class, but unless scholars bring an underlying model to bear against such ideological claims, they are forced to accept or reject them on their own terms. I have deployed the deep-knowledge paradigm in just this sense, as a substrate—to borrow a concept from Creole linguistics—with which ritual ideologies of Africa can be compared.28 Applied to Vodou, I have shown that the most creolized category of spirits is ritually the deepest, conforming to a Dahomean-Yoruba grammar of revision and empowerment that is otherwise invisible or considered unknowable. The question “which idea of Africa?” can thus be answered in relation to a range of positions that frame the major debates. Scholars can, with Herskovits, identify specific elements in Haitian Vodou known to Fon and Yoruba pantheons, although my concern is more with their revisionary principles. In this sense, the notion of a cultural hermeneutic bears some relation to Herskovits’s concept of cultural focus (Herskovits 1966, 59), although more dynamic and dialectical than his posited core values. I also follow Bastide ([1960] 1978, 1971) in emphasizing the “new social frameworks” in which African ritual systems were relocated, and how this relocation in turn shaped their development. By highlighting a broad and flexible Dahomean-Yoruba complex, however, I share with Mintz and Price ([1976] 1992) a more abstract notion of “cognitive orientations” (53) or dispositions transcending specific West African cultures but common to many, emerging as a communicative framework on the slave ships and giving shape to Creole cultural innovations in the politically charged encounters between Africans and New World Europeans. Crucial to Mintz and Price’s argument is a move away from African retention and a view toward creolization and cultural invention according to a general grammar of value systems. But I also can appreciate the displacement of Africa from terra firma into the “tropics of discourse” (H. White 1978). Moving further afield from the concept of origins, and effectively jettisoning it altogether, Scott (1991) assimilates all retrievable African pasts to the ideologies and discourses in which they are expressed. What are significant from this perspective are not knowable connections but the conditions under which they are asserted or denied. And finally, with Gilroy (1993), the very shape of the African diaspora has radically changed,

Creolization and Connaissance / 67

no longer a vector out of Africa to the Americas, but a triangulated field of European, African, and Caribbean confrontations and exchanges bringing the Black Atlantic into graduated focus and form (see also Palmié 2002). If my own application of a West African interpretive paradigm to problems of syncretism and creolization in the New World falls somewhere between Bastide’s sociology and the Mintz and Price encounter model, I do not offer it as the only approach. As the Petwo paradox illuminates the dynamics of creolization in Haiti, it also does so—following Scott—in relation to the ideologies of elites, of Vodou practitioners, and of noirisme alike as part of the picture on the ground. Recall how the Bizongo priest quoted by Wade Davis incorporated noiriste rhetoric into his official séance. Or recall how the threat of the tonton-makouts confronted Ginen on the rural démembré. Ideological inventions of an authentic Africa, whether by nationalist elites or a rural peasantry, bear directly on the politics and practice of Vodou in its most traditional sites and guises. But they do so in relation to a critical substrate that continues to exist even when ideologically denied. Moreover, this substrate model is not committed to simple vectorial dispersion out of Africa. For when readers return, with Gilroy, from Haiti back to Africa, following those circuits of transatlantic trade that creolized the Guinea Coast for at least four centuries, they find the fetishism of the foreign extending well into the hinterland.29 If the problem of African origins has been progressively displaced in studies of the African diaspora, it always returns. And when it does, as the Petwo paradox suggests, critical models from West Africa disclose the negative dialectics of New World inventions. The criteria of recognizing an indigenous Africa may remain under debate, but they do not remain under permanent erasure. Even Mudimbe concludes his critique of the genealogy of gnosis in Africanist discourse with the final judgment that “the space interrogated by the series of explorations in African indigenous systems of thought is not a void” (1988, 200). In this chapter, it is the depth and span of such a nonempty space that I have explored, beneath and across the waters.

THREE

Notes from Ekitiland

Herbalist, diviner (babaláwo), orisha scholar, and devotee Pierre Fatumbi Verger spent the better part of his venerable life on roads less traveled by. I refer not only to the routes than he traveled throughout Yorubaland in Nigeria, Togo, the Republic of Benin (then Dahomey), and across the Atlantic to Bahia in northeast Brazil, where he diligently documented the Yoruba diaspora in what have become classic studies of orisha worship.1 It is true that Verger’s intellectual biography inscribes a cultural and historical geography, enriched by place-names, praise-names (oríkì), sacred centers, and colonial archives that recall the secrets of the living and the dead. But beyond the avenues of extraordinary scholarship, and perhaps less visible to uninitiated eyes, are the roads (ọ na) and pathways that can be opened only by privileged access to a hidden cartography, one that remaps space and time according to coordinates housed within orisha cults navigated by ritual specialists. On these roads Fatumbi also was a traveler, and for every tome of his published research there are others that will never be written. I open thus to praise a great elder who has opened the way for those of us who—in different ways—have followed his lead along the main roads and narrower byways of Yorubaland. The journey is of course endless, for Yoruba culture is, as I have argued elsewhere (Apter 1992), infinitely deep ( jinlẹ̀ ). And it also is regionally diverse. Within the singular designation of a great civilization exist subcultural variations (Oyo, Igbomina, Ijebu, Egbado, Egba, and so on)—what Olabiyi Yai calls “many Yorubalands” (personal communication)—that trace a common genealogy from Ile-Ife, the cosmographic locus of Yoruba kingship and modern ethnic identity. Within these subcultural groups, ritual pantheons vary, reflecting local sociopolitical relations and historical consequences of internecine warfare, the expansion of markets, and considerable internal migration. One of Verger’s

Notes from Ekitiland / 69

great contributions is his careful documentation of local cults, illustrating the historical dialectic between ritual archetypes and their diverse manifestations in various kingdoms and communities. Thus the archetypical Orishanla (Obatala) is variously reconfigured as Osalufon in Ifon, Orisha Oluofin in Iwofin, Orishagiyan in Ejigbo, Orishaoko in Oko, Orishawu in Owo, Orishaeguin in Eguin, Orishajaye in Ijaye, and Obatala in Oba (Verger 1982, 252), revealing how Yoruba deities are ritually incorporated into local kingdoms and centers of power. That the same revisionary dialectic unfolded in the New World, under the very different historical conditions of slavery, class formation, ethnic pluralism, and colonial culture, is illustrated by Verger’s careful comparisons of the orisha on the two sides of the Atlantic, between the orisha cults of West Africa and their reincarnations in, for example, Brazilian Candomblé, Umbanda, and Macumba. The aim of this chapter is to supplement Pierre Verger’s Notes sur le culte des orisha et vodun (1957) with additional notes on orisha cults and Epa masquerades in the Ekiti Yoruba highlands, the northeastern domain of Yorubaland in Nigeria where I served my ethnographic and ritual apprenticeship.2 In one sense my goal is simply documentary, for the religious and political associations of the Ekiti Yoruba region are relatively neglected in any systematic way, with preference generally accorded to the imperial traditions of the Oyo ritual field,3 to other subcultural groups,4 or to Pan-Yoruba theological principles.5 But if new data from Ekitiland add to the archive of Yoruba studies, they also suggest new ways of understanding “la place . . . occupée par l’orisha dans l’organisation sociale” (Verger 1982, 17) and the sociopolitical dynamics of the ritual process itself. In this respect, the Ekiti material has analytical implications that extend throughout Yorubaland and across the ocean to the Americas where the òrìṣ à/orixás/orichas go marching on.

Kingdoms and Cults As Pierre Verger (1982) notes, the variable positions of the orisha throughout Yorubaland are related to varieties of kingship and chieftaincy that have developed in various areas and towns. In the historically traditional kingdoms of ancient Ife and Old Oyo, for example, where authority is still vested in the office of the ọ ba and is administered through lines of civil chiefs and title holders, kings have ruled their territories and subordinate towns with the support of their royal orisha cults. Since all Yoruba kings rule èkejì òrìṣ à or “second to the gods,” it follows that those celebrated kings who shaped the political topography of Yorubaland are associated with the orisha that they served. The prominence of Shango as a Pan-Yoruba deity, reflecting the

70 / Chapter Three

politico-ritual sovereignty of the former Oyo empire (administered as it was by ajẹ́ lẹ́ and ìlárí Shango priests) is perhaps the best known spiritual refraction of imperial power. But the general principle holds that nearly all great orisha are associated with great kings, who enlist the support of the deities and embody their power (àṣ ẹ ) in annual festivals. A glance at the iconographies of cult priests and priestesses reveals further associations between the orisha and kingship—the beaded fringe, calabashes (igbá), coral bracelets, and ritual staffs of the devotees (so beautifully illustrated by Verger’s photographs) recall the king’s beaded crown and insignia of high office, for these ritual officiants are “kings” within their cults and bring the power of their orisha to replenish the body politic. In centralized kingdoms that boast powerful royal cults, such as Oshun in Oshogbo, or Shango in Oyo, there is a tendency to identify a single orisha as the patron of the town, much like the patron saints of Europe and Latin America. But the singularity of these orisha, like that of the kings, should not be overemphasized. If a king is singular, he also is multiple, embodying the town quarters (àdúgbò) and subordinate communities of his civil chiefs (ìwàrẹ̀ fà) and baálẹ̀ , which together compose the political and administrative units of his kingdom. These chiefs, in turn, worship their own specific orisha, mobilizing their devotees into distinct ritual communities. In a general sense, the distribution of political power is configured in the ritual domain, where authority relations are reproduced and transformed by the power of the cults. What happens in communities that are politically less centralized, as in subordinate towns where the baálẹ̀ wears the more modest àkòró crown, or in decentralized village clusters that have no formal king but rotate leadership between titled elders? It is to Verger’s credit that he identified the variable political frameworks in which Yoruba orisha worship is institutionalized: the centralized kingdom, the subordinate town, and the independent village (1982). But his energies remained focused on orisha cult practices and he never really pushed the analytical implications of the sociopolitical contexts themselves. That the organization and distribution of orisha cults throughout Yorubaland corresponds to variable degrees of political centralization and bases of social action, however, suggests a closer association between politics and orisha worship than has generally been recognized. Classic studies of Yoruba religion, including Dennett (1910), Farrow (1926), Frobenius ([1913] 1968), Talbot (1921), and Idowu (1962), identify the family or lineage (“clan” in some terminologies) as the basic or “original” social unit associated with orisha worship. This is of course understandable because the orisha are in many ways conceived as ancestors who

Notes from Ekitiland / 71

became deities after departing from earth, with their descendants serving as their devotees. In colloquial praises (oríkì), a devotee can be identified as ọ mọ (child) of his or her orisha, as in “ọ mọ Ògún” or “ọ mọ Yemoja.” Or, in related usages, all members of a town quarter (àdúgbò) can be addressed as “child” of the orisha that is owned by their chief. But such expressions are metaphorical, deploying the language of ritual—as distinct from social—kinship, and should not be taken too literally. For example, the same devotee can be addressed as ìyàwó (wife) or ọ kọ (husband), ìyá (mother), or yeye (grandmother) of her orisha, as determined by her position within the cult or even by the phase of the ritual process. Indeed, in the metaphorical language of ritual kinship, all cult members belong to a single “family,” as agnates, affines, even as slaves, but the designation does not imply that they are genealogically related, although of course they can be. Thus Bascom (1944) noted in his more sociological study that devotees of the same orisha can marry, whereas descendants of a common ancestor cannot. He concluded his study by classifying orisha cults into two types: what he called “single-sib” (where by “sib” he means lineage or clan) and “multi-sib,” observing that multisib cults could include a ward (àdúgbò) or even the whole town. The point should be clear that orisha worship differs significantly from classic patterns of ancestor worship found in “segmentary” societies such as Tallensi (Fortes 1945) or Lodagaa (Goody 1962), in that nongenealogical principles of recruitment are deployed through divination and dreams. Moreover, had Bascom realized that corporate descent groups (“sibs” in his terminology) represent political segments of quarters, and that quarters represent the political segments of towns and kingdoms, he would have identified political segmentation as the underlying principle of orisha cult organization. Certain cult offices, like the àwòrò or èyémọ lẹ̀ may be vested in specific royal or civil lineages, but inclusive memberships recruit much more widely. To illustrate the close association between politics and orisha worship, I will compare the cult organization and distribution of two northern Ekiti Yoruba kingdoms that differ along the dimensions just outlined. In the relatively decentralized kingdom of Ishan, which exemplifies the ministates more typical of the region (Obayemi 1971, 205–9), elaborate age sets organize corporate patrilineages into loosely federated villages and village clusters (Forde 1951, 80). The articulation of orisha and Epa cults with these patterns of segmentation is quite striking. In the more centralized kingdom of Ayede, however, which was founded by the warlord Eshubiyi, who created a military autocracy around 1845 (midway between the older towns of Ishan and Itaji), both royal and civil orisha cults take a correspondingly different

72 / Chapter Three

Map 3.1. The Kingdom of Ayede, ca. 1878.

form (map 3.1). Elsewhere I have discussed this area in the nineteenth century as a shatter zone between Ibadan’s military encroachments from the west, Ilorin’s incursions from the northwest, and chronic Nupe raids from the northeast (Apter 1992, 35–39; see also Akintoye 1971). My aim here is much more limited; namely, to provide a structural snapshot of the politicoritual configurations of these two kingdoms in the ethnographic present of 1983–84. If this approach appears unduly synchronic, it is not to deny the complex history of this region or the impact of Christianity and British overrule, but to better grasp the dialectics of orisha worship as it both shaped and was shaped by these historical interventions (Apter 1992, 165–211).

North Ekiti Polities Although classic patterns of Yoruba government and town settlement do not conform to north Ekiti kingdoms devastated by warfare and reorganized into defensive enclaves (Lloyd 1954, 1960, 1962), they do illuminate underlying

Notes from Ekitiland / 73

principles of resettlement. In the simplest model of the precolonial polity, a kingdom consists of a town ruled by an ọ ba (king) with his council of ìwàrẹ̀ fà chiefs and is surrounded by lineage-held lands and uncultivated bush. Although the king, who is sacred, stands apart from the royal lineage to represent the kingdom as a whole, his chiefs represent quarters (àdúgbò), with their titles vested in specific lineages. Descent is agnatic, with lineages (ìdílé) internally differentiated into “children of one father” (ọ mọ baba kan) and “children of one mother” (ọ mọ ìyá), who inherit various types of property and define potential lines of fission.6 In a large lineage, title holders of lineage segments may serve as subchiefs of a town chief, and heads of smaller lineages may associate themselves with more powerful lineage chiefs, who come to represent lineage clusters that together constitute quarters. Virilocal residence groups agnates and their wives into compounds (agbo ilé) that represent lineages “on the ground” at depths of three to four generations. According to this classic model of the precolonial polity, simple towns developed into complex kingdoms through lineage fission and out-migration to surrounding farm settlements (Lloyd 1962, 56). As lineages expanded and new land was cleared farther and farther from the town, men who wished to cultivate their own farms established temporary settlements and farm huts (abà) where they lived and worked for prolonged periods of time before returning to their town. Over time, these temporary farm settlements became permanent subordinate towns in relation to the capital town from which they sprang. This organic model of complex kingdoms entails a developmental cycle. When a parent kingdom reached a critical density, and competition for land and titles intensified, lineage segments peeled off to the periphery, where they aggregated to found subordinate towns that reproduced the structure of the metropolitan center but on a smaller scale and were ruled by a baálẹ̀ rather than a sacred king. These subordinate towns, regarded as children of the capital, could continue to grow until, through similar processes, they too formed subordinate settlements of their own. They could, however, also rise up and “slaughter” their “parents”—Yoruba history is full of accounts of how subordinate towns asserted their independence through claims to a beaded crown. The most famous example is that of Old Oyo, which originated as Owu’s subordinate town (S. Johnson 1921, 149; Law 1977, 37; Mabogunje and Omer-Cooper 1971, 31–41) and broke away to establish a ruling empire. But according to Yoruba political ideology, subordinate towns must respect the authority of their father kingdom by virtue of its political seniority and dynastic connection with Ile-Ife. There can be little doubt that subordinate towns did and do develop this way—their Yoruba terms abúlé and eréko also mean “farm shack” and

74 / Chapter Three

“agricultural area”—but kingdoms acquired subordinate towns through conquest and intimidation as well as military alliances. Oral traditions of conquest, however, are generally sweetened by euphemisms, for Yoruba ideas of legitimate authority require an ọ ba to rule by virtue of his royal genealogy and reputable judgment and not by the military prowess of his ancestors. A common euphemism of conquest is that a town invited the conqueror to assume leadership and the former ọ ba stepped down to devote his attention to town rituals (Lloyd 1955b, 24). Thus, if the organic model of subordinate town formation is rooted in the actual developmental cycles of kingdoms, it also is perpetuated by Yoruba ideology, which identifies the unity of the kingdom with the king and traces its origin through his forebears. The centrifugal idiom of complex kingdoms developing from a common center and origin is in fact built into Yoruba creation myths, which tell how Oduduwa descended from heaven (ọ̀ run) to Ile-Ife, where he fathered the first Yoruba kings, who then dispersed to found their own major kingdoms, except for the Ọ́ ọ̀ ni, who remained at Ife. If we accept the classic pattern of Yoruba settlement as a hypothetical baseline, then the effects of raids on the Ekiti slave “reservoir” and the more destructive nineteenth-century wars reversed centrifugal processes of subordinate town formation into centripetal processes of resettlement. Instead of subordinate towns growing out of simple kingdoms, the remnants of vanquished kingdoms came together to form nucleated defensive settlements. Survivors of devastated towns formed quarters (àdúgbò) within refugee kingdoms, reproducing their former social, political, and ritual structures on a reduced scale while preserving their unique identities in praises (oríkì), rituals, and histories of common provenance. Where towns remained intact but sought powerful allies, they became client towns of resettled kingdoms, relinquishing their kingship and providing soldiers and public service to their patron king in exchange for collective security. The northern nineteenthcentury Ekiti warfare state thus developed as an amalgam of towns, with each quarter possessing its distinctive oríkì, chiefs, and orisha. The kingdom of Ishan is in this respect quite typical of the region.

Ishan Kingdom Oral histories (itàn) relate how Ishan consisted of a number of small scattered settlements that combined into one town for mutual protection (see also Oguntuyi 1979, 94–96). Of five such villages, two are still considered strangers: Ilusajumu, which migrated from Iye, and Irefin, itself a composite from Ipole (not to be confused with Ilesha-Ipole) and Ila, the famous

Notes from Ekitiland / 75

Igbomina town (S. Johnson 1921, 423; Pemberton 1979), whereas OkeIshan, Ogilolo, and Adisa villages constituted the original core of settlers. Ishan also has two subordinate towns, Ilemesho and Ilafon, that, according to palace histories, migrated out from the father town. But the people of Ilemesho and Ilafon claim independent origins, the former from Ode-Ekiti and the latter from Ipole (where they broke from Irefin), and explain that their alliance with Ishan was protective. One wonders how much protection Ishan actually offered. The town was sacked first by the Ibadans under Balógun Aganigan, then a few years later, around 1845, by Balógun Ali of Ilorin, and again around 1875 by Balógun Ajayi Ogbori-Efon, the lieutenant of the infamous Àrẹ -ọ̀ nà-kakan`fò of Ibadan.7 In the idiomatic language of Ishan’s historical memory, only “eight people and a dog” survived. The ọ ba or Oníṣ àn of Ishan is among the sixteen Ekiti kings who trace their descent from royal Ife princes who wear beaded crowns and slippers and carry ceremonial staffs of office. Like other Ekiti kingdoms, Ishan is small and compact, with a population of roughly five thousand organized into semiautonomous quarters in the metropolitan capital or its subordinate towns. Whereas the Oníṣ àn represents the unity of the kingdom as a whole, his town chiefs (ìghàrẹ in northern Ekiti dialect, ìwàrẹ̀ fà in standard Yoruba) represent the autonomy of its àdúgbò (quarters), whose interests they serve during periodic meetings in the palace (áòfin in northern Ekiti dialect) with the king. Hence the proverbial wisdom, Ọ ba kìí pe méji l’áòfin, Ìjòyè l’épe mẹ́ fa l’áòfin (There can never be two kings in the palace, but there can be six chiefs in the palace), which not only states the constitutional relationship of a king to his chiefs in council, but also alludes to the fissiparous danger of a chief who aspires to become “king” of his quarter. The rank order of Ishan’s six town chiefs and their quarters is as follows: the Onírèfin of Irefin, the Arunsin of Ogilolo, the Olú of Ilusajumu, the Adàrà of Adisa, the Ejimọ ko of Oke-Ishan, and the Ọ baísà of Ilale, the last of which is also a subquarter of Irefin quarter (fig. 3.1). Each of Ishan’s quarters forms a distinct political and ritual community, internally organized into agnatic landholding groups and stratified into age sets. In Irefin quarter, the Onírèfin presides over three subquarters (also called àdúgbò within this narrower political field): Ilale, Iroyi, and Alewa, each with its own chief and subchiefs. Whereas the Onírèfin meets with his subchiefs to discuss the affairs of Irefin as a whole, each of his subchiefs also meets regularly to discuss the affairs of the subquarter. As chief Ọ baísà explained, “If there is a dispute within Ilale, other members of Irefin [i.e., Iroyi and Alewa] will not hear of the matter.” Within Ilale subquarter, the Ọ baísà, who is always selected

76 / Chapter Three

Figure 3.1. Quarters in Ishan ranked according to senior chiefs.

from Inisa lineage, meets with the Ọ ban´lá of Abudo lineage, the Atẹ genọ of Egenon lineage, and the Ọ bajẹ mọ of Ijemo lineage. Within Iroyi and Alewa subquarters, similar patterns of political segmentation exist between lineage elders and their subquarter chiefs (fig. 3.2). The subchiefs of these three sections of Irefin quarter represent the top grade of elders (ẹ gbẹ́ àgbà) in Ishan’s age-set system as a whole. Below this

Notes from Ekitiland / 77

Figure 3.2. The political morphology of Irefin quarter.

senior grade of elders rank the àrẹ mẹ́ ta, a fusion of three junior age sets that organize younger males for communal work and, in the past, for military service. Although the ẹ gbẹ́ àgbà and àrẹ mẹ́ ta of Ilale, Iroyi, and Alewa sections convene separately as political units to regulate their internal affairs, they join together as one body in Irefin’s dominant Yeyerefin festival or to represent Irefin’s collective interests to the Oníṣ àn. And whereas chiefs Ọ baísà, Ọ basọ́ lọ̀ , and Ọ̀ dọ̀ fin represent the ẹ gbẹ́ àgbà (titled elders) of their subquarters to the Onírèfin, he in turn represents these elders as one body to the king. Below the ẹ gbẹ́ àgbà in the age-set system, the ẹ lẹ́ gbẹ́ —who served as warriors in the nineteenth century and as tax collectors under the British—are organized in a similar fashion. Whereas Ilale, Iroyi, and Alewa subquarters each has its own ẹ lẹ́ gbẹ́ grade led by its senior member, called the Ológun, collectively they are led by the senior Ológun of Irefin quarter, who, together with the senior Ológun of Ishan’s remaining quarters, mobilize as a single group under the Olóro of Ishan (fig. 3.3). If, in the past, the ẹ lẹ́ gbẹ́ of Ishan convened as one body to defend the town, today they do so to undertake major public

78 / Chapter Three

Figure 3.3. The ẹ lẹ́ gbẹ́ age-set organization in Ishan.

works, such as palace and road repairs or to help finance town development projects and cooperative unions. Assemblies of the ẹ lẹ́ gbẹ́ within each quarter, however, regularly occur every fourteen days. The same pattern of political segmentation and age stratification obtains in Ishan’s other quarters, with minor variations. In Ogilolo quarter, the Arunsin heads the ajọ́ àgbà or elders of the major lineages, and the Ológun represents the àrẹ mẹ́ ta of two more junior grades. The Arunsin title is vested in the Irunsin lineage and rotates between its three lineage segments, whereas the Ológun rotates between two separate lineages (Ipara and Irasa),

Notes from Ekitiland / 79

Figure 3.4. Rotating titles in Ogilolo quarter, Ishan.

whose members can marry each other (fig. 3.4). In Ishan’s other quarters, the Ológun title is vested in a single lineage or “ruling house.” Ishan’s two subordinate towns—Ilafon and Ilemesho—are each headed by baálẹ̀ who are subjects of the Oníṣ àn and thus cannot wear the beaded crowns (adé Olókun) of independent kingship, at least in the official public gaze. Both these baálẹ̀ , however, the Aláfọ̀ n of Ilafon and the Obadu of Ilemesho, claim independent origins from Ife as charters of their political autonomy, whereas the Oníṣ àn’s palace historians insist that they migrated out from Ishan. Ilafon’s Ife connection appears at least partially corroborated by its Iro cult, which came from the town of Ijumu and, in the past, sent delegates to the Iro cult in Ife’s Ijumu quarter (Bascom 1944, 31). In any case, the baálẹ̀ of both subordinate towns regard themselves as kings and resent the Oníṣ àn’s overrule. They have never regularly met with the Oníṣ àn but appeal to his authority to solve problems—usually land disputes—that they cannot settle themselves or that involve relations with other towns. In the past, Ilafon and Ilemesho were client towns of Ishan, providing warriors, tribute, and public labor in return for military protection. After the Pax Britannica, they provided public labor and tax in exchange for political and judicial representation in the district. Thus their official subordinate status

80 / Chapter Three

Figure 3.5. Rotating titles in Ilafon.

may well be a consequence of British colonialism, which ossified the more flexible alliances between Ekiti kingdoms into more hierarchical structures of administration (Akintoye 1971, 223–24). Today both towns have their own markets (ọ jà) and farm their own lands. Their relations with Ishan are based more on webs of kinship and affinity that have developed over the years than on political protection and support. The succession of a new baálẹ̀ , however, is still subject to the Oníṣ àn’s approval. The sociopolitical organization of Ilafon and Ilemesho resembles those village clusters (ministates in Obayemi’s terminology [1971]) found among the Akoko and Yagba Yoruba farther northeast. In Ilafon, the baálẹ̀ ’s title of Aláfọ̀ n, and that of his senior chief, the Ejimọ ko, rotate between the two major lineages of the town, and chieftaincy titles are vested in lineage segments (fig. 3.5). Four age sets—which, in descending order, are Kejido, Kemote, Kemeyo, and Eyokiti—articulate with Ilafon’s two maximal lineages and join

Notes from Ekitiland / 81

together during town festivals to proclaim their unity. Whereas Ilafon’s four lineage chiefs represent the elders of Kejido, representatives (olórí ẹ gbẹ́ ) of the more junior grades are elected by their members. The village of Ilemesho exhibits similar principles of political segmentation and age-set stratification but on a larger scale, with eleven lineages in the town.

Cults in Ishan In decentralized kingdoms like Ishan, each quarter (àdúgbò) forms a distinct ritual community, with at least one cult embracing all members of the quarter and several cults articulating with its sections, whether they are subquarters, lineages, or even lineage segments and compounds. Referring to Ishan’s five major quarters and two subordinate towns, one ìghàrẹ chief said, “Every town [ìlú] does its own festivals differently, and there is no combination of towns.” In other words, no two quarters combine within the same orisha cult (the term “town,” [ìlú] can refer to quarters [àdúgbò] in order to emphasize their relational autonomy, although a king would never use the term in this fashion). In Ishan kingdom, each of the five major àdúgbò has a dominant deity that represents the unique identity of the quarter and protects its members with spiritual powers: Oríkì (of all members)

Àdúgbò

Òrìṣ à

1. Irefin

Epa

Ọ mọ Epa

2. Ogilolo

Alakua

Ọ mọ Alakua

3. Ilusajumu

Babatigbo

Ọ mọ Agan

4. Adisa

Yeye-ewu

Ọ mọ Yeye-ewu

5. Oke-Ishan

Olua

Ọ mọ Olua

Membership in these quarter-based cults is exclusive, because members of one quarter cannot belong to the orisha cult of another, although they can attend its festivals as affines and matrilateral kin. The close association between the quarter and its dominant cult is expressed in idioms of common descent, as in “X (the orisha) is the father (or mother) of us all,” or “Y is the ancestor who brought us here,” but such claims are not intended as statements of filiation. It is only metaphorically, in the idiom of ritual kinship and descent, that the members of each quarter and its dominant cult are “of one blood.” The corporate identity of each quarter is further marked by the oríkì (praises) of its members, which include the appellation “child” (ọ mọ ) of the deity that they worship together, and also by a collective taboo against

82 / Chapter Three

eating new yams or selling them in the market, which is not lifted until the following season during the orisha’s annual festival. Orisha cults own property and control ritual power. Ritual paraphernalia— such as masks, effigies, staffs of office, medicines, iron rattles, knives, instruments used to contact the orisha, and sacrificial altars (ojú ẹ bọ )—are housed within a town shrine (ipara òrìṣ à) located in its associated quarter, generally close to the quarter chief’s house. Here the initiated members meet every two markets (that is, every nine days) to feed the deity with sacrifices, discuss cult finances and affairs, and to treat private clients who “beg” the orisha for personal assistance. Quarter cults generally have a bush shrine (igbó imọ lẹ̀ ) where the orisha dwells. This area lies beyond the residential bounds of the town and remains off-limits to outsiders because it is the locus of the orisha’s spiritual power. More important than the shrines themselves, however, is the ritual power (àṣ ẹ ) that they house and the body of techniques that control it. The most highly valued cult property consists of secrets (awo), such as incantations and specialized sacrifices that harness the orisha’s power and direct it toward specific ends. Such property not only is restricted to the cults but also is differentially distributed within them according to ritual offices and levels of seniority. This preliminary sketch illustrates how the quarter and its cult are distinct corporations and that the former does not reduce to the latter. The cult does not consist of the quarter as such but performs its ritual on behalf of the quarter. This distinction is important because it underlies the complementary relationship between political and ritual domains. Whereas the ìghàrẹ chief heads the quarter, the high priest, or àwòrò, heads its dominant cult. The two offices cannot be fused because they control complementary sets of resources. A holder of both titles would simply wield too much concentrated power. Within Irefin quarter in Ishan, I was told that the Onírèfin could not be the àwòrò of either the inclusive Epa cult or the more limited Aku cult because he was a civil chief.8 Chief Ọ baísà of Ilale quarter (actually a subquarter within Irefin) explained his own distance from the Epa cult in the following terms: In my own quarter we have specialists in Epa worship, but I am not a specialist myself, just chief of the quarter. There is nothing that concerns a chief about the worship of Epa, just that I must attend the dancing after worship in the bush. Anytime people in the quarter wish to worship Epa, they will come to me and ask for objects of sacrifice: dog, salt, palm oil, àkàrà [bean cake] and orógbó [bitter kola]. I must give it to them but cannot go myself to the bush.9

Notes from Ekitiland / 83

A chief must provide for a sacrifice to ensure the safety and prosperity of his quarter, but he cannot officiate. He must avoid the bush sacrifice because it lies beyond his jurisdiction. The àwòrò is “chief” in the bush. Although high-ranking chiefs cannot assume the priestly office of an àwòrò, the àwòrò themselves are not without political position and influence, for they occupy nodal links between political and ritual domains. Within each quarter the àwòrò of the dominant cult also is a subchief, combining the patrimonialism of lineage headship with the gerontocracy of age-set elderhood. In most Ekiti kingdoms, the àwòrò also is the Ológun or representative of all age sets in the quarter, described as “the eldest of us all” in terms of social (rather than chronological) age. He is thus involved in three sets of meetings (ìpàdé): those of the quarter chiefs with his subchiefs, those of the senior age set within the quarter and town at large, and those of all àwòrò of a quarter, in which the head priests of lesser cults form a council that meets three times a year with the senior àwòrò to calibrate festival dates. These àwòrò also are kingmakers (afọ bajẹ ) within the quarter, for among their ritual duties are the installation and burial of quarter chiefs. The articulation of àwòrò with formal political arenas is not limited to town quarters but extends to the kingdom at large. Just as the town chiefs lead their àdúgbò and meet with the ọ ba to regulate town affairs, so the dominant àwòrò of each quarter meet with the king’s àwòrò—in Ishan he is called the Ejimọ ko—to regulate the ritual affairs of the kingdom. Here the political and ritual domains of the kingdom intersect, for the Ejimọ ko fuses several roles in one office. He is at once the leader of all age sets in the kingdom, the head of the Oníṣ àn’s kingmakers, the àwòrò of his own quarter cult, and representative of the àwòrò of other cults to the king. If, for example, the àwòrò of the Ogun cult in Irefin quarter prescribes a special sacrifice for the town, he must send a message to the Ejimọ ko, who delivers it to the Oníṣ àn. If the Oníṣ àn complies, he will tax the town through age-set representatives to purchase the required sacrificial animals and offerings, including kola nut and palm wine, which he will then send to the Ogun priest. The Ejimọ ko also meets with the quarter àwòrò to organize festival dates and collective sacrifices for the entire town. He officiates at the Ojo-Apapo festival at which Ishan’s five quarters and two subordinate towns convene to worship Olua, the Oníṣ àn’s orisha. He also will prepare the Oníṣ àn’s corpse with the other kingmakers and will lead the rituals of succession for the next incumbent. In the precolonial era, the council of àwòrò—called the ejio—was empowered to try cases of incest and offenses against the orisha. It also could mediate disputes between civil chiefs. Even today the ritual authority of the king’s àwòrò affords him certain privileges and immunities. Unlike any chief or

84 / Chapter Three

Figure 3.6. The Epa Ekunrin mask of Ilale subquarter, Ishan. Author’s photograph, 1984.

townsman, he does not have to prostrate himself (dọ̀ bálẹ̀ ) before the king; he can, like the king, sit on a leopard skin; and he even can override the king’s jural authority in unusual circumstances. Town historians relate how, in the past, a defendant could escape the king’s adjudication by running into the Ejimọ ko’s compound or by grabbing hold of his leg. However guilty the offender or serious the crime, the king, at least in principle, was required to “forget the matter.” Having sketched the general pattern of politico-ritual organization within Ishan kingdom, we can now zoom in with a more powerful lens to focus on the specific arrangements of orisha cults within the complex quarter of Irefin, the simple quarter of Ogilolo, and the subordinate towns of Ilafon and Ilemesho. Irefin quarter has three major cults—Epa, Aku, and Ogun—that embrace the quarter as a whole but divide into branches that align with its three subquarters: Alewa, Iroyi, and Ilale (see fig. 3.2). In the Epa cult each subquarter has its own masqueraders, organized by age sets and lineages (fig. 3.6). If all Irefin patrikin are praised as “ọ mọ Epa” (children of Epa) in reference to their common spiritual power, each of the subquarters owns specific Epa masks with associated oríkì. Alewa subquarter, where the Onírèfin resides, owns Yeyerefin or “grandmother” (yeye) of Irefin quarter as the dominant Epa mask (fig. 3.7). Iroyi subquarter owns Epa-Iroyi with its associated oríkì and iconography, and Ilale owns Epa-Ilale (also called Ekunrin) who, according

Notes from Ekitiland / 85

Figure 3.7. The Epa Yeyerefin mask of Irefin quarter, Ishan. Author’s photograph, 1984.

to oral traditions, brought the Ilale immigrants from Ila (an Igbomina town; see Pemberton 1979). Thus the Epa cult in Irefin represents the unity of the quarter as an identity-in-difference. Although the subquarters trace different origins—Ilale from Ila, Alewa and Iroyi from Ipole—they have fused to form a single Epa cult with a common shrine. The high priestesses of Epa reflect this fusion in clearly gendered terms. Beneath the àwòrò Epa in the cult’s hierarchy, three women bearing the title Ìyá Ẹ pa (mother of Epa) officiate for the women, including wives and daughters of Irefin quarter, and represent its three subquarters that they serve as female chiefs. The first Ìyá Ẹ pa comes from Idofin lineage in Alewa, the second from Idara lineage in Iroyi, and the third from Isaoye lineage in Ilale. If Irefin’s Epa cult represents a “multisib based cult” (Bascom 1944, 35), its worshippers liken it to a hand (ọ wọ́ ) with fingers (ìka). Irefin’s Aku cult is another such “hand” with “fingers” but, unlike the Epa cult’s public displays, it unites the quarter in hidden ritual. Aku is a mystery of the bush, a type of Oro or bullroarer cult that excludes women and confines them to their compounds during its annual seven-day performance. Although this cult belongs to the Onírèfin, the spiritual power (sometimes called ẹ bọ ra) “chose” the Ọ ban´lá of Abudo lineage in Ilale subquarter as its àwòrò. I was told: Aku is too big to protect only one man [i.e., the Onírèfin] and for only one man to control it, so the àwòrò has to come from Ilale. The person who owns

86 / Chapter Three Aku is at Irefin and the person who Aku chose as his àwòrò is at Ilale, so that should there be a dispute between them they should not split (pín) apart.10

The Aku cult safeguards against fission of Irefin quarter. Its three topranking priests come from Irefin’s three subquarters: the Ọ ban´lá of Ilale, the Ọ̀ rírẹ́ of Alewa, and the Àsábá of Iroyi, who represent the elders. They bring Irefin’s civil chiefs—the Onírèfin, the Ọ baísà, the Ọ basọ́ lọ̀ , and the Ọ̀ dọ̀ fin—to Aku’s bush shrine, where they sacrifice a goat, although in the past they sacrificed a stranger prepared with medicines. The Ọ ban´lá then interprets the voice of Aku through kola nut divination for each chief, advising them of additional sacrifices and temporary taboos that they must observe and extend to their followers. The àwòrò of other cults in Irefin also sacrifice to Aku, for Aku is conceived as their king and as a powerful deity who protected Irefin people from slave raids during the nineteenth century and from smallpox sent by rival quarters and towns. All àwòrò in Irefin meet in Aku’s bush shrine to set the festival dates of their respective cults. In addition to the Epa and Aku cults that embrace Irefin quarter at large, lesser cults serve its subquarters (ọ̀ gbọ́ n) and lineages exclusively. Thus each of the Onírèfin’s three subchiefs has an orisha vested in their lineages and performs rituals on behalf of these more limited jurisdictions. In Ilale chief Ọ baísà “owns” the Oshun cult, with the head (olórí) of Ilale’s age sets presiding as its àwòrò. Today, however, the priesthood is vacant, and chief Ọ baísà complains that his quarter is subsequently weak. Chief Ọ basọ́ lọ̀ of Iroyi subquarter owns an orisha known simply as Orisha-Iroyi, designating the close identification between subquarter and cult, and chief Ọ̀ dọ̀ fin of Ilewa owns Ogun. These subquarter cults are like any lineage cult that performs rituals on behalf of its members, with one important difference. Because they belong to chiefly lineages, their ritual powers and congregations extend beyond the lineage as such to embrace the chief’s political jurisdiction. A quarter or subquarter cult is deemed more powerful, in the ritual sense, than a simple lineage cult because its owner is politically more powerful and commands more followers. As with the quarter cult, ritual and political offices in subquarters are kept apart. Just as the Onírèfin cannot serve as the àwòrò of deity Yeyerefin, so the subquarter chiefs are barred from high ritual office. In simple lineage cults, however, the political and ritual roles of lineage elders are fused. The baálé or lineage head may sacrifice to the lineage orisha, but he is not considered an àwòrò in this more limited ritual capacity. Thus, in Ilale subquarter, while the Oshun cult vested in chief Ọ baísà’s lineage requires a separate àwòrò, the cults of its remaining lineages—orisha Obanifon for Isaoye lineage, orisha Alagba for Egenon lineage, and orisha Atoru

Notes from Ekitiland / 87

and Ogun for Ibore’s two lineage segments (indicating incipient lineage fission)—combine political and ritual roles in their lineage heads (baálé). These lineage heads serve as kingmakers (afọ bajẹ ) for their subquarter chiefs, evaluating eligible candidates for succession and performing limited installation rites. Thus far I have discussed the politico-ritual morphology of a single, albeit complex, quarter in Ishan. Ishan’s remaining quarters exhibit the same principles of orisha cult organization and distribution, with the exception of Ogilolo quarter (see fig. 3.4), which structures political and ritual domains through its rotating title system, a feature found in Ishan’s subordinate towns of Ilafon and Ilemesho as well. In Ogilolo quarter, the Arunsin’s Irunsin lineage owns orisha Alakua, and orisha Adodo is owned jointly by Ipara and Irasa lineages. Idara lineage also possesses its own orisha, called orisha-Idara. According to its pattern of serial rotation, when the Arunsin rotates from (a) to (c), and the Ejimọ ko rotates from (b) to (a), the àwòrò of Alakua rotates from (c) to (b). Similarly, as the Ológun also is the àwòrò of orisha Adodo, the priesthood rotates between Ipara and Irasa lineages. The only nonrotating title in Ogilolo quarter is that of the Oníjòfì, which is perpetually vested in Idara lineage. A similar system of titular devolution occurs in Ishan’s subordinate town of Ilafon. There, if the Aláfọ̀ n (the baálẹ̀ of Ilafon) comes from Oke-Aofin lineage, then the àwòrò from Ilafon’s dominant Iro cult must come from Inisa lineage (see fig. 3.5). Since the next Aláfọ̀ n must be selected from Inisa lineage, the àwòrò of the Iro cult must succeed from Oke-Idofin lineage. Even though Inisa lineage professes to own the Iro cult, and claims that it invited Oke-Idofin to join, the àwòrò of Iro must rotate against the incumbent Aláfọ̀ n in complementary succession. Within both lineages, however, lower political and ritual titles remain fixed. Inisa’s two lineage segments—Ilara and Ilasi—own the orisha Oshun and Osanyin respectively, and Oke-Aofin’s three lineage segments—Aofin, Oke-Omito, and Idemo—own the orisha Oshun-Aofin, Olokun, and Orishanla, respectively. The àwòrò of these three cults also are leaders of the age sets within their lineage segments and serve as the Aláfọ̀ n’s kingmakers. The subordinate town of Ilemesho exhibits a slightly different pattern of titular rotation linked directly to its political organization. Although roughly the same size as Ilafon, with a population of about eight hundred, Ilemesho is a cluster of eleven patrilineages organized into six quarters, each with its political and ritual responsibilities. The baálẹ̀ , titled the Ẹ lẹ́ mẹ̀ ṣ ọ , is primus inter pares, rotating between the six chiefs of each quarter. Each of these chieftaincy titles rotates, in turn, between the lineages of their quarter. The

88 / Chapter Three

Figure 3.8. Rotating titles in Oke-Ade quarter, Ilemesho.

lineages of these quarters own their orisha collectively, with the àwòrò rotating against the quarter chief. Thus Oke-Ade quarter worships orisha Oloroke as its dominant cult. When the quarter chief comes from Oke-Ade lineage, the àwòrò comes from Iwolobo lineage; when chief’s successor comes from Iwolobo lineage, the àwòrò will succeed from Iwoyo lineage (fig. 3.8). This complementary rotation of political and ritual titles binds the lineages of all quarters together. What makes Ilemesho so interesting and unusual are its overlapping politico-ritual jurisdictions that further bind the quarters to each other within the town. Cults of the orisha Iyere, Iro, Iroko, and Oloroke are in fact owned by lineages of different quarters, militating against the ritual autonomy of each quarter as a discrete political unit (fig. 3.9). Thus the orisha Iyere is owned by three lineages—Oke-Ade, Aofin, and Ile-Iya—whose members not only have different duties within the cult but also belong to two separate quarters. Whereas Aofin and Ile-Iya are two lineages within Aofin quarter, Oke-Ade is a lineage within a quarter bearing the same name. The Iro cult cuts across three quarters—Oke-Otun, Ilesire, and Aofin. When I asked whether there were any connection between the Iro cults of Ilemesho and Ilafon, devotees in Ilemesho maintained that the two cults were separate, and that theirs was original from time immemorial (láíláí). Devotees in Ilafon, however, claimed that their Iro cult secrets were stolen by the people of Ilemesho to establish the cult on their own. In Ilemesho, orisha cults that cut across quarters militate against political

Notes from Ekitiland / 89

Figure 3.9. Cross-cutting cults in Ilemesho.

fission and political centralization by uniting them in shared ritual obligations. In such a small village hamlet with a weak and decentralized headship that rotates between its six quarters, each quarter is apt to break away over succession disputes. In a town with no official royal lineage (pace Nigeria’s Morgan Report on “traditional chieftaincies”), where each quarter enjoys the prerogative of eventual headship, competition for high office precipitates open contests for power and influence between chiefs. Since every chief is a prospective king, each quarter can invoke its rights to form a prospective kingdom. A town of Ilemesho’s small size and limited resources, however, cannot afford political fission and out-migration. Cross-cutting political and ritual corporations sustain ties of mutual interdependence. As one Ilemesho man explained with reference to the entire town, “We worship (bọ ) the orisha because we are one family,” a statement that, in strict genealogical terms, is false, but one that represents, in the idiom of ritual kinship, the cohesive role of Ilemesho’s cults. When I asked which orisha cult was in fact the most powerful, the baálẹ̀ explained, “In Ilemesho, no one orisha is more powerful than another. Every one has its own power.”

