Iberian Imperialism and Language Evolution in Latin America 9780226126173

Citation preview

Iberian Imperialism and Language Evolution in Latin America

Iberian Imperialism and Language Evolution in Latin America Edited by s a l i k o k o s . m u f w e n e

The University of Chicago Press Chicago and London

Salikoko S. Mufwene is the Frank J. McLorraine Distinguished Service Professor of Linguistics and the College as well as

professor in the Committee on Evolutionary Biology and the

Committee on the Conceptual and Historical Studies of Science at the University of Chicago. He is the author of several

books, including, most recently, Language Evolution: Contact, Competition and Change.

The University of Chicago Press, Chicago 60637 The University of Chicago Press, Ltd., London © 2014 by The University of Chicago All rights reserved. Published 2014.

Printed in the United States of America 23 22 21 20 19 18 17 16 15 14  1 2 3 4 5 ISBN-13: 978-0-226-12617-3 (cloth)

ISBN-13: 978-0-226-12620-3 (paper)

ISBN-13: 978-0-226-12567-1 (e-book)

DOI: 10.7208/chicago/9780226125671.001.0001 Library of Congress Cataloging-in-Publication Data Iberian imperialism and language evolution in Latin America / edited by Salikoko S. Mufwene.   pages ; cm

 Includes index.

 ISBN 978-0-226-12617-3 (cloth : alk. paper) — ISBN 978-0-226-12620-3 (pbk. : alk. paper) —

ISBN 978-0-226-12567-1 (e-book) 1. Indigenous peoples—

Latin America—Language—History. 2. Spanish language—

Latin America—Influence on foreign languages. 3. Portuguese language—Latin America—Influence on foreign languages. 

4. Spanish language—Latin America—History. 5. Portuguese language—Latin America—History. 6. Languages in

contact—Latin America. 7. Latin America—Colonial

influence. I. Mufwene, Salikoko S., editor of compilation.  PM5009.124 2014  498′.09—dc23

2013039225

This paper meets the requirements of ANSI/NISO Z39.48-

1992 (Permanence of Paper).

contents

Preface vii 1 Latin America: A Linguistic Curiosity from the Point of View of Colonization and the Ensuing Language Contacts 1 salikoko s. mufwene

2 The Many Facets of Spanish Dialect Diversification in Latin America 38

john m. lipski 3 Amerindian Language Islands in Brazil 76 hildo honório do couto

4 Historical Development of Nheengatu (Língua Geral Amazônica) 108

denny moore 5 Language and Conquest: Tupi-Guarani Expansion in the European Colonization of Brazil and Amazonia 143 m. kittiya lee

6 African Descendants’ Rural Vernacular Portuguese and Its

Contribution to Understanding the Development of Brazilian Portuguese 168

heliana mello 7 Brazilian Portuguese and the Ecology of (Post-)Colonial Brazil 186

j. clancy clements

8 Maya and Spanish in Yucatán: An Example of Continuity and Change 205

barbara pfeiler 9 Standard Colonial Quechua 225 alan durston

10 Linguistic Subjectivity in Ecologies of Amazonian Language Change 244 christopher ball

11 The Ecology of Language Evolution in Latin America:

A Haitian Postscript toward a Postcolonial Sequel 274 michel degraff Contributors 329 Subject Index 333 Author Index 341

preface

The present book is the outcome of a workshop I hosted at the University of Chicago on April 13–14, 2007. I owe the event to the Center for Latin American Studies (CLAS), which, thanks to an endowment from the Edward Larocque Tinker Foundation, not only has made it possible for the affiliate faculty over the years to invite scholars on Latin America to spend a quarter or two as fellows at the university, but also has encouraged the local sponsor faculty to organize an academic event on a topic related to the scholarship of the visitor or visitors. In winter 2007, we were especially blessed with the concurrent visits of two linguists, Hildo Honório do Couto, University of Brasília, and Barbara Pfeiler, Universidad Nacional Autónoma de México in Merida, Yucatán. Both scholars had studied various aspects of the contact of European languages with Native American languages, including the endangerment of many of the latter, the curious spread and survival of some of the Native American languages, the ways in which they have been influenced by either Portuguese or Spanish, depending on the country, and how they contributed to shaping a particular colonial variety of the European language. With encouragement from Dain Borges, then Director of the Center for Latin American Studies, I seized this opportunity to convene a two-day workshop on language contact in Latin America; all the contributors to the present volume except Michel DeGraff participated in that workshop. We received complementary funding for the meeting from the Franke Institute for the Humanities, under the directorship of James Chandler. I am grateful to both directors for their enthusiastic support. I am very grateful to Michel DeGraff for agreeing to write a postscript to the volume from a Caribbean perspective, especially that of the former French colony Haiti, whose colonial history, after its expropriation from Spain by the French, ended abruptly and violently at the dawn of the nineteenth century. DeGraff’s contribution reminds us not only that

[ viii ]

preface

there is a French component in Latin America but also that it sheds light on, among other things, why today the evolutionary dynamics of the vitality of “indigenous” languages in the Caribbean are not the same as in continental Latin America. To all the authors: I am very grateful for your confidence in the significance of this volume and for enduring the demanding editorial process. If I bruised the egos of some of you with my repeated requests for clarifications and revisions, I hope the end result gives you some reason to forgive me. I thank you sincerely for cooperating nonetheless. In the same vein, I heartily thank Lois R. Crum for her assiduous copyediting, which has not only helped smooth the prose of this book but also made the presentation of data and usage of particular terminology more uniform from one author to another. I owe a special debt of gratitude to Joshua Beck, then Associate Director of CLAS, and to his assistant Christelle Marpaud, without whose involved logistic assistance the workshop would not have been successful. I am also deeply indebted to Bruce Mannheim, who assisted me in drafting the proposal for this book and with the first round of feedbacks to the contributors. Lastly, I hope the reader will forgive the fact that Latin America is not evenly represented in this volume. Funding limitations and the tight schedule on which I as a non-Americanist was operating in planning the workshop forced me to work with scholars more familiar to me and to our visiting scholars, subject to their availability. My agenda was simply to see whether the meeting on language contact in Latin America, which did not deal with creoles, would enrich current scholarship based on the formation of creoles and would shed additional or different light on a number of “ecological” factors that bear on language evolution in colonial settings. We wound up learning more than I had anticipated, especially regarding language vitality. salikoko s. mufwene

1 * Latin America: A Linguistic Curiosity from the Point of View of Colonization and the Ensuing Language Contacts salikoko s. mufwene

1. Historical Background The starting point is the fifteenth century. Southern Europe was then emerging out of seven centuries of Arab economic hegemony and even, in the case of Spain, political rule. It had become widely accepted that the world was round, the navigational knowledge and technology developed in the Far East had spread, and European merchants wanted to trade directly with the spice growers of Asia and Africa. Just like Iberia (Spain and Portugal) in the west, the Ottoman Empire was then also asserting itself as a new world power around the eastern Mediterranean, in the wake of the weakening Arab domination in the region. Not wanting any military confrontations with the Muslims, the Iberians chose to avoid them by sailing both around Africa and westward across the Atlantic. Nobody anticipated that Christopher Columbus would accidentally discover the Americas, which from the early sixteenth century came to be designated as the New World. But from then on, the Iberians sought new trading commodities across the Atlantic. The Iberians’ ambitions were facilitated by the invention of the caravel ship, equipped with “a large hull for sailing the high seas, a sternpost rudder and triangular sails for directional mobility, and artillery to intimidate those who challenged or refused to cooperate with them” (Eakin 2007, 52). They were now ready to face the rough waters of the Atlantic Ocean, able to transport more goods, including fresh water and nonperishable food, and many people over long distances. With the concurrent invention of capitalism, they were contributing to a mercantile revolution that could be compared to what had initiated modest longdistance trade a couple of millennia earlier: the use of rivers and other waterways as highways (Chaudenson, forthcoming). Similar technological progress with sea navigation had enabled the Arabs and the Chinese to become major mercantile powers centuries earlier, traveling back and forth between East Asia and East Africa. Thanks to their knowledge of monsoon wind cycles, they had learned to sail east or west during the right season and thus minimize the risks of shipwreck (Ansaldo 2009).

[ 2 ]

chapter one

For the first time, the Europeans could now sail to the same markets around Africa and across the Atlantic, in order to open their own trade forts outside Europe. The leading European maritime powers were then Portugal and Castile, which was expanding into today’s Spain. The two territories had then emerged as two autonomous nations, out of the multitude of smaller political/ethnic entities that the Greeks and the Romans had identified collectively as Iberia (Ιβηρία in Greek) and Hispania (in Latin).1 Portugal and Castile had just freed themselves from seven centuries of colonization by the Moors, the Arabs, and the Islamized Berbers, who had worked for the Almoravid Dynasty. They were looking for external resources to solve their economic problems, some of which were consequences of the expensive Reconquista campaign they had concluded. Endeavors to settle and develop economic markets (for grain, spices, gold, and slaves in particular) in northwest Africa and the neighboring islands (Azores, Madeira, and the Canaries, in order to produce sugarcane)2 brought these emergent world powers into conflict with each other, too. Because their geographical expansion also contributed to the spread of Catholicism, popes intervened to resolve the conflicts through successive bulls, the last of which is associated with the Treaty of Tordesillas in 1494 (Clements 2009). According to that agreement, Portugal laid claim to all territories and seas lying east of a north-south line descending from west of Cape Verde across present-day Brazil all the way to modern Indonesia. Spain would claim control of all territories and seas extending from west of that line all the way to the Philippines. Navigators from both maritime powers could thus prove empirically that the world was round by sailing east and west, respectively. The treaty provisions explain why Christopher Columbus, rather than Bartolomeu Dias, discovered the New World. This division also explains why Spain did not acquire exploitation colonies in Africa until the Berlin Treaty in 1885 and why Portugal had only one colony in the New World, Brazil, and wound up with mostly exploitation colonies, typically in continental Africa, and small settlement colonies on Atlantic islands off the western coast of Africa, in India (Gao and Korlai) and Malaysia (Malacca), in Indonesia (Sumatra, Jakarta, Timor), and in China (Macao).3 According to Eakin (2007, 53), the other reason may lie in how the two powers were engaged in their territorial expansions: “Spain was the cutting edge in state formation and nation building in the Renaissance” and may thus have sought to build what was later designated as “New Spain,” whereas “Portugal was at the forefront of the trade revolution” (indeed, it played a more important role than Spain in the slave trade!) and was more inter-

Salikoko S. Mufwene

[ 3 ]

ested in developing coastal trade colonies in Africa and Asia, from which it benefited immensely. Within a century, the Spanish conquistadores had colonized the Caribbean and the lion’s share of the Americas. They first settled Hispaniola (present-day Dominican Republic and Haiti) and then Cuba, from which they colonized the rest of the Caribbean. From Cuba Hernán Cortés launched expeditions to Mexico, and from there the Viceroyalty of New Spain was founded. It included Florida, the Mississippi Valley, the present southwestern United States, Mexico, Central America minus Panama, and the Caribbean.4 The Spaniards also founded concurrently the Viceroyalty of Peru, which included all of present-day South America except for the part of present-day Brazil that had been allotted to and settled by Portugal starting in 1500. As noted previously, the Portuguese were focused on expanding from the Atlantic Ocean all the way to the Pacific. After experimenting with sugarcane cultivation on the Azores and the Madeira archipelagos, they developed, as noted, settlement colonies on islands off the western coast of Africa (Cape Verde, São Tomé, and Principe) and a long chain of trade colonies on the eastern coast of the Atlantic, in the Indian Ocean (on or close to the eastern coast of Africa and on the coast of southern Asia and Southeast Asia) all the way to Indonesia (Java, Sumatra, and Timor) in the Pacific, and on the coasts of China (Macao), and Japan (Nagasaki). The seeds of “Latin America” as a new cultural region, synonymous with “Iberoamerica” or “Hispanic America,” were then planted, although from the seventeenth century onward the Spaniards and the Portuguese faced the competition of England, France, and Holland. These nations had by then also emerged as important maritime and economic powers in their own right and coveted all these territories that their European forerunners had claimed. Latin America itself gradually acquired a geographic definition, at least in the United States: it comprised all parts of the New World south of Anglophone North America, the areas where Romance languages are spoken as dominant vernaculars. Within the next two centuries, the world map had to be redrawn several times to accommodate the additional power players. Although Portugal successfully drove the Dutch and the French out of northeastern Brazil, it lost many of its African and Asian colonies, in particular the Cape of Good Hope (in present-day South Africa), Sri Lanka (then named Ceylon), and Indonesia to Holland. (The Cape of Good Hope was colonized another century later by England.) This hitherto mercantile superpower later also lost Elmina (Ghana), India, and Malaysia to England, as well as Gorée (Senegal) and Indochina to France. No Euro-

[ 4 ]

chapter one

pean nation colonized Japan or China, where Portugal kept Macao up to the late twentieth century and where England founded the trade colony of Canton in the eighteenth century (and later claimed Hong Kong, which it also relinquished to China in the late twentieth century). Spain was not spared the loss of colonies. Central and South American countries such as Belize, Guyana, French Guyane, and Surinam, once colonized by England, France, and Holland, are in the geographic space that the 1494 Treaty of Tordesillas had assigned to Spain. The Louisiana colony was taken by the French at a time when Spain neglected this part of New Spain. The Spanish Empire also lost several Caribbean islands in the same way, although some of them were seized through wars, for instance present-day Haiti, Jamaica, Trinidad, and St. Kitts, as well as Aruba, Bonaire, and Curaçao (also known as the Netherlands Antilles). Later on, the United States acquired Florida from Spain, and Texas, New Mexico, and California were acquired from Mexico in the nineteenth century. Linguistically, the Iberian colonial empire has remained an enigma, especially regarding the rarity of vernaculars that linguists have identified as creoles (thereby disfranchising them from other colonial varieties of European languages spoken predominantly by descendants of Europeans). These vernaculars, now spoken typically by descendants of African slaves and of the nineteenth-century contract laborers from India and West Africa who joined or replaced them on the plantations, have been associated primarily with sugarcane cultivation.5 Brazil had started this industry about a century before the Caribbean colonies, with which our heuristic creole prototypes have been associated, adopted it; but Brazil has no creoles.6 What are the answers to the following relevant questions? Since there is a popular or vernacular Brazilian Portuguese (VBP) that is distinct from (nonstandard) European/Peninsular Portuguese, does this (apparent) enigma mean that we may not have to dwell so much on the distinction between creoles and noncreoles in discussions of the divergence of colonial varieties of European languages from their metropolitan counterparts? Or are there specific ecological reasons for why VBP should not be considered a creole (Mufwene 2001, 2005, 2008)? And why are there no distinct varieties associated with the descendants of African slaves in Brazil, except perhaps for the varieties spoken in the Quilombos (hinterland maroon settlements; see chapter 6)? Why did the New World’s only Portuguese Creole (albeit with Spanish influence), Papiamentu, emerge in the Netherlands Antilles, where the official language has been Dutch and where no sugarcane industry developed? Given the size of the Spanish Empire in the New World, why did the only creole associated with Spanish, Palenquero, emerge, of all places, in the Palenque (settlement of escaped slaves) of San Basilio, Colombia,

Salikoko S. Mufwene

[ 5 ]

where there was no sugarcane industry? Why did Cuba and the Dominican Republic, which engaged in sugarcane cultivation, had large slave populations, and continued the institution of slavery up to the late nineteenth century, not produce any creole (pace Schwegler 2006)? Why did Puerto Rico, which also engaged in the sugarcane industry, not produce a Spanish creole? What specific ecological conditions distinguishing Iberian colonization in the New World from the English, French, and Dutch colonization can shed light on the magic of creole formation—although, as argued by Mufwene (2001, 2005, 2008), there are no restructuring processes or combinations thereof that distinguish the emergence of creole vernaculars from normal language change and speciation? What particular ecological factors in the Portuguese colonies off the shore of the western coast of Africa and in small settlement isolates in Asia favored the emergence of creoles (Kriolu Kabuverdianu, São Tomense, Principense, Korlai Indo-Portuguese, and Macanese/Papia Cristam di Macau) that did not obtain in Brazil? Or is there a problem with the way the putative process of “creolization” is conceived of in linguistics? The same question does not quite arise about Spanish, in part because Spain had very few colonies outside the New World. Those of Africa, namely present-day Morocco, Western Sahara, and Equatorial Guinea, were acquired in the late nineteenth century, during the “Scramble for Africa,” with the Berlin Treaty (1885). Besides, these developed on the exploitation-colony model, which was not conducive to the emergence of creoles, typically associated with plantation settlement colonies.7 An explanation is thus needed for the emergence of Chabacano/Chavacano in the Philippines, if it too qualifies as a creole, as it does according to the stipulations of some linguists. The Spaniards did not colonize the Philippines in the same way as the Portuguese did their Atlantic islands, which the latter peopled with exogenous slaves. Nor did the Spanish colonize the archipelago on the model of Portuguese trade colonies such as Korlai and Macao. So, what in Iberoamerica generally prevented the emergence of creoles? Latin America is a linguistic curiosity in another way, too. In reality, the geographical colonization of Brazil is still in process today, with the ongoing penetration of the Amazon rain forest primarily by Brazilians of European descent. While the earlier Portuguese colonial expansion into the interior was driven especially by the gold rush to Minas Gerais and by later farming settlements in the Southwest, as well as by the cultivation of coffee farther west and by the extraction of “drogas do sertão,” the present penetration of Amazonia has been driven by Hevea cultivation for rubber and, more recently, by lumbering and the exploitation of diamonds. The negative consequences of deforestation caught the atten-

[ 6 ]

chapter one

tion first of environmentalists and later of linguists, who, since the early 1990s, have been decrying the endangerment of Native American and other “indigenous” languages around the world.8 However, as observed by Nettle and Romaine (2000) about this worldwide problem, proportionally fewer Native American languages have been lost in continental Latin America than in Anglophone North America. There are some demographically major languages such as Quechua, Aymara, and Nahuatl, that continue to be spoken by more than 1 million people each in this part of the Americas.9 As discussed by Hildo do Couto and by Denny Moore, Nheengatu, a legacy of the geographical expansion of Tupinambá (also known historically as Língua Geral ‘general/common language’ or Língua Brasílica ‘Brazilian language’), has continued to be spoken, albeit by fewer and fewer people as the most likely speakers have adopted Portuguese as their vernacular. The overall picture is rather unlike that in North America, where linguists now have only memories of major indigenous lingua francas such as Mobilian and Chinook jargons and where former major indigenous vernaculars such as Navajo and Cree are spoken only by a little over one hundred thousand people each. In Brazil, the Portuguese colonial expansion actually contributed to the geographical spread of Língua Geral (Couto and Lee, chaps. 3 and 5, this volume) up to the nineteenth century. For a long time, the Spanish expansion also contributed to the spread or maintenance of Quechua as a lingua franca, a function it had acquired as a unifying language since the Inca Empire.10 And even as Spanish spread in South America, it did not prevent the emergence and spread of Media Lengua ‘middle language’,11 unlike what happened in Anglophone North America, where Michif and Medny Aleut (a.k.a. Copper Island Creole) are now moribund legacies of the European expansion in the social periphery. However, while a general discussion of the present kind can help develop the big picture, it hides a lot of details that set almost every colony apart from the others. This is what the following chapters will help us understand, although many parts of Latin America not covered by the book deserve just as much attention as those included. Unfortunately, all sorts of practical constraints made it difficult to cover additional areas of Latin America. We hope, nonetheless, that the approaches to language contact adopted here, from the colonial to the present period, will inspire future research and more publications. They certainly demonstrate that language contact in the European colonies can be seriously investigated from a non-European perspective; that language evolution did not proceed uniformly even in those contexts where one does not have to distinguish between creoles and noncreoles (see, e.g., Escobar 2012); and that structural change, language speciation, and lan-

Salikoko S. Mufwene

[ 7 ]

guage vitality (including language endangerment and loss) can be discussed as related folds of language evolution. Finally, there is something to be gained from the depth and complementarity of some of the chapters, especially regarding Brazil; this heuristic model can be emulated in future collective publications about Latin America or any other area. The choice of the phrase language evolution in the title of this book was motivated by the need to highlight the fact that many of the ecological factors that account for how particular structural features have been selected into the system of the dominant language also shed light on the dynamics of language vitality in the relevant polities, differentially of course (Mufwene 2001, 2005, 2008). At a time when linguistics has become very much concerned with the endangerment and loss of indigenous languages in former European colonies, it will not surprise the reader that a great deal of the discussion in the chapters deals with this aspect of language evolution. The more involved reader will notice, nonetheless, that the chapters cover much more than this topic. There are many observations about why neither Portuguese nor Spanish has evolved uniformly in the New World. Although, in relation to Native American languages, they have been subjected to typologically similar substrate influences under similar contact conditions (Escobar 2012), there are still regional, even intranational, specificities that set the new Portuguese and Spanish varieties apart from each other. Even variation in the demographic sizes of the indigenous populations (Escobar 2012) and the contrast between residing in the emergent urban centers (where there have been more contacts with Europeans) and in the rural environment (Lipski, chap. 2, this volume) have influenced this differential language evolution. As in Africa, indigenous people living segregated geographically and socioeconomically from Europeans in rural areas have typically been under less influence, if any, to Europeanize culturally. (See also Quijano 2007 and Quijano and Ennis 2000 for such a comparative approach to colonization.)

2. The Chapters In Chapter 2, “The Many Facets of Spanish Dialect Diversification in Latin America,” John Lipski starts by asking why Spanish has diversified so much in Latin America, especially why the Caribbean and the continental varieties are so different from one another, having presumably evolved from the same Castilian language. The question is important because, according to him, this sociolinguistic superstratum in Spanish colonies was relatively homogeneous in the sixteenth century. Also, according to Lipski, “Both peasants and landed gentry were severely underrepresented

[ 8 ]

chapter one

in the colonies, and the first waves of settlers were predominantly members of the skilled trades, small landowners, from marginal zones beset by climatic disasters, and individuals who for whatever reason had not done well economically or socially in Europe” (1994, 40, quoted by MalMolinero 2000, 30–31). The colonization of the New World also started at a time when Castilian was just consolidating itself in the Iberian Peninsula (Mal-Molinero 2000, 27), prevailing over the other Ibero-Romance languages of the emergent unified Spanish kingdom as the official language. Overlooking the facts that all the colonists need not be expected to have spoken Castilian with the same level of fluency and that many of them could have spoken it with substrate influence from Catalan, Galician, Andalusian, or any other Ibero-Romance vernacular, Lipski focuses on the effect of contact with other European languages (such as Italian), with African languages, and with the Native American languages. As explained by Chaudenson (1992, 2001), the latter are the ones that constitute the actual substrata in this colonial scenario, because they are indigenous and prior to all the other languages brought by the immigrants, voluntary and involuntary. However, because the relevant groups did not all come to the colonies at the same time—some of them did not arrive in the relevant critical mass until the late nineteenth or early twentieth century (such as the Italians in Argentina)—one must assume the evolution of Spanish in Latin America to have been gradual, consisting of successive responses to the changing composition of the colonial populations. Such a history of successive and incremental language contacts calls for a periodized account of language evolution, as suggested by Chaudenson (1992, 2002), since the relevant structures were not all in place at the same time, nor did they change overnight and in a wholesale manner. That is, different waves of non-Spanish immigrants, often speaking different languages, must have influenced the evolution of Spanish during various periods of massive language shift from the immigrants’ national vernaculars. The significance of convergent influences (although they may be assumed to have varied from speaker to speaker, according to individual languagelearning skills) also depended largely on the critical mass of the ethnolinguistic groups relative to, and their degree of integration within, the extant Spanish-speaking populations, notwithstanding the extent of typological divergence from the target language. By the founder principle (Mufwene 1996, 2001), those engaged in the process latest and in negligible numbers may be expected to have exerted the least influence, as in the case of several immigrants to Anglophone North America since the nineteenth century, because they typically arrived by increments that were not unified ethnolinguistically or settled on the margins of the An-

Salikoko S. Mufwene

[ 9 ]

glophone populations. While gradually shifting to English as groups, their children selected features of the native Anglophones over the nonnative elements of their parents (Mufwene 2009a). Lipski acknowledges this factor in a different way: “in truth Latin American Spanish is the product not only of its first settlers but of the totality of the population, immigrants and natives alike” (chap. 2, this volume). Chaudenson’s notion of periodization entails that the local ecological conditions change as the composition of the population speaking or shifting to the dominant vernacular varies in time and space. The natures and strengths of the influences exerted on the language vary throughout the colonial period and thereafter. As explained by Mufwene (2001, 2008), the local population structure, with its own particular composition of coexistent languages and varieties thereof, as well as its own patterns of interaction, influences the specifics of language evolution. It appears that the spread of Castilian over such a wide empire as Latin America contained all the ingredients for differential evolution. Thus, Lipski’s question can be reformulated as: What are the specific ecological conditions that account for the speciation of Castilian Spanish in Latin America? What are the common conditions and elements of Caribbean Spanish, on the one hand, and Central and South America, on the other, that account for such a divide enhanced by the sea that separates the islands from the continent? A particular factor relevant to both the comparative approach (as in population studies)12 and the periodization approach, made evident by Lipski’s own discussion and Clements’s (chap. 7), is that the demographic significance of African slaves was not consistent from one colony to another. Adding to complexity in adequate accounts of this differential language evolution is that the ethnolinguistic composition of the African population must not have been uniform either, from one colony to another or even within the same colony, especially the large ones. To the extent that African languages have exerted their share of influence on some of the emergent colonial Spanish and Portuguese varieties (see, e.g., Mello, chap. 6, this volume; and Negrão and Viotti 2012, regarding Brazilian Portuguese), the total substrate element must have varied from one colony or part thereof to another, thus fostering nonuniform evolution of the colonial vernaculars. The question is whether, as an ecological factor, African substrate influence was, overall, significant and uniform enough on the islands to set Caribbean Spanish varieties off from continental varieties. We must bear in mind that no Hispanic Caribbean island has been associated with a creole. Equally significant is that Cuba and the Dominican Republic did not engage in the sugarcane industry until the late eighteenth century, more

[ 10 ]

chapter one

than two centuries after the founding period, during which the European population was the majority and there was no institutional race segregation—which does not mean absence of discrimination (at least in relation to the Africans). Also, these territories maintained the institution of slavery up to the late nineteenth century, long after it had been abolished everywhere else; in fact, they imported the most slaves during that period, to satisfy the demand for cheap labor. Interestingly, it is only in peripheral plantation communities of Cuba that dubious “linguistic creolization,” comparable perhaps to the similar varieties that emerged in the Brazilian Quilombos (Mello, chap. 6, this volume), has been reported (Schwegler 2006). Creolists do not even agree on whether the varieties that emerged in these peripheral communities should be identified as creole, although they are different from the national Spanish and Portuguese varieties.13 To explain why no creole emerged in Cuba or the Dominican Republic, Chaudenson (1992, 2001) argues in the case of Cuba (and the same is true of the Dominican Republic) that these communities arose at a time when there were already a lot of Creole people of African descent who spoke the same colonial Spanish as the Creoles of European descent, and on those plantations there were plenty of them from whom the Bozal slaves could have learned the colonial vernacular adequately.14 By contrast, English, French, and Dutch creole vernaculars emerged at a time when the Bozal slave population was demographically superior to the creole slave population, and the Bozals’ linguistic models included the “seasoned slaves,” who spoke the colonial language nonnatively (Baker 1993; Chaudenson 1992, 2001). Language transmission in plantation settings thus proceeded in a way reminiscent of the exploitation colonies of Africa and Asia, where the colonial varieties of the European languages transmitted through the school system have indigenized and diverged already from the metropolitan models (Mufwene 2005). Thus, the later learners of the European language were not exposed to exactly the same varieties as the earlier learners, who had more access to either native varieties or closer approximations thereof. In the English, French, and Dutch colonies, as opposed to the Spanish colonies, segregation was based on race, rather than just on socioeconomic class; it was part of the population structure since the early stages of the development of the colonies, where the homestead, preplantation society phase did not last long. Just the opposite appears to have been the case in Iberian colonies, if racial segregation was institutionalized at all. As will become evident, Barbara Pfeiler (chap. 8, this volume) contributes yet another dimension to this vexed question. The massive shift to Spanish, which is still in progress today in former colonies of continental

Salikoko S. Mufwene

[ 11 ]

Latin America, did not start at the same time in various colonies. Not all parts of the political/administrative colonies were economically and culturally integrated at the same time. Pfeiler brings to bear the Yucatán Peninsula, where Maya continued as the dominant lingua franca until the twentieth century. The late shift to Spanish was accompanied by substantial substrate influence from Maya. Consequently, there emerged a distinct Yucatán dialect of Spanish that is distinct from Spanish in the rest of Mexico. As reformulated in this introductory chapter, Lipski’s question is thus not a simple and easy one to answer, because it requires a rich understanding of patterns of population growth and human interactions, as well as of significant periods in the spread of Castilian, factors that seem to have varied from one Spanish colony or part thereof to another. For instance, as explained by Lipski himself, Argentina and Uruguay stand out for the extent of Italian influence on the Spanish of Buenos Aires and Montevideo, which is not noticeable in the varieties spoken in the other parts of these countries. An important reason is that, although the Italians arrived late, in the nineteenth century, they quickly constituted critical masses in these emergent urban centers and easily impacted the evolution of Spanish locally. By the same token, one must ask why, although Italians represent the largest ethnic group after the descendants of the Spaniards, Italian has not emerged as an important European vernacular adopted by other Argentinians. Do European Argentinians have a history of language shift that is similar to that European Americans and Canadians? Apparently, the language of the dominant economy was sooner or later learned by all other European immigrants and gradually prevailed over the other European languages, and the more competitive national groups (such as the Germans in North America) were the last to assimilate culturally and give up their vernacular (Mufwene 2009a). Or is the case of the Italian Argentinians similar to that of, for instance, Italian and Lithuanian Americans, who came late, lived in segregated communities, and lost both their national languages and their nonnative varieties of the colonial language as gradually as they became integrated in the dominant population, albeit a dominant population of European descent? In Argentina, what are the social-ecology conditions that prevented Cocoliche, the allegedly pidginlike variety spoken by Italian immigrants in the twentieth century, from evolving into a full-fledged vernacular and therefore a separate, Italian-Argentinian Spanish variety? How similar are these ecological conditions to those that account for the short lives of Italian, German, and Yiddish Englishes in the United States? Lipski does not address these questions, but the reader should bear them in mind in order to assess the complexity and diversity of language con-

[ 12 ]

chapter one

tact in the New World from a comparative perspective, as outlined, for instance, by Diamond and Robinson (2010). Here, Lipski is more concerned with the impact of demographics and the order of speakers’ arrival on the evolution of a language under contact conditions. He asks, “How many speakers of one language or dialect are needed to leave a permanent imprint on the evolving Spanish American varieties?” For him, such demographics, to be found in the large numbers of Andalusians (and Extremadurans [Mal-Molinero 2000, 30]) in the Caribbean in the fifteenth and sixteenth centuries, can explain why Caribbean Spanish is different from continental Latin American Spanish, by the founder effect (Mufwene 1996, 2001). Putatively, the continental varieties bear more influence from Castilian, especially since the seventeenth century, the period during which the Caribbean no longer was the diffusion point of the Spanish colonial expansion. Lipski also speculates on the probable time of the autonomization (Chaudenson 1992, 2001) of Latin American Spanish from Peninsular Spanish: By comparing linguistic innovations occurring in Spain since the early sixteenth century with emerging traits of Latin American Spanish, it is possible to identify with some accuracy the period in which Latin American dialects ceased to reflect major innovations occurring in Spain; essentially, by the 1700s most innovations in Spain did not pass unconditionally to Latin America.

Contrary to much of the literature on the evolution of European languages in the New World, Lipski also argues that Native American languages may have influenced Latin American Spanish not only lexically but also syntactically: “In a racially and socially segregated environment such as existed in colonial Latin America, Spanish is used not only for essential contacts with the population of European descent, but also among members of the same indigenous community.” This substrate influence would have been facilitated by the fact that “during the 16th century and even later, indigenous populations often outnumbered Europeans by a factor of several thousand to one” (chap. 2, this volume). We can learn more about the possibility of such influence from ethnographically informative accounts of how language shift occurred among Native Americans in the same part of the world, where the proportion of those that have survived European colonization to date is larger than in Anglophone North America (Nettle and Romaine 2000). By the founder principle, the later the time of the shift, the less influence the relevant ethnolinguistic groups exerted on the dominant language (Mufwene 2008, chap. 10).

Salikoko S. Mufwene

[ 13 ]

In chapter 3, “Amerindian Language Islands in Brazil,” Hildo do Couto focuses on the impact of Portuguese colonization on indigenous languages. He prefers the notion of language island to that of language enclave, because “it suggests that the territory of the relevant population [. . .] is a kind of island within another population [. . .] analogized to an ocean. Further, it implies that there is a hinterland from which the ‘island’ is somehow detached and to which it remains related” (chap. 3, this volume). On the one hand, this conjures up the question of the ambient ecology that has enabled the islands’ populations to maintain their languages, while the spread of Portuguese has been driving the other languages to extinction. On the other hand, we should also ask whether, in the history of Brazil, Portuguese was the only language to have spread and endangered the other languages. As in the chapters by Denny Moore and by M. Kittiya Lee, we learn that Portuguese colonization helped Língua Brasílica (a.k.a. the Brasílica) and, later, Língua Geral Amazônica (a.k.a. the Amazônica) spread in much of Brazil. The question arises whether this indigenous lingua franca did not contribute to the endangerment and loss of other Native American languages. A population very much engaged in trade even before the arrival of Europeans in Brazil, the Tupinambá became important power brokers, especially after the adoption of their language (identified also as Língua Geral) by Portuguese traders and missionaries as the lingua franca for interactions with the indigenous people and for Christianization (Lee, chap 5, this volume). The advantages that the language conferred because of its association with economic power and, indirectly, with the European civilization appear to have caused language shift, especially in Amazonia. Quite significantly, as noted by Moore (chap. 4, this volume), the “descending” (or forced relocation) of Native Americans to missions or colonists’ settlements, where they were under pressure to communicate in Língua Geral, contributed to the attrition or loss of indigenous languages. Thus, as in Black Africa, European colonization endangered or disadvantaged several indigenous languages not necessarily by promoting and spreading a European language but often, if not typically, by facilitating the further demographic expansion of specific indigenous ones that had already been major or expanding languages before the Europeans arrived. The spread of Língua Geral is indeed comparable to that of, for instance, Wolof, Sango, Lingala, Kituba, and Swahili in Black Africa, although, perhaps because it was colonized typically on the exploitation (rather than the settlement) model, Africa has experienced more multilingualism than language shift, especially in rural areas (Mufwene 2008). Languages, like viruses (Mufwene 2001, 2005, 2008), spread through

[ 14 ]

chapter one

carriers, those who speak them and use them to interact with new individuals, who in turn learn them. In the case of Brazil, Couto explains that the mestizos seem to have played a critical role in spreading Portuguese, especially those mestizos who worked in the colonial administration, whereas, as we also learn from Lee (chap. 5, this volume), Língua Geral was spread by trade, through some European marooned “factors” who had learned the indigenous language, and through Tupinambá interpreters who had learned some Portuguese.15 In Amazonia, Língua Geral, then also called Língua Amazônica, eventually evolved into Nheengatu. Couto is also interested in the ways the meanings of traditional terms were modified to suit the colonizer’s culture. Thus, oka ‘hut’ combined with pukú ‘long’ into oka pukú to designate ‘street’, and tba ‘hamlet’ combined with usú ‘large’ into tabusú to denote ‘town’. These examples show how adaptive a language can be to a new culture that its speakers are brought into contact with. Likewise, Lee (chap. 5, this volume) tells us that Native Americans renamed European products, such as scissors and knives, in their own indigenous terms (thereby extending their meanings), suggesting less tolerance for borrowings or, perhaps, that the intercultural contacts were not intimate enough to lead to borrowings. Indeed, in interpreting Media Lengua as Quichua influenced by extensive lexical borrowings from and structural convergence with Spanish, Shappeck (2011) also underscores the role of increasing contacts between Spanish and Quichua either as speakers of the latter moved to the city or as Spanish gradually spread to rural areas. Couto also reports that the vitality of Língua Geral was not uniform throughout Brazil. The southern variety known as Língua Geral Paulista died in the early nineteenth century, while Língua Geral Amazônica (LGA) was spreading. In rural areas, it died after rural basilectal Portuguese started rooting as the dominant vernacular. (Ball, chap. 10, this volume, gives more information about how late in the twentieth century Portuguese penetrated some parts of Amazonia and the differing ways in which it was received by the indigenous populations in Upper Rio Negro and the Upper Xingu.) An important factor in this ongoing shift to Portuguese seems to be rural exodus, with the city then exerting more pressure on incoming Native Americans to adopt its vernacular: Portuguese. What applies here is the same basic principle that Moore (chap. 4, this volume) associates with the practice of “descending,” in that the people who relocate typically adopt the language of the host population, regardless of whether it is associated with economic power, although the latter factor is far from irrelevant. One is thus reminded not only of language loss in other European settlement colonies, in particular Anglophone North

Salikoko S. Mufwene

[ 15 ]

America and Australia, but also of how the Celtic populations gradually Latinized in the now Romance Europe, first in the city and later in rural areas (Landa 2000). In addition, Couto looks into the reasons why Native American languages are dying today in “population islands” such as the Indian Reservation of the Xingu River. According to him, the relevant languages are typologically and genetically different from one another and no group is interested in the other’s language, while Portuguese is turning into a useful lingua franca. Also, children of interethnic marriages, especially those involving whites, learn Portuguese as their mother tongue. The situation is much the same with other “language islands,” with competence in Portuguese seen as a sign of progress. Urbanization, proceeding at the rate of 70 percent, is also taking its toll, although some ethnic neighborhoods maintain their heritage languages as vernaculars. To be sure, some indigenous languages have emerged as lingua francas, but those who might have been native speakers of the same indigenous languages prefer other languages as their lingua francas. Multilingualism may be a transition toward monolingualism in Portuguese, an evolutionary trajectory that has taken place in other settlement colonies and also in Romance Europe. In general, out of the estimated 1,200 languages that were spoken when the Portuguese first arrived in Brazil, only 180 are still spoken today, typically in “language islands,” where European influence is still marginal. Christopher Ball’s chapter deals with this very subject matter. Denny Moore confirms in Chapter 4, “Historical Development of Nheengatu (Língua Geral Amazônica),” that Língua Geral functioned as the lingua franca in most interactions of the Portuguese with Native Americans in Brazil until the seventeenth century. It may even have functioned as a vernacular in mixed unions. Colonial contacts during the first two centuries, not only between Europeans and Native Americans but also between different indigenous groups (such as in the Christian missions), created ethnographic conditions that were favorable to the demographic and geographical expansion of this indigenous language. Língua Geral was also spread by the Tupinambá and other indigenous populations fleeing northward from the expansion of the Portuguese. Although the size of the ethnic Tupinambá population declined, along with the populations of other Native American groups (just as in other European settlement colonies, largely because the people lacked immunity to illnesses brought from the Old World [Crosby 1992]), Moore notes that the name was extended to other populations culturally related to the Tupinambá, especially those who had also adopted Língua Geral as their vernacular. Missionaries were among the first to learn this and other

[ 16 ]

chapter one

indigenous languages, because competence in them was a more expeditious way of Christianizing the Native Americans than waiting for them to learn Portuguese or relying on linguas ‘interpreters’. Moore is also informative about how geographical ecology influenced the speed of the colonial expansion. The Portuguese gained access to Brazil, a large territory, from the south. It took a whole century for them to reach Amazonia, owing largely to the challenge of penetrating the rain forest. As noted, that region is still being penetrated now, by the use of novel technology that reduces exclusive dependence on rivers as highways. The ecological difficulties also explained why the demographic importance of Portuguese colonists remained very small in Amazonia. An interesting ethnographic division of labor followed, according to which the usage of Portuguese was restricted to senior colonial administration, the original Língua Geral (Brasílica) was used for liturgy, and the Amazônica (the emergent northern variety of the same language) was used for “popular communication” and for “religious instruction” (Freire 2004, 118, cited by Moore).16 Portuguese was apparently so marginal to the lives of people in Amazonia that it exerted little influence, mostly lexical, on the Amazônica, although, according to Denny Moore (pers. comm., Jan. 10, 2013), some syntax of Nheengatu, as the language is called today, eventually converged with that of Portuguese (e.g., the SVO major constituent order and the structure of embedded clauses). However, things changed in the mid-nineteenth century, when steamships started facilitating the penetration of the rain-forest interior of Brazil and urbanization increased and was no longer limited to the coast. Technological and economic changes fueled the penetration of the Amazon by the Portuguese and Brazilians, especially from the south, and therefore encouraged the spread of the Portuguese language at the expense of LGA. This demographic expansion also precipitated the decline of the latter language. Moore observes that “by the end of the century, LGA, or Nheengatu, was mostly spoken in western Amazonia, in the less accessible regions.” His conclusions include the following: “It is clear that the evolution of the language has been complex; each generation of learners has modified it using the patterns available to them. Nheengatu has been more an alternative to Portuguese than an imitation of Portuguese.” What is less evident is whether it is useful to treat LGA as a pidgin or a creole, although it is a contact phenomenon. Readers may decide for themselves. In Chapter 5, “Language and Conquest: Tupi-Guarani Expansion in the European Colonization of Brazil and Amazonia,” M. Kittiya Lee corroborates Couto’s observation that the domination of part of coastal eastern Brazil by the Tupinaé had preceded the arrival of the Portuguese in 1500.

Salikoko S. Mufwene

[ 17 ]

They were in turn dominated by the Tupinambá, a related population, who spread their language, which was later identified as Língua (Geral) Brasílica by the European colonists. This functioned as the unofficial lingua franca between the Native Americans, the Europeans, and the Africans, before Portuguese started to prevail in the seventeenth century, as also noted by Moore. Lee associates the spread and the increasing dominance of Portuguese with the growing number of Portuguese colonists involved in sugarcane cultivation and in the gold rush after gold was discovered in Minas Gerais in 1690. The use of Portuguese in regular interactions with the African slaves was also a significant factor. The other social ecological factor is the emergence of the mestizos as a distinct group (as also noted by Couto in chapter 3); they came to play a more and more important role in the administrative expansion of the colony, thus reducing the status of the Tupinambá as intermediaries between the Europeans and other indigenous populations. In the process of this colonial expansion, the Tupinambá, preceding the Portuguese and their auxiliaries, spread farther and farther into Amazonia, although, according to Denny Moore (pers. comm., Jan. 10, 2013), most of them were exterminated by the late sixteenth century. According to Lee, LGA prevailed because of the prestige of the coastal Tupinambá, as the river transportation system brought the indigenous people more and more into contact with them and the Europeans. According to Denny Moore, the increased contact was primarily the result of the active policy of “descending” indigenous groups noted above. “The captured people had to speak LGA to talk to those controlling them and to each other, since they were from various groups” and spoke different languages (pers. comm., Jan. 10, 2013). Note that it must also have been easier for the other Native Americans to learn LGA than Portuguese, since the indigenous languages are apparently similar typologically. Or could the Native Americans have been better disposed to LGA because their cultures are more similar to one another than to the European culture? These reasons need not be mutually exclusive, in any case. Like the missionaries, who produced catechisms in Língua Geral, the European traders (including some Dutch, French, and Irish) were also less invested in using Portuguese. That LGA interested the European merchants is evidenced by the sheer number of word lists that, based on Lee’s account, were produced during the sixteenth century. The brevity of the lists and their specific contents also suggest that the European traders must have usually relied on interpreters and learned mostly the words for local goods that they needed to know. In fact, they spent most of the time aboard their ships. The interpreters often consisted of “factors” and Tupinambá, the people who had been trading directly with the Por-

[ 18 ]

chapter one

tuguese since the latter arrived in 1500. Confirming Couto’s account of the reluctance of Native Americans to borrow words from European languages, Lee reports that the European word lists include Native American terms for European goods such as knives and scissors. The contacts were undoubtedly too limited or indirect for European lexical items to have made their way into the indigenous languages, at least during the early stages of colonization. According to Denny Moore (pers. comm., Jan. 10, 2013), things apparently changed, and “the whole colony of Brazil [i.e., the parts with Portuguese settlements] was soon bilingual.” As John Lipski points out in chapter 2, the linguistic significance of the presence of Africans in Latin America during the colonial period, variable though it must have been, cannot be overlooked from the perspective of language evolution, especially in places where the slaves constituted an important proportion of the colonial or the postindependence population. This is what Heliana Mello focuses on in chapter 6, “African Descendants’ Rural Vernacular Portuguese and Its Contribution to Understanding the Development of Brazilian Portuguese.” This topic is especially important because the Brazilian colonial population does not appear to have been as segregated as the colonial societies of the English and French Caribbean, for instance. This is apparently why no particularly black Portuguese variety has emerged in Brazil that can be identified as a creole, notwithstanding the absence of particular structural criteria for identifying a particular language variety as “creole” (Mufwene 1986, 2000, 2008; DeGraff 2001, 2005, 2009; contrary to McWhorter 1998). Mello begins by addressing how the Portuguese colonists of the early sixteenth century established communication with Native Americans, since neither group knew the other’s language. She takes us into a critical phenomenon that has too often been overlooked in studies of the emergence of creoles and pidgins, the role played by interpreters (already highlighted by Couto, Moore, and Lee), although this conjures up the issue of how that class of intermediaries emerged in the first place. While Lee speaks of “factors,” Mello invokes the degredados, whose “explicit assignment” was, according to her, “learning the local languages.” (chap. 6, this volume). They are the same kind of individuals identified elsewhere as lançados, undesirables that were expelled from Europe and not welcome to return. They were forced by the circumstances of their existence in exile to learn the language of the host population. Colonization found them useful in return; hence the term factor used by Lee, with the meaning ‘business agent’. These are the unsung heroes of colonization (Mufwene 2005, 2008) who helped facilitate trade and later the domination of the colonies.