90 / Chapter Three

Cult Transformations in the Kingdom of Ayede I have examined the intricate relationship between Ishan kingdom and its ritual associations in what may seem like a tedious and sterile exercise because it demonstrates the systematic complementarity between ritual power and political authority in a little-studied area of Yorubaland, where classic models of Yoruba kingship and political hierarchy give way to more decentralized ministates and village clusters and where age sets and rotating titles figure prominently as mechanisms of politico-ritual organization and devolution.11 What appears complementary from an institutional perspective is of course dialectical from the standpoint of political competition and ritual practice, in that much maneuvering occurs around the fixing of festival dates (Apter 1992, 115), the allocation of ritual roles, the execution of ritual procedure, and the recognition of ritual reciprocities. Indeed, it often is during an orisha festival that a civil chief signals his fission from the kingdom by avoiding the king’s palace, withholding ritual tribute, and even crowning himself king. Similarly, the àwòrò of major orisha cults can signal public disaffection with the king by performing their rituals on a drastically diminished scale and thereby depriving him of their regenerative powers. If in principle ritual is mandated to follow the same routines from time immemorial (láíláí), in practice it changes dramatically, reflecting and even precipitating the revision of authority structures, such as the ranking of civil chiefs or the emergence of new political cleavages. But the conflicts and contradictions that ritual mediates and instantiates are governed by a logic of coimplication, whereby kings, chiefs, subchiefs, quarters, and age sets are ritually reproduced and reconfigured every year. If orisha cult titles are vested in specific lineages, the lineage as such does not establish the social framework of orisha worship, as is so commonly maintained. The precise calibration of political and ritual titles and memberships establishes political segmentation as the underlying principle of orisha cult organization within kingdoms and their subordinate towns. So basic and profound is this underlying politico-ritual complementarity that it has persisted throughout the history of Islamic and Christian conversion as well as colonial overrule and has in fact shaped popular responses to these external interventions. It is with this politico-ritual dialectic in mind that we can travel two kilometers south on the narrow road, once tarred but now reduced to dirt and stubborn potholes, that connects with the neighboring kingdom of Ayede, a historically significant military stronghold that stands out among northern Ekiti kingdoms for its unusually centralized government and elaborate orisha cults.

Notes from Ekitiland / 91

The kingdom of Ayede was founded around 1845 by the warlord Eshubiyi, who migrated about twelve kilometers south from the village of Iye to establish an important refugee settlement and military base for Ekiti resistance against the Ibadans. Listed by Samuel Johnson (1921, 23) among the sixteen “traditional” Ekiti kings, and described by his oríkì as “Lord of the Yagba and Akoko tribes” (Johnson 1921, 403), the Àtá of Ayede attracted settlers from a variety of Yagba towns together with a core of Iye indigenes who today comprise six quarters. The first four quarters represent the original migrants from Iye, organized into Ejigbo, Owaiye, Isaoye, and Ilaaro. Two additional quarters, Omole-Akodi and Egbe-Oba, represent strangers from Iyagba and Ikole towns, who allied with Eshubiyi during his military campaigns and settled to form the “other half” of Ayede, with chiefly representatives in the palace. Because I have described the history and politics of Ayede kingdom elsewhere (1992, 35–93) in considerable detail, documenting the displacement of the former Iye dynasty, the centralization of political power, and the social history of its orisha cults, I will limit my present discussion to a single transformation that occurred in the ritual domain: namely, the consolidation of distinct orisha within Ayede’s dominant cults. Eshubiyi’s rise to power was based on military prowess, political cunning, and the promotion of two orisha cults: Yemoja, which he received from the Olúyọ̀ lé military chief of Ibadan in what proved to be an unsuccessful attempt to secure Eshubiyi’s loyalty, and Orisha Ojuna, a cult from Ikole that Eshubiyi received through his patriline in Iye and developed into a royal cult within Ayede. Local historians (òpìtàn) and Yemoja devotees relate how it was during a Yemoja festival that the warlord Eshubiyi proclaimed himself a traditional king, producing a beaded crown as his supporters announced “Kabíyèsi, Àtá Ayédé!” to herald his paramount status. To this day, the Yemoja cult remains the guardian of the Àtá’s Olokun beaded crown, which is housed in the town shrine (ipara Yemoja) and invested with àṣ ẹ . Although the senior cult offices with the Yemoja cult belong to the king’s Idomogun ruling lineage (the depth and span of which are subject to continuous dispute and litigation), devotees have been recruited through cognatic ties from all of Ayede’s quarters, consolidating ritual support throughout the town. As a result, devotees of distinct orisha within their quarters combined to form a complex cluster of deities within the Yemoja cult itself. Thus beneath Yemoja, who remains paramount and is represented by the Yeyeolokun priestess, are Shango, Orisha Oko, Oshun, Erinle, Olokun, Ogun, and Bayoni, each with a dominant priestess and subordinate “line” of titled devotees. The relationship between these clustered orisha is configured by imputed ties of kinship and affinity (fig. 3.10) and are spatially represented by the relative positions

92 / Chapter Three

Figure 3.10. Kinship and affinity in Ayede’s Yemoja cult pantheon.

of altars within Yemoja’s town and bush shrines. The priestesses of Yemoja are thus priestesses of many deities, which are honored, praised, fed, and incarnated during Yemoja’s annual festival. A similar clustering of deities around a central orisha exists in Ayede’s royal cult of Orisha Ojuna, which also cuts across quarters. To consolidate politico-ritual support from the “stranger” quarter of Egbe-Oba in addition to the Iye quarters of Owaiye and Ilaaro, which previously worshipped their respective orisha (i.e., Oloke, Olua, and Osanyin) separately, Eshubiyi reorganized Orisha Ojuna’s town shrine to house separate chambers for each of these deities and established ritual networks with smaller shrines in various quarters (fig. 3.11). Thus the Olua cult of Owaiye and the Oloke cults of Egbe-Oba can no longer assert their former ritual autonomy but must participate in the worship of Orisha Ojuna, which officially contains them. Although each of these quarters has a recognized civil chief, they do not form distinct ritual communities but have fused their cults within Orisha Ojuna’s inclusive field of ritual command. In contrast to the royal cults of Yemoja and Orisha Ojuna, the civil cult of Orisha Iyagba belongs to the Balógun Áòfin, chief of Omole-Akodi quarter and representative of Yagba towns within the broader kingdom. Worshipping in a distinctive Yagba dialect and attended by ibembe, agere, and ikoko talking drums, Orisha Iyagba’s devotees recall the honor and glory of Eshubiyi’s former war lieutenant and mobilize the Yagba “strangers” in an uneasy alliance with the palace. Playing with the tensions of potential usurpation, the devotees represent ritual warriors who brandish cutlasses and spears that

Notes from Ekitiland / 93

Figure 3.11. Centralized recruitment to the Orisha Ojuna cult in Ayede.

both uphold and threaten the Àtá’s authority (S. Johnson 1921, 156–61). During the reign of the Àtá Omotosho II (1948–91), the female head (Ológun) of ritual warriors was also the king’s fifth wife, bringing Orisha Iyagba into the palace and binding the cult to the king through affinity. But like marriage, ritual alliance is prone to annulment, and the cleavage between Orisha Iyagba and Ayede’s royal cults represents a division of the town into two sides, with the Yagba “strangers” constituting the other half. As a result, the cult of Orisha Iyagba has fused a cluster of its own subordinate orisha owned by separate lineages and settlers of Omole-Akodi quarter. Beneath Orisha Iyagba proper (also called Iyelori), we find Ereo from the Balógun Áòfin’s Iletogun lineage, Ogbon-Ilele and Oloye from the Yagba town of Ipao, Ore from Irele, Iroko from Ejuku, Olooke and Okutaaro from Itapaji, and Agiri from Ogbe, each with its ritual representative and associated oríkì. Thus organized like a ritual kingdom within a kingdom, the cult of Orisha Iyagba actually replicates the pattern of cult fusion and consolidation found in Yemoja and Orisha Ojuna. If Orisha Iyagba is a singular deity, its cult houses a multiplicity of lesser gods and goddesses representing immigrants from towns of Yagba origin (fig. 3.12).

The comparison of Ishan and Ayede kingdoms and their politico-ritual configurations suggests a few new lines of inquiry into Yoruba orisha worship. First, at the level of empirical description, it documents a number of deities

94 / Chapter Three

Figure 3.12. The town shrine (ipara) of Orisha Iyagba in Ayede. Author’s photograph, 1983.

and orisha cults in a relatively neglected region of northeastern Yorubaland, cults associated with specific kingdoms and quarters that were caught in the cross fire of the nineteenth-century wars. The distribution of cults, their ritual practices, and iconographies recall the devastation of towns, the resettlement of refugees, and the political and military alliances that the people of Ekiti formed. Local cults of this region may seem like minor additions to the growing archive of orisha studies, but their significance lies precisely in their peripheral status on the margins of Yoruba identity. I would suggest that future studies of orisha and ebora in the Akoko, Yagba, and Kabba regions will improve our understanding not only of the politico-ritual topography of Yorubaland but also of the historical processes of Yoruba ethnogenesis itself (Peel 1989), as peoples of the periphery came to see themselves as Yoruba. In addition, they will provide a broader body of evidence for documenting local migrations and the role of ritual in generating multiple allegiances, identities, and historical memories in complex kingdoms and quarters. Second, on a more analytical level, our study suggests a powerful if limited method of controlled comparison that can be deployed between kingdoms or even subcultural regions to disclose the specific modalities of cult organization and the underlying principles that govern its variations. Within the decentralized kingdom of Ishan, we saw how the cults of complex quarters and subordinate towns articulated with political segmentation, serial devolution, and age-set stratification with a calibrated precision that underscores the co-implicative dimensions of Yoruba religion and politics and

Notes from Ekitiland / 95

the underlying dialectic that their variations manifest. In this connection, comparison with Ayede provided a significant contrast, revealing the correlative developments of political and ritual centralization in the founding of a military autocracy. If the family or lineage remains a salient idiom for extending ritual identities to broader political jurisdictions within kingdoms, it is as an idiom only. As such, it can be accounted for in ideological terms but should not be mistaken for the dominant framework of orisha cult organization itself, in that the lineage itself is a political and administrative unit within the more inclusive kingdom.12 Third, evidence from Ayede’s dominant orisha cults illustrates an important indeterminacy built into the ambiguous identities of the orisha themselves. If Yemoja, Orisha Ojuna, and Orisha Iyagba represent singular deities, their cults house a multiplicity of orisha clustered within the three town shrines and worshipped by distinct lines of devotees. The consolidation of deities within Ayede’s major orisha cults represents what may well be a common characteristic of orisha cults throughout Yorubaland (see, e.g., Barber 1990), wherein principal deities fragment into multiple spiritual manifestations with associated sacrifices and oríkì. The notion that every Yoruba orisha has a discrete and singular cult is entrenched in earlier Yoruba scholarship, but the notion must be abandoned to account for the complex clustering of distinct orisha around a principal deity. The revision of these last two misconceptions—that lineages establish the social framework of cults and that the relationship between orisha and cult is discrete—has implications that extend beyond Yorubaland proper to illuminate religious developments in the New World. Here we can return to Pierre Verger’s commitment to the broader reaches of the Yoruba universe, primarily within the mysteries of Brazilian Candomblé, which he has studied as a scholar and practitioner. I do not wish to belittle Verger’s painstaking research in sifting through all available historical documents and reconstructing the histories of the terreiros by drawing attention to one misleading contrast that he establishes. Comparing the worship of orisha in Yorubaland with Candomblé ceremonies in Bahia, he states (1982, 72) that La différence entre les cérémonies pour les orisha en Afrique et dans le Nouveau Monde vient surtout de ce qu’en Afrique on évoque, en principe, un seul dieu au cours d’une fête célébrée dans un temple réservé à lui seul, alors que sur le Nouveau Monde, une gamme très étendue d’orisha sont appelés dans un même terreiro au cours d’une même fête. De plus, en Afrique, une telle cérémonie se célèbre généralement pour la collectivité familiale et un seul elégun [possession priest or priestess] est normalement monté par le

96 / Chapter Three dieu, alors qu’au Nouveau Monde, cette collectivité familiale n’existant plus, l’orisha a pris un caractère individuel et il arrive qu’au cours d’une même fête, plusieurs iaôs [ìyàwó, or “wives” (priestesses) of the orisha] soient montés par un même dieu, et ceci, tout à la fois pour le réconfort particulier d’un individu déterminé et la satisfaction additionnelle de tous ceux qui font le culte de cet orisha.13

Evidence from the Ekiti Yoruba highlands suggests that this contrast is overdrawn, because orisha cults are not characteristically discrete—housing one deity only—but, like their Bahian counterparts, consolidate clusters of orisha within town shrines. Furthermore, as the detailed examination of politico-ritual configurations in Ishan and Ayede kingdoms reveals, orisha cults in Africa are by no means limited to collective family members but embrace the dominant political segments of kingdoms and town quarters. Moreover, within such cults, as in the Bahian terreiros, priests and priestesses can be possessed by multiple deities. The implication developed by Verger is that the destruction of the African lineage or collective family in the New World, following from the inhumanities of slavery, created new principles of orisha cult recruitment and thus novel forms of cult organization and religious practice by combining distinct deities within the Candomblé cults. My point is not to deny the radically different social conditions in which Brazilian Candomblé developed, but rather to emphasize the more continuous dialectic of orisha cult clustering in Bahia and beyond. The Yoruba ethnohistorical record reveals that orisha cults were neither singular nor discrete in the first place but performed a critical role in reconfiguring local political relations by consolidating distinct deities within town shrines. That such similar transformations occurred in Bahia, recreating Yoruba kings and their spiritual subjects under radically different conditions of politico-economic subjugation, suggests that the West African baseline of Yoruba orisha worship is much closer to its New World manifestations than has generally been acknowledged. It is with such Atlantic dynamics in mind that detailed studies of local orisha cults at the historic peripheries of Yorubaland can inform the development of Yoruba religion not only in the Americas but as a growing transnational phenomenon (Beliso-De Jesús 2015; Clarke 2004).

FOUR

The Blood of Mothers

The question that I pursue in this chapter is the impact of West African constructions of womanhood and female agency on the plantation societies of the Americas. The character of this influence is complex and variable, stronger in some areas than in others according to changing economic, demographic, and sociopolitical conditions on the ground, not to mention the various slave routes, ports of embarkation, and modes of collecting slaves for the dreaded Middle Passage across the Atlantic. The criteria for assessing this impact also are complex, ranging from what historian Philip Curtin (1969) calls “the numbers game”—determining ethnic populations that were transshipped, in what gender proportions, and to which destinations—to mechanisms of cultural transmission and creolization, which for anthropologists Sydney Mintz and Richard Price ([1976] 1992) is nothing less than the birth of African American culture. To date, the emphasis on African baseline cultures has shifted from the specific tribal origins that Herskovits ([1941] 1958, 33–53) originally invoked in the 1930s and 1940s toward general processes of creolization and ethnogenesis originating in the slave ships and further developing within the new social frameworks of the plantation complex (Mintz and Price [1976] 1992). This change in focus from roots to routes also is marked by a more active understanding of African cultural influences, seen less as survivals and retentions resisting assimilation and more as cultural extensions and inventions, mediating the very processes of creolization and mobilizing resistance against the slaveholding class.1 I will not review these key positions within African diaspora and Black Atlantic studies, save only to highlight the growing focus on women and gender in plantation societies and the West African constructions of womanhood that were variably reworked in the Americas. Since the 1985 landmark publication of Deborah Gray White’s Ar’n’t I a Woman?, women have moved from periphery to center

98 / Chapter Four

in historical studies of Atlantic slavery—not merely added as part of the story but changing our understanding of slavery’s core contradictions.2 My goal is to deepen this focus on gender by developing a West African model of womanhood and agency and applying it to social relations of production and reproduction across a range of plantation societies. In so doing, I hope to get beneath the European gender ideologies that inform so much of the relevant archive in order to approach an African American perspective on the gender dynamics of slave life and community—one that drew upon idioms of fertility and witchcraft and derived from the position of women in markets.3 There are several caveats to this approach that expose my argument to a number of criticisms, not least of which concerns the status of the model that I propose and its representativeness among the slave communities in the Americas. But I offer it less as a definitive interpretation than as an extended hypothesis, a productive line of inquiry to pursue. Clearly, enslaved women were generally “exported” from Africa in lesser numbers than men (but see Beckles 1989, 7–23), from myriad locations, and found themselves in a broader array of New World dislocations as victims of labor exploitation, rape, and ruptured family relations. But it is precisely within the dialectics of social reproduction and transformation that the values of West African womanhood were abstracted and reshaped, mapping onto emerging relations of opposition and mediation, and to some extent informing them. In what ways and degrees are questions that can be pursued in particular cases. My preliminary goal is to set the stage for such research, beginning with material from Yorubaland in Nigeria, where markets are historically female domains.

The Blood of Mothers The model of West African womanhood that took effect in the Americas is associated with the blood of mothers, a highly fetishized, indeed potent substance that accounts for the “secret” of women—that which gives them the ability to conceive and give birth. As I shall argue, the cultural semantics of the blood of mothers is at once concrete and abstract, a distilled essence of womanhood with positive and negative values, as it were; positive in the capacity to mix with male sperm and create new life, negative in the menses, understood as “bad blood” ejected by the womb because it cannot create new life, and feared by men precisely because it can neutralize their most powerful medicines through physical contact (Drewal and Drewal 1983, 79). Among Yoruba-speaking peoples in southwest Nigeria, as well as in the Republic of Benin and even parts of Togo and Ghana, the blood of mothers is a key figure or trope that congeals, through its concentrated array of meanings, a

The Blood of Mothers / 99

number of sociocultural processes and domains that give it much more than biological significance. Focusing on the secrets and mysteries of childbirth, Yoruba ideas of fertility extend to the reproduction of social and political relations, managed by priestesses who invoke the orisha and, during annual festivals of “carrying water,” bring the orisha’s power into the town in ritual bottles and calabashes to revitalize the community (Apter 1992, 97–116). The critical moment of these annual orisha festivals, during Ijọ́ Ipọ nmi, or the Day of Carrying Water, occurs when the possession priestesses (ọ lọ́ ọ̀ tun) “put their water down,” delivering the charged containers of the orisha’s sacred water (omi) into the king’s palace and their associated town shrines. Should the “water” balanced on the devotee’s head fall or spill, havoc will ensue, requiring cleansing rituals to allay the witchcraft that will result. Enemies of the town are thought to use jùjú medicines to sabotage the safe passage of the priestesses by making them stumble and fall, causing the calabash to break apart or, as in one popular account that I recorded, by seizing a container with an invisible power and suspending it in midair (Apter 1992, 235, n. 11). Thus salt is sprinkled to sweeten the route and neutralize any bad medicines that might thwart the priestesses’ safe passage and delivery. Indeed, the entire collective drama is modeled on childbirth. As the priestesses exit the bush shrine for the town, balancing the revitalizing powers of the orisha’s “water” on their heads, they receive the protective blessing, backed by a sacrifice: “May you carry it safely and put it down” (E jẹ́ kẹ gbèé lọ s’ilẹ̀ ). The same blessing is given to a pregnant woman seeking the protection of an orisha, that she carry the fetus safely and deliver it without mishap. Ritual reproduction of the community, like human reproduction, rests on the secret of womanhood—the hidden blood of mothers. There is more to this basic ritual correspondence between human and social reproduction than meets the eye, as the expansive meanings of fertility throughout West Africa suggest. How does the blood of motherhood, with its positive and negative values, relate to the broader social processes that it mediates and embodies? We can fill in the picture by sketching the position of Yoruba women in three related domains: those of the family, or domestic domain; the lineage system, or politico-jural domain; and the market, that historically female sphere of economic exchange organized and controlled by women. Household and Family Like many family forms throughout West Africa, Yoruba households were (and in many areas remain) traditionally polygynous, with one man

100 / Chapter Four

marrying several wives if he was a successful farmer, craftsman, military leader, chief, or entrepreneur, each wife sharing domestic responsibilities while giving birth to as many children as possible. Each wife characteristically has had her own room within the larger compound, where she would sleep, store her trading supplies, and retreat from domestic squabbles. In principle, the wives are ranked by seniority of marriage into the husband’s compound, with the first wife, Ìyá ilé, presiding over her junior cowives, but jealousies and rivalries result in alternative patterns. Status among cowives by seniority is offset by the number of their children, how successful they are, and perceptions of favoritism from their husband. So important are the cowives’ identities as mothers that they refer to each other using teknonyms, as in Ìyá Bímpé (Mother of Bimpe) or Ìyá ìbejì (Mother of twins). A wife who remains childless is not fully incorporated into the family, because she is not contributing to its reproduction and expansion, and she might be sent back to her parents. She is to be pitied, but also resented and despised—a woman not fully realized as a mother, indeed with no proper household name. Likely seen as a victim of witchcraft, she also was destined to become something of a witch herself, taking revenge on her rivals.4 Witchcraft accusations within the domestic group were directed primarily between cowives and focused on their children. If a wife loses a child to disease, or fails to conceive, divination usually reveals a jealous cowife as the cause, one who used her witchcraft to sabotage her rival. As a matter of course, cowives may share the same cooking hearth, but they never prepare food for each other’s children to avoid accusations of poisoning. For the work of the witch is the inverse of fertility. The witch causes, death, impotence, and infertility by consuming the fetus as it develops in the womb, transforming herself into a night bird and sucking the life force from within—as in the very term àjẹ́ , a contraction of ìyá jẹ or “mother eats.” She kills the living children of her cowives by draining their blood in nocturnal feasts with her coven. A witch can even cause male impotence by stealing a man’s penis and using it to have sex with another woman. Moreover, witchcraft is not an aberration that can be eliminated but is a potentiality within all women—an immanent if latent negative appetite opposed to yet embedded within the very life-producing forces of their mysterious blood.5 Lineage Dynamics The negative, life-consuming idiom of Yoruba witchcraft assumes broader significance within the lineage, manifesting divisive processes of segmentation, fission, and optation (Apter 1993, 116–19). Patriarchal and patrilineal

The Blood of Mothers / 101

Figure 4.1. Polygyny and matri-segmentation (from Schwab 1955, 367).

in principle, the Yoruba lineage (ìdílé) consists of groups of households beneath a recognized male elder—often a chief or title holder—with collective rights and resources, including access to farming land and political titles. As strictly exogamous units, the daughters marry out into other lineages, and the wives marry in from other lineages, reproducing, as mothers, the lineages of their husbands. Indeed, the bridewealth paid to the bride’s family has nothing to do with “buying” the wife herself, as is sometimes mistakenly suggested, but establishes rights in her offspring. If a Yoruba mother divorces her husband and wishes to regain custody of her children, her family must return the original bridewealth. It is not, however, through divorce that women destabilize the lineage system, but through the very logic of lineage segmentation, which occurs between half-brothers—sons of the same father but of different mothers. Developing from the structure of polygynous households, the children of one mother (ọ mọ ìyá) differentiate themselves from each other, partly because each mother remains financially responsible for her offspring, but also because each group of full siblings inherits the father’s estate in equivalent portions, per stirpes (fig. 4.1).6 And because such sibling groups compete against each other for influence and control over lineage resources and affairs, they form intergenerational segments defined through brothers of the same father but of different mothers.7 Women as mothers thus become the nodal points through which lineage segmentation occurs, establishing latent lines of cleavage between groups of full siblings and their descendants through their very position as outsiders who reproduce the lineage itself. Moreover, such latent lines of division are activated by competition for impartible resources, most notably civil and even royal political titles that devolve patrilineally, precipitating complete fission when a segment breaks away from the lineage and establishes its

102 / Chapter Four

Figure 4.2. Lineage segmentation and fission (from Schwab 1955, 361).

autonomy through a separate ancestor. Yoruba local histories are full of accounts of brothers who fought over access to a title, or over the division of meat at a funeral, thereby breaking the lineage into two. Inevitably, they are half-brothers, children of the same father but of different mothers, and they represent entire lineage segments at their point of structural division (fig. 4.2). In addition to lineage segmentation and fission, women as mothers could further destabilize the patrilineage through complementary filiation and lineage optation.8 In exceptional cases, a man could make claims on his mother’s patrilineage to gain access to land or political title, including the kingship itself. Leading to lineage optation, when a son opted out of his own patrilineage and into that of his mother, this socially sanctioned transfer of allegiances amounted to a form of lineage cannibalism. Fanning these flames were the winds of mercantile profit. Because mothers, as traders, provided for their children, successful market women could become far wealthier and more influential than their husbands, precipitating the divisive dynamics of lineage segmentation and fission, undermining male authority in the household, and luring sons away from their patrilines with promises of maternal inheritance. Thus the blood of mothers and lineage reproduction had its polluted, negative dimension, the inverse of fertility, undermining the lineage through fission and optation. To appreciate the

The Blood of Mothers / 103

economic dimensions of such dangerous transfusions and transmutations, we can turn to the role of women in markets, where wives and mothers doubled as traders and merchants. Money and Markets Like many women throughout the interlocking network of periodic markets and their ring cycles extending from Senegambia to the eastern Cameroons, Yoruba women historically have controlled town and village markets by organizing into ẹ gbẹ́ associations, pooling credit, and selling agricultural produce—beginning with that of their husbands, who normally provided them with their initial trading capital.9 Occupying stalls at the center of kingdoms and subordinate towns, market women dominate the quintessential space of the public sphere, that center where the roads dividing chiefly jurisdictions converged, where town criers and king’s messengers made important announcements, and where the townspeople also could mobilize against local and government figures, often led by the market women themselves.10 In Yoruba markets, women—to invoke Natalie Zemon Davis (1975, 124–51)—“were on top.” In institutional terms, market women were represented by a female chief, the Ìyálòde (mother of the outside, the public) who was to some degree masculinized by her economic and political power, marked by the coral beads of chieftaincy and man’s hat that in some kingdoms entitled her to a seat on the king’s council. Moreover, because of the association of authority with seniority so characteristic of Yoruba social relations, the more powerful and prosperous market women were older, spending more time trading and less time tending to household duties after their children had grown, married, or moved away. The senior market women were thus typically postmenopausal, a status at once prestigious and respected but also feared and resented because their witchcraft was more powerful. For the blood of these female elders was no longer fertile and no longer flowed. Their barren wombs were seen to trap the menstrual blood that is both polluted and powerful, turning them into vessels of concentrated àṣ ẹ , the vital force of ritual potency and effective verbal command. An illustrative praise-name for these aged mothers is “the one with the vagina that turns upside down without pouring blood” (Drewal and Drewal 1983, 75). The profound association of female blood with motherhood and market women forms a pair of contrastive values. The good blood of procreative fertility and childbirth and its inverse, the infertile blood of menstruation and witchcraft, are present in all women in varying proportions but shift

104 / Chapter Four

toward the negative pole as women pass menopause, gain economic and political autonomy, and accumulate greater wealth through extended market activities. Wealthy market women can undermine male authority in the home, precipitating rivalry between cowives, lineage fission and optation, while increasing their spatial mobility and autonomy by attending markets in neighboring towns after their calibrated cycles of market days. In this capacity their witchcraft is accentuated. Timing of the periodic cycle of Yoruba markets, which occur every “five” days in every town, and coincide with the monthly meetings of the ẹ gbẹ́ associations, mirrors the periodicity of menstruation, as if extending women’s reproductive cycles into the economic cycles of social reproduction through exchange (See Apter 1993, 125–26, n. 2). The work of the market woman is thus necessary but dangerous. Like the mother who converts blood and sperm into new life, the ultimate blessing, the female trader converts money and commodities into new wealth through exchange, producing a surplus that benefits the community at large by adding to the vitality of the market itself. During the major town festivals of the orisha, prayers for agricultural productivity, fertile women, and prosperous markets accompany sacrifices to shrines and altars in the marketplace. An active market animates a kingdom, a weak or enervated market represents a kingdom in decline. But if a healthy market circulates value (money, commodities) throughout the social body, it also can be blocked and sabotaged by the market woman in her capacity as a witch. Successful female traders—sometimes called “Cash Madams” after the huge wads of cash that they store in their bras and wrappers as well as the conspicuous wealth they have accumulated over the years—may be publicly praised, but privately they are often suspected of witchcraft.11 Although their very bodies celebrate corporeal largesse, they are secretly maligned for blocking the flow of money and blood. Successful female traders are resented as hoarders, often accused of hiding scarce commodities to inflate prices and protect themselves against loss. They are believed to violate the rules of the ẹ gbẹ́ association by putting personal profit before collective trust. These same women also are said to be afflicted with abnormal pregnancies. The Cash Madam’s large belly is sometimes rumored to contain a blocked fetus that grows for years. It is only when such a witch is ritually detected and cleansed that her hypertrophic issue (much like Rabelais’s Gargantua) is violently expelled. What I am suggesting is that the bivalent blood of mothers in Yorubaland—and throughout West Africa where market women prevail—is socioculturally and historically associated with their contradictory roles as wives and mothers on the one hand and as traders and merchants on the other, framed in the antithetical idioms of fertility and witchcraft. As wives

The Blood of Mothers / 105

and mothers, they reproduce their husband’s patrilineage, but in this very capacity they compete with their cowives, generating division and witchcraft accusations within the domestic domain, as well as segmentation, fission, and even optation within the lineage system at large. As market women, they sell their husband’s produce for gain, but the profits they accumulate can undermine male authority, individually within the household through economic influence and the prospects of inheritance, as well as collectively in the marketplace, where ẹ gbẹ́ associations consolidate their corporate influence and power. It is surely no coincidence that the same term “ẹ gbẹ́ ” refers to the coven of witches that meets in the market, at the base of the Iroko tree, during nocturnal feasts that drain the blood of their victims—the children of cowives, or their fetuses within. Nor is it surprising that their witchcraft power itself devolves matrilineally from mothers to daughters, an essence that can be allayed but never eliminated from female blood and its generational lines. Moreover, as vessels of transmission, women double as agents of conversion—if their wombs combine male “water” with female blood to create new life for the patrilineage, their trays and calabashes convert commodities into money. Measuring exchange values that circulate throughout the social body, they convert money into “blood” as the flow of socially reproductive value (see also Matory 1986). In their fertile capacities, women as mothers and traders control human and social reproduction. As witches, they subvert fertility by sabotaging, even cannibalizing, procreative vitality and by blocking the flow of money and commodities—hoarding, hiding, and accumulating profit by removing it from circulation. The purest expression of illicit wealth is found in Yoruba idioms of moneymaking magic, which converts the blood of kidnapped children into brand new naira bills. Such “soaked” or “hot” money is typically unproductive and infertile: gained without labor, it is spent without gain.12 A market woman will guard against such nefarious transmutations of value precisely because they represent the negative limits of her witchcraft potentialities—the conversion of vital force into bad blood, of exchange value into unproductive surplus. If actual witches are feared and despised for their cannibalistic appetites and destructive agency, as potential witches Yoruba women are honored and respected, for within their blood lies the secret of their power and value.13

New World Transformations What does this exploratory glimpse into the blood of mothers in Yorubaland have to do with enslaved women in the Americas? By what possible stretch of the diasporic imagination can we relate Yoruba constructions of

106 / Chapter Four

motherhood and womanhood to the radically different historical conditions of plantation societies? First, there were general demographic trends that support such an attenuated cultural connection. From the 1650s to 1750s—as the Atlantic plantation complex coalesced—concentrations of African captives from the Bights of Benin and Biafra represented a five-hundred-mile belt of culturally and economically linked societies surrounding a Yoruba core where households were polygynous, descent was patrilineal, and women dominated a network of periodic markets. Moreover, an uncharacteristically high proportion of these captives (about 40 percent) were women, contrasting with higher male ratios of the Upper Guinea and West Central African trade, and with prevailing higher numbers of male captives during the following one hundred years.14 To be sure, Senegalese and Congolese forced emigrants also were important during this high period of the Atlantic trade, as the wide distribution of “Foulah,” “Mandingo,” and “Congo” ethnonyms in the Americas implies, but their cultural configurations of gender and procreation appear less explicit, sustained by a lower ratio of females and—as with the Afro-Caribbean religions of Vodou, Santería, and Candomblé—largely assimilated to Fon-Yoruba models instead. Second, transcending the question of Yoruba influence as such is the broader significance of those underlying regional models or “grammars” that, following Mintz and Price ([1976] 1992), emerged as salient categories of interaction and communication among captives from various ethnicities in Africa who were thrown together in the holds of the ships and regrouped in the Americas. Crucial to this approach to creolization are two fundamental axes of refabrication in the making of African American culture: the horizontal axis between enslaved Africans whose ethnic differences were unconsciously bridged by underlying values and cognitive orientations; and the vertical domination of Europeans over Africans, framing the complex dialectics of resistance, accommodation, and codetermination that their encounter entailed.15 It is thus within such a sociohistorical matrix that the blood of mothers was remapped, extending beyond the confines of a single ethnic culture to embrace the broader underlying cognitive orientations that it shared among other West Africans, as well as the European ideologies of race, gender, blood, and procreation, that it both resisted and accommodated. Let us begin with this latter set of correspondences between African and European conceptions of black womanhood and sexuality, as it relates to African women. We have seen from the Yoruba model of womanhood and its complementary principles of fertility and witchcraft that female power and agency is both reproductive and transformative, giving birth and taking

The Blood of Mothers / 107

life according to distinctive qualities and dispositions of blood; good blood (procreative) versus bad (menses), circulating versus blocked, generative versus extractive or consumptive. These positive and negative valuations of womanhood, informing regional West African idioms more broadly, resonated with similar contrastive sets deployed within European discourses of primitive barbarity and illicit desire. In her rich exploration of “pornotropical” travel writing from the late sixteenth through the eighteenth centuries, Jennifer Morgan (2004, 12–49) identifies a dominant icon of African womanhood and fertility—the image, often portrayed in engravings—of African mothers with pendulous breasts, emphasizing a hypertrophic physiology of lactation and labor in what became a naturalized charter of female enslavement. In one widely circulating elaboration of this theme, African women working the fields simultaneously suckled children on their backs by throwing their elongated breasts over their shoulders, embodying the conflicting demands of production and reproduction in figurations of the sexual grotesque. In his popular Description (1732) of the Guinea Coast, the slave trader and explorer John Barbot thus wrote of “the poor babes, so carr’d about at their mothers’ backs . . . and how freely they suck the breasts, that are always full of milk, over their mothers’ shoulders, and sleep soundly in that odd posture.”16 Opposing this image of mammarian motherhood was a complementary discourse of “lust and depravity” (Brantlinger 1986, 215) in which “hot constitutioned Ladies” with firm and shapely breasts tested European reserve and resolve (fig. 4.3).17 Whether abhorred as nymphomaniacs or admired as seductive beauties, these figures of African womanhood emphasized desire and sexuality over fertility and reproduction, many portrayed without children as temptresses and concubines. William Smith, who mapped the Gold Coast in 1727 for the Royal African Company, quotes the resident British factor Charles Wheeler, who received a concubine from a local king: Her lovely Breasts, whose Softness to the Touch nothing can exceed, were quite bare, and so was her Body to her Waste [sic] . . . and though she was black, that was amply recompenc’d by the Softness of her Skin, the beautiful Proportion and exact Symmetry of each Part of her Body, and the natural, pleasant and inartificial Method of her Behaviour. She was not forward, nor yet coy, when I pressed her lovely Breasts, she gently stroak’d my Hand and smiling met my Salute with equal Ardour and Fervancy.18

In addition to the community of Afro-European “Creoles” that developed around the slave forts and “castles,” and the “marriages” of convenience

108 / Chapter Four

Figure 4.3. Henry Stanley resisting temptation (from Bruel 1890, 207; reprinted in Brantlinger 1986, 215).

between slave factors and local chiefs, this passage reflects a contrastive register to the “over-the-shoulder breast-feeding mother” (Jennifer Morgan 2004, 49), one that valorizes firm breasts and perfect bodies over the “uberous dugges” of the nursing mother-laborer. Within this bivalent construction of African womanhood, reproduction (fertility) and sexuality (witchcraft) were structurally opposed. The socioeconomic and political meanings of reproduction and transformation throughout the plantation societies of the Americas were framed within racialized relations of production and exchange—of boundaries systematically upheld and transgressed—that will be examined in due course.

The Blood of Mothers / 109

Figure 4.4. Domestic sainthood and social lactation. Left: Hattie McDaniel as “Mammy” in Gone with the Wind (from www.nndb.com/people/077/000063885/). Right: African wet nurse in Brazil (from Freyre, Pinto, and Rodrigues 2005, 35).

At this point we can simply note how the antimonies of black womanhood were stereotyped into familiar figures of loyalty and lust. Nowhere is the opposition more clearly expressed than in antebellum contrasts between Mammy and Jezebel in the American South.19 Evoking the image of the black mammary mother, Mammy represented the loyal domestic slave who served as the lynchpin of the Big House, running the household, attending to her mistress, and manifesting in her bigness and blackness a surrogated form of social lactation. Indeed, mammy as mammary was, in effect, a “surrogate mistress and mother” (D. White [1985] 1999, 49), so devoted to her young white charges that she could double as their wet nurse (fig. 4.4).20 Against this image of black maternalism, the figure of Jezebel posed a sexualized threat, attending not to her mistress’s children but rather to her husband’s needs. Such “Negro wenches” were sold to work as courtesans and prostitutes in the so-called “Fancy Trade,” but within the confines of the plantation and its domestic interiors, they were reputed to exchange sexual favors with the master and his sons for gifts and special treatment. Although in reality they were ravished and raped, such women were constructed as agents of seduction, evoking the image of the African temptress and her natural promiscuity.21 As Deborah Gray White ([1985] 1999, 46) portrays the mythic contrast: On the one hand there was the woman obsessed with matter of the flesh, on the other was the asexual woman. One was carnal, the other maternal. One was at heart a slut, the other was deeply religious. One was Jezebel, the other a Mammy.

110 / Chapter Four

To appreciate the socioeconomic dynamics of reproduction (fertility) and transformation (witchcraft) that this pervasive contrast embodied, and the blood-based idioms with which they resonated, we can turn to the dominant social domains where enslaved women occupied central roles. Household and Family The scholarship on slave families as they developed in the New World reveals a broad range of trends and types correlated with the plantation systems in which they were embedded. I do not have space in this more analytical exercise to adequately navigate this complex terrain, but I will highlight key tensions within family forms that correspond to the values of fertility and witchcraft in Yoruba domestic groups. First, it should be evident that the normative models of West African family organization were severely undermined in the Americas by conditions imposed by the slaveholding class: not just the destruction of families through exploitation and sale, but also through direct interventions into domestic life—arranging slave unions and marriages, lending or renting out enslaved workers to other plantations, drawing some into the Big House or sending them back into the fields, the violation of enslaved girls and women, not to mention the decimation of aging slaves through overwork, punishment, malnourishment, and neglect. Nor was the plantation itself a stable environment—it responded to changing fortunes and market conditions as owners sold enslaved men, women, and children to cover debts and losses, and it was subject to its own dynamics of generational succession when children inherited and divided estates. Nonetheless, African American families did manage to develop and precariously endure, bolstered by the development of provision grounds, serving as havens of protection and survival, and sometimes producing extended compounds connected by polygynous unions. More than any other social domain, the slaves’ households served as repositories of African values for women, not in simplistic terms of passive preservation, but as a selfconscious space of opposition to the Big House and its European codes of domesticity.22 If the Yoruba-based model of household witchcraft derived from tensions generated by polygyny, leading—as we saw—to the divisive dynamics of lineage segmentation, fission, and optation, New World families among the enslaved Africans were rarely granted the freedom to expand along these lines, but displaced and diverted the blood of mothers both inward and outward. To be sure, multigenerational families, polygynous tendencies, and principles approaching patrilineal succession did develop in variable

The Blood of Mothers / 111

Figure 4.5. Nineteenth-century Yoruba compound (from The Gleaner Pictorial Album [London, 1888], 1:13; reprinted in McIntosh 2009, 83).

degrees. Stevenson (1996a, 177) observes that “Virginia’s slaves were part of a variety of marriage, family, and household types, nuclear and extended family structures; monogamous, polygamous, and serial marriages; single and multiple generational households of various combinations of kin, friends, and sometimes strangers,” with variations corresponding to small and large holdings. Bush (1990, 107), drawing on Higman (1975, 1984), notes that in Jamaica, Trinidad, and Barbados, extended or polygynous households were often based on a series of contiguous units built around a yard and enclosed by a fence. The individual units were occupied by mothers and children in the case of polygynous households and by nuclear units in the case of extended family households, an arrangement which is common in West Africa societies.