Salikoko S. Mufwene

[ 19 ]

Another important group in Brazil, at least regarding communication with the African slaves, consisted of negros de reino (from Portugal), whose role Naro (1978, 1988) discussed in relation to the emergence of creoles, and negros ladinos (the acculturated slaves from some Sʋo Tomé). These are distinguished from negros boçais, known in the non-Hispanic literature as Bozals, slaves who had freshly arrived from Africa and had to be acculturated by the creole (locally born) slaves or the “seasoned” ones (acclimated already) who had preceded them; they were typically relegated to hard work in the field (hence the term field hands in the North American literature). Like the Palenque of San Basilio in Colombia, the Brazilian Quilombos were apparently settled primarily by Bozal slaves, although they were joined by other populations. This is one kind of contact setting where strong substratist hypotheses on the emergence of creoles can be tested. In the case of Palenquero, the evidence of such influence on grammar is rather limited (Moñino 2007). Mello concludes that in the case of the Quilombolas’ vernacular Portuguese, the influence is limited to the lexicon, including some idioms. Could Chaudenson’s (1992, 2001) hypothesis on the absence of creoles in Cuba and the Dominican Republic despite the large number of slaves until the late nineteenth century (discussed above) also apply in this case? According to Brazilian scholars such as Hildo do Couto (pers. comm., 2005), Brazilian Portuguese (BP) is now as distinct from European/ Peninsular Portuguese (EP) as American English is from British English; they have developed their own autonomous (standard) norms. In chapter 7, “Brazilian Portuguese and the Ecology of (Post-)Colonial Brazil,” J. Clancy Clements argues that BP and EP are, nonetheless, national varieties of the same language, although they continue to diverge, perhaps also like American and British Englishes. It is irrelevant whether or not they, or some of their subvarieties, are mutually intelligible. According to him, blacks and pardos (African-European mixed-raced people) are the primary reason BP is diverging from EP: “The large number of Africans and their descendants played an integral role in the introduction of new features into VBP [vernacular BP] and in the propagation of these features among the population at large.” However, the question is whether the features associated with the divergence are necessarily of African origins. Or are they features that were simply favored by their congruence with those of some African languages in the way explained by Corne (1999) for French creoles? Noteworthy in this regard is Clements’s own observation that, unlike the situation in the Caribbean, “the concentration of the African population was also in smaller units.” Citing Schwartz (1985, 440), he observes:

[ 20 ]

chapter one

By the advent of the nineteenth century, there were two types of agricultural operations in plantation-heavy areas such as Bahia: the established plantations that depended on the use of large numbers of slaves and the newly formed subsistence farming operations in the Recôncavo area (in Bahia), in which around 70% of the slaveowners held fewer than five slaves each and only twenty-five owners possessed more than twenty slaves. There were few sugar mills in the area; only three slaveowners held more than fifty slaves. (chap. 7, this volume)

This demographic distribution of slaves is very different from that in the English and French Caribbean and in coastal Georgia and South Carolina, where the overwhelming majority of slaves were owned by a small minority of powerful plantocrats who possessed more than one hundred slaves each. There may also have been more reliance on slave labor on English, French, and Dutch Caribbean, circum-Caribbean, and North American plantations, which sometimes grew into huge estates, especially in the Caribbean, than in Brazil. The latter colony is also a historically curious phenomenon because the slaves were not exclusively or predominantly allocated to the agricultural industry. During the eighteenth century, a large proportion of them were employed in the mines of Minas Gerais. Since no variety of VBP has emerged that is exclusively associated with descendants of slaves, unlike what obtains in places where creole vernaculars have emerged, Clements’s chapter, like Lipski’s, raises the question of articulating the particular ways in which the blacks and pardos contributed to the divergence of BP from EP. Note that the decrease of the black and pardo populations in the nineteenth century might not have eroded the contributions of Africans to the structures of BP, any more than the founder effect of the English/Anglo population on the emergence of American versions of English was weakened by the increasing white American populations of continental European origins in the United States (Mufwene 2009a). The reader should also note that, on one hand, unlike Naro and Scherre’s (2007) view, which privileges the EP element, there is significant local-contact-based scholarship that has sought to underscore the contributions of African languages (see, e.g., Florin and Petter 2008; Roncardi and Abraçardo 2008; Lima and Carmo 2008; Galves et al. 2009; Negrão and Viotti 2011). On the other hand, since Clements acknowledges that only one of the ten features he has investigated cannot be traced back to Portuguese, it may not be irrelevant to clarify here that substrate influence need not be reduced to what Allsopp (1977) identifies as “apport,” that is, forms or constructions introduced from the substrate languages but not attested in the lexifier. It may amount to the role that the

Salikoko S. Mufwene

[ 21 ]

substrate languages played in favoring particular options, giving selective advantage to features that were congruent with their own (Mufwene 1990, 2001, 2005, 2008, 2010; Corne 1999; Chaudenson 2001), perhaps especially those that Gilman (1986) identifies as “areal,” in the sense that they spread over geographical areas that straddle different language families. Future research will, hopefully, sort things out. In Chapter 8, “Maya and Spanish in Yucatán: An Example of Continuity and Change,” Barbara Pfeiler takes us to the topic of language coexistence and competition in Mexico. She addresses the question of why Spanish has not replaced Maya in the Yucatán Peninsula. Citing Gabbert (2004, 21) and Farriss (1984, 112), she notes that Maya remained the primary and legal language of the colony until the nineteenth century and was spoken not only by “Indians, mestizos and pardos” but also by members of the “lower castes.” It was also used by the missionaries for the evangelization of the Native Americans. Thus, efforts to spread Spanish in the Yucatán Peninsula started rather late, allowing Maya time to carve its space in the new world order. According to Pfeiler, “The Yucatán Peninsula was not truly integrated into the rest of Mexico until local railroads were joined with the national rail network in the late 1950s. This isolation is one of the most important factors affecting the formation and use of Yucatán Spanish, a variant differing from that of central Mexico” (chap. 8, this volume). From an ecological perspective, the spread of Spanish in Yucatán can be correlated with the expansion of the modern economy, which began with the cultivation of henequen, the building of railroads, and the changes brought about by those events. However, consistent with the emergent population structure of Mexico, Maya continued to be spoken by some Native Americans who were not involved in the industry. The factors that have subsequently affected the demographic attrition of speakers of Yucatán Maya include the increasing immigration of outsiders to Yucatán, thanks to the new economic system, and the emigration of Mayans to urban centers and away from the region to other parts of Mexico (and to the United States) since the mid-twentieth century. While the immigrants contributed to the spread of Spanish, those who left have had to adopt the language of their host populations, Spanish in Mexico and English in Anglophone North America.17 Thus, both the immigration of outsiders to and the emigration of locals from Yucatán have exerted pressure on Yucatecans to adapt to the new socioeconomic ecologies. Both in and outside the homeland, these new ecologies made it hard for parents to transmit their heritage language to their offspring. From the point of view of structure, Maya and Spanish have borrowed words from each other. Maya has indigenized the Spanish loan words,

[ 22 ]

chapter one

enabling a lot of morphological derivations that would not be possible in Spanish. Different parts of Yucatán exhibit varying degrees of borrowings from Spanish. The extent of these can be correlated with the duration and intensity of contact with Spanish. Spanish verbs are seldom used by speakers who have had little exposure to Spanish. However, the influence of Maya on Spanish has contributed to making the local/regional variety of Spanish singular in comparison to other Mexican Spanish varieties. Even function words have been borrowed into Spanish, although nouns are used with Spanish derivational and inflectional morphology. Bilinguals also use the Mayan word order in Spanish, apparently in a manner similar to the convergence that Shappeck (2011) invokes to account for structures of Media Lengua in Ecuador. For instance, adverbs are placed before verbs. Although Maya has survived the competition of Spanish to date, there is every reason to wonder whether it will endure in the future. Note that, despite the large number of its speakers (close to 1 million), the colonial/ imperial language is also spreading as a vernacular. Pressure is indeed increasing to the disadvantage of Maya. More and more parents are transmitting Spanish, instead of Maya, as a mother tongue to their children in order to prepare them adequately for competition in the school system and in the job market. According to Pfeiler, “To a certain extent this is to be expected since, for example, one of the basic requirements for qualified employment or higher education is fluency in Spanish; no such requirement exists for Maya” (chap. 8, this volume). I should clarify, against those who have usually invoked the prestige of the dominant language in such cases, that it is the lucrative aspect of Spanish that appears to drive the shift here. Prestige would be only a secondary reason for a population that is simply adapting to changing socioeconomic conditions. Also, everybody is not learning standard Spanish in any case. As noted above, exodus out of the region is driven by economic pressures, which are working against the maintenance of Maya. Like Maya in the Yucatán Peninsula, Quechua played an important role in the Spanish colonization of parts of South America. Alan Durston treats this topic in chapter 9, “Standard Colonial Quechua.” He is interested in how the Spanish colonists used and reinvented Quechua as a lingua franca, proceeding to develop a “Standard Quechua” out of what appears to have been a cluster of languages.18 According to colonial and Christian records of the sixteenth century, Quechua was variable. The new variety, Standard Colonial Quechua (SCQ), may have been based on what was spoken in Cuzco, the capital of the Inca Empire. Its grammar was simplified and its region-specific lexical items were replaced by those more widely used. Durston observes: “It appears that more than enhancing intelligibil-

Salikoko S. Mufwene

[ 23 ]

ity among indigenous audiences that spoke other varieties of Quechua, SCQ was intended[, ironically,] to be easily learned by Spanish priests.” He also emphasizes that “SCQ was not a pidgin—it retained the full range of grammatical categories present in its base variety. Arguments that SCQ reflects Spanish influence at the grammatical level, or even a failure to understand Quechua grammar, seem unfounded” (chap. 9, this volume). Durston highlights the important role that mining in the Peruvian highlands played in the spread and modification of Quechua as a lingua franca, based on the southern Peruvian variety. “Conciliar Quechua” (fabricated by Spanish missionaries and first used in writing) is what forged its relative homogeneity as a lingua franca, the variety that was further spread in the Spanish colonies. Then arises the question whether SCQ ever became a vernacular. According to Durston, “It appears that in some respects SCQ stood in relation to Central Quechua varieties as [Classical] Latin did to Romance languages in medieval and early modern Europe” (chap. 9, this volume). Thus a diglossic situation existed, instead of competition between vernaculars or lingua francas, and a language shift became less likely to occur. Recall that it was Vulgar Latin that spread as a vernacular, and as it started in the emergent Roman-style urban centers, it was in competition with the indigenous Celtic vernaculars in the relevant part of the former Roman Empire. The Romance languages that Durston refers to are the ultimate outcomes of the indigenization of Vulgar Latin. This is a process that has at times aptly been analogized to the emergence of creoles from the contact of nonstandard varieties of European colonial languages with non-European vernaculars, although one need not call the Romance languages “creoles” at all. The ethnography of this contact situation in the Inca Empire appears to be similar (though only partly and only for a while, as we see below) to, for instance, that in postcolonial Black Africa, where European colonial languages function as official languages and as elite lingua francas that are barely intelligible (or not at all) to the majority of the national populations and are not at all in competition with the indigenous vernaculars. Did the Spaniards ever attempt to impose SCQ in their colonies? asks Durston. “Did they rely on it as an administrative medium? The answer [. . .] seems to be no.” Not every colonist, or even every ecclesiast, learned SCQ. “Lay officials appear to have received no training in Quechua at all. Instead, both lay and ecclesiastical justice and administration relied on interpreters or lenguas, preferably local Spanish ones rather than Indians.” SCQ must have been part of the transition to a Spanish-speaking empire. The Spanish administrators eventually imposed Spanish on the indigenous population partly to get rid of the interpreters. Thus, “As SCQ

[ 24 ]

chapter one

ceased to be widely cultivated by the clergy, its use as a medium of mundane written communication among the indigenous elite also seems to have waned, perhaps because they were no longer exposed to SCQ pastoral imprints.” SCQ was gradually replaced by colloquial and apparently less artificial Quechua. Note, however, that the Spaniards did not fabricate SCQ out of thin air. Some variety of Quechua had apparently functioned as a lingua franca before they colonized the region. However, Durston also remarks: “The term lingua franca evokes a medium that enables exchange between language communities. SCQ does not fit this definition very well; in many respects it was designed as an instrument of control and restriction rather than of wide communication” (chap. 9, this volume). This comment raises the question whether Quechua was the only language to be so used. What about Aymara, which is still spoken (albeit by a decreasing number of speakers), for instance? Was it ever used by the colonists or their auxiliaries for communication with the natives? The significance of the question is underscored by the following observation by Durston: “Varieties of Quechua certainly did function as lingua francas for communication among Indians from different areas who converged in various colonial centers as labor draftees or migrants, as Itier [2000, 2001] has suggested. However, I see little evidence that there was such a thing as a pan-Andean or even pan-Peruvian Quechua lingua franca” (chap. 9, this volume). Durston concludes, on one hand, that “the degree to which the colonial regime effectively promoted Quechua beyond strictly pastoral contexts has been overstated.” On the other hand, it is not evident that the Spaniards, or the Portuguese for that matter, treated all Native American languages equally. Quechua may have been used in the same way Tupinambá, dubbed Língua Geral, was in Brazil. Although in both cases, as in fact in that of Yucatán Maya, the colonial language was promoted to replace the regional language as the lingua franca and ultimately as a vernacular, the early reliance on the indigenous language extended its life. Now it appears logical to ask whether the other indigenous languages were threatened just by the European languages or also by the Native American one promoted by the colonial regime. This consideration does not, of course, lessen the role of colonization as the disturber of the “balance of power” in the indigenous linguascape. Christopher Ball adduces another range of ecological factors that shed more light on the contacts and the ensuing competition between Native American and European languages in remote areas of Brazilian Amazonia. In chapter 10, “Linguistic Subjectivity in Ecologies of Amazonian Language Change,” he focuses on “the speaking individual” as an important ecological factor affecting language evolution. He explains “subjec-

Salikoko S. Mufwene

[ 25 ]

tivity” as follows: “This can be basically stated as equivalent to the claim that in some cultures it might suffice for people to be speakers of only one language to count as fully developed subjects, whereas in another, speaking two, three, or more languages may be a requisite feature of human subjectivity.” In Vaupés, the multilingualism in question does not affect women, those who undergo exogamy and virilocality, in the same way as it does men. In their new residence, women become speakers of minority, nonlocal languages, while their children grow up bilingual, with native competence in both the mother’s and the father’s languages, the father’s language being ethnographically the dominant one. Whether or not the language of a village is endangered ultimately depends on the proportion of men who stay in the village and do not migrate to the city or the Christian mission and, in some ways, on the extent to which children remain attached to their mothers’ languages. However, historical changes that brought forced language mixing in mission schools and increased shifts of linguistic communities to Portuguese and Tukano [the regional major language] have challenged this culturally Vaupés way of becoming a plurilingual person. [. . .] The kind of mixing that was inadvertently encouraged by relocation of children to mission schools [. . .] fostered language change including the rise of Tukano as a regional lingua franca at the expense of smaller Tukanoan varieties. (Ball, chap. 10, this volume)

The new world order introduced by colonization also turned women, still abiding by exogamy and virilocality, into important agents of language shift in favor of Portuguese. This occurred as they moved to the city, married Portuguese-speaking men (or married and then moved), and raised Portuguese-speaking children who were not very interested in the indigenous languages of their mothers. The demographic increase of Portuguese speakers among the natives has meant the decrease of speakers of heritage languages, especially the minor ones. The trend has also been exacerbated by the heightened socioeconomic status of speakers of Portuguese, which has prompted more shifts to it by individual speakers.19 Curiously, the natives blame language mixing, rather than language shift, for the loss of some indigenous languages. In their view, their languages are dying by transformation rather than by shift, the ethnographic process that has preoccupied linguists the most in the past couple of decades.20 The mixing supposedly happened in the Salesian missions, where school children were forced to speak Portuguese but also came into contact with other indigenous languages, which they learned in order to remain “plurilingual.” The languages of minorities in the mission schools

[ 26 ]

chapter one

were influenced by those of larger groups. According to Ball, the facts remind us “that simple equations of the loss of language with the loss of culture are often misguided” (chap. 10, this volume). Thus, contrary to much of the “ecolinguistic” literature (e.g., Nettle and Romaine 2000; Skutnabb-Kangas 2000; Mühlhäusler 2003), speakers who have lost their languages have not necessarily lost their traditional customs, such as exogamy and virilocality, nor the need to learn the language of their new place of residence, as in the present case. The experience of language contact has been different for the Upper Xinguans, who came into contact with the mainstream Brazilian society only recently and have never before been “rigorously missionized or subjected to slave raids and economic exploitation,” Ball points out. They also subscribe to “an ideology that attributes purity of ethnicity to individual monolingualism.” Despite contacts with other ethnolinguistic groups, “plurilingualism as a communicative property of individuals is downplayed, while isolation and ethnolinguistic boundaries are emphasized.” The use of Portuguese is domain-specific and associated with particular expertise, which, while indigenizing local Portuguese, has not affected the practice of most Upper Xinguan languages. This language contact experience has been unlike that of the neighboring Yawalapiti, whose “leadership cultivates a cosmopolitan linguistic identity.” Yawalapiti is thus dying, while the Upper Xinguan languages do not appear to be endangered yet. Ball concludes fittingly that in territories such as Brazil, “we should approach the ecology of language [evolution] with a particularist eye, rather than jumping to generalizations about types, predictive models, and overly powerful metaphors of unidirectional change or loss” (chap. 10, this volume).

3. The Big Picture The demographic and linguistic landscapes of Latin America have certainly changed over the past half millennium. If there is any doubt that Latin America has Indo-Europeanized genetically, because of the Hispanicization phenomenon (a process of race mixing that led Anglophone North Americans to coin the term Hispanic), no question should arise about its linguascape. Latin America has become linguistically Romance, insofar as Portuguese and Spanish are the vernaculars of the majorities of the national populations. One may also say that it has Latinized more faithfully than European Romance countries have, because most of the new colonial offspring of Spanish/Castilian and Portuguese, albeit their standard varieties, are still largely mutually intelligible with their metropolitan counterparts, unlike what one may claim of the European Ro-

Salikoko S. Mufwene

[ 27 ]

mance languages relative to Vulgar Latin, from which they evolved. We must bear in mind that population movements and language contact are as responsible for the divergence of Portuguese and Spanish Romance varieties from Latin as they are for that of Latin American Portuguese and Spanish varieties from their Peninsular ancestors. It is likewise noteworthy that, as pointed out by Lipski, the evolution of Spanish in Latin America, like that of Vulgar Latin in Iberia, has not been uniform. Aside from the fact that Castilian did not come in contact with the same (combinations of) languages, the Castilian monarchs colonized Latin America no more uniformly than the English did their New World colonies or even the now Anglophone North America in particular. Nor was Spanish, like Portuguese in Brazil and English in Anglophone North America, adopted at the same time by all the non-Spanish populations who came to speak it as a vernacular at some point. The target had mutated already, because of previous contacts with other languages, by the time another population was exposed to it, generations later, and of course no new learner would appropriate it faithfully in the first place. Although it is true that varying degrees of assimilation by and absorption into the native-speaking population have eliminated the emergence or maintenance of varieties associated with particular ethnic groups or nations of origin in Europe, critical mass and time of shift among the nonheritage speakers have been important ecological factors in introducing xenolectal elements that would enhance the divergence of an emergent variety away from its kin elsewhere. These factors apparently account for the divergence of Spanish in countries such as Argentina and Uruguay, because of the large proportion of Italian immigrants in the eighteenth and nineteenth centuries (Lipski, chap. 2, this volume), and also in places such as the Yucatán Peninsula, where, according to Pfeiler, its later adoption by Mayan speakers in the twentieth century made allowance for significant substrate influence. But we must also remember that generally the colonial linguistic feature pools did not replicate those of the metropoles and were not identical from one colony to another, either. Thus, the divergence of colonial varieties from the European varieties was to be expected even if Portuguese and Castilian had not come in contact with other languages in the New World colonies. The dynamics of European feature competition were not identical, as is evident from, for instance, Trudgill (2004), regarding what he calls southern hemisphere Englishes (including the varieties spoken in Australia, New Zealand, South Africa, and the Falkland Islands, among others). Generally, the feature recombination dynamics did not remain constant across colonies (Mufwene 2001, 2005, 2006, 2008), an approach more finely articulated by Aboh (2006, 2009).

[ 28 ]

chapter one

There is also something to be learned about the significance of demographic, political, and economic powers in helping the colonial varieties autonomize, thus legitimating their own separate norms, in the way explained by Chaudenson (2001). Former colonies can do this more easily after becoming politically and economically independent, as Latin American countries generally have. Countries such as Brazil, Argentina, and Mexico, whose respective populations are larger and whose economies may now be considered stronger than those of their metropoles, have norms that are treated with as much respect relative to the metropolitan ones as those of American and Australian Englishes relative to British Queen’s English. This is an evolution that is ethnographically both similar to and different from that of English in England’s former colonies. Speakers of “native Englishes” (chiefly the British and North American varieties) are generally less accepting of the “indigenized Englishes” spoken in the former exploitation colonies of Africa and Asia than they are of each other’s, which reflects their recognition of themselves as culturally (thus also politically and economically) more related than they appear to feel toward the former exploitation colonies.21 Such a power-based asymmetrical distinction does not appear to have arisen yet—or, perhaps, to have been articulated so stongly—between European Portuguese and Spanish, on the one hand, and their Latin American counterparts, on the other. Several chapters in this book also highlight the role played by indigenous and European interpreters (called lenguas in the Hispanophone world and often identified as linguists in Anglophone colonial history) in facilitating communication between the Europeans and the native populations during the initial stages of colonization. There is no mystery regarding how the “factors” (Lee, chap. 5, this volume) and degredados (Mello, chap. 6. this volume), as well as the mestizos and the pardos (Clements, chap. 7; and Pfeiler, chap. 8, this volume) learned the indigenous languages: by immersion within the native populations. We still have to find out more about how the Tupinambá traders and other indigenous gobetweens learned the European languages. Is it enough to invoke the fact that some of these “unsung heroes of colonization” (Mufwene 2008) were taken or kidnapped early to Europe as proofs of the explorers’ discoveries (in exchange for the human gauges left behind)? Naro’s (1978, 1988) negros de reino and negros ladinos apply to the African slave populations, not to Native Americans! In any case, it is evident that we do not have to invoke fictional pidginization (beyond the initial attempts by gestures) to account for the initial colonial exchanges between the populations in contact (Chaudenson 2001; Mufwene 2001, 2008, forthcoming). History makes it evident that the European explorers, traders, and early colonists everywhere relied heavily on interpreters, even in the

Salikoko S. Mufwene

[ 29 ]

Americas (see, e.g., Karttunen 1994; Gray and Fiering 2000; Fayer 2003; and Metcalf 2005, among others). As pointed out by Durston, the subsequent imposition of the metropolitan language as the official language, which eventually indigenized while spreading as the vernacular (given the ensuing advantages in the new world order), was intended to do away with the need for interpreters and to establish direct contact with the indigenous populations, to the extent that the Europeans wanted to involve them as partners in the emergent socioeconomic systems. The question is now a more vexed one; and it is also another matter whether everybody complied, including the Jesuit fathers in Brazil, who were more interested in proselytizing in the languages most familiar to the natives and often denounced the atrocities of the colonial rule. It is also evident that some major indigenous languages were widely used and spread, especially through the interpreters and the missionaries, during the early stages of colonization. Noteworthy agents of the spread were, according to Denny Moore, “the offspring of marriages to native women [i.e., the mestizos] or children raised by native maids” (pers. comm., Jan. 10, 2013). As noted above, some of those languages have survived to date. An important question suggested especially by Couto, Durston, Lee, and Moore in their chapters is whether these major indigenous languages did not play a role in driving other indigenous languages to extinction. This means that while colonization was the ultimate disruptor of the balance of power in the precolonial linguascapes, the immediate cause of the extinction of indigenous languages was not always the spread of the European languages. Ball’s and Pfeiler’s chapters also show that the indigenous populations did not all come into (regular) contact with the Europeans at the same time. Geographical or social isolation has protected some of them until recently; and some languages are being endangered now not only because outsiders are penetrating their traditional habitats but also because some who would otherwise be speakers are leaving home for economic opportunities in places where they cannot sustain the use of their heritage languages.22 There are also the questions why the indigenous languages have almost all vanished from the Caribbean islands and why the few that survive are moribund. Other than the atrocities committed by the Spaniards during the early stages of colonization, including mass murders and deportations (Casas 1992a, 1992b), does the current situation also reflect, on the one hand, the additional effect of the ills brought by the European colonists and, on the other, the small size of the territories and of the affected populations? Can the impacts of these ecological factors be sorted out? There is thus a complex nonlinear and nonuniform history of language

[ 30 ]

chapter one

contact and its consequences in Latin America that students of the endangerment of indigenous languages, especially advocates for their revitalization, should better understand. The comparative assessment provided by Michel DeGraff in chapter 11: “The Ecology of Language Evolution in Latin America: A Haitian Postscript toward a Postcolonial Sequel,” is very informative. No remedy can promise success unless the prescribers understand how the epidemic is spreading. I hope the reader will appreciate the wealth of information provided in the following chapters about language evolution in parts of Latin America during both the colonial and the postcolonial periods, and that this volume will function as an invitation for more comparative studies of similar phenomena in other parts of Latin America and other parts of the world. Notes I am grateful to Yves Moñino, Denny Moore, Michel DeGraff, and our anonymous reviewers for useful feedback on earlier drafts of this chapter. I assume full responsibility for all the remaining shortcomings. 1. According to Clements (2009), the Lusitanians had just chosen to form a separate kingdom, while the other counties were merged into the Kingdom of Castile and, by the fifteenth century, into the Kingdom of Spain. This merger cum Castilian political expansion proceeded through inter-Crown wars and marriages and through the Reconquista as the northern kingdoms helped free their weaker, southern counterparts from the Moors’ domination. Note, however, that to date Catalan is still considered a separate language from Spanish/Castilian. Like the Basques, the Catalans feel colonized and promote wider usage of their heritage language in official business. It is noteworthy that the spread of Castilian as the official colonial language, concurrently with the emergence of an integrated kingdom by the name of España, may have marked the birth of the Spanish language (español), since no Iberian language had been identified by this name before. 2. According to Schwartz (1985), these islands are where the Iberians first experimented with sugarcane cultivation, including the institution of plantation slavery. This was different from the domestic slavery in mainland Europe. Places like Brazil, with its far larger expanses of land and a geographic ecology (climate and soil) in the northeast that was more favorable, enabled the Portuguese in particular to develop sugarcane cultivation at an industrial level that became more lucrative. 3. This is a deliberately oversimplified history because our focus is the Americas. Up to the eighteenth century, the Portuguese had an important trade empire around the coast of Africa, on the coast of southern Asia and Southeast Asia, and all the way to China and Japan (Ostler 2005). After England, France, and Holland joined in the colonial venture from the seventeenth century onward, the Portuguese were unable to develop important colonies in South Africa, India, Sri Lanka, Malaysia, or Indonesia because of the competition. Portugal maintained its presence only in the small settlements mentioned in the text and the exploitation colonies of Guinea Bissau, Angola, and Mozambique. 4. To be sure, New Spain also included the Philippines, the Mariana Islands, the

Salikoko S. Mufwene

[ 31 ]

Caroline Islands, Taiwan, and parts of the Moluccas, but these territories are peripheral to the discussions in this book. Eakin (2007, 80) includes the Caribbean in the Viceroyalty of New Spain (his map 4), but he also says on page 83 that this constituted one of three “core” regions from which the Spanish Empire in the Americas expanded. It is not clear whether the name New Spain also started the trend for including “New” in the names of colonies, such as New England, New Netherland, and New France, all the way to New Zealand, and names of smaller settlements, such as New York, New Orleans, and New Hebrides. Such names appear to confirm Crosby’s (1986) thesis that the European settlers intended to create “new Europes,” better ones than they were leaving behind. 5. Gullah, spoken in coastal South Carolina and Georgia, appears to be exceptional in being associated with rice cultivation. Its basilect is also considered less divergent from its lexifier, nonstandard English, than those of other English creoles. 6. Convenience has traditionally included in this set not only creoles spoken in the Caribbean Sea (e.g., Haitian, Martiniquais, Jamaican, Bajan, and Papiamentu) but also those in Guyana and Surinam (e.g., Saramaccan and Sranan), which, strictly speaking, are in South America. I follow this tradition here. 7. Guinea Bissau’s creole appears to be related to Kabuverdianu in the same way that Sierra Leone’s Krio is related to Jamaican Creole, although, to be sure, these African mainland varieties are also local evolutions from varieties that had evolved earlier on the islands, corresponding thus to what Chaudenson (1979, 2001) identifies as “second-generation creoles/varieties.” 8. I use indigenous in quotation marks here simply to flag and disavow the frequent use of it by linguists as an absolute (rather than a relative) term that applies only to populations and languages outside Europe that are not of European descent. This nonrelative usage reflects especially the colonial perspective of Europeans who dispersed outside Europe from the fifteenth century on and encountered other populations they considered inferior. It suggests incorrectly that Europe does not, or did not, have populations and languages indigenous to it. English and the Romance languages are as indigenous to Europe as Afrikaans is to South Africa or as many creoles are to the Caribbean. They are all local formations from other languages that they survived or evolved from. Immigrants to Europe can justifiably refer to native (white) Europeans as indigenous. 9. Yucatán Maya, spoken by about seven hundred thousand people (Lewis 2009, accessed Dec. 24, 2011), can be added to the list. It is also significant that as a group the Mayan languages, spoken by a total of about 6 million speakers (Lewis 2009, Statistical Summaries, accessed Dec. 24, 2011), have all survived in Mesoamerica. 10. The fortunate, if not exceptional, fate of these major languages does not contradict the other fact that several indigenous languages were decimated along with the genocides of their speakers, under the atrocious regime of early Spanish colonization, as decried by, e.g., Casas (1992a, 1992b), especially in the Caribbean, where virtually no indigenous languages remain that are not moribund. 11. According to Shappeck (2011, ii), Media Lengua is the outcome not of relexification of Quichua/Quechua by Spanish, as argued earlier by Muysken (1981), but rather of “adlexification [i.e., lexical borrowings], code-mixing, and structural convergence.” It would thus be not an autonomous language with its own separate norm but the result of something similar to “code-switching” phenomena, much discussed

[ 32 ]

chapter one

in the linguistics literature over the past thirty years or so. Varieties similar to Media Lengua are now attested in rural Ecuador that may be characterized as “Hispanicized Quichua” and are concomitant with the gradual shift to Spanish (Shappeck 2011, iv). 12. Comparative linguists have typically been interested in identifying especially lexical and phonological “correspondences,” or shared forms, between languages. The correspondences enable them to produce cladograms, or taxonomic trees (Stammbaums in linguistics), which show whether the relevant languages are related genetically and, if they are, how. The kinds of genetic ties that obtain vary according to how clades representing the different languages are related, directly or indirectly, on the cladogram. In contrast, the practice of the comparative approach in population studies is more concerned with identifying ecological factors that account for why populations that appear to have similar histories and inhabit similar geographical habitats have evolved culturally (including evolution in their economic developments) in different ways. See, e.g., Diamond and Robinson (2010). 13. The fact that a language variety that is spoken primarily by descendants of African slaves diverges from the “mainstream” national or regional variety, spoken primarily by people of European or mixed descent, is not a sufficient condition for stipulating that it is a creole or has emerged by “creolization.” The putative process appears to be social and political, as there is no particular combination of structural features that can be claimed to be uniquely “creole” (Mufwene 1986) or any particular combination of structural processes that can be identified as “creolization” (Mufwene 2000). What, much to the embarrassment of creolistics, has implicitly justified singling out particular “contact languages” as “creoles” (as if there were any modern languages whose evolution has not involved contact) is divergence from the variety/ varieties considered “normal” evolutions and the association of their emergence with exogenous/nonnative non-European slaves (around the Atlantic Ocean and in the Indian Ocean) or contract laborers (in the Pacific). See elaborate discussions in Mufwene (2001, 2005, 2008) and DeGraff (2003, 2005). 14. The reader should remember that the term creole arose originally (in the late sixteenth century) in reference to people of non-European descent born in Iberian colonies; it was applied to language varieties much later, in the late eighteenth century, in non-Iberian colonies. Creole people, especially those of the homestead phase (i.e., when the relevant colonies had not yet developed plantations and the European and African populations were not yet segregated, either), did not necessarily speak creole language varieties (Chaudenson 1992, 2001). 15. As explained by Lee, “factors” were Crown-appointed local agents whose responsibility was “to negotiate all affairs between the [natives] and the Europeans” and whose “tasks included identifying profitable goods, persuading local headmen to trade, negotiating prices with [native] merchants, and arranging for the delivery of the [goods] to the factory,” their trade posts (chap. 5, this volume). This was apparently not a peculiarity of Brazil, since “factories” were also opened in every location where the Portuguese and other Europeans traded on the coasts of Africa and Asia, before these continents were subjected to exploitation colonization. The factors were typically Europeans who had been stranded in the new territories after surviving shipwrecks and had learned a local language, in this case Língua Geral. Along with the interpreters, they eliminated the need for incipient jargons or pidgins during these early contacts of Europeans with non-Europeans, contrary to the literature in creolistics, a point I will elaborate in future work (e.g., Mufwene, forthcoming).

Salikoko S. Mufwene

[ 33 ]

16. I am personally reminded here of my own experience in elementary school in the Bandundu Province of the now Democratic Republic of Congo in the 1950s. The variety of Kikongo that missionaries used for liturgy was called Kikongo-Kisantu (literally ‘sacred Kikongo’). They used it in the headquarters mission they had founded earlier in Kisantu (sacred place) on the Lower Congo River, where varieties of ethnic Kikongo are spoken. It was very different from the vernacular/vehicular Kikongo spoken locally/regionally, which is now identified in the creolistics literature as a creole and by the name (Kikongo-)Kituba (Mufwene 2009b). Kikongo-Kisantu was apparently also concocted by the missionaries themselves, based on a stereotypical canon of Bantu morphosyntax, as they thought that the contact-based Kituba was too bastardized and inadequate for the Bible and liturgy. It is not evident that KikongoKisantu preserved the complex tonal patterns of ethnic Kikongo any more than Kituba did; the European colonizers were generally not particularly good with our tonal systems. 17. This process conjures up Denny Moore’s invocation of the “descending” of Brazilian Native Americans to work away from home. Relocation forced them to adopt the language of the locals, contributing to the demographic attrition of their native vernaculars. 18. Other powerful outsiders or rulers in other circumstances have manipulated local languages in similar ways. For instance, what is usually referred to as Kikongo in central Africa is a construct of convenience for a cluster of related languages spoken in the former Kingdom of Kongo. The same thing happened in southern Africa in the nineteenth century; numerous minor Bantu languages were lumped together as Setswana, Shona, Sotho, and so forth. It seems to be true of the cluster of languages treated indiscriminately by outsiders as Chinese, a confusion well exploited by the Chinese government for national integration. 19. A uniformitarian must wonder to what extent these social practices may shed light on the dynamics of the spread of Vulgar Latin in southwestern Europe. Current and recent histories must also prompt interesting questions about some historical events, regarding which, to be sure, the information (made) available to us is usually incomplete, especially regarding language evolution. We must remember that the shift to Vulgar Latin, and later to the emergent Romance languages, was not necessarily as abrupt or as wholesale as typically suggested in (Romance) historical linguistics, in which gradual changes have typically been telescoped. 20. See Mufwene (2005; 2008, chap. 10) regarding the alleged “decreolization” cum “debasilectalization” of creoles as a form of language death, albeit by transformation, through the adoption of structures of the lexifying language. 21. It would of course be naive to believe that the reasons articulated here for this ethnographic distinction are the only ones (Mufwene 2001, 2008). Linguists who have embraced this line of obviously ethnic/race-based discrimination about language evolution have as much soul-searching to do in the practice of their research as those who reject such ecological considerations out of hand. It is noteworthy that in the by now traditional distinction between “native” and “indigenized” Englishes, the status of the varieties spoken in the former English plantation settlement colonies of the Caribbean is not clear, owing largely to the issue of whether English creoles should be considered new English dialects or separate languages that may not even be Germanic (Mufwene 2009c). 22. Massive relocation to new places where one cannot continue speaking one’s

[ 34 ]

chapter one

heritage language accounts for the endangerment of Gullah (mistakenly associated with “decreolization” [Mufwene 1997, 2008]). According to Hazaël-Massieux (1999), this would also account for the endangerment of French creoles in French overseas departments, as those who would be speakers migrate to the French metropole, often identified as the “Hexagone.” However, not being fully integrated in the host, Hexagonal society has often led many of these “migrants” to become more loyal to Créole and speak it more fervently as an identity marker or an in-group variety.

References Aboh, Enoch. 2006. The role of the syntax-semantics interface. In L2 acquisition and creole genesis, ed. Claire Lefebvre, Lydia White, and Christine Jourdan, 253–275. Amsterdam: Benjamins. ———. 2009. Competition and selection: That’s all! In Complex processes in new languages, ed. Enoch Aboh and Norval Smith, 317–344. Amsterdam: Benjamins. Allsopp, Richard. 1977. Africanisms in the idioms of Caribbean English. In Language and linguistics problems in Africa, ed. Paul F. Kotey and Haig Der-Houssikian, 429–441. Columbia, SC: Hornbeam Press. Ansaldo, Umberto. 2009. Contact languages: Ecology and evolution in Asia. Cambridge: Cambridge University Press. Baker, Philip. 1993. Assessing the African contribution to French creoles. In Africanisms in Afro-American language varieties, ed. Salikoko S. Mufwene, 123–155. Athens: University of Georgia Press. Casas, Bartolomé de las. 1992a. In defense of the Indians. Trans. and ed. Stafford Poole, C.M., from a 1552 Latin manuscript. DeKalb: Northern Illinois University Press. ———. 1992b. A short account of the destruction of the Indies. Trans. and ed. Nigel Griffin. London: Penguin Books. Chaudenson, Robert. 1979. Les créoles français. Paris: Fernand Nathan. ———. 1992. Des îles, des hommes, des langues: Essais sur la créolisation linguistique et culturelle. Paris: L’Harmattan. ———. 2001. Creolization of language and culture. London: Routledge. ———. Forthcoming. A água e as línguas. In Línguas: Contatos e/ou fronteiras, ed. Pierre Guisan and Thomas Finbow. São Paulo: Editora Parábola. Clements, Clancy. 2009. The linguistic legacy of Spanish and Portuguese: Colonial expansion and language change. Cambridge: Cambridge University Press. Corne, Chris. 1999. From French to Creole: The development of new vernaculars in the French colonial world. London: University of Westminster Press. Crosby, Alfred W. 1986. Ecological imperialism: The biological expansion of Europe, 900–1900. Cambridge: Cambridge University Press. ———. 1992. Ills. In Atlantic American societies: From Columbus through abolition 1492–1888, ed. Alan L. Karras and John Robert McNeill, 19–39. London: Routledge. DeGraff, Michel. 2001. On the origin of creoles: A Cartesian critique of neo-Darwinian linguistics. Linguistic Typology 5:213–310. ———. 2003. Against creole exceptionalism. Discussion note. Language 79:391–410. ———. 2005. Linguists’ most dangerous myth: The fallacy of creole exceptionalism. Language in Society 34:533–591. ———. 2009. Language acquisition in creolization and, thus, language change: Some

Salikoko S. Mufwene

[ 35 ]

Cartesian-uniformitarian boundary conditions. Language and Linguistics Compass 3–4:888–971, www3.interscience.wiley.com/journal/122464861/abstract. Diamond, Jared, and James A. Robinson, eds. 2010. Natural experiments of history. Cambridge, MA: Harvard University Press. Eakin, Marshall C. 2007. The history of Latin America: Collision of cultures. New York: Palgrave Macmillan. Escobar, Anna María. 2012. Spanish in contact with Amerindian languages. In The handbook of Spanish linguistics, ed. J. I. Hualde, I. Olarrea, and E. O’Rourke, 65–88. Malden, MA: Wiley-Blackwell. Farriss, Nancy M. 1984. Maya society under colonial rule. Princeton, NJ: Princeton University Press. Fayer, Joan. 2003. African interpreters in the Atlantic slave trade. Anthropological Linguistics 45:281–295. Florin, José Luiz, and Margarida Petter, eds. 2008. Africa no Brasil: A formação vo da língua portuguesa. São Paulo: Editora Contexto. Freire, José Bessa. 2004. Rio Babel—a história das línguas na Amazônia. Rio de Janeiro: EdUERJ. Gabbert, Wolfgang. 2004. Becoming Maya: Ethnicity and social inequality in Yucatán since 1500. Tucson: University of Arizona Press. Galves, Charlotte, Helder Garmes, and Fernando Rosa Ribeiro, eds. 2009. AfricaBrasil: Caminhos da lingua portuguesa. Campinas, Brazil: Editora da Unicamp. Gilman, Charles. 1986. African areal characteristics: Sprachbund, not substrate? Journal of Pidgin and Creole Languages 1:33–50. Gray, Edward G., and Norman Fiering, eds. 2000. The language encounter in the Americas, 1492–1800. New York: Bergham Books. Hazaël-Massieux, Marie-Christine. 1999. Les créoles: L’indispensable survie. Paris: Editions Ententes. Itier, César. 2000. Lengua general y quechua cuzqueño en los siglos XVI y XVII. In Desde afuera y desde adentro: Ensayos de etnografía e historia del Cuzco y Apurímac, ed. Luis Milliones, Hiroyasu Tomoeda, and Tatsuhiko Fujii, 47–59. Osaka: National Museum of Ethnology. ———. 2001. La propagation de la langue générale dans le sud du Pérou. In Le savoir, pouvoir des élites dans l’empire espagnol d’Amérique, 63–74. Centre de Recherche sur l’Amérique Espagnole Coloniale, Travaux et Documents, no. 3. Paris: Université Sorbonne Nouvelle Paris III. Karttunen, Frances. 1994. Between worlds: Interpreters, guides, and survivors. New Brunswick, NJ: Rutgers University Press. Landa, Manuel de. 2000. A thousand years of nonlinear history. New York: Swerve Editions. Lewis, M. Paul, ed. 2009. Ethnologue: Languages of the world. 16th ed. Dallas, TX: SIL International. Online version, www.ethnologue.com/. Lima, Ivana Stolze, and Laura do Carmo, eds. 2008. Historia social da lingua nacional. Rio de Janeiro: Casa de Rui Barbosa. Lipski, John. 1994. Latin American Spanish. London: Longman. Mal-Molinero, Clare. 2000. The politics of language in the Spanish-speaking world. London: Routledge. McWhorter, John H. 1998. Identifying the creole prototype: Vindicating a typological class. Language 74:788–818.

[ 36 ]

chapter one

Metcalf, Alida C. 2005. Go-betweens and the colonization of Brazil. Austin: University of Texas Press. Moñino, Yves. 2007. Les rôles du substrat dans les créoles et dans les langues secrètes: Le cas du palenquero, créole espagnol de Colombie. In Grammaires créoles et grammaire comparative, ed. K. Gadelii and A. Zribi-Hertz, 49–72. Saint-Denis, France: Presses Universitaires de Vincennes. Mufwene, Salikoko S. 1986. Les langues créoles peuvent-elles être définies sans allusion à leur histoire? Etudes Créoles 9:135–150. ———. 1990. Transfer and the substrate hypothesis in creolistics. Studies in Second Language Acquisition 12:1–23. ———. 1996. The founder principle in creole genesis. Diachronica 13:83–134. ———. 1997. The ecology of Gullah’s survival. American Speech 72:69–83. ———. 2000. Creolization is a social, not a structural, process. In Degrees of restructuring in creole languages, ed. Ingrid Neumann-Holzschuh and Edgar Schneider, 65–84. Amsterdam: John Benjamins. ———. 2001. The ecology of language evolution. Cambridge: Cambridge University Press. ———. 2005. Créoles, écologie sociale, évolution linguistique. Paris: L’Harmattan. ———. 2006. “New-dialect formation” is as gradual as “creole-development”: A response to New-dialect formation: The inevitability of colonial Englishes, by Peter Trudgill. World Englishes 25:177–186. ———. 2008. Language evolution: Contact, competition, and change. London: Continuum Press. ———. 2009a. The indigenization of English in North America. In World Englishes: Problems, properties, prospects: Selected papers from the 13th IAWE Conference, ed. Thomas Hoffmann and Lucia Siebers, 353–368. Amsterdam: Benjamins. ———. 2009b. Kituba, Kileta, or Kikongo: What’s in a name? In Le nom des langues III: Le nom des langues en Afrique sub-saharienne: Pratiques dénominations, catégorisations, ed. Carole de Féral, 211–222. Louvain-la-Neuve, Belgium: Peeters, BCILL 124. ———. 2009c. Some offspring of colonial English are creole. In Vernacular universals vs. contact-induced language change, ed. Juhani Klemola, Markku Filppula, and Heli Paulasto, 280–303. London: Routledge. ———. forthcoming. Globalisation économique mondiale des XVIIe–XVIIIe siècles, émergence des créoles, et vitalité langagière. In Proceedings of the colloquium of the Comité International des Etudes Créoles, Mauritius 2012, ed. Arnaud Carpooran. Mühlhäusler, Peter. 2003. Language of environment, environment of language: A course in ecolinguistics. London: Battlebridge. Muysken, Peter. 1981. Halfway between Quichua and Spanish: The case for relexification. In Historicity and variation in creole studies, ed. Arnold Highfield and Albert Valdman, 52–78. Ann Arbor, MI: Karoma. Naro, Anthony Julius. 1978. A study on the origins of pidginization. Language 54: 314–47. ———. 1988. A reply to “Pidgin origins reconsidered,” by Morris Goodman. Journal of Pidgin and Creole Languages 3:95–102. Naro, Anthony Julius, and Maria Marta Pereira Scherre. 2007. Origens do português brasileiro. São Paulo: Parábola Editorial. Negrão, Esmeralda, and Evani Viotti. 2011. Epistemological aspects of the study of the participation of African languages in Brazilian Portuguese. In Portugais et

Salikoko S. Mufwene

[ 37 ]

langues africaines: Etudes afro-brésiliennes, ed. M. Petter and M. Vanhove. Paris: Karthala. ———. 2012. Em busca de uma história linguística. Revista Estudos da Linguagem 20:309–342. Nettle, Daniel, and Suzanne Romaine. 2000. Vanishing voices: The extinction of the world’s languages. Oxford: Oxford University Press. Ostler, Nicholas. 2005. Empires of the word: A language history of the world. New York: Harper Collins. Quijano, Aníbal. 2007. Coloniality and modernity/rationality. Cultural Studies 21, nos. 2–3: 168–178. Quijano, Aníbal, and Michael Ennis. 2000. Coloniality of power, Eurocentrism, and Latin America. Views from South 1:533–580. Roncardi, Claudia, and Jussara Abraçardo, eds. 2008. Português brasileiro. Vol. 2, Contacto linguistico, heterogenidade e hist∴ria. Rio de Janeiro, Brazil: Editora da Universidade Federal Fluminense. Schwartz, Stuart B. 1985. Sugar plantations in the formation of Brazilian society: Bahia, 1550–1835. Cambridge: Cambridge University Press. Schwegler, Armin. 2006. Bozal Spanish: Captivating new evidence from a contemporary source (Afro-Cuban “Palo Monte”). In Studies in contact linguistics: Essays in honor of Glenn G. Gilbert, ed. Linda L. Thornburg and Janet Fuller, 71–101. New York: Peter Lang. Shappeck, Marco. 2011. Quichua-Spanish language contact in Salcedo, Ecuador: Revisiting Media Lengua syncretic language practices. PhD diss., University of Illinois–Urbana-Champaign. Skutnabb-Kangas, Tove. 2000. Linguistic genocide in education—or world-wide diversity and human rights? Mahwah, NJ: Lawrence Erlbaum. Trudgill, Peter. 2004. New-dialect formation: The inevitability of colonial Englishes. Oxford: Oxford University Press.

2 * The Many Facets of Spanish Dialect Diversification in Latin America john m. lipski

1. Introduction: The Sources of Spanish Dialect Differentiation Five hundred years ago, a rather homogeneous variety of Spanish spoken by a few thousand settlers was scattered across two continents. Although many regional languages were spoken in fifteenth-century Spain (and most are still spoken even today), only Castilian took root in the Americas, in itself a remarkable development. More remarkable still is the regional and social variation that characterizes Latin American Spanish today; some of the differences among Latin American Spanish dialects are reflected in dialect divisions in contemporary Spain, while others are unprecedented across the Atlantic. In accounting for dialect diversification in Latin American Spanish, three main factors come into play, in addition to the inevitable linguistic drift. The first is the Peninsular roots of Latin American Spanish, meaning the varieties spoken by Spanish settlers from all over peninsular and insular Spain over a period of more than four centuries. The second is contact with other languages, these being principally the indigenous languages of the Americas spoken in the major Spanish colonies, but also African languages spoken by hundreds of thousands of slaves brought to the colonies and, to a lesser extent, languages of voluntary immigration in later centuries, mainly Italian, English, Cantonese Chinese, and Afro-European creole languages of the Caribbean, such as Haitian Creole, Jamaican Creole, and Papiamentu. The third factor is the catalytic effect that emerging cities in Spanish America exerted on regional varieties of Spanish, which ultimately spread far beyond the pale of the cities to become regional, national, and transnational standards. All three factors had their impact at one point or another, but central to all three themes is the question of how much linguistic influence a given group of individuals exerted on the Spanish language at particular times. That is, how many speakers of one language or dialect are needed to leave a permanent imprint on the evolving Spanish American varieties? Is the lemma “first is best” the appropriate slogan, or is “safety in numbers” (or, in the

John M. Lipski

[ 39 ]

case of involuntary servitude, “misery loves company”) a more fitting label? Like the questions asked by journalists and detectives, the “who,” “where,” “why,” and “when” must be determined in order to account for the “what” of language diversification.

2. The Dichotomy “Demographic Strength versus Chronological Primacy” In searching for the roots of Latin American Spanish dialectal variation, proposals have grouped around two opposing viewpoints, as regards the relative importance of demographic strength versus chronological primacy (reviewed in Lipski 2002). The first proposal is that uniquely defining characteristics of a given dialect are directly correlated with the demographic proportions of groups—whether they are speakers of other varieties of Spanish or of other languages—assumed to have contributed the features in question.1 The opposing postulate holds that the first settlers—the “founders”—exercise a permanent influence on the subsequent development of the dialect in a fashion far out of proportion to their demographic strength, an impact that continues past the time when descendants of the original founders enjoyed any special prominence.

3. The “Founder Principle” and the “Antillean Period” Of the theories seeking to establish the roots of Latin American Spanish in the speech of the earliest settlers, the most influential is the “Antillean period” from 1493 to 1519 (e.g., Boyd-Bowman 1956; Catalán 1958; Guitarte 1980; Rosenblat 1977, 20; also Lockhart and Schwartz 1983, chap. 3). During this period Spain consolidated its settlements on Hispaniola and Cuba and launched expeditions to Central and South America. Santo Domingo was the point of departure for the first expeditions to Puerto Rico, Cuba, Trinidad, Jamaica, Darién, the Caribbean coasts of Venezuela and Colombia, and Mexico. According to one line of thought, the Andalusian influence became decisive during the early decades of the sixteenth century. As a result of the Spanish government’s procedures for emigration, future settlers from all regions of Spain submitted their applications for passage at the Casa de la Contratación in Seville (at the time the major city in Andalusia and the second-largest city in Spain); they often waited a year or more before embarking for Spanish America. This waiting period provided the first occcasion for dialect leveling and the incipient formation of a supraregional koiné, presumably dominated by Andalusian traits. The long passage to the Americas provided addi-

[ 40 ]

chapter two

tional opportunities for the assimilation of Andalusian linguistic features, since ships’ crews were often recruited from Andalusia and the Canary Islands. At a time when Spanish settlements in the New World were entirely sustained by maritime contact with Europe, successive arrivals who participated in explorations and settlements of the mainland were, it is claimed, immersed in the prevailing speech patterns of the American insular settlements and in turn carried this form of speech to colonies established on the mainland. Although Spanish trade with mainland colonies soon bypassed the Antilles, except for purposes of loading supplies, the seeds of “Andalusian-American” Spanish would have been sown. Boyd-Bowman’s “Antillean period” theory is an instantiation of Mufwene’s “Founder Principle,” a hypothesis applied to the origin and development of creole languages, in which it is claimed that “structural features of creoles have been predetermined to a large extent . . . by characteristics of the vernaculars spoken by the populations which founded the colonies in which they developed. European colonies often started with large proportions of indentured servants and other low-class employees of colonial companies, thus by speakers of nonstandard varieties of the creoles’ lexifiers” (Mufwene 1996, 84). Unlike Boyd-Bowman’s theory for the emergence of (Antillean) Latin American Spanish, the Founder Principle does not ascribe any special prestige to the creators of a creole language; indeed, they often represent the lowest social classes and marginalized groups, whose very marginality in a colonial setting gives precedence to their erstwhile nonprestigious speech forms, propelling them into a new linguistic norm. Both approaches coincide in attributing a significant proportion of major traits of a new language or dialect cluster to the earliest speakers, transplanted from a metropolis or from peripheral zones to a location where their languages and dialectal traits come together for the first time.

4. The Chronology of Spanish Evolution It is often stated that Latin American Spanish is “Andalusian” in character, as opposed to “Castilian,” but when comparisons are made with the contemporary dialects of Spain, only the Spanish dialects of the Caribbean basin truly sound “Andalusian” in the modern sense, while highland dialects, such as of central Mexico, Colombia, Peru, and Bolivia, in many ways resemble “Castilian” Spanish. Spanish continued to evolve in Latin America whether or not in contact with European innovations. All dialects of Latin American Spanish acquired most of the major linguistic innovations that occurred in Spain at least through the end of the seven-

John M. Lipski

[ 41 ]

teenth century, and some more recent Peninsular phenomena were also transferred to Latin America. Among the pan-Hispanic changes occurring well past the first century of Spanish-American colonization are the following. (1) In 1492, Spanish contained six sibilants: /s/ (ss), /z/ (s), /ts/ (ç), z /d / (z), /ʃ/ (x), and /ʒ/ (g/j). /s/ and /z/ were apicoalveolar, like contemporary Castilian /s/. There is some indication that merger of the alveolar fricatives and the affricates was already beginning in Andalusia by the end of the fifteenth century, but the change was not complete (Catalán 1956–57).2 In no Spanish dialect had devoicing of the voiced sibilants even begun. In Latin America, early Spanish borrowings into Nahuatl, Quechua, and Guaraní verify that Spanish colonists still maintained the contrast in voicing. Within Spain, devoicing of /z/ and /dz/ was complete by the end of the sixteenth century (Catalán 1957), even in Andalusia. If Latin American Spanish had received an Andalusian imprint during the “Antillean period,” we should expect a voicing distinction between /s/ and /z/ to have remained indefinitely, as it has in Sephardic (Judeo) Spanish, which was delinked from other Peninsular varieties at the beginning of the sixteenth century. Instead, Latin American Spanish kept pace with both Castile and Andalusia in devoicing all sibilants, at approximately the same time as this was occurring in Spain. In the New World and in western Andalusia and the Canary Islands, all the anterior sibilants fell together to /s/. In the remainder of Spain, the reflex of /ts/-/dz/ became an interdental fricative /θ/. (2) As another part of the general devoicing process, Spanish /ʃ/ and /ʒ / merged to a voiceless fricative, which later velarized to /x/, and the change was complete by the middle of the seventeenth century (Lapesa 1981, 379). Judeo-Spanish still retains the phonemes /ʃ/ and /ʒ/ and has no velar fricative /x/. Early borrowings into Native American languages give proof that /ʃ/ was still a prepalatal fricative during the first century of Spanish settlement in the New World (and the word Chicano, from the old pronunciation of mexicano, bears witness to this early colonial sound), but it too followed the dialects of Spain. (3) /b/ and /v/ were still separate phonemes in Spain during the “Antillean period” of Latin American settlement. Spanish words taken into Native American languages during the sixteenth century reflect this difference. /b/ and /v/ subsequently merged in all Peninsular and Latin American dialects. (4) At the time of the first Spanish settlements in the Americas, the formal second person pronoun usted (s.) < vuestra merced ‘your mercy’ and the analogical ustedes (pl.) had not yet emerged (and neither is found

[ 42 ]

chapter two

in Judeo-Spanish). In Spain, these pronouns did not come into general use until the end of the seventeenth century; Latin American Spanish acquired the pronouns at the same time. The preceding survey amply demonstrates that early-sixteenth-century Spanish of the “Antillean period,” or even the Spanish brought to colonies founded throughout the seventeenth century, is vastly different from all modern varieties of Spanish, in Spain and Latin America; only Sephardic Spanish (representing a quasi-fossilized early-sixteenth-century Spanish with some noticeably Andalusian traits) might be an approximation of what Caribbean Spanish might actually be like if the Founder Principle and the “Antillean period” model were viable hypotheses for the formation of modern Latin American Spanish dialects. Despite the vast cultural differences separating the Sephardim in exodus and Spanish settlers in the American colonies—not the least of which is that most Sephardic varieties were cut off from other dialects of Spanish—the retention of the “founders’” traits for the better part of five centuries can serve as a demonstration of what the diachronic results of the Founder Principle might look like. Models of dialect formation that limit the formative period to the first half century or even full century of colonial settlement are unrealistic; linguistic cross-fertilization between Spain and Latin America extended over several centuries. In any nation arising from colonization, the speech and cultural patterns of the first settlers retains a nostalgic significance that transcends any objective contribution this group might have made. In reconstructing the true history of a nation, colonial heroes assume larger-than-life proportions, and the spirit of the original colonists is seen embodied in the current population. These sentimental issues rarely hold up under serious linguistic scrutiny, and in truth Latin American Spanish is the product not only of its first settlers but of the totality of the population, immigrants and natives alike. If the crucial defining traits of contemporary Latin American Spanish were not forged during the early sixteenth century, then attention must be shifted to later events, from the late sixteenth century to the first decades of the twentieth century. In shaping the eventual form of Latin American Spanish dialects, language contact and the emergence of urban speech communities played decisive roles; these influences will be treated in turn.

5. In Search of Alternative Models: The Role of the City For at least two centuries, Spanish settlement of the New World was planned in Castile, engineered in Andalusia, and aided by the Canary

John M. Lipski

[ 43 ]

Islands. Administrative matters involving the American colonies were handled by the Consejo de Indias, eventually moved from Castile to Seville. Future settlers made application for passage at the Casa de la Contratación in Seville and often waited a year or more before embarking for Spanish America. The Consulado de Sevilla, dominated by Seville merchants, long enjoyed a monopoly on trade with the Americas. Ships’ crews were recruited from Andalusia and the Canary Islands. Many ships left directly from Seville; others departed from the Andalusian ports of Cádiz, San Lúcar, and Huelva. Prevailing winds and sea currents, as well as partially fortuitous Spanish colonizing patterns, shaped preferential routes into and out of the Caribbean. Ships arriving from Spain entered the southern Caribbean, often stopping at Jamaica or another eastern island, and docked at Cartagena de Indias, which became the major South American port and trade zone. Ships carrying goods and passengers bound for the Pacific coast of South America put in at Portobelo, Panama, whence cargo was transferred to Panama City on the Pacific side by a combination of mule trains and river boats. Guayaquil, El Callao, and later Valparaiso were the major Pacific ports, and once Spain began sending galleons to the Philippines, Acapulco was added to the list. On the Caribbean coast of Mesoamerica, Veracruz was the main point of entry. Ships returning to Spain from Portobelo usually put in again at Cartagena, then headed for the northern Caribbean. Havana became the foremost port of supply for returning ships, while other Caribbean towns such as Santo Domingo, the first Spanish city in the Americas, quickly lost their early importance. Except for a few of the earliest towns, such as Nombre de Dios and Portobelo, Panama, which were soon abandoned in the Spanish colonial scheme, the hubs of Spanish colonial society have evolved into large urban masses, each with several million inhabitants. Each city is a complex sociolinguistic microcosm, and it is difficult to imagine how any external linguistic force could have a significant impact on the thriving Spanish dialects. The notion that the idiosyncrasies of literally a handful of people, no matter how rich or powerful, could permanently transform the speech of an entire city, region, or nation lies beyond belief. Aside from the internal dynamics of large urban areas, the only major linguistic shifts occurring in modern Latin America result from rural migration to the cities. Matters were not always as they are today; the explosive demographic growth that has turned former colonial centers into impersonal urban sprawls has occurred within the past century or less. During the time when the foundations for Latin American dialects were laid, the major cities and towns were a tiny fraction of their present size, and models of

[ 44 ]

chapter two

language change unthinkable today were viable options in past centuries. Moreover, the population did not always increase across time; the Spanish colonies were afflicted with epidemics and plagues that sometimes reduced the population of a given area by half or more. As a result, some cities experienced no net growth over a period as long as two centuries. If the “Antillean period” prior to 1530 is considered crucial, then only a handful of island villages with a total population of a few thousand colonists are at stake. If the entire sixteenth century is taken into account, few cities in Spanish America achieved a population of five thousand or more inhabitants. Some of today’s major population centers, embodying national dialects, had not yet been founded. When one considers that a typical fleet arriving at Cartagena, Portobelo, or Lima might bring several hundred settlers, the possible linguistic effects of a contingent of new settlers on an evolving dialect could be considerable. A single fleet could, under some circumstances, bring new arrivals who amounted to nearly half the resident population, and even if not all new settlers remained in the port of entry, their linguistic contributions would not be inconsequential.

6. The Emerging Critical Mass of Spanish American Cities Until at least the middle of the eighteenth century, the principal cities of Spanish America were small and relatively isolated, and they contained speech patterns that could easily be influenced by rather small numbers of incoming settlers and immigrants. By comparing linguistic innovations occurring in Spain since the early sixteenth century with emerging traits of Latin American Spanish, it is possible to identify with some accuracy the period in which Latin American dialects ceased to reflect major innovations occurring in Spain; essentially, by the 1700s most innovations in Spain did not pass unconditionally to Latin America. At the same time the first quintessentially Latin American innovations emerged as distinctive dialectal features. A comparison of the time line (figure 2.1) of changes in Spain and Latin America with the demographic patterns of Spanish American urban zones—ports and capital cities—reveals that once cities reached a critical mass of several tens of thousands (which usually occurred during the late eighteenth or early nineteenth century), these speech communities effectively resisted full incorporation of language changes that occurred in Spain and arrived with new settlers. There are, at the same time, instances in which growing urbanization in colonial Spanish America is directly correlated with linguistic innovations. Most noteworthy is the žeísmo or groove fricative pronunciation

Figure 2.1. Time line for changes affecting Castilian, Andalusian, and Latin American Spanish

[ 46 ]

chapter two

[ʒ ] of /j/ typical of the Río de la Plata area (Buenos Aires and Montevideo), which, according to Fontanella de Weinberg (1987), appears to have emerged in the early to mid nineteenth century, a time when Buenos Aires took the first of many enormous demographic leaps, to become one of the largest cities of the Americas. The demographic and economic strength of Buenos Aires, to which Montevideo can be added, consolidated this feature—which occurs at the sociolinguistic margins of other Spanish-speaking areas such as Seville and parts of Mexico—into a prestigious mainstream trait. Subsequently the devoicing of the same sound to [ʃ] evidently originated in the late nineteenth or early twentieth century in Buenos Aires, whence it slowly spread to other areas of Argentina and to Montevideo. Mexico City experienced a similar growth spurt during the same time period, and it is likely that the characteristic central Mexican realization of phrase-final /r/ as a sibilant [ɹ] arose at this point. The highly fronted posterior fricative /x/ realized as palatal [ç] in the Spanish of Santiago de Chile, now affecting most of the country and most noteworthy before front vowels as in gente ‘people’ and ajeno ‘foreign’, also appears to correlate with the demographic leap of Santiago during the nineteenth century, as does the highly fronted realization of /tʃ/ as in Chile.

7. Language Contact: Indigenous Languages Language contact phenomena are, beyond any reasonable doubt, the most important factors responsible for the diversification of Spanish across the entire American continent. In chronological order—and probably also in order of overall impact—these involve contact with indigenous languages, with languages of involuntary immigration (enslaved Africans), and with languages of voluntary immigration, mostly from Europe. Beginning with the first category, aside from indigenous lexical items and toponyms, there is no consensus on the effects of Native American languages on Spanish. The Spanish of Latin America is widely varied, including features not attested in Spain. In pronunciation and syntax, many Latin American dialects present systematic innovations that are not easy to explain away as linguistic drift, the inheritance of Spanish settlers, or borrowing from neighboring dialects. Particularly in areas where the indigenous population has remained demographically and ethnically prominent, it is not unreasonable to suppose that some unique features of regional Spanish dialects are attributable to prolonged contact with indigenous languages. Few claims of indigenous influence have been accompanied by a demonstration of the purported substrate patterns, nor

John M. Lipski

[ 47 ]

of the opportunity for second-language or ethnically marked varieties to percolate upward into regional dialects of Spanish. Too often, the mere demographic presence of a large indigenous or mestizo population has uncritically been taken as the source of “peculiarities” of a given dialect zone, without verifying either the viability of the hypothesis in linguistic and historical terms or the existence of alternative explanations. The case for an indigenous influence on nonlexical features of Latin American Spanish must be presented as in a court of law, demonstrating motive, method, and opportunity. During the sixteenth century and even later, indigenous populations often outnumbered Europeans by a factor of several thousand to one, and yet the nature of Spanish settlement was not always conducive to substratum influences. For an indigenous language to permanently influence colonial Spanish required a special set of conditions, which were not present in all colonies or at all times. During the early colonial period, the entry of indigenous people into Spanish cities was carefully controlled, and most Native Americans could not reside or even spend the night inside the cities. Gradually a class of bilingual intermediaries arose, to supply the goods and services required by the Spanish colonists; the Spanish interlanguages as spoken by indigenous bilinguals did not begin to have a permanent effect on regional Spanish dialects until those speakers became absorbed into the urban setting. Native Americans who use Spanish only occasionally, having learned it as a second language past childhood, speak varieties of Spanish (with considerable idiolectal variation) in which the phonology, morphology, and syntax of the native language are superimposed on Spanish patterns. Today such speech can be heard in indigenous redoubts throughout the Amazon Basin, the Andes, and Mesoamerica; in the past, it existed in nearly every Spanish colonial settlement. Even when Spanish is used on a daily basis, between workers and employers or between rural residents and priests, fluency may never rise above the level of a rough pidgin. Such indigenous-flavored Spanish has no ready way of expanding beyond the group that has created it, and ordinarily it leaves no traces on natively spoken Spanish. For a partially restructured form of Spanish to permanently penetrate regional varieties of Spanish, a major sociolinguistic shift must break the equilibrium that sustains the contact-influenced variety and confines it to second-language learners. Speakers of the indigenousinfluenced varieties need to occupy positions in which their speech becomes the norm. Such speakers must be present in great enough numbers to make their speech patterns demographically prominent. The partially restructured variety, by definition the result of having learned Spanish

[ 48 ]

chapter two

as a second language, must gradually become a first language, without shedding the indigenous accretions. This requires insulation from normative standards or a social environment in which such standards are no longer relevant. In a racially and socially segregated environment such as existed in colonial Latin America, Spanish is used not only for essential contacts with the population of European descent but also among members of the same indigenous community. Mestizos provide a bridge between the cultures and facilitate language transfer and the development of a stable ethnically marked variety of the colonial language. To exemplify potential consequences of contact with indigenous languages, I will mention four cases, all from the Andean linguistic zone. The first is found only among bilingual speakers for whom Spanish is nondominant, while the other three phenomena embrace a much wider cross-section of Andean Spanish.

8. An Example of Contact-Influenced Spanish: Double Possessives In her analysis of Andean Spanish, Escobar (1994) describes the distinction between contact phenomena found only among bilingual indigenous speakers and those constructions that are found throughout the Andean region even among monolingual Spanish speakers. Bilingual Spanish combinations are nearly always highly stigmatized, and they connote lack of formal education and imperfect acquisition of Spanish. One of the prime shibboleths is the use of double possessive constructions involving both the preposition de ‘of’ and the third person possessive determiner su, especially with the possessor coming before the possessed object: De Juan su mamá ‘John’s mother,’ del perro su rabo ‘the dog’s tail’. Only slightly more acceptable are double possessives with the opposite word order: su marido de Juana ‘Juana’s husband’. These constructions are clear calques of Quechua and Aymara and are readily produced and understood by bilingual speakers of these languages. A representative Quechua construction is Mariya-x

María -poss

wasi -n

house-poss

‘Mary’s house’

The monolingual Spanish speaker lacking any knowledge of Quechua structures, one whose grammar includes only the combination la casa de María, will be at a disadvantage in terms of rapid interpretation. This is not unlike what occurs when the bilingual Spanish-English speaker in the United States says te llamo para atrás ‘I’ll call you back’ instead of the

John M. Lipski

[ 49 ]

Spanish-only te vuelvo a llamar, creating an equivalent for the English postverbal particle.

9. Contact-Induced Phenomena in Monolingual Andean Spanish Varieties Given the heavy social stigma of anything that smacks of indigenous culture, such transparent calques are among the first linguistic elements to be shed en route to escape the dreaded classification of cholo. Many more subtle contact-induced phenomena have penetrated virtually all monolingual sociolects of Andean Spanish and are responsible for giving this dialect zone its unique characteristics. Three are worth mentioning as exemplars of the permanent imprint of indigenous languages on Latin American Spanish: clitic doubling, crypto-evidentials, and pitch accents.

clitic doubling Andean Spanish permits, and for large numbers of speakers actually requires, clitic doubling of inanimate [+definite] direct objects, a feature not found in other varieties of Spanish, where direct object clitics can combine only with animate direct object NPs or pronominals. At the most vernacular level, the Andean doubled clitic is invariant lo, that is, without the usual inflection for gender and number. Some examples follow. Peru:

Le pedí que me lo calentara la plancha (Pozzi-Escot 1972, 130) I asked her to heat up the iron for me Se lo llevó una caja (Luján 1987, 115) She took a box

Northwestern Argentina (Gómez López de Terán and Assís 1977; Rojas 1980, 83):

¿Me lo va a firmar la libreta?

Will you sign the book for me? Bolivia:

Tú lo tienes la dirección (Stratford 1989, 119) You have the address

Cerralo la puerta (Justiano de la Rocha 1986, 29) Close the door

Mientras tanto, vémelo el asado (Mendoza 1991) Meanwhile, watch the roast for me

[ 50 ]

chapter two

Ecuador (Suñer and Yépez 1988): Le veo el carro I see the car

Even a cursory glance at Quechua and Aymara grammar suffices to demonstrate that direct object clitics of the sort used in Spanish do not occur; moreover, the usual object-verb word order precludes the canonical linear combinations found in Spanish. At the same time, the facts that clitic doubling occurs only in dialect zones characterized by extended language contact, and that among Spanish-recessive bilinguals the invariant clitic lo functions more as a transitivity marker than as a true object clitic, motivate the search for subtle contact-induced transfer. Quechua marks direct object nouns with the suffix -ta (or -man if following a verb of motion). This suffix is invariable, cliticizes to all direct object nouns whether definite or indefinite, and even attaches to questions and relative clauses, as shown by the following (Peruvian) examples (an approximation in “Andean” Spanish is given in parentheses). T’ika-ta

kuchu-ni

Flower-acc

cut-1sg = ‘I cut the flower’ (lo corto la flor)

ima-ta

kuchi-ni?

Challwa-ta

apa –nki

Asta-ni

unu –ta

What-acc

Fish-acc

Carry-1sg

cut-1sg = ‘What do I cut?’ (¿qué lo corto?)

carry-2sg (fut) = ‘You will carry fish’ (lo llevarás pescado)

water-acc = ‘I carry water’ (lo acarreo agua)

The accusative marker -ta does not occupy exactly the same syntactic position as the invariable lo of the corresponding Andean Spanish sentences, which would be roughly as indicated above. However, it would be easy for a speaker of a second-language variety of Spanish to interpret the clitic lo, statistically the most frequent among the many Spanish object clitics, as some sort of direct object marker comparable to Quechua -ta. Although in Quechua this element is always attached to the direct object noun, in a canonical Quechua SOV transitive sentence where the direct object immediately precedes the verb, -ta coincidentally comes just before the verb, that is, in the identical position as the Spanish proclitic lo.3 It is not irrelevant that Spanish lo itself marks an accusative relationship, albeit not in the fashion of Quechua -ta. A speaker of the emergent

John M. Lipski

[ 51 ]

indigenous-influenced variety, encountering preverbal lo only in clearly transitive sentences (including the possibility of clitic doubling with human DOs, as in the Southern Cone), would be all the more likely to overgeneralize lo for all transitive clauses. Since the quintessential Quechuainfluenced varieties of Spanish maintain an O-V word order, Spanish lo would at first be misanalyzed as a case marker attached to the noun, in a direct calque of Quechua -ta: el poncho-lo tengo. As second-language speakers develop greater fluency in Spanish, word order gravitates to the more usual V-O for nonclitic DOs. It is at this stage that lo, now implicitly recognized as an object clitic, remains behind in proclitic position, yielding the stable Andean Spanish doubled-clitic pattern. This pattern of events is admittedly speculative, but it does correlate well with observations on the development of Spanish proficiency among Quechua speakers (see also Muysken 1984).4

crypto-evidentials in bolivian spanish Both Quechua and Aymara have morphological evidential markers, which indicate whether the information conveyed by a speaker is based on first- or secondhand knowledge. Spanish has no similar grammatical construction; paraphrases such as dizque ‘it’s said that’, tengo entendido que ‘I’ve heard that’, and so forth, are normally used. In Bolivian Spanish and marginally, perhaps, in other Andean dialects, the Spanish pluperfect indicative has lost its usual meaning of past with respect to another past point and has acquired the meaning of secondhand reporting. This allows for contrasts of evidentiality, particularly with respect to the past. The speaker who says llegaste a las ocho ‘you arrived at 8:00’, using the preterite tense, is indicating personal knowledge of the time of arrival, whereas habías llegado a las ocho, with the pluperfect indicative, reflects secondhand knowledge only. The pluperfect indicative can also be used epistemically to express the result of a deduction and as a reaction of surprise at learning a previously unsuspected fact. Thus, upon hearing a friend speaking French for the first time, one might exclaim habías aprendido francés, roughly ‘I didn’t know that you knew French.’ A Bolivian herbal healer who was surprised at my awareness of local customs said habías vivido en Bolivia ‘you must have lived in Bolivia (previously)’. Unlike the syntactic calques and clitic doubling, for which a plausible template for language transfer can be postulated, the evidential use of the pluperfect indicative must be approached circumstantially: it occurs only in contact with Aymara and occasionally Quechua, languages that have morphological markers of evidentiality, and is not found anywhere else in the Spanish-speaking world. The precise means by which the Spanish

[ 52 ]

chapter two

pluperfect indicative developed the nuance of secondhand reporting in the Andean region is not known. The temporal relations expressed by the Romance pluperfect do not map directly onto Quechua and Aymara patterns; in fact, the entire range of Spanish compound verbs differ typologically from structures found in Andean languages. It may be that indigenous speakers acquiring Spanish initially analyzed compound verbs as involving some kind of particle. The Spanish present perfect, for example, has been generalized to take over most of the functions of the simple preterite, for example, in situations excluding the present moment. In this sense the Andean Spanish present perfect is similar to the French passé composé and the standard Italian present perfect, representing a more advanced evolution than is found in Peninsular Spanish dialects. In the Andean region, one can say nos hemos conocido el año pasado ‘we met for the first time last year’, a combination that would not be acceptable in the remainder of Latin America. The notion of grammatical encoding of evidentiality is presumably compelling to speakers of Aymara and Quechua, as also attested by the expanded use of dizque ‘it’s said that’, diciendo ‘saying’, and other Spanish markers in Andean Spanish (Laprade 1976, 1981; Stratford 1989, 1991).

early high peak alignment in andean spanish Phonetic and phonological influences of Native American languages on Spanish have been postulated for many speech communities, and in contemporary second-language and ethnically marked varieties clear cases of transfer can still be heard. As for monolingual Spanish dialects resulting from previous contacts between Spanish and indigenous languages, the evidence is less clear, and unsubstantiated claims abound. The most convincing cases can be made in those regions where bilingual speakers with varying degrees of proficiency in Spanish can still be found, alongside monolingual Spanish speakers. One promising area of research involves intonational patterns. Spanish intonational contours vary widely across dialect regions and according to pragmatic factors, such as focus, level of politeness, and so on. There are some rather robust common denominators as regards the pitch accents that accompany stressed vowels. Spanish signals word-level stress through a combination of phonetic features, which include lengthening, a greater intensity across a broad spectral range (“spectral tilt”), and especially the use of a rising tone correlated with the accented syllable. Intuitively one might suppose that the high tone coincides with the accented syllable, but this is usually the case only in the nuclear accent, that is, the final accent of the phrase. Across a wide range of dialects, in prenuclear accented syllables the high tone occurs

John M. Lipski

[ 53 ]

toward the end of the tonic syllable or on the immediately following syllable; this is late-peak alignment, and it is typical of nonfocused constituents (Beckman et al. 2002; Face 2000; Prieto et al. 1995; Sosa 1999). At the same time, there is a downdrift of high tones, so that the highest pitch accent is usually the first in the intonational phrase and each successive pitch accent is lower than the preceding ones. In one of the first studies of Spanish intonational patterns in bilingual environments, O’Rourke (2005) has demonstrated that in Peru, the Lima variety is characterized by the more typical late alignment of high tones with respect to prenuclear stressed syllables. This is an area where Quechua influence was historically minimal, although currently there are many Quechua-speaking migrants from the highlands. In Cuzco, the seat of the ancient Inca Empire and an area where Quechua still maintains vitality, intonational patterns are quite different, with a significant number of instances in which high tones coincide with prenuclear stressed syllables. Significantly, the latter pattern typifies regional Quechua. These patterns occur even among university-educated monolingual Spanish speakers, those least likely to be affected by neighboring Quechua-influenced speech varieties. The Cuzco data, in conjunction with data from dialects of Spanish in contact with northern Basque pitch accents (Elordieta 2003), as well as with data from other bilingual communities, suggest that prolonged bilingualism can alter Spanish pitch accents in subtle ways, not necessarily by directly copying patterns of indigenous languages, but rather through the creation of hybrid configurations that expand the monolingual Spanish possibilities.

10. Contact with Languages of Voluntary Immigration: Italians in the Río de la Plata Among the many languages other than Spanish carried by voluntary immigrants to Spanish America, few produced lasting imprints on Spanish, largely owing to the relatively small numbers of speakers involved in comparison with the already established Spanish dialect zones and the rapidity with which subsequent generations were integrated into the local speech communities, acquired Spanish, and lost their ancestral languages. In some cases, the typological distance between the immigrant languages and Spanish also represented an impediment to the retention of contact-induced variants (although very intense contact can ultimately override typological mismatches in the direction of quasi-universal tendencies of language change [e.g., Heine and Kuteva 2005]). A significant exception to this trend is the case of Italian immigration to Buenos Aires and Montevideo, a massive demographic displacement whose linguistic effects are readily apparent.

[ 54 ]

chapter two

To give an idea of the magnitude of this immigration, nearly 2.3 million Italians emigrated to Argentina alone between 1861 and 1920, and more than half of them arrived after 1900; they constituted nearly 60 percent of all immigration to Argentina. Most of the immigrants ended up in greater Buenos Aires (Bailey 1999, 54) and accounted for between 20 percent and 30 percent of that city’s population. As a result of immigration, largely by Italians, the population of greater Buenos Aires (including the surrounding countryside) grew from 400,000 in 1854 to 526,500 in 1881 and 921,000 in 1895 (Nascimbene 1988, 11). Similar proportions, scaled down to size, characterize Montevideo for the same time period. Italian immigrants were not speakers of standard Italian, which resulted from language-planning efforts that had not yet begun in the late nineteenth century; they spoke regional dialects and languages, mostly from southern Italy, and among the immigrants some dialect leveling inevitably took place, as it does in Italy. Given the partially cognate status of Spanish and Italian, contact-induced varieties developed that freely combined both Spanish and Italian elements, as well as many innovations based on analogy and language transfer. It may well have been the possibility for achieving meaningful communication with Spanish speakers by making only relatively small departures from their native Italian dialects that resulted in long-lasting acquisitional plateaus among Italian immigrants in Buenos Aires and Montevideo. The impact of Italian dialects on the Argentinian Spanish lexicon, beginning with the underworld slang known as lunfardo and passing into everyday usage, is undisputed. Two other less easily traceable features are also worth considering, one segmental and the other suprasegmental.

the “long fall” pitch accent of buenos aires and montevideo In the area of pronunciation, while claims of Italian-like prosody are frequently aired, only recently has empirical research been brought to bear on this topic. In particular, the notably rising+falling pitch accent on final stressed syllables in Buenos Aires and Montevideo Spanish—and now extending to provincial varieties in both countries—is impressionistically similar to stereotypical Italian patterns. Kaisse (2001) describes the quintessential Argentinian “long fall,” in which the stressed syllable is significantly lengthened and the tone drops sharply across the elongated vowel. This distinctive pattern is combined with early peak alignment of high tones on prenuclear stressed syllables, similar to that found in Andean Spanish (O’Rourke 2005). Colantoni and Gurlekian (2004) provide a more detailed acoustic analysis of Buenos Aires pitch accents and

John M. Lipski

[ 55 ]

combine these results with a sociohistorical overview of the Italian presence in Buenos Aires beginning in the late nineteenth century. According to Argentinian observers from the time periods in question, the typical porteño (Buenos Aires and Montevideo) intonation pattern did not exist prior to the late nineteenth century, which coincides chronologically with the enormous surge in Italian immigration. At the same time studies of Italian intonation patterns (e.g., D’Imperio 2002 and references therein) confirm patterns congruent to those of modern Buenos Aires Spanish. The circumstantial evidence thus strongly points to an Italian contribution to Buenos Aires–Montevideo intonation, not as a simple transfer but, as in the case of Andean Spanish, via the creation of innovative hybrid patterns that could not easily be extrapolated in the absence of a sustained language contact environment.

loss of word-final /s/ in porteño spanish The other area in which the Italian-Spanish interface may be implicated in Buenos Aires–Montevideo Spanish is the realization of word-final /s/. Dialects of Spanish represent a cline of pronunciation patterns, ranging from the full sibilant realization of syllable- and word-final /s/ (the etymologically “correct” pronunciation) to nearly complete elimination of all postnuclear /s/. The intermediate stages, which represent the majority of the Spanish-speaking world, involve some kind of reduced pronunciation, usually an aspiration [h]. In nearly all of Argentina, as in the remainder of the Southern Cone (Chile, Uruguay, and Paraguay), syllable-final /s/ is weakened or elided. Final /s/ is retained as a sibilant in a shrinking area of Santiago del Estero and in a tiny fringe along the Bolivian border in the far northwest. Among educated speakers from Buenos Aires, aspiration predominates over loss, which carries a sociolinguistic stigma (Fontanella de Weinberg 1974a, 1974b; Terrell 1978). In word-final prevocalic position (e.g., los amigos ‘the friends’), sibilant [s] predominates among more formal registers and in the upper socioeconomic classes. Aspiration or elision of prevocalic /s/ carries a sociolinguistic stigma in Buenos Aires, although this configuration is the logical result of /s/-weakening, following the route taken by many other Spanish dialects (Lipski 1984). Preconsonantal /s/ is routinely aspirated in all varieties of Argentinian and Uruguayan Spanish, with the exception of the varieties spoken in northern Uruguay along the Brazilian border, where a stronger final /s/, influenced by the neighboring Portuguese dialect, still prevails. However, the complete loss of syllable- and word-final /s/ continues to be highly stigmatized in Buenos Aires and Montevideo and is immediately associated with uneducated rural and marginalized urban speakers. The

[ 56 ]

chapter two

interface with speakers of Italian dialects is at least partially responsible for the extraordinary range of /s/-reduction in Río de la Plata Spanish. None of the Italian dialects implicated in contact with Río de la Plata Spanish contain word-final consonants, although word-initial and wordinternal /s/ + consonant clusters are common. Moreover there are many near-cognates with Spanish in which the typical difference is the presence of a final /s/ in Spanish and the absence of a consonant in Italian; this includes the first person plural verb endings (-mos in Spanish, -iamo in Italian), and meno/menos ‘less,’ ma/mas ‘but,’ sei/seis ‘six,’ and many others. These similarities provided a ready template for Italian speakers to massively eliminate word-final /s/ in Spanish, while retaining at least some instances of word-internal preconsonantal /s/. At the same time, the aspirated realization of syllable-final /s/ in Argentinian and Uruguayan Spanish does not correspond to any regional Italian pronunciation, and it presents a challenge to phonological interpretation. Whereas speakers of Río de la Plata Spanish dialects routinely perceive aspirated [h] as /s/ and are often surprised to realize that they are equating sibilant and aspirated variants, speakers of languages where syllable-final aspiration does not occur more often perceive the aspiration as a total absence of sound, and they reanalyze the Spanish words as not containing /s/. Italian immigrants typically dropped final /s/ in such items, even when regional varieties of Spanish realized final /s/ as an aspiration.5

11. Contact with Languages of Involuntary Immigration: The African Diaspora in Latin America By far the largest non-Spanish demographic and linguistic presence to reach Latin America was carried by the more than 8 million African slaves who for nearly four centuries provided much of the labor force in colonial and postcolonial Spanish America. In much of colonial Spanish America, populations of African origin equaled or surpassed the European population up to the time of colonial independence in the early 1800s, particularly in large urban areas. Those areas include cities that are not currently identified with significant Afro-Hispanic populations, such as Mexico City, Puebla, Asunción, and especially Buenos Aires and Montevideo, whose black populations were between 30 percent and 40 percent. Despite hundreds of literary and folkloric documents describing the halting Spanish of Africans in Spanish American colonies, as long as these populations remained in rural areas (originally working in mining, later in plantation agriculture), their speech had little effect on urban language. Only when Africans and their descendants moved to cities—to work as servants, la-

John M. Lipski

[ 57 ]

borers, and, once freed, as artisans and entrepreneurs—was it possible for their language to be heard and to exert a slight but palpable influence on the surrounding Spanish dialects. In the cities, many African-born bozales (speakers of pidginized Spanish) worked as street vendors, crying out their wares in distinctive songs or pregones, and their approximations to Spanish were often imitated in popular culture (Lipski 2001, 2005). In the Southern Cone, African vocabulary items became implanted in Argentinian and Uruguayan Spanish, the most common being mucama ‘female domestic servant’, evidently coming from Kimbundu (spoken in Angola) mukamba, meaning female attendant of a queen or female slave held in high regard by the master (Leal 2001, 292; Pereda Valdés 1965, 185). The formerly popular Argentinian dance milonga is also evidently derived from Kimbundu mu-longa (plural mi-longa) meaning ‘multitude of words’ (Pereda Valdés 1965, 185; Lienhard 2008, 111), as are other more local words. In the Caribbean, the African population was largely concentrated in rural plantations, especially in Cuba. Although dozens of authors imitated their bozal speech, it had no impact on Caribbean Spanish until freed Africans moved to the cities and their speech and music was absorbed by rebellious youth, always eager for novelty and iconoclastic behavior. This was particularly true in those regions in which large numbers of enslaved Africans arrived toward the end of the slaving period, most notably Cuba, Argentina, and Uruguay. Ultimately, the overwhelming torrent of African words and even some grammatical patterns became entrenched in the popular imagination (including the quintessentially Caribbean word chévere ‘great, fantastic’ as well as the modern Cuban asere ‘friend’), using the centrifugal force of urban speech and, later, the potent international outreach of recorded music, to spread Afro-Cuban language to people with no African heritage. In wealthier families, children were cared for by black servants. The white children learned the language of their black caretakers and the caretakers’ children and, as occurred in the southern United States, grew up in effect bidialectal. During most of the colonial period, only in a few of the largest cities, such as Havana and Cartagena, was there even a minimal amount of ghettoization, which may have fostered the retention of certain ethnically marked words or pronunciation; such an influence was seen in inner-city neighborhoods in the United States and in the townships of apartheid-era South Africa. In the remaining places, the ratio of African-born workers who learned Spanish as a second language was always small in comparison to the native Spanish-speaking population, black and white. Matters changed rapidly following the Haitian Revolution, which began in 1791. The French half of the island of Hispaniola, then known as Saint-

[ 58 ]

chapter two

Domingue, was by far the world’s largest sugar producer at the end of the eighteenth century. After the slave revolt and the establishment of the free nation of Haiti at the turn of the nineteenth century, sugar production dropped almost to zero, and Cuban sugar planters moved in to supply the world market. This required the immediate importation of hundreds of thousands of additional laborers, the majority of whom came directly from Africa; subsequently they were brought also from other Caribbean colonies. Nearly 86 percent of the more than 1.3 million African slaves taken to Cuba arrived during the first half of the nineteenth century. Newly arrived African workers were highly concentrated on sprawling sugar plantations, housed in barracks, and deprived of the broad-based contact with native speakers of Spanish that earlier generations of Africans had experienced. The last African arrivals in Cuba were split between speakers of Yoruba and related Nigerian languages, on the one hand, and speakers of Kikongo, on the other, and in both instances substrate-induced phenomena survived the first-generation pidgin stage and left traces in Afro-Cuban ritual language (Fuentes Guerra and Schwegler 2005; Castellanos 1990; Lipski 2000a). African-born bozales rarely communicated with white plantation owners or even working-class whites, but rather with a small group of free black or mulatto foremen, slave drivers, and overseers. These free blacks spoke Spanish natively, although given their own relative isolation from wider segments of the Spanish-speaking population, they may have used an ethnically marked variety. The large slave plantations prevented most of the African-born workers from acquiring full native competence in Spanish, but even though some African languages were used on the plantations, the slaves inevitably had to use Spanish with the overseers, as well as with some of the other Africans.

12. Search for African Influences on Caribbean Spanish The list of potential traces of African influence on Caribbean Spanish is quite lengthy; three frequently mentioned candidates are the loss of syllable-final /s/, atypical intonation patterns, and double negation.

loss of syllable-final /s/ Some scholars have tried to trace the massive elimination of syllableand word-final consonants in all Caribbean Spanish dialects to an African substrate, but in fact these pronunciation patterns are the direct inheritance of southern Spain and the Canary Islands, regions that supplied the

John M. Lipski

[ 59 ]

majority of settlers in Caribbean colonies. In one of the few attempts to refine the search for an African imprint on Latin American Spanish pronunciation, Megenney (1989) notes the high degree of overlap between total loss of word-final /s/ (as opposed to aspiration or other forms of consonantal reduction) and majority Afro-Hispanic populations in the Caribbean basin. The late Cuban scholar Figueroa Arencibia (1994, 1995, 1998) made similar claims for the Spanish of eastern Cuba, but without supporting evidence. Given that rates of deletion of syllable-final /s/ reach 100 percent in southern Spain and the Canary Islands, for which no African influence can be postulated and which are strongly implicated in the formation of Caribbean Spanish dialects, the case for an African contribution to /s/-elision in the Caribbean is tenuous at best. As occurred elsewhere in Latin America, speakers of African languages that contained predominantly open syllables tended to overlook weakly pronounced syllable-final consonants in regional varieties of Spanish, therefore possibly extending to the logical extreme processes of phonetic reduction already in progress.

multiple h* peaks and little downstep More likely candidates for African-influenced pronunciation patterns involve intonation and pitch accents, which only recently are emerging as objects of empirical study. Megenney (1982) noted that the vernacular speech of predominantly black communities in the Dominican Republic was characterized by unusual intonational patterns, with declarative utterances ending on a mid tone rather than the usually falling tone associated with other Spanish dialects. Subsequent work by Willis (2003a, 2003b, 2006) has confirmed typologically unusual phrase-final patterns for Dominican Spanish. In a recent study of the Afro-Iberian creole language Palenquero, Hualde and Schwegler (2008) also demonstrate intonational contours that are atypical of any Latin American Spanish dialects. In particular, all prenuclear stressed syllables receive a uniformly high tone, as opposed to the more usual downdrift and alignment of prenuclear high tones with the immediately posttonic syllable. They note that although contemporary Palenquero is a pitch-accent language like Spanish and not a tone language, unlike Kikongo and similar Bantu languages known to have participated in its formative period, “at some point in the past Palenqueros reinterpreted Spanish stress as requiring an association with a lexical H tone” (26). My own research on Afro-Hispanic speech communities reveals similar intonational contours, all of which depart from other regional varieties

chapter two

[ 60 ]

of Spanish, suggesting a common historical influence. The patterns all involve a series of early-aligned H* tones and minimal downstep across nonexclamatory nonfocused declarative utterances. In addition to Palenquero, the following Afro-Hispanic dialects routinely exhibit multiple early-aligned H* peaks. Colombia: Chocó H

H

H

H

H

po ke la pri me ra pri me ra e a

H

po a ki

Because the first one was here, down below Venezuela: Tacaragüita (Barlovento) H

H

H

jo me ka se

a ba xo

H

koŋ koŋ la ma ma mi si xo

I married the mother of my children

Mexico (Costa Chica of Guerrero and Oaxaca): San Nicolás (Guerrero) H

o ri

ta

H

hwis te

pan

H

ta

You just went to my place

mi

Afro-Bolivian Spanish (Lipski 2006a, 2006b, 2006c, 2007, 2008a). Dorado Chico (Nor Yungas, La Paz), Bolivia: H

H

ka da

H

do

H H H

H

se ma na noh to

ka

it was our turn every week

H

ba

The negros congos of Panama (Lipski 1989, 1997, 2009b, 2011): H

ko mo

H

no se

H

ma

What’s your name?

ma

u

te

H

ne

Afro-Cuban imitation of bozal Spanish from Ortiz López (1998; see also Castellanos 1990; Fuentes Guerra and Schwegler 2005): H

mu

H

tʃa ko sa

H

ta

ol bi

H

H

H

da pa loh ne gro

Black people forget many things

Ecuador: Mascarilla, Chota Valley (Lipski 1987, 2008b) H

H

H

H

H

el pri mer dwe ɲo e ra un se ɲor xe sus The first owner was Mr. Jesús Jácomi

H

xa ko mi

John M. Lipski

[ 61 ]

Equatorial Guinea, Malabo (Lipski 1985, 1990, 2000b, 2004c; Quilis and Casado-Fresnillo 1995): H

H

H

H

H

H

es te pi tʃI sur xjo kwan do bi nje ron los ni xe rjanos a gi ne a This pidgin English came out when the Nigerians arrived

From the perspective of comparative Afro-Hispanic intonation, it is noteworthy that H* pitch accents are aligned with all prenuclear stressed syllables and that typically there is no downstep of pitch accents across the expanse of an utterance. This adds to the circumstantial evidence that contact with African languages with lexical tone permanently influenced the development of Afro-Hispanic speech communities.

double negation Negation in Spanish exhibits relatively little variation over the Spanishspeaking world, and the same is true of the remaining Ibero-Romance languages. One exception to the generally unremarkable behavior of negative structures in Spanish is double negation, typically represented by the combination of preposed and postposed no with no inflection suggesting reflection or focus. This construction is confined to a couple of dialects that are characterized by a significant historical presence of African languages. Double negation is also typical of vernacular Brazilian Portuguese and of the vernacular Portuguese of Angola, Mozambique, and São Tomé and Príncipe. In Spanish America double negation is found in the vernacular speech of the Dominican Republic (Benavides 1985; Jiménez Sabater 1975, 170; Megenney 1990, 121–28; Schwegler 1996b) and in the Chocó region of northwestern Colombia (Schwegler 1991; Granda 1977; Ruíz García 2000). Chocó:

Yo no aguanté el calor de allá no (Schwegler 1991, 97) I couldn’t stand the heat there

él no ha vuelto no (Ruíz García 2000) he hasn’t returned

Dominican Republic (Schwegler 1996b): Yo no estoy llegando tarde no I’m not arriving late

The same construction is attested for nineteenth-century Afro-Cuban Spanish:

[ 62 ]

chapter two

yo no so pobre, no (Benítez del Cristo 1930) I’m not poor

No é mío, no (Cabrera 1976, 44) It’s not mine

no señó, yo no soy cuchara, no (Cabrera 1983, 443) No sir, I’m not a spoon

El amo no quiere matar Eugenio, no (Malpica la Barca 1890) The master doesn’t want to kill Eugenio

Yo no bebe guariente, no (Fernández 1987, 96) I don’t drink liquor

yo pensá que mama suyo que lo parí nelle no lo va a cuñusé, no (Cruz 1974, 231) I think that the mother that gave birth to you won’t recognize you

alma mio no va a juntar no, con cuerpo de otra gente . . . (Laviña [1797] 1989, 89)

My soul won’t join the body of another person

That these Cuban literary examples are not simple inventions is revealed by the unpublished correspondence between the Cuban scholar José de la Luz Caballero and the American encyclopedist Francis Lieber, from around 1830.6 Lieber queried whether Afro-Cubans spoke a creole language. Among other things, Luz Caballero commented on the use of double negation: como ya dije en mi respuesta, hay algunos modos de corromper el idioma empleado generalmente por todos los bozales, pero estos se refieren mas bien á las construcciones que no á la pronunciación. . . . 10ᵒ Repiten los negros casi siempre la negativa asi dicen vg. “no va á juntar no” “no va á salir no.” [as I already said in my reply, there are some means of corrupting language that are generally used by bozales, but these are mostly constructions and not pronunciation. . . . The blacks almost always repeat the negative and say “I’m not going to get together,” “I’m not going to leave.”]

For scholars seeking an African source for double negation in Dominican, Chocó, and earlier Afro-Cuban Spanish, the most likely suspect is

John M. Lipski

[ 63 ]

Kikongo, which was clearly in the right place at the right time, at least in Cuba and Colombia. In Cuba Kikongo speakers formed the palo mayombe cult, which survives to this day, including many Kikongo linguistic elements (Fuentes Guerra and Schwegler 2005). In Colombia the creole language Palenquero has a strong Kikongo component (Schwegler 1996a). Kikongo, together with some other Bantu languages, shows “double negation,” typically with ke . . . ko (cf. Bentley 1887, 607): ke be- sumba ko

neg cl. buy neg They do not buy.