This passage could apply to a Yoruba compound, with its separate rooms and units for each cowife and her children arranged around the husband (or other senior males) within a single walled domicile (fig. 4.5). And notwithstanding the imposition of matrilineal descent through official European legal codes, according to which a slave’s status devolved through its mother, patrilineal trends asserted themselves when collective property was at stake.23 In eighteenth century Jamaica, “rights to provision grounds and houses were passed on by slave fathers to their families,” with priority accorded to the eldest son (Bush 1990, 93).24 But despite this evidence of incipient patrilineal

112 / Chapter Four

group incorporation, slave households tended toward matrifocal and nucleated forms, with shallow generations, conjugal pairs, or “single” mothers with “away” husbands living on distant properties or plantations (Morrissey 1989, 85–96). It is therefore not surprising that D. White ([1985] 1999, 135) highlights witchcraft conflicts between bonded women competing over men throughout the antebellum South, not as institutionalized cowives as such, but as proto-polygynous formations. With family fission resulting from forced sales and relocations, however, the blood-based antimonies of witchcraft and fertility were concentrated onto the mother-child bond itself, wherein procreation and protection vied with abortion and infanticide. Motherhood for enslaved women was ambivalent and fraught, embodying the contradictions of production and reproduction—between labor units and breeding units (Bush 1990, 133, citing Gautier 1983)—within the womb’s generative powers. Consistently low fertility rates among enslaved women in the Americas have been attributed to a variety of socioeconomic factors, ranging from malnutrition, unsanitary conditions, traumatic stress, and the physiology of exhausted bodies to systematic abortion and infanticide that amounted to nothing less than a gynecological revolt (Jennifer Morgan 2004, 11; Patterson 1982a, 261) against slavery itself. Planters were explicitly preoccupied with low fertility and high child mortality, favoring “breeding wenches” over barren women through widely distributed rewards and sanctions. Moitt (2001, 94–95) notes that in Jamaica and Barbados as well as in the French Antilles, female slaves were given money and cloth for each successful delivery, with work exemptions accruing for every additional child. Bush (1996, 199) reports that Jamaican rewards included larger rooms and houses for slave mothers and infants. Deborah Gray White ([1985] 1999, 100) describes how, in Georgia and South Carolina, pregnant slaves and nursing women were given lighter workloads on “trash gangs,” receiving additional clothing and an extra week of rations for each birth. But if successful childbirth brought material rewards, failed reproduction exacted its costs, including forced sale, whippings, and even the iron collar (Moitt 2001, 95). Planters blamed pregnant slaves and midwives for miscarriages, stillbirths, and infant deaths, either indirectly from the bad blood of their imputed promiscuity, or directly, as willful killers. The Jamaican plantocrat Edward Long believed that “most black women [were] subject to obstructions of the menstrua” (monthly periods), leading to “incurable” sterility (Bush 1996, 198). Indeed, “menstrual maladies” were a common affliction in the plantation south, including “amenorrhea (lack of menstrual flow), abnormal bleeding between cycles (sometimes caused by benign or malignant tumors), and abnormal discharges (resulting from such conditions as

The Blood of Mothers / 113

gonorrhea, tumors, and prolapsed uterus),” in addition to complications with pregnancy such as “convulsions, retention of the placenta, ectopic pregnancy, breech presentation, premature labor, and uterine rigidity” (D. White [1985] 1999, 83). Infant disease such as neonatal tetanus (tetanus nascentium) also was widespread, known as lockjaw, “lowjaw” (Bush 1990, 143– 46), and “mal à mâchoir” (Moitt 2001, 97) because in paralyzing the infant’s larynx it prevented the newborn from feeding at the breast. This blocked flow of mother’s milk resonates with blood-draining idioms of West African witchcraft, particularly because the disease was attributed to the midwives’ mistreatment of the umbilicus, that irreducible channel between mother and fetus that must be properly severed for a safe delivery. But if bad blood and low fertility were connected morally and medically in the minds of the planters, it was widespread evidence of secret abortions that most directly implicated mothers and midwives as fetus-destroying witches. Planters were convinced, not without reason, that enslaved women aborted with the assistance of midwives, whose roles as herbalists and even hospitalières provided access to a range of abortifacients. Tennessee physician John H. Morgan identified “the herbs of tansy and rue, the roots and seed of the cotton plant, pennyroyal, cedar berries, and camphor, either in gum or spirits” (D. White [1985] 1999, 85), whereas one Dr. Thomas Dancer in Barbados recorded “Cerasee, Barbados Pride, Wild Passion Flower, Water Germander and Wild Tansey” as indigenous plants inducing labor in women (Bush 1990, 141).25 Although distributions of abortion by African ethnicity are impossible to determine with any precision, it is significant that in colonial Saint Domingue, with its historically “Congolese” or Central African imports, it was “les femmes Aradas” from the Dahomean slave trade who were particularly associated with abortion and were punished with iron collars like recidivist maroons (Debien 1974, 365).26 Moitt (2001) describes similar associative afflictions on Flauriau Plantation in Saint Domingue, where one enslaved midwife named Arada—as if personifying her “nation”—was forced to wear a rope collar with seventy knots, “each knot representing a child she had allegedly killed” (63). There is no question that many innocent midwives were falsely accused of abortion and infanticide, reflecting the fears and anxieties of planters unable to reproduce their workforce from within; but there also is ample evidence that enslaved women conspired to subvert their reproductive potential, thereby empowering themselves as witches against the unyielding demands of the master class. Two disparate images suggest how abortion and infanticide assumed witchcraft proportions for whites and blacks alike. In one, Deborah Gray ([1985] 1999, 126) quotes an enslaved Christian convert

114 / Chapter Four

seeking redemption in the plantation south: “I was carried to the gates of hell and the devil pulled out a book showing me the things which I had committed and that they were all true. My life as a midwife was shown to me and I have certainly felt sorry for all the things I did, after I was converted.” In the other, Debien (1974, 365) cites Moreau de Saint-Méry’s observation that by the end of the eighteenth century in Saint Domingue, punishment for abortion shifted from the iron collar to a form of social ostracism, in which “on lui imposait un billot de bois sur le dos pour la punir et l’humilier par ce simulacra d’un enfant.”27 If the first image doubles as a kind of witchcraft confession in which the guilt-ridden midwife seeks divine forgiveness and cleansing, the second unwittingly reveals a Yoruba response to children lost in childbirth, when bereft mothers carry ìbejì dolls to placate the dead child’s jealousy of the living and prevent any mystical retaliation. The “sticks of wood” that Moreau de Saint-Méry saw as signs of public humiliation may well have doubled as ritual paraphernalia to protect slave mothers and complicit midwives from the avenging spirits of the aborted and killed.28 Plantation Dynamics We have seen how the domestic dynamics of witchcraft were channeled “inward” within slave families, focusing on the mother-child bond as a form of “gynecological revolt.” But if the slave family was fractured and truncated through the forced management, sale, and separation of kin, its witchcraft was not so easily contained, but also was channeled outward, mapping onto the broader divisions within the plantation system at large. Methodologically, we can replace the Yoruba lineage with the patriarchal plantation as the dominant New World corporate group, identifying gendered patterns of segmentation and fission within racialized idioms of production and stratification. Again, we must emphasize that the schematic simplification of complex variations into a number of key patterns and contradictions is oriented only toward systemic trends underlying a broader range of plantation societies and slaveholding estates. Let us return to the dynamics of witchcraft between Yoruba cowives within a polygynous household. Recall how patterns of lineage segmentation took root between brothers of different mothers vying for resources and political influence, splitting over time into distinct branches and separating entirely through lineage fission. In each stage of this process, women occupied the nodal points of differentiation, generating womb-linked matri-segments within the patrilineage at large. Although such differentiating trends were cut short within the coercive confines of slave households

The Blood of Mothers / 115

and families, they emerged within the Big House, where a kind of de facto polygyny developed between the master and his servant-concubines.29 The Big House was by no means the exclusive site of interracial sexual relations, because white indentured workers also engaged in “servant breeding,” much to the management’s disapproval (Beckles 1989, 94, 132). But in the Big House, white male access to African and African American women remained an implicit seigneurial right and something of an open secret in matters of “colored” progeny. From a Yoruba perspective, we might reconsider the planter’s household as polygynous, the white wife as first wife (ìyálé)—if there was one—ruling over her African American cowives (ìyàwó). Even if empirically this structure was more immanent than manifest, it gave the blood of enslaved mothers a broader range of subversive potentials with respect to domestic reproduction and the devolution of the estate. Within the more intimate confines of the master’s house, the loyal “mammy” who upheld boundaries met the coquettish “Jezebel” who crossed them, sleeping with the master—whether forced or by design—and ultimately bearing children.30 The question of whose children, with what rights, was of course crucial. Although some slaveholders disavowed their mixed-race children, treating them like any other slave, others accorded them preferential treatment, or even legal recognition. Bush (1990, 155) writes of eighteenthcentury Jamaica that “wealthier white men often bequeathed their estates to black or coloured ‘wives’ and their offspring or made provisions for their manumission in their wills. Despite the harsh, unequal and callous nature of slave society, close, loving bonds between black and white did exist and the degree of money and property left by whites to mulattoes was a cause for concern and controversy in plantocratic circles.” Control over interracial marriage and inheritance would be imposed by the courts as the caste and class structures of plantation societies consolidated within the broader Atlantic system, reflecting rising fears of “degeneration” and the loss of white supremacy. But even the most draconian laws were never uniformly enforced. Black lovers and concubines threatened to subvert the system from within, mingling the bloodlines of masters and slaves into lighter progeny with claims on the estate and possible emancipation. Thus the blood of slave mothers was not only lightened, or “cleansed” in the language of racial caste, but was converted into money and honor through pathways of upward mobility.31 If generational succession and devolution of the estate could follow interracial lines within the Big House, it precipitated a very different pattern of family fission among enslaved workers’ households. When the master died and his widow and children inherited the plantation, slaveholding

116 / Chapter Four

properties were regularly divided and sold, often separating slave families in the process. Here the “blood” of mothers was structurally salient because some owners sold husbands and fathers before rupturing the bonds between mothers and children, recapitulating a pattern from the slave markets— where mothers and children were more often sold together—that reinforced matrifocal family patterns. Reasons for favoring the mother-child bond were hardly altruistic but followed from observations that mothers who lost their children through sale became demoralized, listless, and less productive as workers. They also were more prone to flight or marronage. And within the maternal sibling unit, females were favored over males, possibly as better caretakers for their aging parents. Analyzing the dispersal of the Gaillard estate in South Carolina, historian Cheryll Ann Cody (1982, 207–8) found that “daughters were twice as likely as sons to stay with their mother,” adding that “because so many women with children were entrusted with the care of elderly parents, the resultant pattern reinforced the maternal kinship ties and tended to produce three-generational matrifocal families.” Such trends toward matrifocality, privileging the mother-daughter bond at its core, echo not only the logic of matri-segmentation within the Yoruba patrilineage, but also the transmission of witchcraft from mothers to daughters in submerged maternal lines. As Cody (1982, 210) argues, slave naming practices favored patrikin precisely because paternal and fraternal ties were most vulnerable to separation. Episodic divisions associated with the dispersal of an estate had a matrifocal impact on slave family structure, but occurred intermittently, with variable outcomes depending on the planter’s longevity, the number of recognized children, the size of plantation, and the profitability of his crop. Daily contradictions, however, were built into women’s competing roles as laborers and mothers, so perversely represented by Matthew Lewis’s complaint that overseers and bookkeepers “kick black women in the belly from one end of Jamaica to another” (cited in Bush 1990, 45). If planters encouraged slave fertility to replenish the labor force, those responsible for labor productivity tried to minimize the costs of pregnancy and childrearing. Clearly the imperatives of production and reproduction were at odds, pitting the demands of the field against those of the house. Pregnant women seeking special dispensations could be flogged; most worked long hours in gangs “where they performed the same field labour as men . . . from sun-up till sun-down” (Beckles 1989, 12). In this capacity enslaved women were effectively defeminized, converting reproductive value into labor power and thereby minimizing their caring capacities as mothers in the domestic domain. Whereas the Yoruba mother was masculinized in the market, seeking opportunities and

The Blood of Mothers / 117

accumulating profits at the expense of her domestic obligations, the enslaved mother was masculinized in the field, where even the flow of her breast milk was blocked. In one of the most profound manifestations of the split between labor units and breeding units, many plantations had “weaning houses” where specialized “sucklers” would wet-nurse newborns while their mothers toiled in the fields. Small wonder that overworked and exploited enslaved women resorted to abortion and infanticide; in West African terms, these women were forced into a witchlike mode of plantation production that converted mother’s milk into the master’s monetary gain. It was only in markets that such extreme exploitation of laboring women was partially redressed. Huckstering and Markets The development of slave provision grounds and an internal market system for cultivated goods is a well-known chapter in the history of New World plantation societies beginning in the seventeenth century (Morrissey 1989, 50–61). It charts significant gains for enslaved women who took charge of an emergent sphere of circulation and exchange. According to the standard narrative, planters introduced provision grounds to offset the costs of feeding enslaved workers—and thus of maintaining the labor force by allowing them to augment minimal rations with staples like root crops, corn, and plantain, as well as vegetables and domestic animals raised on garden plots. The grounds could be worked collectively on larger plots located miles from their homes, as in the Jamaican polink, or individually, as in the “small house spots” (Mintz and Hall [1960] 1974, 191) of Antigua and Barbados.32 In both cases, provision grounds produced enough agricultural surplus to generate barter and monetary transactions in markets that were typically held on Sundays. Over time these markets developed into differentiated systems of internal commodity circulation in which Africans and “coloreds” sold their goods directly, retailed as “middlemen,” and traded with free blacks and whites of all strata. They also were characteristically dominated by “aggressive” women who drew upon a West African heritage of marketeering to outmaneuver the white competition. According to Bush (1990, 49), enslaved women “carried these skills with them to the Caribbean and, in addition to participating fully in the cultivation of provision grounds, they became prominent as market sellers and ‘higglers’ or commercial intermediaries who sold the crop surplus of other slaves for a small profit.” Beckles (1989, 73) similarly argues that in opposing British legislation that criminalized their market activities, “slaves fought for the right to be legitimate, autonomous

118 / Chapter Four

economic agents, as this was the only way to preserve aspects of the commercial heritage they had brought to the New World,” adding that “it was slave women, African-born and Creole, who, from the beginning, dominated numerically the huckstering business in Barbados.” Nor was this trend limited to the British West Indies, but extended throughout the French Antilles and in areas of the US South, where “the market scene was dominated by women” (Moitt 2001, 56). There is no question whether enslaved workers—particularly women— were empowered by these more independent forms of production and exchange (Mintz 1971, 248). On a basic level, they reappropriated their labor power to work and sell for themselves, accumulating money and profits that could lead to manumission, as evidenced by free alongside enslaved women in the markets. But short of acquiring freedom through the courts, they pushed against the limits of bondage by exercising greater agency through spatial mobility and economic bargaining. Like Yoruba women pulled further afield through interlocking ring markets in West Africa, enslaved hucksters enjoyed a freedom of movement at odds with their restrictive plantation lives. Where such Yoruba mobility challenged male authority in the management of household and lineage affairs, slave women participating in Sunday markets threatened the patriarchal authority of the master class (Olwell 1996). On a concrete level, female hucksters were physically liberated during their weekly sojourns to the market, escaping the confines of the plantation to enter a broader public sphere. In the French Antilles, where movement was more regulated, marketers were required to carry passes or “tickets” signed by their owners, giving rise to a black market in false passes that destabilized white control. In the British West Indies, enslaved workers traveling to market were exempted from pass requirements otherwise in place, indicating the recognized importance of internal marketing at large. But if planters valued this weekly marketing as an external subsidy, they also suspected higglers and hucksters of stealing goods from the plantation stores and converting them into trading capital, thereby gaining socially and financially at the owner’s expense. In Barbados, the legislature criminalized the slaves’ market trade as theft, first in the 1688 Slave Code, again in 1708, and in 1733 through the draconian “Act for the Better Governing of Negroes, and the More Effectual Preventing of the Inhabitants of this Island, from Employing their Negroes or Other Slaves in Selling or Bartering” (Beckles 1989, 76). Even with such laws on the books, however, the slave huckster’s marketeering prevailed, harnessing a movement and momentum that continued to grow with the rise of rural

The Blood of Mothers / 119

peasantries after abolition (Mintz 1955). If the Yoruba market woman in West Africa profited by “eating” her children and family, New World hucksters “ate” the plantation, diverting its productive surplus into mercantile pathways of commodity exchange. The spatial mobility of enslaved traders defined a moral topography of the social landscape, correlating distance from the plantation “center” with immorality at the margins. It is no coincidence that in Georgia’s Low Country, the slave women who dominated the informal markets were accused not just of hoarding goods and inflating prices, but also of selling their bodies. Savannah’s “Cake Wenches” were singled out by whites as “African harpies” whose sexual behavior would “vitiate the morals and dilapidate the constitutions of young [white] men,” thereby “posing a particularly invidious threat to the social and moral fabric . . . of the white community” (Wood 1990, 321).33 That the author of this particular diatribe signed himself “antimulatto ” in capital letters graphically highlights the fears of interracial sexual relations generated by successful market women, establishing an implicit homology between immoral economic and sexual exchange.34 The demonic figure of the harpy, part woman and part bird, not only recalls the image of the Yoruba witch-bird, but also captures the hybrid “monstrosity” of racial and legal boundary crossing. This latter confounding of legal categories went beyond the material opportunities that huckstering provided for augmenting agency and buying freedom; it was rooted in the very dialectics of commodity exchange. In his study of markets in antebellum South Carolina, appropriately titled “Money Knows No Master,” McDonnell (1988, 34) explores the phenomenological implications of transacting value in market contexts, highlighting the principle of commensuration governing exchange relations between subjects and objects: “Exchange, in Aristotle’s words, ‘treats all parties as equals’; as Marx put it, the ‘social relationship between the two owners is that of mutual alienation . . . each exists as his own surrogate (equivalent) and as the surrogate of the other.’” Thus, through the exchange of money and commodities, African Americans and whites became socially equivalent, not in a fully fledged legal sense, but phenomenologically, through intersubjective recognition and realignment. “In the marketplace,” McDonnell writes, “not only were blacks raised, but whites were lowered” (35), a leveling that manifested the spatiotemporal coordinates of the market system itself. Thus, according to Charles Ball, the enslaved “became a kind of freeman on Sunday all over the southern country” (quoted in McDonnell 1988, 36). It is therefore no wonder that whites were threatened by the growing influence of

120 / Chapter Four

enslaved hucksters and hawkers, for their presence “as equals” confounded the categorical distinctions that upheld plantation slavery. Like their Yoruba counterparts, New World market women mixed money and blood, converting the general equivalent of commodity value into the social equivalence of generalized exchange. In the white backlash that their success inspired, they were portrayed as “serpents gnawing at the vitals of plantation society” (McDonnell 1988, 36), recapitulating with striking symmetry the blooddraining profits of the Yoruba witch.

Throughout this somewhat experimental chapter I have explored New World plantation societies through the lens of Yoruba gender, focusing on the blood of mothers and its bivalent antinomies of witchcraft and fertility. I have justified the relevance of this ethnomodel to a wide range of cases in the Americas not because the Yoruba as such predominated among African captives, but with reference to Mintz and Price’s generative approach to the historical dynamics of cultural abstraction. I have used Yoruba idioms of female power as something of an ideal type—a set of symbolic relations, transpositions, and conversions—that extends beyond Yorubaland to designate a broader regional “grammar” in West Africa, those “deep-level cultural principles” (Mintz and Price [1976] 1992, 14) underlying a transethnic zone from Ouidah to Calabar and their associated hinterlands. The salience of this model as a common denominator among captive Africans exported from this region emerged through the dialectics of enslavement and encounter precisely because structures of reproduction and transformation governed the core dynamics of creolization.35 Within this regional grammar—what I would recast as a regenerative scheme—the blood of mothers mediated basic contradictions within the households, plantations, and emerging markets of the Americas: within slave families, the “witchcraft” of low fertility, infanticide, and domestic fission; within the Big House, the “witchcraft” of concubinage, interracial sexual contacts, and the appropriation of the planter’s blood-based patrimony; throughout the plantation, the conflicting demands of production and reproduction; and in the internal marketing system, the progressive empowerment of enslaved women as higglers and hucksters. It is important to emphasize that these blood-based idioms—and the women who embodied them—did not just reflect the core transformations of plantation societies but took an active role in shaping them, an activism and agency that had witchcraft overtones for Africans and Europeans alike: whether in overt accusations of sorcery, theft, and infanticide; more muted conceptions

The Blood of Mothers / 121

of motherhood and promiscuity; or more general discourses of race and sexuality framing encounters from “above.” Against the racial ideologies and gender discourses of the master class, the blood of mothers provided a critical counterpoint, a space of body politics and black female agency that is gaining new attention in Atlantic history.

FIVE

Ethnogenesis from Within

Essence is expressed by grammar. —Ludwig Wittgenstein

It is now an anthropological truism that ethnic identity is “other”-oriented, such that who we are rests on who we are not. If there is anything primordial about ethnicity, it is not the blood-based body of affective affiliations that plagued an earlier generation of modernization theorists, but rather, as John Comaroff (2010, 531), revisiting Barth (1969), maintains, “the act of drawing boundaries among populations.” The contexts in which such demarcations occur are of course crucial to their social and historical significance, whether they are motivated by politics, resource competition, class formation, marketing, immigration, religious encounter, or the apocalyptic violence of state-directed genocide. Such contexts also illuminate the cultural “stuff”—such as language, ritual, kinship, or costume—that is selected and produced as the content of ethnicity. But whatever its significance and cultural material, ethnicity rests on prior conditions of differentiation and othering, forming “a dialectic of identification and contrast” (Comaroff 2010, 531) that accounts for its fluid and dynamic characteristics. In its widespread Hegelian variations, this dialectic surfaces in self-designating ethnonyms originally generated by ethnic outsiders, as external coinages. The development of Yoruba identity in the late nineteenth century falls well within this view from without. According to the standard narrative, “Yoruba” was a Hausa or Fulani term designating Old Oyo, later extended to its southern vassals and neighbors who were otherwise organized into subcultural groups such as Egba, Egbado, Ijesha, Ijebu, Ondo, Ekiti, and

Ethnogenesis from Within / 123

Akoko but lacked any overarching identity as such. Law (1997, 206, 216, n. 11) ventures the earliest appearance of the ethnonym in a West African Arabic language source from 1615 (Ahmed Baba’s treatise on enslavable infidels), and cites Bowdich (1819, 208–9) to document the use of the term among Muslim residents of Kumasi in 1817.1 In his monumental History of the Yorubas, Samuel Johnson (1921, 5) quotes a “copious extract” from Denham and Clapperton (1826, app. 12, p. 22), who translated an Arabic manuscript by Sultan Mohammed Bello of Sokoto that describes “Yarba” as “an extensive province” whose people bought slaves from the north and “resold them to the Christians” who transshipped them from the coast. The appropriation of “Yoruba” as a conscious term of self-identification also occurred from without, among the Aku community of liberated slaves in Freetown, where the Church Missionary Society with protégé Samuel Ajayi Crowther standardized Yoruba language and orthography for what became the elite consolidation of a Pan-Yoruba identity (Ajayi 1960). The consolidation occurred slowly and unevenly, beginning with Oyo as “Yoruba proper” and gradually expanding into a wider regional identity through the ideological framework of a Christian nation (Peel 1989, 2000). More recently, Matory (1999, 2005) has judiciously demonstrated that a crucial contribution to modern Yoruba identity came from late nineteenth-century African Brazilian travelers to Lagos, whose manifold impact on the Lagosian Renaissance left a lasting legacy of cultural nationalism.2 It is not my intention to challenge the specifics of these externalist perspectives, but rather to complicate the very distinction between inside and outside that they presuppose.3 If any prior demarcation of boundaries involves “a dialectic of identification and contrast,” then how are these identifications made and contrasts drawn? By what cultural logic and semantic principles are objects, entities, particulars, and collectivities constituted in the first place, in a sense “from within”? What I am suggesting in this exploratory chapter is, first, that the primitive function of Yoruba negation is not in fact unconditioned, but is already embedded within a cultural semantics of quantification; and second, that such demarcating and objectifying modalities played an important if largely overlooked role in the mediation and incorporation of those “external” factors and diasporic influences that eventually gave rise to Yoruba ethnicity. My aim is to both complement and destabilize externalist perspectives by turning Yoruba lineage theory on its head—relating descent, residence, kinship, and kingship in Nigeria to their reconstituted ritual frameworks in Cuba and Brazil. Following Barber (1991) on praise-poetry and Verran (2001) on Yoruba quantification,

124 / Chapter Five

I reexamine the semantics of the category ilé in the emergence of Lucumí and Nagô houses in order to provide an internal counterperspective on Yoruba ethnogenesis from within. If such an exercise helps explain the rise of Yoruba ethnicity, it does so circuitously, by focusing on two specific trajectories in Cuba and Brazil that shaped the transnational field of Yoruba social and religious capital—an important episode in the broader consolidation of pan-Yoruba identity in Nigeria (Matory 2005). Whereas the first trajectory charts the emergence of ritual lineages out of Catholic brotherhoods, the second relates colonial ideologies of racial stratification to growing concerns with ritual purity. My reanalysis of these trajectories through a Yoruba lens suggests historical continuities between racial whitening and ritual purification in Santería and Candomblé.

The Home in the Town The basic Yoruba concepts of “home” (ilé) and “town” (ìlú) are marked by considerable socio-semantic ambiguity, embracing diasporic idioms of “homeland,” ritually reconstituted centers, kingdoms, and quarters, as well as houses in which people live. “Home” or “house” can shift between narrow conceptions of a localized residential family group, a collection or compound of such groups (agbo ilé, lit., “flock” of houses), or in a more genealogical idiom, between a core of agnatic lineal descendants (ọ mọ baba kan) and the broader range of cognates, affines, strangers, and, in the past, pawns and domestic slaves that would become attached and in some ways “absorbed” within the lineage (ìdílé) or compound (ilé). Such absorption is rarely absolute, in that “strangers” may be barred from inheriting lineagevested titles; they may commemorate distinctive orisha that they “brought” to the compound generations before; or, as Barber (1991) so cogently demonstrates, their praises (oríkì) can set them apart from the “core” by resurrecting associations with alternative origins. But even the agnatic core is open to negotiation, since sons seeking titles through complementary filiation—that is, through their mother’s connection to the title-holding patrilineage—can “opt” into their mother’s agnatic line.4 I knew a family in Ayede-Ekiti that seemed to have two cognomens—they were ilé Ilétogun (see how ilé is already nested in a prior “Ogun” line) and ilé Balógun, holders of the Balógun war-title of Ayede. A member of the family explained that her paternal uncle received the title from his mother’s Balógun line, with which his own patrikin became genealogically associated.5 It was precisely such slippage and ambiguity that gave rise to a debate about Yoruba descent that has never been fully resolved (Barber 1991,

Ethnogenesis from Within / 125

156–58; Watson 2003, 6–9). Peter Lloyd (1966) noted that genealogical claims through the mother’s line for access to corporately held titles and land were more common in southeastern Yoruba areas such as Ondo, and he concluded that Yoruba descent itself took different forms—agnatic in some regions, cognatic in others. This conception was based on certain assumptions about the character of Yoruba lineage structure and descent; not only that it was primarily patrilineal, but also that localized corporate groups of families and compounds formed the primary building blocks of Yoruba social and political organization. Given the Fortesian dogma of the day, Lloyd saw domestic groups as reproductive cells of the social order, clustered within lineages that served as dominant segments of Yoruba kingdoms. If the king (ọ ba) in council mediated the political interests of his ìwàrẹ̀ fà chiefs, the chiefs promoted the interests of the “houses” or “lineages” through which their corporately held titles devolved. There is already significant slippage within the model itself, because the jurisdictions of civil chiefs typically extend beyond the descent group as such to clusters of associated lineages and houses. In the Ekiti region where I worked, lineages were grouped into wards or quarters (àdúgbò), and these were the recognized domains of chiefly jurisdiction. But here, too, the semantic slippage was extreme. In the relatively decentralized kingdom of Ishan, for example, each quarter was named after the dominant lineage or ilé (house) in which its chieftaincy was vested (Apter 1995, 376–78). But more inclusively, these quarters also could be considered towns in their own right, marked by distinctive dialects and orisha (deities) within the kingdom at large. I wondered why Lloyd (1955a, 1968) had overlooked these broader political sodalities, which were so central to the political organization of towns like Ado-Ekiti where he had worked. In a sense, Lloyd abstracted the simplest descent-framed paradigm as the primary political unit and saw more inclusive coresidential arrangements as corporate, and in some cases cognatic, manifestations of the lineage principle itself.6 In what might seem like a minor footnote in Yoruba kinship studies, Bender (1970) challenged Lloyd’s agnatic and cognatic variations on methodological criteria. Whereas for Lloyd, and for much British social anthropology of the time, the corporate lineage was structurally primary, for Bender, with a US culturalist emphasis, the patrilineal idiom was logically prior to the actual social groupings that it generated on the ground. He argued that whatever the localized corporate groupings we may encounter empirically— predominately agnatic in some areas, cognatically mixed in others—we should not conflate descent as a symbolic idiom of genealogical reckoning with the residential, economic, and demographic characteristics of the actual

126 / Chapter Five

groups. The debate is significant because it approaches socio-semantic slippage between meaning and reference as a problem to be resolved one way or the other, rather than—as I shall suggest in due course—as central to the very deployment of the category ilé. Despite their different approaches to the shifting meanings of descent, however, both Lloyd and Bender saw it as a building block of Yoruba political and social organization, generating those patterns of opposition and inclusion that characterized the kingdomin-council. The dominant conception that households and lineages exist in towns, and build higher-level political relations between their representative chiefs, remained unchallenged. In a break from this foundationalist vision of descent, Eades (1980), following Fadipe (1970, 97–118), turned the genealogical meaning of ilé on its side, treating it as a residential group or compound first and foremost, with lineage groupings within its walls, rather than as the residential correlate of a prior corporate lineage. Such a view not only matched Yoruba conceptions of the compound as the primary unit of affiliation and corporate organization within towns, Eades argued, but also directly resolved the agnatic–cognatic debate by identifying a bilateral kinship ideology that is effectively pushed in a patrilineal direction because of the primary factors of virilocal residence and economic cooperation between fathers and sons (Eades 1980, 49–54).7 The agnatic descent group as household core emerges, in Eades’s view, as a secondary consequence of primary residence patterns. Households as primary building blocks “are best seen not as localised families, but as groups of co-residents, some of whom are related” (51). The secondary character of agnatic descent is further evidenced by the “Hawaiian” terminology of Yoruba kinship, which makes no distinctions between lineal and collateral relations.8 The admixture of relational types within the compound—residential and lineal, consanguineal and affinal, core and stranger, freeborn and slave— is further complicated by patterns of internal segmentation around halfbrothers (ọ mọ baba), offspring of the same father but of different mothers, reinforced by inheritance rules. If groups of full siblings (ọ mọ ìyá) form minimal units of segmentary opposition vis-à-vis domestic resources, over generations the same principle of differentiation extends throughout the more inclusive compound (ilé) of smaller households consolidated within larger political units. Not only are the primary factors “governing” composition of the ilé ambiguous—agnatic for Bender, agnatic and cognatic for Lloyd, and residential for Eades—but the referential scope of ilé remains fluid, opening up to frame larger political groupings of related compounds and narrowing all the way down to distinguish sibling groups within

Ethnogenesis from Within / 127

polygynous households. Indeed, the core unit of Yoruba social and political organization has defied clear demarcation and definition.

The Town in the Home In her monumental study of oríkì orílẹ̀ , the attributive praise-poetry that invokes origins, Karin Barber (1991) clarifies the fluid and dynamic parameters of the Yoruba ilé through the very discourse genre that governs its composition, since oríkì “play a part in the actual definition and constitution of groups” (138). Barber’s subtle and nuanced exegesis of shifting origins and boundaries within the category ilé in Okuku reveals a complex, labile, and negotiable social field in which descent, residence, patrilineal and matrilateral ties, and cooperative arrangements between attached lineage segments and fictive kin are variably invoked and adjusted to include or exclude according to context. From this more actor-oriented perspective grounded in social and discursive practice, the debate over the primacy of descent versus residence becomes moot. After reviewing the arguments of both camps, Barber (1991) concludes as follows: In Okuku, it was not possible, in the end, to propose either the “compound” or the “lineage” as the fundamental social unit. Rather, the principle of descent and the principle of residence were entwined and interpenetrated at every level, down to the foundations of social identity. And this identity was continually redefined according to the circumstances, giving rise to different “groups,” differently recruited in different situations, so that no single definition of a primary social unit was in the end possible. One term—ilé—was used for almost all significant groups: but it turned out to refer to different kinds of units in different circumstances. (158)

From this more situated perspective, no objective inventory of ilé within any town is possible because the units themselves are flexible and fluid. As Barber (1991, 159) notes, her table of twenty-nine ilé in Okuku should not be seen in terms of “solid and permanent social units” but rather “as an indication of the range of possibilities open to social groupings as they adjusted their boundaries according to context.” Indeed, to treat the table as an objective inventory would succumb to what Bourdieu (1977) calls the synoptic illusion, when practical schemes of social classification that make sense only when strategically deployed are abstracted into fixed hierarchies that violate the actual economy of logic on the ground (97–109). It is not Barber’s move toward a theory of practice, however, that I wish to

128 / Chapter Five

highlight, important as it is, but an even more radical paradigm shift that her rich ethnography suggests. As we shall see, the fluidity of the category ilé is not a function simply of its strategic deployment and shifting boundaries on the ground, but also of a deeper mode of semantic configuration. We begin with Barber’s paradoxical insight that the very oríkì orílẹ̀ with which “related” members of an ilé are praised, and which shape the flexible parameters of the compound unit, invoke not common ancestors on a genealogical tree but the towns of origin from which their forbears originally migrated: Oriki orile are one of the principal means by which groups of people who regard themselves as kin recognise each other and assert their unity. But they do so in terms of a common town of origin, and not, in the first instance, in terms of ancestry. The key emblems in oriki orile are always associated with the names of places. Oriki orile do include allusions to illustrious men and women among the ancestors of the group, but these allusions are attached to the notion of town of origin. Oriki orile do not trace genealogies, nor do they revolve around the notion of a lineage founder. The members of a group assert that they are “one” because they all came from the same place of origin, and distinguish themselves from people coming from other places. (Barber 1991, 145)

Barber reflects on how this association between kin groups and emblematic towns of origin might have developed, beginning with “primordial towns of origin” in which all members shared the same oríkì orílẹ̀ vis-à-vis outsiders from other towns, and distinguished themselves internally by lineage oríkì corresponding to “notional patrilineal kin groups” (148). According to this primordial starting point, the ilé must have originally designated agnatic lineages and lineage segments and would only later become associated with particular towns of origin through the subsequent population movements associated with political fission, competition for resources, slave-raiding campaigns, and the nineteenth-century wars. As some towns expanded and others were sacked, regrouping under warriors and strong men who offered protection and war booty, a resettlement pattern emerged of reconstituted “houses” defined primarily by their towns of origin. There is no question whether the turbulent warfare of the nineteenth century following the collapse of Old Oyo circa 1836 generated much population movement and resettlement, giving rise to new sociopolitical amalgamations. My own case study of the kingdom of Ayede provides an example of just this process, of stranger lineages and quarters settling around an original core to defend against Ibadan and Nupe predators (Apter 1992, 36–54). But

Ethnogenesis from Within / 129

I also would argue that such resettlement patterns are built into the cultural frameworks of Yoruba homes and towns in the first place. Barber appears to acknowledge this possibility when she reflects on the earlier forms of ilé: It is quite possible that before the nineteenth century, there was less diversity in the ile and in the relationships between them, fewer “stranger” groups and weaker bonds between and across compounds. Perhaps ile were more unitary and more strongly bounded. On the other hand . . . , [b]earing in mind the probable high degree of population mobility before the nineteenth century, one may suspect that flexibility of group boundaries, and the possibility of invoking a variety of principles of recruitment, was already present in the social system, to be drawn on in different ways and with increasing intensity as the need increased. (1991, 164)

It is just this latter consideration that I would like to push further. Rather than speculate about primordial towns in a somewhat mythic past, I would argue that the primary “place” of the town (ìlú) within the home or compound (ilé), as invoked by oríkì orílẹ̀ , is built into the cultural category of ilé—that the very location of the “town” in the “home” is always already primordial. If such a radical reconceptualization appears counterintuitive at first, it more accurately reflects Yoruba concepts of group membership based on principles of semantic designation and logical quantification that differ significantly from their English language counterparts (Verran 2001). One of the first implications of this paradigm shift—placing the town within the home—concerns notions of socio-logical priority. As long as we see lineages-cum-residential ilé as primary units of sociopolitical organization, as building blocks of the segmented groups that vie against each other for political power and control, then it makes no sense to say that they come from elsewhere, from outside the system itself. If towns are made up of houses, then how are houses made up of towns? Yet from a Yoruba perspective this conundrum can be solved. Not only does the town in the home make perfect sense, but once grasped, it resolves the agnatic versus cognatic debate and the lineage versus residence debate and obviates appeals to primordial towns. Getting there, however, requires a quick detour through Yoruba number theory.

The Whole in the Part “The different practices in classification that underlie the generation of predicating terms in English and Yoruba create different types of referring categories” writes Helen Verran (2001, 186) in her brilliant study of the

130 / Chapter Five

logic of Yoruba quantification. There is no way I can do justice to the depth and sophistication of the argument she develops, which itself instantiates the very recursive tallying embedded in Yoruba enumeration. Rather, I shall grossly simplify one of her major insights in order to illuminate an important difference between the English and Yoruba semantics of number. In English, and “Western” epistemology more generally, things, objects, and numbers in the world are conceived as “spatiotemporal particulars,” individual entities that form collections of specific kinds and types—in more formal terms, as members of abstract sets. The number six is a collection of six objects forming a group or set, or more abstractly, six units of one in an extended series on a number line; a family of four is composed of four individually related persons; a collection of residents living in houses within a spatiotemporal area combine to form a neighborhood. In Yoruba language and culture, Verran shows, things, objects, and numbers in the world are conceived as “sortal particulars,” qualitative sorts of “thinghood” that infuse the universe and that manifest themselves in different modes at particular times and places. Sortal particulars can manifest themselves within a plurality of objects that form what “we” would see as members of a set, but the objects themselves are secondary to the sortal particular that they instantiate. “Number, in Yoruba language talk, is a degree of dividedness” (Verran 2001, 198). Things, objects, and numbers in the world are modes of manifesting sortal particulars in a given situation, time, or place. Five oranges are not five individual oranges forming a group, but “orangeness” divided into a plurality of five. Set membership is not additive; rather it is differentiating or decompositional—it starts with the whole and breaks it up into parts. A family of four is the sortal particular of familyhood broken into four related persons; a neighborhood is a sortal particular of neighborhoodness broken down into residents and their homes, according to its mode of manifestation at a particular time and place. A few well-known Yoruba motifs and practices illustrate this logic of objectification and quantification very clearly. There is the myth of Orishanla as the primordial orisha who was sabotaged by his slave when the latter rolled a large boulder onto his master and smashed him into hundreds of pieces. When Orunmila tried to collect the pieces and put “Orisha” back together again, those fragments left behind became the various orisha in Yorubaland (Beier 1980, 6–7; Idowu 1962, 59–60). There are many interesting interpretations of this classic myth, especially as a master–slave dialectic, but the point I want to highlight here is how the very multiplicity of the orisha themselves represents a mode and degree of dividedness relative to a primordial whole. Naming ceremonies also illustrate individuation from a prior whole.

Ethnogenesis from Within / 131

Much is made about how Yoruba naming ceremonies held seven days after the birth of a child weigh against infant mortality before assigning a social identity, but I would underscore how the ceremony itself is a ritual of nominalization, individuating the child from the house or lineage from which he or she emerges. In a sense, the child is not born into a house or lineage but is born out of it, coming to manifest the lineage through the ritual mode of naming itself, as in the proverb Ilé ni à n´ wò k’á tó s’ọ mọ l’órúkọ (It is to the house that we look before we name the child).9 Oríkì themselves, as “modes of attribution,” also are central mechanisms of manifesting sortal particulars—the very shifting boundaries of inclusion and exclusion invoked by paternal lines, maternal lines, towns of origin, and significant landmarks instantiate “homeness” or “compoundness” in its plurality of forms at particular times and places. The meaning of ilé becomes less a question of what it “is” than of how—in what mode—it manifests. When we return to the place of the Yoruba town in the home, with Verran’s manifesting modes in mind, we can see that the building-block approach to household and lineage, in which individual ilé combine to form increasingly inclusive compounds and quarters, is doomed from the start. The search for the primary principles of these primary units, be they descent (agnatic or cognatic), virilocal residence, or some basic entanglement of the two, approaches the problem from the wrong direction. The very origin of ilé as a prior town—the town of origin indicated by oríkì orílẹ̀ —implies that the home or compound is a manifestation of the town; it is a mode of dividing the town into ilé, or a mode of dividedness of the town itself. We thus start from the town as it subdivides into ilé, not from the ilé as they combine to form towns. From the standpoint of Yoruba quantification, whereby objects emerge from prior wholes, every ilé is a manifestation of ìlú, a microcosm, as it were, of the kingdom at large.10 And just as every kingdom has a royal lineage with exclusive claims to the kingship, so every ilé has an agnatic core with exclusive claims to a lesser political or ritual title. The primordial ilé must have nonagnatic extras in the form of maternal kin, attached strangers, or fictive kin because if every ilé is potentially the ìlú that it manifests, it must contain within itself the basis of differentiating royal from nonroyal lines. In this model, the core ilé must contain both agnates and nonagnates, or outsiders living at the very center of the compound. Historically, the dynamics of political fission illustrate how this actually occurred. Disaffected princes, ambitious civil chiefs, or entrepreneurial Big Men who accumulated clients and resources frequently broke away to found their own kingdoms, establishing ruling dynasties with beaded crowns that eventually invoked the authority of Ile-Ife. Or within a kingdom, a rival

132 / Chapter Five

house could usurp the kingship to inaugurate a new ruling line. And if a powerful “house” could expand into the very kingdom that it “manifested,” so a “kingdom” could “shrink” into a quarter or house within a town. In Ayede, when the Eshubiyi line rose up in the nineteenth century to usurp the kingship from the Olua house, the latter dwindled to a shadow of its former glory, although its members retained ritual icons of sovereignty associated with their orisha and still refer to themselves as a town (ìlú) (fig. 5.1; Apter 1992, 45–54). Local histories often recall how subordinate towns that seceded from the center eventually rose up to displace their former rulers, reconstituting, as it were, the town in every home. The same cosmological whole in the part governs the manifesting modes of orisha worship as it transposes shifting boundaries of ilé and ìlú between social, political, and ritual domains. As I have argued elsewhere (Apter 1992, 149–61), the orisha are simultaneously one and many, allowing followers to form around a lineage core, lineages to combine within the jurisdiction of a quarter, or emerging royal lines to incorporate the gods of their displaced predecessors. But there is one ritual transposition I would like to highlight because it explains how a “town of origin” associated with an ilé becomes, or comes to manifest, the town where it resides. For during the celebration of its orisha, the house qua town shrine (ilé òrìṣ à) invokes its forbears and literally occupies the town, bringing its power from the outside bush into the town shrine, dominating the streets and crossroads, entering into other shrines and houses, stopping traffic, imposing fines, and taking possession of public space. During such ritual takeovers, the house of the orisha becomes the palace and kingdom that it manifests and assumes control over the town.11 Its priests and priestesses represent the kings as well as the military and civil chiefs of the former kingdoms and towns from which they migrated. Such ritual maneuvers are powerful because they manifest the potential of a real political takeover as well. Town shrines as ritual houses are internally organized both as palaces (ààfin) and kingdoms (ìlú), mirroring the actual palace of the kingdom with its expansive courtyard, verandah, parlor, and inner chambers, while protecting within its walls the ritual crowns, calabashes, and deities of the orisha’s town of origin. In principle, all kings, crowns, and orisha come from Ile-Ife as the sacred locus of original kingship, and in this sense all shrines manifest Ile-Ife within their cores. Historically, however, they represent the towns and quarters from which they migrated, as strangers joining a group of indigenes, disaffected political factions founding new kingdoms, or persecuted communities that were forced to relocate during the turbulent nineteenth-century “intertribal” wars. In Ayede-Ekiti, the three dominant

Figure 5.1. Priest of Orisha Olua, representing the former ruling house in Ayede. Author’s photograph, 1984 (from Apter 1992, plate 15).

134 / Chapter Five

orisha cults consolidated in the 1850s represent the composite character of the political community as it was centralized under the warlord Eshubiyi. The royal Yemoja cult, with its associated orisha, represents Ayede’s link with Ibadan, from which the Àtá Eshubiyi received his crown, and reassembles the prior constellation of Oyo-centric deities that the Ibadan warlord Olúyọ̀ lé reconstituted, according Oyo’s Shango a central place. As the Yemoja and Shango priestesses lead the ritual procession throughout Ayede, they carry the beaded crowns and calabashes of kingship, which they incarnate through spirit possession. The priestesses receive salutations of “Kabiyesi!” (“Your highness!”), to which they respond “Ẹ ṣ ẹ un!” in a hypercorrect accent that explicitly marks Ibadan provenance.12 The cult of Orisha Ojuna represents immigrants from Ikole, with an altar dedicated to the Elékọ̀ lẹ́ , recalling the Àtá Eshubiyi’s town of origin before he married into Iye and founded Ayede. The nonroyal cult of Orisha Iyagba represents immigrants from the Yagba kingdom of Alu and its associated towns, glorifying the military chieftaincy of the Balógun Áòfin with incipient icons of rival kingship. If Orisha Iyagba is a nonroyal cult within Ayede, it nonetheless manifests Iyagba kingship in its festival praises and ritual iconography. I have discussed the political dynamics of ritual mobilization in greater depth elsewhere (Apter 1992), but here I wish to highlight the ritual manifestation of the whole in the part in that town shrines double as kingdoms unto themselves. Indeed, town shrines are internally organized as miniature governments. As Fadipe (1970) explains, “The priesthood of every òrìsà is organized on the model of the political system. There is not only a hierarchy of officials, but these officials also bear titles that have been adopted from the secular government” (284–85). This model of the orisha cult as ritual monarchy or kingdom is not an aberration of an “original” family or lineage-based cult that evolved over time into complex forms, as one eminent scholar of Yoruba religion has maintained (Idowu 1962, 130–32), but rather constitutes the primordial form of the orisha cult house as it manifests the whole in the part, from which the genealogically defined lineage or family emerges. Nowhere is this mode of particularization more clearly evidenced than in Cuba and Brazil, where the Lucumí and Nagô “nations” developed into genealogically defined ritual “houses” and “families,” replete with initiatory bloodlines and branches.