Like Palenquero, the second negator (ko) occurs phrase-finally, allowing for intervening objects and adjuncts: ke be kuenda malembe ko

They don’t walk slowly (A.M.D.G. 1895, 24) ke tukwendanga lumbu yawaonso ko

We do not go every day (Bentley 1887, 607)

This is a promising candidate for substratal influence on Spanish and Portuguese, given that the placement of ko sentence-finally correlates with the position of the second negator in Afro-Iberian double negation constructions. Since the final particle ko may be optionally absent in Kikongo (in which case the sentence carries an element of surprise [A.M.D.G. 1895, 23]), convergence with Spanish and Portuguese could be further facilitated. The possibility for an African component to double negation is circumstantially plausible in the case of Colombia and Cuba, but significantly less so for the Dominican Republic. Unlike Cuba, Santo Domingo did not receive a massive surge of slaves in the early nineteenth century; most Africans arrived in Santo Domingo early in the colonial period, after which the arrival of African-born slaves slowed to a trickle (Lipski 2004b). The major extra-Hispanic influence on nineteenth- and early-twentiethcentury Dominican Spanish has been Haitian Creole, carried first by invading Haitian armies, then by settlers who arrived from the western end of the island during the Haitian occupation, and in the twentieth century by migrant sugar plantation laborers. Examples of double negation have also been recorded among elderly Haitians living in rural eastern Cuba, by Ortiz López (1998, 1999a, 1999b, 2001); similar constructions may be heard among Haitians living in the Dominican Republic:

[ 64 ]

chapter two

Cuando yo iba venil pa cá mi familia no quiere venil pa cá no

When I was going to come here my family did not want to come here La hija mía no entiende nada lo que yo hablo con él. No entiende no.

My daughter doesn’t understand anything that I say to her. She doesn’t understand.

The Cuban-Haitian data, when combined with the frequent use of double negation in rural regions of the Dominican Republic, suggest that a Haitian influence may be at least partially responsible. Haitian Creole is noted for use of a sort of double negation, combining the usual preverbal pa with cliticized phrase-final -non (ending affirmative sentences with cliticized -wi is an even more common strategy). Some of the CubanHaitian tumba francesa songs exemplify this (Alén Rodríguez 1986, 57; 1991):7 yo di mué contan

they say I am happy

mué pa capa ri no

I can’t laugh

mué pa capa contan no . . .

I can’t be happy

The presence of numerous other Haitian Creole items in rural Dominican Spanish (Lipski 1994, 2004b) reinforces the notion that Haitian Creole has penetrated Dominican Spanish for at least two centuries, despite the traditional hostility between the two peoples. Speakers of Haitian Creole were also in the right places at the right time to have influenced the formation of double negatives in Afro-Cuban bozal Spanish. The possibility that Haitian Creole has influenced (if not actually caused) double negation in Dominican Spanish is further enhanced by the existence in both kreyòl and Dominican Spanish of double affirmation (Toribio 2002), a trait not found in any other Spanish dialect. Haitian Creole: m’ byen wi

I’m doing fine ou gen pwoblèm wi papa

you’ve got problems, man Dominican Spanish:

Ella trabaja bien duro sí she works really hard

John M. Lipski

[ 65 ]

El gallito pinto puede ganar sí the spotted rooster can win

13. Summary and Conclusions The preceding sections have exemplified factors that at various times and under varying circumstances molded the Spanish language as originally transplanted to the Americas by Spanish settlers. The two main currents of influence can be grouped under the headings of sociodemographic profiles and language contact, with specific factors and relative proportions varying widely across time (five centuries) and space (two continents). Even though the majority of the early settlers came from southern and southwestern Spain (Andalusia and Extremadura), dialect leveling was already occurring among the first Spanish colonists, as evidenced by the noteworthy lack of highly regionalized features in early Latin American Spanish, as well as the absence of Ibero-Romance languages other than Castilian (e.g., Galician, Asturian/Leonese, and later Catalan and Aragonese) in the linguistic mix that set the stage for the development of Latin American Spanish. In reality it was not Andalusian Spanish that directly provided the scaffolding upon which Latin American dialects were constructed, but rather a set of common denominators resulting from dialect leveling, which, because they embodied fewer phonological oppositions and more simplified syllable structure than found in the entirety of Peninsular Spanish dialects, bear greater resemblance to Andalusian Spanish than to Castilian varieties. It is therefore difficult to separate the effects of the Founder Principle, namely the predominance of Andalusians among the earliest Spanish colonists, from the effects of dialect leveling among subsequent arrivals from diverse areas of Spain. Regardless of the overall efficacy of the Founder Principle in Spanish America, it was tempered by the rapid switch of colonial administrative control centers from Andalusia (Seville) to Castile and the subsequent arrival of large contingents of colonists speaking nonsouthern varieties of Spanish. The constant renewal of Peninsular administrative and clerical personnel, as well as continued immigration from Spain, provided a ready conduit for the transmission of linguistic innovations to Spanish American colonies, through the end of the seventeenth century. These innovations made their way from port cities and administrative capitals (the latter nearly always located inland) to smaller regional population centers, but they frequently failed to reach far-flung outposts and isolated enclaves. The pathways of diffusion of Peninsular Spanish innovations, or the lack thereof, can be traced through the presence of clusters of archa-

[ 66 ]

chapter two

isms in various Latin American regions, combined with the absence of traits arising in Spain and found in larger Latin American settlements. An example of this dichotomy is the retention of the subject pronoun vos (originally a second person plural pronoun) for familiar second person singular address, as opposed to modern Spanish tú. Vos and accompanying verb forms predominate in regions of Latin America that were geographically and sociolinguistically distant from colonial administrative centers (Páez Urdaneta 1981; Benavides 2003); this includes all of Central America, much of the Andean region, the Southern Cone, and isolated areas of Cuba. By the early eighteenth century, numerous Spanish-American speech communities had reached a critical mass in terms of demographic strength, economic and social self-sustainability, and linguistic self-awareness and no longer echoed prevailing speech patterns from Spain. From this time onward, distinctly American varieties of Spanish coalesced in both urban and rural areas. Although many factors intervened in the formation of Latin American Spanish dialects, including the inevitable effects of stochastic variation in speech communities scattered across such an immense geographical area, contact with other languages provided the most powerful shaping force in most regions. Indigenous languages, never displaced from their heartlands and spoken natively by demographic majorities, provided the backdrop for partial restructurings of Spanish. These restructured elements percolated their way upward from the secondlanguage speech of indigenous residents on the sociolinguistic periphery of colonial settlements to natively acquired Spanish, at times even as spoken monolingually by individuals with no indigenous language background. In the Andean region, Mexico, Paraguay, and much of Central America, contact with indigenous languages continues to shape local and regional Spanish dialects and is largely responsible for the unique traits of these varieties. Immigrant languages provided comparatively fewer innovations, owing principally to the relatively small numbers of immigrants at any one time and place and the typical abandonment of the ancestral language after a single generation. Italian immigration to Argentina and Uruguay was exceptional in its massive demographic impact on Buenos Aires and Montevideo, as well as in the high proportion of recognizable cognate patterns that facilitated the creation of intermediate Italo-Spanish hybrid sociolects. Also to be reckoned with is the arrival, around the turn of the twentieth century, of thousands of creole English-speaking West Indians in Costa Rica and Honduras to work in the banana industry, in Panama to participate in the construction of the Panama Canal, and in Nicaragua across a longer time span and for various reasons. To this day, the re-

John M. Lipski

[ 67 ]

gional dialects of Spanish along the Caribbean coast of Central America reflect this bilingual contact (Lipski 1986). Languages arriving with enslaved, indentured, and kidnapped laborers, from Africa, China, and Polynesia, respectively, seldom left more than occasional lexical traces, except in extraordinary cases such as the creation of Maroon communities, syncretic religious cults, or popular music and dance. In its totality, the linguistic impact of coerced servitude is smaller than might be expected given the numbers of individuals involved (several million Africans, several hundred thousand Chinese, and several thousand Pacific islanders); but when added to the crucible of language evolution, these languages provided yet another force of change (Lipski 1999; Maude 1981; McCreery and Munro 1993). In a few areas of Spanish America, language contacts in border regions have resulted in microdialects that occasionally extend beyond the immediate border zones. The impact of Haitian Creole on the Spanish of the Dominican Republic is a case in point. More regionally confined contact phenomena can be found in Spanish-speaking regions near the border with Brazil: in northern Uruguay, northeastern Argentina, northern Bolivia, and eastern Paraguay (Lipski 2008c, 2009a; Elizaincín et al. 1987), where various Spanish-Portuguese hybrid varieties can be found. In perhaps the most extensive border contact environment, namely the vast Mexican border with the United States, the impact of English on the Spanish of Mexicans who work and study in the United States but return to Mexico daily or weekly has yet to be studied in depth, despite the vast quantity of research on Mexican Spanish in the United States. In summary, it is not surprising that Spanish, a language spoken by some 400 million people spread over every continent, has diversified over the past five centuries; it would be quite surprising if this had not occurred. In Latin America, the particular trajectories of the emergent dialects were set by a unique combination of language contacts—some coincidental and others the result of deliberate practices—and the idiosyncrasies of Spanish imperialism. The preceding discussion is meant to be suggestive of the possibilities, a glimpse into both the predictable and the unexpected results of the multilingual and multicultural encounters that gave rise to the syncretic society known as Latin America. Notes 1. Such claims must confront obvious contradictions within the data of Latin American Spanish; thus, while Basque influence has been suggested for retention of the phoneme /ʎ/ (written ll) in Paraguayan Spanish (e.g., by Granda 1979), other traits of Paraguayan Spanish, such as the weak aspirated pronunciation of final /s/, stand in sharp contrast to the consonant-strong Spanish of the Basque Country. More-

[ 68 ]

chapter two

over, Basque influence was even stronger in colonial Venezuela, where the Compañía Guipuzcoana was once the major economic force, and yet Venezuelan Spanish bears absolutely no resemblance to the Spanish of the Basque region of Spain. New Mexico was also settled largely by Basques (including the founder of the first colony, Juan de Oñate), but New Mexican Spanish is vastly different from any variety heard in northern Spain. Similarly, although the early presence of Andalusian farmers in Costa Rica is undisputed, central Costa Rican Spanish is among the least “Andalusian-like” varieties of Latin American Spanish. In another striking demonstration, by 1898, on the eve of the Spanish-American War, nearly half of the Cuban population had been born in insular or peninsular Spain, and nearly 25 percent of the Cuban population came from areas of Spain where final /s/ resists effacement and where the phoneme /θ/ (zeta) is opposed to /s/, and yet this massively un-Cuban speech community left absolutely no trace on subsequent incarnations of Cuban Spanish. However, the arrival of hundreds of thousands of Italian immigrants to Buenos Aires and Montevideo beginning in the late nineteenth century left numerous traces, as will be shown here. 2. Evidence from early borrowings into Native American languages is ambiguous as regards the status of Spanish ç and z but is most consistent with the proposal of a fricative articulation (e.g., Canfield 1934, 1952; Mannheim 1988). This suggests that dialect leveling may have resulted in the selection of the fricative variants from among the variable realizations still found in late-fifteenth-century Spain. The same leveling probably occurred in Sephardic (Judeo) Spanish, a series of archaic varieties resulting from the forced diaspora that began in 1492. 3. This analysis is pursued further in Lipski 2004a. In Quechua, the case marker -ta has other functions, including adverbial and locative ones. It is also used to signal direct objects in certain double-object constructions involving verbs of helping and teaching. In nearly all(?) instances, however, ta does not appear in immediate preverbal position, nor in any other single canonical position that might cause -ta to be calqued by an object clitic in Andean Spanish. Postnominal -ta may also be followed by other enclitic particles in nondative constructions, in effect being “buried” among the clitics and not corresponding in any clear way with a Spanish element. Only in the case of accusative -ta is the linear order convergent enough with Spanish Clitic + Verb combinations to make transfer feasible. 4. A referee has suggested that Quechua “polyperson” constructions might lie at the root of clitic doubling in Andean Spanish. It may be that such configurations, together with the aforementioned reanalysis of ta, converged to push Andean Spanish in the direction of consistent clitic doubling. 5. Lavandera (1984, 64–66) confirmed that in the pronunciation of Italian immigrants in Argentina, word-final /s/ completely disappears, while preconsonantal /s/ (which is normally an aspirated [h] in Argentine Spanish), is retained as a sibilant [s]. This treatment of /s/, which departs drastically from Argentinian Spanish, duplicates Italian patterns. Meo Zilio (1989, 214) notes the widespread elimination of word-final /s/ among Italians in the Río de la Plata, except for some central-northern Italians, who sometimes added a paragogic vowel: ómnibus > onibusse. The veracity of the cocoliche literary texts can also be put to the test by comparing them with contemporary Italo-Spanish contact language. Italian immigration surged in Montevideo in the mid-twentieth century, around 1950. Some examples were collected by Barrios (1996, 1999, 2003, 2006), Barrios and Mazzolini (1994), Barrios et al. (1994),

John M. Lipski

[ 69 ]

Ascencio (2003), and Orlando (2003) among Italian immigrants in Montevideo, all of whom had emigrated from southern Italy in the 1950s. 6. I am grateful to Clancy Clements for providing me with the text of this fascinating document. 7. An additional bit of suggestive evidence comes from the Spanish dialect spoken in Güiria, Venezuela, on the Paria Peninsula near Trinidad (Llorente 1994, 1995). In this community Spanish is in contact with Trinidad French Creole, known as patois, a variant of Lesser Antilles Creole French. Double negation is found both in patois and in Güiria Spanish, but not in any other Venezuelan dialect, once more implicating Creole French as the source of double negation: yo no estoy yendo no ‘I’m not going’.

References Alén Rodríguez, Olavo. 1986. La música de las sociedades de tumba francesa en Cuba. Havana: Ministerio de Cultura. ———. 1991. The tumba francesa societies and their music. In Essays on Cuban music: North American and Cuban perspectives, ed. Peter Manuel, 77–85. Lanham, MD: University Press of America. A.M.D.G. 1895. Elements de la langue congolaise. Brussels: Desclée, de Brouwer. Ascencio, Pilar. 2003. La oración de relativo en lenguas de contacto: el cocoliche. In Aspectos de la cultura italiana en el Uruguay, ed. G. Barrios, 107–124. Montevideo: Ministerio de Educación y Cultura. Bailey, Samuel. 1999. Immigrants in the lands of promise: Italians in Buenos Aires and New York City, 1870–1914. Ithaca, NY: Cornell University Press. Barrios, Graciela. 1996. Marcadores lingüísticos de etnicidad. International Journal of the Sociology of Language 117:81–98. ———. 1999. Etnicidad y lenguaje: Le aculturación sociolingüística de los inmigrantes italianos en Montevideo. PhD diss., Universidad Nacional del Sur, Bahia Blanca, Argentina. ———. 2003. Mantenimiento y cambio de lengua en los inmigrantes italianos residentes en Montevideo. In Aspectos de la cultura italiana en el Uruguay, ed. G. Barrios, 43–87. Montevideo: Ministerio de Educación y Cultura. ———. 2006. Identidad etnolingüística y estrategias de asimilación: Inmigrantes italianos en Montevideo. In Estudios sobre contacto inter- e intralingüístico, ed. Yolanda Hipperdinger, 1:35–58. Bahía Blanca: EDIUNS (Universidad Nacional del Sur, Departmento de Humanidades). Barrios, G., and S. Mazzolini. 1994. El proceso de acomodación lingüística en los migrantes italianos residentes en Montevideo. In Analis del discurso, ed. A. Elizaincín and I. Madfes, 57–69. Montevideo: Facultad de Humanidades y Ciencias de la Educación. Barrios, G., S. Mazzolini, and V. Orlando. 1994. Lengua, cultura e identidad: Los italianos en el Uruguay actual. In Presencia italiana en la cultura uruguaya, ed. S. Álvarez de Lasowski, 97–115. Montevideo: Facultad de Humanidades y Ciencias de la Educación. Beckman, Mary, Manuel Díaz-Campos, Julia Tevis McGory, and Terrell Morgan. 2002. Intonation across Spanish, in the tones and break indices framework. Probus 14:9–36.

[ 70 ]

chapter two

Benavides, Carlos. 2003. La distribución del voseo en Hispanoamérica. Hispania 86:612–623. Benavides, Celso. 1985. El dialecto español de Samaná. Anuario de la Academia de Ciencias de la República Dominicana 9:297–342. Benítez del Cristo, Ignacio. 1930. Los novios catedráticos. Archivos del Folklore Cubano 5, no. 2: 119–146. Bentley, W. Holman. 1887. Dictionary and grammar of the Kongo language. London: Tróbner. Boyd-Bowman, Peter. 1956. Regional origins of the earliest Spanish colonists of America. PMLA 71:1152–1172. Cabrera, Lydia. 1976. Francisco y Francisca: Chascarrillos de negros viejos. Miami: Editorial C. R. ———. 1983. El monte. 3rd ed. Miami: Editorial C. R. Canfield, D. Lincoln. 1934. Spanish literature in Mexican languages as a source for the study of Spanish pronunciation. New York: Instituto de las Españas en los Estados Unidos. ———. 1952. Spanish American data for the chronology of sibilant changes. Hispania 35:25–30. Castellanos, Isabel. 1990. Grammatical structure, historical development, and religious usage of Afro-Cuban bozal speech. Folklore Forum 23, nos. 1–2: 57–84. Catalán, Diego. 1956–57. El çeçeo-zezeo al comenzar la expansión atlántica de Castilla. Boletim de Filologia 16:306–334. ———. 1957. The end of the phoneme /z/ in Spanish. Word 13:283–322. ———. 1958. Génesis del español atlántico (ondas varias a través del océano). Revista de Historia Canaria 24:233–242. Colantoni, Laura, and J. Gurlekian. 2004. Convergence and intonation: Historical evidence from Buenos Aires Spanish. Bilingualism 7, no. 2: 107–119. Cruz, Mary. 1974. Creto Gangá. Havana: Instituto Cubano del Libro Contemporáneos. D’Imperio, Mariapaola. 2002. Italian intonation: An overview and some questions. Probus 14:37–69. Elizaincín, Adolfo, Luis Behares, and Graciela Barrios. 1987. Nos falemo brasilero. Montevideo: Editorial Amesur. Elordieta, Gorka. 2003. The Spanish intonation of speakers of a Basque pitch-accent dialect. Catalan Journal of Linguistics 2:67–95. Escobar, Anna María. 1994. Andean Spanish and bilingual Spanish: Linguistic characteristics. In Language in the Andes, ed. Peter Cole, Gabriella Hermon, and Mario Daniel Martín, 51–73. Newark: Latin American Studies, University of Delaware. Face, Timothy. 2000. Prosodic manifestations of focus in Spanish. Southwest Journal of Linguistics 19:45–62. Fernández, Francisco. 1987. El negro cheche. In Teoría y práctica del catedratismo en Los negros catedráticos de Francisco Fernández, ed. Matías Montes Huidobro. Miami: Editorial Persona. Figueroa Arencibia, Vicente Jesús. 1994. La marca de plural (sintagma nominal) en el español no estándar de una barriada de Santiago de Cuba. Anuario de Lingüística Hispánica 10:103–121. ———. 1995. La marca de plural (sintagma nominal) en el español no estándar de Santiago de Cuba. In Estudios de literatura y cultura colombianas y de lingüística afro-

John M. Lipski

[ 71 ]

hispánica, ed. Peter Konder, Mathias Perl, and Klaus Pörtl, 185–204. Frankfurt: Peter Lang. ———. 1998. Un rasgo semicriollo en el español no estándar en la región suroriental cubana: El tratamiento de /-s/. PhD diss., Universidad de La Habana. Fontanella de Weinberg, M. 1974a. Aspectos sociolingüísticos del uso de -s en el español bonaerense. Orbis 23:85–98. ———. 1974b. Comportamiento ante -s de hablantes femininos y masculines del español bonaerense. Romance Philology 28:50–58. ———. 1987. El español bonaerense: Cuatro siglos de evolución lingüística (1580–1980). Buenos Aires: Hachette. Fuentes Guerra, Jesús, and Armin Schwegler. 2005. Lengua y ritos del Palo Monte Mayombe: Dioses cubanos y sus fuentes africanas. Frankfurt: Vervuert/Iberoamericana. Gómez López de Terán, Noemi, and Mirta Estela Assís. 1977. Uso del pronombre personal átono “lo.” In Primeras jornadas nacionales de dialectología, 299–302. Tucumán: Facultad de Filosofía y Letras, Universidad Nacional de Tucumán. Granda, Germán de. 1977. Estudios sobre un área dialectal hispanoamericana de población negra: Las tierras bajas occidentales de Colombia. Bogotá: Instituto Caro y Cuervo. ———. 1979. Factores determinantes de la preservación del fonema /ll/ en el español del Paraguay. Lingüística Española Actual 1:403–412. Guitarte, Guillermo. 1980. Para una periodización de la historia del español de América. In Perspectivas de la investigación lingüística hispanoamericana, ed. Juan Lope Blanch, 119–137. Mexico: Universidad Nacional Autónoma de México. Heine, Bernd, and Tania Kuteva. 2005. Language contact and grammatical change. Cambridge: Cambridge University Press. Hualde, José Ignacio, and Armin Schwegler. 2008 Intonation in Palenquero. Journal of Pidgin and Creole Languages 23:1–31. Jiménez Sabater, Max. 1975. Más datos sobre el español en la República Dominicana. Santo Domingo: Ediciones Intec. Justiano de la Rocha, Dora. 1986. Apuntes sobre las lenguas nativas en el dialecto español de Bolivia. La Paz: Instituto Internacional de Integración del “Convenio Andrés Bello.” Kaisse, Ellen. 2001. The long fall: An intonational melody of Argentinian Spanish. In Features and interfaces in Romance, ed. Julia Herschensohn, Enrique Mallén, and Karen Zagona, 148–160. Amsterdam: John Benjamins. Lapesa, Rafael. 1981. Historia de la lengua española. 9th ed. Madrid: Gredos. Laprade, Richard. 1976. Some salient dialectal features of La Paz Spanish. Master’s thesis, University of Florida. ———. 1981. Some cases of Aymara influence on La Paz Spanish. In The Aymara language in its social and cultural context, ed. M. J. Hardman, 207–227. Gainesville: University Press of Florida. Lavandera, Beatriz. 1984. Variación y significado. Buenos Aires: Hachette. Laviña, Javier, ed. [1797] 1989. Doctrina para negros: Explicación de la doctrina cristiana acomodada a la capacidad de los negros bozales, de Nicolás Duque de Estrada. Barcelona: Sendai. Leal, Gilberto. 2001. Farígá/Ìfaradà: Black resistance and achievement in Brazil. In African roots/American cultures: Africa in the creation of the Americas, ed. Sheila Walker, 291–300. Oxford: Rowman and Littlefield.

[ 72 ]

chapter two

Lienhard, Martin. 2008. Milonga: The “dialogue” between Portuguese and Africans in the Congo and the Angola Wars (sixteenth and seventeenth centuries). In AfriAmericas: Itineraries, dialogues, and sounds, ed. Ineke Phaf-Rheinberger and Tiago de Oliveira Pinto, 91–124. Frankfurt: Vervuert/Iberoamericana. Lipski, John. 1984. On the weakening of /s/ in Latin American Spanish. Zeitschrift für Dialektologie und Linguistik 51:31–43. ———. 1985. The Spanish of Equatorial Guinea. Tübingen: Max Niemeyer. ———. 1986. Mexican-American English and Central American English: Sociolinguistic mirror images. In Focus on the Caribbean, ed. Manfred Görlach and John Holm, 191–208. Amsterdam: John Benjamins. ———. 1987. The Chota Valley: Afro-Hispanic language in highland Ecuador. Latin American Research Review 22:155–70. ———. 1989. The speech of the negros congos of Panama. Amsterdam: John Benjamins. ———. 1990. El español de Malabo: Procesos fonéticos/fonológicos e implicaciones dialectológicas. Madrid/Malabo: Centro Cultural Hispano-Guineano. ———. 1994. A new perspective on Afro-Dominican Spanish: The Haitian contribution. Research Paper no. 26, University of New Mexico Latin American Institute. ———. 1997. El lenguaje de los negros congos de Panamá y el lumbalú palenquero: Función sociolingüística de criptolectos afrohispánicos. América Negra 14:147–165. ———. 1999. Chinese-Cuban pidgin Spanish: Implications for the Afro-creole debate. In Creole genesis, attitudes, and discourse, ed. John Rickford and Suzanne Romaine, 215–233. Amsterdam: John Benjamins. ———. 2000a. Bozal Spanish: Restructuring or creolization? In Degrees of restructuring in creole languages, ed. Ingrid Neumann-Holzschuh and Edgar Schneider, 55–83. Amsterdam: John Benjamins. ———. 2000b. The Spanish of Equatorial Guinea: Research on la hispanidad’s bestkept secret. Afro-Hispanic Review 19:11–38. ———. 2001. Panorama del lenguaje afrorrioplatense: Vías de evolución fonética. Anuario de Lingüstica Hispánica 14:281–315. ———. 2002. The role of the city in the formation of Spanish American dialect zones. Arachne 2:1, http://arachne.rutgers.edu/vol2_1lipski.htm. ———. 2004a. El español de América: Los contactos bilingües. In Historia de la lengua española, ed. Rafael Cano, 1117–1138. Barcelona: Ariel. ———. 2004b. Nuevas perspectivas sobre el español afrodominicano. In Pensamiento lingüístico sobre el Caribe insular hispánica, ed. Sergio Valdés Bernal, 505–552. Santo Domingo: Academia de Ciencias de la República Dominicana. ———. 2004c. The Spanish of Equatorial Guinea. Arizona Journal of Hispanic Cultural Studies 8:115–130. ———. 2005. A history of Afro-Hispanic language: Five centuries and five continents. Cambridge: Cambridge University Press. ———. 2006a. Afro-Bolivian language today: The oldest surviving Afro-Hispanic speech community. Afro-Hispanic Review 25, no. 1: 179–200. ———. 2006b. Afro-Bolivian Spanish and Helvécia Portuguese: Semi-creole parallels. Papia 16:96–116. ———. 2006c. El dialecto afroyungueño de Bolivia: En busca de las raíces el habla afrohispánica. Revista Internacional de Lingüística Hispanoamericana 3, no. 2: 137–166.

John M. Lipski

[ 73 ]

———. 2007. Afro-Yungueño speech: The long-lost “black Spanish.” Spanish in Context 4:1–43. ———. 2008a. Afro-Bolivian Spanish. Frankfurt: Vervuert/Iberoamericana. ———. 2008b. Afro-Choteño speech: Towards the (re)creation of a “Black Spanish.” Negritud 2:99–120. ———. 2008c. Searching for the origins of Uruguayan Fronterizo dialects: Radical code-mixing as “fluent dysfluency.” Journal of Portuguese Linguistics 8:5–46. ———. 2009a. “Fluent dysfluency” as congruent lexicalization: A special case of radical code-mixing. Journal of Language Contact 2:1–39. ———. 2009b. Tracing the origins of Panamanian Congo speech: The pathways of regional variation. Diachronica 26:380–407. ———. 2011. El habla de los Congos de Panamá en el contexto de la lingüística afrohispánica. Panama: Instituto Nacional de Cultura. Llorente, M. L. 1994. Materiales para el estudio del patois de Güiria. Licenciatura thesis, Universidad Católica Andrés Bello, Caracas. ———. 1995. El patois de Güria: Una lengua criolla del estado Sucre. Montalbán 28:7–19. Lockhart, James, and Stuart Schwartz. 1983. Early Latin America. Cambridge: Cambridge University Press. Luján, Marta. 1987. Clitic-doubling in Andean Spanish and the theory of case absorption. In Language and language use: Studies in Spanish, ed. Terrell Morgan, James Lee, and Bill VanPatten, 109–121. Washington, DC: University Press of America. Malpica la Barca, Domingo. 1890. En el cafetal. Havana: Tipografía de “Los niños huérfanos.” Mannheim, Bruce. 1988. On the sibilants of colonial southern Peruvian Quechua. International Journal of American Linguistics 54:168–208. Maude, H. E. 1981. Slavers in paradise: The Peruvian slave trade in Polynesia, 1862– 1864. Stanford, CA: Stanford University Press. McCreery, David, and Doug Munro. 1993. The cargo of the Montserrat: Gilbertese labor in Guatemalan coffee, 1890–1908. Americas 49, no. 3: 271–295. Megenney, William. 1982. Elementos subsaháricos en el español dominicano. In El español del Caribe, ed. Orlando Alba, 183–201. Santiago de los Caballeros: Universidad Católica Madre y Maestra. ———. 1989. An etiology of /-s/ deletion in the Hispanic Caribbean: Internal process or substratum influence? In Estudios sobre el español de América y lingüística Afroamericana, 300–327. Bogotá: Instituto Caro y Cuervo. ———. 1990. Africa en Santo Domingo: La herencia lingüística. Santo Domingo: Museo del Hombre Dominicano. Mendoza, José. 1991. El castellano hablado en La Paz: Sintaxis divergente. La Paz: Universidad Mayor de San Andrés. Meo Zilio, Giovanni. 1989. Estudios hispanoamericanos: Temas lingüísticos. Rome: Bulzoni. Mufwene, Salikoko S. 1996. The founder principle in creole genesis. Diachronica 13:83–134. Muysken, Pieter. 1984. The Spanish that Quechua speakers learn: L2 learning as norm-governed behavior. In Second languages: A cross-linguistic perspective, ed. Roger Andersen, 101–119. Rowley, MA: Newbury House.

[ 74 ]

chapter two

Nascimbene, Mario. 1988. Los italianos y la integración nacional. Buenos Aires: Ediciones Selección Editorial. Orlando, Virginia. 2003. Cambio de código, estrategia comunicativa y marcador sociolingüístico de identidad de los italianos residentes en Montevideo. In Aspectos de la cultura italiana en el Uruguay, ed. G. Barrios, 89–106. Montevideo: Ministerio de Educación y Cultura. O’Rourke, Erin. 2004. Peak placement in two regional varieties of Peruvian Spanish intonation. In Contemporary approaches to Romance linguistics (LSRL 33), ed. Julie Auger, J. Clancy Clements, and Barbara Vance, 321–341. Amsterdam: John Benjamins. ———. 2005. Intonation and language contact: A case study of two varieties of Peruvian Spanish. PhD diss., University of Illinois Urbana-Champaign. Ortiz López, Luis. 1998. Huellas etno-sociolingüísticas bozales y afrocubanas. Frankfurt: Vervuert. ———. 1999a. El español haitiano en Cuba y su relación con el habla bozal. In Lenguas criollos de base lexical española y portuguesa, ed. Klaus Zimmermann, 177–203. Frankfurt: Vervuert. ———. 1999b. La variante hispánica haitianizada en Cuba: Otro rostro del contacto lingüístico en el Caribe. In Estudios de lingüística hispánica: Homenaje a María Vaquera, ed. Amparo Morales et al., 428–456. Río Piedras, Puerto Rico: Editorial de la UPR. ———. 2001. El sistema verbal del español haitiano en Cuba: Implicaciones para las lenguas en contacto en el Caribe. Southwest Journal of Linguistics 20, no. 2: 175–192. Páez Urdaneta, Iraset. 1981. Historia y geografía hispanoamericana del voseo. Caracas: Casa de Bello. Pereda Valdés, Ildefonso. 1965. El negro en el Uruguay pasado y presente. Montevideo: Revista del Instituto Histórico y Geográfico del Uruguay, no. 25. Pozzi-Escot, Inés. 1972. El castellano en el Perú: Norma culta nacional versus norma culta regional. In Escobar 1972, 125–142. Prieto, Pilar, Jan van Santen, and Julia Hirschberg. 1995. Tonal alignment patterns in Spanish. Journal of Phonetics 23:429–451. Quilis, Antonio, and Celia Casado-Fresnillo. 1995. La lengua española en Guinea Ecuatorial. Madrid: Universidad Nacional de Educación a Distancia. Rojas, Elena. 1980. Aspectos del habla en San Miguel de Tucumán. Tucumán: Facultad de Filosofía y Letras, Universidad Nacional de Tucumán. Rosenblat, Angel. 1977. Los conquistadores y su lengua. Caracas: Universidad Central de Venezuela. Ruíz García, Marta. 2000. El español popular del Chocó: Evidencia de una reestructuración parcial. PhD diss., University of New Mexico. Schwegler, Armin. 1991. El español del Chocó. América Negra 2:85–119. ———. 1996a. “Chi ma nkongo”: Lengua y rito ancestrales en El Palenque de San Basilio (Colombia). 2 vols. Frankfurt: Vervuert. ———. 1996b. La doble negación dominicana y la génesis del español caribeño. Hispanic Linguistics 8:247–315. Sosa, Juan. 1999. La entonación del español. Madrid: Cátedra. Stratford, Billie Dale. 1989. Structure and use of Altiplano Spanish. PhD diss., University of Florida.

John M. Lipski

[ 75 ]

———. 1991. Tense in Altiplano Spanish. In Sociolinguistics of the Spanish-speaking world, ed. Carol Klee, Luis Ramos-García, 163–181. Tempe, AZ: Bilingual Review Press. Suñer, Margarita, and María Yépez. 1988. Null definite objects in Quiteño. Linguistic Inquiry 19:511–519. Terrell, Tracy. 1978. Aspiración y elisión de /s/ en el español porteño. Anuario de Letras 16:45–66. Toribio, Almeida Jacqueline. 2002. Focus on clefts in Dominican Spanish. In Structure, meaning, and acquisition in Spanish: Papers from the 4th Hispanic Linguistics Symposium, ed. James Lee, Kimberly Geeslin, and J. Clancy Clements, 130–146. Somerville, MA: Cascadilla Press. Willis, Erick. 2003a. The intonational system of Dominican Spanish: Findings and analysis. PhD diss., University of Illinois. ———. 2003b. Prenuclear tone alignment in Dominican Spanish. In 15th International Conference of Phonetic Sciences, 2941–2944. Barcelona: Universidad Autónoma de Barcelona. ———. 2006. Utterance signaling and tonal levels in Dominican Spanish declaratives and interrogatives. Journal of Portuguese Linguistics 5:179–205.

3 * Amerindian Language Islands in Brazil hildo honório do couto

1. Introduction The term language island was coined in German dialectology in 1847 to designate the language of a small community in the neighborhood of Königsberg (the hometown of the philosopher Immanuel Kant), in East Prussia. The object the term designates had been studied much earlier, regarding two villages, Siebenburgen in the sixteenth century and Zips in the seventeenth century. The neogrammarian Jost Winteler had dedicated his Ortsgrammatik Die Kerenzer Mundart des Kanton Glarus (1876) to a few language islands. Even Jakob Grimm collected a few words from language islands. The concept was definitively incorporated in German linguistics tradition by about 1900. In general, a language island (LI) designates a group of immigrants forming a relatively small speech community within the territory of speakers of another language. In the German philological tradition, it is used to designate both German LI’s outside the German territory and foreign LI’s within. One of the most famous examples of the latter is Lusatia (Lausitz) or Sorbia, the Slavic LI of the Wends or Sorbs (German die Wende/wendisch). As examples of the former, we could mention the German LI’s of Friul (Sauris), Romania, Poland, and Russia, and several others. Outside of Europe, we can find German LI’s in Argentina, Australia, Brazil, Canada, Paraguay, the United States, and other countries (Clyne 1975; Rosenberg 2001, 2003). The expression language island is not very popular in the Anglophone and Romance linguistics traditions, where the term enclave is generally used to designate roughly the same thing. However, I think language island is preferable to enclave because it suggests that the territory of the relevant population (with its language) is a kind of island within another population (with its culture and language) analogized as an ocean. Further, it implies that there is a hinterland from which the “island” is somehow detached and to which it remains related. Interestingly, enclave was first used in linguistics to translate the German word Sprachinsel ‘language island’ (Mattheier 1996; Wiesinger 1973). An LI constitutes a good test case for contact linguistics. Because it

Hildo Honório Do Couto

[ 77 ]

is generally associated with a relatively small speech community, the contact with the surrounding language (for instance, the contact of Portuguese with Tupi-Guarani language varieties during colonial period) is much more amenable to direct observation than most other instances of language contact. Some of the Tupi-Guarani language varieties are now lost forever. Language obsolescence is also easily observable in an LI. In summary, I think LI’s are convenient observatories for the study of language contact (Rosenberg 2001, 2003). My main objective in this chapter is to show that in many countries colonized by European powers, such as Brazil, there are today several Amerindian LI’s. The irony is that the Amerindians did not migrate to the colonizer’s territory. In contrast to all the instances of LI’s mentioned above, Amerindian peoples became minorities in their own territories. Since, as Mufwene (2001) correctly says, a language is a parasitic species whose existence depends on the population that speaks it, what come into contact directly are not languages but the populations (or better, parts thereof) that use them. The next section of the chapter lays down the framework of language contact in which LI’s will be studied. Section 3 continues the preparation of the scenario by giving some ideas about the linguistic situation in Brazil before the arrival of the Portuguese. Section 4 rounds out this scenario, through an exposition of the first contacts of Amerindian languages with Portuguese. Sections 5, 6, and 7 discuss the concept of Língua Geral as well as two of its manifestations, Língua Geral Paulista and Língua Geral Amazônica. Sections 8 through 11 focus on my main theme, articulating various types of LI’s and what usually happens to them, namely, obsolescence, which frequently leads to language death. Section 12 contains concluding remarks.

2. Language Islands in the Framework of Language Contact In order to understand how LI’s emerge, it is necessary to briefly outline the ecological approach to language contact that I adopt. It is called ecological because it takes concepts from biological ecology in order to build its epistemological foundations, as Mufwene (2001, 2005) does in dealing with the ecology of language evolution and contact. From the early 1990s onward, this approach has been called ecolinguistics, defined as the study of the relationships between language and its environment (Haugen 1972; Salzinger 1979; Fill 1993; Couto 2007). In this case, the first step is to look for the most important concepts of ecology. These are the ecosystem, which comprises both the organisms living in it and the interrelationships that obtain among them, on the one hand, and between

[ 78 ]

chapter three

them and their environment, on the other. The organisms of the same species that live together in a certain territory are called a population (Odum 1971). The second step is to look for the equivalents of these concepts in the domain of language, beginning with ecosystem. Whenever laypersons hear the name of a language unknown to them, their first question is, “Which people speak this language?” Upon hearing the answer, the second question is, “Where do these people live?” That is to say, it is common sense to assume that, for a language (L) to exist, there must be a people or population (P) living in a specific place or territory (T) that uses it. The system formed by these three elements constitutes what has been called the Fundamental Ecosystem of Language (FEL), or Community. Precursors of this concept can be found in works of the German linguist Wilhelm Trampe, the Danish Jørgen Døør and Jørgen Christian Bang, and the psychologist Kurt Lewin, as well as in Charles Sanders Peirce’s triadic sign. It is important to point out that the idea of FEL is especially applicable to the languages of small ethnic groups, such as the surviving Amerindian populations. The environment of the population is part of the ecosystem. Inside this ecosystem (made up of L, P and T), PT constitutes the Fundamental Environment of Language. On closer inspection, however, we can see that FEL may be broken into three smaller ecosystems, together with their subenvironments. The first one emerges when L is considered in relation to the totality of the members of P as an organized whole, that is, society. This is the Social Ecosystem of Language, inside which society is the Social Environment Language.1 Most studies in ecolinguistics deal with phenomena of this domain, such as language contact, multilingualism, language and its environment, discourse analysis, ecological correctness, and so on. The subject of LI belongs to this domain, too, because most Amerindian people (if not all) were prototypical representatives of FEL. Their languages entered a process of decay precisely when this prototypicality began to be disturbed by the colonizers, or, more precisely, the invaders. The second subecosystem of language consists of L in relation to the brains of the individuals that compose P, in which case we have the Mental Ecosystem of Language, formed by L and its mental environment. The Mental Environment of Language is the brain, or the network of neuronal connections, with their synapses. Phenomena pertaining to this domain have been studied in psycholinguistics, neurosciences (especially neurolinguistics), and connectionism, among others. Some of the main questions traditionally addressed here are the formation, storage, processing, and decay of language in the neuronal network of the brain. The third subecosystem of language is the Natural Ecosystem of Lan-

Hildo Honório Do Couto

[ 79 ]

guage, namely the totality formed by language and the world as well as their interrelationships. The “world” constitutes the Natural Environment of Language. The study of this domain goes as far back as the Greek philosophers. For instance, in the famous dialogue Cratylus by Plato, the character Cratylus defended the thesis that the shapes of words are determined by the things they refer to. The opposite point of view was sustained by Hermogenes, who said that words do not have anything in common with the things they designate. They are related to things by convention. The whole philosophical tradition called reference deals with this type of relationship. In linguistics it is not very popular. However, Haiman (1980) is of the opinion that even grammar reflects the world up to a certain point. In view of the definition of ecolinguistics given above, there are some questions that should be asked, including the following: What is a language? What is a population? What is a territory? What is the environment of a language? What is the nature of the relationships that hold between them? There is not enough space here to discuss all of these questions. Suffice it to give a brief delimitation of what is intended by language. I understand by language the way the members of a specific community communicate verbally among themselves.2 That is, I emphasize performance over competence or grammar, especially in the way this is understood in generative grammar. From an ecolinguistic point of view, grammar is what makes the interrelationships, or interactions, between any two members of the Community successful. Grammar is only one part of language, though a very important one. It is a necessary, but not a sufficient, condition for the existence of language. For more discussion of these and related subjects, see Fill (1993) and Couto (2007). Since an LI emerges when members of a P migrate from their original territory to the territory of another P (Couto 2009), I would like to describe at least four types of migration in order to emphasize which of them are of interest in LI studies. Recall that it is not languages that enter in contact directly, but peoples/populations or parts thereof. This fact is consistent with Mufwene’s (2001, 2005) claim that L is a parasitic species whose life depends on the life of the members of P. Weinreich (1953) was one of the first, if not the very first, to say that the locus of language contact is the speakers’ minds/brains.

[ 80 ]

chapter three

The first situation refers to cases where members of a certain P and its language (PL2) migrate to the territory of another P that is relatively well established in its own territory and speaking its own language (PL1). Here PL1 is the “stronger” side of the contact and PL2 is the “weaker” one. Examples of this type of migration are the Turks in Germany, the Arabs in France, and the Hispanics in the United States, with some exceptions in the South. What happens linguistically in this case depends on the specific ecology of each contact. For example, if just one family migrates, the tendency is toward what has been called the three-generations theory. The adults will acquire at most an L2 variety of the local language, and their children will tend to be bilingual, fluent in both their parents’ language and in the language of the host P. The third generation will tend to be monolingual in the host P’s language and have only passive knowledge of their ancestors’ language. However, if the number of immigrants is higher, they may constitute an LI in the host people’s territory. The vitality of the language also depends on the ecology of each case. The tendency is for the immigrant language to undergo attrition as the host language prevails. Ultimately, it may disappear. The second situation represents cases where it is members of a more prestigious or a militarily or economically “stronger” population and its language (PL1) that move to the territory of a “weaker” one (PL2). The most conspicuous examples are the invasion and colonization of Africa, the Americas, and parts of Asia by European powers. The linguistic outcome of this process depends on the ecology. However, in many cases, European colonizers imposed their language on the local people or, according to Mufwene (2005), created an economic system in which natives became dependent in order to earn their living and thereby were pressured to shift vernaculars. What happened in many cases was that the colonists actually killed the natives and consequently their languages, as occurred in Cuba, Hispaniola (today’s Haiti and Dominican Republic), and several other Caribbean islands. In other cases, the native populations were pushed to marginal tracts of land, where they were reduced to linguistic minorities (and LI’s). Most of the still surviving Amerindian languages (if not all of them) are undergoing obsolescence owing to contact with a European language now spoken by the majority. This is what happened in Brazil and the United States, for instance. The third and fourth situations are not conducive to the formation of LI’s. The former obtains when both PL1 and PL2 migrate to a third territory, in which case both of them are exogenous, in the sense of Chaudenson (1989). This occurred, for example, in the Cape Verde Islands and in Mauritius as well as in Curaçao, where creole language varieties emerged.

Hildo Honório Do Couto

[ 81 ]

The latter refers to cases in which both PL1 and PL2 remain in their own territory, but members of each go back and forth between the two territories. This is the case of the Russians who in the late nineteenth century used to spend the summer months in northern Norway in order to sell their products and buy fish. The result of this contact is what came to be known in creole studies as Russenorsk. Something similar happened with the Hiri travels in Papua New Guinea, even before the arrival of the Europeans. Border situations where there is a river or a mountain chain separating the two territories may also fall into this category.

3. The Linguistic Situation in Brazil before the Arrival of the Portuguese The ancestors of the South American Indians passed the isthmus of Panama in around 12,000 BC. The Portuguese colonizers arrived in AD 1500, and it is impossible even to know the exact number of languages that existed before that relatively early date. However, employing the model of Labov (1977), using the present to explain the past, we can at least make some estimates. Some of the visitors to the newly discovered land wrote reports about the then existing ethnic groups. As early as 1531, there is information about the Tupi language, which was spoken all along the coast near what is São Paulo today. Varieties of the same language were spoken by several groups from there up to the Amazon. One of them was Tupinambá, which the Portuguese adopted as a lingua franca to communicate with the indigenous populations, using Tupi/Tupinambá interpreters. For this reason, the Portuguese referred to these language varieties as a mais geral pela costa ‘the most generally spoken all along the coast’. The expression língua geral was so frequently used that later on it became the name of a restructured variety of Tupi and Tupinambá. However, even in those times, the Europeans came into contact with several other languages, which Tupinambá speakers themselves called Tapuian languages, that is, languages of the enemies. Around 1584, Fernão Cardim counted 76 Tapuian, that is, non-Tupi/ Tupinambá, languages. In 1639 the famous priest Antônio Vieira said that more than 150 languages were spoken in the Amazon alone. Among the languages bearing Tupi-Guarani names, 61 different peoples were identified in the eighteenth century: Pixuna ‘the blacks’, Guaranaguat ‘Guarana eaters/drinkers’, Pyraguat ‘fish eaters’, and so on. Others bore Maué names such as Aitoria and Apanaria. Among the other languages mentioned in colonial writings, we could include Kariri (states of Bahia and Sergipe), Yatê (Pernambuco), and Xokleng (state of Santa Catarina). Maromirim/Guarulho is sometimes also mentioned. Other languages that

[ 82 ]

chapter three

were mentioned a little later include Manau (which supplied the name of the capital of Amazonas state, Manaus), Makuxi, Tikuna, Baré, and Baniwa (Amazon). We could add to these languages all those mentioned below. Taking all this data into consideration, as well as the geographical distribution of the languages about which we have some kind of information, Rodrigues (1993) estimated that the number of languages spoken in what is Brazil today at the time of the arrival of the Portuguese may have ranged between 1,175 and 1,273.

4. The First Contacts between Portuguese and Amerindian Languages The Portuguese arrived in what is now Brazil on April 23, 1500. There is a relatively detailed description of what happened during this very first contact between the Portuguese and the Amerindians. In the famous Carta de Caminha (‘letter of Caminha’) written by the scribe of Pedro Álvares Cabral’s fleet, Pero Vaz de Caminha explains that the very first contact was relatively successful. For example, when the Portuguese saw about eight Indians walking along the seashore, Captain Álvares Cabral sent Nicolau Coelho to meet them and explore the land. Other Indians appeared, so that at the time of getting together they were about twenty. Since they were coming in the direction of the Portuguese, Coelho “asked them through signs to put their bows and arrows down, and so they did” (Caminha 1943, 134). Caminha said several times that the inhabitants of the land were a sort of people “whom nobody can understand,” but even so, several other interactions took place, mainly through gestures. The two parties exchanged a lot of products from their respective cultures, without words. In a few days the Amerindians were helping the Portuguese carry fresh water to their ships in order to fill the barrels of water for the continuation of the trip toward India. This latter interaction is emblematic of what happened later, that is, many Amerindians were enslaved in order to work for the Portuguese. Occasionally, a few Amerindians sang and danced together with the Portuguese, a symbiotic interaction I have called elsewhere communion (Couto 2001).3 This hospitality on the part of the Amerindians facilitated the colonization of the whole country by the Portuguese a few centuries later as well as the near extermination of the natives. Three decades later, the Diario da navegação by Pero Lopes de Sousa (1531) says that on the coast of Cananéia, south of the present-day state of São Paulo, there were about six Europeans and that at least one of them was a good translator into and from the local language, namely, Tupi, which, as noted above, was spoken all along the coast, up to the

Hildo Honório Do Couto

[ 83 ]

twenty-seventh latitude (Gândavo 1575), albeit under various names. In 1557 the Suma da doutrina cristã was translated into the Tupi language by João Azpilcueta Navarro. The priest Manoel da Nóbrega said, in Cartas do Brasil (1549–60), that the Jesuits always tried to learn the local languages. Although in their opinion the Amerindians were “so brute that they [did] not even have words” (apud Anchieta 1990, 140), the Jesuits wanted to teach the Christian doctrine to the Amerindians. In 1595 the first grammar of this language appeared in Coimbra, the famous Arte de grammatica da língua mais usada na costa do Brasil (Art of the grammar of the most used language along the coast of Brazil), written by José de Anchieta, who had come to the New World to replace Navarro. In the same year Anchieta also wrote Dialogo da Doutrina Cristã in “língua brasílica” (the Brazilian language), a variety of what later became known as Língua Geral, as explained below. A little later, there were many mestizos in São Vicente (near what is São Paulo city today). Both these mestizos and their mothers spoke only Tupi, albeit an altered variety of it, as a result of the contact with Portuguese and other languages. Every literate newcomer to São Vicente had to read Anchieta’s grammar. As a matter of fact, what the Portuguese called Língua Brasílica or Língua Geral became so important in the first days of Portuguese colonization that Luís da Grã (1523–1609) imposed its study among the Jesuits of the present state of Pernambuco. One century later, Antônio Pais de Sande (1693–94), said, “The Lingua Geral do Brasil [the general language of Brazil] is a door through which we can understand the other languages” of the land. According to him, “The children of the people from São Paulo speak only the Indian language,” that is, Língua Geral (apud Holanda 1979, 89). Around 1768, Língua Geral was still the language of women in São Paulo, and of their children, of course, who had at most a passive knowledge of their fathers’ language. Hércules Florence heard Língua Geral from old people around the years 1825–29. This is one of the last mentions of the use of this language. It disappeared suddenly.