Atlantic Transformations: Lucumí and Nagô There is no question whether the lexicon of African “nations” that emerged in the era of Atlantic slavery played an important if complex role in the

Ethnogenesis from Within / 135

development of ethnic identities in the Americas as well as in West and Central Africa. We know that much was invented on both sides of the Atlantic (Palmié 2010), as well as during the Middle Passage (Mintz and Price [1976] 1992), making the identification of New World “nations” with prior source populations in Africa problematic at best. Curtin (1969, 183–90) gives three reasons for confusion, stemming from the interests of slavers and planters: (1) the ethnic lumping of disparate groups in Africa under an umbrella ethnonym, such as Mandingo, (2) the identification of Africans with their ports of embarkation (e.g., Mina) rather than the hinterlands from which they may have arrived, and (3) the association of slaves with behavioral stereotypes in the colonies, such that “rebelliousness,” “docility,” physical strength, or entrepreneurial propensities became ascribed “national” characteristics by which they were classified. Yet rather than occlude ethnic origins as such, these factors help us understand how African “national” identities emerged out of Afro–European encounters, subsuming specific populations within more broadly named collectivities. Much also has been said, following Bastide ([1960] 1978, 1971), of “the new social frameworks” that reshaped neo-African communities in the Americas, including patterns of labor segmentation, color stratification, and the underlying dynamics of class formation that motivated re-Africanization among free blacks. Throughout the Iberian imperial landscape, the religious brotherhoods of the Catholic Church stand out as especially formative institutional loci in the making of New World African identities, serving as “generative bases” (D. Brown 2003, 34) or “incubating cells” (Palmié 1993, 341) of ethnically denominated socioreligious groups such as Congo, Hausa, Ibo, Carabalí, Jeje, Mina, Fanti, and Ganga, among which the Lucumí of Cuba and the Nagô of Brazil emerged as preeminent ethnic “nations.” The historical sociology of New World Catholic brotherhoods—cofradías and cabildos in Cuba, irmandades in Brazil—involves a complex web of economic, sociopolitical, and religious connections that begin with sixteenthcentury colonization, when religious fraternities of patron saints incorporated enslaved Africans into religious sodalities that doubled as social clubs and mutual-aid societies. Based primarily in cities and provincial towns, the cabildos de nación organized African-born slaves according to their professed nations of origin, establishing crucibles of syncretism and creolization as African-derived religious and cultural practices coalesced within a Catholic frame. If, as administrative arms of church and state, these brotherhoods recognized the authority of the governor, they also pushed toward political autonomy as self-governing corporations that financed manumission for their members, and in extreme cases, fomented rebellion.13 It is not the

136 / Chapter Five

politics of the brotherhoods as such that I wish to emphasize here, however, but their development from miniature kingdoms into ritual houses during the nineteenth and twentieth centuries. Lucumí Counterpoint Like the orisha cults of Yorubaland, the cabildos de nación in Cuba were modeled on the political system in which they were embedded, incorporating monarchic, military, and religious orders within their iconography and institutional organization. As David Brown (2003) explains in his formidable study of Cuban Santería, all the cabildos and their associated ethnic “nations,” whether Lucumí, Congo, or Carabalí, initially emerged within the same organizational framework, even if they differed in influence and prestige: “The cabildo de nación borrowed the title of official state, church, and civic institutions, especially from the monarchy and the military, to describe their offices. Cabildo kings . . . were supported by a ‘court,’ consisting of a ‘queen’ or two (second in rank), and lower-ranking officers, such as the abanderado (standard bearer) and the mayor de plaza (chief of ceremonies)” (35). Cabildos were, in effect, “miniature neo-African monarchies” (D. Brown 2003, 35), referred to as reinados (kingdoms) in popular discourse. Represented by flags that linked African “nations” to Catholic saints, their members paraded under the umbrella of the church, taking to the streets during Carnival, saint’s days, and the Day of Kings, while demanding small payments (aguinaldo) from onlookers as a form of ritual tribute and protection. Entering Havana from beyond the city walls (extramuros), the “African” kings and queens in procession constituted something of an occupying force: “The group literally took over the street from curb to curb in a public space, where its collective identity and complex hierarchy were on display” (D. Brown 2003, 36–37). Fusing European codes of royalism and militarism with African drum rhythms and associated choreographies, the cabildo processions established performative arenas of symbolic appropriation and sociopolitical empowerment, where enslaved blacks could flex some muscle within limits uneasily tolerated by the authorities.14 There is no question whether European markers of hierarchy, power, and socioeconomic value fed into cabildo iconography and processional conventions, producing an Atlantic Creole assemblage that circuitously fed back not only to Freetown and Lagos, but also into the Nigerian hinterland. If the vectors of transmission are not always clear, the circum-Atlantic parameters are expansive. Nunley (1987, 131) shows how the Ode-Lay masquerades of

Ethnogenesis from Within / 137

Freetown were embellished with a “fancy” and “fierce” aesthetic that carried with it New World finery; Roach (1996) traces performative chains of ritual surrogation that link Mardi Gras pageants to the Elizabethan stage. Indeed, David Brown (2003, 39) cites a contemporaneous account of an 1856 Day of Kings festival in which a Havana cabildo ruler sported “a genuine costume of a king of the Middle Ages, a very proper red, close coat, velvet vest and a magnificent gilt paper crown.” Yet such carnivalesque phantasms of feudal monarchy are not so easily reduced to plantocratic kitsch, for they reference a broader royal lexicon imbued with deadly authority. In the orisha festivals of Ayede-Ekiti, for example, the Àtá (ọ ba, king) wears not only a European-style beaded crown, but also royal robes of embroidered àrán cloth made from an imported European red velvet brocade reserved exclusively for kings and their ritual representatives (Apter 1992, 108–9).15 My point is not to reduce sacred kingship to African replicas of European replicas “all the way down,” but rather to emphasize the African grammars of sovereignty motivating the appropriation of European emblems and signs, a coterminous process of royalist codification that occurred on both sides of the Atlantic.16 With this “deeper” cultural logic in mind, we can revise the standard depiction of cabildos as miniature Bourbon states—providing safe haven for African gods and kings—by emphasizing a much greater continuity with orisha cult performance and organization than has generally been acknowledged. If much has been written on the dynamic correspondences between Yoruba deities and Catholic saints, less has been said of their isomorphic performance genres; between saint festivals and the Day of Kings in the Catholic calendar, and the public annual festivals culminating in the Day of Carrying Water (Ijọ́ Ipọ nmi) for the orisha. In formal terms, when Yoruba priestesses mobilize their cults to take over the town, they take possession of space and time. As they process from bush shrine to the market and palace, they honor political patrons and historic locations with propitiations and oríkì, demanding tribute from those praised, fining cars and trucks that interrupt them on the roads, and reinscribing public space with the history of the cult. “Today does not own itself,” they sing, “today belongs to X,” naming the principal deity of their orisha cult. In Ayede, the Yemoja and Orisha Iyagba processions intersect, waging muted ritual battle as both groups take over the town. Iyagba warrior priestesses brandish cutlasses and spears, mobilized by military rhythms as they confront Yemoja’s avatars from Ibadan and Oyo. Each cult empowers the king and the kingdom in its own name, activating its distinctive past to bear upon the present. Each cult also manifests “hot” and “cool” gods with characteristic choreographies and rhythms—measured and

138 / Chapter Five

staid versus rapid and staccato—emphasizing reproduction and transformation, authority and power. During the orisha’s festival, the kingdom that it reproduces and transforms becomes the cult’s kingdom of origin. The processions of Catholic virgins and saints mirrored the outings of the orisha with striking symmetry, converging within the cabildos de nación in both formal and substantive terms. Both traditions carry icons invested with power through public streets and pathways before delivering them to their sacred destinations. As in the cabildo processions, the orisha priestesses take over public space, exacting economic tribute for cult support and protection just as aguinaldos were demanded in Havana. Like the miniaturized court societies of the cabildos, the orisha cult priestesses in procession formed ritualized kings in council, with associated chiefs and warrior priestesses among the inner entourage. Furthermore, the “fancy” and “savage” aesthetic opposition that characterized European versus “African” codes (D. Brown 2003, 47–51) in Cuba, contrasting formal decorum and comportment against those raffia fringes and animal horns that dominated periodic breakout sessions of animated drumming and dancing (corros), were already embedded within orisha cult registers that indexed the “cool” hierarchy of the kingdom against the “hot” powers of transformation and subversion. Even specific ritual syntagms carried over. Describing a royal cabildo procession in 1844, the US physician J. G. F. Wurdemann noted, “The whole gang was under the command of a negro marshall, who, with a drawn sword, having a small piece of sugar-cane stuck on its point, was continually on the move to preserve order in the ranks” (cited in D. Brown 2003, 37). For David Brown, this telling motif suggests a potential assault on slavery itself: “What emblem could better embody the bitter-sweetness of black carnival in the period’s premier sugar-producing slave society than this precious detail? The entire Cuban ‘sugar palace,’ at its height between 1820 and 1860 . . . revolved around that piece of cane, now sword pierced and held aloft by the formally dressed and officially empowered black officeholder, who was also a slave. Did his performance slyly comment on ‘cane’ as the ‘signifying’ slave songs of the antebellum United States ‘sang’ upon ‘corn’?” (D. Brown 2003, 47–48). Without denying the plausibility of this interpretation, in light of the surplus of meanings generated by carnival, I would only underscore that this “precious detail” also is found in Ayede’s orisha festivals, where, among the category of ológun “field marshals,” priestesses waging ritual battle carry spears and cutlasses tipped with kola nut pieces—instruments of death capped by gifts of life.17 There, the kola nut serves as a minimalist protective cover, highlighting the fragile containment of militant dismemberment so easily unleashed.

Ethnogenesis from Within / 139

My intention is not simply to pinpoint the Yoruba origins of specific ritual patterns and practices, but to show that the core ritual systems within orisha cults and at least some of the cabildos de nación were virtually the same in the nineteenth century—they replicated within their restricted codes and corporate organization the larger monarchies in which they were embedded, while also repossessing the public sphere during processions of collective renewal and empowerment. Moreover, the quantification of the whole in the part informing the òrìṣ à-ocha system helps account for two trends that have inspired considerable debate. The first concerns the scope and character of Lucumí influence in the colonial period vis-à-vis other ethnic nations and religious societies that developed coterminously. The second addresses the historic transformation of Lucumí cabildos into “house-temples” (casa templos). The question of Lucumí influence is all the more puzzling when one considers that their cabildos were fewer in number than those of the Congo and Carabalí nations (D. Brown 2003, 63; López Valdés 1994). How did a minority ethnic nation emerge as such a dominant Afro-Cuban identity? Part of the answer seems to lie with its religious culture’s absorptive capacity. Characteristically open to “stranger” deities, the Yoruba-Fon-related religions of Cuban Santería and Brazilian Candomblé have incorporated European, Congolese, Indian, and Creole gods and spirits within their houses and altars, thereby gaining ground in the field of Afro-Caribbean religious production while setting the stage for subsequent contests over purity and authenticity. The same principle of ecumenicalism seems to have applied to Lucumí organizations that not only consolidated those subgroups (Oyo, Ketu, Ijesha, etc.) speaking dialects of what we now call the Yoruba language, but also came to claim as its ritual descendants non-Yoruba-speaking nations as well. In what may well have been a remodeling of Spanish hyphenated naming conventions, we find Lucumí-Eyó (Oyo), -Egua (Egba), -Agguado (Egbado), -Iyecha (Ijesha) also extending to Lucumí-Achanti (Ashanti), -Fanti, -Popos (Dahomey), -Araras, -Benin, -Magi (Mahi), -Bariba, -Jausa (Hausa), -Ibo, and -Tapa (Nupe).18 In addition to boosting Lucumí recruits and affiliates, this expansive Lucumí association with other nations earlier on may have further influenced the organization of their distinctive cabildos by transposing the modular organization of the miniature kingdom into black cabildos across the board.19 Such a broadening of Lucumí socio-religious jurisdiction might be cast as strategic consolidation at a time when the population of African-born blacks was declining and government assaults against black cabildos and people of color were on the rise, triggered by the 1886 emancipation of slaves. But the

140 / Chapter Five

cultural conditions of such broad consolidation were built into the concept of a Lucumí totality, which came to manifest its segmented particulars both laterally, as subethnic affiliations, and lineally, as members of initiatory lineages. As social kinship gave way to ritual kinship between the 1880s and the 1920s, the kingdom-cum-cabildo was reconstituted within the house or home. David Brown (2003, 62–66) identifies three historical trajectories by which the cabildo de nación gave way to the early twentieth-century casa templo: one of continuous transformation; one of covert derivation; and one of more independent inauguration by ex-slaves and creoles who established ritual family lines. But in many ways these different trajectories represent manifestations of an underlying process of ritual segmentation and differentiation, giving rise to a transitional cabildo—an “umbrella organization connecting numerous emergent houses and ritual family lines”—that culminated in fully fledged family fission. Thus emerged the modern twentiethcentury “‘house’ of Ocha . . . composed of a single extended ‘ritual family’ (familia de santo) directed by a single priestly elder who practices within a private domicile in more or less discrete or underground fashion” (D. Brown 2003, 67). Critical to Brown’s analysis is how ritual kinship and kingship are mutually sustained. If ritual kinship is traced through the padrino (godfather) or madrina (godmother) as house owner and elder, actual initiation is performed by the ọ bá-oriaté (lit., “king,” “master of ceremonies”) in the preparation and consecration of the neophyte’s head.20 Moreover, icons of kingship are intimately associated with the ritual prerogatives of the ocha themselves, whether channeling past kings of Oyo with Chango or by symbolic and iconographic association with royal crowns, cloths, scepters, and tributes. Revising Ortiz’s theory of cabildo origins for the Cuban Lucumí-Santería religion, Brown places greater emphasis on eponymous ancestors as “root founders” (raíces) of houses: A deeper “starting point” of the Lucumí religion lies not so much in the cabildos as in revered personages who founded and led the religion’s Ifá and Ocha “houses” (casas or ilés) and gave birth to their corresponding ramas (“branches”). A casa (ilé in Lucumí) refers not merely to the physical domicile of a priest, but to a ritual “family” of priests initiated by an elder of Ocha. . . . Rama refers to the genealogical lineage or lineage tributary from which priests (and their houses) descend. Casas are the nodes, as it were, of the ramas, and these terms together constitute the sacred genealogical organizing principles of the Lucumí tradition. . . . Upon consecrating the “head” of an initiate, a ritual elder—the “owner” (dueño) of a “house”—becomes

Ethnogenesis from Within / 141 the “godparent” (padrino or madrina) of a new “godchild” (ahijado or ahijada). The new member of the house comes to have “brothers” and “sisters” (hermanos and hermanas) among the house’s other initiated priests, both alive and deceased (eguns). The term “house” can also subsume collegial or “working” relationships with priests of other houses or ramas.21 (D. Brown 2003, 74)

I have quoted this passage at length because it so clearly identifies the genealogical matrix within which Lucumí houses and their ritual lineages were produced, mirroring the characteristic principles of fission and fusion in segmentary lineage systems more generally. Although house lineages and their segments appear to continuously divide and break free over time, such fissiparous pressures are counterbalanced by consolidating trends of lineage fusion, when, as Brown points out, the house subsumes “‘working’ relationships with priests of other houses or ramas.” My point in highlighting this countervailing trend is not merely to complement fission with fusion, but to reestablish the immanent frame of the Lucumí casa (ilé) as whole-inthe-part. What Brown sees as a basic break between the old cabildos and the founding of new houses can be recast in much more continuous terms, as a mode of sortal—in this case genealogical—particularization, privileging an initiatory principle of lineal descent that came to dominate the town in the home. If the miniature kingdoms of the nineteenth century no longer shaped corporate architecture and organization, they nonetheless persisted within the domestic iconography of the casa templos, manifesting through the obá’s rites of consecration, as well as in royal and warrior ochas, and their emblems, altars, and shrines.22 Nagô Themes and Variations The development of the Nagô nation within Brazilian Candomblé is so embedded within twentieth-century ideologies of Yoruba purity and authenticity that is difficult to excavate its genesis from within the lay Catholic brotherhoods (irmandades) of the colonial period. In his foundational study of the African religions of Brazil, Roger Bastide ([1960] 1978) explored the multifarious affinities between the Fon and Yoruba divinities of the Jeje and Quêto/Nagô nations and their saintly counterparts in Catholicism, locating them within a broad range of class contexts and regional variations. Salvador de Bahia represented the genuine or “pure” Candomblé for Bastide, with “corrupted” and “degraded” forms spreading inland and southward, reflecting the “anomic” consequences of industrialization and labor migration.

142 / Chapter Five

Areas like Minas Gerais, where gold mining in the eighteenth century supplanted the declining sugar economy in the northeast, were hardly worthy candidates for the true Candomblé, where instead one found the so-called Bantu “sects” of “congadas,” festival groups organized by such black brotherhoods as Our Lady of the Rosary, St. Balthazar, Saint Benedict, Saint Iphigenia, the Girdle of Saint Francis, and so on. But it is here, in the mining communities of the colonial period, that Bastide developed his analysis of Brazil’s “two Catholicisms,” one white and European, the other black and Africanized, ritually reconstituting black kingship and court society. Indeed, these brotherhoods in Bastide’s discussion parallel the Cuban cabildo in crucial ways. Both served as mutual aid societies that rotated credit, guaranteed funerals and burials, and financed the manumission of a certain number of slaves each year (Bastide [1960] 1978, 116). Both fomented a sense of shared ethnic kinship. Both took over the streets in public processions during holy days. But the most striking congruity lies in the royal offices and titles and the protocols that these latter commanded: the still vivid memories of African kingship emerge even more plainly in the congadas. These festivals accepted the continuance of a monarchic regime for Brazilian Negroes—in an adulterated form, of course, and incorporated into the worship of Our Lady of the Rosary. The earliest mention of a congada is in the town of Iguarassu in Pernambuco in 1700, but it already existed, at least in fragmentary form, in the middle of the seventeenth century, and its origin can be traced back to Portugal. Pereira da Costa tells us that each parish had its king, queen, secretary of state, marshal, herald of arms, ladiesin-waiting, etc., who were addressed as “Your Majesty,” “Your Excellency” or “Madam.” The election was held on the feast day of Our Lady of the Rosary and was the occasion for dancing, which varied in type according to the ethnic origin of the king. (Bastide [1960] 1978, 120)

Here we confront the miniature monarchy in motion, replete with entourage and royal salutations within a “Bantu” rather than Mina or Nagô ethnic nation. But such designations had a fluid content, indicated by the variable ethnicity of the king and associated choreographic styles. The festivals culminated in ritual coronation, processing to the chapel, where “the priest consecrated the man whom the brotherhood had chosen by placing a cardboard crown on his head” (Bastide [1960] 1978, 121). There is no question that Congolese slaves—especially those from the Kingdom of Kongo—would have recognized the marriage of Catholicism and monarchy in the brotherhoods from their own dynastic conversions,

Ethnogenesis from Within / 143

beginning with Nzinga a Nkuwu’s baptism in 1491, and consolidated under his son Alfonso, who developed Christianity into a royal Kongo cult (Fromont 2014; Thornton 1984, 148). Though many of the African coastal societies were touched by missionaries during the centuries of the Atlantic slave trade, none appropriated the theocratic models of Catholicism as fully as did Kongo in the sixteenth and seventeenth centuries (see also Fromont 2014). Yet it would be a mistake to attribute the Minas Gerais ritual monarchies to Congolese influences alone, because the majority of enslaved Africans during the formative years of the brotherhoods, from 1690–1750, came from the Mina coast and were primarily Fon and Yoruba (Kiddy 2005, 39– 45). It was only after 1750 that Congolese slaves came to dominate demographically, putting their “Bantu” ethnic stamp on the black brotherhoods of Minas Gerais.23 As Kiddy explains, The documentation demonstrates that Brazilians had started importing Mina slaves into the port of Rio de Janeiro at the beginning of the eighteenth century, and many of those slaves were destined for the mines. These slaves became the earliest members of the rosary brotherhoods in Minas Gerais and participated in shaping the heterogeneous communities that the brotherhoods would become. Within the brotherhoods, West African slaves could begin to reconstruct their worlds around common cultural elements such as kinship, kingship, expressions of hierarchy, and a link with the ancestors (49).

In other words, the formative framework of black brotherhood royalism was significantly Fon and Yoruba in origin, establishing a West African politicoritual grammar that was subsequently filled with Central African content, as in songs, rhythms, names, and even nations. I would argue, however, that beneath the “Bantu” exteriors that built up around it, a more fundamental Yoruba-Fon substrate endured, motivating the very formal transposition of whole into part, of kingdom into brotherhood.24 I am aware that this line of interpretation can be pushed too far, succumbing to a cultural chauvinism not unknown among Yoruba specialists.25 However, it is precisely the more abstract and formal features of generative cultural models that are most structurally determinative yet most difficult to perceive. The complementary trajectory is easier to discern—how the Yoruba framework of gods and kings directly associated with the Nagô nation came to take over Bantu and other non-Nagô Candomblé houses, whether by absorption, imposition, or conscious appropriation. In Bahia, Bastide explains, despite the multiplicity of Congo and Angola houses, a general Yoruba paradigm

144 / Chapter Five

prevailed: “The Yoruba imposed their divinities and the structure of their ceremonies on the other ‘nations.’ The result is that today everyone worships the same gods in his own language, with his own music, in structurally similar ceremonies. . . . Throughout this region the prestige of the Nagô was so high that rivalry compelled the other ‘nations’ to borrow the organization system of their cult, along with their orixás, which they identified with their own voduns or spirits. They borrowed not only the essential features of the Nagô rites but even their priestly hierarchy” ([1960] 1978, 194–95, emphasis added). From this Yoruba-centered perspective, all gods are orisha (orixa), all rituals, sacrifices, and invocations follow Nagô conventions, and all Candomblé nations replicate the core features of the ritual Nagô nation form. Even the cult of Indian spirits adopted Nagô norms, giving rise to the caboclo Candomblé houses. Whether rightly or wrongly, Bastide attributes such mimetic appropriation to Nagô’s ruling status in the hierarchy of Bahian Candomblé houses, a position earned by its imputed fidelity to an authentic African tradition: “Thus the prestige of the ‘Nagô’ finally won out everywhere. Their prestige derived from their having upheld the ancestral religion most faithfully in the original form in which it had been brought to America by Ketu priests captured by the Dahomans and sold into slavery in Bahia. Thanks to the initiation of generation after generation of new filhas de santo, vestals of the sacred fire, the tradition has been maintained without any adulteration or falsification” (Bastide [1960] 1978, 197).26 Bastide’s line of argument reveals not only the modular reproduction of the Nagô ritual nation throughout non-Nagô Candomblé houses, but also the complicity of the anthropologist in ratifying claims of ritual authenticity in what would become a religious purification movement.27 Central to Bastide’s perspective, however, is a model of ritual reproduction through initiatory lineages that consolidates the bloodlines of Nagô purity within a competitive field of Candomblé houses, designated by ilé (house) in Yoruba and either casa (house) or terreiro (yard) in Portuguese. We are now in position to trace the trajectory of Candomblé’s privileged Nagô line from Afro-Catholic irmandades to the genealogically constituted houses and families that—in Yoruba terms—they came to manifest. The emergence of Candomblé houses within Afro-Bahian society is difficult to pinpoint, but Parés (2004, 189) argues that by the early nineteenth century a number of ethnic “nations,” particularly Angola, Jeje, and Nagô, were coalescing around work crews (cantos), Catholic brotherhoods, secular dance-drum gatherings (batuques), and what were already identified as Africa-derived religious congregations. Parés further suggests that a predominantly Jeje (Fon-Dahomean) religious template had already formed when

Ethnogenesis from Within / 145

the relatively late and concentrated influx of Nagô slaves in the 1820s began in Bahia, revaluing the Jeje-based “vudum” deities with the Nagô-Yoruba “orixá” (190). By the 1830s, Nagô ritual dominance was portended with the founding of the Casa Branca do Engenho Velho, also known as Ile Iyá Nassô after the mother of its putative founding priestess, which became known as the original Candomblé house from which the “pure” Nagô branches have descended. The dual naming is significant because it points to two important originary frames: the first, as “the White House of the Old Sugar Mill,” invokes nothing less than the Great House of the sugar plantocracy, relating colorations of whiteness and purity that, I will argue, took on racial overtones. To this day, this “house” and those branches founded by its ritual descendants are known as the “great houses” of Candomblé (Matory 2005, 125). The second designation as the house of Iyá Nassô frames the temple with reference to its charismatic founder, Marcelina-Obatossi, whose mother held the title of Iyá Nassô in the Nagô-Yoruba Xango (Shango) cult.28 Variant accounts of how they returned together to Ketu and thereafter brought back the “true,” Oyo-based, tradition to Bahia established an important return-to-the-homeland precedent for re-Africanizing Nagô-Candomblé lines according to authoritative standards of ritual purity.29 But it is the logic of eponymous nominalization and the lineages that founding ancestors inaugurate that—as with the Lucumí casas de templos—I wish to emphasize at this juncture. The very plurality of names—the one institutionally framed, the other focused on personages—captures a movement of particularization in which lineal modes of genealogical reckoning emerge from the house and rise to the fore. If cosmologically the Candomblé houses manifested Oyo and Ketu kingship within their walls as condensed mystical geographies, sociologically the two dominant “wholes” were Catholic brotherhoods and sugar plantations, representing the urban and rural “niches” in colonial Brazil where African religions reemerged (Bastide [1960] 1978). Like the congadas of Minas Gerais further south, the Candomblé houses developed from brotherhoods. According to Paul Johnson, “As a community . . . the terreiro Engenho Velho had its roots in the Catholic brother- and sisterhoods (irmandades), the men with Our Lord of Martyrs, the women with the order of Our Lady of the Good Death (Nossa Senhora da Boa Morte)” (2002, 75). It was out of this sisterhood that Marcelina-Obatossi emerged, founding the terreiro named after her “mother,” Iyá Nassô (P. Johnson 2002, 75) to inaugurate a lineage model of what Matory (2005, 125) calls “initiatic families,” thus setting the stage for segmentation and fission within the ritual lineage at large. As Paul Johnson explains:

146 / Chapter Five With the death of the priestess and leader, Marcelina-Obatossi, succession disputes led to fission and the splintering off of two new terreiros: one in the neighborhood of Gantois, Iyá Omi Ase Iyámase, usually referred to simply as Gantois; the other called Axé Opô Afonjá (the force of the staff of Afonjá). These three houses, along with Alaketu, comprise the traditional houses of the Nagô-Ketu nation, the trunk of the tree from which thousands of descendant houses would branch and flower. They serve both as the genealogical progenitors of many terreiros in Brazil and as the authoritative model of tradition and correct liturgy for many more. (2002, 76).

Here we see in full evidence how the blood-based logic of ritual initiation within the “royal” line of Marcelina-Obatossi took precedence and continued to shape relations within and between Candomblé houses, how blood became the dominant idiom of the “kingdom” within the cult. The connection between ritual kinship and the collective power of the terreiro is nicely captured by Matory’s discussion of the Yoruba concept of àṣ ẹ (axé in Brazil), a concept in Yorubaland that relates the animating force of sacrificial blood and associated medicines to personal power, discursive agency, authority, and kingly command, but which in Brazil has come to further specify the very shared genealogical substance of initiatic families: In Brazil, axé can be a countable thing. An axé is the membership of a temple (or a family of temples), united by the same continuous ritual transmission of axé from a founding priest or priestess to the “children-in-saint” whom they have initiated and so forth. Thus, one might ask a worshiper which axé he or she belongs to. In reply, he or she might identify the initiatic lineage by the name of its founder, its temple of origin, or its currently ranking temple. An axé (temple community or family of temples) is held together by shared axé (ritually constituted life-force) and, ideally, by a shared set of ritual conventions that are supposed to have remained unchanged since the founding of that axé (family of temples). (2005, 124)

Through the lens of axé, we can see how the great White House of the Old Sugar Mill, sometimes referred to as just Casa Branca (White House), came to manifest that of its founding priestess, Marcelina-Obatossi, disseminating through lineages of her initiated godchildren. The very names of those houses that broke away through fission are at once “linked to the axé of Iyá Nassô” (Matory 2005, 125) yet emphasize their own distinctive identities as crucibles of their own axé. The “Iyá Omi Ase” of Gantois means “mother of the water of àṣ ẹ ” in Yoruba, at least initially referencing the ritually loaded

Ethnogenesis from Within / 147

term “water,” which in orisha worship refers to the reproductive blood of mothers as a potent manifestation of collective àṣ ẹ (see Apter 1992, 97–116), whereas the “Axé Opô Afonjá” references the àṣ ẹ of the historic Àrẹ -ọ̀ nàkakan`fò of Ilorin who, as head of the army of the Oyo Empire, led a rebellion against its king (S. Johnson 1921, 191–92), thereby serving as an effective charter for a Candomblé house that was born of fission. But as Matory has also emphasized in his quotation, “axé can be a countable thing,” a statement that accurately captures, from a Yoruba perspective, the emergence of quantity out of quality or thinghood; that is, from an encapsulating category or totality of axé rather than additive quanta forming aggregate sets. That the sense or reference of this totality may change according to spatial location and historical context does not detract from its countability through particularization but only underscores the variable wholes through which its parts—as cult initiates—are manifestly made. The members of a Candomblé house, like those of an orisha cult or Lucumí casa, are not recruited or assembled as a subset of a given population; rather they are generated by axé, ritually crowned and reborn through initiation. One vector of transmission, the town in the home, ritually transposed as the kingdom in the cult, was manifested through the historical emergence of the Candomblé house from the Catholic brotherhood, in which the various kingdoms of God, the colonial state, and the Dahomean-Yoruba homelands were reinscribed. Another powerful vector of transmission motivating discourses of ritual purification, during what Parés (2004, 191–98) identifies as the first and second phases of Candomblé’s Nagôization (1870s–1930s), appears to derive from the earlier sugar plantations of the nineteenth century and their caste-like ideologies of racial purity and miscegenation. Like the Great Houses of Nagô Candomblé, the casa grande of the sugar plantocracy was obsessed with hierarchy, rank, and purity. Although much has been written both for and against the myth of racial democracy in Brazil, and its antecedent ideologies of color stratification under slavery, there is no question that ideas about honor, status, and blood naturalized productive relations on the sugar plantations. Bastide ([1960] 1978) describes the casa grande as the hub of the plantation system at large: The master’s family was endogamous; it wanted no black blood in its veins. A wife was chosen with an eye to her racial purity and her fitness to bear her husband’s children and propagate his line. . . . The intermediate class consisted of poor whites who could survive only by integrating themselves as dependents within the only stable units in the colony, the big landowning families, and of mulattoes or free Negroes almost totally assimilated into the

148 / Chapter Five Portuguese civilization. The house slaves were selected for their beauty, intelligence, health, and cleanliness from the Creole blacks or the Mina or Nagô Africans, i.e. almost exclusively from the West African group. The field hands were usually Bantu or semi-Bantu. In short, social status increased with proximity to European values as represented by the master and his wife. (68)

Even if we allow for a certain degree of stereotyping in this portrayal of the casa grande, the language of racial purity was clearly central to an ethnically stratified productive regime ruled by white blood, followed by descending orders of status and labor correlated with increasing degrees of “polluting” blackness.30 Upward mobility was extremely limited for blacks within this hegemonic system, resulting in strategic miscegenation from below, which, again following Bastide, was “characterized by the expression ‘limpar o sangue,’ the purging of the blood by sleeping with whites and producing children with lighter skin, whose white fathers would help them along and who might in this way be freed from the yoke of slavery and enjoy an advantage when it came to economic competition” (Bastide [1960] 1978, 68). From the standpoint of black bondage, if such “cleansing” led toward freedom, for white planters it threatened the categorical separations that supported their power and privilege. Thus the fear and threat of racial mixing encouraged the policing of bloodlines from above, excluding mulattos from the inner circles of elite marriage, education, and society while criminalizing blackness at the margins. The Brazilian Creole was paradoxically a sign of “progress” and a creeping threat to white overrule.31 When seen against the historical backdrop of the casa grande and its logic of cleansing, the cultivated purity in the “great houses” of Bahian Candomblé resonates with striking symmetry. They maintain the “pure” NagôKetu-Oyo tradition while other houses are syncretistic, mixing categories and practices from dubious sources among mestizo-caboclo spirits and their devotees. Nagô orthodoxy and ritual purity not only claim fidelity to African precedents but also refer to the quality of a house’s axé, that ritual blood of the initiatic family that is channeled by lineages and produces new “children.” Like the white plantation overlords, the matriarchs and patriarchs of the Nagô houses regulate the propagation of pure lines of descent, uncorrupted by the dubious ancestors (egum) and mixed spirits (caboclos) that are kept at bay (Matory 2005, 129). A systematic inversion and displacement is at work that in some ways reverses the casa grande, since the telos of ritual purification is embedded in a project of re-Africanization. Yet to say that blackness triumphs over whiteness in this context misconstrues the colorations of race and nation in Candomblé, for ritual whiteness sustains the

Ethnogenesis from Within / 149

language of purity in the cooling white (funfun) deities of Oxala and his congeners. To be sure, the mulatto mixing of black and white is displaced onto the mestizo mixing of Indian and white in the unstable category of caboclo, which in Nagô temples embodies the essence of ritual pollution and includes Bantu admixtures as well. But such shameful taints can be overcome through healing rites of cleansing (limpeza) that send meddlesome egums and caboclos away (Matory 2005, 130–31) or by seeking connections with a purer house. As Matory explains, “Any given priest and his or her temple might abandon their original axé and seek affiliation with an older and more prestigious axé” (125), engaging in a form of ritual upward mobility that further benefits one’s initiate-offspring. What I am proposing as an extended hypothesis is that the logic of ritual purity in the great houses of Nagô Candomblé reworked the racial ideologies of blood and stratification in the casa grande of the colonial sugar estates, transposing principles of kinship and descent from social to ritual domains.32 To be sure, the diasporic returns and commercial interests of free black Bahians in the nineteenth century, and the codifications and ratifications of anthropologists in the 1930s, were crucial to the Nagôization of Candomblé, as has been demonstrated by Capone ([1999] 2010), Paul Johnson (2002), Matory (1999, 2005), and Parés (2004), and this helps explain how Nagô temples converted their ritual resources into social and political capital. But the idioms of royalism and genealogy that mediated and motivated these broader exchanges were generated by Yoruba scheme transpositions “from within.” It is not the cultural models of kingdom (ìlú) and house (ilé) as such that I have emphasized, but the modes of manifesting the former within the latter, a process that has brought the parallel developments of Lucumí and Nagô houses into clearer focus.

Throughout this chapter I have argued that an important strand of Yoruba ethnogenesis was fundamentally grounded in a cultural mode of quantification and boundary construction in West Africa and the Americas.33 For all the interest in how Yoruba identity developed from without, beginning in the mid-nineteenth century with Christian missionaries and diasporic returnees, the “making” of the Yoruba was deeply embedded in the generative semantics of sortal particularization (Verran 2001) and the transpositions of house and home through which idioms of genealogical descent emerged. In Yorubaland proper, I reversed received approaches to the building blocks of the social order by placing the town or kingdom as prior to and thus structurally immanent within the residential compound or home. Such a

150 / Chapter Five

counterintuitive approach to the town in the home—at least from the conventions of Western social science—was inspired by Barber’s exposition of oríkì orílẹ̀ , that form of praise-poetry celebrating lineage identities through towns of origin rather than eponymous ancestors (1991, 135–53). Moreover, once grasped as kingdoms in microcosm, Yoruba homes and their lineage dynamics can be seen as instantiating a range of organizational modes, including those made manifest through the very performances of oríkì themselves, rather than as primary principles of a specific form and type—residential versus genealogical, agnatic versus cognatic, corporate versus symbolic, which the lineage debates could never pin down.34 When recast as emergent sortal determinations rather than primary structural types, Yoruba descent and its residential dimensions become manifestations of wholes in parts. If in Yorubaland we saw how orisha worship effects such transpositions through collective rites of renewal, the same operational logic also was discerned at the core of Cuban Santería and Brazilian Candomblé. In each of these contexts, cult houses manifested prior kingdoms of origin that repossessed towns during public festivals, as cabildos de nación and irmandades that were ritually remade into neo-African monarchies. Within the emerging fields of socio-religious production in colonial Cuba and Brazil, the Lucumí and Nagô ethnonyms came to dominate along two axes of ritual accommodation and resistance: one lateral, assimilating non-Yoruba ethnic nations within their expansive brotherhoods; the other vertical, filtering the iconography of European royalism through Yoruba grammars of kingship and sovereignty. As the Lucumí and Nagô nations expanded, largely through religious avenues after the 1870s and 1880s, their social capital increasingly accrued through ties of ritual rather than ethnic kinship. The capitalization of ethnic entrepreneurship that followed in Cuba and Brazil, and the consolidation of a Yoruba ethnicity in Nigeria that it energized, belong to the longer vector of Yoruba ethnogenesis in transatlantic perspective (Matory 2005, 38–148). My goal has been to identify a critical cultural modality—a distinctive semantics of quantification—that rendered these developments both possible and intelligible, while also bringing novel historical interpretations to light.35 In Cuba, the transition from cabildo de nación to casa de ocha was a continuous development of town into home— largely in response to state persecution—as ritual ramages and lines of consecration remade neo-African monarchies in genealogical terms. In Brazil, the emergence of Candomblé houses from black brotherhoods and sisterhoods resonated with the historic casa grande, transposing plantocratic fears of racial mixing into obsessive concerns with ritual purity. Thus when Matory (2005, 115) asks, “Why is it that Brazilian Nagô Candomblé and the Cuban

Ethnogenesis from Within / 151

Lucumí Regla de Ocha pursue ritual objectives of ‘purity’ and ‘cleansing’ that are virtually absent from the cognate Nigerian òrìṣ à religions that are typically regarded as their origins?” we can point to the ritual reworking of color stratification within the New World cults. That the considerable literature on Nagô purification has overlooked this specific transposition of racial and ritual bloodlines attests to the insights afforded by Yoruba modes of demarcation and boundary formation in the plantation societies where neo-African identities initially emerged.36 If such motivating logics are difficult to perceive from without, they provide new perspectives on the diasporic trajectories of Yoruba ethnogenesis from within.

A F T E R WO R D

Beyond the Mirror of Narcissus

The preceding explorations of the “Yoruba-Atlantic” reveal systematic patterns and historical trajectories that resonate with my Yoruba research in Nigeria, highlighting revisionary strategies and regenerative schemes that—I have argued—are grounded in Yoruba culture. Against the seasoned trend of the last quarter century “decommissioning” if not deconstructing African origins in the Black Atlantic, I have sought something of a restoration—not of a cultural ancien régime harking back to bygone privileges, but of a critically reformulated culture concept, in this case distinctively Yoruba, which designates something real, somewhat knowable, eminently historical, and even indispensable for certain kinds of Atlantic research. I have proceeded by selective illustration, revisiting classic topoi in Afro-American studies such as Herskovits’s syncretic paradigm, the Petwo paradox in Haitian Vodou, the historical conditions of orisha cult clustering, remappings of gender in plantation societies, and the rise of Lucumí and Nagô houses in Cuba and Brazil, in each case offering new interpretations based on cognate dynamics in Yorubaland. In the beginning of this odyssey, my goal was merely corrective. It was clear to me from my Nigerian research that many of the foundational studies of Yoruba-related religions in the Americas were based on a false comparison with a fixed hierarchy of deities and a clear one-godper-cult correspondence that never actually existed in Yorubaland, but which served as a foil for patterns of New World fragmentation and indeterminacy that remain deeply ingrained within African diaspora narratives. Rather, such patterns of fusion and fragmentation, I have argued, are fundamental characteristics of orisha cults and pantheons mediating the dialectics of political competition both within and between historical kingdoms. As my subsequent Atlantic forays have revealed, the very deconstructive and reconstructive strategies at the core of Yoruba ritual reproduction and cosmological

Beyond the Mirror of Narcissus / 153

renewal remade specific spaces of death and transfiguration in the Americas into habitable worlds of Afro-Creole agency. And as this cultural perspective solidified, my goals shifted from corrective to prescriptive. Which draws attention to an ironic twist in the current state of AfroAtlantic historical anthropology. If the first round of “New World Negro” scholarship got the African baseline cultures “wrong” by reifying fixed tablets of tradition from Senegambia and the Bights of Benin and Biafra down to the ports of Luanda and Benguela, the current decomposition of African origins into various inventive rhetorics and registers consigns Africa to the dustbin of speculative mythmaking—unless of course we turn our attention to the history of the myths themselves.1 But even a compelling history of mystification requires comparative touchstones of the real, whether applied to the myths of nationalizing elites or those of subaltern ritual practitioners. In clearing the path for my culturalist restoration, I would like to identify an unholy trinity of “isms” that have troubled the Afro-Atlantic waters but need not undermine all scholarly investigations into African cultural legacies in the Americas. The first, that of essentialism, is indeed a problem when understood in its Marxian sense of transforming social and political relations of inequality into natural relations of the given order—through such invocations as the sloth and promiscuity of the laboring classes, the limited capacities of women and minorities, or the innate superiority of the super-rich. Clearly this logic of naturalization is a powerful mechanism of ideological mystification, pervading popular and social science discourses alike, and needs to be exposed for what it is. Within Africanist anthropology, as I have argued elsewhere (Apter 1999), the culture concept has a highly problematic genealogy, descending from racially marked notions of tribe and tribal cultures that essentialized the inequalities of the civilizing mission and, under the British, shaped the administrative framework of indirect rule. But does the colonial invention of tribe in Africa, so cogently exposed by Terence Ranger (1983), vitiate the very idea of African cultures? I think not, but one has to proceed critically in reading the colonial codifications of tradition against the grain, separating the flexible principles of kinship, gender, law, politics, and above all, group affiliation, from the ideological fixtures of tribal custom. In the Yoruba case, there is legendary variation between subcultural groups, dialects, patterns of descent, artistic styles, political organization, culinary preferences, urban settlement patterns, and ritual performance genres, as well as multiple affiliations and flexible boundaries that not only defy rigid codification but also represent a dynamic field of cultural principles and practices with a distinguished precolonial history. What unifies these related variations under a rubric that has come to be called “Yoruba”?

154 / Afterword

Here I have identified a number of “deep” cultural schematisms or structuring structures, multiply transposable in both collective histories and bodily practices, that are in some sense essential to Yoruba culture, underlying its regional variations, with identifiable modalities in the Yoruba-Atlantic. They are not reifications of European hegemony but have resisted, appropriated, and transformed that hegemony.2 The second problematic “ism” associated with the culture concept in Africa, that of ahistoricism, portrays worlds of fixed custom and timeless tradition, typically narrated in the ethnographic present, that perdure through the ages and lack historical consciousness beyond a few ancestral generations. If such a characterization borders on caricature, it still echoes throughout the halls of academe, where African cultures inhibit innovation, resist change, reveal the wellsprings of primordial violence, or exhibit the virtues of noble savagery. It is now an anthropological truism that cultures— for those who still accept the term—not only have histories, but shape them as well, but the methodological challenge remains for any cultural argument how to specify the vectors of determination, or the conditions under which cultural orders become historically salient.3 How do we position Yoruba culture as a shaper of change without removing it from history as its prior condition? Or at the other methodological extreme, how do we locate it within the flux of change without dissolving into a Heraclitan river such that we can never step into the same culture twice? Here I have taken a dialectical approach that acknowledges the relative autonomy of structuring structures while acknowledging their historicity and susceptibility to change. The structuring structures highlighted throughout these chapters are revisionary and regenerative schemes, incorporating prior transformations within their reproductive logics, thereby manifesting history while shaping it.4 We have seen that the Yoruba cultural resources brought to the Americas were not merely symbolic but regenerated black communities and sovereignties under abject conditions of exploitation and social reproduction. But in what sense are these “Yoruba” cultural resources? Here we encounter the perils of anachronism, in retrojecting recent data from Nigeria and the Republic of Benin into the eighteenth- and nineteenth-century Americas, and by calling the resulting correspondences and trajectories by a term that only gained panethnic currency at the beginning of the twentieth century. Like its sibling “isms,” anachronism is another symptom of colonial ethnography, not only presenting functioning “tribal cultures” as relics of a pristine past, but also denying coevality by consigning the primitive to the “there and then” in contrast to modernity’s “here and now” (Fabian 1983, 32, 33, 37– 38).5 Clearly anachronism carries a lot of ideological baggage. Its ghosts are

Beyond the Mirror of Narcissus / 155

particularly drawn toward portrayals of African rituals and religions as timehonored traditions that preceded Islam and Christianity, offering privileged windows into authentic cultures and cosmologies. To avoid these pitfalls, recent Yoruba research has abandoned such pursuits, shifting attention away from culture as such toward what Peel (1989, 2000) has termed the cultural work of Yoruba ethnogenesis, exploring the historical conditions in which “the Yoruba” emerged rather than the cultural conditions through which that history was shaped (see also Matory 1999, 2005). Yet even within this progressive development, something extremely important slips away; what we might call the formative powers of Yoruba culture, located not just in ritual templates, or manifestations of àṣ ẹ , but also (as I argued in chapter 5) in the semantics of quantification and its primordial principles of boundary formation. Here we can navigate the errors of anachronism, and even identify “Yoruba” cultural determinations, by bearing in mind the relative distinctions between terms, concepts, and their referential objects, or denotata.6 Terms (names, nouns, designators, signifiers) and concepts (meanings, ideas, intensions, signifieds) are not commensurate, but they generally work together. Since Saussure ([1916] 1966) we have known that their relation is arbitrary, fixed by convention rather than nature, and necessary for meaningful communication. If the term “Yoruba” emerged in the late nineteenth century and developed further in the twentieth to designate shared characteristics of a semibounded “people,” it hardly conjured them into being ex nihilo but organized them within a nominal class or category. To be sure, the recognized “nations” of the Anagos, Oyos, Ijeshas, Ijebus, and so forth were not yet “Yoruba” during the seventeenth, eighteenth, and most of the nineteenth centuries, which is why we find no such ethnonym among enslaved and manumitted Africans in the Americas. But their shared, reconstituted, and reconstituting cultural attributes were no less related than the mutually comprehensible dialects that they spoke, serving as an “umbrella of language” (Lovejoy 2004, 41; see also Eltis 2004). The point is basic but important. To call such eighteenth-century cultural and linguistic commonalities Yoruba is technically anachronistic because as an inclusive term it developed much later. But it is not a big deal if we acknowledge from the start that used analytically, it designates a cluster of intensions (meanings) and spatiotemporal extensions (geographic and historical) that give it some measure of objectivity. Moreover, real problems arise when we conflate these distinctions, focusing so exclusively on Yoruba ethnogenesis and the dialogical work of codifying “Yoruba” culture that we end up confusing culture with ethnicity; that is, with the self-consciously motivated politics of naming. This is not to deny the dialectical feedback loops between the

156 / Afterword

codifiers of culture (be they cultural nationalists, British colonial officers, or motley crews of anthropologists) and the culture they codified, nor the necessity of taking them seriously. But it bears worth repeating that what we call culture—its intensional meanings and extensional horizons—is not fully reducible to modern ethnogenesis, even when culture and ethnic consciousness become intimately entangled.7 I have focused on the Yoruba-Atlantic not because the “Yoruba” had the greatest influence in the Americas, but because their cultural history is paradigmatic of the kinds of Atlantic trajectories that bring Africa back in. The metadiscursive strategies of decomposing African origins are revealing and important, but there is more to the idea of Africa in the Americas than a history of circulating representations and positional rivalries within emergent fields of religious, political, and intellectual capital. If we take Africanist ethnography seriously and read against the textual grain, we can catch glimpses of the “there” both then and now.

NOTES

INTRODUCTION

1.

2.

3.

4.

5.

6.

7.

See Clapperton and Lander (1829), Frobenius ([1913] 1968); and S. Johnson (1921). For recent art historical studies focused on Ife, each with its particular virtues, see Abiodun (2014) and Blier (2015). For religious studies, see Olupona (2011); for histories, Law (1977) on Oyo and Akintoye (2010) on Ife. See Smith (1963), a life history of Grenadian prophet Norman Paul, originally serialized in four parts, appearing in Caribbean Quarterly, 1957–60. Smith’s field notes for two other studies of Carriacou and Grenada (Smith 1962, 1965) are full of ethnographic descriptions of Shango worship and Spiritual Baptists, little of which made it into his monographs. I thank Mary F. Smith for giving me copies of his field notes in 1995. Gilroy (1993, 4) was the first to characterize the Black Atlantic as the “rhizomatic, fractal structure of [a] transcultural, international formation,” which he opposed to the naturalized (and thus racialized) ethno-national categories that dominated African diaspora scholarship—hence his shift from rooted to “rhizomorphic, routed, diasporic cultures” (28). Mishra (2006, 82) writes that “by 1994, the non-teleological, rhizomatic, fractal approach to diaspora criticism had clearly become the ascendant framework.” For a Fanonian synthesis of black diasporic political consciousness with Third Cinema and the Lacanian imaginary, see Hall (1989). See also Pérez (2016) and Hayes (2011), respectively, for more micro-ethnographic foci on food (Chicago) and sexuality (Rio de Janeiro) in the “Black Atlantic”; P. Johnson (2014) and Parés and Sansi (2011) for focusing on the historical modalities of fetishism in Afro-Atlantic religions; Guedj (2012) for Akan ritual repertoires of black nationalism in the United States; and Castor (2013, 2017) for multicultural registers of citizenship in Trinidad Orisha. The orisha festivals had by no means died out in Ayede, but recruitment was declining, and relations between evangelicals and traditionalists were tense, particularly as children pressured their aging parents to leave the cults. See particularly Kant ([1787] 1973, 137), where he characterizes the a priori transcendental condition of the object, and thus of objectivity in general, as “the transcendental object = x.” For a classic statement of Weber’s neo-Kantian approach to historical objectivity via the ideal type, see Weber ([1904] 1949).

158 / Notes to Pages 7–17 8.

9.

10.

11.

12.

13.

14.

For a useful survey of this emerging literature, see Kohn (2015). For a clear methodological statement of the ontological turn, see Henare, Holbraad, and Wastell (2007). For an application of this method to Ifa divination in Cuba, see Holbraad (2012). For analyses of enregisterment and the reflexive models of language (registers) that it deploys, see Agha (2003, 2005). For the concept of indexical order from which enregisterment derives its sociolinguistic force, see Silverstein ([1996] 2003). In fairness to Palmié, Quine, whose ontological definition I invoked earlier, would agree. If as a scientist (or social scientist) Quine might valorize cultures over unicorns, his ontological relativity (“to be is to be the value of a bounded variable”) would recognize that in principle their difference is one of degree rather than kind, and that it is not inconceivable that a future (and more adequate) theory of the world would privilege unicorns (or Pegasus, or Homeric gods) over cultures, as the objectivity of both are “cultural posits” (Quine 1951, 41). Thus, I invoke the ontological priority of culture over unicorns as a social scientist, not as a philosopher. For a more general discussion of entextualization in relation to contextualization and the production of culture, both in interactional real time and by anthropologists producing and analyzing texts, see Silverstein and Urban (1996). For a more technical discussion of entextualization in its various (denotational, metasemantic, and metapragmatic) modalities, see Silverstein (2003). These databases have come a long way since Curtin’s Census, both in expanding and aligning categories of data, improving our understanding of the slave trade voyages and extending our purview into postabolition legacies. See, for example, the Emory University website “Voyages: The Trans-Atlantic Slave Trade Database,” with its associated African Origins Project, at www.slavevoyages.org; Walter Hawthorne and Gwendolyn Hall, “Slave Biographies: The Atlantic Database Network,” at http:// history.msu.edu/research/projects/data-archives/slave-biographies/; and Catherine Hall, “Legacies of British Slave-Ownership,” at www.ucl.ac.uk/lbs/. For Jakobson’s classic 1956 article on syntagmatic and paradigmatic axes in language, see his “Two Aspects of Language and Two Types of Aphasic Disturbances,” reprinted in Jakobson (1971, 239–59). For his famous elaboration of “poetic function” as the transposition of one axis onto the other, see his “Linguistics and Poetics,” in Sebeok (1960, 350–77). During my interviews and oral histories of the crowns, often hidden, in Ayede’s subordinate kingdoms, and in autonomous neighboring kingdoms, the stock answer to my questions about a crown’s provenance was “Lat’Ilé-Ifè l’ó ti wá,” or “It came from Ile-Ife,” even if the migration routes through which it traveled were complex. My exposure to Santería shrines in Union City is thanks to David H. Brown, who invited me along during several of his dissertation research trips in 1986. For a rich conceptual and ethnographic exploration of such Atlantic relocalization within the Orisha tradition, see Argyriadis and Capone (2011). CHAPTER ONE

1.

2.

The term “New World,” which denotes the post-Columbus Americas, is full of ideological problems of possession and temporality. I retain the term uneasily for the sake of historiographic continuity, with the qualification that invisible quotes surround each of my usages to bracket its pejorative connotations. See, for example, Du Bois 1939; C. Woodson 1936; Price-Mars 1983; Ortiz 1916; Hurston 1938; Ribeiro 1952; Ramos 1937; and Freyre 1956. For the conceptually

Notes to Pages 18–39 / 159 foundational writings of earlier black intellectuals outside the mainstream academy, and thus largely erased from the diasporic canon, see Kelley (1999, 2000). 3. The Nigerian fieldwork on which this argument is based took place from October 1982 to December 1984 and during three months of summer 1990. 1 gratefully acknowledge funding from Fulbright-Hays, the Social Science Research Council, the American Council of Learned Societies, and the American Philosophical Society. 4. See, for example, S. Barnes 1989; D. Brown 1989; Murphy 1988; and Thompson 1984. Of these, only Murphy (1988, 120–24) discusses syncretism. “Syncretism” has been critically reappraised, however, since the original 1991 publication of this chapter as an essay, both as a condition of cultural identity and alterity more generally (Amselle 1998), a space of complex religious negotiations (Stewart and Shaw 1994), and in relation to the re-Africanization of Afro-Atlantic religions (Capone 2001, 2005, 2008, 2014, 2016b; Palmié 1995). For a retheorization of syncretism in African prophetic Christianity, see Mary (2000). 5. These conferences have been held in such cities as Ife (Nigeria), Salvador (Brazil), San Francisco (California), São Paulo, Havana, Rio de Janeiro, Port of Spain, and New York City. They mark the self-conscious transnationalism of the òrìṣ à tradition as a world religion promoted by COMTOC (Conférences mondiales sur la tradition et la culture des Orisha). See Capone (2005, 279–97 and color plates 36–40). 6. For a sustained and rigorous critique of the rhetoric and ideology of Africanist discourse, see Mudimbe (1988). 7. I use the term syncretic paradigm to identify the larger model (and its additional concepts) within which the more specific meaning of syncretism proper is located. 8. Fernandez (1990, 150–51) recounts Herskovits’s affiliation with the NAACP after expressing an initial reluctance. 9. For a glimpse of the ideological conflict that Herskovits experienced with the Carnegie Corporation, as well as the corporation’s colonial epistemology, see Jackson (1986a; 1986b, 117–18). 10. The “Fon” (also called Dahomeans by Herskovits) and the “Yoruba” are missionarycolonial ethnic designations that emerged in the nineteenth century to refer to peoples of what is today the southern half of the Republic of Benin and southwest Nigeria. The infamous slave port of embarkation was at Ouidah, controlled for a long time by the Portuguese. 11. Thus in Ayede, the Yemoja cult houses the additional deities Orisha Oko, Shango, Ogun, Oshun, Oya, and Olokun. 12. For a more detailed version of this story, see Métraux ([1959] 1972, 42–43). 13. In her discussion of Tshidi Zionists, Jean Comaroff notes how the intent “to deconstruct existing syntagmatic chains, to disrupt paradigmatic associations, and, therefore, to undermine the very coherence of the system they contest” inevitably reproduces, on a formal level, aspects of the symbolic order that it reconfigures, so that “subversive bricolages always perpetuate as they change” (Comaroff 1985, 198). It is her association of syncretistic movements with subversive bricolages that I am calling “critical practice.” C H A P T E R T WO

1.

I use the term Afro-American studies inclusively, as an initial academic designation that later developed into rubrics such as Black Studies, Africana Studies, African American Studies, African Diaspora Studies, Afro-Atlantic Studies, and Black Atlantic Studies.

160 / Notes to Pages 41–52 2.

See Bascom (1944) for the first sustained attempt to identify the extralineage bases of orisha cults. 3. Pemberton (1977, 12–14) has noted multiple orisha represented on the altars of town shrines in Ila Orangun, such as Oshun, Osanyin, and Shopona under Obatala but has neglected the structural bases of such clusters. Barber describes the orisha as “at once fragmented and fused” (1981, 736) but attributes this property to “the intense personal nature of the òrìṣ à-devotee relationship” as modeled on the politics of Big Manship. See also McKenzie (1997, 477–87), and chapter 3 in this volume. 4. See Barber (1981, 738–40) and Apter (1998) for discussions of Yoruba secrecy as a critical ritual resource and Buckley (1976, 1985) for an extended discussion of secrecy within a dialectic of revelation and concealment in Yoruba medicine. 5. When applied to the Dogon case, as in Apter (2005a), this revisionary approach to deep knowledge vitiates Van Beek’s (1991) critique of Griaule because the content of the esoterica is contextually and historically specific. 6. The fundamentos or fundamentals of Cuban Santería, for example, refer to the blood, stones, and herbs of ritual sacrifice and investiture that embody the concentrated essence or vital power of the deity. See Bascom (1950). 7. One of the earliest written accounts of the Bois Caïman sacrifice is Dalmas (1814, 117), based on the testimony of Ignace, a slave from one of the burned plantations of Gallifet, who was captured and tried before the court of Cap-Français (Laguerre 1989, 61). Whether myth, history, or most likely some combination of both, the event was enshrined by Price-Mars (1928) and became something of a charter for noirisme. See also Fick (1990, 92–94, 104–5). 8. Karen Brown (1991, 100–101) associates the Rada and Petwo distinction with “two archetypal social groups: family members and foreigners, insiders and outsiders, the oppressed and their oppressors,” an interesting position that works to a limited extent, because Petwo are outside official channels and are worshipped outside the family hounfor (temple), but Brown’s position cannot account for the revolutionary spirits among the Petwo. She further posits Ogou as a mediate lwa, a position I criticize while engaging de Heusch (1995, 113). 9. According to John Goldsmith, the phonological shift from /di/ to /ti/ in Romance and Bantu languages alike is a virtual impossibility (personal communication, November 6, 2000). 10. More recently, Geggus (2014, 24–25) provides evidence that Dom (not “Don”) Pedro was likely a Kongolese maroon so-named by former Portuguese masters, and that Moreau de Saint-Méry’s attribution of his Spanish provenance was mistaken. 11. Dunham ([1969] 1994, 27) refers to this highest stage of initiation as “prix-des-yeux,” parallel to “prix-de-cloche” in Congo Vodou. She provides the gloss “‘prize-of-eyes’ or ‘price-of-eyes’; clairvoyance” (279). 12. Métraux ([1959] 1972, 68) reports that priests who lack training and claim knowledge directly from the lwa are known as “hungan-macoutte,” a disparaging term that curiously resonates with the sorcery of those Bizongo priests who became tontonmakouts under François Duvalier (discussed later in this chapter), suggesting the illegitimacy of power outside of authority. For the parallel development of both modes of acquiring ritual knowledge in Brazilian Candomblé—by inspired gift as opposed to formal initiation, associated with magic versus religion—see Sansi (2007, 22–23), building on Boyer (1996). 13. The same principle of inheritance and associated forms of segmentation and fission

Notes to Pages 52–64 / 161

14. 15.

16.

17. 18.

19.

20.

21.

22. 23.

24.

are found in Yorubaland, where individually acquired land devolves per stirpes, among children of cowives (Lloyd 1962, 296–300). For an intensive analysis of such developmental cycles and the branches generated by segmentation, see Larose (1978). Karen Brown’s (1991) rich account of Vodou based on the priestess Mama Lola reveals “a matrilineal family tree” establishing a “parallel kinship structure buried beneath the official versions” (16–17), one that may well reflect the de facto importance of property-owning matri-segments. See Schwab (1955) for a Yoruba parallel, also discussed in chapter 4. Laguerre (1989, 45–47, 72, 79–81) locates the birth of secret societies such as Bizongo with the rise of colonial maroon communities and groups, some of which were nomadic, raising the interesting possibility that the sorcery of dispossession and landlessness was historically associated with nomadic maroons. He also argues that after the revolution peasants joined Bizongo societies to protect those lands initially allocated by Dessalines’s agrarian reform (1804–6) from reappropriation by the black and mulatto elites, framing the authority of the lakou externally, in relation to an extractive class. If true, this latter argument complicates the opposition between Ginen and sorcery within the lakou by relating the tensions generated by its developmental cycle to the incursions of this extractive class. According to Laguerre, Bizongo “belongs to the family of strong spirits, born in time immemorial in Africa” (74), thereby assuming the character of Ginen with respect to authority, a view very much at odds with Larose (1977) but possibly explained by such externally protective functions. For an extension of this model throughout the transnational circuits of Haitian migrant workers in the United States, see Richman (2005, 2014). Laguerre (1989) witnessed the return of François Duvalier after his death as “loa 22,” dressed “in a dark suit and black hat, wearing heavy reading glasses and holding a pistol in his right hand,” so named to commemorate Duvalier’s lucky number, 22 (118). See also P. Johnson (2006). Laguerre (1989, 101–20) usefully distinguishes between the political support Duvalier gained from Vodou temple networks in his electoral victory and his use of Vodou secret societies (such as Bizongo) to extend his network of secret police and thus strengthen his authoritarian control. Many scholars, including Laguerre (1980; 1989, 74), derive Bizongo from the spirit protector of Bissagot slaves. I believe that such morphologically tenuous derivations are overly speculative when unsupported by additional evidence. As Hurston ([1938] 1990) explains, “There is a long list of [Petwo] spirits who have the same names as the Rada gods except that the second name distinguishes them from the Rada. ‘Ge-rouge’ after a name places that god in the Petros or the Congos” (167). Rada names not only function as root morphemes but also refer to spirits that Haitians call “root lwa” (lwa rasin), thus underscoring their marked Petwo forms (Karen Brown 1991, 100). In this respect, the relation of Ginen to Petwo follows Dumont’s (1970) concept of hierarchy as “the encompassing of the contrary” (240). See Bastide for a similar division of the nations in Brazil “for the purpose of fomenting interethnic rivalries and so preventing the development of class consciousness and a general revolt of the blacks against the whites” ([1960] 1978, 61). Corroborative evidence of this status hierarchy within the slave community comes from Debien (1974, 50), who cites similar passages from Degrandpré (1801, 75)

162 / Notes to Pages 64–69

25. 26.

27.

28.

29.

and Malenfant (1814). Larose claims, however, that “Radas were considered the most intelligent group: hard workers they were said to learn easily and be good domestics, which quite agrees with their present status as ‘white’ spirits” (1977, 102). I would suggest that Larose is confusing their position in the slave community with their status in Vodou (see the following discussion). Blier (1995a, 20) privileges a psychological over a political approach to the power of secrecy, although many of her readings support the deep-knowledge paradigm. In her superb ethnography, Rosenthal (1998) shows how the Ewe are centrally placed within this regional cultural field, with a form of Vodou (Gorovodu) associated with Banguele slave spirits that are clearly related to the Haitian Petwo. Again, it is not the direct link I am looking for, but a translocal and transethnic regional framework. In this respect, it is more significant that Ewe, Fon, and Yoruba share the same complex of ritual schemas and deities that characterize the cultural history of the slave coast. See also Blier (1995b). For an extension (or, perhaps, retraction) of this paradigm into the body, in terms of the hot and cool valences of circulating blood associated with vital disequilibrium and equilibrium, see Brodwin (1996, 8–96), Buckley (1985), and Laguerre (1987). We can explore these associations within a cultural economy of the body. Lefebvre (1986, 1993) and Singler (1993) argue that Haitian Creole has a Fon grammatical substrate, a position opposed by Chaudenson (1990) in his “superstrate thesis.” See Mufwene (1990, 1993) for a clarification of these debates. For a different structural model derived from creolistics, see Drummond (1980). Drummond applies Bickerton’s (1975) concept of an “intersystem” to capture the dynamic patterning of ethnic designations in Guyana. The relevance of creolistics to models of cultural creolization in the Caribbean is of great theoretical and comparative significance, notwithstanding the perils of simplistic analogizing (Palmié 2006, 443–47). For an illuminating linguistic application to “spirit migration” in Asogwe Vodou in Haiti, see Hebblethwaite (2015). European imports from muskets and gin to cloth, coins and even crowns have been incorporated into Yoruba ritual and royal regalia since at least the early nineteenth century, as Hugh Clapperton’s first visit to Old Oyo in 1826 revealed (Clapperton and Lander 1829, 37–60). The Benin bronzes and carved elephant tusks are famous for incorporating Portuguese powers into the heads of kings, and at least one carved royal leopard (now in the British Museum) is spotted with European cartridge shells. For nineteenth-century accounts of gift giving and ceremonial exchange between European traders and explorers and African chiefs along the Niger Delta coast and up the Niger River, see Allen (1848), Baikie (1856), Jones (1963), Laird and Oldfield (1837), Lander and Lander (1832), and Mockler-Ferryman (1892). For creolized ritual forms and associations developed by freed slaves settling in Freetown, Sierra Leone, see Nunley (1987). CHAPTER THREE

1. 2.

See, for example, Verger (1957, 1968, 1982). Fieldwork on orisha worship in Nigeria from October 1982 to December 1984 was funded by fellowships from Fulbright-Hays and the Social Science Research Council. Subsequent visits from August to October 1990 and June to December 1993 were funded by the American Philosophical Society, the American Council of Learned Societies, and a Fulbright-CIES fellowship. Support for this research is gratefully acknowledged.

Notes to Pages 69–98 / 163 3.

See, for example, Abimbola (1976), Agiri (1975), Ajayi (1974), Barber (1991), Beier (1959), Belasco (1980), S. Johnson (1921), Matory (1994). 4. See, for example, Bascom (1944), Drewal (1992), Drewal and Drewal (1983), Ogunba (1967), Olupona (1991, 2011). 5. For example, Awolalu (1979), Idowu (1962), Thompson ([1971] 1976, 1984). 6. Important studies of Yoruba kinship and descent include Bender (1970), Eades (1980, 37–64), Fadipe (1970, 97–146), Lloyd (1955a; 1962, 279–307; 1966), and Schwab (1955, 1958). 7. A. C. C. Swayne, “Intelligence Report. Ayede District of Ekiti Division,” 1936, p. 3, Ondo Province, Nigerian National Archives (Ibadan). 8. Although civil chiefs are barred from high ritual office, military chiefs in the past were not. At Ijaye in the mid-nineteenth century, the Àrẹ -ọ̀ nà-kakan`fò “bolstered up his power not only by judicious feasting of the masses every fifth day but also by usurping the headship of the different cults, particularly that of Shango” (Ajayi 1964, 67). 9. Interview with chief Ọ baísà, December 18, 1983, Ishan-Ekiti. 10. Interview with chief Ọ baísà, December 18, 1983, Ishan-Ekiti. 11. Elaborate age sets and rotating title systems are generally identified as Bini influences among the Ekiti Yoruba (Forde 1951, 15). 12. See Smith (1956) for this pathbreaking political analysis of lineage segmentation within a general theory of government. 13. “The difference between the ceremonies for orisha in Africa and the New World stems mainly from the fact that in Africa, in principle, only one god is evoked during a festival celebrated in a temple reserved for him alone, whereas on the New World, a very wide range of orisha are called in the same terreiro during the same festival. Moreover, in Africa, such a ceremony is generally celebrated for the family community and a single elégun [possession priest or priestess] is normally mounted by the god, whereas in the New World, where this family community no longer exists, the orisha has taken on an individual character and so it happens that during the same festival, several iaôs [ìyàwó, or “wives” (priestesses) of the orisha] are mounted by one and the same god, to the particular benefit of a specific individual and the additional benefit of all those who worship this orisha” (my translation). CHAPTER FOUR

1.

2.

3.

See Herskovits ([1941] 1958, 1966), Bastide ([1960] 1978, 1971), Mintz and Price ([1976] 1992), and Gilroy (1993) for the basic paradigm shift from roots to routes. For more recent Atlantic case studies, see Capone (2005), Clarke (2004), P. Johnson (2007), Matory (2005), and Otero (2010). For a precise methodological reformulation, see Palmié (2010). See, e.g., Beckles (1989, 2000), Berry (2007), Kathleen Brown (1996), Bush (1990), Campbell, Miers, and Miller (2008), Fox-Genovese (1988), Gaspar and Hine (1996), Gautier (1985), Mintz (1983), Moitt (2001), Jennifer Morgan (2004), Morrissey (1989), and Stevenson (1996b). Bush (1990), Jennifer Morgan (2004, 50–68), and Robertson (1996) are among the few studies that deploy sociocultural material on West African womanhood and agency to illuminate gender dynamics in plantation societies of the Americas. For gendered dimensions of African American conjuring and midwifery, see Fett (2002). It could be argued that the English term “witchcraft” itself belongs to the European ideology I am trying to transcend. African languages, however, have terms for malign female doubles who consume children, as in the Yoruba term àjẹ́ —first recorded

164 / Notes to Pages 100–109

4.

5.

6. 7. 8. 9.

10. 11. 12. 13. 14. 15.

16.

17.

18. 19.

by Samuel Ajayi Crowther in 1843—which contracts ìyá jẹ , or “mother eats.” See Crowther (1843, 81), Prince (1961, 797). See Fadipe (1970, 65–146) for one of the best accounts of Yoruba kinship and family organization, published posthumously from his 1939 (University of London) PhD dissertation. The literature on Yoruba witchcraft is extensive. Key descriptions and discussions include Apter (1993), Drewal and Drewal (1983), Hallen and Sodipo (1997), Idowu (1970), Matory (1994), Morton-Williams (1956), and Prince (1961), whose original conjecture that àjẹ́ may be a contraction of ìyá jẹ has been cited by many others. For the application of the legal term per stirpes to Yoruba lineages segments of full siblings, see Lloyd (1962, 296–300). For intricate discussions of such characteristic patterns of lineage segmentation and fission, see Lloyd (1955a, 1966) and Schwab (1955). The concept of complimentary filiation was first developed by Fortes (1949); that of optation was first developed by J. Barnes (1962). For the basic organization and transactional dynamics of Yoruba markets, see Fadipe (1970, 159–63), Hodder (1969), Sudarkasa (1973), and Trager (1976). For the ritual representations of value and profit associated with market responses to Atlantic trade, see Belasco (1980). For a detailed analysis of such popular mobilization led by Yoruba women, see Apter (1987b). For popular depictions of the Yoruba Cash Madam, see Barber (2000, 271) and Ogunyemi (1996, 171). For Yoruba idioms and dramas of moneymaking magic, see Apter (1998, 84–86; 2005b, 248–49), Barber (1995), and Matory (1994, 124). For an ethnographic analysis of the secret dimensions of Yoruba witchcraft and ritual reproduction, see Apter (1991). See Jennifer Morgan (2004, 58–61), Nwokeji (2001), and Geggus (1989). Here I am following Deborah Gray White ([1985] 1999, 211, n. 65), who writes, “In researching links to the African past it is probably unwise to look for one-to-one correlations between African societies and black American society. It seems best to take the approach of anthropologists Sidney Mintz and Richard Price. They argue that the African-American’s adaption to Western cultural mores was governed by ‘unconscious “grammatical” principles,’ which provided the framework for the development of new institutions and served as a catalyst in their development.” Cited in Jennifer Morgan (2004, 41). Barbot’s manuscript was completed in 1688 but was published posthumously in English by the Churchill brothers in 1732. Although based on his travels from 1678 to 1682, it also drew upon other travel writings, resulting in an intertextual composite not unusual within the genre. For the definitive philological history of this text, see Hair, Jones, and Law (1992). For a philological analysis of its historiographical value, see Law (1982). The phrase “hot constitution’d Ladies,” broadly quoted, comes from William Smith, A New Voyage to Guinea (London, 1744, 51–52), who adds that they are “continually contriving stratagems how to gain a lover.” Quoted in D. White ([1985] 1999, 29). See William Smith, A New Voyage to Guinea (London, 1744), 142–43, quoted in Jennifer Morgan (2004, 46). But see also Perez (2010) for the religious opposition in Cuba between the black Virgen de Regla (Yemaja), invoked as “the world’s wet nurse,” and the sexualized mulatta Virgen de Caridad del Cobre (Ochun).

Notes to Pages 109–117 / 165 20. How often surrogate motherhood extended from mythic to actual wet-nursing is difficult to say, but Beckles (1996, 121) maintains that “elite white women in Barbados commonly preferred black nannies to nurse their children,” and provides a vivid vignette from a visitor in the 1790s who witnessed, with some discomfort, “a white child sucking . . . the long breasts of the slave.” See also Freyre ([1933] 1986), who discusses ideologies of race, breast milk, and the impact of black wet nurses (macama, mucamba) on white patriarchal sexual culture and psychology in Brazil; and WallaceSanders (2008). 21. Stevenson (2005, 355) writes, “The southern ‘slave woman’ was popularly thought of as an evil, manipulative temptress who used her insatiable sexual appetite for personal gain. She was seducer, adulteress, whore for hire, all wrapped into one—the bane of her mistress, the damnation of her master, or any man who fell under her spell,” adding that “the lore about the sexual prowess of slave women pervaded every rank of southern slave society.” 22. See, e.g., Bush (1990, 98), as well as Stevenson (1996a, 173–74), who argues that—in antebellum Virginia—a more African-derived code of motherhood in the fields emphasizing procreation and protection opposed the Victorian code of the mistress and her more assimilated slaves and servants in the Big House. 23. Euro-American institutions also could reinforce African-derived family structures. Blassingame (1979, 178) notes that Episcopalians strengthened extended families among slaves by requiring godparents for all children baptized, a trend presumably echoed throughout the Catholic plantation societies as well, although significant distinctions between ritual and social kinship should not be overlooked. See also Roberts (2004, 254). 24. The evidence comes from Beckford (1790, 2:323–24) and Long (1774). 25. For White’s reference, see John Morgan (1860); for Bush’s, see Dancer (1809). 26. The association of Arada ethnicity in colonial Saint Domingue with West African cultural practices was enshrined by the distinction between Rada (Arada) and Petwo (as in Creole or Congo) deities (lwa) in Vodou (Apter 2002). 27. “She was made to carry a stick of wood on her back, to punish and humiliate her with the simulacrum of a child” (my translation). 28. Evoking the most punitive antiwitchcraft sanctions, male and female slaves accused of poisoning whites and blacks alike were burned at the stake in the colonial South. See P. Morgan (1998). 29. Beckles (1989, 117–18), however, argues that polygyny remained the norm from 1627 to 1780 among enslaved families in Barbados, and that “it took many generations of Creolization and pressure from largely European Christian sources before black males were apparently reconciled to abandoning the practice.” The higher incidence of polygyny than is normally acknowledged in slave family scholarship raises the correlative hypothesis that witchcraft between cowives also was widespread. 30. These stereotypical labels are not to be taken literally, but rather as mythic projections of gendered relations of social domination at the intersection of household, caste, and class. 31. From the standpoint of white female hegemony, the black seductress or colored “coquine” would be seen as something of a blood-sucking witch, “polluting” the white patriline while converting its blood into money and property, a hypothesis worth testing with reference to contemporaneous white women’s writings. 32. According to Mintz and Hall ([1960] 1974, 187), the etymology of “polink” evokes a semantics of empowerment and emancipation: “Th[e] use of the term polink, which

166 / Notes to Pages 119–131 H. P. Jacobs (personal communication) believes to be related to the Spanish palenque (a palisade or palisaded village . . . ; later, a fortified runaway slave village, as in Cuba or Columbia), is interesting in itself, symbolizing the link between independent cultivation and the status of the slaves.” 33. Wood (1990, 329, n. 41) cites the Savannah Republican of August 26, 1817, for this passage. She also cites other diatribes by “anti-mulatto” in the same newspaper published on August 16 and 19, 1817; September 6, 20, and 27, 1817; and January 17, 1818. 34. See also Olwell (1996), who explores eighteenth-century discourses of immorality regarding enslaved market women in the Charleston marketplace and emphasizes the more general contradiction between patriarchy and the circumscribed “freedoms” of the market. 35. One would of course expect a greater correspondence between Yoruba idioms and plantation dynamics in the Iberian colonies of Brazil and Cuba, and in Spanish Trinidad, where large numbers of Yoruba slaves, particularly in the nineteenth century, had a more recognizable “Yoruba” influence, as in Candomblé, Santería, of Trinidad Shango. But it is precisely because the model is salient where it might not be expected—the US plantation south, Barbados, Jamaica, and less surprisingly, the French Antilles—that a case for a broader regional West African influence can be made. CHAPTER FIVE

1.

See also Lovejoy (2004), who argues that Ahmad Baba’s 1613 reference implies an even earlier popular usage among northern Muslims referring to their southern neighbors. He also challenges the idea that “Yoruba” first applied to Oyo, arguing that the term “suggests a country, not necessarily a political state” (41), and adds that Oyo was only a minor polity at the time. For more on Ahmad Baba and his West African ethnic nomenclature, see Lovejoy (2003, 12–15). 2. See also Law (2004), Lindsay (1994), and Verger (1968). 3. See Ojo (2009b) for a similar shift back to internal Yoruba ethnogenesis, which he attributes to the population displacements of the nineteenth-century wars; and Ojo (2009a) for the role of the orisha in consolidating Yoruba identity in the diaspora. See also Lovejoy and Ojo (2015). 4. See Lloyd (1955a, 1962, 1966, 1968) and Schwab (1955). 5. The reason given, quite typical of such cases, was that no qualified successor could be found in the male line. 6. See also Peel (1983), who notes, “In Lloyd’s case, politics is reduced to kinship, for the rules of kinship are treated as producing forms (i.e., the lineages) anterior to and determinative of politics” (10). 7. Eades’s argument confuses kinship, which is always bilateral and egocentric, with descent, which is ancestor-oriented and thus lineal. See Scheffler (1966). 8. Louis Henry Morgan was the first to identify “Hawaiian” kinship as the simplest type of classificatory (as opposed to descriptive) kinship system, merging relatives into a small number of categories irrespective of the father’s and mother’s “sides.” See Morgan (1871, 451–57). As a kinship type, the Hawaiian system refers to rules of classifying kin, not to ethnolinguistic provenance. 9. See Akinnaso (1980, 279) and Johnson (1921, 81) for more conventional interpretations of this proverb. 10. If Peel (1983, 50) likens “the household’s overall unity . . . to a small polity” in his discussion of domestic economic roles, I am further arguing that it manifests the larger polity in which it is embedded.

Notes to Pages 132–144 / 167 11. 12.

13.

14. 15.

16.

17. 18. 19.

20. 21. 22.

23.

24.

25. 26.

27.

Thus the same greeting for entering the king’s palace, “Ẹ bọ ààfin,” is used while entering a shrine during its festival. I owe this insight to my research assistant Olu Ibitoye, who first pointed out that the Yemoja priestesses pronounce Ẹ ṣ ẹ un in an exaggerated fashion “to show they come from Ibadan.” See, for example, the role of cabildos in the Escalera and Ponte rebellions (Childs 2006; Paquette 1998). See also H. Lovejoy (2012). On the “ambiguous role” of cabildos in both accommodating and resisting the absolutist colonial polity with their own limited sovereignty in El Cobre, see Díaz (2000, 262–84). The cabildos de nación were expelled from Havana to the barrios extramuros in 1792 (D. Brown 2003, 28). When Hugh Clapperton and his servant John Lander made the first European visit to the Alaafin of Oyo in 1826, the Alaafin was wearing “the imitation of an European crown of blue cotton covered over pasteboard, made apparently by some European and sent up to him from the coast” (Clapperton and Lander 1829, 37). For a rigorous methodological elaboration of this coterminous perspective on transatlantic ritual entanglement, see Palmié (2010), who focuses on the Ekpe–Abakuá connection in relation to the Middle Passage. Circulating through networks of ritual exchange, kola nuts embody the generative value of reciprocity, hence the proverb “S/he who brings kola brings life.” These examples, many from López Valdés (1986), are listed in David Brown (2003, 316, n. 67, and 319, n. 19). This suggestion remains hypothetical, but it questions the standard assumption that the administrative structures shared by all black cabildos derived from European society and the Catholic Church. The obá-oriaté (king-master of the divination mat) of the regla de Ocha is distinguished from the obá (king) of the regla de Ifa, but both offices manifest kingship. For Ortiz’s cabildo derivation argument, see Ortiz ([1906] 1973, 1921), cited in David Brown (2003, 62). David Brown (2003, 210–86) discusses the “palace” of the Obá Lucumí as a mythic template of the kingdom or palace within the casa de ocha, one which resonated with the Spanish king’s representative in Cuba while invoking Yoruba kingship as “a diasporic chronotope” (211). Kiddy (2005, 94) discusses a registration book for a rosary brotherhood in Mariana from 1754, which listed the ethnicity of 85 percent of its members, of whom 62 percent were West African, mostly identified as Mina. On the referential ambiguities of the ethnonym Mina, see Gwendolyn Hall (2003) and Law (2005). For the variably Yoruba component of Mina, see Lovejoy (2004, 42); Reis and Mamigonian (2004), who discuss the “Mina-Nagô” outside Bahia; and Carvalho Soares (2004), who argues that “Mina” in Rio de Janeiro became increasingly Yoruba in the nineteenth century. See Heywood and Thornton (2007, 214–15) for Kongo kings already present in the Brotherhood of Our Lady of the Rosary during the 1620 Festival of Xavier in Luanda. For a more historically grounded account of the development of Nagô, Jeje, and Angola Candomblé nations as meta-ethnic denominations with shifting conceptual referents, see Parés (2008). For a discussion of the more recent “re-Bantuization” of the Candomblé Angola contra Yoruba hegemony, see Capone (2016b, 479–80). For a related discussion of spirit migration in Haitian Vodou, see Hebblethwaite (2015). See Capone ([1999] 2010), Paul Johnson (2002), Matory (2005), Parés (2004), Sansi

168 / Notes to Pages 145–149

28.

29.

30.

31.

32.

33.

(2007), and Palmié (2013) for discussions of this complicity and its impact on the politico-ritual capital of the cults. For studies of “impure” varieties of “syncretistic” houses and practitioners, see Hayes (2011), Wafer (1991), and the caboclo-oriented illuminations of Brazeal (2003, 2008, 2010, 2014). For cogent discussions of the “good” (“mosaic”) and “bad” (“fusional”) types of syncretism developed in Bastide’s broader oeuvre, establishing a foundational framework from which the re-Africanization of Candomblé houses (followed by a transnational form of ritual Pan-Africanism, with associated “Afro-African” and “Afro-Occidental” syncretic modalities) developed throughout the second half of the twentieth century, see Capone (2008, 2014). There is considerable inconsistency on the precise relationship between the woman known as Iyá Nassô and her “daughter,” who may have been her initiated godchild, but in this case the kinship idiom is more important than the filial status of the maternal bond. Paul Johnson (2002, 75–76) and Parés (2004, 193–94) discuss various accounts of this voyage from Bastide ([1960] 1986), Carneiro (1961), and Verger (1981). See Matory (2005, 115–48) for a broader discussion of the transnational dynamics and mercantile motives of the Candomblé purification movement in both ritual and ideology. Bastide’s placement of Nagô house slaves above Bantu field slaves within the plantocratic social hierarchy may represent his own Nagô bias as a Candomblé quasi-initiate rather than the historical ideology that was actually in play in colonial Brazil, where the more assimilated Creoles and Congos (Bantus) served as house slaves in opposition to enslaved Nagôs working in the fields. Like the Rada–Petwo ritual reversal in Haitian Vodou (discussed above in chapter 2), the Nagô Candomblés may well have reversed the contemporaneous social hierarchy within the cults by putting Nagôs on top. See also Capone (2000, 68–70), who locates Bastide’s evaluative Yoruba-Bantu hierarchy within the legacies of pseudoscientific evolutionism. For preoccupations with racial purity and white marriage within the planter class of colonial Bahia, see Schwartz (1985, 269–75). For an extended discussion of how the tensions of miscegenation and whitening played out in Brazilian abolitionism and the nationalisms that followed, see Skidmore ([1974] 1993). I would further suggest that the Candomblé casa transposes the caste ideology of the colonial casa grande into the characteristically female-headed slave household, ritually transforming the slave house into the Great House, and by implication, slaves into masters. Support for this hypothesis is indicated by the predominance of female leadership in the Candomblé houses, the combination of aristocratic and slave references in the devotees’ sartorial codes, and the historical fact, discussed in Parés (2010), that in the mid-nineteenth century manumitted Candomblé leaders purchased slaves to augment their followers. That the actual founders of Candomblé houses were free urban blacks rather than plantation house slaves in no way precludes their appropriation of casa grande domestic frameworks within the terreiros, particularly in light of the salience of the Big House in the Brazilian popular imagination (Freyre [1933] 1986). For the idea that Candomblé ritual purity and its African lineage was valorized against Brazilian anxieties of racial degeneration associated with “mixed and syncretistic cults” in the work of Arthur Ramos (1937) on a national scale, see Capone (2008, 269) and Sansi (2011, 35). To equate a mode of quantification such as sortal particularization with a “timeless cultural essence” is misleading because it is has no semantic content in and of itself, and hence no substantive meaning, but it establishes the formal ground of generating

Notes to Pages 150–156 / 169 substances and meanings. The extent to which the same or similar quantifying logic pervades related language and cultural groups in West Africa cannot be determined without the type of technical analyses that Verran (2001) has done for Yoruba enumeration. 34. For example, Barber (1991, 168–72) shows how oríkì orílẹ̀ of the mother demarcate cognatic descent lines vis-à-vis agnatic cores. 35. My focus on the logic of boundary construction from within is therefore the same kind of argument that Dumont (1970) makes for explaining caste in India, although he generalizes “holistic hierarchy” into the organizing principle of South Asian society. Furthermore, my use of sortal particularization illuminates why ritual purity within the orisha-based tradition emerged in Cuba and Brazil and not in Nigeria. 36. I should emphasize that this ritual transposition of the logic of whitening does not deny the impact of transnational discourses of Yoruba purity that emerged in opposition to European and North American racism in the late nineteenth century, as discussed by Matory (2005, 57–61), but helps to explain how such purity was shaped and incorporated within the dominant Candomblé temples. A F T E R WO R D

1. 2. 3. 4. 5. 6.

7.