5. Língua Geral As suggested in section 4, the Portuguese expression língua geral (plural línguas gerais) and the Spanish equivalent lengua general were very commonly used in the early sixteenth century to refer to Amerindian languages that were widely spoken and were often used as lingua francas. In Spanish America, Quechua was referred to as la lengua general de los indios del reyno del Perú ‘the general language of the Indians of the Kingdom of

[ 84 ]

chapter three

Peru’. Guarani was called la lengua general guarani ‘the general Guarani language.’ In Portuguese America, more than one designation were used. The Tupi-Guarani language variety spoken from the Rio de Janeiro coast up to the Northeast and the North (Pará) was called Tupinambá from the eighteenth century onward, in order to distinguish it from the mixed variety (later called Língua Geral) that was already arising. The term Tupi was applied to the variety spoken in São Vicente, near São Paulo. There were also the Tupiniquim of Espírito Santo and the Potiguara of the northeastern state of Paraíba. In modern times linguists generally refer to all these language varieties as Tupi, although the Tupis themselves lived only along the São Paulo coast. The varieties of the extreme South and of Paraguay are called Guarani. Then, the Tupinambá/Tupi language varieties were diversely called língua do Brasil ‘language of Brazil’, língua da terra ‘language of the land’, and língua do mar ‘language of the sea’. In the seventeenth century, they collectively received the name Língua Brasílica (Brazilian language), which is the designation used by the historian M. Kittiya Lee (chap. 5, this volume). According to Rodrigues, “in the 16th century the Portuguese began learning the Língua Brasílica, because they [the Portuguese] were relatively a small minority in relation to the Tupinambás. Since most colonizers came to Brazil without women, they used to live with Indian ones. One result of this inter-mixing of races is that the Língua Brasílica (Tupinambá) became the mother tongue of their children.” This was especially true in Bahia, where the whole population became part of the colonial system. Rodrigues adds, “It was to this popular language, which was widespread among acculturated Indians and non-Indians, that the name Língua Geral was systematically applied. The systematic use of this name began in the second half of the 17th century.” (1986, 101). Something similar happened on the plateau of Piratininga, now the city of São Paulo. Nowadays, the expression língua geral is applied at least to two language varieties. The first one is the restructured variety of Tupi that was spoken in São Paulo, including the town of São Vicente and the Upper Tietê River. This is the language used by the bandeirantes (a kind of pioneers), who took it to the present states of Minas Gerais, Goiás, and Mato Grosso, as well as to the south of Brazil. The apogee of this Língua Geral lies between the mid-seventeenth and the mid-eighteenth centuries. By the early nineteenth century, it had practically disappeared. The second variety of Língua Geral developed out of northern Tupinambá, especially in the states of Maranhão and Pará, from the second half of the seventeenth century onward. This language was taken to the valley of the Amazon River by missionaries and troops. It became the lingua franca

Hildo Honório Do Couto

[ 85 ]

of the region and is still spoken along the Rio Negro basin, including the Uaupés and Içana Rivers. According to Rodrigues (1996), a third Língua Geral developed in Paraguay, which he called Guarani Criollo (Creole Guarani); it was locally known as avañe’en ‘men’s language’ (avá ‘man’ + ñe’en ‘language’). I will not discuss this language here. Suffice it to add that it became the general language of the mestizo population as well as a lingua franca of interethnic communication. Modern varieties of it, highly influenced by Spanish, received the name Jopará (mixture); it is not the same language as the basilectal varieties of Guarani still found in some pockets of rural Paraguay. Together, Guarani and Jopará are the languages of about 60 percent of the Paraguayan people. They are still very similar to other Tupi-Guarani language varieties, including the Ancient Tupi that was once spoken in São Paulo and neighboring areas. Meliá (1974) offers an interesting interpretation of Jopará. Lustig (1996) provides a more recent overview of it, including many bibliographic references. The topic of Língua Geral in Brazil is not yet exhausted. Rodrigues (1983) says that there were also a Nagô (Yoruba) Língua Geral in the state of Bahia and a Kimbundu or Congolese Língua Geral north and south of Bahia. Finally, there was also an Ewe-based Língua Geral in the region of the gold mines of Minas Gerais, for which we have a vocabulary and a few grammatical notes written in 1741 and published in 1945 (Peixoto 1945). Rodrigues (2003) critically evaluates this work, while Souza (2003) analyzes its structure. However, as Rodrigues (2003) himself points out, this is an undue extension of the expression língua geral. In the linguistics literature, Língua Geral generally refers only to what Rodrigues (1986, 1996) called Língua Geral Paulista and Língua Geral Amazônica, the latter being also known as Nheengatu. The expression is not a synonym for lingua franca, nor is it a pidgin or a creole (see note 9); it is just the name of two specific lingua francas of Tupi and Tupinambá origin, respectively. The former disappeared, whereas the latter is still spoken in the Amazon region (see also Denny Moore, chap. 4, this volume).

6. Língua Geral Paulista It is important to emphasize that this Língua Geral is a restructured variety of the Tupi language that was spoken in São Vicente, Piratininga, and along the Tietê River in the state of São Paulo. In order to distinguish it from other varieties of Língua Geral, Rodrigues (2003) called it Língua Geral Paulista (LGP), that is, “Língua Geral of São Paulo” (in reference to the state, not the city). According to this author, only single

[ 86 ]

chapter three

men came to the province in the first half of the sixteenth century. In this scenario, not only local wives spoke Tupi but also their mestizo children and, to some extent, their Portuguese husbands. This situation remained unchanged for many of the following years. In 1694, a Jesuit priest said, “The language spoken in these families is the Amerindian one, whereas Portuguese is learned only at school” (Rodrigues 1996, 8). The original Tupi population was exterminated little by little, survived basically by mestizo (Mameluco) families, who spoke the restructured Tupi. Eventually the children and their fathers became bilingual in Portuguese and Tupi, the former more than the latter. This Tupi was the language more of the emergent mixed population than of the Amerindians. Through the intermediacy of the bandeirantes, LGP was spread to the present states of Minas Gerais, Goiás, Mato Grosso, and Mato Grosso do Sul, as well as to the north of Paraná. For this reason, we can find several Tupi place names and names for specimens of fauna and flora even where Tupi or LGP was never spoken. Unfortunately, we do not have many records of LGP. What we do have is a small dictionary of verbs, a still smaller list of words as well as some remarks made by foreign visitors to the region in those times, and the marks LGP left on Portuguese, which I discuss below. The dictionary, a manuscript probably from the eighteenth century, was published by Von Martius, in Glossaria linguarum brasiliensium (1867), under the name of Tupi Austral. The small list of words was found in the twentieth century in a community of Bororo Indians, blacks, and mestizos near the town of Araguari, in the present state of Minas Gerais. Tupi underwent several changes to serve the new society. The changes were more visible at the level of vocabulary. For example, the word Tupã, originally used to denote the genius of thunder and lightning, was used by the Jesuits to translate the Portuguese Deus ‘God’. Nossa Senhora ‘Our Lady’ was translated as Tupã Sï ‘Tupã’s mother’. The word ïbaka meant ‘firmament’, but the Portuguese missionaries used it to designate the Christian céu ‘heaven’. In some cases, the adjective eté ‘true’ had to be added to some words whose original Tupi meaning had changed. Here are some examples: Original  word cauim

Original

 meaning

Indian maize  wine

European

 meaning Wine

jaguara

ounce (animal)

Dog

tapiira

anta, tapir

Ox

Recovered

 original meaning cauim-eté ‘true

 cauim’

jaguar-eté ‘true  jaguar’

tapiir-eté ‘true tapir’

Hildo Honório Do Couto

[ 87 ]

The lexical items that Tupi borrowed from Portuguese include kabaru (< cavalo) ‘horse’, kurusa (< cruz) ‘cross’, and kabara (< cabra) ‘goat’ (feminine), as well as proper names like Peró (< Pedro) ‘Peter’, Panacicu (< Francisco) ‘Francis’, and Dui (< Luis) ‘Louis’. All borrowings were adapted to Tupi syllable structures, which are primarily CV and CVC. Tupi did not have any abstract words that matched the needs of the missionaries. To meet their communicative needs, the missionaries resorted to concrete Tupi words and shifted their meaning to cover the desired European abstraction, though following extant Tupi patterns. For example, obá-juba ‘yellow face’ > ‘frightened’, apïsá-bïra ‘erect ear’ > ‘attentive’, nambi-bebé ‘flying ear’ > ‘rapid’, and pï-atã ‘hard foot’ > ‘strong’. In other cases, there were semantic changes. For day, month, and year, the words ara ‘light’, jasï ‘moon’, and akaju ‘cashew’ were used. The reason for the latter example is that the cashew grows only once in a year. Even some not-so-abstract concepts were built using Tupi resources. Thus, tesá ‘eye’ plus ï ‘water’ became tesaï ‘tear’; kam ‘breast’ plus ï gave kambï ‘milk’ (lit., ‘breast’s water’). To designate objects unknown to the original Tupi culture, still other words changed their meanings metaphorically or metonymically: oka ‘hut’ plus pukú ‘long’ gave oka pukú, which was used to designate ‘street’, and tabusú (taba ‘hamlet’ + usú ‘large’) denoted ‘town’. The population that spoke LGP used many nicknames of Tupi origin, even for people born in Europe and bearing a Portuguese name. Here is a short list of names followed by the nicknames: Domingos Leme da Silva = o Botuca (the Botuca)

Francisco Dias de Siqueira = o Apuçá (the Apuçá)

Gaspar Vaz da Cunha = o Jaguaretê (the Jaguaretê)

Francisco Ramalho = o Tamarutaca (the Tamarutaca)

Salvador de Oliveira Leme = o Sarutaiá (the Sarutaiá)

Even the governor António da Silva Caldeira Pimentel had a nickname, Casacuçu ‘big coat’, because he always wore a coat (casaco in Portuguese). At least one example of Portuguese nicknames has been recorded, namely, Perna-de-Pau ‘peg leg’. By the late eighteenth and early nineteenth centuries, these nicknames were becoming more frequent. While the use of Tupi nicknames was declining, there was an emergence of Portuguese alternatives. For instance, Via-Sacra, Ruivo ‘red-haired’, Orador ‘speaker’, Cabeça do Brasil ‘head of Brazil’, and Pai da Pátria ‘father of the homeland’. During the transition between the two periods, there was a combination of Portuguese names and Tupi morphemes, for example, Mecia Fernandes (female), was known as Meciaçu ‘big Mecia’ and Pedro Vaz de Barros as Pedro Vaz Guaçu ‘Big Pedro Vaz’.

[ 88 ]

chapter three

LGP disappeared by the end of the eighteenth century, leaving many traces not only in local Portuguese but also in Brazilian Portuguese in general. These traces were mainly lexical, thousands of them, in the domains of flora, fauna, and place names. Many of them are still in use, for example, ibirá-pitanga (later pau-brasil ‘brazilwood’), and capivara (a big rodent), Paraná ‘sea, lake, river’, the name of a Brazilian state and of a river. There are a few others, such as capoeira (< kaa ‘bush, forest’ + puera ‘extinct’, namely, ‘hencoop’, later, name of a foot-fighting dance), toca (< t-oka ‘the house’, i.e., ‘small animal’s burrow’), xará (< xe rera ‘my name’, i.e., somebody’s homonym), peteca (< petek ‘to beat with the palm’, i.e., ‘shuttlecock’). The Tupi/LGP adjective mirim ‘small, little’ entered everyday Portuguese in constructions like secretário/a-mirim ‘little boy/girl serving as a secretary’ and guarda-mirim ‘little boy/girl serving as a guard’. The opposite of mirim, namely, (g)uaçu/uçu ‘large, big, great’, was not borrowed. However, it occurs in several names for flora, fauna, and toponymy, as in imbiruçu ‘tree whose bark is used to make ropes’, jararacuçu (jararaca ‘a type of snake’) ‘a big jararaca’, Mogi-Guaçu, a town near São Paulo City, so named because there was a smaller one called Mogi (< mboi ‘snake’ + ï ‘water, river’) ‘river of the snake’. Sampaio (1955) supplied one of the first inventories of terms for fauna, flora, and place names of Tupi/LGP origin, whereas Cunha (1978) attempted to develop a general inventory of words from this origin. As I have tried to show (Couto 2000), in present-day Brazilian Portuguese there are words like natureba (a person who tries to live according to the principles of nature) and decoreba in place of de cor ‘by heart’. Nowadays, we can see words with this suffix appearing from time to time. All of them have a mildly pejorative connotation. My thesis is that this comes from Língua Geral. Finally, I would like to remark that at the same time LGP was fading away, the so-called Caipira Dialect, a very basilectal variety of local rural Portuguese, was emerging. Furthermore, it was emerging roughly in the same region where LGP had been used. However, it would not be correct to say that the dialect is an offspring of LGP. The main reason for this is that its most conspicuous features are roughly the same as those of other rural dialects of Brazilian Portuguese spoken where LGP had never been known.4

7. Língua Geral Amazônica, or Nheengatu As noted above, there is a second variety of Língua Geral, which developed in the states of Maranhão, Pará, and Amazonas, around the begin-

Hildo Honório Do Couto

[ 89 ]

ning of the second half of the seventeenth century, after the expulsion of the French, who dominated São Luís (capital of the state of Maranhão) from 1611 to 1614 (see also Moore and Lee, chaps. 4 and 5, this volume). Tupinambá was the main language of this region, reaching the mouth of Tocantins River. As a result of the intermarriages of Portuguese men and Tupinambá women, a mestizo population arose, whose children usually spoke the mother’s language. The consequence was the emergence of a mixed language variety that began to differ from the original ethnic Tupinambá. This scenario is very similar to that in which LGP emerged, which is why it received the same name, Língua Geral, in addition to Brasiliano ‘Brazilian’, Língua Brasílica ‘Brazilian Language’ and Tupi Colonial ‘Colonial Tupi’.5 In the second half of the nineteenth century, it received the name Nheengatu (nheen ‘language’ + katu ‘good’), which is very common nowadays. In Colombia and Venezuela it is called Lengua Yeral. To distinguish it from the southern Língua Geral, based on Tupi proper, Rodrigues (1986) called it Língua Geral Amazônica ‘Amazonian Língua Geral’ (LGA). As explained in section 5, this language was spread all along the valley of the Amazon River by missionaries and troops, going on to the Amazon’s tributaries Solimões and Rio Negro and from the latter to the Içana and Uaupés Rivers.6 On the Solimões (called Marañón in Colombia), it was used at least in the town of Tefé. In the middle Amazon, it was very common in the town of Itacoatiara.7 According to some investigators, on the Xié River it replaced the Uerequena language, while in some places on the Içana River it was replaced by Baniwa. Nowadays, Nheengatu has practically disappeared in Maranhão and Pará as well as along the Lower and Middle Amazon River. Its use is now restricted to the Upper Rio Negro, around the town of São Gabriel da Cachoeira. It is possible that there are still some old speakers along the Madeira, Solimões, and Mawé Rivers. On the Lower Içana River, Nheengatu tended to replace native Baniwa. However, even in Upper Rio Negro its use is continually declining. In the late eighteenth century, the use of Nheengatu (or of any other local language) was forbidden by the Portuguese administrator the Marquis of Pombal, who also threw the Jesuits out. This prohibition included the use of Nheengatu names, which was very common in those times, for towns and peoples. All Nheengatu names of the time had to be replaced with Portuguese ones. For this reason, today there are towns named after Portuguese places such as Santarém, Alenquer, and Óbidos. However, the language did not disappear. During the Revolução da Cabanagem ‘Cabanagem Revolution’ (1834–40) against the political and economic domination of the region by the Portuguese, Nheengatu became the symbol of

[ 90 ]

chapter three

the insurgents, made up principally of Caboclos and Tapuias, all speakers of this lingua franca. A consequence of their defeat was the strengthening of European occupation, leading to a depopulation of the region. To overcome this problem, the authorities stimulated the immigration of people from the dry and poor northeastern part of Brazil. The presence of these immigrants was decisive in the decline of Nheengatu, which was pushed upstream on Rio Negro, so that today it survives mainly around the town of São Gabriel da Cachoeira, on the Upper Rio Negro, influencing and being influenced by Baniwa and Tukano, besides Portuguese. Inside neighboring Colombia, Venezuela, and Peru, we can find some evidence of it. Rodrigues (1996) presents some LGA (and LGP) characteristics that set them apart from Tupinambá/Tupi. Included are the preference for the SVO sentence constituent order instead of the original SOV; the disappearance of the argumentative, the predicative, and other cases; a restriction of deictic distinctions; and a reduced use of the evidential particles. As suggested above, the morpheme etá ‘many’ became a plural marker, as in jaguar + etá ‘jaguars’. Numerals above four were borrowed from Portuguese, alongside some conjunctions. At the phonological level, there are several cases of restructuring. The prenasalized stops tended to be replaced with the respective nasal consonants, so that mbaé ‘thing’ became maé and nde ‘you 1pl’ became ne, although there were also changes that favored the retention of the stop over that of the nasal consonant, as when mboi ‘snake’ became búia. The unrounded central high vowel [ï] was replaced by [i].8 However, the Portuguese superstrate influence is more visible at the level of the vocabulary. Examples include páia (< pai) ‘father’, surára (< soldado) ‘soldier’, pudéi (< poder) ‘power’, convidá (< convidar) ‘to invite’, viagem ‘trip, travel’, festa ‘feast’, and bicho ‘beast, animal’. Of course, all names of technological products are also borrowed from Portuguese. There were also cultural influences of Portuguese on Nheengatu. Since no Amerindian language had had a specific word used exclusively for year, the word akajú was used for it because the cashew grows only once in a year. For month the word jasï ‘moon’ was used, whereas day was rendered as ára ‘sun, light’, exactly as in LGP. All Christian terms had to be adapted. For example, church became Tupã-oka ‘house of God’ and angel was rendered as karaí-bebé ‘flying white’. When the missionaries tried to translate parts of the Bible, they were confronted with a series of problems because they had to find ways of expressing all Christian concepts in LGA, and in other languages, for that matter. Nheengatu did not have the same impact on Brazilian Portuguese as did Tupi or LGP in the South. However, in Amazonian Portuguese there are some vocabulary items whose origins may be traced to it. For ex-

Hildo Honório Do Couto

[ 91 ]

ample, the suffix -rana ‘similar, pseudo’ is attested in words such as brancarana ‘pseudo-white’ (< branco ‘white’), tupinambarana ‘Tupinambás that migrated to the island of the same name’, and campinarana ‘prairie that begins to be filled with weed’ (< campina ‘prairie’), as well as some names of plants and animals typical of the region. In earlier times, there were roughly the same naming strategies we saw above in relation to LGP. For example, an individual was known as Paurú Membéca, in Portuguese Paulo Mole ‘Soft Paul’. Another nickname was Paxikú Pukú (Francisco Comprido) ‘Long Francis’. A third example is Peró Guasu (Pedro Grande) ‘Big Peter’. These names were recorded in the town of Itacoatiara by the late eighteenth century. Nheengatu has also functioned as the native language of some groups, such as the Baré. We can therefore ask whether it does not qualify as a creole language. As a matter of fact, some creolists claim that pidgins and creoles emerged in a similar scenario, that is, out of the contact of mutually unintelligible languages that led to a sort of mixed variety called pidgin, which putatively becomes a creole after it is acquired as a native language by children, or, according to other creolists, once it becomes the primary vernacular of a community.9 The evolution in the use of nicknames in LGP mentioned above would follow the pidginization > creolization > decreolization cycle that some creolists claim obtains in the evolution of creole languages. However, we have no convincing reason for considering either LGP or LGA as a pidgin or a creole. To put it simply, they are simply contact-based varieties of southern Tupi and of northern Tupinambá, respectively, that became lingua francas for speakers of diverse languages. Because Nheengatu is fundamentally linked to the occupation of Amazonia, speakers of other Amerindian languages were, until the early twentieth century, encouraged to use it.10 From the language of a dominated people (the Tupinambás) it became an instrument of colonization through the missionaries, colonists, and others. In fact, it became an emblem of European occupation. For Amerindians it was the language of the colonizers. For the latter it was considered a mixed Tapuia language or nheengaíba ‘bad language’ used by local peoples. In spite of the steady decline in its use, Nheengatu still has a considerable presence in the Amazon region. It was the most used language during a meeting of Indian teachers in Manaus in 2006, where the participants were speakers of several mutually unintelligible languages, such as Cambeba, Tikuna, and Apurinã. Finally, it is noteworthy that Freire (2000) established five periods in the existence of LGA/Nheengatu. The first is the period of the interpreters, during the sixteenth century; the second is that of its implantation,

[ 92 ]

chapter three

from 1616 through 1686; the third involves its expansion, which goes from 1686 to 1757; the fourth corresponds to its Lusitanisation, from 1757 to 1850; and the fifth spans the period from 1850 to the present. This latter period coincides with the hegemony of the Portuguese language, which has reduced its importance.

8. Amerindian Language Islands I: The Parque Indígena do Xingu The Parque Indígena do Xingu (Indian Reservation of the Xingu River) is the most conspicuous conglomerate of Amerindian groups in Brazil (see map 3.1). It was officially created in 1961, to protect the fauna and flora as well as the indigenous populations against the expansion fronts that were encouraged by the foundation of Brasília, the capital of Brazil, a little earlier (1959). The Parque Indígena do Xingu (PIX) is in the northern part of the state of Mato Grosso, along the upper part of the Xingu River, including most of its affluents (Seki 1993; Emmerich 1992). PIX in fact constitutes a “language archipelago,” consisting of various LI’s. Although many of the languages spoken in PIX belong to different families and stocks, they share a considerable amount of their culture, at least in the southern part of PIX. This has been true since before the arrival of the whites and the foundation of PIX. However, each group considers its own language “better” and “more beautiful” than those of the other groups. Many individuals stick to their own language even in intergroup interactions, mixing it with gestures and mute bartering. This situation has favored the emergence of Portuguese as a lingua franca. Emmerich (1992) mentions seven Amerindian groups in PIX. Seki (1993) claims there are seventeen of them. This difference in numbers reflects the fact that groups were added (e.g., the Txikão in 1967 and the Tapayuna in 1969–70), either by moving some of them to the area or by enlarging the reservation’s domain. The ethnic groups now living in PIX are the following: Tupi stock: Kamayurá and Kayabi (Tupi-Guarani family); Juruna (Juruna family); Aweti (Aweti family) Aruak family: Waurá, Yawalapiti, and Mehinaku Karib family: Kuikuro, Kalapalo, Matypu, Nahukwá, and Txikão Jê family: Suyá, Tapayuna, Mentuktire, and Panará Isolated language: Trumai In 1990 the total number of inhabitants of PIX was 3,154. Table 3.1 indicates the numbers by ethnic group, in descending order.

Hildo Honório Do Couto

[ 93 ]

Kayapó

RR

AP

AM

PA

Kayapó

MA

CE RN PB PE AL SE BA

PI AC

TO

RO

T.I CAPOTO/JARINA

MT

GU

GO

DF

MG

ES

MS

XIN

SP

RJ

PR SC

RIO

RS

Kayapó Yudát Kayabi Yudja Kayabi Yudá Kayabi

LEGENDA Aldeias/línguas Rios Principais

Kayabi Kisêdjê Kayabi T.I PARQUE DO XINGU Kayabi Kayabi Kisêdjê Kayabi Kayabi

Kayabi

Kayabi

Kisêdjê

10000

0

10000

20000

30000

metros

Escala: 1:1.250.000

Projeção Universal Transversa de Mercador Datum: SAD_69 Zona:22

Kayabi Ikpeng

T.I WAWI Trumai Trumai Kamaiurá

Kisêdjê

Yawalapiti Kamaiurá Nahukuá Steinen Steinen Kalapalo Waurá Matipu Yawalapiti Kuikuro Aweti Mehinaku

Kalapalo Kalapalo

Kuikuru Kalapalo

Map 3.1. Languages of the Upper Xingu River. Legend: Aldeias/Línguas = Villages/ Languages; Rios Principais = Main Rivers.

Yawalapiti was already obsolescing by the late nineteenth century. Its remaining speakers lived among the Kuikuro, the Waurá, the Mehinaku, the Kamayurá, and the Aweti. With the “help of the Villas-Boas brothers, the group was restructured as a tribal unit through the incorporation of people from other ethnic groups as well as marriage with Kamayurá and Kuikuro women and men” (Emmerich 1984, 24). For this reason, at least three languages are spoken in the Yawalapiti village, Kamayurá, Kuikuro, and Yawalapiti, besides the language of the Caraíba ‘white man’, that is,

[ 94 ]

chapter three

Table 3.1. Population of PIX Ethnic group

Men

Women

Total

Kayabi Mentuktire Kamayurá Kuikuro Kalapalo Waurá Suyá Txikão Yawalapiti Juruna Mehinaku Panará Trumai Matypu/Nahukwá Aweti Tapayuna Total

272 221 142 153 129 99 99 84 75 66 77 56 54 50 38 22

253 228 145 124 120 88 66 62 70 65 51 66 55 52 45 27

525 449 287 277 249 187 165 146 145 131 128 122 109 102 83 49 3,154

Portuguese. Thus, it is an LI that was devastated, a bit later recomposed itself, and is undergoing a new flood. Yawalapiti is likely to die much sooner than most of the other PIX languages, because, according to some investigators, it has the lowest number of monolingual speakers. The Trumai live in two villages. In 1952 the group included only 18 persons. In the early 1990s, however, their number increased to 109, although not all of them were truly Trumai. In one of the villages, there was only one person that could be considered monolingual, an old woman whose parents were both Trumai. Except for her and one child, everybody knows more than one language. In mixed marriages, men and women speak Portuguese very frequently. The young are forgetting Trumai and learning Portuguese. According to Emmerich (1984), Portuguese is the lingua franca within PIX. In principle, this would not be a problem. However, it is being learned subtractively, not additively (Skutnabb-Kangas 2001, 400). It is gaining ground not only among the Yawalapiti and the Trumai but also among the other groups. Moreover, some groups are so small that their survival is greatly in peril. This is the case, above all, for the Aweti and the Tapayuna. Nevertheless, each people group uses its own language, except for the Trumai and the Yawalapiti. The former is one of the most affected by the presence of Portuguese, because the Trumai village lies near the headquarters of PIX administration. The Yawalapiti’s language is the most endangered of all of them.

Hildo Honório Do Couto

[ 95 ]

Many groups have already lost both their languages and their territories altogether. For example, the Tsuva have been assimilated by the Kuikuro, the Naravúte by the Kalapalo, and the Nahukwá by the Matypu. Some groups no longer exist as distinct groups. Their members have dispersed into other groups, as in the case of the Kustenau, whose survivors live with the Waurá, the Anumaniá, whose last members live with the Mehinaku, and the Manitsawá, whose few remaining individuals live among the Aweti and the Suyá. As noted above, Portuguese functions as the lingua franca in CaraíbaIndian interactions and frequently also in Indian-Indian communication. Consequently, some languages of this linguistic archipelago are moribund, whereas others are just endangered. However, because of the increasing spread of Portuguese, no PIX language is in a comfortable position. All of them will certainly have the same fate, extinction, sooner or later.

9. Amerindian Language Islands II: The Upper Rio Negro Region Another region with an archipelago of LI’s is Upper Rio Negro, in the northeastern part of the state of Amazonas, which is probably the most multilingual in Brazil (see map 3.2). It comprises not only Rio Negro but also some of its affluents, such as the Içana and the Uaupés, as well as the Uaupés affluents: the Tiquié and the Papuri. Rio Negro proper is the river that, together with the Solimões, forms the Amazon River when they meet near Manaus. The Amazon flows from here downstream to its mouth near Belém and the huge island of Marajó, which is larger than Switzerland. There are three language families in the Upper Rio Negro region: Eastern Tukano,11 Aruak, and Maku. Eastern Tukano is composed of the languages Tukano (formerly called Betoya), Desana, Tuyuka, Arapasu (extinct), and Wanana, besides those spoken in Colombia. As for the Aruak family, it comprises the languages Baniwa, Koripako, Baré, Warekena, and Tariana. Baniwa is spoken mainly on the Içana River. It is split into several dialects, some of which have specific names, for instance, Karutana for those living on the Lower Içana, and Koripaka (Koripako) for the group on the Upper Içana. Baré and Werekena are predominant along the Xié River. The languages of the Maku family are spoken by groups living in the area surrounded by the Uaupés, Negro, and Japurá Rivers. There are at least six groups of this family in Brazil. They are Bará, Hupda, Yahup (Yaho), Nadëb (Nadob), Kumã, and Guariba. Yahup and Hupda are very similar. Perhaps they could be considered two dialects of the same language. The Guariba are referred to in Nheengatu by the name Wariwa-

Map 3.2. Upper and Mid Negro River. Legend: Famílias Linguísticas = Language Families; Tukano Oriental = Oriental Tukano: Tukano, Desana, Tuyuka, Wanana; Aruak: . . . ; Língua Geral: Nheengatu (language names); Predomínio do Nheengatu de S. Gabriel da Cacheira e Cucul = Predominance/supremacy of Nheengatu of São Gabriel da Cachoeira and Cucul; Terras Indígenas = Indian/Amerindian Lands; Rios = Rivers.

Hildo Honório Do Couto

[ 97 ]

Tapuya ‘Tapuya people’. According to the 2000 census, in the entire region of Upper Rio Negro and environs there are 31,625 Amerindians in Brazil, 26,281 in Colombia, and 7,290 in Venezuela. Some investigators include in the languages of this region some of the Yanomami family, formerly known as Xiriana. However, their speakers live somewhere east and northeast of it, mainly in the states of Roraima and Amazonas as well as in Venezuela. There is a trend toward urbanization; 70 percent of the Amerindians of the region live in towns. Even some indigenous villages, for example Iauaretê, look like towns. The biggest urban center is São Gabriel da Cachoeira, with its ten thousand inhabitants. Of these, 95 percent are Indians, mostly Tukano. There are three Indian neighborhoods, Dabaru, Areal, and Tirirical. In these neighborhoods, each individual keeps his paternal language, even though the neighborhoods are multilingual. In Areal, most of them speak Tukano, even if they belong to another ethnic group. Younger people also speak Portuguese and, sometimes, Nheengatu. Others speak Baniwa. One Tuyuka man is Tukanized because he married a Tukano woman and speaks her language at home. A Desano also speaks Tukano, whereas one Caboclo speaks Portuguese, and his neighbor (a Karapanã) speaks Spanish. There are three lingua francas in this town, Nheengatu, Tukano, and Baniwa. In accordance with Law 145, dated December 11, 2002, these three languages have been declared official in the municipality of São Gabriel da Cachoeira, in addition to Portuguese, which keeps spreading its domain. So the practical effect of this law has been nil. The migrations toward the towns are prompted by the search for jobs and education. This is the common cause of the loss of ancestral languages. Unfortunately, some parents are forcing their children to learn Portuguese because it is the symbol of progress, whereas the ethnic languages are associated with “backwardness.” The lingua francas have a wider domain, however. The Maku speakers of the Uaupés region use Tukano as a lingua franca, so that they are at least bilingual. In the indigenous land of Balaio, the Koripako speak the language of the Baniwa. The Arapasu and the Tariana have shifted to Tukano, which is the lingua franca of the Uaupés, Tiquié, and Papuri Rivers, although there still are some people who know their ancestral languages. Some investigators say that there has been a process of Tukanization in this region, whereas the Tukano who moved to Middle Rio Negro have adopted Nheengatu. This Tukanization process has worked as a buffer against Nheengatu, and against Portuguese for that matter. The Upper Rio Negro region is largely multilingual, in the sense of territorial multilingualism. However, there is also individual multilingual-

[ 98 ]

chapter three

ism. Further, owing to the practice of linguistic exogamy, many individuals speak their fathers’ and mothers’ languages, in addition to Tukano, Nheengatu, Baniwa, or Portuguese, or even Spanish in some cases. Almost everybody is at least bilingual or trilingual. Many are real polyglots. In this multilingualism and polyglotism, the ethnic languages, typically the fathers’, are identity markers. Interestingly, despite the existence of twenty-three ethnic groups (twenty-seven if we include those of Venezuela and Colombia), there is a cultural unity, not only in terms of exchanges of goods but also in terms of material and spiritual culture as well as worldview, exactly as with the groups of the PIX discussed in the previous section.12

10. Amerindian Language Islands III: Moribund Languages The expansion of the Portuguese language and European culture in Brazil eroded the structures of Amerindians’ languages, drove them to the condition of LI’s, and also undermined the survival of their speakers as separate groups with their own cultural identities. In other words, most Amerindian languages are now endangered, and many of them are moribund. I consider a language moribund when there are only a handful of people, usually elderly, who use the language rarely in everyday communication with fellow members of the Community. All such speakers have an L2 as their vernacular or primary language. The Laboratory of Amerindian Languages of the University of Brasília has studied at least seven moribund languages in the above sense. The first is Kokáma, in the Western Amazonian region, probably of the TupiGuarani family. By the end of the 1990s, it had only eleven speakers, all of old age, along the Solimões River. According to Cabral (2003), this language shows characteristics of creoleness, perhaps having evolved from a Tupi-Guarani language, but its contact history has not been documented. The Kanoê language, a member of an isolated family, was spoken by four to five people in 1990. They lived in the towns of Guaporé and GuajaráMirim, also in the Amazonian region. Some languages have a more comfortable situation, if we consider the number of people who still speak them. They include Mehinaku, of the Aruak family. However this ethnic group (of PIX) has only about 150 speakers. Another group is Yuhup, of the Maku family, located on Upper Japurá and Upper Rio Negro. The number of its speakers is about 400. However, pressure from Nheengatu, Tukano, and Portuguese is continually eroding the vitality of this language. In other words, although these languages are not moribund, they are evolving toward moribundity.

Hildo Honório Do Couto

[ 99 ]

There are many other moribund languages in Brazil, such as Krenak, in the state of Minas Gerais. It is still spoken by some speakers, but there are so few of them, and the language is so rarely spoken, that its future is highly uncertain. Among the languages that are not yet moribund but are on the way to moribundity we could add the following: Trumai, isolated family (PIX), fewer than 40 speakers Mynky, isolated, Upper Juruena river, Mato Grosso, fewer than 30 speakers Koaiá, isolated, whose last speakers live with the Aikaná group, in the southeast of the state of Rondônia (they lost their Fundamental Ecosystem of Language [FEL]) Jabuti, or Djeoromitxi, whose last 40 members live with the Makurap, at the Posto Indígena Guaporé, near Guajará-Mirim In summary, it would not be far-fetched to say that all surviving Amerindian languages in Brazil are endangered. Many of them are already dying. One of the causes is that they became LI’s after the arrival of the European colonizers.

11. Amerindian Language Islands IV: Dead Languages Moribundity is the penultimate stage in the obsolescence process. Language death is the last stage. If language is how members of the Community (FEL) communicate among themselves, we may conclude that a language is dead when it has fewer than two speakers. There must be at least two people for communication to take place. Furthermore, this is the minimum that an FEL (i.e., a minicommunity) may consist of. Languages may disappear through at least three processes. The first is the extermination of its speakers. The second happens when the speakers shift to another language, because of invasion or conquest or both. A third way to language death is the dispersal of speakers to several different places, so that they do not have the opportunity to talk to one another, again in accordance with FEL. The second and third processes begin when speakers stop using their language in some contexts, such as in public. As the number of domains increases in which the language is no longer used, the language dies. In the second and third processes, each individual ceases to use her original language as soon as she moves to a host Community/ FEL. That is, for a language to be considered alive, the mere existence of some individuals who know it is not enough. They must use it in everyday interaction. In fact, there are many individuals around the world who know Latin, Sanskrit, or Classical Arabic. However, these languages are

[ 100 ]

chapter three

considered dead, because they are not used in day-to-day interactions. In Brazil all of these processes obtained.13 The two most famous dead languages of Brazil are Tupi and Tupinambá. The first was replaced by LGP, which eventually died out. The second was replaced by LGA-Nheengatu, which is still spoken, especially in the Upper Rio Negro region. We could say that these languages underwent “death by transformation,” a category suggested by Salikoko Mufwene in comments on this chapter. A third dead language of the same period could be added: Old Guarani, which was spoken in Paraguay, parts of southern Brazil, Uruguay, and northern Argentina. As we saw in section 5, there are some remnants of Guarani, above all in Paraguay, where it changed to Jopará as a result of contact with Spanish. It also survived among some small groups, such as the Mbyas in Paraguay and Brazil. However, these present-day varieties are not the same language as Old Guarani, which is dead. One of the languages that disappeared through language shift in the Amazon is Baré. Its speakers have shifted completely to Nheengatu because of contact with the missionaries, although there still are some speakers in Venezuela. Interestingly, both groups have also forgotten that they once had their own ethnic language, of the Aruak family. They now claim that their ethnic language is Nheengatu. The Tariana (also of the Aruak family) of the Uaupés River have shifted to Tukano. The Tukano who migrated to Middle Rio Negro have shifted to Nheengatu, but they remember that the ancestral language was Tukano. Meader (1978) presents an overview of research conducted in the 1960s with fifteen groups in the northeastern part of Brazil that had lost practically everything of their language. One of the his general conclusions is that “none of these groups uses their ethnic language as a means of everyday communication. Few individuals are able to remember some words of their ancestors’ language” (8); that is, these languages were already dead. He asserts, “Despite their probable several flaws, the lists of words we are collecting here are probably the last linguistic information we will ever be able to have about these languages” (11). The Pataxó of the state of Bahia, who are now where the Portuguese encountered the Tupinambá in 1500, have also lost their language. When they were visited in the 1960s, they lived at the Posto Caramuru of the Service for Protection of the Indians, near the town of Itaju, Bahia. They were entirely dependent on the Posto, which gave them a weekly stipend for their survival. They did not need to work. Two of the older ones were genuinely Amerindians, namely, Raco (almost 100 years old) and Tçitçi’a (around age 50). Raco, though physically well, had lost a lot of his mental capacity. When he tried to tell a story, his Portuguese was unintelli-

Hildo Honório Do Couto

[ 101 ]

gible. He was unable to give a single word in his ancestors’ language. As for Tçitçi’a, the most active of his group, he remembered many isolated words, although he was unable to form phrases with them (Meader 1978, 17). In other words, not only their language is a bygone but also their own FEL, which has already fallen apart. Meader’s (1978) book contains a list of 163 words proffered by Tçitçi’a. As usual in such cases, it includes a lot of variants. Although the surviving vocabulary is small, there are some synonyms. For example, there are two words for ‘meat’ and for ‘hair’, four for ‘monkey’ (one of them designates ‘big monkey’), and not fewer than five for ‘wild animal’ (bicho). There are also many polysemous and polyfunctional items, as well as several circumlocutions, which is typical of incipient pidgins. In this case, dying languages are a kind of mirror-image of incipient languages. There is also the interesting case of the Xukuru, in the town of Pesqueira, Pernambuco. According to Lacerda (2001, 3), this population had been prohibited to practice their religion and to speak their own language. This led to the death of their language. However, they later “managed to retrieve around 150 lexical items, with which they tried to speak the language again,” interestingly following the structures of Portuguese grammar and phonology. However, a small list of words known by a few people (but not used in everyday communication) is not a language. The Xukuru language may thus be considered dead. These words are like linguistic fossils. A concomitant of language loss is loss of ethnic identity in this case, which happened because the Xukuru no longer constitute an FEL, since an FEL requires a P living in a T and speaking their own L. The L side of the triad is missing. There is another interesting case of death of a language by shift to another language, in the north of the state of Amapá. The group known as Karipuna originally spoke a language of the Tupi-Guarani family. They shifted to the language of neighboring Guiana, the French-lexifier creole known as Guyanais. Now they think this creole variety is their ethnic language.14 The list of dead languages in Brazil is a long one indeed. As noted above, there were about 1,200 languages spoken in Brazil when the Portuguese arrived in 1500. Today there are fewer than 180 languages left, amounting to a loss of more than two languages per year.

12. Concluding Remarks All the processes discussed in previous sections were conducive to the formation of LI’s and, eventually, to language death. It is true that some languages became lingua francas, such as Língua Geral Paulista, Língua Geral Amazônica, and, in the Rio Negro region, Tukano and Baniwa.

[ 102 ]

chapter three

One might think that lingua franca formation would go in the direction of language survival, not of language death. However, LGP disappeared in the early nineteenth century. LGA, or Nheengatu, is also disappearing even though it is still spoken by many Upper Rio Negro groups. The Tukano lingua franca will certainly disappear, as urbanization progresses and promotes communication in Portuguese. Thus, the indigenous lingua francas have had ephemeral lives. This means that the emergence and life of Língua Geral are not exceptional regarding language death. The language appears to have followed the normal trajectory under the circumstances of its practice. Indeed, Língua Geral Paulista and Língua Geral Amazônica/Nheengatu emerged because the Amerindian peoples who had spoken Língua Geral had to adapt to a new language ecology in which the colonizers and their European language prevailed. Thus, the Amerindians had to shift from their indigenous languages, at least in part in the case of the ethnic ones. The practice of Língua Geral was already a process moving away from ethnic languages, which eventually led to their extinction. This situation is anything but favorable to the vitality of the indigenous languages. As far back as the late nineteenth century, many Amerindians of several ethnic groups migrated to towns and cities, where they generally built a miniature of their original FEL, an evolution facilitated by a tendency they have to live in the same neighborhood, where mini-LI’s emerge. Even in the huge city of São Paulo, we can find some in-migrant Amerindians, although the numbers are unknown. The total number of Indians living in urban centers is estimated to be one hundred thousand, or onefourth of the whole Amerindian population in the country. As expected, these migrations are additional causes of the endangerment of the Amerindian languages, because the urban ecologies favor deculturation and usage of Portuguese. It would be no exaggeration to say that all remaining Amerindian languages in Brazil are reduced to the condition of LI’s surrounded everywhere by the present dominant language, Portuguese. Becoming LI’s certainly implied a weakening of their vitality, because the process meant a reduction of the land their speakers could use. Ironically, however, precisely this situation is now a way of retarding the death of their languages. In fact, in an urban LI, individuals of the same ethnic group are able to interact with one another in everyday life. This interaction favors the maintenance not only of the language but of the whole culture, enhancing the integrity of the group in question qua group. In other words, in the beginning, the condition of LI was bad for Amerindians because it was caused by the invasion of most of their lands, the T of their FEL. Since the geographical spread of Europeans has been steamroller-like, at

Hildo Honório Do Couto

[ 103 ]

present it appears that living in an LI is better than having some individuals scattered amid the towns and cities of the descendants of the invaders. Unfortunately, it also appears that even the remaining LI’s will disappear. That is, the best ecologies for maintaining Amerindian languages and cultures are undoubtedly (something close to) those that preceded Portuguese colonization. It is true that there are a few groups that have not yet been in contact with Europeans and other outsiders, such as the one that was recently seen from an airplane in the state of Acre. It is also true that the government is trying to demarcate the areas inhabited by some remaining groups.15 The Raposa do Sol Reservatrion in the state of Roraima is a case in point. However, many politicians (including the governor of the state) and agriculturalists argue that the demarcated area is too large, that it would be too much land for too few Indians. These politicians and agriculturalists seem not to remember that it is their ancestors that invaded the land of the Amerindians. Language death represents a great loss for linguists and for linguistics. Since South America is a big peninsula, it is relatively isolated from the rest of the world. The local languages show a series of features that are found only among them. Pirahã, spoken near the Madeira River in the Amazon, has the smallest inventory of phonemes of the world: six consonants and three vowels plus the glottal fricative. Among its phonemes, Pirahã has the sound [D] (a kind of lateral fricative), not found in any other language. It also has a voiced bilabial fricative, which is very rare. Maxakali of Minas Gerais is the only language to have vocalic variants for all consonants. As has been known since the colonial times, Tupi has an inclusive and an exclusive possessive pronoun in the first person plural. The Karitiána (Tupi family) of the state of Rondônia indicates negation in sentences through suppression of the morphemes of aspect and tense in the verb. Hixkaryana (Karib family) is an object-first language. And so on. All this wealth and diversity will eventually disappear, as a consequence of Iberian imperialism in Brazil. Notes 1. The expression “environment of language” is ambiguous, because sometimes it means “ecosystem of language.” We have seen that the former is part of the latter, which is formed by the totality of “language” and its “environment” together with their interrelationships. However, as in ecology, this ambiguity does not cause misunderstandings. The context makes it less ambiguous, so that we may use Fundamental Ecosystem of Language and Fundamental Environment of Language interchangeably, without affecting clarity. The same is true of social, mental, and natural ecosystems/ environments of language. However, there are instances in which it is necessary

[ 104 ]

chapter three

to distinguish them, as, for instance, when we are talking about words and their referents. In this case, Natural Environment of Language is not the same as Natural Ecosystem of Language, because it has to do with the things words refer to. 2. Mufwene says that “to the lay person the term language means something like ‘way of speaking.’ Thus English originally meant ‘the way the English people speak’ and kiSwahili ‘the way the waSwahili speak’” (2001, 1). In Latin, it was not possible to say “I speak Latin.” Instead, Romans had the phrase latine loqui ‘to talk like the Latins/Romans’. This interpretation of language is not totally new; Coseriu (1952) defends this particular thesis. 3. The concept of communion is reminiscent of Jakobson’s (1960) phatic function of language. Couto 2001 is a relatively detailed analysis of the interaction that took place between the Portuguese and the Amerindians. 4. Holanda 1979 is one of the best overviews of the history of LGP. 5. Freire 2004 lists more than fifteen different names by which it has been identified. 6. According to Taylor (1985), some time ago Nheengatu was also spoken in the towns of Barcelos, Tapurucuara, Ilha Grande, and Taracuá, among other localities. 7. The Jesuit Anselm von Eckart lived in Itacoatiara (then Aldeia de Abacaxis) from 1753 to 1759. He wrote Specimen linguae brasilicae vulgaris based on data collected there. The examples he gives are very similar to those of southern Tupi/LGP, and of Tupinambá, for that matter. 8. As a matter of fact, this sound was generally replaced either by [i] or by [u], especially in the South. The ending -tïba (‘collection/plantation of’) was rendered in Portuguese either as -tiba or -tuba, as in place names like Curitiba (capital of the state of Paraná) ‘plantation of kori [pine tree]’ and Ubatuba (coastal town of São Paulo state) ‘plantation of fruit [ïbá] trees’. 9. Some creolists say that Língua Geral is either a pidgin or a creole. For example, Hall says, albeit in passing, that “Língua Gêral” [sic] is a ‘Pidgin Tupí-Guaraní’” (1966, 4, 18). Smith states that “Lingua Geral Amazonica/Nheengatu (based on Tupinamba)” is “a creole based on the now extinct Tupinamba. Spoken in the Rio Negro valley by more than 3000 bilingual speakers.” He also says that “Lingua Geral Paulista/ Tupi Paulista (Brazil/Paraguay)” is “a Tupi-lexifier pidgin or creole of the São Paulo region. The last speakers died at the beginning of the 20th century” (1994, 363). Hancock claims that “Lingoa Gêral [Língua Geral] occasionally known as Ava’-neé or Tupïhaïa, is a rudimentary pidgin based on the Tupi-Guarani languages of South America, at one time in extensive use in coastal and inland Brazil” (1977, 387). Apparently, none of these three authors had direct contact with the language or its most authoritative bibliography. Probably all of them based their statements on secondhand sources. 10. Denny Moore says that “Nheengatu was discouraged after the mid-18th century and was receding after that” (pers.comm., Jan. 3, 2013), but he does not mention his sources. It has indeed been receding, like any other Amerindian language. However, this does not mean that it was (or is) a lingua franca for many speakers of other languages. 11. Eastern Tukano languages are also spoken in Colombia, Equador, and Peru. 12. Besides the works by Aryon Dall’Igna Rodrigues, there is much information on indigenous languages of this region and of Brazil at large at www.povosindigenas .org.br. 13. For a language to be considered alive, the existence of an FEL seems to be indispensable, even if it had died out previously and was resurrected later, as is claimed

Hildo Honório Do Couto

[ 105 ]

of Israeli Hebrew. Artificial languages like Esperanto also have several speakers dispersed in several countries who know it. However, it is not a natural language. Sign language apparently would be in the same situation, but only apparently. First, it was not fabricated by a polyglot like Ludwig L. Zamenhof, but emerged naturally out of domestic signs; second, its users use it in everyday acts of communication whenever they have the chance, and they frequently have this chance in their associations. 14. For a critical appraisal of this view of Karipuna’s history as well as its linguistic situation, see Röntgen (1998). By the way, Karipuna is the only creole language spoken in Brazil. 15. Many of the groups of PIX and Upper Rio Negro regions discussed previously fall into this category.