See, e.g., Herskovits ([1941] 1958, 33–53) as opposed to Palmié (2013). For a concise exposition of such resistance in Bahia, see Reis (2011). The most explicit theorizations of this problem are found in Sahlins (1985) and Sewell (2005, 197–224). The foundation of this dialectical concept comes from Bourdieu’s (1977) theory of practice. In these passages, Fabian actually reworks anachronism into the more rigorous and systemic concept of “allochronism.” The basic epistemology I have in mind has its roots in Gottlob Frege’s ([1892] 1997) important distinction between “sense” and “reference,” first published in 1892, reworked by many analytical philosophers, and reformulated by Carnap (1947) into “intension” and “extension.” For an illuminating representation of the ongoing tension between these two interpretive polarities, see the “Scholarly Exchange” debate in The Americas between J. Lorand Matory (2015), who subjects Luis Nicolau Parés (2013) to an overwrought if principled anti-Herskovitsian critique, and the latter’s principled, more measured reply (Parés 2015).

REFERENCES

Abimbola, Wande. 1976. Ifá: An Exposition of Ifá Literary Corpus. Ibadan, Nigeria: Oxford University Press. Abiodun, Rowland. 2014. Yoruba Art and Language: Seeking the African in African Art. New York: Cambridge University Press. Afolabi, Niyi. 2009. Afro-Brazilians: Cultural Production in a Racial Democracy. Rochester, NY: University of Rochester Press. Agha, Asif. 2003. “The Social Life of Cultural Value.” Language and Communication 23, nos. 3–4: 231–73. ———. 2005. “Voice, Footing, Enregisterment.” Journal of Linguistic Anthropology 15, no. 1: 38–59. Agiri, B. A. 1975. “Early Oyo History.” History in Africa 2:1–16. Ajayi, J. F. Ade. 1960. “How Yoruba Was Reduced to Writing.” Odu 8:49–58. ———. 1964. “The Ijaye War, 1860–65.” In Yoruba Warfare in the Nineteenth Century, edited by J. F. A. Ajayi and R. S. Smith, 59–149. Cambridge: Cambridge University Press. ———. 1974. “The Aftermath of the Fall of Old Oyo.” In History of West Africa, edited by J. F. A. Ajayi and M. Crowder, 2:129–66. New York: Longmans. Ajayi, Jacob F. A., and Robert S. Smith. 1964. Yoruba Warfare in the Nineteenth Century. Cambridge: Cambridge University Press. Akinjogbin, I. A. 1967. Dahomey and Its Neighbours, 1708–1818. Cambridge: Cambridge University Press. Akinnaso, F. Niyi. 1980. “The Sociolinguistic Basis of Yoruba Personal Names.” Anthropological Linguistics 22, no. 7: 275–305. Akintoye, Stephen A. 1971. Revolution and Power Politics in Yorubaland, 1840–1893: Ibadan Expansion and the Rise of the Ekitiparapo. London: Longmans. ———. 2010. A History of the Yoruba People. Dakar, Senegal: Amalion. Allen, William. 1848. A Narrative of the Expedition Sent by Her Majesty’s Government to the River Niger in 1841. 2 vols. London: Richard Bentley. Alonso, Ana M. 1990. “Men in ‘Rags’ and the Devil on the Throne: A Study of Protest and Inversion in the Carnival of Post-Emancipation Trinidad.” Plantation Society in the Americas 3, no. 1: 73–120. Amselle, Jean-Loup. 1998. Mestizo Logics: Anthropology of Identity in Africa and Elsewhere. Translated by Claudia Royal. Stanford, CA: Stanford University Press.

172 / References Apter, Andrew. 1987a. “The Historiography of Yoruba Myth and Ritual.” History in Africa 14:1–25. ———. 1987b. “Things Fell Apart? Yoruba Responses to the 1983 Elections in Ondo State, Nigeria.” Journal of Modern African Studies 24, no. 3: 489–503. ———. 1991. “The Embodiment of Paradox: Yoruba Kingship and Female Power.” Cultural Anthropology 6, no. 2: 212–29. ———. 1992. Black Critics and Kings: The Hermeneutics of Power in Yoruba Society. Chicago: University of Chicago Press. ———. 1993. “Atinga Revisited: Yoruba Witchcraft and the Cocoa Economy, 1950–1951.” In Modernity and Its Malcontents: Ritual and Power in Postcolonial Africa, edited by J. Comaroff and J. L. Comaroff, 111–28. Chicago: University of Chicago Press. ———. 1995. “Notes on Orisha Cults in the Ekiti Highlands: A Tribute to Pierre Verger.” Cahiers d’études africaines 35, nos. 138–39: 369–401. ———. 1998. “Discourse and Its Disclosures: Yoruba Women and the Sanctity of Abuse.” Africa 68, no. 1: 68–96. ———. 1999. “Africa, Empire, and Anthropology: A Philological Exploration of Anthropology’s Heart of Darkness.” Annual Review of Anthropology 28:577–98. ———. 2002. “On African Origins: Creolization and Connaissance in Haitian Vodou.” American Ethnologist 29, no. 2: 233–60. ———. 2005a. “Griaule’s Legacy: Rethinking ‘la parole claire’ in Dogon Studies.” Cahiers d’études africaines 45, no. 177: 95–129. ———. 2005b. The Pan-African Nation: Oil and the Spectacle of Culture in Nigeria. Chicago: University of Chicago Press. ———. 2007. Beyond Words: Discourse and Critical Agency in Africa. Chicago: University of Chicago Press. Apter, Andrew, and Lauren Derby. 2010. Introduction to Activating the Past: History and Memory in the Black Atlantic World, edited by Andrew Apter and Lauren Derby, xiii–xxxi. Newcastle upon Tyne, UK: Cambridge Scholars. Argyriadis, Kali. 2000. “Des noirs sorciers aux babalaos: Analyse du paradoxe du rapport à l’Afrique à la Havane.” Cahiers d’études africaines 40, no. 160: 649–74. Argyriadis, Kali, and Stefania Capone, eds. 2011. La religion des orisha: Un champ social national en pleine recomposition. Paris: Hermann. Awolalu, J. 1979. Yoruba Beliefs and Sacrificial Rites. London: Longmans. Baikie, William B. 1856. Narrative of an Exploring Voyage up the Rivers Kwo’ra and Bi’nue in 1854. London: John Murray. Bakhtin, Mikhail. 1984. Rabelais and His World. Translated by Hélène Iswolsky. Bloomington: Indiana University Press. Barber, Karin. 1981. “How Man Makes God in West Africa: Yoruba Attitudes toward the Òrìsà.” Africa 51, no. 3: 724–45. ———. 1990. “Oríkì, Women and the Proliferation and Merging of the Òrìsà.” Africa 60, no. 3: 313–37. ———. 1991. I Could Speak until Tomorrow: Oríkì, Women, and the Past in a Yoruba Town. Washington, DC: Smithsonian Institution Press. ———. 1995. “Money, Self-Realization, and the Person in Yoruba Texts.” In Money Matters: Instability, Values, and Social Payments in the Modern History of West African Communities, edited by Jane Guyer, 205–24. Portsmouth, NH: Heinemann. ———. 2000. The Generation of Plays: Yoruba Popular Life in Theater. Bloomington: Indiana University Press. Barnes, John A. 1962. “African Models in the New Guinea Highlands.” Man 62:5–9.

References / 173 Barnes, Sandra T., ed. 1989. Africa’s Ogun: Old World and New. Bloomington: Indiana University Press. Barth, Frederick. 1969. Introduction to Ethnic Groups and Boundaries: The Social Organization of Cultural Difference, edited by Frederick Barth, 9–38. Boston: Little, Brown. Bascom, William. 1944. “The Sociological Role of the Yoruba Cult Group.” American Anthropologist, n.s., 46, 1, no. 2. Memoirs, 63:1–75. ———. 1950. “The Focus of Cuban Santeria.” Southwestern Journal of Anthropology 6, no. 1: 64–68. Bastide, Roger. (1960) 1978. The African Religions of Brazil: Toward a Sociology of the Interpenetration of Civilizations. Translated by Helen Sebba. Baltimore: Johns Hopkins University Press. Citations refer to the 1978 edition. ———. (1960) 1986. Sociología de la religión. Madrid: Jucar. ———. 1971. African Civilizations in the New World. New York: Harper Torchbooks. Bastien, Rémy. 1966. “Vodoun and Politics in Haiti.” In Religion and Politics in Haiti, edited by H. Courlander and R. Bastien, 39–68. Washington, DC: Institute for Cross-Cultural Research. Beauvoir-Dominique, Rachel. 1995. “Underground Realms of Being: Vodoun Magic.” In Sacred Arts of Haitian Vodou, edited by Donald Cosentino, 153–77. Los Angeles: University of California, Los Angeles Fowler Museum. Beckford, William. 1790. A Descriptive Account of the Island of Jamaica. 2 vols. London: T. and J. Egerton. Beckles, Hilary McD. 1989. Natural Rebels: A Social History of Enslaved Black Women in Barbados. New Brunswick, NJ: Rutgers University Press. ———. 1996. “Black Female Slaves and White Households in Barbados.” In More Than Chattel: Black Women and Slavery in the Americas, edited by D. B. Gaspar and D. C. Hine, 111–25. Bloomington: Indiana University Press. ———. 2000. “Female Enslavement and Gender Ideologies in the Caribbean.” In Identity in the Shadow of Slavery, edited by Paul Lovejoy, 163–82. London: Continuum. Beier, Ulli. 1956. “Before Oduduwa.” Odu 3:25–31. ———. 1959. A Year of Sacred Festivals in One Yoruba Town. Lagos: Nigeria Magazine Special Publication. ———. 1980. Yoruba Myths. Cambridge: Cambridge University Press. Belasco, Bernard. 1980. The Entrepreneur as Culture Hero. New York: Praeger. Beliso-De Jesús, Aisha M. 2015. Electric Santería: Racial and Sexual Assemblages of Transnational Religion. New York: Columbia University Press. Bender, D. R. 1970. “Agnatic or Cognatic? A Reevaluation of Ondo Descent.” Man, n.s., 5, no. 1: 71–87. Berry, Daina Ramey. 2007. “Swing the Sickle for the Harvest Is Ripe”: Gender and Slavery in Antebellum Georgia. Urbana: University of Illinois Press. Bhabha, Homi. 2004. The Location of Culture. 2nd ed. New York: Routledge. Bickerton, Derek. 1975. Dynamics of a Creole System. Cambridge: Cambridge University Press. Blassingame, John W. 1979. The Slave Community: Plantation Life in the Antebellum South. 2nd ed. New York: Oxford University Press. Blier, Suzanne P. 1995a. African Vodun: Art, Psychology, and Power. Chicago: University of Chicago Press. ———. 1995b. “Vodun: West African Roots of Vodou.” In Sacred Arts of Haitian Vodou, edited by Donald Cosentino, 61–87. Los Angeles: University of California, Los Angeles Fowler Museum.

174 / References ———. 2015. Art and Risk in Ancient Yoruba: Ife History, Power, and Identity, c. 1300. New York: Cambridge University Press. Bohannan, Laura. 1952. “A Genealogical Charter.” Africa: Journal of the International African Institute 22, no. 4: 301–15. Bourdieu, Pierre. 1977. Outline of a Theory of Practice. Translated by Richard Nice. Cambridge: Cambridge University Press. Bowdich, Thomas E. 1819. Mission from Cape Coast to Ashantee: With a Statistical Account of That Kingdom, and Geographical Notices of Other Parts of the Interior of Africa. London: John Murray. Boyer, Véronique. 1996. “Le don et l’initiation.” L’Homme 138 (April–June): 7–24. Brandon, George E. 1983. “‘The Dead Sell Memories’: An Anthropological Study of Santeria in New York City.” 2 vols. PhD diss., Rutgers University. ———. 1993. Santeria from Africa to the New World: The Dead Sell Memories. Bloomington: Indiana University Press. Brantlinger, Patrick. 1986. “Victorians and Africans: The Genealogy of the Myth of the Dark Continent.” In “Race,” Writing, and Difference, edited by H. L. Gates Jr., 185–222. Chicago: University of Chicago Press. Brazeal, Brian. 2003. “The Music of the Bahian Caboclos.” Anthropological Quarterly 73, no. 4: 639–69. ———. 2008. “Dona Preta’s Trek to Cachoeira.” In Africas of the Americas: Beyond the Search for Origins in the Study of Afro-Atlantic Religions, edited by S. Palmié, 223–53. Leiden, Netherlands: Brill. ———. 2010. “A Goat’s Tale: Diabolical Economies of the Bahian Interior.” In Activating the Past: History and Memory in the Black Atlantic World, edited by Andrew Apter and Lauren Derby, 267–93. Newcastle upon Tyne: Cambridge Scholars. ———. 2014. “The Fetish and the Stone: A Moral Economy of Charlatans and Thieves.” In Spirited Things: The Work of “Possession” in Afro-Atlantic Religions, edited by P. Johnson, 131–54. Chicago: University of Chicago Press. Brodwin, Paul E. 1996. Medicine and Morality in Haiti: The Contest for Healing Power. Cambridge: Cambridge University Press. Brown, David H. 1989. “The Garden in the Machine: Afro-Cuban Sacred Art and Performance in Urban New Jersey and New York.” PhD diss., Yale University. ———. 2003. Santería Enthroned: Art, Ritual, and Innovation in an Afro-Cuban Religion. Chicago: University of Chicago Press. Brown, Karen McCarthy. 1991. Mama Lola: A Vodou Priestess in Brooklyn. Berkeley: University of California Press. Brown, Kathleen. 1996. Good Wives, Nasty Wenches, and Anxious Patriarchs: Gender, Race, and Power in Colonial Virginia. Chapel Hill: University of North Carolina Press. Bruel, James William. 1890. Heroes of the Dark Continent. San Francisco: Pacific. Buckley, Anthony. 1976. “The Secret: An Idea in Yoruba Medicinal Thought.” In Social Anthropology and Medicine, edited by Joseph B. Loudon, 396–421. London: Academic. ———. 1985. Yoruba Medicine. Oxford: Clarendon Press. Bush, Barbara. 1990. Slave Women in Caribbean Society, 1650–1838. Bloomington: Indiana University Press. ———. 1996. “Hard Labor: Women, Childbirth, and Resistance in British Caribbean Slave Societies.” In More Than Chattel: Black Women and Slavery in the Americas, edited by D. B. Gaspar and D. C. Hine, 193–217. Bloomington: Indiana University Press. Campbell, Gwyn, Suzanne Miers, and Joseph C. Miller, eds. 2008. Women and Slavery. Vol. 2, The Modern Atlantic. Athens: Ohio University Press.

References / 175 Capone, Stefania. (1999) 2010. Searching for Africa in Brazil: Power and Tradition in Candomblé. Translated by L. L. Grant. Durham, NC: Duke University Press. ———. 2000. “Entre Yoruba et Bantou: L’influence des stéréotypes raciaux dans les études afro-américaines.” Cahiers d’études africaines 50, no. 157: 55–77. ———. 2001. “Regardes croisés sur le bricolage et le syncrétisme: Le syncrétisme dans tous les états.” Archives de sciences sociales de religions 114 (April–June): 42–50. ———. 2005. Les Yoruba du nouveau monde: Religion, ethnicité et nationalisme noir aux ÉtatsUnis. Paris: Karthala. ———. 2008. “Transatlantic Dialogue: Roger Bastide and the African American Religions.” In Africas of the Americas: Beyond the Search for Origins in the Study of Afro-Atlantic Religions, edited by S. Palmié, 255–92. Leiden, Netherlands: Brill. ———. 2014. “Les babalawo en quête d’une Afrique ‘universelle’ ou le syncrétisme revisité.” In Mobilité religieuse: Retours croisés des Afriques aux Amériques, edited by Philippe Chanson, Yvan Droz, Yonata N. Gez, and Edio Soares, 95–114. Paris: Karthala. ———. 2016a. “The Pai-de-santo and the Babaláwo: Religious Interaction and Ritual Rearrangements within Orisha Religion.” In Ifá Divination, Knowledge, Power, and Performance, edited by Jacob K. Olupona and Rowland O. Abiodun, 223–45. Bloomington: Indiana University Press. ———. 2016b. “Re-Africanization in Afro-Brazilian Religions: Rethinking Religious Syncretism.” In The Brill Handbook of Contemporary Religions in Brazil, edited by S. Engler and B. Schmidt, 473–88. Leiden, Netherlands: Brill. Carnap, Rudolph. 1947. Meaning and Necessity: A Study in Semantics and Modal Logic. Chicago: University of Chicago Press. Carneiro, Edison. 1961. Candomblés da Bahia. Rio de Janeiro: Conquista. Carvalho Soares, Mariza de. 2004. “From Gbe to Yoruba: Ethnic Change and the Mina Nation in Rio de Janeiro.” In The Yoruba Diaspora in the Atlantic World, edited by Toyin Falola and Matt D. Childs, 231–47. Bloomington: Indiana University Press. Castor, N. Fadeke. 2013. “Shifting Multicultural Citizenship: Trinidad Orisha Opens the Road.” Cultural Anthropology 28, no. 3: 464–74. ———. 2017. Spiritual Citizenship: Transnational Pathways from Black Power to Ifá in Trinidad. Durham, NC: Duke University Press. Chaudenson, Robert. 1990. “Du mauvais usage de comparativisme: Le cas des études créoles.” Travaux du Cercle Linguistique d’Aix-en-Provence 8:123–58. Childs, Matt D. 2006. The 1812 Aponte Rebellion in Cuba and the Struggle against Atlantic Slavery. Chapel Hill: University of North Carolina Press. Clapperton, Hugh, and Richard Lander. 1829. Journal of a Second Expedition into the Interior of Africa, from the Bight of Benin to Soccatoo. London: John Murray. Clarke, M. Kamari. 2004. Mapping Yoruba Networks: Power and Agency in the Making of Transnational Community. Durham, NC: Duke University Press. Cody, Cheryll Ann. 1982. “Naming, Kinship, and Estate Dispersal: Notes on Slave Family Life on a South Carolina Plantation, 1786–1833.” William and Mary Quarterly 39, no. 1: 192–211. Comaroff, Jean. 1985. Body of Power, Spirit of Resistance: The Culture and History of a South African People. Chicago: University of Chicago Press. Comaroff, Jean, and John L. Comaroff. 1991. Of Revelation and Revolution. Vol. 1, Christianity, Colonialism, and Consciousness in South Africa. Chicago: University of Chicago Press. Comaroff, John L. 2010. “The End of Anthropology, Again: On the Future of an In/ Discipline.” American Anthropologist 112, no. 4: 524–38.

176 / References Courlander, Harold. (1960) 1973. The Drum and the Hoe: Life and Lore of the Haitian People. Berkeley: University of California Press. ———. 1966. “Vodoun in Haitian Culture.” In Religion and Politics in Haiti, edited by Harold Courlander and R. Bastien, 1–38. Washington, DC: Institute for Cross-Cultural Research. Crowther, Samuel Ajayi. 1843. Vocabulary of the Yoruba Language. London: Church Missionary Society. Curtin, Philip. 1969. The Atlantic Slave Trade: A Census. Madison: University of Wisconsin Press. Dalmas, Antoine. 1814. Histoire de la révolution de Saint Domingue. Paris: Mame Frères. DaMatta, Roberto. (1979) 1991. Carnivals, Rogues, and Heroes: An Interpretation of the Brazilian Dilemma. Translated by J. Drury. Notre Dame, IN: University of Notre Dame Press. Dancer, Thomas. 1809. The Medical Assistant; or, Jamaica Practice of Physic Designed Chiefly for the Use of Families and Plantations. Kingston, Jamaica: Alexander Aikman. Davis, Natalie Zemon. 1975. Society and Culture in Early Modern France: Eight Essays by Natalie Zemon Davis. Stanford, CA: Stanford University Press. Davis, Wade. 1988. Passage of Darkness: The Ethnobiology of the Haitian Zombie. Chapel Hill: University of North Carolina Press. Dayan, Joan. 1998. Haiti, History, and the Gods. Berkeley: University of California Press. Debien, Gabriel. 1974. Les esclaves aux Antilles françaises (XVII–XVIII Siècles). Basse-Terre, Guadaloupe: Société d’histoire de la Guadeloupe. Degrandpré, L. 1801. Voyage à la côte occidentale d’Afrique en 1786 et 1787. Paris: Dentu. Denham, D., and H. Clapperton. 1826. Narrative of Travels and Discoveries in Northern and Central Africa in the Years 1822, 1823, and 1824. London: John Murray. De Heusch, Luc. 1995. “Kongo in Haiti: A New Approach to Religious Syncretism.” In Slavery and Beyond: The African Impact on Latin America and the Caribbean, edited by Darien J. Davis, 103–19. Wilmington, DE: Scholarly Resources. De Man, Paul. (1971) 1983. Blindness and Insight: Essays in the Rhetoric of Contemporary Criticism. Minneapolis: University of Minnesota Press. Dennett, R. E. 1910. Nigerian Studies. London: Macmillan. Derby, Lauren. 1994. “Haitians, Money and Magic: Raza and Society in the HaitianDominican Borderlands.” Comparative Studies in Society and History 36, no. 3: 487–526. Díaz, María Elena. 2000. The Virgin, the King, and the Royal Slaves of Cobre: Negotiating Freedom in Colonial Cuba, 1670–1780. Stanford, CA: Stanford University Press. Drewal, Henry, and Margaret Drewal. 1983. Gelede: Art and Female Power among the Yoruba. Bloomington: Indiana University Press. Drewal, Margaret. 1992. Yoruba Ritual: Performance, Play, Agency. Bloomington: Indiana University Press. Drummond, Lee. 1980. “The Cultural Continuum: A Theory of Intersystems.” Man 15, no. 2: 352–74. Du Bois, W. E. B. 1939. Black Folk, Then and Now: An Essay on the History and Sociology of the Negro Race. New York: Henry Holt. Dumont, Louis. 1970. Homo Hierarchicus: An Essay on the Caste System. Translated by M. Sainsbury, L. Dumont, and B. Gulati. Chicago: University of Chicago Press. Dunham, Katherine. (1969) 1994. Island Possessed. Chicago: University of Chicago Press. Eades, Jeremy S. 1980. The Yoruba Today. Cambridge: Cambridge University Press. Eltis, David. 2004. “The Diaspora of Yoruba Speakers, 1650–1865: Dimensions and Implications.” In The Yoruba Diaspora in the Atlantic World, edited by Toyin Falola and Matt D. Childs, 17–39. Bloomington: Indiana University Press.

References / 177 Fabian, Johannes. 1983. Time and the Other: How Anthropology Makes Its Object. New York: Columbia University Press. Fadipe, Nathaniel A. 1970. The Sociology of the Yoruba. Ibadan, Nigeria: Ibadan University Press. Falola, Toyin, and Matt D. Childs, eds. 2004. The Yoruba Diaspora in the Atlantic World. Bloomington: Indiana University Press. Farrow. Stephen S. 1926. Faith, Fancies and Fetich; or, Yoruba Paganism. New York: Macmillan. Fernandez, James W. 1990. “Tolerance in a Repugnant World and Other Dilemmas in the Cultural Relativism of Melville J. Herskovits.” Ethos 18, no. 2: 140–64. Fett, Sharla M. 2002. Working Cures: Healing, Health, and Power on Southern Slave Plantations. Chapel Hill: University of North Carolina Press. Fick, Carolyn E. 1990. The Making of Haiti: The Saint Domingue Revolution from Below. Knoxville: University of Tennessee Press. Forde, Darryl. 1951. The Yoruba-Speaking Peoples of Southwestern Nigeria. London: International African Institute. Fortes, Meyer. 1945. The Dynamics of Clanship among the Tallensi. London: Oxford University Press. ———. 1949. The Web of Kinship among the Tallensi. London: Oxford University Press. ———. 1969. Kinship and the Social Order: The Legacy of Lewis Henry Morgan. Chicago: Aldine. ———. 1971. Introduction to The Developmental Cycle in Domestic Groups, edited by Jack Goody, 1–14. Cambridge: Cambridge University Press. Fox-Genovese, Elizabeth. 1988. Within the Plantation Household: Black and White Women of the Old South. Chapel Hill: University of North Carolina Press. Frazier, E. Franklin. 1939. The Negro Family in the United States. Chicago: University of Chicago Press. Frege, Gottlob. (1892) 1997. “On Sense and Reference.” Translated Max Black. In The Frege Reader, edited by M. Beany, 151–71. Oxford: Blackwell. Freyre, Gilberto. (1933) 1986. The Masters and the Slaves [Casa-Grande & Senzala]: A Study in the Development of Brazilian Civilization. Translated by S. Putnam. Berkeley: University of California Press. Freyre, Gilberto, Estêvão Pinto, and Ivan Wasth Rodrigues. 2005. Casa-Grande & Senzala em Quadrinhos. Recife, Brazil: Fundação Gilberto Freyre. Frobenius, Leo. (1913) 1968. The Voice of Africa: Being an Account of the Travels of the German Inner African Exploration Expedition in the Years 1910–1912. Translated by Rudolf Blind. New York: Benjamin Blom. Fromont, Cécile. 2014. The Art of Conversion: Christian Visual Culture in the Kingdom of Kongo. Chapel Hill: University of North Carolina Press. Garrigus, John D. 2006. Before Haiti: Race and Citizenship in French Saint-Domingue. New York: Palgrave Macmillan. Gaspar, D. B., and D. C. Hine, eds. 1996. More Than Chattel: Black Women and Slavery in the Americas. Bloomington: Indiana University Press. Gates, Henry Louis. 1988. The Signifying Monkey: A Theory of Afro-American Literary Criticism. New York: Oxford University Press. Gautier, Arlette. 1983. “Les esclaves femmes aux Antilles françaises, 1635–1848.” Historical Reflections. Réflexions Historiques 10, no. 3: 409–33. ———. 1985. Les sœurs de solitude: La condition féminine dans l’esclavage aux Antilles du XVIIe au XIXe siècle. Paris: Éditions Caribéennes.

178 / References Geggus, David. 1989. “Sex Ratio, Age and Ethnicity in the Atlantic Slave Trade: Data from French Shipping and Plantation Records.” Journal of African History 30, no. 1: 23–34. Geggus, David, ed. 2014. The Haitian Revolution: A Documentary History. Indianapolis: Hackett. Gilroy, Paul. 1993. The Black Atlantic: Modernity and Double Consciousness. Cambridge, MA: Harvard University Press. Goody, Jack. 1962. Death, Property and the Ancestors: A Study of the Mortuary Customs of the Lodagaa of West Africa. London: Tavistock. Griaule, Marcel. 1965. Conversations with Ogotemmêli. Translated by Ralph Butler. London: Oxford University Press. Griaule, Marcel, and Germaine Dieterlen. 1991. Le renard pâle. Paris: Institut d’Ethnologie. Guedj, Pauline. 2012. Panafricanisme, religion akan et dynamiques identitaires aux États-Unis. Paris: L’Harmattan. Hair, P. E. H., Adam Jones, and Robin Law, eds. 1992. Barbot on Guinea: The Writings of Jean Barbot on West Africa, 1678–1712. 2 vols. London: The Hakluyt Society. Hall, Gwendolyn Midlo. 2003. “African Ethnicities and the Meanings of Mina.” In TransAtlantic Dimensions of Ethnicity in the African Diaspora, edited by P. Lovejoy and D. Trotman, 65–81. London: Continuum. Hall, Stuart. 1989. “Cultural Identity and Cinematic Representation.” Framework 36:68–82. Hallen, Barry, and J. Olubi Sodipo. 1997. Knowledge, Belief, and Witchcraft: Analytic Experiments in African Philosophy. Stanford, CA: Stanford University Press. Hayes, Kelly E. 2011. Holy Harlots: Femininity, Sexuality, and Black Magic in Brazil. Berkeley: University of California Press. Hebblethwaite, Benjamin. 2015. “Historical Linguistic Approaches to Haitian Creole: Vodou Rites, Spirit Names and Songs: The Founders’ Contributions to Asogwe Vodou.” In La Española—Isla de Encuentros; Hispaniola—Island of Encounters, edited by Jessica Stefanie Barzen, Hanna Lene Geiger, and Silke Jansen, 65–86. Tübingen, Germany: Gunter Narr. Henare, Amiria, Martin Holbraad, and Sari Wastell. 2007. Introduction to Thinking through Things: Theorising Artefacts Ethnographically, edited by A. Henare, M. Holbraad, and S. Wastell, 1–31. London: Routledge. Herskovits, Melville J. 1928. The American Negro: A Study of Racial Crossing. New York: Knopf. ———. 1937. “African Gods and Catholic Saints in New World Negro Belief.” American Anthropologist, n.s., 39:635–43. ———. (1938) 1967. Dahomey: An Ancient West African Kingdom. 2 vols. Evanston, IL: Northwestern University Press. ———. (1941) 1958. The Myth of the Negro Past. Boston: Beacon. ———. 1966. The New World Negro: Selected Papers in Afroamerican Studies. Bloomington: Indiana University Press. Heywood, Linda, and John Thornton. 2007. Central Africans, Atlantic Creoles, and the Foundation of the Americas, 1585–1660. Cambridge: Cambridge University Press. Higman, Barry W. 1975. “The Slave Family and Household in the British West Indies, 1800–1834.” Journal of Interdisciplinary History 6, no. 2: 261–87. ———. 1984. Slave Populations of the British Caribbean, 1807–1834. Baltimore: Johns Hopkins University Press. Hodder, B. W. 1969. “Markets in Yorubaland.” In Markets in West Africa: Studies of Markets and Trade among the Yoruba and Ibo, edited by B. W. Hodder and U. I. Ukwu, 3–109. Ibadan, Nigeria: Ibadan University Press.

References / 179 Holbraad, Martin. 2012. Truth in Motion: The Recursive Anthropology of Cuban Divination. Chicago: University of Chicago Press. Horton, Robin. 1983. “Social Psychologies: African and Western.” In Oedipus and Job in West African Religion, edited by Meyer Fortes and Robin Horton, 41–82. Cambridge: Cambridge University Press. Hurston, Zora Neale. (1935) 1978. Mules and Men. Bloomington: Indiana University Press. ———. (1938) 1990. Tell My Horse: Voodoo and Life in Haiti and Jamaica. New York: Harper and Row. Idowu, E. Bolaji. 1962. Olódùmarè: God in Yoruba Belief. London: Longmans. ———. 1970. “The Challenge of Witchcraft.” Orita 4, no. 1: 3–16. Ivy, Marilyn. 1995. Discourses of the Vanishing: Modernity, Phantasm, Japan. Chicago: University of Chicago Press. Jackson, Walter. 1986a. “The Making of a Social Science Classic: Gunnar Myrdal’s An American Dilemma.” Perspectives in American History 2:43–61. ———. 1986b. “Melville Herskovits and the Search for Afro-American Culture.” In Malinowski, Rivers, Benedict, and Others: Essays on Culture and Personality, edited by George Stocking Jr., 95–126. Madison: University of Wisconsin Press. Jakobson, Roman. 1971. Selected Writings, vol. 2. The Hague: Mouton. Johnson, Paul Christopher. 2002. Secrets, Gossip, and Gods: The Transformation of Brazilian Candomblé. Oxford: Oxford University Press. ———. 2006. “Secretism and the Apotheosis of Duvalier.” Journal of the American Academy of Religion 74, no. 2: 420–45. ———. 2007. Diaspora Conversions: Black Carib Religion and the Recovery of Africa. Berkeley: University of California Press. ———, ed. 2014. Spirited Things: The Work of “Possession” in Afro-Atlantic Religions. Chicago: University of Chicago Press. Johnson, Samuel Ajayi. 1921. History of the Yorubas: From the Earliest Times to the Beginning of the British Protectorate. Lagos, Nigeria: CMS Bookshops. Jones, G. I. 1963. The Trading States of the Oil Rivers: A Study of Political Development in Eastern Nigeria. London: Oxford University Press. Kant, Immanuel. (1787) 1973. Immanuel Kant’s Critique of Pure Reason. Translated by Norman Kemp Smith. London: Macmillan. Kelley, Robin D. G. 1999. “‘But a Local Phase of a World Problem’: Black History’s Global Vision, 1883–1950.” Journal of American History 86, no. 3: 1045–77. ———. 2000. “How the West Was One: On the Uses and Limitations of Diaspora.” Black Scholar 30, nos. 3–4: 31–35. Kiddy, Elizabeth W. 2005. Blacks of the Rosary: Memory and History in Minas Gerais, Brazil. University Park: Pennsylvania State University Press. King, Stewart R. 2001. Blue Coat or Powdered Wig: Free People of Color in Pre-Revolutionary Saint Domingue. Athens: University of Georgia Press. Kohn, Eduardo. 2015. “Anthropology of Ontologies.” Annual Review of Anthropology 44:311–27. Laguerre, Michel S. 1980. “Bizango: A Voodoo Secret Society in Haiti.” In Secrecy: A CrossCultural Perspective, edited by Stanton K. Tefft, 147–60. New York: Human Sciences Press. ———. 1987. Afro-Caribbean Folk Medicine. South Hadley, MA: Bergin and Garvey. ———. 1989. Voodoo and Politics in Haiti. New York: St. Martin’s. Laird, MacGregor, and R. A. K. Oldfield. 1837. Narrative of an Expedition into the Interior of

180 / References Africa, by the River Niger, in the Steam-Vessels Quorra and Alburkah, in 1832, 1833, and 1834. 2 vols. London: Richard Bentley. Lander, Richard, and John Lander. 1832. Journal of an Expedition to Explore the Course and Termination of the Niger; with a Narrative of a Voyage down That River to Its Termination. 3 vols. London: John Murray. Larose, Serge. 1977. “The Meaning of Africa in Haitian Vodu.” In Symbols and Sentiments: Cross-Cultural Studies in Symbolism, edited by Ioan Lewis, 85–116. London: Academic. ———. 1978. “The Haitian Lakou, Land, Family and Ritual.” In Family and Kinship in Middle America and the Caribbean, edited by Arnaud Marks and René Römer, 482–512. Willemstad: Institute of Higher Studies in Curaçao. Latour, Bruno, and Steve Woolgar. 1979. Laboratory Life: The Social Construction of Scientific Facts. Beverly Hills, CA: Sage. Law, Robin. 1973. “Traditional History.” In Sources of Yoruba History, edited by S. O. Biobaku, 25–40. Oxford: Clarendon. ———. 1977. The Oyo Empire c. 1600–c. 1836: A West African Imperialism in the Era of the Atlantic Slave Trade. Oxford: Clarendon. ———. 1982. “Jean Barbot as a Source for the Slave Coast of West Africa.” History in Africa 9:155–73. ———. 1997. “Ethnicity and the Slave Trade: ‘Lucumi’ and ‘Nagô’ as Ethnonyms in West Africa.” History in Africa 24:205–19. ———. 2004. “Yoruba Liberated Slaves Who Returned to Africa.” In The Yoruba Diaspora in the Atlantic World, edited by Toyin Falola and Matt D. Childs, 349–65. Bloomington: Indiana University Press. ———. 2005. “Ethnicities of Enslaved Africans in the Diaspora: On the Meanings of ‘Mina’ (Again).” History in Africa 32:247–67. Lefèbvre, Claire. 1986. “Relexification in Creole Genesis Revisited: The Case of Haitian Creole.” In Substrata versus Universals in Creole Genesis, edited by P. Muysken and N. Smith, 279–300. Amsterdam: John Benjamins. ———. 1993. “The Role of Relexification and Syntactic Reanalysis in Haitian Creole: Methodological Aspects of a Research Program.” In Africanisms in Afro-American Language Varieties, edited by Salikoko Mufwene, 254–79. Athens: University of Georgia Press. Lindsay, Lisa. 1994. “‘To Return to the Bosom of Their Fatherland’: Brazilian Immigrants in Nineteenth-Century Lagos.” Slavery and Abolition 15, no. 1: 22–50. Lloyd, Peter C. 1954. “The Traditional Political System of the Yoruba.” Southwestern Journal of Anthropology 10:366–84. ———. 1955a. “The Yoruba Lineage.” Africa 25, no. 3: 235–51. ———. 1955b. “Yoruba Myths—A Sociologist’s Interpretation.” Odu 11:20–28. ———. 1960. “Sacred Kingship and Government among the Yoruba.” Africa 30, no. 3: 221–38. ———. 1962. Yoruba Land Law. London: Oxford University Press. ———. 1966. “Agnatic and Cognatic Descent among the Yoruba.” Man, n.s., 1, no. 4: 484–500. ———. 1968. “Conflict Theory and Yoruba Kingdoms.” In History and Social Anthropology, edited by I. M. Lewis, 25–62. London: Tavistock. ———. 1971. The Political Development of Yoruba Kingdoms in the Eighteenth and Nineteenth Centuries. Occasional Paper no. 31. London: Royal Anthropological Institute. Long, Edward. 1744. The History of Jamaica; or, General Survey of the Antient and Modern State

References / 181 of the Island: With Reflections on Its Situation, Settlements, Inhabitants, Climate, Products, Commerce, Laws, and Government. 3 vols. London: T. Lowndes. López Valdés, Rafael L. 1986. “Notas para el estudio etnohistórico de los esclavos Lucumí de Cuba.” Anales del Caribe 6:55–74. ———. 1994. “Cabildos de africanos y religiones afro-cubanas: Un nuevo enfoque.” Unpublished manuscript. Lovejoy, Henry. 2012. “Old Oyo Influences on the Transformation of Lucumí Identity in Colonial Cuba.” PhD diss., University of California, Los Angeles. Lovejoy, Henry B., and Olatunji Ojo. 2015. “‘Lucumí,’ ‘Terranova,’ and the Origins of the Yoruba Nation.” Journal of African History 56, no. 3: 353–72. Lovejoy, Paul E. 2003. “Ethnic Designations of the Slave Trade and the Reconstruction of the History of Trans-Atlantic Slavery.” In Trans-Atlantic Dimensions of Slavery in the African Diaspora, edited by P. Lovejoy and D. Trotman, 9–42. London: Continuum. ———. 2004. “The Yoruba Factor in the Trans-Atlantic Slave Trade.” In The Yoruba Diaspora in the Atlantic World, edited by Toyin Falola and Matt D. Childs, 40–55. Bloomington: Indiana University Press. Mabogunje, Akin L., and John D. Omer-Cooper. 1971. Owu in Yoruba History. Ibadan, Nigeria: Ibadan University Press. Malenfant, C. 1814. Des colonies: Et particulièrement de celle de Saint-Domingue. Paris: Audibert. Malinowski, Bronislaw. (1922) 1961. Argonauts of the Western Pacific. New York: Dutton. Mars, Louis. 1953. “Nouvelle contribution à l’étude de la crise de possession.” Mémoires de l’Institut français d’Afrique noir 27:213–33. Mary, André. 2000. Le bricolage africaine des héros chrétiens. Paris: Cerf. Mason, Michael. 1970. “The ‘Jihad’ in the South: An Outline of the Nineteenth Century Nupe Hegemony in North-Eastern Yorubaland and Afenmai.” Journal of the Historical Society of Nigeria 5, no. 2: 193–209. Matory, J. Lorand. 1986. “Vessels of Power: The Dialectical Symbolism of Power in Yoruba Religion and Polity.” MA thesis, University of Chicago. ———. 1994. Sex and the Empire That Is No More: Gender and the Politics of Metaphor in Oyo Yoruba Religion. Minneapolis: University of Minnesota Press. ———. 1999. “The English Professors of Brazil: On the Diasporic Roots of the Yoruba Nation.” Comparative Studies in Society and History 41, no. 1: 72–103. ———. 2005. Black Atlantic Religion: Tradition, Transnationalism, and Matriarchy in the AfroBrazilian Candomblé. Princeton, NJ: Princeton University Press. ———. 2006. “The ‘New World’ Surrounds an Ocean: Theorizing the Live Dialogue between African and African American Cultures.” In Afro-Atlantic Dialogues: Anthropology in the Diaspora, edited by Kevin A. Yelvington, 151–92. Santa Fe, NM: School of American Research Press. ———. 2015. Review of The Formation of Candomblé: Vodun History and Ritual in Brazil, by Luis Nicolau Parés. The Americas 72, no. 4: 609–27. McDonnell, Lawrence T. 1988. “‘Money Knows No Master’: Market Relations and the Slave Community.” In Developing Dixie: Modernization in a Traditional Society, edited by Winfred B. Moore Jr., Joseph F. Tripp, and Lyon G. Tyler Jr., 31–44. New York: Greenwood. McIntosh, Marjorie K. 2009. Yoruba Women, Work, and Social Change. Bloomington: Indiana University Press. McKenzie, Peter. 1997. Hail Orisha! A Phenomenology of a West African Religion in the MidNineteenth Century. Leiden, Netherlands: Brill.

182 / References Métraux, Alfred. (1959) 1972. Voodoo in Haiti. Translated by H. Charteris. New York: Schocken. Mintz, Sidney W. 1955. “The Jamaican Internal Marketing Pattern: Some Notes and Hypotheses.” Social and Economic Studies 4:95–103. ———. 1971. “Men, Women and Trade.” Comparative Studies in Society and History 13, no. 3: 247–69. ———. 1983. “Caribbean Marketplaces and Caribbean History.” Radical History Review 27:110–20. Mintz, Sidney W., and D. Hall. (1960) 1974. “The Origins of the Jamaican Internal Marketing System.” In Caribbean Transformations, edited by S. W. Mintz, 180–213. Baltimore: Johns Hopkins University Press. Mintz, Sidney W., and R. Price. (1976) 1992. The Birth of African-American Culture: An Anthropological Perspective. Boston: Beacon. Mishra, Dudesh. 2006. Diaspora Criticism. Edinburgh: Edinburgh University Press. Mockler-Ferryman, Augustus F. 1892. Up the Niger: Narrative of Major Claude MacDonald’s Mission to the Niger and Benue Rivers, West Africa. London: George Philip and Son. Moitt, Bernard. 2001. Women and Slavery in the French Antilles, 1635–1848. Bloomington: Indiana University Press. Moore, Robin. 1994. “Representations of Afrocuban Expressive Culture in the Writings of Fernando Ortiz.” Latin American Music Review 15, no. 1: 32–54. Moreau de Saint-Méry, Louis-Elie. 1797. Description topographique, physique, civile, politique et historique de la partie française de l’île de Saint Domingue. 2 vols. Paris: L. Guérin. Morgan, Jennifer L. 2004. Laboring Women: Reproduction and Gender in New World Slavery. Philadelphia: University of Pennsylvania Press. Morgan, John H. 1860. “An Essay on the Production of Abortion among our Negro Population.” Nashville Journal of Medicine and Surgery 19 (August): 117–23. Morgan, Louis Henry. 1871. Systems of Consanguinity and Affinity of the Human Family. Washington, DC: Smithsonian Institution. Morgan, Philip D. 1998. Slave Counterpoint: Black Culture in the Eighteenth-Century Chesapeake and Lowcountry. Chapel Hill: University of North Carolina Press. Morrissey, Marietta. 1989. Slave Women in the New World: Gender Stratification in the Caribbean. Lawrence: University Press of Kansas. Morton-Williams, Peter. 1956. “The Atinga Cult among the South-Western Yoruba: A Sociological Analysis of a Witch-Finding Movement.” Bulletin de l’IFAN 18, nos. 3–4: 315–34. Mudimbe, Valentin Y. 1988. The Invention of Africa: Gnosis, Philosophy, and the Order of Knowledge. Bloomington: Indiana University Press. ———. 1990. “Which Idea of Africa? Herskovits’s Cultural Relativism.” October 55:93–104. ———. 1994. The Idea of Africa. Bloomington: Indiana University Press. Mufwene, Salikoko. 1990. “Transfer and the Substrate Hypothesis in Creolistics.” Studies in Second Language Acquisition 12:1–23. ———. 1993. “African Substratum: Possibility and Evidence.” In Africanisms in AfroAmerican Language Varieties, edited by Salikoko Mufwene, 192–208. Athens: University of Georgia Press. Murphy, Joseph. 1988. Santería: An African Religion in America. Boston: Beacon. Murray, Gerald F. 1977. “The Evolution of Haitian Peasant Land Tenure: A Case Study in Agrarian Adaptation to Population Growth.” PhD diss., Columbia University. Nicholls, David. 1979. From Dessalines to Duvalier: Race, Colour and National Independence in Haiti. New York: Cambridge University Press.

References / 183 ———. 1985. Haiti in Caribbean Context: Ethnicity, Economy and Revolt. New York: St. Martin’s. Nunley, John W. 1987. Moving with the Face of the Devil: Art and Politics in Urban West Africa. Urbana: University of Illinois Press. Nwokeji, Ugo G. 2001. “African Conceptions of Gender and the Slave Traffic.” William and Mary Quarterly 58, no. 1: 47–68. Obayemi, A. 1971. “The Yoruba and Edo-Speaking Peoples and Their Neighbours before 1600.” In History of West Africa, edited by J. F. Ade Ajayi and Michael Crowder, 1:196– 263. New York: Longmans. Ogunba, Oyin. 1967. “Ritual Drama of the Ijebu People: A Study of Indigenous Festivals.” PhD diss., University of Ibadan. Oguntuyi, A. 1979. History of Ekiti. Ibadan, Nigeria: Bisi. Ogunyemi, Chikwenye Okonjo. 1996. Africa Wo/Man Palava: The Nigerian Novel by Women. Chicago: University of Chicago Press. Ojo, Olatunje. 2009a. “‘Heepa (Hail) Òrìsà’: The Òrìsà Factor in the Birth of Yoruba Identity.” Journal of Religion in Africa 39, no. 1: 30–59. ———. 2009b. “The Root Is Also Here: The Nondiaspora Foundations of Yoruba Ethnicity.” In Movements, Borders, and Identities in Africa, edited by Toyin Falola and A. Usman, 53–80. Rochester, NY: University of Rochester Press. Olupona, Jacob K. 1991. Kingship, Religion, and Rituals in a Nigerian Community: A Phenomenological Study of Ondo Yoruba Festivals. Stockholm: Almqvist and Wiksell. ———. 2011. City of 201 Gods: Ile-Ife in Time, Space, and the Imagination. Berkeley: University of California Press. Olwell, Robert. 1996. “‘Loose, Idle and Disorderly’: Slave Women in the EighteenthCentury Charleston Marketplace.” In More Than Chattel: Black Women and Slavery in the Americas, edited by David Barry Gaspar and Darlene Clark Hine, 97–110. Bloomington: Indiana University Press. Ortiz, Fernando. 1916. Hampa afro-cubana: Los negros esclavos, estudio de sociológica y derecho pública. Havana: Revista Bimestre Cubana. ———. 1921. “Los cabildos Afrocubanos.” Revista Bimestre Cubana 16:5–39. ———. 1940. “Los factores humanos de la cubanidad.” Revista Bimestre Cubana 45:161–86. ———. (1906) 1973. Los negros brujos: Apuntes para un studio de etnología criminal. Miami: Ediciones Universal. Otero, Solimar. 2010. Afro-Cuban Diasporas in the Atlantic World. Rochester, NY: University of Rochester Press. Palmié, Stephan. 1993. “Ethnogenetic Processes and Cultural Transfer in Afro-American Slave Populations.” In Slavery in the Americas, edited by W. Binder, 337–63. Würtzburg, Germany: Königshausen und Neumann. ———. 1995. “Against Syncretism: ‘Africanizing’ and ‘Cubanizing’ Discourses in North American Òrìsà Worship.” In Counterworks: Managing the Diversity of Knowledge, edited by Richard Fardon, 73–104. London: Routledge. ———. 2002. Wizards and Scientists: Explorations in Afro-Cuban Modernity and Tradition. Durham, NC: Duke University Press. ———. 2005. “The Cultural Work of Yoruba-Globalization.” In Christianity and Social Change in Africa: Essays in Honor of J. D. Y. Peel, edited by Toyin Falola, 43–81. Chapel Hill, NC: Carolina Academic Press. ———. 2006. “Creolization and Its Discontents.” Annual Review of Anthropology 35:433–56. ———. 2008. Introduction to Africas of the Americas: Beyond the Search for Origins in the Study of Afro-Atlantic Religions, edited by S. Palmié, 1–37. Leiden, Netherlands: Brill.

184 / References ———. 2010. “Ekpe/Abakuá in Middle Passage: Time, Space and Units of Analysis in African American Historical Anthropology.” In Activating the Past: History and Memory in the Black Atlantic World, edited by Andrew Apter and Lauren Derby, 1–44. Newcastle upon Tyne, UK: Cambridge Scholars. ———. 2013. The Cooking of History: How Not to Study Afro-Cuban Religion. Chicago: University of Chicago Press. Paquette, Robert L. 1988. Sugar Is Made with Blood: The Conspiracy of La Escalera and the Conflict between Empires over Slavery in Cuba. Middletown, CT.: Wesleyan University Press. Parés, Luis Nicolau. 2004. “The ‘Nagôization’ Process in Brazilian Candomblé.” In The Yoruba Diaspora in the Atlantic World, edited by Toyin Falola and Matt D. Childs, 185– 208. Bloomington: University of Indiana Press. ———. (2006) 2013. The Formation of Candomblé: Vodun History and Ritual in Brazil. Translated by Richard Vernon and L. N. Parés. Chapel Hill: University of North Carolina Press. ———. 2008. “Ethnic-Religious Modes of Identification among the Gbe-Speaking People in Eighteenth and Nineteenth Century Brazil.” In Africa, Brazil and the Construction of Trans-Atlantic Black Identities, edited by Boubacar Barry, Elisée Soumonni, and Livio Sansone, 179–207. Trenton, NJ: Africa World Press. ———. 2010. “Memories of Slavery in Religious Ritual: Comparing Benin Vodun and Bahian Candomblé.” In Activating the Past: History and Memory in the Black Atlantic World, edited by Andrew Apter and Lauren Derby, 71–97. Newcastle upon Tyne, UK: Cambridge Scholars. ———. 2015. “Reply to the Review of J. Lorand Matory.” The Americas 72, no. 4: 628–41. Parés, Luis Nicolau, and Roger Sansi, eds. 2011. Sorcery in the Black Atlantic. Chicago: University of Chicago Press. Park, Robert E. 1919. “The Conflict and Fusion of Cultures with Special Reference to the Negro.” Journal of Negro History 4:111–33. Parker, Richard G. 1991. Bodies, Pleasures and Passions: Sexual Culture in Contemporary Brazil. Boston: Beacon. Parrinder, E. G. 1967. The Story of Ketu: An Ancient Yoruba Kingdom. 2nd ed. Ibadan, Nigeria: Ibadan University Press. Patterson, Orlando. 1982a. “Recent Studies on Caribbean Slavery and the Atlantic Slave Trade.” Latin American Research Review 17, no. 3: 251–75. ———. 1982b. Slavery and Social Death: A Comparative Study. Cambridge, MA: Harvard University Press. Peel, John D. Y. 1983. Ijeshas and Nigerians: The Incorporation of a Yoruba Kingdom, 1890s– 1970s. Cambridge: Cambridge University Press. ———. 1989. “The Cultural Work of Yoruba Ethnogenesis.” In History and Ethnicity, edited by E. Tonkin, M. O. McDonald, and M. Chapman, 198–215. London: Routledge. ———. 2000. Religious Encounter and the Making of the Yoruba. Bloomington: Indiana University Press. Pemberton, John. 1977. “A Cluster of Symbols: Orisha Worship among the Igbomina Yoruba of Ila-Orangun.” History of Religions 17, no. 1: 1–28. ———. 1979. “Sacred Kingship and the Violent God.” Berkshire Review 14:85–106. Pérez, Elizabeth. 2010. “The Virgin in the Mirror: Reading Images of a Black Madonna through the Lens of Afro-Cuban Women’s Experiences.” Journal of African American History 95, no. 2: 202–28. ———. 2016. Religion in the Kitchen: Cooking, Talking, and the Making of Black Atlantic Traditions. Chicago: University of Chicago Press.

References / 185 Price, Richard. 1976. The Guiana Maroons: A Historical and Bibliographic Introduction. Baltimore: Johns Hopkins University Press. ———. 1983. First Time: The Historical Vision of an Afro-American People. Baltimore: Johns Hopkins University Press. Price, Sally. 1984. Co-Wives and Calabashes. Ann Arbor: University of Michigan Press. Price, Sally, and Richard Price. 1980. Afro-American Arts of the Surinam Rain Forest. Berkeley: University of California Press. ———. 1991. Two Evenings in Saramaka. Chicago: University of Chicago Press. Price-Mars, Jean. 1928. Ainsi parla l’oncle: Essais d’ethnographie. Port-au-Prince: Compiègne. Prince, Raymond H. 1961. “The Yoruba Image of the Witch.” Journal of Mental Science 107:795–805. Quine, Willard V. O. 1948. “On What There Is.” Review of Metaphysics 2, no. 5: 21–38. ———. 1951. “Two Dogmas of Empiricism.” Philosophical Review 60, no. 1: 20–43. Ramos, Arthur. 1937. As culturas negras no novo mundo. Rio de Janeiro: Civilização brasiliera. Ramsey, Kate. 2011. The Spirits and the Law: Vodou and Power in Haiti. Chicago: University of Chicago Press. Ranger, Terence. 1983. “The Invention of Tradition in Colonial Africa.” In The Invention of Tradition, edited by Eric Hobsbawm and Terence Ranger, 211–62. New York: Cambridge University Press. Reis, João José. 2011. “Candomblé and Slave Resistance in Nineteenth-Century Bahia.” In Sorcery in the Black Atlantic, edited by L. N. Parés and R. Sansi, 55–74. Chicago: University of Chicago Press. Reis, João José, and Beatriz G. Mamigonian. 2004. “Nagô and Mina: The Yoruba Diaspora in Brazil.” In The Yoruba Diaspora in the Atlantic World, edited by Toyin Falola and Matt D. Childs, 77–110. Bloomington: Indiana University Press. Ribeiro, René. 1952. Cultos afrobrasileiros de Recife: Um estudo de ajustemento social. Special issue of Boletim do Instituto Joaquim Nabuco. Recife, Brazil: Recife Instituto Joaquim Nabuco de pesquisas sociais. Richman, Karen E. 2005. Migration and Vodou. Gainesville: University Press of Florida. ———. 2014. “Possession and Attachment: Notes on Moral Ritual Communication among Haitian Descent Groups.” In Spirited Things: The Work of “Possession” in Afro-Atlantic Religions, edited by P. C. Johnson, 207–23. Chicago: University of Chicago Press. Roach, Joseph. 1996. Cities of the Dead: Circum-Atlantic Performance. New York: Columbia University Press. Roberts, Kevin. 2004. “Yoruba Family, Gender, and Kinship Roles in New World Slavery.” In The Yoruba Diaspora in the Atlantic World, edited by Toyin Falola and M. Childs, 248–59. Bloomington: Indiana University Press. Robertson, Claire. 1996. “Africa into the Americas? Slavery and Women, the Family, and the Gender Division of Labor.” In More Than Chattel: Black Women and Slavery in the Americas, edited by D. B. Gaspar and D. C. Hine, 3–40. Bloomington: Indiana University Press. Rosenthal, Judy. 1998. Possession, Ecstasy, and Law in Ewe Voodoo. Charlottesville: University of Virginia Press. Rotberg, Robert I. 1976. “Vodun and the Politics of Haiti.” In The African Diaspora: Interpretative Essays, edited by Martin L. Kilson and Robert I. Rotberg, 342–65. Cambridge, MA: Harvard University Press. Ryle, Gilbert. 1949. The Concept of Mind. Chicago: University of Chicago Press. Sahlins, Marshall. 1985. Islands of History. Chicago: University of Chicago Press.

186 / References Sansi, Roger. 2007. Fetishes and Monuments: Afro-Brazilian Art and Culture in the 20th Century. New York: Berghahn. ———. 2011. “Sorcery and Fetishism in the Modern Atlantic.” In Sorcery in the Black Atlantic, edited by Luis Nicolau Parés and Roger Sansi, 19–39. Chicago: University of Chicago Press. Saussure, Ferdinand de. (1916) 1966. Course in General Linguistics. Translated by Wade Baskin. New York: McGraw Hill. Scheffler, Harold. 1966. “Ancestor Worship in Anthropology; or, Observations on Descent and Descent Groups.” Current Anthropology 7, no. 5: 541–48. Schwab, William B. 1955. “Kinship and Lineage among the Yoruba.” Africa 25, no. 4: 352–74. ———. 1958. “The Terminology of Kinship and Marriage among the Yoruba.” Africa 28, no. 4: 301–13. Schwartz, Stuart B. 1985. Sugar Plantations in the Formation of Brazilian Society: Bahia, 1550– 1835. Cambridge: Cambridge University Press. Scott, David. 1991. “That Event, This Memory: Notes on the Anthropology of African Diasporas in the New World.” Diaspora 1, no. 3: 261–84. ———. 1997. “‘An Obscure Miracle of Connection’: Discursive Tradition and Black Diaspora Criticism.” Small Axe 1:19–38. Sebeok, Thomas. 1960. Style in Language. Cambridge, MA: Technology Press of the Massachusetts Institute of Technology. Sewell, William H. 2005. Logics of History: Social Theory and Social Transformation. Chicago: University of Chicago Press. Silverstein, Michael. 1993. “Metapragmatic Discourse and Metapragmatic Function.” In Reflexive Language: Reported Speech and Metapragmatics, edited by John Lucy, 33–58. Cambridge: Cambridge University Press. ———. (1996) 2003. “Indexical Order and the Dialectics of Social Life.” Language and Communication 23, nos. 3–4: 193–229. Silverstein, Michael, and Greg Urban. 1996. “The Natural History of Discourse.” In Natural Histories of Discourse, edited by M. Silverstein and G. Urban, 1–17. Chicago: University of Chicago Press. Singler, John V. 1993. “African Influence upon Afro-American Language Varieties: A Consideration of SocioHistorical Factors.” In Africanisms in Afro-American Language Varieties, edited by Salikoko Mufwene, 235–53. Athens: University of Georgia Press. Skidmore, Thomas E. (1974) 1993. Black into White: Race and Nationality in Brazilian Thought. Durham, NC: Duke University Press. Smith, M. G. 1956. “On Segmentary Lineage Systems.” Journal of the Royal Anthropological Institute 86, no. 2: 39–80. ———. 1957. “The African Heritage in the Caribbean.” In Caribbean Studies: A Symposium, edited by Vera Rubin, 34–46. Mona, Jamaica: University College of the West Indies, Institute for Social and Economic Research. ———. 1962. Kinship and Community in Carriacou. New Haven, CT: Yale University Press. ———. 1963. Dark Puritan: The Life and Work of Norman Paul. Kingston, Jamaica: University of the West Indies, Department of Extra-Mural Studies. ———. 1965. Stratification in Grenada. Berkeley: University of California Press. ———. 1975. Corporations and Society. The Social Anthropology of Collective Action. Chicago: Aldine. Smith, William. 1744. A New Voyage to Guinea. London: John Nourse. Spivak, Gayatri C. 1988. “Can the Subaltern Speak?” In Marxism and the Interpretation of

References / 187 Culture, edited by C. Nelson and L. Grossberg, 271–313. Urbana: University of Illinois Press. Stevenson, Brenda E. 1996a. “Gender Convention, Ideals and Identity among Antebellum Virginia Slave Women.” In More Than Chattel: Black Women and Slavery in the Americas, edited by D. B. Gaspar and D. C. Hine, 169–90. Bloomington: Indiana University Press. ———. 1996b. Life in Black and White: Family and Community in the Slave South. New York: Oxford University Press. ———. 2005. “‘Marsa Never Sot Aunt Rebecca Down’: Enslaved Women, Religion, and Social Power in the Antebellum South.” Journal of African American History 90, no. 4: 345–67. ———. 2007. “The Question of Slave Female Community in the American South: Methodological and Ideological Approaches.” Journal of African American History 92 (Winter): 74–95. Stewart, Charles, and Rosalind Shaw, eds. 1994. Syncretism/Anti-Syncretism: The Politics of Religious Synthesis. New York: Routledge. Sudarkasa, Niara. 1973. Where Women Work: A Study of Yoruba Women in the Marketplace and in the Home. Anthropological Papers, no. 53. Ann Arbor: University of Michigan, Museum of Anthropology. Talbot, Amaury S. 1921–26. The Peoples of Southern Nigeria. 2 vols. London: Oxford University Press. Taussig, Michael T. 1987. Shamanism, Colonialism, and the Wild Man: A Study in Terror and Healing. Chicago: University of Chicago Press. Thomas, Hugh. 1971. Cuba: The Pursuit of Freedom. New York: Harper. Thompson. Robert F. (1971) 1976. Black Gods and Kings: Yoruba Art at UCLA. Bloomington: Indiana University Press. ———. 1984. Flash of the Spirit: African and Afro-American Art and Philosophy. New York: Vintage. ———. 1995. “From the Isle beneath the Sea: Haiti’s Africanizing Vodou Art.” In Sacred Arts of Haitian Vodou, edited by Donald Cosentino, 91–119. Los Angeles: University of California, Los Angeles Fowler Museum. Thornton, John. 1984. “The Development of an African Catholic Church in the Kingdom of Kongo, 1491–1750.” Journal of African History 25, no. 2: 147–67. Trager, Lillian. 1976. “Yoruba Markets and Trade: Analysis of Spatial Structure and Social Organization in the Ijesaland Marketing System.” PhD diss., University of Washington. Trouillot, Michel-Rolph. 1990. Haiti: State against Nation. New York: Monthly Review Press. Van Beek, Walter E. 1991. “Dogon Restudied: A Field Evaluation of the Work of Marcel Griaule.” Current Anthropology 32, no. 2: 139–58. Verger, Pierre F. 1957. Notes sur le culte des orisha et vodun. Dakar: Institut français d’Afrique noire. ———. 1968. Flux et reflux de la traite des nègres entre le golfe de Bénin et Bahia de Todos os Santos du dix-septième au dix-neuvième siècle. Paris: Mouton. ———. 1981. Orixás: Deuses Iorubás na Africa e no novo mundo. Salvador, Brazil: Corrupio. ———. 1982. Orisha: Les dieux Yorouba en Afrique et au nouveau monde. Paris: A. M. Métailié. Verran, Helen. 2001. Science and an African Logic. Chicago: University of Chicago Press. Wafer, Jim. 1991. The Taste of Blood: Spirit Possession in Brazilian Candomblé. Philadelphia: University of Pennsylvania Press. Wallace-Sanders, Kimberly. 2008. Mammy: A Century of Race, Gender, and Southern Memory. Ann Arbor: University of Michigan Press.

188 / References Watson, Ruth. 2003. ‘Civil Disorder Is the Disease of Ibadan’: Chieftaincy and Civic Culture in a Yoruba City. Athens: Ohio University Press. Weber, Max. (1904) 1948. “‘Objectivity’ in Social Science and Social Policy.” In The Methodology of the Social Sciences, edited and translated by E. A. Shils and H. A. Finch, 49–112. New York: Free Press. White, Deborah Gray. (1985) 1999. Ar’n’t I a Woman? Female Slaves in the Plantation South. New York: W. W. Norton. White, Hayden. 1978. Tropics of Discourse: Essays in Cultural Criticism. Baltimore: Johns Hopkins University Press. Wirtz, Kristina. 2007. Ritual, Discourse, and Community in Cuban Santería: Speaking a Sacred Word. Gainesville: University Press of Florida. ———. 2008. “Divining the Past: The Linguistic Reconstruction of ‘African’ Roots in Diasporic Ritual Registers and Songs.” In Africas of the Americas: Beyond the Search for Origins in the Study of Afro-Atlantic Religions, edited by S. Palmié, 141–77. Leiden, Netherlands: Brill. ———. 2014. Performing Afro-Cuba: Image, Voice, Spectacle in the Making of Race and History. Chicago: University of Chicago Press. Wood, Betty. 1990. “‘White Society’ and the ‘Informal’ Slave Economies of Lowcountry Georgia, c. 1763–1830.” Slavery and Abolition 11, no. 3: 313–31. Woodson, Carter G. 1936. The African Background Outlined. Washington, DC: Association for the Study of Negro Life and History. Woodson, Drexel. 1990. “Tout Mounn Sé Mounn, Men Tout Mounn Pa Menm: Microlevel Sociocultural Aspects of Land Tenure in a Northern Haitian Locality.” PhD diss., University of Chicago. Yelvington, Kevin. 2006a. Introduction to Afro-Atlantic Dialogues: Anthropology in the Diaspora, edited by K. Yelvington, 1–32. Santa Fe, NM: School of American Research Press. ———. 2006b. “The Invention of Africa in Latin America and the Caribbean: Political Discourse and Anthropological Praxis, 1920–1940.” In Afro-Atlantic Dialogues: Anthropology in the Diaspora, edited by K. Yelvington, 35–82. Santa Fe, NM: School of American Research Press.

INDEX

ààfin (palace[s], Standard Yoruba), 132 abà (farm huts), 73 abortion, 112, 117; midwives assist in, 113; punishment for, 114. See also child/children; midwives; women abúlé (farm shack), 73–74 adé Olókun (beaded crown), 79 Ado-Ekiti, 125 àdúgbò (quarters), 1, 29, 41, 70, 71, 73, 74, 75, 81, 82; in Ekiti region, 125; as towns, 125 afọ bajẹ (kingmakers), 83, 87 Africa/African: and black Americas, 18; as imprecise designation, 26, 28–29; studies, 39, 159n1. See also blacks; Negro(es); racism “African Harpies,” stereotype of enslaved women traders, 119. See also women Africanism/Africanity, 18; as acculturative process (Herskovits), 25–29; Herskovits’s continuum of New World and, 20, 21, 21t, 22–23; in New World, 23–24, 32, 110, 117–18, 119–20, 166n35 (see also secrecy); and Vodou, 34. See also Africanization; religion(s) Africanization: of Christianity in New World, 35, 45; redone, as purification, 148. See also Africanism/Africanity; diaspora, African; Nagô-Candomblé lines; whites agbo ilé (compounds), 42, 73, 124. See also house/home aguinaldos (economic tribute, in Cuba), 138 ahistoricism, 154

Ajayi, Jacob Ade, 2 àjẹ́ (witch, malign female double), 100, 163–64n3 ajẹ́ lẹ́ Shango priests, 70. See also Shango ajiaco (folkloric soup), 4 àkòró crown, 70 Aku cult, 85–86. See also cult(s); orisha Aláfọ̀ n (baálẹ̀ of Ilafon), 87; àwòrò of, serve as kingmakers for, 87 Allada (Arada), 36, 46 Alu (Yagba kingdom), 134 Americas. See New World anachronism, 154–55 analogy: of cool to right and hot to left (deities), 36, 46–47; European/civilization and African/barbarianism, 63; fancy/savage and European/African, 138; of hot/cool blood and equilibrium/disequilibrium, 162n27; sexuality/ witchcraft and fertility/motherhood, 110; spears/kola nut and death/life, 138; of transformation in Petro, 47–48; of water/Rada and fire/Petwo, 60–61. See also deities; orisha; Petwo (Petro) deities; Rada deities; Vodou (Haitian) ancestors: egum (dubious, in Candomblé), 148; origin of eponymous nominalization and lineages, 145; as “root founders” (raíces) of houses (in Santería), 140–41; worship of, 70–71, 150. See also descent; genealogy(ies); house/ home; kingdom(s); lineage(s), Yoruba; origin(s) Angola, 50; Candomblé nations in, 167n26

190 / Index animals: deities demand as sacrifice, 47; metamorphosis of, 47–48. See also deities; sacrifice(s) Anthony, Saint, 25 anthropology: apparatus of, 7; as “science of man,” 24–25 áòfin (palace, Ekiti dialect), 75. See also ààfin Arabia, 13 àrán cloth, embroidered, festival, 137 àrẹ mẹ́ ta (three junior age sets), 77 Àrẹ -ọ̀ nà-kakan`fò: of Ibadan, 75; of Ijaye, 163n8; of Ilorin, head of Oyo army, 147 Aríbúṣ ọ lá (she-who-makes-the-deep-aplace-of-honor), 51. See also ibú Aristide, Jean-Bertrand, 55 Aristotle, on commensuration, 119 Àsábá of Iroyi, 86 àṣ ẹ (Brazilian axé, power), 9, 13, 44, 70, 91; levels of (axé), 149; mothers’ collective manifestation of, 146–47; personal, 146; of postmenopausal mothers, 103; purity and (axé), 148. See also blood; Brazil; king(s); orisha; power; sacrifice(s) Àtá (king [ọ ba] in kingdom of Ayede), 91, 92, 137 Àtá Eshubiyi, 134. See also Elékọ̀ lẹ́ Àtá Omotosho II, 93 authority, opposed to power, 42–43, 42f, 49 awo (secrets), 82; political influence of, 83– 84. See also secrecy; secret(s) àwòrò (cult office/high priest), 71, 82–83; in Irefin, 86. See also cult(s); orisha; priests; religion(s) àwòrò Epa, 85 Axé (Àṣ ẹ ) Opô Afonjá, 147 axis: horizontal/vertical of Afro American culture, 106, 150; of substitution, 12; syntagmatic, 11, 13. See also culture Ayede, kingdom of, 2, 44, 71–72, 72m, 128–29, 132, 134; cult transformations in, 90–93; founding of, 134; link with Ibadan, 158n14 baálẹ̀ (chiefs of subordinate towns), 70, 73; as diminutive kings, 79, 80; of Ilafon and Ilemesho, 79, 87; of Ilemesho, 89. See also chief(s); elder(s) baálé (lineage elder[s]), 41, 86, 87 Baba, Ahmed, 123, 166n1

babalao (diviner, Cuban Lucumí), 14 babaláwo (diviner, Yoruba), 68 Bahia: concerned with purity, 168n31; emergence of Candomblé houses in, 144; Nagô-Yoruba politico-cultural framework in, 143–45; “true,” Oyobased, tradition brought to, 145. See also Brazil Ball, Charles, 119 Balógun (Ayede) war title, 124 Balógun Áòfin, chief of Ayede Omole-Akodi quarter, 92, 134 Bantu, syncretism with Portuguese Catholicism, 35 Barbados, 118. See also New World; slave(s) Barbara, Santa, 25 Barber, Karin, 124, 127–29, 150, 160n3 Barbot, John, 107, 164n16 Bascom, William, 6, 28, 71 Bastide, Roger, 20, 39, 135, 141–42, 143– 44, 148, 161n23, 167–68n27 batuques (secular dance-drum gatherings), 144. See also dance; music Beauvoir-Dominique, Rachel, 47–48 Beckles, Hilary McD., 117–18 Beliso-De Jesús, Aisha, Electric Santería, 6 Bello, Sultan Mohammed of Sokoto, 123 Bender, D. R., 126 Benin (formerly Dahomey), 11, 13 Big House (casa grande, plantation mansion), polygynous slave–master relations within, 114–15. See also great houses of Candomblé; plantation(s); polygyny, New World; slave(s); whites Bizongo society, 34; birth of, 161n16; and Bissagot slaves, 161n20; as extreme form of Petwo, 49, 57; forms coalition with tonton makout and VSN to protect Duvalier, 56, 57; leader of, quoted, 47; meaning of name, 47–48, 161n20; as opposite of Rada, 49, 52, 58, 59; opposites at core of, 59. See also Duvalier, François (“Papa Doc”); Haiti; Sans Poils black market, 118. See also markets, town and village blacks: appropriation of power by, in Caribbean, 36; “pollution” of, in Brazil, 148; “purity” through miscegenation, 148. See also casa grande; Negro(es); power; racism blessing: female fertility as ultimate, 104;

Index / 191 to pregnant women from orisha, 99. See also orisha; religion(s) blood: black Brazilian, “cleansed” through miscegenation, 148; mothers’, 97–121. See also blood, mothers’; casa grande; plantation(s) blood, mothers’: base of royal Candomblé line, 146; “cleansed” through miscegenation, 115; destructive when undermining husbands, 102–3, 104; embodies contradictions, 107–21; fetishized, 98–99; good vs. infertile or bewitching, 103–4; as ingredient to power, 146; manifests collective àṣ ẹ , 146–47; postmenopausal as “bad,” 103. See also mothers; witch(es); women bọ (worship), 89 Boas, Franz, 20 Bois Caïman, 46, 48, 58, 160n7 Boukman, 32, 46, 58 Bourdieu, Pierre, 127 Brazil, 20, 24, 26, 161n23; Nagô nations and houses in, 134; and syncretism of Christianity and Africanism, 35; Yoruba ethnicity reconstituted in, 123. See also New World; Yoruba; Yorubaland bridewealth, assures wives’ custody of children if divorced, 101. See also wives brotherhoods, black, 142–43; of Minas Gerais, 143; royalism within, 143. See also brotherhoods, Catholic brotherhoods, Catholic, 124, 135, 141; Candomblé house emerged from, 147; predominant in Candomblé, 145; trace slaves’ places of origin, 135; work on slaves’ behalf, 135. See also brotherhoods, black; Catholicism; irmandades Brousard, Carl, 56 Brown, David, 136, 138, 140–41 Bush, Barbara, 117 cabildos de nación (Cuban Catholic brotherhoods), 150, 167n13, 167n14, 167n19; defined, 140; evolve into casa templo, 140–41; of Lucumí, 139, 141; of Lucumí, “preserve” slave origins, 135–36, 138, 139, 152; and Nagô black brotherhoods, 142. See also brotherhoods, Catholic caboclos (mixed spirits), 148, 149. See also spirits

“Cake Wenches,” stereotype of enslaved women traders, 119. See also women calabash (igbá), as symbol, 44, 99. See also symbol(s) Candomblé, Brazilian, 18, 28, 32, 151, 160n12, 168n32; casa, 168n32; and “cool” and “hot” deities, 36; development of Nagô nation within, 141–49; emergence from Catholic brotherhood, 147; members of house, generated by axé (àṣ ẹ ), 147; Nagôization of, 147–49; orisha cults of, 96; and syncretism, 19, 25, 29, 44, 124; “true,” 141–42. See also house/home; New World; syncretism cannibalism, metaphorical, 102, 118–19 cantos (work crews), 144 Capone, Stefania, 4, 149 Carnegie Corporation, 21 carnival, meanings of, 137–38. See also ceremony(ies); orisha Casa Branca do Engenho Velho (Ile Iyá Nassô), original Candomblé house, source of “pure” Nagô branches, 145 casa de ocha, 167n22; Cuban cabildo de nación becomes, 150. See also house/ home; kingdom(s) casa grande (Brazilian Big House), European values of, 147–49. See also Big House; plantation(s) casa templo (Lucumí), 140–41 “Cash Madam(s),” as blocking money and blood, 104. See also women Catholicism: Africanized, 35; brothers and slave origins, 135; government in postcolonial Haiti, 33; religion of New World slaveholders, 25; requires baptized slaves to have godparents, 165n23; two, white and black, 142; “uneasy” in New World, 32; and Vodou, 32–33, 34. See also brotherhood, Catholic; religion(s); saints, Catholic; Vodou (Haitian) Catholic New World, 22 ceremony(ies): of cabildos, 135–36; for orisha, 163n13; semantics of naming, 130–31. See also number; orisha; religion(s) Césaire, Aimé, 39 chains, as mythic markers, 13. See also Oduduwa charms. See fetish(es)

192 / Index chiasmus, in creation myth, 11. See also myth(s) chief(s) (ìghàrẹ , ìwàrẹ̀ fà): civil, and kings, 69, 70, 73, 75, 83–84, 131–32; civil, and orisha cults, 41, 43, 71, 82–83; civil, and political role reversals, 42–43, 44, 75; functions of, 125; Ilafon lineage, 80– 81; of market women, 103; merge with priestly functions (fusion), 86–87; qualify for high ritual office, 163n8; ranked, in Ishan, 75–77, 76f, 77f, 78f; signal fission from orisha cults, 90; subchiefs, 73, 75, 76, 83, 86, 90. See also baálẹ̀ child/children, 71; mixed-race, 115; and stigma of loss of, 100. See also mothers; ọ mọ ; women choreography(ies), 136, 137. See also dance Christianity, 12; becomes royal Kongo cult, 143; enslaved woman converted to, 113–14. See also Catholicism Christophe, Henri, 32–33, 55. See also Haiti; Vodou (Haitian) Church Missionary Society, 12 “civilizing mission,” Europe’s, 31–32 clairvoyance, highest stage of initiation, 50, 160n11. See also initiation Clapperton, Hugh, 1 class, formation of, 17, 45, 69, 122, 135; and Rada-Petwo divide, 65 cleansing. See limpeza; purification/purity Cody, Cheryll Ann, 116 cofradías (Cuban Catholic brotherhoods), preserve slave origins, 135. See also brotherhood, Catholic; Catholicism; Cuba colonialism, British, 79–80 color, politics of, 54–55. See also blacks; mulattos, Haitian; racism; slave(s); whites Comaroff, Jean, 7; subversive bricolages, 163n8 Comaroff, John, 122 commensuration, as exchange, in trading contexts, 119 confiance, 50 congadas (Bantu sects), 142; origins and social structure of, 142 Congos (the people): preferred for house service, 63; “true,” 63. See also Kongo; slave(s) connaissance, and creolization, 39–67. See also imọ jinlẹ̀ ; knowledge, deep

Cosmas, Saint, 25, 26. See also saints, Catholic coups d’état, Yoruba, 42. See also power cowives, 100; and matri-segmentation, 101; and witchcraft, 100; as witches in plantation house family structure, 114–15. See also family(ies); witch(es); witchcraft; women creation. See myth(s) Creoles: Atlantic, produced by brotherhoods, 136–37; origins of, 59 (see also Dahomey; Kongo); revolutionary, identify with “low” Petwo line, 54, 59; threaten Brazilian whites, 148. See also brotherhoods, Catholic; creolization; miscegenation; Petwo (Petro) deities; Rada deities créolité, 39 creolization: and connaissance, 39–67; cultural, 162n28; dynamics of, 51–55, 63–64, 67; in Guiana, 20; and nanchon, 62–64. See also Creoles Crowther, Samuel Ajayi, 123 Cuba, 24, 26, 63; Lucumí and “nations” in, 134; as “sugar palace,” 138; and syncretism of Christianity and Africanism, 35; transformation in, 14; Yoruba ethnicity reconstituted in, 123. See also festivals; Santería (Cuban) cult(s), 2, 3; adaptability of, 44; in Ayede, 90–93, 92f, 93f; Congolese becomes Christian, 143; consolidation in Ayede, 90–93; in Ishan, 84–89; kingdom in, 147; of Marcelina-Obatossi, 146; Nagô-Yoruba Xango (Shango), 145; orisha of, 29–30, 58, 69–96, 132, 134, 137; political logic of, 41–42, 44, 45; royal, 42, 69, 70, 143; slaves “retain” in New World, 25; sociopolitical bases of, 41–42; of Yoruba priestesses, 137. See also orisha; politics; religion(s); “sib”/“multi-sib” (Bascom); and individual names culture, 12–13; African as pristine, 32; birth of Afro American, 105–21; meaning of, 155–56; over unicorns, 158n10; Yoruba, 7–8, 10, 13, 28, 68–96; Yoruba, critical reconstruction of, 152–54. See also cults(s); myth(s); religion(s); ritual(s); slave(s); Yoruba Curtin, Philip, 97

Index / 193 da Costa, Pereira, 142 Dahomean-Yoruba lands, 147; cosmology of, 60, 61 Dahomey (present-day Benin), 11, 36, 46; as origin of Petwo deities, 59; as origin of Rada deities, 47. See also Creoles; Kongo Damballa (Damballa Wedo/Ouida), 25– 26; Petwos recreate as (hot) Damballahflangbo, 59 Damian, Saint, 25, 26 dance, 46; as veiled resistance, 35; violent, makes birth of Petwo nation, 48–49. See also music Dancer, Thomas, 113 Danse à Don Pèdre/The Don Pedro, 48–49. See also dance; music daughters: care for aged mothers, 116; strictly exogamous, 101. See also family(ies); metaphor(s); mothers; women Davis, Natalie Zemon, 103 Davis, Wade, 57, 59, 67 Day of Carrying Water (Ijọ́ Ipọ nmi), 99, 137; paralleled with saint festivals and the Day of Kings, 137. See also festivals de Bahia, Salvador, 141 decomposition, of African origins, 152–53, 156. See also origin(s) de Heusch, Luc, 60, 61, 62 deities: Bizongo, 59; “cool” vs. “hot,” 36, 43, 46–47, 48, 59; demand animal sacrifice, 47; of orisha, 41; represent Ishan quarters, 81–82; town shrines house, 42; vudum, 144–45; white vs. warrior, 43; Yoruba, 29–30, 69. See also Bizongo society; gods, African; hegemony; Petwo (Petro) deities; Rada deities; religion(s) démembré(s), 51, 64. See also lakou Denis, Lorimer, 55 de Saussure, Ferdinand, 155 descent: ambiguity between agnatic and cognatic, 124–25; Barber on, 127–29; Bender on, 125–26; Eades on, 126; essential to Yoruba social structure, 126; intertwined with residence, 127; vs. residence, debate over, 127, 129. See also family(ies); house/home; lineage(s), Yoruba; mothers; Yoruba; Yorubaland Dessalines, Jean-Jacques, 32–33, 48, 55, 161n16. See also Haiti; Vodou (Haitian)

determinism, cultural, Herskovits on, 23. See also culture; Herskovits, Melville J. diaspora, African, 2–3, 4, 6, 11, 17, 23, 25, 66–67, 151, 154, 156; Old World origins of, 38, 124; research in, 40–67, 152–56, 157n3; as slaves in New World, 107–21; syncretism in, 34–36; truly African within, 37; Yoruba, 26, 28, 154. See also slave(s); Yoruba divorce, 101. See also marriage, always exogamous djab (spirit devils), 46 dọ̀ bálẹ̀ (prostrate oneself), 84 Dogon, 45 dowry. See bridewealth drums. See music Du Bois, W. E. B., 39 Duvalier, François (“Papa Doc”), 4, 32, 41, 161n18, 161n19; as Machiavellian, 37; regime in Haiti, 55–58; religion under, 33–34, 55–56; support of, 56. See also Haiti; Vodou (Haitian) Eades, Jeremy S., 126, 127 ẹ bọ ra (spiritual power), 85, 94. See also power ẹ gbẹ́ àgbà (titled elders), 76. See also baálé; elder(s) ẹ gbẹ́ associations: as coven of witches, 105; women’s means of controlling market, 103–5. See also witch(es); women egum (dubious ancestors), 148, 149. See also miscegenation Ejimọ ko (Ishan kings’ àwòrò), 83, 84. See also àwòrò ejio (council of àwòrò), 83. See also àwòrò; Oníṣ àn; orisha èkejì òrìṣ à (second to the gods), 69 Ekiti Yoruba, 1, 15, 28, 41, 68–96; history and cults/rituals, 93–94. See also Yoruba; Yorubaland elder(s): godfather/godmother as, 140; initiation of, 45; lineage based, 41, 78, 81, 86, 101; political, 1, 75, 76; power of market women as, 103; priest as, 140; priests represent, 86; ritual, 141. See also baálẹ̀ ; chief(s); politics; power elders, female, as traps of menstrual blood, 103. See also women ẹ lẹ́ gbẹ́ (various roles), 77–79, 78f Elékọ̀ lẹ́ (Àtá Eshubiyi’s town of origin), 134

194 / Index Ẹ lẹ́ mẹ̀ ṣ ọ , baálẹ̀ of Ilemesho, 87 emancipation of slaves, consolidates Lucumí, 139–40. See also Lucumí; markets, town and village; slave(s) endogamy, in Brazilian casa grande, 147. See also miscegenation Epa cult, 84–85, 84f. See also cult(s); orisha Epa-Ilale (Ekunrin), 84 epistemology: extremes, 7; and ontology, 7–9 eréko (agricultural area), 73–74 Escalera, La, 35, 46 Eshubiyi (warlord and king of Ayede), impact on Ayede’s political/ritual structure, 91–92, 132, 134 Eshu/Eshu-Elegba, 25, 28 essentialism, 153 ethnicity: as “other” oriented, 122; of slaves ambiguous, 135; Yoruba, 122–23 ethnogenesis, Yoruba, 122–52, 154–55. See also Yoruba; Yorubaland “ethnohistorical method” (Herskovits), 19 Europeans. See whites Ewe, and Fon and Yoruba religions, 162n26. See also Fon; religion(s); Yoruba ẹ̀ wọ̀ n (chain), 10 exchange, market, as setting for racial equality, 119–20. See also markets, town and village; racism; slave(s); women exoteric/esoteric knowledge, 44. See also knowledge, deep èyémọ lẹ̀ (cult office), 71. See also cult(s) Fadipe, Nathaniel A., 134 familia de santo, 140. See also family(ies) family(ies): broken up, 115–16; cowives in, 100; emerges from orisha cult house, 134; as metaphor within ritual kinship, 71, 81, 140; patriarchal, 100–101; and polygyny, 99–100; shared genealogical substance in, 146; slaves’, 110, 115–16; as town, 89; witchcraft in, 114–15, 116. See also house/home; slave(s); witchcraft; women Fernandez, Jim, 3 fertility: low rate of, in New World, 112; West African meanings of, 99–105. See also blood; blood, mothers’; infertility; women

festivals: Catholic and African, 137; of orisha, 41, 43, 99, 104, 137, 157n5; of orisha, as assault on slavery, 138. See also orisha; ritual(s) fetish(es), 50; of state, 55 fetishism: in Afro-Atlantic religions, 157n4; of commodified labor, 53; of the foreign, 67; of women’s blood, 98 fetus, “blocked” in “Cash Madams,” 104. See also infanticide filhas de santo (vestals of the sacred fire), preservers of Nagô, 144. See also Nagô nation; religion(s); Yoruba fission, 53, 61; family, 115–16, 120, 140; and lineage, 52, 73, 87, 100, 101–3, 102f, 104, 105, 110, 112, 114, 141, 146– 47, 160–61n13; of orisha cults, 90; Petwo entails, 58; political, 14, 29–30, 44, 73, 88–89, 90, 128, 131–32. See also family(ies); fusion; slave(s) Fon, 31, 159n10; god(s) of, 25–26, 28, 141; origin of name, 159n10; and Yoruba, 28, 65, 66, 106, 139, 143, 162n26; and Yoruba slaves, 143. See also gods, African; religion(s) Frazier, E. Franklin, 19, 40 Freyre, Gilberto, 37 Frieda, Erzulie, 53 Frobenius, Leo, 1 funfun (white, white deities, cool), 43, 149. See also Nagô-Candomblé lines; racism; whites fusion: of chiefdoms with orisha functions, 87; cultural, 25; and lineage, 141; political, 29–30, 76–77; religious, 85, 93, 152; and syncretism, 167– 68n27; theological, 31. See also fission; lineage(s), Yoruba; power gender. See womanhood, in New World; women genealogy(ies), 13–15, 30, 39, 67, 125, 141, 145; of Brazilian terreiros, 146; as charter, 13, 44; cultural, 20–21, 28, 37, 153; dynastic, 37, 42, 74, 125, 150; as geographic, 128; of gods, 29; of Haitian lwa, 59; idioms of, 149; of initiatic families, 146; of “many Yorubalands,” 68; and meaning of ilé, 126; as metaphor, 71, 89, 140–41; of orisha, 134; of Petwo