References Anchieta, José de. 1990. Arte de grammatica da língua mais usada na costa do Brasil. São Paulo: Loyola. Originally published in 1595. Cabral, Ana Suelly A. C. 2003. Evidências de crioulização abrupta em Kokáma? Papia 13:180–186. Caminha, Pero Vaz de. 1943. A Carta de Pero Vaz de Caminha. Ed. Jaime Cortesão. Rio de Janeiro: Livros de Portugal. Originally published in 1500. Chaudenson, Robert. 1989. Créoles et enseignement du français. Paris: L’Harmattan. Clyne, Michael. 1975. Forschungsbericht Sprachkontakt. Kronberg, Germany/Ts: Scriptor Verlag. Coseriu, Eugenio. 1952. Sistema, norma y habla. Revista de la Faculdad de Humanidades y Ciencias 9:113–181. Couto, Hildo Honório do. 2000. O sufixo -eba do português brasileiro e os conceitos de produtividade, gramática e língua. In O português brasileiro: pesquisas e projetos, ed. Sybille Grosse and Klaus Zimmermann, 427–450. Frankfurt: TFM Verlag. ———. 2001. A interação entre portugueses e ameríndios em Porto Seguro em 1500. Pesquisa lingüística 6, no. 2: 7–28. ———. 2007. Ecolinguística: Estudo das relações entre língua e meio ambiente. Brasília: Thesaurus Editora. ———. 2009. Ecologia, linguística e ecolinguística: Contato de línguas. São Paulo: Editora Contexto. Cunha, Antônio G. 1978. Dicionário das palavras portuguesas de origem tupi. São Paulo: Melhoramentos/Editora UnB. Emmerich, Chalotte. 1984. A língua de contato no Alto Xingu. PhD diss., Universidade Federal do Rio de Janeiro. ———. 1992. O português no Parque Indígena do Xingu, Mato Grosso, Brasil Central. Estudos linguísticos e literários 13:57–90. Fill, Alwin. 1993. Ökologie: Eine Einführung. Tübingen: Gunter Narr Verlag. Freire, José R. Bessa. 2000. Da ‘fala boa’ ao português na Amazônia brasileira. Amazônia em cadernos 6:1–45. Originally published in Amerindia 8 (1983). ———. 2004. A extensão da língua geral amazônica no século XIX e a política de línguas. Revista internacional de lingüística iberoamericana 1, no. 3: 9–22. Gândavo, Pero Magalhães. 1575. Historia da Provincia de Santa Cruz. Lisbon: Officina Antônio Gonçalves. Haiman, John. 1980. The iconicity of grammar: Isomorphism and motivation. Language 56, no. 3: 515–540.

[ 106 ]

chapter three

Hall, Robert A., Jr. 1966. Pidgin and creole languages. Ithaca, NY: Cornell University Press. Hancock, Ian F. 1977. Repertory of pidgin and creole languages. In Pidgin and creole linguistics, ed. Albert Valdman, 387. Bloomington: Indiana University Press. Haugen, Einar. 1972. The ecology of language. In The ecology of language, ed. A. Dil, 325–339. Stanford, CA: Stanford University Press. Holanda, Sérgio Buarque de. 1979. A língua geral em São Paulo. Appendix to Raízes do Brasil, 88–100. 13th ed. Rio de Janeiro: Livraria José Olympio. Originally published in 1936. Jakobson, Roman. 1960. Closing statements: Linguistics and poetics. In Style in Language, ed. Thomas A. Sebeok, 350–377. Cambridge, MA: MIT Press. Labov, William. 1977. On the use of the present to explain the past. In Linguistics at the crossroads, ed. Adam Makkai, Valerie B. Makkai, and Luigi Heilman, 226–261. Lake Bluff, IL: Jupiter Press. Lacerda, Rosely. 2001. Xukuru: Um povo à busca de sua língua? Paper read at the Encontro da Associação: Crioulos de Base Lexical Portuguesa e Espanhola, Coimbra, June 28–29. Lustig, Wolf. 1996. Mba ‘eichapa oiko la guarani? Guarani y jopara en el Paraguay. Papia 4, no. 2: 19–43. Mattheier, Klaus J. 1996. Methoden der Sprachinselforschung. In Kontaktlinguistik, ed. Hans Goebl, Peter H. Nelde, Starý Zdenek, and Wolfgang Wölck, 812–818. Berlin: Walter de Gruyter. Meader, Robert S. 1978. Índios do Nordeste. Série Lingüística no. 8. Brasília: Summer Institute of Linguistics. Meilá, Bartomeu. 1974. Hacia una ‘tercera lengua’ en el Paraguay. Estudios paraguayos 2, no. 2: 31–71. Mufwene, Salikoko S. 2001. The ecology of language evolution. Cambridge: Cambridge University Press. ———. 2005. Créoles, écologie sociale, évolution linguistique. Paris: L’Harmattan. Odum, Eugene P. 1971. Fundamentals of ecology. 3rd ed. Philadelphia: W. B. Saunders. Peixoto, António da Costa. 1945. Obra nova de lingua geral de Mina. Lisbon: Agéncia Geral das Colónias. Rodrigues, Aryon Dall’Igna. 1986. Línguas brasileiras: Para o conhecimento das línguas indígenas. São Paulo: Edições Loyola. ———. 1993. Línguas indígenas: 500 anos de descobertas e perdas. DELTA 9, no. 1: 83–103. ———. 1996. As línguas gerais sul-americanas. Papia 4, no. 2: 6–18. ———. 2003. Obra nova da língua geral de Mina: A língua ewe nas Minas Gerais. Papia 13:92–96. Rodrigues, José Honório. 1983. A vitória da língua portuguesa no Brasil colonial. Humanidades 1, no. 4: 21–41. Röntgen, Karl-Heinz. 1998. L’origine contestée d’une communauté créolophone: Les Karipuna au Brésil. Etudes créoles 21, no. 2: 36–64. Rosenberg, Peter. 2001. Deutsche Minderheiten in Lateinamerika. Europa-Universität Viadrina-Frankfurt/Oder, www.kuwi.euv-frankfurt-o.de/~sw1www/publikation/ lateinam.htm (accessed May 21, 2005). ———. 2003. Vergleichende Sprachinselforschung: Sprachwandel in deutschen Sprachinseln in Russland und Brasilien. Linguistik online 13, nos. 1–3: 273–323.

Hildo Honório Do Couto

[ 107 ]

Available at www.linguistik-online.de/13_01/rosenberg.pdf (accessed April 13, 2005). Salzinger, Kurt. 1979. Ecolinguistics: A radical behavior theory approach to language behavior. In Psycholinguistic reasearch: Implications and applications, ed. Doris Aaronson and Robert W. Rieber, 109–129. New York: Erlbaum. Sampaio, Teodoro. 1955. O tupi na geografia nacional. 4th ed. Salvador: Câmara Municipal. Reprint, with the title Vocabulário geográfico brasileiro (São Paulo: São Paulo University Press, 1970). Seki, Lucy. 1993. Notas sobre a história e a situação lingüística dos povos indígenas do Parque do Xingu. In Lingüística indígena e educação na América Latina, ed. Lucy Seki, 89–117. Campinas, Brazil: Editora da UNICAMP. Skutnabb-Kangas, Tove. 2001. Linguistic human rights in education for language maintenance. In On biocultural diversity: Linking language, knowledge, and the environment, ed. Luísa Maffi, 397–411. Washington, DC: Smithsonian Institution Press. Smith, Norval 1994. An annotated list of creoles, pidgins, and mixed languages. In Pidgins and creoles, ed. Jacques Arends, Pieter Muysken, and Norval Smith, 363. Amsterdam: John Benjamins. Souza, Silvia M. C. 2003. A predicação na língua geral de Mina. Papia 13:97–106. Taylor, Gerald. 1985. Apontamentos sobre o Nheengatu falado no rio Negro, Brasil. Amerindia 10:5–23. Weinreich, Uriel. 1953. Languages in contact: Findings and problems. The Hague: Mouton. Wiesinger, Peter. 1973. Deutsche Sprachinseln in Mitteleuropa. In Lexikon der germanistischen Linguistik, ed. H. P. Althaus, H. Nenne, and H. E. Wiegand, 367–77. Tübingen: Max Niemeyer Verlag.

4 * Historical Development of Nheengatu (Língua Geral Amazônica) denny moore

1. Introduction One of the most fascinating cases of language contact, the evolution of Nheengatu, spanning five hundred years in Brazil, has received relatively little international attention. The indigenous language commonly spoken on the coast of Brazil in 1500, the precursor of Nheengatu, was widely learned in the colony in the sixteenth century, abetted by Jesuit use as an instrument of catechism and of control of native peoples and by favorable government opinion. It was called the Brasílica in the first two centuries, or Tupinambá, the name of one of the indigenous nations that spoke it, or coastal Tupi. A version of the language was prevalent in the seventeenth and eighteenth centuries in the Portuguese Amazonian colonies of Maranhão and Grão-Pará, although it declined in official favor and importance after the mid-eighteenth century. It evolved rapidly as a result of historical events and language contact. By the mid-eighteenth century, the colloquial variety was recognized as different and referred to as Língua Geral or, as designated by Lee (2005), the Vulgar. The term Língua Geral Amazônica (LGA) is used to distinguish the northern version from Língua Geral Paulista, a similar lingua franca spoken in the South for a time. In the nineteenth century the autodenomination Nheengatu became more common. An overview of the historical development of this language is presented here, with an attempt to interpret its changes within the framework of the study of languages in contact, especially using the notions of borrowing versus substratum interference (Thomason and Kaufman 1988) and the replication of models (Heine and Kuteva 2005). The findings are similar to those of Moore, Facundes, and Pires (1993), whose suggestions are consistent with subsequent research. Not all the research on the language and its earlier stages can be included here. Nheengatu has a notable charm. People delight in learning it and regard it with affection. It and its earlier versions offer a rich field for philological investigation. It is also important in Brazilian national history. The Portuguese-Tupi interaction is explored by Noll and Dietrich (2010). A certain amount of folklore has grown up around Nheengatu, in

Denny Moore

[ 109 ]

its present and former phases, including the idea that it was invented by the Jesuits or is impure Tupi. The changes that have occurred in the evolution of Nheengatu are complex and are what one might expect given the nature of the language situations through which it passed in the often tragic history of its speakers. A brief sketch of the modern Nheengatu of the Upper Rio Negro, based on original research, is presented in section 5, to give a specific, systematic overview of the language, however brief. There has been surprisingly little description of modern Nheengatu dialects. Fortunately, a large description of the Nheengatu spoken by the Baré, Warekena, and Baniwa, based on field research and natural texts, has recently appeared (Cruz 2011). It includes observations on the evolution of the language. Aside from various evident dialect differences, the analysis presented in that work differs in some respects from the sketch presented here. Only a couple of differences, relevant for the question of the effects of contact, will be dealt with at this time.

2. Language and the Early Settlement of the Coast of Sixteenth-Century Brazil At the time of the first Portuguese contact in 1500, the east coast of what is now Brazil was populated by native peoples speaking a chain of dialects of a language of the Tupi-Guarani branch of the Tupi linguistic family. These tribes were at war with one another and with non-Tupian groups (referred to collectively as Tapuia) in the region. First contacts produced astonishment and wonder on both the European and the indigenous sides. They led to exchanges, in which the native peoples sought European goods, especially metal tools, and the Europeans sought brazilwood for making red dyes, as well as food. The French and the Dutch also established outposts along the coast. Portuguese dominance was not established until the middle of the seventeenth century. European men often fathered children with native women and established close relations with indigenous groups in this manner, producing mestizo children who spoke the native language. Hemming observes, “From the outset, colonists surrounded themselves with native women, and the families descended from Caramuru and Ramalho or Jerônimo de Albuquerque were proud of their Indian blood; such families even bolstered land claims by boasting of descent from the original inhabitants of Brazil!” (1978, 179). This was true of all regions of the colony (cf. Rodrigues 1996): of these three examples of Portuguese men who married into indigenous groups, Caramuru lived in Bahia, Ramalho in São Vincente, and Jerônimo de Albuquerque in Pernambuco.

[ 110 ]

chapter four

Europeans who lived for some time in the new land learned the language, which gave them influence as middlemen, especially since there was relative linguistic uniformity along the coast. According to Lee, “early colonial sources show no indication that ‘mixed’ languages played any role in inter-ethnic linguistic communication. Instead, local tongues, namely mutually comprehensible tribal and regional dialects of the TupiGuarani language family, came to mediate early interlingual experiences” (2005, 21). Some clarification is perhaps in order here about the term TupiGuarani language family, which Lee (chap. 5, this volume) considers “a single entity and linguistic class that brings together smaller speech communities that reflect ethnic and regional varieties.” This definition could fit any genetic grouping of linguistic variants—a stock, family, or language. In its normal usage, the Tupi-Guarani language family (or the TupiGuarani branch of the Tupi language family) designates a large group of distinct languages and groups (including, for example, Xetá, Sirionó, Araweté, Ka’apor, Kamayurá, Guajá, and Tapirapé) who had split from each other perhaps a thousand or more years before the first Portuguese arrival in Brazil. These varied groups live in different regions with differing sociocultural systems. It would be a challenge to specify an ethos common to all of these groups and distinct from, say, Arawak or Carib groups. If the ethos of what is called the coastal Tupi were one of a tendency toward alliances with outsiders, one would expect that their expansion would have been characterized by cooperative relations with neighboring tribes, which does not seem to have been the general case. One of the coastal tribes, the Tupinambá, were said to have descended from the interior of Bahia onto the coast and displaced the Tupinaé (Lee 2005, 79). They played a considerable role in the history of the colony of Brazil. By the end of the seventeenth century, with the Tupinambá tribe greatly diminished, the name began to be employed more generally to designate groups that were linguistically and culturally related. The use of the native language throughout the colony continued after the Portuguese established permanent settlements beginning in the decade of the 1530s. This is similar to the spread and predominance of Maya in the colonial period (Pfeiler, chap. 8, this volume). In the 1540s sugar plantations were established, and they had a great need for labor. Before European contact, the native tribes had taken war captives to be later executed. With European presence some of these were traded as slaves to the whites. The Portuguese and their French rivals encouraged warfare, with each other and between tribes, creating more captives. Soon slave raids by Europeans, abetted by indigenous and mixed-blood allies, were directed against indigenous groups (Hemming 1978, 34–44).

Denny Moore

[ 111 ]

Legally, indigenous slavery was outlawed by the king of Portugal in 1570. However, it continued in practice, since slaves could be seized in “just wars” against hostile indigenous tribes or could be “ransomed”—a term applied to native captives supposedly about to be executed by the tribes who were saved by colonialist forces. Natives could also sell themselves into slavery. To neutralize hostile native populations and obtain workers, expeditions to “descend” native peoples were undertaken, often giving gifts and using mestizos to persuade tribal groups to resettle among the colonists, where many were enslaved or forced to work under harsh conditions and died from disease. Missionaries, settlers, and government officials participated in this system. Each of these three parties had its own fluctuating view of the moral questions involved. Each party pursued its own individual interests, often disputing with the other two parties to further its own purposes. Disease and overwork greatly reduced the indigenous population in contact. One example given in Hemming states: “In the early 1560s there were over 40,000 converts in the Jesuit villages near Bahia. By 1585 Anchieta reported that, despite infusions of thousands more from the interior, these were reduced to ten thousand” (1978, 144). This enormous mortality stimulated more expeditions to capture indigenous people and also the importation of (relatively expensive) African slaves. In 1610 a French visitor in Salvador estimated that the administrative area of Bahia contained 2,000 whites, 3,000–4,000 black slaves, 7,000 Indian and black slaves on the sugar plantations, and 8,000 free natives in the missions (86). The first Jesuits arrived in Brazil in 1549, in Bahia. The Jesuits founded mission villages for indigenous Christians, in theory for their safety, training, and moral education. They soon favored forced resettlement, since the native peoples were doubtful converts with little interest in forced labor. Hemming notes that “by the end of the century there were only 128 Jesuits in Brazil, but they controlled virtually all Indians under Portuguese rule there” (1978, 98). At that time they administered about thirty mission villages (179). Missions were arranged on European patterns, with a church, living quarters for missionaries, and buildings for mission activities. According to Hemming, “The community was run entirely by two Jesuit fathers, a vicar and a curate, always dressed in long black habits and four-cornered caps.” These controlled indigenous subordinates: “They appointed Indians to serve for life as chiefs, and annually as magistrates, bailiffs and other ecclesiastical offices” (1978, 115). Children were separated from their parents and housed in “colleges.” The missions could be quite prosperous in their economic activities, based on indigenous labor. In this

[ 112 ]

chapter four

“curious world of detribalized Indians” (109), the language spoken was the Brasílica. The Jesuits were more active learners of the native language than were the other missionary orders. They made extensive use of interpreters, called línguas. Their earliest interpreters “were selected from colonists of Portuguese birth living in Brazil before the arrival of the Jesuits” (Lee 2005, 152). All these earliest colonists spoke the indigenous language. In the second half of the sixteenth century, the Jesuits produced language manuals for practical use in communication and religious instruction, copied by hand, since there were no printing presses in Brazil at the time. The best-known linguistic work of the sixteenth century in Portuguese Brazil, Arte da Grammatica da Lingoa mais usada na costa do Brasil, was written by Father José de Anchieta (1595), who visited a number of tribes and was attentive to variation. It was composed between 1555 and 1556 in six months and published in 1595 (Lee 2005, 132), consisting of sixty pages, using Latin as a descriptive model. A catechism in the native language was published by Father Antônio de Araújo in 1618. A Jesuit who learned the language of the indigenous villages of Bahia, Father Luis Figueira, published Arte da Língua Brasílica in 1621. The author “sought out rural Indians and great missionary linguists born and raised among the Indians to consult” about the language (Lee 2005, 138). (Note that these sources exclude mestizos, possibly indicating that their speech was already considered less authentic.) The Jesuits were also active in education. Their schools included indigenous, mestizo, and white students. This study and use of the indigenous language parallels that of the Maya language “used by missionaries and plantation owners in their efforts to achieve social and religious domination” (Pfeiler, chap. 8, this volume).

3. Amazonia: The Second Colony and the Evolution of Língua Geral Amazônica Portuguese settlement of the Amazon region began more than a century after the first contacts on the east coast of Brazil. The French abandoned Fort St. Louis (in modern São Luis, Maranhão) in 1615, and the Portuguese went on to establish a wooden fort in what is now Belém, near the mouth of the Amazon River, in 1616. This region became a new colony in 1621, separate from the colony of Brazil to the south. The new colony, the state of Maranhão, included Grão-Pará (extending west up the Amazon River) and Ceará (later ceded to Pernambuco). It was subordinated directly to Portugal, partly for ease of maritime travel. Its capital, São Luis, was transferred to Belém in 1737. Amazonia was less attractive for settlement than Brazil, given the great

Denny Moore

[ 113 ]

distances and difficulty in transportation and communication. Its rainforest environment was very different from Portugal. And it was linguistically much more diverse. A Spanish Jesuit who traveled the Amazon River counted more than 150 different languages along the banks of the Amazon and the mouths of its principal tributaries (Acuña [1641] 1941, 199, cited in Freire 1983, 42). Colonization by Europeans and by African slaves was later and slower in Amazonia, impeding the spread of the Portuguese language. There were only about 150 Portuguese speakers settled in Belém in 1616. By 1720 there were still only about 1,000 Portuguese in Amazonia, in contact with about 75,000 free and enslaved natives (Freire 2004, 54). During the sixteenth century, Amazonia had received migrations of Tupinambá groups fleeing Brazil. Immigrants from settlements in Brazil increased the number of the speakers of this indigenous language: “Indians, whites, blacks, mulattos and mestizos leaving their homes in Brazil to populate the northern colony brought the Brasílica with them” (Lee 2005, 165). Mestizo children learned the language from their indigenous mothers; white children learned it from their indigenous nursemaids. Given the comparative unavailability of Portuguese speakers, the Brasílica was the natural choice for the colonization of Amazonia. The practices developed in the colony of Brazil were applied in Amazonia: slave raids and “descents” of tribal groups into mission and government villages, with forced labor, disease, and immense mortality rates. For the first three decades, there was little control over the settlers. In 1665 a law dealt the Jesuits control over all indigenous villages (Hemming 1978, 324), although settlers resisted, with some success. In 1686 the Regimento das Missões gave the Jesuits complete control of the indigenous population. This legal measure also permitted more “descents” of indigenous groups. In only four years, between 1687 and 1690, the missionaries “descended” 184,040 natives for Church and king (Baena 1831, 247, cited in Freire 1983, 50). These groups spoke many different languages, of diverse language families, such as Arawak, Carib, and Tukano. In 1689 a Carta Régia specified Língua Geral as the official language of Maranhão and Grão-Pará. This official support for the language continued until 1727, when another Carta Régia prohibited Língua Geral and promoted Portuguese. The prohibition had little effect, and in 1750 the whole colony spoke Língua Geral Amazônica, except for some in the colonial administration coming from Portugal. During the first two centuries of Jesuit activity in Brazil (1549–1750) and then in Maranhão and Grão-Pará, the language that was called the Brasílica, essentially Tupinambá, continued as a liturgical language with apparently little change. However, this language as it was used in daily affairs by common people seems to have changed greatly by the mid-

[ 114 ]

chapter four

eighteenth century. Lee makes the strong claim that “one hundred years after the initial colonization of the State of Maranhão and Pará, the Brasílica had become incomprehensible” (2005, 165). Since the language was still actively used in liturgy, some classic linguistic works about the Brasílica were republished. For example, the 1621 grammar by Figueira reappeared in 1685 and the 1616 catechisms of Antônio de Araújo were reedited in 1686. A catechism by Bettendorff was published in 1687, but, having observed the gap between the Brasílica and the language commonly spoken, Bettendorff produced another version in the colloquial form of the language, Língua Geral. Freire observes that “Tupinambá was the language of the ritual acts, LGA the language of popular communication and, therefore, of religious instruction. The relation between Tupinambá and LGA was similar to the relation between Latin and Portuguese in Portugal and in other parts of Brazil” (2004, 118). Durston uses the very same analogy in the Quechua case, stating that “in some respects [Standard Colonial Quechua] stood in relation to Central Quechua varieties as Latin did to Romance languages in medieval and modern Europe” (chap. 9, this volume). Pfeiler discusses a similar contemporary diglossia situation, contrasting “pure” Hach Maya with the vernacular Xe’ek’ (chap. 8, this volume). Monserrat found the language in the documents from the eighteenth century to be very different from the language of the two earlier centuries (2003, 188). Lee discusses language manuals from the eighteenth century that describe the colloquial Língua Geral, in contrast to the Brasílica (2005, 205–208). There seems to be general agreement that by the mid-eighteenth century Língua Geral Amazônica was quite different from Tupinambá, although it is hard to say whether they were truly mutually incomprehensible. Another question is whether the alterations in the Brasílica were not already in progress in the colony of Brazil in the century preceding the colonization of Amazonia. Lee states, “In its early Amazonian form, the Brasílica exhibited minor differences in pronunciation and vocabulary but remained comprehensible to its speakers in the State of Brazil” (2005, 169). Given the lack of records of colloquial speech of the sixteenth century, it is hard to estimate the degree of change during that century. If Língua Geral Amazônica and Língua Geral Paulista were spoken by distinct speech communities after the first century or two of contact, it might be possible to spot shared innovations in the two languages (not caused by parallel evolution or borrowing) that would date those changes to the time before their separation. The fact that Figueira took care to avoid the common speech of the settlements as a basis for his grammar hints at changes there. One would imagine that the borrowing

Denny Moore

[ 115 ]

of Portuguese words for Western items and practices had begun in the sixteenth century. Any changes in the common speech in the colony of Brazil would presumably have been brought to the new colonies of Maranhão and Grão-Pará by people arriving there from Brazil. If the principal mechanism modifying the Brasílica had been borrowing (as opposed to substratum interference), then the colony of Brazil would, arguably, have had more favorable conditions for this than would Amazonia, since Brazil, by the end of the sixteenth century, had a considerably higher concentration of Europeans (and Africans). But there would have been a countervailing conservative tendency because of the presence in the colony of Brazil of many speakers of Tupinambá and related dialects, providing continuous input of the relatively unaltered form of the language. Since the indigenous groups captured or “descended” in the colony of Brazil often spoke dialects close to the Brasílica, there would have been little substratum interference in the speech of the mestizos and detribalized peoples. In Amazonia in the seventeenth and eighteenth centuries, the conditions for language change would have been the reverse of those in the colony of Brazil in the sixteenth century. There were relatively few speakers of Portuguese. The Tupinambá themselves were becoming nearly extinct in the eighteenth century, and their speech would no longer serve as ballast for the Brasílica. At the same time, relatively enormous numbers of speakers of non-Tupi-Guarani languages were brought into contact in resettlement villages, and some form of the Brasílica or, later, LGA, was the means of communication with each other and with authorities. According to Lee, “the Jesuit missionary João Daniel observed aldeias [villages] in which 30 to 40 groups, each speaking different languages, lived side by side” (2005, 183). There is evidence for the lack of any preexisting trade language or lingua franca in the Amazonian colony. For example, the contact with the Nheengaíba of Marajó was possible using a mestizo whose mother spoke the language; there was no shared language (Barros 2003, 95). The difficult communication with the Aimoré likewise indicates a lack of a lingua franca used between indigenous groups with different mother tongues in the colony of Brazil (Hemming 1978, 172). Durston (chap. 9, this volume) doubts a direct relationship between Standard Colonial Quechua and any precontact indigenous lingua franca, which has sometimes been assumed. The scenario in which speakers of many indigenous languages had to learn the lingua franca, LGA, in the Amazonian resettlement villages is exactly what Thomason and Kaufman describe as “imperfect group learning during a language shift” (1988, 37). In fact, the rapid evolution

[ 116 ]

chapter four

of LGA is just what would be expected, according to those authors, who state, “In fact, substratum effects are more likely to enter a TL [Target Language] rapidly than slowly: if the shift takes place over long centuries, then the shifting population is more likely to be truly bilingual in the TL” (41). One indication of the mechanism of language change (borrowing versus substratum interference) is the amount of lexical borrowing relative to structural effects. Thomason and Kaufman observe, “The interesting point here is that in borrowing proper many words will be borrowed before any structural interference at all occurs, but in substratum interference . . . structural interference comes first” (1988, 21). Borrowings from the various languages being replaced do not appear to have been numerous, although they are not easy to detect. Moore, Facundes, and Pires regard the facts of the rapid changes in LGA at this time as highly consistent with substratum interference from indigenous groups learning the language from indigenous and mestizo speakers and less consistent with lexical or structural borrowing from Portuguese. They note, “As would be expected, borrowings from Portuguese were limited, but the grammar was altered by many new speakers” (1993, 115). In fact, even in texts from the latter half of the nineteenth century, Portuguese lexical borrowings are few, and they must have been even fewer a century earlier. For example, a Christian text 411 words long, “A Criação de todas as Cousas” (The creation of all things), transcribed by Lourenço Costa Aguiar, contains only three obvious borrowings: santasána ( N

nã́mbi-púra

‘earring’

N + N > Adj N + Adj > N

Ptc + Ptc > Ptc not-to

foot-nail pain-day

manioc-soft

know-well

ear-part.inside čí-arã́ma ‘to not’

5.2.3 Inflection and Derivation The inflectional and derivational affixes of modern Nheengatu are all Tupian. Even recent Portuguese borrowings can accept person prefixes. Verb infinitives can accept indigenous affixes, rather like the Spanish

Denny Moore

[ 129 ]

Table 4.5. Inflectional and derivational affixes in modern Nheengatu Inflection

Derivation

Verbal Series a-

1 sg

muyu-

transitivizer intransitivizer/ reflexivizer ‘semi, almost’ ‘someone with tendency for . . .’

reu-

2 sg 3

-̃to -mã́ỹã

yape(aẽ˜ta-)u-

1 pl 2 pl 3 pl

-sára ~ -gára -éra ~ -wéra -wára -ĩ́ma

agent ‘habitual doer of . . .’ ‘characterized by’ ‘without’

Nominal Series seneiyanépeaẽt̃ á-itá -ã́na ~ -wã́na -rẽ́

1 sg 2 sg 3 unspecified 1 pl 2 pl 3 pl pl perfective imperfective

-mirí -asú

diminutive augmentative

infinitives borrowed into Maya (Pfeiler, chap. 8, this volume). Some modern affixes seem to be the result of grammaticalization of what were formerly lexical items. For examples of modern affixes, see table 4.5. Reduplication to indicate repetitive action has been retained as a morphological process in Nheengatu, for example, ya-yapí ‘throw or shoot repeatedly’, pi-píka ‘drizzle’. Reduplication was present in Tupinambá, as in most Tupian languages.

5.3 syntax 5.3.1 Matrix Clause Composition The matrix clauses are in some respects similar to Portuguese and in other respects similar to Tupian languages. The morphemes associated with syntactic operations are mostly indigenous. There are three sentence types in Nheengatu, in embedded as well as matrix clauses. Verbal sentences consist of an optional subject followed by one or more VPs containing verbs carrying a subject prefix. Multiple VPs are characteristic of Tupian languages (Moore 1994). These verbs

chapter four

[ 130 ]

may be intransitive, transitive, or stative. Transitive verbs are optionally followed by an object, as in Portuguese. In the examples below, embedded rather than matrix clauses are given as illustrations if the text examples of the latter are lacking or unclear: (1)

[ya-mũỹṹ]VTrans 1pl-make

čĩmbiʔú

[ya-pinačíka]VIntr

food

1pl-fish

[ya-mũỹṹ]VTrans 1pl-make

‘We make food, we fish, we make chicha.’ (Note three VPs.) (2)

[yã-mbúri]VTrans 1pl-put

maniáka

paranã́

manioc

upé

river

in

[i-mẽmbéka]VStat 3-be.soft

‘We put the manioc in the river to become soft.’

kaširí

chicha

arã́ma

in.order.to

There are two verbs that might be considered auxiliaries; they occur after the main verb, contrary to the order in Portuguese: putái ‘want’ and ikú ‘be’. The former can occur without a subject prefix, forming a complex verb. The latter can be preceded by a verb, an adjective, or a postpositional phrase, as illustrated below: (3)

[a-yuwíri

putái]V’

se-retã́ma

1sg-return want

kití

1sg-land

to

‘I want to return to my land.’ (4)

yãndé

[ya-ikú]Aux

[[ya-purĩngitá]V

we

 1pl-speak

 1pl-be

‘We are talking Nheengatu.’ (5)

išé

[se-rúka

I

1sg-house

‘I’m in my house.’

upé]PP

in

yeʔẽngatú]VP

Nheengatu

a-ikú 1sg-be

The copula sentence type consists of an obligatory subject followed by a predicate noun phrase or adjective phrase. Unlike Portuguese, Nheengatu has no overt copula, as in the following example, in which the subject-predicate order is inverted: (6)

rẽ-mbeʔú

2sg -tell

aẽta-supé 3pl-for

‘Tell them that I’m fine.’

[purã́nga  good

išé]S’Cop

I

The third sentence type consists of a predicate with no subject. The predicate is composed of a predicative particle followed by an NP or by a

Denny Moore

[ 131 ]

clause with an overt subject. These are perhaps replicating Portuguese impersonal constructions, except that the predicative particle shows no verbal characteristics. However, first and second position particles are common in Amazonian languages. Cruz considers the predicative particles as first position (2011, 358), although they do not necessarily occur initially in the sentence, e.g., her example (905). At least one of the predicative particles, presízu ‘it is necessary . . .’, is borrowed from Portuguese É preciso; and the first syllable of aikwé ‘there is’ looks like Portuguese aí ‘there’: (7)

[aikwé]Ptc

kašuéira

there.be

waterfall

[presízo]Ptc

aẽtá

‘There are waterfalls.’ (8)

is.necessary they

u-ištudái

pohtugéš

upé

3-study

Portuguese

in

‘It’s necessary that they study Portuguese.’

5.3.2 Syntactic Processes in Matrix Clauses The major syntactic processes affecting matrix clauses look more indigenous than European. Negation: Verb phrases can be individually negated with the particle čí: (9)

čí

not

[a-pitá]VP

1sg-stay

[a-iwír

1sg-return

kwá-kití]VP

this-toward

‘I don’t stay, I come back to Belém.’

The negative particle can occur in the beginning of the clause, negating all of it. It can also form a negative focus construction with a fronted NP: (10)

[čí

tapiʔíra]Focus

  not tapir

apigáwa

man

u-yuká 3-kill

‘It was not the tapir that the man killed.’ (elicited)

Topicalization: Noun phrases can be topicalized, leaving behind third person copies: (11)

[ỹãʔã́ that

yawára,]Topic dog

aʔé it

u-suʔú 3-bite

‘That dog, it bit the man.’ (elicited)

apigáwa

man

chapter four

[ 132 ]

Questions: Polar questions can be formed by intonation. (12)

ĩndé you

re-murái

apekatú

2sg-live

kwá-suʔí

far

this-from

tetã́ma

city

‘Do you live far from here from this city?’

suí?

from

Interrogative word questions are formed using indigenous interrogative words and the particle taʔá. (13)

mãʔã́

taʔá

re-wasẽ́mu

pušuéra?

what

Q

2sg-find

ugly

‘What do you find ugly?’

As in Portuguese, the interrogative word need not be fronted and can remain in situ. (14)

taína

child

u-mãʔã́

mãʔã́?

3-see

what

‘The child saw what?’

Adverbial movement: Sentence-level adverbials can be fronted or placed between phrases. (15)

[kušiʔĩ́ma]Adv formerly

aikwé

there.be

yepé

a

feičiséiru

shaman

a-koñeséi

1sg-know

‘Formerly, there was a shaman whom I knew.’

waʔá

relz

Some common syntactic processes in Portuguese, such as passives or clefts, do not occur in Nheengatu. 5.3.3 Embedded Clauses Nheengatu embedded clauses are especially noteworthy in that they show three different patterns: (i) Subordinate clauses formed on an abstract model, perhaps Portuguese, using indigenous morphemes (ii) Subordinate clauses formed on a specific Portuguese model, but using indigenous morphemes (iii) Frank borrowings from Portuguese, with accompanying Portuguese grammatical morphemes

Denny Moore

[ 133 ]

In the first pattern, the clause contains a subordinating particle immediately after the head of the predicate, that is, after the main verb, after the predicate nominal or predicate adjectival, or after the predicating particle, according to the type of the predicate. That is, a paradigm for subordination has been created, not just a collection of individual cases of subordination. These particles include waʔá relativizer, ramẽ́ time, arã́ma purpose, and čí-arã́ma negative purpose. (This last particle occurs clause initially). The relative clauses can have an external head and a corresponding empty internal extraction site: (16) a-yururé

se-mã́ỹã

u-pitá

́ a arãm

yane-rẽndá

1sg-ask

1sg-mother

3-stay

purpose

1pl-farm

1sg-grandpa

3-leave

relz

us

purpose

[se-ratíwa

u-šári

waʔá

yãndé

arã́]S’Rel

upé

in

‘I asked my mother to stay in our farm that my grandfather left for us.

In that example, note that the relative clause modifying ‘farm’ has been extraposed from inside the postpositional phrase to the end of the sentence. Extraction and backing of NPs is common in Nheengatu and appears to be an innovation, not occurring in Portuguese and not generally reported for Tupi-Guarani languages. Alternatively, the relative clauses may be headless, with one missing argument: (17) aẽtá u-kõtái

[Ø u-akõteséi

they 3-tell

Ø 3-happened

waʔá

garapé

apíra

kití]     S’Rel

relz stream headwaters toward

‘They would tell us what happened on the headwaters of the stream.’

The time, purpose, and negative purpose clauses formed by ramẽ́, arã́ma, and čí-arã́ma, respectively, distribute like adverbials or adjectivals: (18) aẽtá u-pisíka paʔá

yãndé [ya-ú

they 3-catch they.say us

ramẽ́

č ĩmbiʔú irusã́nga]S’Adv

1pl-eat time food

‘They would catch us when we ate cold food.’ (19) yã-mbúri 1pl-put

maniáka

manioc

paranã́

river

upé

in

[i-mẽmbéka 3-be.soft

cold

arã́ma]S’Adv purpose

‘We put the manioc in the river in order for it to become soft.’

(20) ya-ú

čĩmbiʔú, sakú, [čí-arã́ma

1pl-eat food

hot

kurupira-itá

u-rasú yãndé]S’Adv

neg-purpose Kurupira-pl 3-take us

‘We would eat hot food for the Kurupira not to take us away.’

chapter four

[ 134 ]

(21)

ya-mãỹã́

čĩmbiʔú

‘

We saw the food for the man to eat.’

1pl-see

[apigáwa

food

u-ú

man

3-eat

arã́ma]S’Adj

purpose

In the second pattern, a subset of the Nheengatu WH words (MA words in Nheengatu) are used in embedded clauses in a manner similar to that of Portuguese. The MA words are awá ‘who(m)’, mãʔã́ ‘which, what, that’, mairamẽ́ ‘when’, marã́ma ‘because’, mamẽ́ ‘where’, and mayé ‘how’. The relative clauses with awá and mãʔã́ cannot have external heads, unlike their Portuguese counterparts: (22)

[mãʔã́

u-yururé

what

i-tupã́na

3-asked

3-god

u-yũmbuʔé

3-pray

tupã́na

god

‘What he asked (from) his god, praying to his god . . .’ (23)

*apigáwa man

[mãʔã́ who

u-yururé

3-asked

supé . . .]S’Rel

to

i-tupã́na . . .]S’Rel

3-god

‘The man who/that asked his god . . .’

These are strikingly similar to Tariana relative clauses with an indigenous interrogative pronoun head, also formed on a Portuguese model, reported by Aikhenvald (2002, 181). The clauses formed by the other MA words distribute as adverbials or adjectivals: (24)

aẽtá

u-mãʔã́

they

úka

3-see

house

[mamẽ́

where

‘They saw the house where I live.’ (25)

išé I

čí

not

a-sasá

1sg-pass

i-puší

3-bad

a-murái]S’Adj

1sg-live

[mayé how

‘I’m not having a bad time like they say.’

aẽtá

they

ũ-mbeʔú]S’Adv 3-say

Embedded questions also follow the Portuguese pattern, but using indigenous MA words: (26)

. . . čí

aẽtá

not they

u-kwá

3-know

[mãʔã́ what

kurupíra-itá

Kurupira-pl

u-mũỹṹ 3-do

‘. . . they didn’t know what the Kurupira would do to us.’

yane-irṹ]  S’Q 1pl-with

Some transitive verbs can take unmarked sentential complements:

Denny Moore

[ 135 ]

Table 4.6. Portuguese markers of subordination and coordination in modern Nheengatu Function

Nheengatu

Portuguese

English

complementizer conjunction disjunction negative disjunction conditional

kí i u nẽ́ sá

que e ou nem se

that and or neither/nor if

(27)

. . . nẽ́

išé

a-mãnduʔái

nor I

[a-mãʔã́

1sg-think

ĩndé]S’Comp

1sg-see

you

‘. . . nor I think of seeing you.’

In the third pattern listed above, subordination is marked. Obvious borrowings from Portuguese for subordination and coordination are listed in table 4.6. (28)

. . . [re-murái

2sg-live

iké]VP

here

u

[re-murái

or

ĩterió

2sg-live

interior

‘. . . you live here or in the interior . . .’

kičí . . .]VP

toward

5.3.4 Phrases The structure of phrases is rather conservative. Two major changes from the indigenous pattern are the order Verb Object in the VP and the greater elaboration of adjectival and adverbial phrases as incorporation within the verb declined. Noun phrases retain the order Genitive Noun: [yane-yeʔẽ́nga]NP 1pl-language

[karíwa

white.man

‘our language’ yeʔẽ́nga]NP

language

‘white man’s language’

Also attested is the order Noun Adjective when the latter is attributive: (29)

[čí

not

ya-pudéi

1pl-can

[[[ya-ú

[čĩmbiʔú

1pl-eat food

We cannot eat cold food . . .’

And Demonstrative NP:

irusã́nga]NP ]VP ]S’Comp ]VP cold

chapter four

[ 136 ]

(30)

[[kuʔá this

[se-awá-itá

1sg-hair-pl

purã́nga . . .]NP ]NP

pretty

‘This pretty hair of mine . . .’

There is a position after the head of the predicate that contains aspectual suffixes, subordinating particles, and auxiliaries: (31)

a-mũỹã́

1sg-do

pãỹẽ́

all

mãʔã́

what

mamẽ́

where

[a-purakí

waʔá

1sg-work

relz

‘I do everything where I am working.’

(32)

[[a-síka

ramẽ́]VP]S’Adv

   1sg-arrive time

se-mbíra-itá

1sg-child-pl

a-ikú]VP

1sg-be

[u-kíri-ãna

u-ikú]VP

3-sleep-already 3-be

‘When I arrive, my children are already sleeping.’

Nheengatu retains postpositions (common in Amazonia), in contrast to the prepositions of Portuguese. As is characteristic of Tupian languages, postpositional phrases have a strictly adverbial distribution, never modifying nouns. (33)

a-morái

1sg-live

ramẽ́

time

[[se-páya]NP

1sg-father

‘When I lived with my father . . .’

irṹ . . .]PP with

5.4 text fragment of modern nheengatu from the upper rio negro This is the beginning of a text, “Conversation in Belém between Two People from the Upper Rio Negro,” which was recorded and transcribed in 1988, in the Museu Goeldi in Belém. The two speakers are Lenir da Silva, a young woman in her thirties from the region of the Upper Rio Negro, trilingual in Nheengatu, Portuguese, and Spanish, and Gerson, a somewhat younger man from a Baniwa community who is bilingual in Nheengatu and Portuguese and who lives in the city of San Gabriel da Cachoeira. In this small text fragment, seventy-six words long, there are seven loan words from Portuguese: ‘live’, ‘here’, ‘city’, ‘nine’, ‘until’, ‘family’, and ‘or’. Gerson: (34)

ĩndé you

muʔí

how.many

akayú years

taʔá Q

re-morái

2sg-live

iké

here

‘For how long have you lived here in this city?’

kwá

this

sidádi

city

upé

in

Denny Moore

[ 137 ]

Lenir: (35)

išé I

akayú

nóvi

year

nine

akayú-ãna

year-already

a-yuwíri

1sg-return

‘It has been nine years that I live in the city.’

(36)

išé I

a-yupukwá

se-retã́ma

1sg-city

suʔí

from

iké

1sg-accustom

here

‘I got used to this place.’ (37)

išé I

čí

a-mãnduʔái

not

a-yuwíri

1sg-think

se-família-itá

1sg-return

rúka

1sg-family-pl

house

‘I don’t think of returning to my family’s house.’ (38) a-kwakatú

išé

1sg-believe

čí

I

a-yupukwá

not

kití

to

a-kití

1sg-accustom

there-to

‘I think I cannot accustom myself to that place anymore.’ (39)

a-pitá

kurí

iké

até

kumairamẽ́ Tupã́na-itá

1sg-stay future here until

when

kurí

God-pl

future

u-kwá

3-know

‘Only God knows how long I’m going to stay here.’ (Literally, ‘(I) will stay here until when God will know.’) (40)

mayé

taʔá

how

a-mãʔã́

a-sú

Q

ãỹṹ

1sg-go arã́

1sg-see

a-watá

only 1sg-walk se-anã́ma-itá

purpose

se-retã́ma 1sg-city

kití to

1sg-family-pl

‘How can I go back to that city only to see my family?’ (41)

išé I

čã́

not

a-mãnduʔá 1sg-think

a-yuwíri

1sg-return

‘I don’t think of going back there.’ (42) a-yuwíri

1sg-return

kurí

future

ãỹṹ

only

a-mãʔã́

1sg-see

a-kití

there-to

arã́

for

‘I will go back there just to visit my family.’

se-anã́ma-itá

1sg-family-pl

Gerson: (43)

kušiʔima ̃́

formerly

re-yúwi

ramẽ́

kwá-kití

2sg-come time that-to

mayé-ta

re-yúwi

how-Q

2sg-come purpose

‘Formerly, when you came here, how did you come?’

arã́

chapter four

[ 138 ]

(44)

aikwé

awá

there.be who

u-rúi

3-bring

ĩndé

you

u

or

re-yúwi

2sg-come

putái

want

te

real

ne-rupí

2sg-by

‘Was there anybody to bring you or did you yourself want to come?’

6. Conclusion and Prospects for the Indigenous Languages in Brazil What I have presented in this chapter is only an outline of the historical development of Nheengatu up to the present, which has been driven by economics, politics, religious expansionism, demography, geography, and technological change. There are various parallelisms to the cases of Quechua and Yucatán Maya. It is clear that the evolution of the language has been complex; each generation of learners has modified it using the patterns available to them. Nheengatu has been more an alternative to Portuguese than an imitation of Portuguese. It was always the native language of some speakers, with continuity of transmission, though strongly influenced by historical events and language contact. It was never a pidgin or a creole. This language provides an excellent opportunity for studies of languages in contact, since there are documents from each century. I have given here a sketch of the present form of the language in one region, with special attention to the effects of language contact on its structure. Some suggestions in terms of borrowing, substratum interference, and model replication have been offered. Substratum influence from indigenous peoples learning the language seems to have been the principal mechanism of the rapid change in LGA as diverse groups were abruptly and often brutally assimilated during the seventeenth and eighteenth centuries. In addition, some model replication and structural diffusion from Portuguese took place along the centuries of contact. Lexical and grammatical borrowing from Portuguese increased in the past century, with increased bilingualism and more dislocation and mixing of native groups for economic reasons, especially during the two rubber booms. At present the priority for research is the location and documentation of the modern dialects of Nheengatu. This is urgent, given the threat of their extinction, which already appears to have happened to some of the varieties. Studies of the historical stages of the language need to be systematic and informed by modern knowledge of diachronic linguistics and of the effects of languages in contact; lists of changes since Tupinambá are not sufficient. Nheengatu was declared one of the official languages of the municipality of São Gabriel da Cachoeira in 2002, but that measure has not been implemented.