Index / 195 spirits, 48; ritual, 11; Yoruba, 144, 149, 150. See also family(ies); lineage(s), Yoruba; metaphor(s) Ginen (Guinea), 46, 49, 51, 59, 64; connaissance linked to, 50; deities span entire spectrum, 61, 62, 62f; opposite of Petwo, 50, 52, 57, 161n22; and Petwo/ Rada distinction, 61, 62, 62f, 160n7; realm of, under waters, 51. See also Petwo (Petro) deities gods, African, 25–26; associated with revolution, 45–46; confused with Catholic saints, 28; cosmology of, 60; vs. “Creole” gods, 54; Fon, 25–26, 28, 141; Ginen, elite, 61; paralleled with Catholic saints, 18, 19, 21, 25–29, 27t, 34–37, 45; Petwo, 60; renamed with European names, 59; West African vary, 30; Yoruba, 25–26, 29, 60. See also deities; Petwo (Petro) deities; Rada deities; religion(s); Vodou (Haitian) Golden Age, theological, 28, 37 government(s): of Duvalier, 55–58; Ishan, structure of, 74–81, 76f, 77f, 78f, 79f, 80f, 87; orisha cult structure modeled on secular, 134; town shrines as, 134; Yoruba, 42, 42f. See also king(s); kingdom(s); orisha; politics; power; quarter(s); town(s) Gray, Deborah, 113–14 great houses of Candomblé, 145; maintain “pure” Nagô-Ketu-Oyo tradition, 148, 149. See also Big House; casa grande; house/home Griaule, Marcel, 45 Guiana: Africanism “purest” in, 20; creolization in, 20. See also Ginen (Guinea) Guinée, 18, 34, 52. See also Vodou (Haitian) Haiti, 20, 24, 26; black elites’ rivalry with mulatto elites, 54–55; history of, 55; land tenure in, 51; loa in, 30–31, 46–47; Rada as elites in, 54; Vodou in, 32–34, 55–56. See also religion(s); Vodou (Haitian) hegemony: deep knowledge opposes, 37; Oyo’s, 1, 11; Petwo opposition to, 61; Rada association with, 61, 64; unmade and remade through religious syn-

cretism, 35–36; Yoruba, 37, 153–54, 167n26. See also Catholicism; power; whites hermeneutics, 2, 3, 23, 31; of power, 32; of revision, 37, 44, 45. See also Herskovits, Melville J. hero, ancestral, becomes a god, 13. See also ancestors; gods, African Herskovits, Melville J., 2, 3–4, 97; on African acculturation in the New World, 39; on conscious and unconscious retentions, 23–24; criticized, 17–38; on syncretism, 17–38. See also culture; religion(s); syncretism histories, oral, 1, 74–75, 158n14. See also myth(s) hounfor, 49 houngan (Vodou priest), 47, 48, 52, 57 household: African structure maintained in New World, 110–11, 111f; composition of, 126; Eades redefines, 126. See also family(ies); house/home; ilé house/home (ilé): Afro-Catholic irmandades as antecedents to Nagô, 144–45; ambiguous definition of, 124, 126–27; ancestors as “root founders” (raíces) of, 140–41; Barber on, 127; caboclo Candomblé, 144; caboclo Candomblé, source of “pure” Nagô branches, 145; chiefs of, 125; growth into kingdom, 131–32; as microcosm of kingdom, 131, 149; system of Lucumí, 140–41; as town, 131; town becomes, 150; town in, 127–29, 131, 132, 141, 147, 149–50; as town shrine, 132. See also chief(s); family(ies); household; ilé; ilé òrìṣ à; kingdom(s); town(s) hungenikon (Vodou choirmaster), 50 hunsi (Vodou initiate), 50 Hurston, Zora Neale: on Haitian loa, 30– 31; on Rada vs. Petwo, 47, 49, 54 husbands: polygamous, 100; supply seed money for wives in marketplace, 103. See also men; wives Ibadan(s), 2, 91, 134, 137, 167n12 Ibeji (Yoruba twins), 26 ìbejì dolls, for atonement for abortion (in colonial Saint Domingue), 114 Ibitoye, Olu, 5, 167n12

196 / Index Ibo, 63 ibú (the deep), 51. See also Aríbúṣ ọ lá icons, religious, 26, 138. See also cult(s); religion(s); ritual(s) identity(ies): invented, 18; pan-Yoruba, 124 Ìdì (tied up), as fetishized text (Qur’an, Bible) in Oduduwa myth, 12 ìdílé (lineage), 41, 73, 101, 124 Ifa (Fa), 26, 27t, 28 Ife, 1, 10, 11, 69, 74, 75, 79; as foil to Oyo, 11 igbá (calabash), 70. See also calabash (igbá), as symbol; metaphor(s) igbá Yemoja (Yemoja’s calabash), 44 igbó imọ lẹ̀ (bush shrine), 82 Ighare (ìghàrẹ ): (town chiefs), 75, 82; chiefs and quarters, 89, 89f. See also chief(s); town(s) ìjinlẹ̀ (dangerous and deep), 9. See also jinlẹ̀ ; knowledge, deep Ijọ́ Ipọ nmi (Day of Carrying Water), 99, 137. See also festivals ìka (hand), 85 Ikole, 134 Ilafon, 75, 79–81, 80f, 84, 86; Iro cults of, stolen, 88; lineage of officials, 87. See also Ilemesho (Ishan town) ìlárí Shango priests, 70. See also Shango ilé (casas/houses), 123–24, 141; Lucumí, 141; as manifestation of ìlú, 131; multiple applications of term, 124, 126, 131; towns located within, 129, 131, 132. See also household; house/home; town(s) ilẹ̀ (earth), 11, 14. See also Ogboni Society Ile-Ife, 10, 11, 68, 73, 74, 131; locus of original kingship, 132 Ile Iyá Nassô (mother of founding priestess of Casa Branca do Engenho Velho), 145 Ilemesho (Ishan town): baálẹ̀ of, 89; politico-ritual structure of, 79–81, 87– 89, 88f, 89f; power of orisha in, 89; rotating titles in Oke-Ade quarter, 88f. See also Ilafon ilé òrìṣ à (house qua town shrine), 132. See also household; house/home; ipara; shrine(s); town(s) ìlú (kingdoms/towns), 14, 29, 41, 124, 129, 132; ambiguity of meanings, 81, 124. See also kingdom(s); town(s) imọ jinlẹ̀ (deep knowledge), 2, 30, 37. See also knowledge, deep

impotence, male, caused by witchcraft, 100. See also infertility infanticide, 112, 117; caused by witchcraft, 113–14. See also witch(es); witchcraft infants, suffer from mother’s malnutrition, 112–13. See also child/children; infanticide; mothers; women infertility, 100; punishment for, 112. See also fertility; women inheritance, 160–61n13; agnatic lineage, sons opt into mothers’, 124; patrilineal, source of discord, 101–2. See also family(ies); matrilineality; patrilineality initiation, 14, 140, 146, 147; of filhas de santo, 144; of priests, 50; as prix-des-yeux, 160n11. See also ritual(s) ìpàdé (meetings), 83 ipara (orisha cult-house), 82, 91, 94f. See also shrine(s); town(s) Irefin, three major orisha cults within, 84– 87. See also cult(s); orisha irmandades (Brazilian Catholic brotherhoods and sisterhoods), 150; “preserve” slave origins, 135, 141; as roots of Nagô line, 144. See also brothers, Catholic; Catholicism Ishan, kingdom of, 2, 74–89; cults in, 81– 89; politico-ritual structure of, 76f, 77f, 78f, 79–90, 79f, 80f, 84f, 85f, 88f, 89f Itaji, kingdom of, 2, 71 itàn (oral histories), 74. See also òpìtàn Ivy, Marilyn, 13 ìwàrẹ̀ fà (civil chiefs), 70, 73, 75, 125. See also chief(s); Ighare ìyá (mother), 71. See also mothers; witch(es); witchcraft; women Ìyá Ẹ pa (mother of Epa), 85 Iyagba, warrior priestesses of, 137 Ìyá ìbejì (mother of twins), 100 Ìyá ilé/ìyálé (first wife), 100. See also family(ies) Ìyálòde (female chief of market women), 103 Iyá Nassô, 168n28 Iyá Omi Ase, 147 ìyàwó (wife), 71, 115. See also cowives; family(ies); women Jackson, Walter, 23 Jamaica: abuse of slaves in, 116; loving bonds with masters in, 115; patrilineality in, 111–12. See also New World

Index / 197 Jim Crow, 39 jinlẹ̀ (deep), 68. See also ìjinlẹ̀ ; imọ jinlẹ̀ Johnson, Paul Christopher, 4, 145–46, 149 Johnson, Samuel Ajayi, 1, 6, 10, 12 jùjú medicine, 98, 99. See also medicines Kant, Imanuel, 6–8; relativized, 9 Ketu kingship, 145 Kiddy, Elizabeth W., 143, 167n23, 167n24 king(s) (ọ ba), 5, 10, 42, 43, 73; baálẹ̀ as, 80; and chiefs, 75, 83–84, 89, 125, 131– 32; Ishan, descent of, 75; as “master of ceremonies,” 140; and orisha, 69–72; paraphernalia of, 140; power exceeded by Ejimọ ko, 83–84. See also chief(s); kingdom(s); kingship; ọ ba; orisha; politics; power kingdom(s): àwòrò helps regulate, 83–84; in cult, 147; cults in, 29–30, 69–72, 147; development of, 72–74; development of, and Yoruba ideology, 74; expansion and shrinking of, 131–32; ilé as microcosm of, 131, 149; Ishan and Ayede compared, 74–96; miniature, 140, 141, 142; orisha in, 74–96; quarters (àdúgbò) of, 81–82; repossessing towns, 150; ritual houses as, 132; shrines as ritual, 13, 134; shrinking into houses/towns, 132, 140, 141; towns become, 72–73, 132. See also family(ies); house/home; king(s); kingship kingship, 11, 12, 167n20; African, and congadas, 142; crowns of (adé Olókun), 79; dialectics of, 44; house expands into, 131–32; Ile-Ife as origin of, 132; Ketu, 145; Oyo, 145; system reduplicated in cult rituals, 139; Yoruba, 167n22. See also king(s); kingdom(s); monarchy; myth(s); Oyo kinship, 80; based on places of origin, 128– 29; Hawaiian type, 166n8; ritual, and collective power of terreiro, 146; ritual, metaphors within, 71, 81; social goes to ritual (Lucumí and Candomblé), 140, 149; structure of Lucumí houses (casas), 140–41. See also family(ies); household; house/home; kingdom(s) knowledge, deep (imọ jinlẹ̀ ), 30, 37–38; carried by slaves from Africa, 50; Petwo and, 58; revisionary logic of, 41–46; ritual, 160n12. See also connaissance,

and creolization; hegemony; Petwo (Petro) deities; power kola nut, as symbol, 86, 138, 167n17. See also symbol(s) Kongo, 49–50; as origins of Petwo deities, 59–66; ritual enclaves in, 61. See also Creoles; Dahomey (present-day Benin); Petwo (Petro) deities lactation/labor duality, of African women, 107. See also mothers; women Lagos, 123 Lagos Colony, 12. See also Nigeria Lagosian Renaissance, 123 láíláí (time immemorial), 9, 88, 90 lakou, 51, 64, 161n16. See also démembré(s) Lamurudu, 12 land acquisition: among démembrés and lakous, 52–53; and lineage, 53 Lander, Richard, 1 langaj (Vodou ritual language, possessed speech), 48 language(s): about, 8; of race, 21; Yoruba idiom, 166n35. See also Yoruba; and individual examples la-place (master of ceremonies, Vodou), 50 Larose, Serge, 51–52, 53–54, 57, 58, 63, 161–62n24 Lavalas movement, 55 Law, Robin, 123 Legba, 25–26 Les Griots (magazine), 55 Lewis, Matthew, 116 limpeza (cleansing of “blood”), 149. See also miscegenation; purification/purity lineage(s), Yoruba (ìdílé): fission and fusion in, 29–30, 141; as households led by chief, 100; in Ilafon, 80–81; in Ogilolo, 87; optation by males as “cannibalistic,” 102; ritual, of Lucumí, 140–41; segmentation and fission of, 102, 102f; structure of, 80–81, 101, 125. See also descent; family(ies); house/home; inheritance; orisha; Yoruba Lloyd, Peter, 43, 125, 126 logic: of orisha political organization of cults, 41–42; of Yoruba quantification, 129–32 L’Ouverture, Toussaint, 32–33, 48. See also Haiti; Vodou (Haitian)

198 / Index “lowness thesis,” “heterodox” Petwo vs. “hegemonic” Rada, 54, 58, 61. See also Petwo (Petro) deities; Rada deities Lucumí, 8, 9, 14, 18; consolidated by emancipation, 139–40; influence in New World, 139–40; origins of, 140–41 lwa (loa) (gods and goddesses), 29, 30, 31, 46, 47, 50, 51, 52, 54, 55, 59, 60–61; hot/cool distinction among Haitian, 60. See also gods, African; orisha madrina (godmother), 140. See also family(ies); slave(s) magic, 51; “confronting Guinea,” 52 Magloire, Clement, 56 Makandal, François, 58 Malinowski, Bronislaw, 22 mambo (Vodou priestess), 50. See also priestesses; priesthood; priests; priests and priestesses; religion(s); Vodou (Haitian) manumission, possible for slave women employed as marketers, 118 Marassa twins, 25, 26 Marcelina-Obatossi (founder of Casa Branca do Engenho Velho), 145, 146 markets, town and village: immorality against women sellers, 166n34; women control (can block), 103–5; women control, in New World, 117–18. See also slave(s); women maroons, nomadic, 161n16 marriage, always exogamous, 101. See also cowives; polygyny, African; polygyny, New World; women Marx, Karl, 153; on commensuration, 119 masculinization: of African women, 103–5; of slave women, 116–17. See also markets, town and village; women mask(s), Epa, 84f, 85f Matory, J. Lorand, 4, 12, 15, 105, 123, 124, 145, 146, 147, 148, 149, 150, 155 matrilineality, 116, 161n15. See also mothers; patrilineality; women matrix, cultural, 14; genealogical, of Lucumí, 140–41. See also family(ies); house/home; origin(s) McDaniel, Hattie, 109f. See also slave(s); women McDonnell, Lawrence T., 119 Mecca, 12

medicines: as ingredient to power, 146; jùjú, 99. See also blood; power men, medicines neutralized by menstrual blood, 98. See also family(ies); husbands; jùjú medicine; patrilineality; polygyny, African; polygyny, New World menopause, liberates women to market work, 103 menses, 98; cycle reiterated by aged market women, 104; trapped within aged marketwomen, 103. See also blood; fertility; women mestizaje, 39 mestizos, 149. See also mulattos, Haitian metamorphosis: of humans to animals, 47–48; political implications of Petwo, 48. See also animals; religion(s); sacrifice(s) metaphor(s): cannibalism as, 102, 118– 19; in ritual kinship, 71, 81. See also symbol(s) Métraux, Alfred, 46, 50, 51, 54 midwives: blamed for defective infants in New World, 113; punished for abortions, 114. See also child/children; womanhood, in New World; women Mina, 167n24 Mintz, Sydney, 97 miscegenation: as “cleansing” of slave blood, 115, 147–49; effects of (Herskovits), 20; fear of, 119; threatens Brazilian whites, 148. See also Creoles; polygyny, New World; racism; whites missionaries, Catholic, resonate in Congo, 143. See also Catholicism; religion(s) mocambos (black republics of escaped slaves), 35 Moitt, Bernard, 112, 113 monarchy: of Minas Gerais, 143; miniature, 142; origins of Yoruba, 10. See also king(s); kingship; Oduduwa; politics; Yoruba Moreau de Saint-Méry, Louis Elie, 48–49, 63–64, 114 Morgan, Jennifer, 107 Morgan, John H., 113 Morgan, Louis Henry, 166n8 Morgan Report on “traditional chieftaincies” (Nigeria), 89. See also Nigeria motherhood: African vs. Victorian models, 165n22; in New World, 112. See

Index / 199 also New World; womanhood, in New World; women mothers: aged, as trapping menstrual blood, 103; aged, cared for by daughters, 116; as agents of lineage discord, 102, 104; blood of, 97–121; masculinization of, 116–17; sons opting into agnatic line of, 124; strictly exogamous, 101. See also blood; family(ies); matrilineality; witch(es); witchcraft; women Mudimbe, Valentin Y., 31–32, 65, 67 mulattos, Haitian, 33. See also blacks; Haiti; mestizos; Negro(es) “multi-sib cults,” 29, 71, 85. See also cult(s) music, of (talking) drums, 92, 136 myth(s): creation, 10, 11, 12, 74; migration, 10–12; and origins of king/kingdoms, 74; of re-Africanizing Nagô-Candomblé lines, 145; substitutions in, 12 Nagô-Candomblé lines: maintaining “purity” of, 148; re-Africanized, 145. See also Candomblé, Brazilian; Nagô nation; purification/purity Nagôization (1870s–1930s) of Candomblé, 147, 148, 149 Nagô-Ketu nation, three houses of, 146 Nagô nation, 18, 141–49, 168n30. See also Yoruba Nagô Ogun deities, 60 Nagô-Yoruba Xango (Shango) cult, 145 names, many, of Casa Branca do Engenho Velho, 145. See also naming naming: politics of, 155; Yoruba ceremony distinguishes one from the many, 130– 31; Yoruba hyphenated like Spanish, 139. See also names; number nanchon (nation), 62–64. See also Haiti; Petwo (Petro) deities; Rada deities Nansourkry, 61 négritude, 39; Vodou as model of, 36. See also noirisme; power; Vodou (Haitian) Negro(es): darker-skinned more “African” than lighter-skinned, 20–21; Herskovits on, 19. See also blacks; power; racism; slave(s) New World, 158n1; African adaptation to, 164n15; plantation societies in, 105–21; relationship with West Africa, 37; religion in, 95, 145; retention of Africanism within, 23, 24, 32, 110, 117–18, 119–20,

166n35; women dominate markets in, 117–18, 119–20. See also Africanism/ Africanity; Africanization; religion(s); slavery; syncretism; Vodou (Haitian) Nigeria, 9, 13; march to independence, 37; Yoruba pantheons vary within, 30. See also religion(s); Yoruba; Yorubaland; and individual place-names Nimrod, 12 noirisme, 39, 55, 56, 58, 67. See also négritude North American Lowcountry, ix number: English system (composition), 130; semantics of Yoruba (dividedness/ decomposition), 130 Nupeland, 10, 11 ọ ba (king), 5, 10, 42, 43, 69, 141. See also king(s); kingdom(s); obá-oriaté Ọ baísà, chief, 75, 86; explains necessary distance from cult functions, 82–83 Obá Lucumí, palace of, 167n22 Ọ ban´lá (Abuda àwòrò), 85, 86 obá-oriaté, 167n20 Ọ basọ́ lọ̀ (Irefin civil chief), 86 Obatala, 30, 36, 43 ocha (orisha), 140, 141 Ọ̀ dọ̀ fin (Irefin civil chief), 86 odù (highly coded divination verses), 11 Oduduwa, 10; ancestor of Yoruba kings, 74; chain of, 10–14. See also king(s); myth(s); Yoruba ọ̀ gbọ́ n (subquarters), 86 Ogboni Society, 59 Ogilolo, cults within, 87. See also orisha Ogun (Gu), 18, 26, 27t, 28, 41, 43, 86. See also Nagô Ogun deities Ogun-Badagary (Ginen), 59. See also deities; gods, African; Nagô Ogun deities ọ jà (market[s]), 80 Ojo-Apapo (festival of Ishan’s five quarters), 83. See also cult(s) ojú ẹ bọ (sacred altars), 82 Oke-Ade quarter, rotating titles in, 88, 88f ọ kọ (husband), 71 Okuku, Barber on social structures in ilé of, 127–29 Olodumare, 10, 12 Ológun (representative of all age sets in Ishan quarter, ritual and political), 77–79, 83, 87, 93, 138. See also orisha; politics; power

200 / Index Olokun, 43, 44 ọ lọ́ ọ̀ tun (possession priestesses), 99 olórí (head, as in leader), 86 olórí ẹ gbẹ́ (representatives), 81 Olua (house), 132, 133f Olua (Oníṣ àn’s orisha), 83 Olua orisha (in Ayede), 133f Olúyọ̀ lé (military chief of Ibadan), 91, 134 Omi ([sacred] water), 51, 99 ọ mọ (child), 71, 81–82. See also child/children ọ mọ baba (half-brother), 126 ọ mọ baba kan (children of one father), 73, 124 ọ mọ Epa (children of Epa), 84 ọ mọ ìyá (children of my [or one] mother), 5, 73, 101; as full siblings, 126 ọ na (roads), 68 Oníjòfì, 87 Onírèfin, 84, 86 Oníṣ àn (ọ ba) (king), 75 ontology, and epistemology, 7–8, 9, 10 Ọ́ ọ̀ ni, 74 òpìtàn (historians), 91. See also itàn Oranyan, migration myth, 11 orientations, cognitive (Mintz and Price), 66 origin(s): African, 39, 40–67, 152–53, 156; of (Nagô) cabildo, 142; chains as markers of, 13; of Cuban Lucumí-Santería religion, 140–41; cultural, and syncretism, 24–31; mythic (see myth[s]); of Petwo line, 59; places of, 128; of Portuguese congada, 142; slaves’ places of, unknown, 135; tribal, 97; of Yoruba kings, 74. See also brotherhoods, Catholic; myth(s); religion(s) oríkì (praises), 13, 45, 68, 71, 74, 93, 95, 127, 131, 137, 150; of strangers, 124 oríkì orílẹ̀ (praise poetry on origins), 127, 128, 129, 150 Ọ̀ rírẹ́ of Alewa, 86 òrìṣ à funfun (white deities), 43 òrìṣ à-ocha system, 139 orisha (òrìṣ à/orixás/orichas), 86–87; African and New World (Verger), 69; altars of, 160n3; as ancestors become deities, 70– 71; association with “cool” and “hot” deities, 36, 45, 49, 137–38; ceremonies for, 95–96, 163n13; consolidates in Iyagba cult, 93; consolidation of, in

Ayede, 90–93, 92f; cult organization of, in Ayede, 90–93, 138; cult organization of, in Ishan, 81–82, 90; deities of, 41; devotees of, 70–71; as one and many, 132; origin of, 70–71, 130; “owned” by chiefs, 86–88; and political power, 43, 45, 69–96; secrets of, 82; tradition, 18; water, as mothers’ blood in, 146–47; worship of, 1–2, 5–6, 11, 30, 31, 69–96, 150, 151; worship of, vs. ancestor worship, 71. See also cult(s); deities; politics; religion(s); ritual(s); syncretism orisha Adodo (for Irasa and Ipala lineages), 87 orisha Alagba (for Egenon lineage), 86 orisha Alakua, 87 (for Irunsin lineage), 87 orisha Atoru (for Ibori lineage), 87 orisha Idara (for Idara lineage), 87 Orisha-Iroyi, 86 Orisha Iyagba (Iyelori), 92, 93, 95, 137; nonroyal cult, 134; town shrine (ipara) of, 94f Orishanla (Obatala), 69; first orisha, 130 orisha Obanifon (for Isaoye lineage), 86 orisha Ogun (for Ibori lineage), 87 orisha Ojuna, 91–92, 95; cult of, 134; deities and orisha in, 92, 93f orisha Yemoja, 91 Ortiz, Fernando, 37 ọ̀ run (heaven), 10, 11, 74 Orunmila, 130 Oshun, 43 Our Lady of the Good Death (Nossa Senhora da Boa Morte, sisterhood), 145 Our Lord of Martyrs (brotherhood), 145 ọ wọ́ (hand), 85 Oyo, 1, 11, 37, 69–70, 73, 122, 128, 137; head of army, 147; head of army, rebels against king, 147; Ife as foil to, 11; kingship, 145; as “Yoruba Proper,” 123, 166n1 Oyo-Yoruba, as “Yoruba” linguistic and cultural model, 28 padrino (godfather), 140. See also family(ies) “paganism,” 5 paket kongo, 50 palace(s): details of, 132, 167n11; of Obá Lucumí, 167n22; ritual houses as, 132. See also house/home; ritual(s)

Index / 201 Palmares, 35, 46 Palmié, Stephan, 8, 9, 12–13, 14 paradigm, syncretic (Herskovits’s), 20–24, 25–26, 27t, 28–29. See also Herskovits, Melville J.; syncretism paradox, Petwo, 46–51, 67. See also “lowness thesis”; Petwo (Petro) deities; Vodou (Haitian) Parés, Luis Nicolau, 144–45, 147 Park, Robert E., 19 patriarchy, “threatened” by miscegenation, 118. See also family(ies); marriage, always exogamous Patrick, Saint, 25, 26 patrilineality: in naming of slaves, 116; in New World, 111–12; in Yoruba social structure, 126. See also matrilineality; men; New World Patterson, Orlando, 25 Pax Britannica, 79 peasantry, landless, 53 Pedro, Don (Dom), 48–49, 59, 160n10 Peel, John D. Y., 12, 155 performance(s): cabildo, Day of Kings, 137; orisha, of Ayede-Ekiti, 137 Pétion, Alexandre, 48, 55 petro, 46 Petro, Jean, 48 Petwo (Petro) deities, 36, 40, 59–60, 162n26; debate over provenance from Kongo or Creole deities, 59; and deep knowledge, 58; harnessed by Duvaliers to prevent bourgeois revolution, 56–58; Kongo elements in, 61–62; opposed to Rada deities, 46–47, 49–50, 51, 58, 60, 62, 62f, 63–64, 160n8, 165n26; origins of, 59–60, 61; paradox, 46–51; paradox, solution to, 54, 58; repository for all non-Rada deity families, 62–64; resemble West African secret societies, 59; role reversal with Rada, 63–64; and syncretism, 60. See also deities; gods, African; power; Rada deities; religion(s); syncretism; Vodou (Haitian) “Petwo sauvage,” transmogrified wanderers, 53. See also Petwo (Petro) deities plantation(s): as patriarchal corporate group (ẹ gbẹ́ ), 114; predominant in Candomblé houses, 145. See also Candomblé, Brazilian plantation society, in New World, 105–

21, 163n2. See also motherhood; New World; women planters. See slaveholders, New World plantocracy, sugar, 145, 147, 149. See also plantation(s); plantation society, in New World; slave(s) plants, abetting abortion, 113 poetics, Jakobsonian, 11 polink (provision grounds of enslaved Jamaicans), 117. See also Jamaica political action, bases of, 46. See also politics; revolution politics: of color, 54–55; and orisha worship, 70–72, 72m, 82; parallel in Yorubaland and Bahia, 96; of revision, 47– 48; of Yoruba government, 27t, 41–43, 44. See also chief(s); hegemony; king(s); kingdom(s); orisha; power; ritual(s); Yoruba polities, precolonial, 72–74. See also kingdom(s); politics polygyny, African: and families, 100; and matri-segmentation, 101, 101f. See also polygyny, New World polygyny, New World: Herskovits on, 21, 22; between master and enslaved women, 114–15; among slaves, 165n29. See also cowives; marriage, always exogamous; polygyny, African poor (people): rural/urban, 53, 55; as source of political power, 55 Portugal, 142 power: of àwòrò vs. ìghàrẹ chiefs in Ishan, 82–84; blacks appropriate, 36; categories of (class-determined), 36; Catholicism as, 34–36; class-based (Haiti) vs. politically based (Africa), 54; “cool” and “hot,” 36, 44, 137; of deep knowledge, 44–45, 50; of deities, 45–47; elites reclaim, 37; fight for, 14; hermeneutics of, 2, 32; Ife’s, 10; ingredients of, 146; of Ishan Ejimọ ko, 83–84; of mothers’ blood, 97–121; opposed to authority, 49; of orisha, 69–96; of Petwo vs. Rada, 49, 51, 53, 58, 60, 62, 62f, 64; in Vodou, 46–51; of women working in markets, 103–5, 117–19; in Yoruba politico-ritual system, 41–43, 42f. See also blood, mothers’; Catholicism; hegemony; politics; polygyny, African; polygyny, New World; slaveholders, New World; whites; women

202 / Index Price, Richard, 97 Price-Mars, Jean, 33, 36 priestesses: possession by Yoruba (ọ lọ́ ọ̀ tun), 99, 137; Shango, 134; Yemoja, 134, 167n12. See also orisha; priests; priests and priestesses priesthood, accession of devotees to, 50. See also imọ jinlẹ̀ priests: hold deepest form of recollected knowledge, 50–51; “hungan-macoutte,” 160n12; and “upward mobility,” 149; Vodou, 52. See also connaissance, and creolization; imọ jinlẹ̀ ; knowledge, deep; orisha; priestesses; priests and priestesses; religion(s) priests and priestesses: go to lwa to increase powers, 51; iconography of, 70; represent multiple entities, 132. See also mambo (Vodou priestess); orisha; priestesses; priesthood; priests prise des yeux, la, 50; “prix-des-yeux,” 160n11 processions, cabildo, 135; of Catholic virgins and saints with orisha outings, 138 purification/purity: within dominant Candomblé temples, 169n36; as reAfricanization, 148; ritual, standards of, 145, 147–48, 149, 168–69n35; Yoruba, 169n36. See also whites pwen (points or manufactured deities), 46, 52 quantity (countability), emergence of concept, 147, 168–69n33 quarter(s) (àdúgbò), 124, 131; àwòrò of, 83– 89; in Ayede, 91–93, 128; cults among, 81–82, 94; defined, 1; deities vested in, 30; in Ishan, 81–89; kingdom shrinks to, 132; king embodies, 70; lineages and, 125, 132; orisha and, 30, 41, 71, 82, 96; orisha cults within, 85–89; and political segmentation, 71, 73, 74; power transforms, 42; rank order of Ishan, 75–79, 76f, 77f, 78f, 79f; subquarters, 75, 77, 82, 84, 85, 86; as towns, 125. See also chief(s); king(s); kingdom(s); orisha quilombos (black republics of freed slaves), 35, 46 races. See blacks; casa grande; miscegenation; mulattos, Haitian; whites racism, 19, 23–24; in Brazilian casa grande,

147–48; among Europeans, 12, 13–14, 169n36; among white women, 165n31. See also slave(s); whites rada, 46 Rada deities, 36, 161–62n24; as conservative, 49; as elites, 53–54; hot and cold among, 60; opposed to Petwo deities, 46–47, 49–50, 51, 58, 60, 62, 62f, 63– 64, 160n8, 161n21, 165n26; perspective on Petwo and Bizongo, 58; role reversal with Petwos, 63–64. See also deities; gods, African; Petwo (Petro) deities; power; Vodou (Haitian) Ramos, Arthur, and syncretism, 19. See also syncretism Ranger, Terence, 153 rebirth, 147; supersedes birth, 14 recording, 10; of festival cycles, 5 Regla de Ifa, 167n20 Regla de Ocha (Lucumí), 140–51, 167n20 reinados (kingdoms). See cabildos de nación relexification, strategies of, 59–60 relic, sacred, Qur’an as, 12 religion(s): Cuban Lucumí-Santería, 140; deep knowledge empowers, 37–38; importance to New World Africans, 22– 23, 25; orisha, 150, 151; re-emergence of African in New World, 145; West African, 28, 37–38; Yoruba, 23, 31, 37. See also brotherhoods, Catholic; Catholicism; gods, African; myth(s); orisha; ritual(s); saints, Catholic; and names of individual gods resettlement: causes of, 128–29; patterns of Yoruba, 72–74, 128–29. See also kingdom(s); town(s) residence: vs. descent, debate over, 127, 129; intertwined with descent, 127. See also household; house/home; lineage(s), Yoruba retention(s): of African linguistics, music (dance), etiquette, motor habits, 23–24; and modalities of revision and resistance, 38. See also culture revision and resistance: politics of, 47–48; and West African history, 38. See also retention(s) revolution: associated with African gods, 45–46; associated with witchcraft, 113– 14; Haitian, 48, 64; women’s, against slavery, 112

Index / 203 Rigaud, André, 48 ritual(s), 2, 37; of Aku cult, 86; European imports used in, 162n29; Ishan, paraphernalia for, 82; power of Yoruba, 41– 43; of purity, 150–51, 169n35; royal, 44; simulating reproduction, 99; systems, Yoruba, 41–43, 139; takeovers occur during, 43, 150; of “upward mobility,” 149; Vodou, 32–33; water spirits important in Yoruba, 50–51, 146–47. See also cult(s); myth(s); politics; purification/ purity; religion(s); Yoruba Roach, Joseph, 14, 137 sacrifice(s), 99, 104; of animals, 47, 48, 86; blood in, 146; Bois Caïman, 160n7; objects for, 82, 86. See also deities; gods, African; orisha Saint Domingue, 63 saints, Catholic: African nations linked to, 136; paralleled with African gods, 18, 19, 21, 25–29, 34–37, 45; paralleled with orisha, 70. See also Catholicism; gods, African; religion(s) salt, used in festivals, 99 Samedi, Baron, “Papa Doc” takes as model, 55 Sans Poils, 52–53. See also Bizongo society Santería (Cuban), 18, 25, 28, 32, 160n6; and “cool” and “hot” deities, 36; and syncretism, 25, 29, 44, 124; “Yoruba” language in, 8. See also syncretism secrecy, 5–6; in honoring òrìṣ à, 30; maintained by African cults in New World, 25–26, 37, 44–45; power of, 162n25; safeguarding deep knowledge, 44. See also gods, African; religion(s); saints, Catholic; secret(s); slave(s) secret(s): within connaissance, 50; and deep knowledge, 43–46; of Kongo and Angola, 50; of women, 98–99. See also connaissance, and creolization; imọ jinlẹ̀ ; secrecy secret police. See tonton-makout Secte Rouge society, 34, 59. See also Haiti segmentation, political, 41–42, 42f, 94–95; and orisha cult, 71, 90. See also orisha; politics Senghor, Léopold, 39 Shango, 11, 25, 30, 36, 43, 69–70 shrine(s) (ipara): as governments, 134;

mark deification of heroes, 13; of towns, 82, 93–94, 94f, 95, 96, 160n3; of towns, as kingdoms, 134; of towns, house clusters of deities, 42. See also ilé òrìṣ à; myth(s); religion(s); ritual(s) “sib”/”multi-sib” (Bascom), 29, 71, 72, 85. See also cult(s); family(ies); metaphor(s); orisha Simon, Antoine, 33, 55; regime “iconic of high and low,” 55 singing, in orisha cult ceremonies, 137, 138. See also music slave(s): assimilation of determines status, 63; baptism of, in New World, 25, 35; carried knowledge from Africa to New World, 50; categories of, 63–64, 116; Congolese, 142–43; Congolese, conversion to Christianity, 142–43; emancipation of, 139–40; Fon and Yoruba, 143; godparents required of, by Christians, 165n23; households of, broken up, 115–16; households of, preserve West African women’s culture, 110; household vs. field, 116, 168n30; labor vs. breeding conflict, 116–17; liberation of through huckstering, 118, 119–20; loving bonds with masters, 115; matrifocal patterns when sold, 116; Mina integrate African culture, 143; Nagô, 168n30; naming of, 116; places of origin determined by Catholic brotherhoods, 135; places of origin unknown, 135; polygyny among, 165n29; Rada preferred, 161–62n24; raids on Ekiti, 74; religion(s) of, 25–26; republics of escaped, 35, 165–66n32; revolutions in Haiti, Cuba, and Brazil, 35; songs of, 138; stereotyped, 135, 165n21, 165n31; women as, in New World, 105–21, 165n28; women preferred by buyers, 116. See also Catholicism; religion(s); slaveholders, New World; syncretism; witchcraft; women slaveholders, New World: polygynous with enslaved women, 114–15; reward enslaved women for newborn children, 112; undermine, abuse, and exploit slaves, 110, 116. See also Big House; polygyny, New World; slave(s); whites slavery: forms of, in New World, 23, 25; New World festivals as assault on, 138; and religion, 25–26, 31

204 / Index Smith, M. G., 2–3, 19, 40, 43, 157n2, 163n12 Smith, William, 107 sorcery, 61. See also magic; Petwo (Petro) deities Soulouque, Faustin, 33, 55 spirits: Petwo, as Creole, 48–51; “root lwa” (lwa rasin), 161n21; water, 50–51. See also deities; gods, African; orisha Spivak, Gayatri, 12 Stanley, Henry, 108f stereotyping, 119, 120, 135, 153, 165n21, 165n31. See also slave(s); women subquarters, 75, 77, 82, 84, 85, 86. See also kingdom(s); quarter(s); town(s); village(s) substitutions, in myth, 12. See also myth(s) sugar plantocracy, 145, 147, 149. See also miscegenation; whites symbol(s): calabash as wide range of, 44; foreign, 44; as icons of terror, 56; kola nut as, 138; Petwo as evil, 49. See also metaphor(s) syncretic paradigm (Herskovits’s), 17–20 syncretism: and Africanity in the New World, 17–18, 20, 46, 96, 137, 143; and Afro-Caribbean religious production, 139; based on commonalities, 143; cabildos and, 137; and Candomblé, 45, 96, 139; comparison between Brazilian, Cuban, and Haitian forms of, 31; cultural origins and, 24–31, 60, 135–36; defined, 21, 22; “double,” 34; “good” and “bad,” 168n28; and hegemony, 36; Herskovits’s theory of, 17–38; Herskovits’s theory of, recast as political, 45; in Santería, 25, 45, 139; Vodou and Catholicism in Haiti, 33, 45. See also gods, African; Herskovits, Melville J.; saints, Catholic takeovers, ritual, 43, 132, 137, 150. See also orisha; ritual(s) Taussig, Michael T., 48 terreiros (yards): Bahian, 95, 96, 144; collective power of, 146 Ti Jean Petro, renamed African god, 59 Tisma, Madame, 51 tonton-makout, 52, 56, 160n12; favored by Duvalier, 57. See also Bizongo society; Duvalier, François (“Papa Doc”); Haiti

town(s), 74–75, 79, 83–84, 91, 93; becomes home, 150; becomes kingdom, 72–73, 79, 132; as child, 71; as family, 89; fission of, leads to kingdom, 73; formation and dissolution of, 74; in the home (ilé), 124–27, 131, 141, 147, 149– 50; house as, 131; in ilé, 131; in Ishan, 74–75; lineages in, 126; of origin, 128, 150; as quarters, 81, 125; references of word for, 81; repossess kingdoms, 150; and ritual takeover, 132, 137; shrine(s) in, 82, 91–92, 95, 96, 132, 134, 160n3; subordinate, 70, 74, 75, 79–80, 83, 84, 87, 90, 94, 103; Yoruba markets in, 104. See also chief(s); house/home; kingdom(s); shrine(s); village(s) transposition, ritual, 132. See also ritual(s) Trouillot, Michel-Rolph, 56 United States, 20, 138. See also North American Lowcountry “upward mobility” (via limpeza), 149. See also purification/purity; racism; whites Verger, Pierre Fatumbi, 39, 68–69, 95–96 Verran, Helen, 129–30, 131 village(s): clusters of, 2, 41, 70, 80, 90; in Ishan, 74–75. See also Ilemesho (Ishan town); markets, town and village; quarter(s); town(s) violence, in dance, 48–49. See also dance; Pedro, Don (Dom) Virgen de Regla, 25 Vodou (Haitian), 18, 32–34, 67, 162n26; basic cult units of, 51; becomes state religion, 33–34; and Catholicism, 32– 33, 34; ceremonies feared by palace elite, 55; on connaissance, 50; in Haitian politics, 55–58; history of, 34; Hurston on, 30–31; idioms of, 52; matrilineal, 161n15; Petwo paradox, 46–51; Rada of Ginen powerful within, 54; revolution and role reversal within, 64; in rural Haiti, 33; status elevated by Duvalier, 55–56, 161n19; and syncretism, 25, 44. See also Haiti; Petwo (Petro) deities; Rada deities; religion(s); revolution Volontaires de la Sécurité Nationale (VSN), 56. See also Bizongo society; Haiti vudum deities, revalued by Nagô-Yoruba “orixá,” 144–45

Index / 205 wangole (fetishes), 50 warfare, Yoruba (nineteenth century), 37, 74, 94, 128, 132 water/fire overlap among deities, 60–61. See also analogy; deities; Ginen; gods, African; Petwo (Petro) deities; Rada deities wet nurse(s), 109, 109f, 117; black, 165n20. See also womanhood, in New World; women Wheeler, Charles, 107 White, Deborah Gray, 97, 109, 112, 113–14, 164n15 “White House of the Old Sugar Mill” (Casa Branca White House): as Great House of the sugar plantocracy, 145; originary frame of Casa Branca do Engenho Velho (Ile Iyá Nassô), 145. See also Big House; casa grande; great houses of Candomblé; plantocracy, sugar whites: associated with purity, 124, 145, 168n31; control of, threatened by hucksters, 118, 119–20; displaced by mulattos in Haiti, 54, 106–7; identified with Rada nation, 53–54; as slaveholders (see Big House; slave[s]); supremacy of, in New World, 115, 147–48. See also Big House; casa grande; Catholicism; hegemony; New World; power; racism; slave(s); slaveholders, New World Wirtz, Kristina, 8–9, 12–13 witch(es): black seductress as, 165n31; work inverse to fertility, 100, 112–14. See also witchcraft witch-bird, Yoruba, parallels “African Harpies” stereotype, 119, 120 witchcraft, 98, 163–64n3, 165n28; destructive market women accused of, 104; and infertility, 100; matrilineal, 105, 116; menopausal power of, 103; in slave families/New World, 111–16; slaves resort to, 117. See also magic; Vodou (Haitian) wives, only exogamous, 101. See also cowives; family(ies); women womanhood, in New World, sexual (Jezebel)/motherly (Mammy) dichotomy, 107–10, 108f, 109f, 111–12, 115, 163n2, 164n17, 164n19. See also polygyny, New World; slave(s); women women: antithetical powers (source of

fertility/witchcraft) of, 104–5, 106–7, 112; as caretakers of mothers, 116; control town and village markets, 102–5, 117–18; “eat” family and plantations, 118–19, 164n3; enslaved, stereotypes, 165n21, 165n31; enslaved in New World plantation societies, 105–21; as negative force, 104–5, 118–19; as potential witches, 100, 104, 105, 165n31. See also “African Harpies”; “Cash Madam(s)”; cowives; midwives; mothers; polygyny, African; polygyny, New World; slave(s); witch(es); witchcraft; womanhood, in New World Wood, Betty, 119 Woodson, Carter G., 39 worship, ancestral vs. orisha, 71. See also cult(s); orisha; religion(s); ritual(s) Wurdemann, J. G. F., 138 Yagba, 80, 91–94, 134 Yai, Olabiyi, 68 Yemoja (Yemanja), 25, 30, 43, 44, 51, 91– 92, 95, 132, 137; priestesses of, 92 yeye (grandmother), 71 Yeyeolokun, 44 Yeyerefin (grandmother deity [Epa mask] of Irefin quarter), 84, 85f Yoruba: cosmology, 60; culture, 7–8, 10, 13, 28, 68–96, 152–55, 166n1; defined, 8, 153–54, 155, 159n10; ethnicity of, 122–23, 166n1; etymology of name, 122–23; father of, 13; and Fon, 28, 65, 66, 106, 139, 143, 162n26; and Fon slaves, 143; gods, 25–28, 27t, 69 (see also gods, African; and individual names); government, 42–43; hermeneutics of, 3, 23, 44; influence on New World idioms, 166n35; kingship, 167n22; language standardized, 123; Nagô-associated politico-ritual system, 144, 145; names hyphenated, 139; nineteenth-century wars, 37, 74, 94, 128; origins, 1, 4, 8, 9, 10–14, 27t (see also descent); origins of name, 159n10; perspective on African diaspora and slavery, 107–21; priests and priestesses, 5–6, 49, 51 (see also religion[s]); religion in, 23, 31, 37 (see also religion[s]); research on, 41, 152–55; ritual, 43, 50–51; “tradition,” 8; as “Yarba,” 123.

206 / Index Yoruba (continued) See also Africanism/Africanity; culture; diaspora, African; Ekiti Yoruba; Fon; knowledge, deep; language(s); Nagô nation; Nigeria; orisha; religion(s); Yorubaland; and individual cults and place names Yorubaland, 11, 14, 28; northeastern (Ekiti-

land), 68–96; orisha worship in, 150. See also Ayede, kingdom of; Ishan, kingdom of; Yagba; Yoruba Zaka (Azaka), 56–57. See also gods, African zinglins (paramilitary thugs), 55 zombie, Haitian, 57. See also Haiti; Vodou (Haitian)