Denny Moore

[ 139 ]

From the history of Nheengatu we can imagine that the same causal factors will determine the fate of the native languages of Brazil. Some of these factors continue as in the past. The insistent missionary drive to eliminate native religion (and the music, narratives, and other verbal art associated with it) continues in full force, though with fundamentalist Protestant sects now as the principal agents, often using language study as a means of control and indoctrination. One recent brute example of this drive, well known in Brazil, is the pseudodocumentary video Hakani created and distributed by JOCUM (Youth with a Mission) missionaries. The video portrays the Suruwahá shaman as causing suicides and the live burial of children. It has been denounced by Survival International (www .survivalinternational.org/about/hakani). Even the Xingu Park, conceived as an area where indigenous culture would be protected (Ball, chap. 10, this volume) is now actively targeted by missionary forces. Economics continues to be a primary determining factor, and it is urgent to develop appropriate economic alternatives on the indigenous reserves, which potentially offer good opportunities for extractivist activities, although this potential is seldom realized. Politically, contemporary Brazilian government policies are increasingly enlightened. Protection for recently contacted indigenous groups is much improved. Bilingual education is required, although such requirements are often of limited success in practice. International documentation programs, especially the Endangered Languages Documentation Programme and the DOBES (Documentation bedrohter Sprachen) program, have sponsored many projects in Brazil, conducted by Brazilian linguists or linguists residing in Brazil. Partly through the influence of these programs, a government program, ProDocLin (Projeto de Documentação de Línguas Indígenas), administered by the Museu do Índio and coordinated by an experienced field linguist, Bruna Franchetto, is documenting thirteen languages and stimulating language maintenance. The movement in favor of language documentation using recent methods is more advanced in Brazil than in the rest of South America. A first step in language preservation is knowledge of the situation of the individual languages. Earlier language surveys (for example, Rodrigues 1986) confused languages and ethnic groups and also confused population figures with speaker numbers. As a result of this last confusion, the degree of language endangerment was seriously underestimated for years. Recent surveys, while still deficient, give a more accurate picture, for example the UNESCO Atlas of the World’s Languages in Danger (www .unesco.org/culture/languages-atlas), Moore (2006), and Moore, Galucio, and Gabas (2008). A large-scale survey of all the languages of Brazil, the Inventário Nacional de Diversidade Linguística, is in the initial stages of

[ 140 ]

chapter four

implementation. It potentially will gather enormous amounts of useful data, if designed and implemented in an intelligent manner. The 2010 Brazilian national census contained a question about the indigenous language spoken. However the census had the limitations inherent in autodeclaration, aside from some confusion about what a “language” is. Part of Brazilian linguistic folklore is that there are 180 indigenous languages in Brazil, although no one repeating this number can say what it means— what the criteria are for grouping variants into languages. Given the current interest by native groups in language documentation and maintenance and given the progressive government policies, it is too early to be defeatist about the prospects of survival of at least the more vital languages. Reading colonial history, we distinguish between those figures whose action was favorable to the native peoples and those whose action was unfavorable. Someday the history of our own epoch will be written, and it will be seen who was useful to the indigenous peoples (and their languages), who was harmful, and who just sat there talking. Notes 1. Americanist phonetic symbols are commonly used for Nheengatu. Nonobvious symbols used are pl (plural), Q (question), S’Rel (relative clause), S’Comp (sentential complement), relz (relativizer), PP (postpositional phrase), S’Adv (adverbial clause), S’Cop (copula complement clause), V’ (complex verb), and Ptc (particle). 2. Eternal gratitude to Lenir da Silva for her patient assistance.

References Acuña, Cristobal de. [1641] 1941. Novo Descubrimento do Grande Rio das Amazonas. In Descubrimentos do Rio das Amazonas, ed. Carnaval et al., 127–294. São Paulo: Cia Editora Nacional. Aikhenvald, Alexandra. 2002. Language contact in Amazonia. New York: Oxford University Press. Anchieta, Joseph de. 1595. Arte de Grammatica da Lingua mais Usada na Costa do Brasil. Coimbra, Portugal: Antonio Mariz. Baena, Antonio Ladislau Monteiro. 1831. Representação ao Conselho Geral da Provincia do Pará Sobre a Especial Necessidade de um Novo Regulamento Promotor da Civilização dos Índios da Mesma Provincia. Annaes da Bibliotheca e Arquivo Público do Pará - Belém 2:241–292. Barros, Maria Cândida D. M. 2003. Notas sobre a política jesuítica da língua geral na Amazônia (séculos XVII–XVIII). In Freire and Rosa 2003, 85–112. Bettendorff, João Felipe, S.J. 1678. Compendio da Doutrina Christam Na lingua Portugueza, & Brasilica: Em que se comprehendem os principaes mysterios de nossa Santa Fe Catholica, & meios de nossa Salvação. Lisboa, Portugal: Na Officina de Miguel Deslandes. Borges, Luiz C. 1991. A língua geral amazônica: Aspectos de sua fonêmica. Master’s thesis, Universidade Estadual de Campinas, Campinas, São Paulo.

Denny Moore

[ 141 ]

Cabral, Ana Suelly A. C. 1995. Contact-induced language change in the western Amazon: The non-genetic origin of the Kokama language. PhD diss., University of Pittsburgh. Cruz, Aline. 2007. O estatuto das fricativas na Língua Geral Amazônica. In Afinal, o que, nós, lingüistas, fazem? Seleção de textos proferidos durante o IX Encontro dos Alunos de Pós-Graduação em Linguística da Universidade de São Paulo, ed. Suzi Lima. Available at www.fflch.usp.br/dl/ixenapol/Trabalhos/cruzaline.pdf. ———. 2011. Fonologia e gramática do Nheenagatú: A língua geral falada pelos povos Baré, Warekena e Baniwa. Utrecht, Netherlands: LOT. Daniel, João. 1976. Tesouro descoberto no Rio Amazonas. 2 vols. Rio de Janeiro: Biblioteca Nacional. Felix, Maria Invanete de Santana. 2002. A Língua geral amazônica: Contributions para o estudo de suas variedades dialetais faladas ao longo do Rio Amazonas e seus tributários, nos séculos XIX e XX. Master’s thesis, Universidade Federal do Pará. Figueira, Luís. 1621. Arte da Língua Brasílica. Lisboa, Portugal: Manoel da Silva. New edition in 1687, by Miguel Deslandes: Arte de Grammatica da Lingua Brasilica. Facsimile by J. Platzmann in 1878: Grammatica da Lingua do Brasil, Leipzig B. G. Teubner. More recent reedition by Emílio Allain: Arte de Grammatica da Lingua Brasilica, Lombaerts and C., Rio de Janeiro, s.d. Freire, José Bessa. 1983. Da “fala boa” ao português na amazônia brasileira. Amérindia 8:1–83. Paris. ———. 2004. Rio Babel—a história das línguas na Amazônia. Rio de Janeiro: EdUERJ. Freire, José Bessa, and Maria Carlota Rosa, eds. 2003. Línguas gerais: Política lingüística e catequese na América do Sul no período colonial. Rio de Janeiro: EdUERJ. Grenand, Françoise, and Epaminondas Henrique Ferreira. 1989. Pequeno Dicionário da Língua Geral. Série Amazonas-Cultura Regional, 6. Manaus, Brazil: SEDUC. Heine, Bernd, and Tania Kuteva. 2005. Language contact and grammatical change. Cambridge: Cambridge University Press. Hemming, John. 1978. Red gold: The conquest of the Brazilian Indians. London: Macmillan. Lee, M. Kittiya. 2005. Conversing in colony: The Brasílica and the Vulgar in Portuguese America, 1500–1759. PhD diss., Johns Hopkins University. Leite, Yonne. 2003. A Arte de gramática da língua mais usada na costa do Brasil e as línguas indígenas brasileiras. In Freire and Rosa 2003, 11–24. Michael, Lev. 2010. The pre-Columbian origin of a diachronic orphan: The case of Omagua. Paper presented at the Berkeley Linguistic Society 36, February 7, 2010. Monserrat, Ruth M. F. 2003. O tupi do século XVIII (tupi-médio). In Freire and Rosa 2003, 185–194. Moore, Denny. 1994. A few aspects of comparative Tupi syntax. Revista Latinoamericana de Estudios Etnolingüísticos 8:157–168. ———. 2006. Brazil: Language situation. In Encyclopedia of Language and Linguistics, ed. Keith Brown, 2:117–128. 2nd ed. Oxford: Elsevier. Moore, Denny, Sidney Facundes, and Nádia Pires. 1993. Nheengatu (Língua Geral Amazônica), its history, and the effects of language contact. In Survey of California and other Indian languages, ed. Margaret Langdon, 93–118. Report 8, Proceedings of the Meeting of the Society for the Study of the Indigenous Languages of the Americas, July 2–4, 1993, and Hokan-Penutian Workshop, July 3, 1993, Columbus, OH.

[ 142 ]

chapter four

Moore, Denny, Ana Vilacy Galucio, and Nilson Gabas Jr. 2008. O desafio de documentar e preservar as línguas amazônicas. Scientific American (Brasil) Amazônia (A floresta e o futuro) 3:36–43. A revised version is available at www.etnolinguistica. org/media:set2008. Noll, Volker, and Wolf Dietrich, eds. 2010. O português e o tupi no Brasil. São Paulo: Editora Contexto. Rodrigues, Aryon. 1986. Línguas brasileiras: Para o conhecimento das línguas indígenas. São Paulo: Edições Loyola. ———. 1990. You and I = neither you nor I: The personal system of Tupinambá (Tupi-Guaraní). In Amazonian linguistics: Studies in lowland South American languages, ed. Doris L. Payne, 393–405. Austin: University of Texas Press. ———. 1996. As línguas gerais sul-americanas. Papia 4:6–18. Schleicher, Charles O. 1998. Comparative and internal reconstruction of the TupiGuarani language family. PhD diss., University of Wisconsin–Madison. Silverstein, Michael. 1996. Dynamics of language contact. In Languages: Handbook of North American Indians, ed. William C. Sturtevant, 17:117–136. Washington, DC: Smithsonian Institution. Stradelli, Ermana. 1929. Vocabularios da Lingua Geral Portuguez-Nheêngatu e Nheêngatu-Portuguez. Revista do Instituto Histórico e Geographico Brasileiro,Tomo 104, vol. 158: 5–768. Taylor, Gerald. 1985. Apontamentos sobre o nheengatu falado no Rio Negro, Brasil. Amérindia 10:5–23. ———. 1988. Ortografia do nheengatu: Proposta de um sistema gráfico para transcrever a Língua Geral do Rio Negro, Amazonas, Brazil. Unpublished manuscript, Centre National de la Récherche Scientifique, Paris. Thomason, Sarah Grey, and Terrence Kaufman. 1988. Language contact, creolization, and genetic linguistics. Berkeley: University of California Press.

5 * Language and Conquest: Tupi-Guarani Expansion in the European Colonization of Brazil and Amazonia m. kittiya lee 1. Introduction: Tupi-Guarani Colonists and a New Coastal Language Ecology Around 1587 in Salvador, “the oldest of Indians” recalled that even prior to the arrival of the Portuguese in 1500, the Tupinaé had reached the Bay of All Saints and had driven away its inhabitants, speakers of Gê languages whom they called Tapuia. After many years of living by the sea, the Tupinaé were ousted by another group of newcomers, the Tupinambá. Attracted by the “richness and fertility of this land,” the invaders erected villages, planted gardens, enjoyed excellent hunting, and traded with neighbors, eventually populating 100 to 150 independent villages along over four thousand kilometers of coastline (Soares de Sousa [1587] 1989, 215, 216;1 Monteiro 1997, 977). See map 5.1. The settlement of the Tupinaé and the Tupinambá in the best agricultural grounds of lowland eastern South America begins the story of expansion and conquest for the coastal clans of the Tupi-Guarani language family. This story was repeated in the colonial history of Portuguese America (1500–1822) and thus shaped that history. The descendants of the newcomers were said to have greeted the crew sailing under Pedro Álvares Cabral, the Portuguese commander credited with the official discovery of Brazil in 1500. Since the days of early communications between indigenous societies and European visitors, and ongoing through much of the colonial era, the languages of coastal Tupi-Guarani-speaking peoples were the first ones used between and among Indians, Europeans, Africans, and their American-born kin. Writers, observers, and Jesuit missionaries recognized the widespread use of Tupi-Guarani speech and produced grammars, catechisms, and other didactic texts that codified the ethnic tongues into one single language format. They called it a língua geral Brasílica, which literally means “the Brazilian lingua franca.” In the sixteenth and seventeenth centuries in the colony, the state of Brazil, it served as unofficial lingua franca (Rodrigues 1986, 21). The Brasílica gained a new role in Portugal’s lesser known colony: the state of Maranhão and Pará, or Amazonia. About a half century following foundation in 1621, the Brasílica was adopted legally as the language of colonization in 1686.

[ 144 ]

chapter five

Map 5.1. Theories about Tupi-Guarani origins and early population movements (c. sixth to sixteenthth centuries). Map drawn by L. W. Sora. Sources: Brochado 1984; Métraux 1927; Heckenberger et al. 1998; Noelli 1998; Fausto 1992, 381–396.

By the early eighteenth century, the Brasílica had changed, according to observers, into a vulgar, or vernacular, lingua franca; by midcentury, it became known simply as “the Vulgar.” As for the Brasílica, it lost official status in Amazonia in 1757. Instead, residents of Amazonia spoke the Vulgar as their first language until as late as the nineteenth century (note that the Vulgar is known as Nheengatu or Língua Geral Amazônica in the specialized literature [see Edelweiss 1947; Bessa Freire 1983, 40, 46–49; Couto, chap. 3, this volume; Moore, chap. 4, this volume]). I argue in this chapter that the Brasílica evolved as a colonial lingua franca largely because its native-speaking communities, the Tupi-Guarani Indians, were undergoing processes of conquest and expansion. According to the specialized literature, the varieties spoken by ethnicities of

M. Kittiya Lee

[ 145 ]

Tupi-Guarani Indians spread as the lingua franca as a matter of historical serendipity. The geographical prominence along the coast and the major rivers brought native speakers of the language family into contact with the European colonizers. Furthermore, writers perceived that closely related features of the varieties facilitated the use of Tupi-Guarani in interethnic communications (Rodrigues 1986, 18, 29, 32). Hence, because Europeans’ encounters with Indians were often with the Tupi-Guarani, writers referred to the varieties as the língua geral, or the lingua franca of the indigenous inhabitants of Brazil and Amazonia (Soares de Sousa [1587] 1989, 44). I avoid using the term Língua Geral and the regional variants Língua Geral Paulista (LGP), Língua Geral Amazônica (LGA), and Nheengatu used by Hildo do Couto, Denny Moore, and other linguists (chaps. 3 and 4 in this volume; Edelweiss 1947; Bessa Freire 1983; Rodrigues 1986). My primary reasons for doing so are, first, that this chapter focuses on the early days when the Brasílica was used in interethnic communication before LGP and LGA/Nheengatu were documented. Second, my goal of identifying the Tupi-Guarani’s influence on colonial affairs applies equally to the Brasílica, LGP, and LGA/Nheengatu, since these are all languages based on Tupi-Guarani ethnic varieties. Third, as a historian, I prefer to leave questions of language classification—what linguistic features distinguish the Brasílica from LGP or LGA/Nheengatu?—to trained specialists. Instead, and in the interest of transparency, I employ in this chapter the most common name given by the colonial writers who authored the language records this study is based on and follow the original, broad meaning of the writers when they used “the Brasílica” to refer to the Brazilian lingua franca they learned to speak or heard spoken between linguistically dissimilar peoples. Doing so not only makes clear that my arguments apply specifically to the historical language documented by contemporary observers. It also elides any confusion provoked by use of the generic label língua geral, which colonial writers also used for other unrelated lingua franca. For example, in the eighteenth century, sufficient numbers of the slave and free black populace understood the African language they called Mina to warrant its function as lingua franca in the captaincy of Minas Gerais (Yai 2000; Castro 2002). Elsewhere in colonial Brazil, other African tongues also earned the same denomination as “língua geral” (Couto, chap. 3, this volume). Lastly, I understand the critical characteristics of “Língua Geral” as defined by linguists—its territorial dispersion and its common linguistic features—are shared by other languages and so are not unique features of the Brasílica. For example, that the native speakers of Romance tongues formed the largest part of the European settler community is reflected

[ 146 ]

chapter five

in the use of Portuguese, increasingly diffuse in towns they dominated (which were few and limited to the coastline) from the seventeenth century onward. However, observers never qualified Portuguese as a “língua geral,” as they did for the African language Mina. Moreover, Portuguese, Mina, and other languages used in communications between speakers of diverse languages never served in the colony-wide context, as the Brasílica did in rural, urban, and indigenous (including non-Tupi-Guarani) European, African, and mixed settlements, at least not during the colonial era. Portuguese and Mina prevailed, respectively, in the towns populated predominantly by the Portuguese and in the enclaves of Africans and their descendants within the settlements, plantations, and ranches run by the Europeans. (These enclaves can perhaps also be identified as “language islands,” to borrow a term used by Couto, chapter 3, this volume.) Thus, a broader focus of the current work questions why the Brasílica, over other important local or regional language varieties, dominated both colonies of Portuguese America during three centuries of European colonization. Section 2 of the chapter introduces the first written sample of the Brasílica: a 1519 vocabulary of the commodities exchanged between an unspecified ethnicity of Tupi-Guarani-speaking Indians and European mariners. The text, although rarely studied in the scholarship on Brazil, is significant as the earliest known record of any language of the land. For this chapter, its importance lies in its lexical content, which evinces the Tupi-Guarani linguistic affiliation of the Indians that scholars have long assumed but have not proved.2 Thus, this text establishes at an early date a commercial relationship uniting the Tupi-Guarani and the Europeans. Analysis of the vocabulary sheds light on the spoken communications, the conditions under which the lexical data was collected, and the material culture of the interethnic relations. These issues introduce the commerce in brazilwood that initially enticed Europeans to the colony. Read alongside coeval sources, they reveal the faintest outlines of what may be called a Tupi-Guarani ethos in outsider relations. This ethos propelled numerous clans of the language family to seek out powerful foreigners as allies, to establish themselves as indispensable to the success of their allies, to foster what became long-standing partnerships, and to maintain their positions as key players in the shifting tides of human history. Section 2 identifies this Tupi-Guarani spirit in the interrelations the Tupi-Guarani sustained with European and indigenous outsiders from the sixteenth through the eighteenth centuries. Although Brazil and Amazonia were colonized officially by the Portuguese, the broader term European is preferred because of the international dimension of the Spanish, French, Dutch, Irish, and English presence in and competition for the colonies. Representatives from these other European kingdoms fought

M. Kittiya Lee

[ 147 ]

each other and the Portuguese for control, almost always in partnership with Tupi-Guarani speakers. Section 3 then casts comparative light on the distinct behaviors of historical indigenous groups with foreigners. The juxtaposition of contrasting modes of interrelationships that American Indian language groups adopted with others—or the peoples they did not consider to be their own—prominently features an unmistakable TupiGuarani ethos of engagements with outsiders and strangers. Not too long ago, the Tupi-Guarani were themselves foreigners in new lands. The coastal prominence of the language family by the start of the colonial period resulted directly from the eastward relocations of the speech communities.3 As early as the sixth and seventh centuries, Tupi-Guarani peoples had begun to extend beyond original settlements. Map 5.1 summarizes two major schools of thought about the origins of the group and the routes they traveled to settle new territory. Generally, scholars concur that the Tupi-Guarani established themselves in the lands they crossed until they reached the Atlantic (Bruhns 1994, 266; Noelli 1998, 663; Scheel-Ybert et al. 2008, 768). By 1300, Tupi-Guarani speakers had established sites in the River Plate estuary, formed by the Paraná, Paraguay, and Uruguay Rivers (Bruhns 1994, 266). By the sixteenth century, they had dispersed over the most abundant lands of eastern South America (see map 5.2). Ethnic clans lorded over stretches along four thousand kilometers of the coast, lining the banks of the continent’s two largest fluvial systems, from the mouth of the Amazon River to the River Plate (Monteiro 1997, 977; Dean 1995, 29, 38). Around this time, the Tupi-Guarani began to receive the Europeans who arrived from transAtlantic crossings, ready to swap the wares they transported for local items.

2. The Seaside Barter for Brazilwood From 1500 until the 1530s, the Europeans traveled to the colony in search of brazilwood (Caesalpinia sappan) (Marchant 1942). Since the Middle Ages, European cloth makers had prized the dyes extracted from this tree, which yielded tones ranging from orange to vermillion to purple (Schneider 1978, 420). These shades were derivatives of the color red, associated with fire, ferocity, and fortitude, and for such reasons they were viewed as indicators of noble qualities at least by the fourteenth century. Royalty, courtesans, and the nobility favored garments tinged with deep, warm hues, which they wore to distinguish themselves from the drab, dark dress of ordinary folk (Gage 1993, 89; Wiesner-Hanks 2009, 108). When the Portuguese identified the dyewood in timber samples collected during the official Portuguese discovery of Brazil, King Manuel lost no

[ 148 ]

chapter five

Map 5.2. Population movements to/from indigenous settlements (early colonization: Brazil, 1500–1560, Amazonia, 1600–1650s). Map drawn by L. W. Sora.

time in monopolizing the trade. As early as 1502, the Portuguese king contracted New Christian merchants to explore and claim three hundred leagues of the coast annually, to build and maintain one new trading post (feitoria) every three years, and to deliver brazilwood on each return voyage (Marchant 1942, 29). Although the Treaty of Tordesillas (1494) had given Portugal a claim over the lands that later included Brazil and Amazonia, King Manuel honored contracts with other nations so long as they remitted a share of the profits to the royal treasury. But the lure of the greater wealth to be gained from contraband trade was too much to resist. The French were notorious in the illegal traffic of brazilwood; other important contenders included the Spanish, who moored regularly in Brazilian harbors en route to colonial possessions by the River Plate (Capistrano de Abreu 1997,

M. Kittiya Lee

[ 149 ]

52–68; Johnson and Silva 1992, 95; Staden [1557] 2008, 30–34, 37–38). Consequently, European traders returned heavy with logs after visiting Rio de Janeiro before 1510, Bahia in 1510 and 1526, Cabo Frio in 1511, and Pernambuco in 1520, 1522, 1526, 1527, and 1531 (Marchant 1942, 29–30). Such recorded voyages represented only a partial picture of what was a truly burgeoning trade accessed by the Europeans (see map 5.2). This was the context of European engagement with coastal Brazil into which the Italian Antonio Pigafetta stepped in December 1519. By then, the Bay of Guanabara had become known for more than the native groves of dyewood. Abundant ship-quality timber allowed the crewmen to fix problems that had affected the vessels since departure from Europe. Local communities willing to hew, shape, haul, and deliver logs, as well as to barter food, provided provisions for the ships destined for the rich profits to be made in the silk and spice trade in the Far East or the mineral wealth of the River Plate. Thus, the Bay became an ideal port of call. As a volunteer on the Spanish expedition commanded by the Portuguese navigator Ferdinand Magellan, Pigafetta was one of eighteen lucky survivors aboard the first known vessel to circumnavigate the globe (1519–22). Taking on the task of scribe, the Italian regularly jotted down the sights he saw and the names of things he heard about. The travel narrative he published around 1525 brought to life for European audiences the peoples, customs, material wealth, flora, fauna, and languages he encountered across the world.4 Among this information is the first known Tupi-Guarani vocabulary, in a document entitled, “Some Words of the Peoples of Verzin,” or Brazil.5 Although Pigafetta did not reveal the identity of his informants, the glossary suggests they were the Indians living around the Bay who spoke Tupi-Guarani varieties. Travel accounts in the 1530s and 1550s identify them by the ethnonyms Tupi, Tupinambá, and Tupinikin (Schmidel [t. 1535] 1997, 20; Thévet [1575] 1953, 225; Léry [1578] 1990, 56; Staden [1557] 2008, 105).6 Table 5.1 is based on the modern transcription of the Tupi-Guarani vocabulary included in the Beinecke-Yale manuscript (Pigafetta [t. 1519] 1994, 17–18). Scholars consider the French-language document to be the most accurate version of the original.7 The left column lists the English terms (translated from French in Pigafetta [t. 1519] 1994, 45), and the right column lists the words that the Italian writer presumably heard in the barter between the Indians and the Europeans. The narrow commercial focus of the eight terms befits the interactions Pigafetta must have witnessed in an era when the European presence remained tied largely to the brazilwood trade. The document provides a unique perspective regarding the earliest of interethnic trade relations. However, it remains underutilized in the scholarship on Brazil, perhaps because the work de-

[ 150 ]

chapter five

Table 5.1. The first known Tupi-Guarani vocabulary [corn,] millet flour fish hook knife comb scissor bell good, better

maiz huy pinda [i]taisse chiguap pirame iteumaraca [ga]tum, maragatum

Source: “Some words of the Peoples of Verzin,” in Pigafetta [t. 1519] 1994, 45.

votes only a few pages to describing Portuguese America, or perhaps because specialists question the reliability of the information it contains, a matter I discuss below. Instead, the specialized literature draws from later works that focus exclusively on Portuguese America. Pigafetta’s vocabulary, compared to the accounts by French royal cosmographer André Thévet (1556), German artilleryman Hans Staden (1547–48, 1549–55), and French Calvinist pastor Jean de Léry (1556–58), is the shortest in length. In all four documents, the amount of linguistic information provided correlates with the time each author passed in the colony. Thévet’s ten-week visit surpassed Pigafetta’s by two months: hence it contains several pages of ethnographic data about the Indians, including a transcription of Christian doctrine in the Tupinambá language. Léry, who resided in the same fledging settlement as Thévet, interacted with the Tupinambá through French interpreters for “about a year or so” (Léry [1578] 1990, 56). Accordingly, he was able to publish an eighteen-page “Colloquy”8 made up of words and dialogues about salutations, colors, animals, plants, and cultural and material items (178–195). The “Colloquy” forms one chapter in History of a Voyage to the Land of Brazil, which itself is peppered with Tupi-Guarani terms and anecdotes about conversations with its speakers, scenes witnessed among them, and comments on their habits. Staden’s True History and Description of a Country Populated by a Wild, Naked and Savage Man-munching People, situated in the New World, America was composed following two voyages. After a short first trip, Staden remained for nine years the second time; during that period he lived as a prisoner-of-war among affiliated Tupinambá clans around Bertioga and São Vicente. The extensive linguistic content of Staden’s work reflects the fluency that he attained during a time when the only languages he heard were the ethnic

M. Kittiya Lee

[ 151 ]

tongues of each clan and, in far fewer instances, French, which he did not speak, and Portuguese, which he had acquired (Staden [1557] 2008). Whereas the work by Staden reflects linguistic acquisition through immersion and long-term residence, Pigafetta’s “Some words” captures the low proficiency of most European visitors. Travel literature indicates that the merchants and crewmen tended to sleep, eat, and tend to affairs on the vessels when calling at the Brazilian harbors (Caminha [1500] 2002; Schmidel [t. 1535] 1997). Hence, the Europeans interacted with the Indians over the course of days or on and off throughout the weeks necessary to collect wood and replenish supplies. Such limited exposure is unlikely to have resulted in the Europeans’ adoption of the language, excepting individual talents. The experiences of the “master’s mate,” who was a repeat visitor to the colony, is typical of most Europeans. When called upon by the crew to mediate in the barter with Tupinambá Indians, he could “stammer out a few words in [the Tupinambá] language” (Léry [1578] 1990, 26). Gestures and facial expressions must have been helpful, and, along with them, lists like “Some words” introduced key terms and phrases to uninitiated travelers anticipating direct and profitable conversations with the Indians. Notably absent in the glossary is the item that so motivated the Portuguese, French, and Spanish to cross the ocean. Perhaps the Indians had become all too familiar with the European craze for brazilwood, so that there was no need to name it. This was the case in 1531, when an unspecified group of Indians invited Pero Lopes de Sousa to trade with an impressive show of athletic prowess. Swimming with the swift pace of the ship on which the Portuguese rode, the Indians asked about his wish to trade for brazilwood—although the author does not specify exactly how they did so (Marchant 1942, 33). In Pigafetta’s case, the Indians who met with the crew may have simply arrived bearing logs to hand over in return for European merchandise. Pigafetta may not have heard the Tupi-Guarani term for the wood if the haggling for its exchange occurred before he arrived. That the ship Bretôa waited for deliveries of brazilwood from April 17 until May 12, 1511, exemplifies the lengthy period required for the logs to be prepared. Indian and African workers spent almost one month identifying and traveling to the grove of brazilwood trees, felling them with the metal tools provided by the Portuguese, shaving the bark, dividing the trunks into sections, and rounding the logs (Marchant 1942, 37). The logs then had to be carried to the factory or trading post (feitoria) near the coast, presumably by the same laborers (34–41). If Magellan purchased as many logs as the Bretôa did, then the wait time would have exceeded the two weeks that Pigafetta remained in the colony.

[ 152 ]

chapter five

The final explanation for the absence of “brazil” on Pigafetta’s list is the high probability that Magellan followed the commercial protocol established by the Portuguese. By law, a Crown-appointed agent known as the “factor” (feitor) was charged with negotiating all affairs between the Indians and the Europeans. His tasks included identifying profitable goods, persuading local headmen to trade, negotiating prices with Indian merchants, and arranging for the delivery of the timber to the factory, where the factor stored the goods until the arrival of Portuguese traders. Therefore, local knowledge was crucial to the factor’s ability to realize the obligations. Although linguistic aptitude was not a requirement, the biographies of the few known factors indicate fluency in the local speech. In some cases, the men were selected from the tiny population of Europeans already living on the land as survivors of shipwrecks, as deserters from earlier ships calling at port or as exiled convicts (Johnson and Silva 1992, 93; [Lopes de Sousa] [1530–32] 1968, 51; Marchant 1942, 35, 38–39; Couto 1995, 194–195; Cortesão et al. 1960–62, 43). Although the factor had been transferred prior to Magellan’s arrival, the commander called upon João Lopes de Carvalho, a Portuguese sailor who learned the local language variety during years of residence in the Bay of Guanabara (Parr 1953, 282). The glosses for foodstuffs and merchandise in “Some words” suggest the continuing fame of the Bay of Guanabara as a trading post among European mariners. That the vocabulary features comestibles and small trade items infers that the factor bartered for the major commodity of brazilwood while the sailors negotiated directly with the Indians for personal consumption. For example, sixteenth-century travel writers regularly mentioned manioc flour, a staple of coastal indigenous societies, as an important provision for ships, which often ran low on food rations after the two to three months typically required to cross the Atlantic Ocean (Caminha [1500] 2002, 56; Staden [1557] 2008, 27, 30, 38; Léry [1578] 1990, 26, 69–72). The second starch listed in the vocabulary presents the single linguistic anomaly in a document otherwise restricted to TupiGuarani lexica. Pigafetta notes for ‘corn’ (Zea mays) the word maiz, a term of Arawak origin. Coastal speakers possessed their own word for the plant, so loan words are improbable (Léry [1578] 1990, 71–72). Moreover, no Arawak clans were known to have occupied the eastern shores (Urban 1992, 95), nor is the Arawak word documented at any other time in the sixteenth century by coastal Indians. The most plausible explanation would be an oversight on the author’s part. Scholars have suggested that portions of the narrative drew from the writings of Christopher Columbus, the first European to document American corn, and of Amerigo Vespucci, who introduced to European readers many Indian-language terms.

M. Kittiya Lee

[ 153 ]

As the authors of well-circulated travel accounts about the New World in the late fifteenth and early sixteenth centuries, both men popularized words used by the Arawak-speaking Taino of the Caribbean islands (Pigafetta [t. 1519] 1995, xxiii; Warman 2003, 12, 37).9 Although portions of the ethnohistorical information in Pigafetta’s First Voyage around the World may be unreliable for this reason, my own comparison of “Some words” with records of Tupinambá, Tupinikin, and Brasílica vocabularies from 1540 until 1795 confirms what is otherwise the faithful recording of Tupi-Guarani lexica in “Some words.” The Italian writer specifies by name the merchandise that the Europeans brought with them in anticipation of barter with the Indians. “Fish hook,” “knife,” “comb,” “scissor,” and “bell” were prized by the autochthonous societies that did not possess metal implements prior to trade with the Europeans. Their standard reference in sixteenth-century accounts identifies the items as the staples of Indian-European seaside commerce (Staden [1557] 2008, 76). However, their notable absence from a 1500 report of the first known encounter between coastal Indians and Portuguese mariners reveals that by 1519 an evolution had occurred in the class of goods supplied by the Europeans. The metal tools in Pigafetta’s glossary were more expensive for the Europeans to acquire than the scraps of clothing and “worthless old” headgear (plucked from the very heads of the Portuguese sailors) that they gave to the Indians in 1500 (Caminha [1500] 2002, 46, 48, 53). It appears that the Europeans’ hunger for brazilwood—not to mention the actual starvation they endured as food rations diminished toward the end of the long Atlantic crossing—drove them to accommodate the demands of their indigenous commercial partners. Something can be said about Tupi-Guarani ingenuity in packaging their services into the sale of the item so hotly contested by the Europeans. By demanding the tools and by providing the labor necessary to fell, chop, and transport the logs, the Indians themselves became indispensable to all steps of the commercial transaction. Perhaps, too, the Europeans attempted to conserve whatever leverage they could by refining their persuasion skills. For instance, Pigafetta picked up the phrase “very good,” which he might have heard in sales pitches to Magellan’s crew, much in the fashion of the enthusiastic Indian vendor reported in the 1550s (Léry [1578] 1990, 27). The brevity of “Some words” further suggests that limited conversations between the sailors and the Indians occurred as they swapped products for personal consumption. It served as a reference guide for first-time sailors to know what foods to request. Staden witnessed one such conversation: “When the ships come to bargain in this manner, one or two savages row out in a canoe and hand over their goods as fast as they can.

[ 154 ]

chapter five

Then they demand what they want in return, which the Portuguese then give them” (Staden [1557] 2008, 76). André Thévet also recalled transactions in “this manner [of speech]: ‘give me that, and I will give thee this,’ without any further talk” ([1575] 1953, 225). The modest lexica, consisting of limited grammar and a few terms adequate for efficient identification of the trade commodities, exhibit the features linguists have identified as common to many trade languages (Crystal 1997, 336, 439). This low level of proficiency, best represented by itinerant European merchants or sailors, contrasts with the fluency observed in men like Hans Staden, João Lopes de Carvalho, the anonymous factors referenced in historical documents (Johnson and Silva 1992), and the many unnamed mestizo interpreters born of Tupi-Guarani mothers and European fathers (Moore, chap. 4, this volume; Metcalf 2005). Unique to the Brazilian case is the exclusive reliance on indigenous lexica. “Fish hook,” “comb,” “scissor,” and “knife” are named in the local tongue, suggesting that borrowings from European languages were not widespread even for objects that Europeans introduced. Even in the second half of the century, when the European presence had grown frequent, interethnic transactions occurred in the Tupi-Guarani coastal varieties (Cardim [1583] 1980; Knivet [t. 1591] 2008; Anchieta [1595] 1933). The absence of recognizable European loanwords stands in contrast with what has been described in the European trade colonies in Africa and in the Pacific in the nineteenth century. Though separated by three hundred years, colonies in sixteenth-century Portuguese America and in nineteenth-century Africa and the Pacific Ocean shared commonalities. Both were situated along trade routes and near forts. Encounters between the natives and the Europeans were largely commercial, similarly occasional, and depended on the length of time that each ship moored nearby. But unlike early Portuguese America, interethnic spoken communication in the African and Pacific colonies took place in the nonstandard vernaculars of European merchants, which the native traders acquired expressly for commercial purposes (Mufwene 2001, 8).

3. A Tupi-Guarani Ethos of Partnership: A Historical Survey The European reliance on the Tupi-Guarani varieties continued as a long-standing practice throughout the sixteenth century. Their initial use evolved as a practical consequence of the Europeans’ desire for brazilwood, which limited affairs involving the Indians to the stretch of the shore where it grew natively. Recall that these were the lands recently occupied by settler Tupinae and Tupinambá, whose descendants ventured out to meet the first Europeans along the coast. Inland and farther north,

M. Kittiya Lee

[ 155 ]

interpreters who accompanied European writers into the Amazon entered landscapes dominated by the same language group (Medina [1542] 1988). Diogo Nunes likened the customs and languages of the riverine inhabitants he encountered around 1538 to that of the Tupi-Guarani by the coast, where he was born and raised (Nunes [c. 1552] 1993, 33; [c. 1552] 2000, 4). The Tupi-Guarani situation along the Atlantic shore, the Amazon River, and the River Plate, all navigable waterways accessible to seagoing European vessels, determined that the experiences of the Europeans almost always involved Tupi-Guarani speakers, a matter reflected in all sixteenth- and early-seventeenth-century travel literature. Although it must have seemed to the Europeans that everywhere they explored, they encountered Tupi-Guarani speakers, in fact, by the 1530s, explorers had learned about non-Tupi-Guarani Indians. During the 1530–32 expedition to erect the first Portuguese settlement of São Vicente, Pero Lopes de Sousa noted that the speech of the Gê language family (see map 5.2) sounded like “the chatter of the Moors [of North Africa or living in Iberia],” and the Portuguese claimed they “did not understand [it], nor was it like that of Brazil [the Brasílica]” ([Lopes de Sousa] [1530–32] 1968, 78, 97). These first encounters were fleeting, as were all subsequent meetings between the Gê and the Europeans for one century following recorded contact (Pompa 2003, 202–206; Puntoni 2002, 52–53; Langfur 2006, 24, 56). The contrast between the Tupi-Guarani and the Gê modes of interaction with outsiders, a matter discussed below, emphasizes the Tupi-Guarani ethos, which was evident in their consistent willingness to make and sustain contact with the Europeans and to render themselves crucial to the success of European endeavors. The historiography on colonial Brazil has interpreted this Tupi-Guarani spirit as one of amicability. Narrations of the early colonial phase describe the “friendly” Tupi-Guarani Indians who helped the Portuguese gain a foothold in the colony despite the threats posed by competing European kingdoms (Capistrano de Abreu 1997; Johnson and Silva 1992; Marchant 1942).10 Once the Portuguese founded permanent settlements (the first, in 1531–32 in São Vicente, by permission of the Tupinikin of Piratininga) and erected a provincial capital at Salvador in 1549 (a task enabled by the Tupinambá of the Bay of All Saints), the introduction of sugarcane prompted authorities and colonists to turn to African slavery (Schwartz 1986). In addition to the epidemics that decimated Indian populations, war, enslavement, environmental degradation, and territorial encroachment, other upheavals brought on by European occupation induced the remaining indigenous societies to flee. By the end of the sixteenth century in the urban centers of Brazil, autonomous Tupi-Guarani Indians had been exterminated, had escaped inland, or had been absorbed into

[ 156 ]

chapter five

the fold of descendants claiming mixed indigenous, European, and African heritages. From the seventeenth century onward, the former lords of the coast were relegated to remote hinterlands (see map 5.3); in history books, too, they were glimpsed only in the foliage and on the rivers of Amazonia, or scattered throughout the vast interior. In this manner the traditional historiography about the late colonial period presents the involvement of the Tupi-Guarani Indians in colonial affairs as having diminished greatly from the influential roles they had once played.

Map 5.3. Expeditions of contact, exploration, and enslavement (c. 1550s to 1750s). Map drawn by L. W. Sora. Sources: Monteiro 1994; Sweet 1974.

M. Kittiya Lee

[ 157 ]

Archaeologists and anthropologists have built on this historical narrative of the disappearance of the Indians from the core regions of the colony. They have characterized as “migration” or “expansion” the population movements of the language group before and after the arrival of the Europeans (Noelli 2008, 660–663). The westward migrations of the TupiGuarani away from the coast are viewed as the Indians’ abandonment of original settlements, which had fallen under European control starting in the mid-sixteenth century (660). By contrast, the geographic proliferation referred to as “expansion” is said to have occurred in pre-Columbian times when Tupi-Guarani movements resulted in demographic growth, the breakup of villages, and forestry management, which brought speakers to the Atlantic coast while they still retained influence over original sites (Brochado 1984; Noelli 2008, 660). The work of ethnohistorians, colonialists, and scholars of the indigenous peoples of Brazil have provided fresh perspectives on the degree to which the Tupi-Guarani, despite or perhaps because of their mobility, created, sustained, and determined the outcome of colonial affairs (Carneiro da Cunha 1992; Metcalf 2005). Monteiro (1997) argues that the seventeenth-century expansion of the colony’s frontier and the indigenous slave market, the economic engine of the captaincy of São Paulo, relied extensively on individuals who successfully played off the political rivalries that divided Tupi-Guarani clans, unaffiliated native peoples, Luso-Brazilians, the Portuguese, and the Spanish (see map 5.3). Scholars of the Jesuits have emphasized the importance of Tupi-Guarani Indians in creating and renewing affiliations between the clans of the language family and the Catholic missionaries that endured for two and a half centuries throughout the far-flung provinces of Portuguese America. In particular, the elevation of the Tupi-Guarani languages from the mother tongues of coastal Indians to the unofficial lingua franca of the colony and the subsequent proliferation of grammars and catechisms in the Brasílica underscore the ongoing involvement throughout the sixteenth and seventeenth centuries of native speakers of Tupi-Guarani with religious agents of European colonization (Araújo 1618; Araújo and Leam 1686; Figueira 1621, 1687; Bettendorff 1678; Leite 1938–50; Moore, chap. 4, this volume; Couto, chap. 3, this volume). Native speakers led campaigns known as “flying missions” (missões volantes), which crisscrossed the colony and introduced Christianity and the Brasílica to indigenous peoples who had hitherto lived far from the Portuguese settlements and Jesuit missions. Thus, they edged the latter into the fold of colonial authorities (Leite 1938–50, vol. 2; Barros 2001; Pompa 2003; Bessa Freire 2004). These works emphasize the consistency with which Tupi-Guarani peoples sustained engagements with the Portuguese and, later, with settlers and

[ 158 ]

chapter five

colonial authorities. As they did so, member clans of the language family remained central and in many cases enabled the continued Portuguese expansionist projects in the Americas, even in the late colonial era. But it would be difficult to argue that the Tupi-Guarani invested solely in the Portuguese. In fact, if the Portuguese believed that the TupiGuarani were everywhere to be found, other indigenous peoples must have concluded that wherever the Tupi-Guarani arrived, the Europeans and Luso-Brazilians soon followed (see maps 5.2 and 5.3). The collaboration between numerous Tupinambá tribes and the French allowed the latter a generous helping of the profits from the illegal traffic of brazilwood around the Bay of Guanabara in the 1540s and 1550s and near Belém and São Luís in the 1600s (Daher 2007; Abbeville 1945). Recall also André Thévet and Jean de Léry, the French clergy who accompanied the 1550s attempt to erect the settlement of “France Antarctique.” The French presence in the Bay of Guanabara would not have been possible except for the permission granted by the local Tupinambá. Throughout the 1640s, Potiguar Indians and the Dutch carved out colonies in the northeastern captaincies of Paraíba and Pernambuco, much to the alarm of the Tupinambá and their Portuguese allies (Capistrano de Abreu 1997, 69–90). Near the River Plate from the late sixteenth until the mid-eighteenth century, the Carijó (or the Guarani, as they were known in the literature on colonial Paraguay) interacted with settlers, enslavers, and church and colonial authorities on both sides of the frontier that divided Portuguese and Spanish America (Monteiro 1994; Ganson 2003). And clans in parts of Amazonia welcomed Dutch, French, and Irish traders and colonists throughout the seventeenth and eighteenth centuries (Evreux [1614] 1929; Sweet 1974; Lorimer 1989; Farage 1991; Dominguez 2000). While it may appear that the Tupi-Guarani sought to associate with the Europeans to compete against indigenous neighbors and foes, or against Portuguese and Luso-Brazilian territorial encroachment, it is important to remember that for the most part, the battles played out between member tribes of Tupi-Guarani languages. As colonial writers and modern historians and anthropologists have shown, the bellicose relations that marked affairs between clans were and continue to this day to be distinct features of the Tupi-Guarani, recorded by observers from the sixteenth through the twentieth centuries (Staden [1557] 2008; Soares de Sousa [1587] 1989; Bettendorff [1694–98] 1990; Fernandes 1970; Clastres 1995; Viveiros de Castro 1992). Perhaps the Tupi-Guarani commitment to sustained and meaningful relationships with the Europeans reflects their desire to preserve themselves as the lords of the land, which they believed themselves to be. (For a similar phenomenon elsewhere, see Ball, chap. 10, this volume.)

M. Kittiya Lee

[ 159 ]

The rivalries internal to the language family often did not characterize Tupi-Guarani relations with non-Tupi-Guarani Indians. The history of the Tupinambá of coastal Pernambuco echoes the language family’s tendency toward alliance with outsiders. According to longtime Amazonian resident and Jesuit missionary Father Felipe Bettendorff, the residents of eighty-four Tupinambá villages left Pernambuco between the 1520s and the 1530s (see map 5.2). Seeking refuge from the slaving, violence, ecological destruction, and other mayhem caused by Portuguese colonization, they pressed westward into Amazonia, eventually settling the banks of the Madeira River (see Tupi-Guarani migration on map 5.2). The whip of an angry Spanish man set them in flight again, this time to an island located twenty-eight leagues from the mouth of the Madeira (Bettendorff [1694–98] 1990, 56–57).11 There, the Tupinambá inserted themselves into local political and commercial circuits, forged commercial alliances with the “seven Provinces” of peoples to their north, and collected tribute from the “Mataieces” (57; Maranhão 1947, 196).12 Through intermarriages, they linked their own lineage with those of their neighbors (Fernandes 1963, 55). Neighboring native societies must have understood the benefits to be gained by allying themselves with the expansionistic, warrior peoples. Around 1613, Caeté headman Arraia Grande expressed regret, saying that his people would have fared better had the Tupinambá been their allies (Evreux [1614] 1929, 172). Where confrontations occurred, the Tupinambá gained the respect and fear of their neighbors, who watched as they “consum[ed] entire nations [of resistors]” in warfare (Bettendorff [1694–98] 1990, 57). The integration of Tupi-Guarani newcomers into regional Amazonian networks calls for revision of the expansion/migration concepts scholars have used to categorize the territorial relocations of the language group through time. Although the Tupinambá of coastal Pernambuco left their homes, they created relationships with neighbors that heightened their social, political, and economic status in the new human landscape. This spirit of partnership forged with outsiders and foreigners—in contrast to the warring nature that characterized the interactions between clans— forms part of the elusive Tupi-Guarani ethos that anthropologists have sought to identify (Fernandes 1963; Clastres 1995; Viveiros de Castro 1992). That the Portuguese followed closely along the trails blazed by the Tupi-Guarani throughout the lowlands of eastern South America (Sampaio 1987) raises the issue that Tupi-Guarani expansion remained in full swing throughout the colonial era and in fact determined where, when, and to some degree how the Europeans sought to fulfill their imperial ambitions. The tendency of the Tupi-Guarani to make themselves essential part-

[ 160 ]

chapter five

ners is especially salient when viewed in contrast with the actions of the Gê-speaking Indians featured in Hal Langfur’s study of the Eastern Sertão.13 From the 1750s until the 1830s, colonization of the region brought “local elites, slaves, impoverished settlers from other parts of the colony, and seminomadic indigenous peoples into a contest for land, labor, and resources, radiating outward from the mining district’s major towns” (Langfur 2006, 9). Especially in the second half of the 1600s, such contests broke out as armed conflict, war, raiding, and kidnapping. Though “unequal and often coerced . . . [these contests remained] intensive exchange nevertheless” and formed “an essential means of [interethnic] communication and exchange” (214, 261). Langfur’s study suggests another means by which Indians who met with outsiders chose to relate and to incorporate the latter, by interrelationships of avoidance, violence, and competition. The ancestors of the Gê Indians of the Eastern Sertão are believed to have been the non-Tupi-Guarani Indians who fought Tupinaé and Tupinambá invaders on the coast prior to the arrival of the Portuguese in 1500. Gê territorial occupation interrupted Tupi-Guarani coastal dominance, and as dwellers by the sea, the former Indians had made contact with European visitors by the 1530s (recall the Portuguese observer who compared their speech to that of North African Moors). Because the Gê also lived by the groves of brazilwood, they were well placed to compete against the Tupi-Guarani in establishing trade agreements with the Europeans, had they so desired (see maps 5.2 and 5.3). Yet they did not, because they conceptualized exchange with the other in terms of avoidance and violence, as testified by sixteenth-century writers (Léry [1578] 1990, 29). Hence, one may even argue that the Tupi-Guarani became the favored indigenous trading partners for the Europeans simply because other Indians shunned the opportunity.

4. Conclusion: Language and Conquest In this chapter I have attempted to argue for a Tupi-Guarani epoch of conquest and expansion, one driven by a spirit particular to the language family in its manner of engaging with and confronting outsiders. Although in general, human relations invariably span a range of acts from accommodation and collaboration to avoidance and confrontation, what distinguishes the Tupi-Guarani is the tendency to define themselves as essential to the successes of their partners and to renew and sustain these alliances through the long term. In the case of the coastal Tupi-Guarani, their arrival to the shores of eastern South America shortly before the

M. Kittiya Lee

[ 161 ]

landing of the Europeans on the same lands resulted in associations that bound the groups for over two centuries. In arguing for an indigenous age of conquest and for the evolution of the Brasílica as an expression of that Tupi-Guarani expansion, this chapter may seem at odds with the ideas articulated in other chapters. In fact, I have sought to understand how Tupi-Guarani conquerors and migrants recognized and incorporated the European outsiders, who were also in the midst of political and territorial expansion. In all instances, the central positions played by the Tupi-Guarani in the early sixteenth century evolved into long-standing relationships wherein the Europeans depended on Tupi-Guarani speakers to achieve their imperial ambitions. In the different and far-flung lands of Portuguese America in the seventeenth and eighteenth centuries, configurations of alliance divided specific ethnicities of Tupi-Guarani speakers among themselves but united them with what each clan considered to be the most prominent of the Europeans. The episodes of collaboration, whether military, commercial, religious, or political, reveal that the distinct ethnicities of the TupiGuarani language group colluded with the imperial and colonial designs of the European rulers, colonial authorities, and settlers and, later, of Luso-Brazilians. Underlying the concerns discussed here is the larger issue that questions why the Brasílica endured as the most important lingua franca for the two colonial provinces of Portuguese America. Recall that in Brazil, the lingua franca fell second to Portuguese by the late colonial era; in Amazonia it persisted as the first language of most families and as the primary lingua franca for all who lived in colonial villages and towns. While it might be true that the Brasílica spread as far and for as long as it did under the patronage of Iberian colonial and church authorities, the long Tupi-Guarani history of engagement with foreigners suggests an additional reason. Its expansion from the sixteenth through the nineteenth century serves as one more reflection of the Tupi-Guarani success in continually defining and redefining themselves as essential actors whose participation would ensure success in local human endeavors. Notes Part of the writing and research for this chapter was made possible by the Mellon Postdoctoral Fellowship in the History Department at the University of Chicago, the American Historical Association, the Center for Historical Research at Ohio State University, and the National Endowment for the Humanities at the John Carter Brown Library. The sources consulted for map 5.1 include Brochado (1984), Métraux (1927), Heckenberger et al (1998), Noelli (1998), and Fausto in Carneiro da Cunha (1992, 381–396) and for

[ 162 ]

chapter five

map 5.3 are Monteiro (1994), and Sweet (1974). I am grateful to the participants of the conference, especially to Salikoko S. Mufwene and Bruce Mannheim, as well as the anonymous reviewers, for their probing comments and generosity of time. Their input has made this chapter a much better work; any errors or omissions herein are fully my responsibility. 1. For all historical sources written during the colonial era from 1500 to 1822, I have placed in brackets the original year of writing, if it is known, or the year of first publication. If no date is available for the year of writing, or if the text remained as a manuscript until publication in the nineteenth through twenty-first centuries, then the year of travel to the colony is given in brackets. The letter t serves to distinguish the date of travel from the year of writing or publication. Following each bracketed date is the year of the modern publication consulted. In order to not confuse readers with the diversity of sixteenth- and seventeenth-century spellings of indigenous groups’ names, I have made use of modern Luso-Brazilian orthographic conventions for the names of Indian clans, Portuguese settlements, and geographic features; in bibliographic citations and in quoted material, the original spelling is maintained. 2. The 1519 year of the vocabulary predates the usual start date for the historiography, which begins earnest discussion of the colonial period in the 1550s (and in some cases, in the 1530s or the 1580s). If Indians in the early sixteenth century are addressed, they are considered the ancestors of the Indians living in the same region and identified by mid-sixteenth-century writers (see discussion in the chapter about André Thévet, Jean de Léry, and Hans Staden). 3. I follow the specialized literature (for example, Brochado [1984]; Rodrigues [1986]; Bessa Freire [2004]) and regard the Tupi-Guarani language family as a single entity and linguistic class that brings together smaller speech communities that reflect ethnic and regional varieties. For this reason, I refer to the Tupi-Guarani language family in the singular (i.e., the Tupi-Guarani language family expands) but use the plural when mentioning the peoples, speakers, clans, and so forth, of the linguistic affiliation (i.e., the Tupi-Guarani Indians settle). 4. Pigafetta’s record was titled Le voyage et navigation faict par les Espaignolz es isles de Mollucques, des Isles quilz ont trouve audict voyage, de roys dicelles, de leur government et maniere de vivre, avec plusiers autres choses and carried the following approximate translation in English: The Voyage and Navigation of the Spaniards among the Moluccas, the Islands that they found during said voyage, together with many other things (Pigafetta [t. 1519] 1995, xlvii). Subsequent partial and complete editions of the work were published under different titles. 5. The edition printed in Venice in 1536 (Pigafetta 1536) carries the title “Alcune parole che vsano le genti ne la terra del Bresil” (“Some words that the people of the land of Brazil use”). In addition to the Tupi-Guarani vocabulary, Pigafetta’s First Voyage around the World (Voyage and Navigation) also includes word lists in the following languages: Tehuelche of Argentina and Patagonia; Visayan of the south Pacific and the Philippine Islands; and Malay of the Malay Straits (Pigafetta [t. 1519] 1995, 10, 12, 55). 6. These Tupinambá were probably neither the same peoples as the migrants mentioned in this chapter’s introduction, who settled in the Bay of All Saints before the arrival of the Portuguese, nor the Tupinambá discussed later in this chapter who fled from Pernambuco to Amazonia. Scholars remain uncertain about the affiliations of the numerous clans that colonial writers named equally as “Tupinambá,” and scholars have sought to distinguish between them according to geographic location; the ones

M. Kittiya Lee

[ 163 ]

referenced by this note are the Tupinambá of the Bay of Guanabara. In the seventeenth and eighteenth centuries, writers began to use the term generically to refer to all Indians (see Moore, chap. 5, this volume). 7. Though written in French and not in Pigafetta’s native Italian, the BeineckeYale manuscript is thought by specialists to be an authentic work, the one composed closest to the dates (March 1523 to April 1524) when the original manuscript was written (Pigafetta [t. 1519] 1994, 17–18, 27). If these estimations are correct, then Pigafetta would have drafted the Beinecke-Yale document nearly one to two years after returning to Europe. 8. The full title of the chapter is “Colloquy upon the entry or arrival in the land of Brazil among the people of the country called Tupinamba and Tupinikin: In the savage language and in French” (Léry [1578] 1990, chap. 20). 9. Additional words throughout Pigafetta’s narrative support the hypothesis: the words king, canoe, and hammock are of Arawak origin. That the Arawak terms do not appear in subsequent Tupi-Guarani vocabularies suggests that Pigafetta borrowed from the works of Columbus and Vespucci, either intentionally or accidentally. 10. This perspective is not limited to historians. See, for example, Rodrigues (1986) and Couto (chap. 3, this volume). 11. This is probably the island known as “Ilha dos Tupinambarana.” 12. The tribute-paying peoples called “Mataieces” by Bettendorff may be the ancestors of the modern-day Pano-speaking Matís Indians or the Matsés Indians, all settled near to the Javari River. 13. The lands known during the colonial era as the Eastern Sertão today make up the state of Minas Gerais and the borders it shares with the states of Rio de Janeiro, Espírito Santo, and Bahia.

References Abbeville, [Friar] Claude d’, O.F.M. [1614] 1945. História da missão dos Padres Capuchinhos na Ilha do Maranhão e suas circumvisinhaças. São Paulo: Livraria Martins Editora. Anchieta, José de. [1595] 1933. Arte de gramática da Lingua Mais usada na costa do Brasil. Rio de Janeiro: Imprensa Nacional (facsimile edition). Araújo, Antonio de, S.J. 1618. Catecismo na lingoa brasílica. Lisboa, Portugal: Pedro Craesbeeck. Araújo, Antonio de, S.J., and Bertholameu de Leam, S.J. 1686. Catecismo Brasilico da Doutrina Christãa, com o ceremonial dos sacramentos, & mais actos parochiaes. Emendado nesta segunda impressaõ Pelo P. Bertholameu de Leam da mesma Companhia [Emended in this second edition by P. Bertholameu de Leam of the same Company (of Jesus)]. Lisboa, Portugal: Miguel Deslandes. Barros, Maria Cândida Drumond Mêndes. 2001. The office of Língua: A portrait of the religious Tupi interpreter in Brazil in the sixteenth century. Itinerario 25, no. 2: 110–140. Bessa Freire, José. 1983. Da ‘fala boa’ ao português na amazônia brasileira. Amerindia 8:39–83. ———. 2004. O Rio Babel: A história das línguas na Amazônia. Rio de Janeiro: Editora da Universidade do Estado do Rio de Janeiro, Editora Atlântica. Bettendorff, João Felipe, S.J. 1678. Compendio da Doutrina Christam Na lingua Portu-

[ 164 ]

chapter five

gueza, & Brasilica: Em que se comprehendem os principaes mysterios de nossa Santa Fe Catholica, & meios de nossa Salvação. Lisboa, Portugal: Na Officina de Miguel Deslandes. ———. [1694–98] 1990. Crônica da missão dos padres da Companhia de Jesus no Estado do Maranhão. 2nd ed. Belém: Fundação Cultural do Pará Tancredo Neves; Secretaria de Estado da Cultura. Brochado, José Joaquim Justiniano Proenza. 1984. An ecological model of the spread of pottery and agriculture into eastern South America. PhD diss., University of Illinois at Urbana-Champaign. Bruhns, Karen Olsen, ed. 1994. Ancient South America. New York: Cambridge University Press. Caminha, Pero Vaz de. [1500] 2002. First letter from Brazil. In Colonial Latin America: A documentary history, ed. Kenneth Mills, William B. Taylor, and Sandra Lauderdale-Graham, 43–58. Wilmington, DE: SR Books. Capistrano de Abreu, João. 1997. Chapters of Brazil’s colonial history, 1500–1800. Trans. Arthur Brake. Oxford: Oxford University Press. Cardim, Fernão. [1583] 1980. Tratados da terra e gente do Brasil. Belo Horizone: Itatiaia; Editora da USP. Carneiro da Cunha, Manuela, ed. 1992. História dos Índios no Brasil. 2nd ed. São Paulo: Companhia das Letras/Secretaria Municipal de Cultura/FAPESP. Castro, Yeda Pessoa de. 2002. A língua mina-jeje no Brasil: Um falar africano em Ouro Preto do século XVIII. Belo Horizonte: Fundação João Pinheiro, Secretária da Cultura do Estado de Minas Gerais. Clastres, Hélène. 1995. The land-without-evil. Tupí-Guaraní prophetism. Trans. Jacqueline Grenez Brovender. Urbana: University of Illinois Press. Cortesão, Armando, A. Teixeira da Mota, and Comissão Executiva das Comemorac̜ões do V Centenario da Morte do Infante D. Henrique, eds. 1960–62. Portugaliae monumenta cartographica. 6 vols. Lisboa, Portugal: Imprensa Nacional-Casa de Moeda. Couto, Jorge. 1995. A construção do Brasil: Ameríndios, Portugueses e Africanos, do Início do povoamento a finais de quinhentos. Lisbon: Cosmos. Crystal, David, ed. 1997. The Cambridge encyclopedia of language. 2nd ed. Cambridge: Cambridge University Press. Daher, Andrea. 2007. O Brasil francês: As singularidades da França equinocial, 1612– 1615. Rio de Janeiro: Civilização Brasileira. Dean, Warren. 1995. With broadax and firebrand: The destruction of the Brazilian Atlantic forest. Berkeley: University of California Press. Domingues, Angela. 2000. Quando os índios eram vassalos: Colonização e relações de poder no norte do Brasil na segunda metade do século XVIII. Lisboa, Portugal: Comissão Nacional Comemorações dos Descobrimentos Portugueses. Edelweiss, Frederico G. 1947. Tupís e Guaranís: Estudos de etnomínia e linguística. Bahia: Museu da Bahia, Secretaria de Educação e Saúde. Evreux, Ivo d’. [1614] 1929. Viagem ao norte do Brasil feita no[s] annos de 1613 a 1614 pelo Padre Ivo D’Evreux religioso capuchinho. Trans. Cezar Augusto Marques. Rio de Janeiro: Depositarios Freitas Bastos. Farage, Nádia. 1991. As muralhas dos sertões: Os povos indígenas no rio Branco e a colonização. Rio de Janeiro: Paz e Terra. Fausto, Carlos. 1992. Fragmentos de história e cultura tupinambá: Da etnologia como instrument critic de conhecimento etno-histórico. In Carneiro da Cunha 1992, 381–396.

M. Kittiya Lee

[ 165 ]

Fernandes, Florestão. 1970. A função social da guerra na sociedade Tupinambá. 2nd ed. São Paulo: Editora Globo. ———. 1963. Organização social dos Tupinambá. São Paulo: Difusão Européia do Livro. Figueira, Luís. 1621. Arte da Lingva Brasilica. Lisboa: Manoel da Silva. ———. 1687. Arte de grammatica da lingua Brasilica. Lisboa, Portugal: Miguel Deslandes. Gage, John. 1993. Color and culture: Practice and meaning from antiquity to abstraction. Berkeley: University of California Press. Ganson, Barbara. 2003. The Guarani under Spanish rule in the Rio de la Plata. Stanford, CA: Stanford University Press. Heckenberger, Michael J., Eduardo G. Neves, and James B. Peterson. 1998. De onde surgem os modelos? As origens e expansões Tupi na Amazônia central. Revista de Antropologia 41, no. 1:69–96. Johnson, H. B., Jr., and Maria Nizza da Silva, eds. 1992. Nova história da expansão Portuguesa: O império Luso-Brasileiro (1500–1620). Lisbon: Editorial Estampa. Knivet, Anthony. [t. 1591] 2008. As incríveis aventuras e estranhos infortúnios de Anthony Knivet: Memórias de um aventureiro ingles que em 1591 saiu de seu país com o pirata Thomas Cavendish e foi abandonado no Brasil, entre índios canibais e colonis selvagens. Ed. Sheila Moura Hue. Trans. Vivien Kogut Lessa de Sá. 2nd ed. Rio de Janeiro: Jorge Zahar Ed. Langfur, Hal. 2006. The forbidden lands: Colonial identity, frontier violence, and the persistence of Brazil’s eastern Indians, 1750–1830. Stanford, CA: Stanford University Press. Leite, Serafim, S.J. 1938–50. História da Companhia de Jesus no Brasil. 10 vols. Lisbon: Livraria Portugália; Rio de Janeiro: Instituto Nacional do Livro. Léry, Jean de. [1578] 1990. History of a voyage to the land of Brazil, otherwise called America. Trans. Janet Whatley. Berkeley: University California Press; Oxford: Oxford University Press. [Lopes de Sousa, Pero.] [1530–32] 1968. Diário da navegaço de Pêro Lopes de Sousa (1530–1532). Lisboa, Portugal: Agência-Geral do Ultramar. Lorimer, Joyce. 1989. English and Irish settlement on the River Amazon, 1550–1646. Cambridge, UK: Hakluyt Society. Maranhão, Francisco de N.S. dos Prazeres. 1947. Poranduba Maranhanse. São Luís, Brazil: n.p. Marchant, Alexander. 1942. From barter to slavery: The economic relations of Portuguese and Indians in the settlement of Brazil, 1500–1580. Baltimore: Johns Hopkins University Press. Medina, José Toribio, ed. [1542] 1988. The discovery of the Amazon [according to Friar Gaspar de Carvajal, and others]. Trans. Bertram T. Lee. Ed. H. C. Heaton. New York: Dover. Metcalf, Alida C. 2005. Go-betweens and the colonization of Brazil, 1500–1600. Austin: University of Texas Press. Métraux, Alfred. 1927. Migrations historiques des Tupi-Guarani. Journal de la Société des Américanites de Paris 29:1–45. Monteiro, John M. 1994. Negros da terra: Índios e bandeirantes nas origens de São Paulo. São Paulo: Companhia das Letras. ———. 1997. The crises and transformations of invaded societies: Coastal Brazil in the sixteenth century. In The Cambridge History of the Native Peoples of the Ameri-

[ 166 ]

chapter five

cas, ed. Frank Saloman and Stuart Schwartz, vol. 3, pt. 1, 973–1023. Cambridge: Cambridge University Press. Mufwene, Salikoko. 2001. The ecology of language evolution. Cambridge: Cambridge University Press. Noelli, Francisco Silva. 1998. The Tupi: Explaining origin and expansions in terms of archaeology and of historical linguistics. Antiquity 72:648–663. ———. 2008. The Tupi expansion. In Handbook of South American Archaeology, ed. Helaine Silverman and William H. Isbell, 659–670. New York: Springer. Nunes, Diogo. [c. 1552] 1993. Letter to King D. João III in Lisbon. In As crônicas do Rio Amazonas, ed. Antônio Porro, 33–35. Petrópolis, Brazil: Vozes. ———. [c. 1552] 2000. Letter to King D. João III in Lisbon. In O Novo Éden: A fauna da Amazônia brasileira nos relatos de viajantes e cronistas desde a descoberta do rio Amazonas por Pinzón (1500) até o Tratado de Santo Ildefonso (1777), ed. Nelson Papavero, Dante Martins Teixeira, William Leslie Overal, and José Roberto PujolLuz, 3–5. Belém: Museu Paraense Emílio Goeldi. Parr, Charles McKew. 1953. So noble a captain: The life and times of Ferdinand Magellan. New York: Thomas Y. Crowell. Pigafetta, Antonio. [t. 1519] 1994. Magellan’s voyage: A narrative account of the first circumnavigation. Trans. and ed. R.A. Skelton. New York: Dover. ———. [t. 1519] 1995. The first voyage around the world (1519–1522): An account of Magellan’s expedition. Ed. Theodore J. Cachey, Jr. New York: Marsilio. ———. 1536. Il viaggio fatto da Gli Spangivoli a torno al mondo. [Venice]. Pompa, Cristina. 2003. Religião como tradução: Missionários, Tupi e Tapuia no Brasil colonial. Bauru, Brazil: EDUSC; ANPOCS. Puntoni, Pedro. 2002. A guerra dos bárbaros: Povos indígenas e a colonização do sertão nordeste do Brasil, 1650–1720. São Paulo: HUCITEC; Editora da Universidade de São Paulo; FAPESP. Rodrigues, Aryon Dall’Igna. 1986. Línguas brasileiras: Para o conhecimento das línguas indígenas. São Paulo: Edições Loyola. Sampaio, Teodoro. 1987. O Tupi na geografia nacional. São Paulo: Editora Nacional. Scheel-Ybert, Rita, Kita Macario, Angela Buarque, Roberto M. Anjos, and Mariana Beauclair. 2008. A new age to an old site: The earliest Tupiguarani settlement in Rio de Janeiro State? Anais da Academia Brasileira de Ciências 80, no. 4: 763–770. Schmidel, Ulrich. [t. 1535] 1997. Viaje al Río de la Plata. Buenos Aires: Emecé. Schneider, Jane. 1978. Peacocks and penguins: The political economy of European cloth and colors. American Ethnologist 5, no. 3: 413–447. Schwartz, Stuart B. 1986. Sugar plantations in the formation of Brazilian society: Bahia, 1550–1835. Cambridge: Cambridge University Press. Soares de Sousa, Gabriel. [1587] 1989. Notícia do Brasil. Lisboa, Portugal: Alfa. Staden, Hans. [1557] 2008. Hans Staden’s true captivity: An account of cannibal captivity in Brazil. Ed. and trans. Neil L. Whitehead and Michael Harbsmeier. Durham, NC: Duke University Press. Sweet, David. 1974. A rich realm of nature destroyed: The middle Amazon Valley, 1640–1750. PhD diss., University of Wisconsin–Madison. Thévet, André, O.F.M. [1575] 1953. Les Français em Amérique pendant la deuxième moitié du XVIe siècle. Le Brésil et les Brésiliens. Ed. Suzanne Lussagnet. Paris: Presses Universitaires de France. Urban, Greg. 1992. A história da cultura brasileira Segundo as línguas nativas. In Carneiro da Cunha 1992, 87–102.

M. Kittiya Lee

[ 167 ]

Viveiros de Castro, Eduardo. 1992. From the enemy’s point of view: Humanity and divinity in an Amazonian society. Trans. Catherine V. Howard. Chicago: University of Chicago Press. Warman, Arturo. 2003. Corn and capitalism: How a botanical bastard grew to global dominance. Chapel Hill: University of North Carolina Press. Wiesner-Hanks, Merry. 2009. The marvelous hairy girls: The Gonzales sisters and their world. New Haven, CT: Yale University Press. Yai, Olabiyi. 2000. Texts of enslavement: Fon and Yoruba vocabularies from eighteenth- and nineteenth-century Brazil. In Identity in the shadow of slavery, ed. Paul E. Lovejoy, 103–125. London: Continuum.

6 * African Descendants’ Rural Vernacular Portuguese and Its Contribution to Understanding the Development of Brazilian Portuguese heliana mello

1. Introduction Differences between European and Brazilian Portuguese (BP) have been noticed and reported upon for more than a century, as noted by Clancy Clements (chap. 7, this volume). Of course, initially these reports were outside the scope of scientific investigation; today they mostly help us fit missing pieces into a huge puzzle. Travelers’ reports, short notes in diaries, papers, and narratives in fiction have all been scrutinized. However, a great deal of information is still missing about the evolutionary trajectory of BP and the populations that ultimately developed this language variety. In this chapter, taking into account demographic, historical, and linguistic arguments, I present a broad view of today’s vernacular BP (VBP) along with a brief discussion of the linguistic peculiarities of populations of African descent in rural areas. It is hoped that factoring the Portuguese variety of rural Afro-Brazilians into the broader sociohistorical and linguistic scenario will shed some light on the evolution of BP. Since the early twentieth century, various authors have attempted descriptive accounts of the linguistic situation in Brazil (e.g., Silva Neto 1986). Today, several scholars in Brazil and abroad who have developed different explanatory hypotheses on the evolution of BP seem to agree that contact among various ethnic groups in colonial and imperial Brazil was a central factor in this process. Historical evidence has increasingly been adduced to support this apparent consensus. What seems to be an issue is whether and to what extent the speech of the majority population of Brazil reflects this contact history.1 Views differ markedly. At one extreme stand the by now little-heard cries of a previous creolization and subsequent decreolization of Portuguese in Brazil (e.g., Guy 1989). At the other stand those who argue adamantly that drift is the general force that drove BP to its current form (e.g., Naro and Scherre 2007). In between these two poles, one can identify many views that in one way or another argue that some degree of contact among languages had an effect at different levels of BP. This diversity of hypotheses underscores the need for

Heliana Mello

[ 169 ]

further linguistic and historical research to uncover more facts about the emergence of BP. About the paucity of scientific studies, Mattos e Silva laments, “The Portuguese language, from its origins . . . up till today has never had a circumstantiated historical report, grounded on theoretical and empirical bases” (1998, 21). The complex sociocultural and linguistic phenomena that over the centuries led to the divergence of BP from its European ancestor may never be totally extricated. There are, however, some indicial elements that lead us to hypothesize how the regional and social continuum of lects that BP consists of emerged. It is hoped that those hypotheses will be confirmed, disconfirmed, or elaborated as new studies are undertaken.

2. The Makeup of the Brazilian Population The arrival of the Portuguese on the coast of Bahia, on April 22, 1500, was certainly not the beginning of the complex societal multilingualism that obtained in Brazil. When the Portuguese arrived, a wide range of Amerindian languages were spoken; some of them even functioned as lingua francas and were used in the initial contacts between the Portuguese colonists and the native peoples of Brazil. The introduction of Portuguese and African languages just increased the complexity of the linguascape. Note that language contact in these circumstances involved only oral discourse. This has a much greater impact on the possible outcomes of contact, because there were no fixed norms regimenting the paths taken by languages being used in interactions between ethnolinguistic groups, especially in trade, proselytizing, and colonial administration (Villalta 1997). To propose models for the development of BP, it is necessary to take a look at the three most important linguistic inputs that made up the initial contact situation in Brazil: Amerindian languages, Portuguese, and African languages, along with the demographic numbers associated with them (Mattos e Silva 1991). The indigenous languages spoken on the Brazilian coastline at the time of the arrival of the Portuguese have now been classified as belonging to the Karib, Macro-Arawak, Macro-Tupi, and Macro-Jê groups (Rodrigues 1994). During the first three decades of contact in the sixteenth century, the Portuguese interacted primarily with indigenous populations speaking the Tupi-Guarani languages (generally classified in the MacroTupi group) and the Jê languages. Travelers’ reports of that time mention mostly varieties of Tupinambá and Tupiniquim, which were so close to each other that they could be considered dialects of the same language.

[ 170 ]

chapter six

One of the first publications that mentioned the commercial jargons used on the Brazilian coast by Europeans and local peoples was written by Antonio Pigafetta and published in Venice in 1536. In this report, the author mentions a few lexical items belonging to the semantic fields of food and common objects. The document leads us to think that the trade interactions between the Europeans and the natives, basically for brazilwood, were of very limited linguistic complexity. Most probably, these incipient varieties consisted of a few syntactic structures and a restricted lexicon centered around the strictly necessary interactions between the two populations. It is also important to mention that the Portuguese at the time were in the habit of leaving degredados behind, with the explicit assignment of learning the local languages.2 This practice is documented in the letter written by Pero Vaz de Caminha, between April 26 and May 2, 1500, upon the discovery of Brazil by Pedro Álvares Cabral. The role played by the cultural interpreters and the degredados must have been important in the initial decades of the exploitation of brazilwood. This scenario changed as time passed and points of occupation were established and consistently maintained by the Portuguese. A result was the development of the so-called Língua Brasílica, also known as Lingua Geral, a koiné based on Tupi-Guarani (as discussed by Hildo do Couto, Denny Moore, and M. Kittiya Lee, chaps. 3, 4, and 5, this volume). The relevance of this lingua franca was notoriously documented by the Jesuit Father Anchieta, who published in 1595 a book entitled Arte de gramática da lingua mais usada na costa do Brasil (The art of grammar for the most used language on the Brazilian coast), which is no less than a grammar of ancient Tupi—indicating the importance that this language had at that time in Brazil. The Brasílica language remained an important lingua franca in Brazil until the eighteenth century, when an official law passed in Portugal established Portuguese as the legally mandatory language.3 Nonetheless, a variety of the Brasílica, known as Nheengatu, is still spoken today in the Amazon region, although it is dying. (See also chapters 4 and 3, by Moore and Couto, this volume.) Another important contribution to the formation of BP comes from the African languages spoken by the slaves, the first of whom arrived in 1538 (Mattoso 1989). It is estimated that about 4 million Africans were brought to Brazil over the next three centuries. They spoke languages belonging in two macro-families: the Afro-Asiatic and the Congo-Kordofanian. The first group, including speakers of especially Hausa (a Chadic language), were restricted to Bahia and practiced Islam. The second group is the most important one, comprising the great majority of Africans brought to

Heliana Mello

[ 171 ]

Brazil. They spoke languages from the same linguistic macro-family, the Niger-Congo, a family represented mostly by Kwa and Bantu languages. The Kwa languages were Ewe, Fon, and Mahi (Jeje, Mina, Ijo and Yoruba), while the Bantu languages included especially Kikongo, Umbundu, and Kimbumdu (Castro 2001). The Bantu speakers far outnumbered the Kwa speakers. They arrived earlier and were spread to various parts of Brazil, unlike the Kwa speakers, who were brought much later to Bahia (Castro 2001). According to Mattoso (1989), some Africans arrived in Brazil directly from Portugal (negros do reino ‘blacks from the kingdom’). This suggests that they may have learned Portuguese there and spoke a variety of it before reaching the colony. Mello (1997) also notes that some sugarcane plantation owners in São Tomé relocated to Brazil, bringing their creoleor Portuguese-speaking slaves with them. Depending on their level of acculturation and proficiency in Portuguese, slaves were classified as negros boçais ‘bozal blacks’, which literally means ‘stupid blacks’ and refers to those Africans who did not speak any Portuguese, or as negros ladinos ‘acculturated blacks’, literally ‘pure, genuine blacks’. The first could not express themselves in Portuguese, while the latter had a good command of the language. The ladinos had higher status owing to their linguistic skills, which were an important asset and helped some move up in the colonial population structure. The degredados were left behind in Brazil by Álvares Cabral to ensure communication between the newly discovered territory and Portugal. Because they learned the indigenous languages, they were used as liaisons between the natives and the colonizers. In the years that succeeded the discovery, the Portuguese made several trips to the Brazilian coast in search of riches. The occupation of the territory did not begin, de facto, until 1530 (Wehling and Wehling 1994). The first Portuguese who came to Brazil belonged to the masses of the population; most of them were men, economic destitutes who looked for opportunities for a new beginning (Freire 1977). We may assume that they spoke nonstandard vernacular varieties of European Portuguese. Although many of them probably came from port cities, many others may also have emigrated from rural areas. At the time, most people had no formal education and were illiterate. Their origins included Madeira, the Azores, Oporto, the Alantejo, Minho, and Lisbon (Silva Neto 1986). This diversity of points of origin appears to have prevented any particular variety from prevailing over the others. As in other colonies, such as the United States and Australia, koineization (characterized by many sociolinguists as dialect leveling) occurred (Holm 1992). Intermarriages between Portuguese men and indigenous and, later,

[ 172 ]

chapter six

African women also produced a highly mixed society and had additional linguistic consequences. Lingua Geral became the native language of a large contingent of mixed Brazilians, especially those with indigenous mothers. These were mostly concentrated in the captaincy of São Vicente and were known as bandeirantes. Later on, as they expanded their territory in search for riches, they spread their language. Although they were slaves, mulattoes (slaves of mixed white and black ancestry) had better chances of buying their freedom and rising in the social ladder, especially because they could speak Portuguese. As the colonial economy evolved from a strictly exploratory one, based primarily on the exploitation of brazilwood, and became agricultural, the number of Portuguese working-class immigrants increased substantially. The colonial system and its ties to the metropolis also became stronger. Little by little, the European Portuguese input not only increased but started to become a stable contender in the complex linguistic setup of colonial Brazil, in which there was already a strong presence of Lingua Geral, along with several varieties of Portuguese spoken as a second language by Africans and Native Americans. The input of more standardized varieties of European Portuguese did not occur until the nineteenth century, with the deployment of the royal family from Portugal to Brazil, fleeing from the Napoleonic invasion (Rodrigues 1985). Along with the royalty, many artists, intellectuals, and well-to-do Portuguese came to Brazil and brought with them an appreciation for the fine arts, books, theater, and so on. There followed a surge of newspapers, the creation of schools and universities, and the emergence of new social habits inspired by the European nobility. At least in the urban areas, there was a significant linguistic impact and an increase in schooling.

3. Demographic Data Brazil has a history of population contacts that spans more than five hundred years. Initially the contacts involved Amerindians and the Portuguese, but later the picture became more complex with the importation of Africans and the inception of mixed populations, mulattoes, and mestizos (persons of mixed European and indigenous parentage). This scenario was made more complex from the mid-nineteenth century well into the twentieth century by the significant immigrations of Italians, Germans, eastern Europeans, Japanese, and Arabs.4 The Brazilian ethnolinguistic history is far from completely understood. However, there are several data that can be considered firmly established and can inform our hypotheses about the development of BP.

Heliana Mello

[ 173 ]

The data presented here are provided by the Instituto Brasileiro de Geografia e Estatística (Brazilian geography and statistics institute) and are probably the most reliable data available to researchers at this time. Table 6.1 presents data comparing the indigenous populations in the sixteenth century with the populations today. A drastic decrease can be observed. It is not possible to infer from these figures the percentage of the indigenous population that was in contact with the Portuguese and consequently was exposed to the Portuguese language in colonial Brazil. Neither is it possible to draw any inferences about the mixed-population indexes; thus any discussion about the percentage of the population that spoke Lingua Geral is restricted. Table 6.2 represents the approximate numbers of African arrivals in Brazil. The numbers speak for themselves and point toward a potential for cross-linguistic influences in a contact situation, since there were breaks in the transmission of African languages once Africans were taken away from their homelands. That is, the inputs from native varieties of African languages to locally born children were drastically reduced, or

Table 6.1. Indigenous population of Brazil today and in the sixteenth century (selected groups) Indigenous population estimates Selected groups and localization Acre (Rio Purús): Not less than 16 groups Amazonas (Rio Branco): 9 groups Tocantins: 19 groups Nordeste, coast: 7 groups Nordeste, interior: Not less than 13 groups Maranhão: 14 groups Bahia: 8 groups Minas Gerais: 11 grupos groups Espírito Santo (Ilhéus): 9 groups Rio de Janeiro: 7 groups São Paulo: 8 groups Paraná e Santa Catarina: 9 groups Rio Grande do Sul: 5 groups Mato Grosso do Sul: 7 groups Mato Grosso Central: Not less than 13 groups Others Total

Modern

16th century

3,000–5,000 11,000–16,000 5,000–5,600 1,000

30,000 33,000 101,000 208,000 85,000 109,000 149,000 91,000 160,000 97,000 146,000 152,000 95,000 118,000 71,000 786,000 2,431,000

2,000–6,000 0–200

3,200–4,200 6,200–8,200 1,900–2,900

Source: Instituto Brasileiro de Geografia e Estatística 2000. Note: Empty cells indicate the lack of reliable statistical data.

[ 174 ]

chapter six

Table 6.2. Approximate numbers of African arrivals in Brazil, 1531–1855 Periods

Numbers

Periods

1531–1575 1576–1600 1601–1625 1626–1650 1651–1670 1676–1700 1701–1710 1711–1720 1721–1730 1731–1740 1741–1750 1751–1760 1761–1770 1771–1780 1781–1785

10,000 40,000 10,0000 10,0000 18,5000 17,5000 153,700 139,000 146,300 166,100 185,100 169,400 164,600 161,300 63,100

1786–1790 1791–1795 1796–1800 1801–1805 1806–1810 1811–1815 1816–1820 1821–1825 1826–1830 1831–1835 1836–1840 1841–1845 1846–1850 1851–1855 Total

Numbers 97,800 125,000 108,700 117,900 123,500 139,400 188,300 181,200 250,200 93,700 240,600 120,900 157,500 6,100 4,009,400

totally eliminated in some cases, while new communicative situations demanded the use of new languages, notably Portuguese and Lingua Geral, which most likely were learned as second languages by the adult slaves, at varied levels of proficiency. Table 6.3 presents estimates of the arrivals of the Portuguese in Brazil for the period extending from 1500 to 1991. Note that during the crucial periods of population growth in colonial Brazil, especially in the eighteenth and early nineteenth centuries, the African importations clearly outnumbered Portuguese immigrations. However, while the massive importation of Africans ended in the late nineteenth century, Portuguese immigration continued well into the late twentieth century, albeit in small numbers. The available population estimates reveal that during the colonial period the Portuguese were a significant minority in Brazil. The larger population contingents were Amerindians initially, and later Africans, their descendants, and people of mixed ancestry (Mussa 1991). This demographic growth pattern favors the coexistence of various interlingual levels of Portuguese with other languages during the colonial period, before Portuguese emerged as the dominant language of Brazil. In section 4 I provide an overview of data regarding rural Portuguese spoken by people of African descent, with the aim of showing that in that group, despite their isolation over centuries, the patterns found are very similar to those of a broad vernacular Portuguese in Brazil.

Heliana Mello

[ 175 ]

Table 6.3. Estimated Portuguese immigration to Brazil, 1500–1991 Period 1500–1580 1581–1640 1641–1700 1701–1760 1808–1817 1827–1829 1837–1841 1856–1857 1881–1900 1901–1930 1931–1950 1951–1960 1961–1967 1981–1991

Portuguese America

Colonial empire

100,000

280,000 300,000 120,000

600,000 24,000 2,004 629 16,108 316,204 754,147 148,699 235,635 54,767 4,605

Annual average: Portugese America 500 10,000 2,666 668 125 8,054 15,810 25,138 7,434 23,563 7,823 406

Annual average: Colonial empire 3,500 5,000 2,000

Source: Instituto Brasileiro de Geografia e Estatística 2000.

4. Rural Brazilian Portuguese among Descendants of Africans As noted, the role played by the Africans and their descendants in the shaping and diffusion of VBP must definitely be considered. The sheer numbers of this population throughout the colonial period, the decrease and progressive isolation of the Amerindian population, and the irregularity of the Portuguese immigration patterns suggest that Africans and their descendants must have had an important role in the emergence of VBP. The debate on the origins of the distinctive features of VBP (in relation to European Portuguese) has revealed no evidence for assuming that a Portuguese creole was spoken in Brazil during the colonial period. However, the contribution of people of African descent need not have produced a creole, especially since the colonial social ecology of Brazil was not the same as in the Caribbean or the Indian Ocean. In the absence of diachronic evidence of the Portuguese spoken by the Africans of the colonial and imperial periods, it seems advisable to resort to apparent time and examine the speech patterns of the elderly and rural populations, especially in isolated communities. So in the paragraphs that follow, I focus on the speech of some isolated descendants of Africans in rural communities and discuss the extent to which their speech differs from, or resembles, general VBP. Reports abound about runaway slaves and the rural, isolated communi-

[ 176 ]

chapter six

ties they created. These were usually set up in areas difficult to reach and became homes of resistance to the colonial status quo. According to historians, there were thousands of such communities, some of which lasted for brief periods of time and others for decades or even centuries. Some of them still survive (Santos and Camargo 2008). Almost twelve hundred such communities, called quilombos (and their inhabitants, quilombolas), have been identified in Brazil today. According to some reports, the quilombos of Ambrosio (Minas Gerais), Quariterê (Mato Grosso), Limoeiro (Turiaçu-Maranhão), and Lagoa Amarela (Chapadinha-Maranhão) were some of the most troublesome for the colonial regime, as they not only engaged in active resistance but also waged localized wars against government authorities and landowners. In 1630, as a result of constant slave rebellions, many sugarcane plantation owners in the state of Pernambuco, in northeastern Brazil, left their properties. Their departure encouraged the fleeing of a large number of slaves, who sought protection in the celebrated Quilombo dos Palmares, in the state of Alagoas. In 1670 there were allegedly around fifty thousand runaway slaves and their descendants living in Palmares. The war against the Palmarinos led by Zumbi lasted five years, until the quilombo finally was destroyed by Portuguese forces. Apparently, the quilombos were much more than short-term groupings of slaves; they were organized societies, with their own autonomous economic systems. There are no reports that deal specifically with the language varieties spoken in the quilombos during the colonial period. We can conjecture that these organized spaces brought together people who had spoken different native languages and spoke Lingua Geral and various Portuguese interlanguages. Most likely, in the communities that lasted, a contact variety of Portuguese emerged, was adopted as the lingua franca, and eventually became the vernacular of the quilombolas. From the evidence encountered today, in some quilombos, some forms of Portuguese that included several words from African languages were used as secret languages. Good examples of these are found in Tabatinga, Bom Despacho, in the present state of Minas Gerais (Queiroz 1998); and in the famous Cafundó area in the state of São Paulo (Vogt and Fry 1996). The best-studied variety of rural Portuguese among descendants of Africans in Brazil is spoken in Helvécia, in the state of Bahia.5 This variety was first recorded in the late 1960s by Carlota Ferreira, who was then associated with a dialectology project, and later, in the 1980s and 1990s, by the creolists Alan Baxter and Dante Lucchesi. But Helvécia was never a quilombo. Instead, it originated in lands donated to former slaves by Swiss, French, and German landowners, who left the region when coffee cultivation declined in the late nineteenth century. However, as we shall

Heliana Mello

[ 177 ]

see, the Helvécia dialect presents the same features found in broad VBP as described in Mello (1997). This fact points toward a core of uniformity surrounded by peripheral variation in BP. Most of its distinct features are part of the various lects spoken broadly in Brazil. There are, however, features that have higher frequency in Helvecian than in other rural VBP varieties. It is as if Helvecian were a radical examplar of general VBP, with some of its features occurring with higher frequency (Lucchesi 2001). This has led some scholars to claim that Helvecian may have once been a creole (Ferreira 1984–85; Pereira n.d.). Among the best-studied morphosyntactic features of Helvecian are variable gender concord in NPs, variable number concord in NPs, and variable concord between subject NPs and the verb. These are reviewed here, using examples provided by Lucchesi (2001) as a starting point. Mello (1997) provides an extensive overview of lexico-semantic, phonological, and morphosyntactic features found in VBP, highlighting those that have been identified in the literature as more specific of Afro-Brazilian dialects. There are several general features that probably deserve more extensive discussion than the brief mention they receive in this chapter. Afro-Brazilian VBP also includes a large component of African-based lexicon, which has been investigated by scholars such as Castro (2001) and Megenney (1970), although they focus on items of West African origin that have been adapted to Portuguese morphology. There are also entire dictionaries dedicated to this issue (e.g., Lopes 2003). The African-based lexicon and idiomatic calques have a higher percentage in areas where there are also religious rituals of African tradition, although they are attested elsewhere in the Brazilian territory. Most of these items have been traced to Bantu languages, suggesting a significant Bantu (cultural) influence in Brazil. In phonology VBP exhibits the CV syllabic structure found universally as a preferred phonotactic configuration among languages of the world. Therefore, in Brazilian Portuguese, words such as pneu [pnew] ‘tire’ are overwhelmingly pronounced [pinew], with an epenthetic vowel breaking consonant clusters. Its prosodic features have been less studied from a comparative perspective, but they have been mentioned in the literature as a probable area of interest for a better understanding of the evolution of VBP.6 The morphosyntax of VBP has received the most attention from scholars. This might be because grammar is the domain that has been the focus of most studies of standard language varieties; indeed grammar has been the foundation of linguistic theories over the past fifty years. Among the reported features of VBP that set it apart from standard Brazilian Portuguese as well as from European Portuguese, the following figure predominantly: negation patterns, NP gender and number agreement,

chapter six

[ 178 ]

relativization strategies, subject-verb agreement, word order, topicalization, pronominal forms, reflexivization, and the structure of verbal arguments. Prominent in the studies referred to above is gender and number concord within the NP. This ensues from the fact that Portuguese presents a rather complex structure for NPs, which can be roughly represented as follows: NP → {(Todo)(Determiner)(Numeral)(Intensifier)(Numeral)(Possessive)

(quantifier) (prenucleus 1)(prenucleus 2) N (postnucleus 1)(postnucleus 2)}

As can be seen, the Portuguese NP presents a nucleus with optional leftward and rightward extensions. Some of the extensions can be recursive. In any case, determiners, possessives, and adjectives must agree with the nucleus in number and gender. The other elements (quantifiers, intensifiers, numerals, and adverbs) are invariable. Number and gender agreement in standard Portuguese is obligatory and is determined by the NP nucleus, namely the head noun. If, for example, the nucleus is marked [+fem, + pl], so must the determiners, possessives, and adjectives, as in as minhas lindas irmãs [the-fem-pl myfem-pl beautiful-fem-pl sisters] ‘my beautiful sisters’. In VBP number and gender concord within the NP is variable and presents some tendencies that are identified below. The most striking feature of the NP in Helvecian and other AfroBrazilian communities, such as Calunga in the state of Goiás, is the variability, but not neutralization, of gender agreement within the NP. It happens to be attested in some cases but not in others, as exemplified below: (1) Lack of gender agreement within the NP: Ele He

era

was

uma

a[fem]

He was a well liked person.

pessoa

person[fem]

(2) Presence of gender agreement within the NP: Eu I

comprei essa

bought this[fem]

I bought that house

casa

house[fem]

muito very

querido.

dear[masc]

aí

there

Lucchesi (2001, 2005a) argues that different factors determine the presence or lack of gender agreement within the NP. According to him,

Heliana Mello

[ 179 ]

only 5 percent of NPs present lack of gender agreement and this is conditioned by (1) the level of complexity of the NP (the simpler it is, the more gender agreement is likely to be expressed), (2) the leftward or rightward positioning of nominal modifying elements (elements to the right of the noun tend not to show gender agreement), and (3) salience (nouns that are morphologically marked for gender tend to attract gender marking to their modifying elements, e.g., irmã ‘sister’, while nouns that are genderneutral favor lack of gender agreement, e.g., foto ‘photo’). Besides these characteristics, it was observed that the newer generations of Helvecians, who tend to leave the community in search of jobs and better life opportunities, exhibit a higher degree of gender agreement than the elders who have always lived in the community. Although variable NP number agreement has been associated with African-Brazilian VBP in isolated communities, it is widespread and highly frequent in general VBP. It has been amply investigated by many Brazilian sociolinguists working in the variationist paradigm, who claim that this feature is attested even in urban settings (e.g., Scherre and Naro 1993, 2001). According to these scholars, what distinguishes Helvecian from general VBP is that it exhibits a much higher frequency of lack of number agreement within the NP. This peculiarity can be attributed to its having evolved in relative isolation from other BP varieties, therefore excluded from pressure from the more prestigious sociolects. The linguistic factor that seems the most significant in conditioning number marking is the relative position of determiners and modifiers within the NP. There is a marked tendency for the most leftward element to be marked for plurality; the more rightward the position of a constituent is, the less likely it is for number agreement to be expressed in it. Therefore, determiners usually get marked and postnominal modifiers are usually not marked: (3) Umas Some[pl]

coisa

besta

thing

foolish

Some foolish things (cf. Portuguese: Umas coisas bestas)

Following the simplicity principle, the shorter the NP is, the more likely it is for all its elements to be marked for number. The more complex the NP is, the more likely it is for only its leftward elements to be marked for plural. The same tendency holds for the younger generation, who mark NP number agreement more frequently than do the older generation in Helvecian (see, e.g., Baxter 2004; Lucchesi 2005b). The third feature that has a higher frequency and is more common in Helvecian than in other VBP varieties is subject-verb concord. This fea-

[ 180 ]

chapter six

ture too has been extensively studied in the variationist sociolinguistics paradigm. Several variables have been identified that influence verbal morphology, among which are the phonetic salience of verb forms (irregular forms tend not to lose morphology), verb type, cooccurrence of subject forms, the presence or absence of the subject (with the presence of subject favoring loss of verb morphology), the semantic content of the subject (with [+human] favoring verb morphology), number agreement within the subject NP (it favors verb morphology), and person (with 3rd plural favoring loss of morphology). The following two examples will suffice to illustrate the lack of subject-verb agreement patterns found in Helvecian: (4) As

The-pl

mulhé

woman

vai

go-3sg

The women go (cf. Portuguese: As mulheres vão) (5) Nóis num we.neg

qué

not

isso

want-3sg

this

We don’t want this (cf. Portuguese: Nós não queremos isso)

Like the other vernacular features discussed above, verb morphology is more frequently expressed by the younger population than by the older one, because of access to other BP norms. The movement, then, has been one of acquisition of standard BP features by the Afro-Brazilian communities as these have more contact with the outside society, as has been pointed out by different authors (e.g., Lucchesi and Baxter 1995; Lucchesi 2001; Mello 2002). This can be appreciated in table 6.4, which shows an increase in subject-verb agreement among the Helvecian population between ages twenty and forty. Table 6.5 shows a comparison between different population groupings and their tendency toward change in the use of subject-verb agreement. As can be seen, for both rural VBP and Afro-Brazilian VBP, the tendency is the same, with an increase in subjectverb agreement for the younger population. For urban VBP the adult population, with more schooling, is the one with the highest agreement percentages. For standard urban BP, it is the oldest generation that holds the highest scores for agreement. Adults and young people have a lower rate, showing a tendency toward convergence with VBP. These data illustrate that the linguistic features that have been studied in contemporary Afro-Brazilian communities are also very much present in broader VBP, spoken across the entire nation of Brazil, in both rural and urban areas. The only feature that has been attested in Afro-Brazilian

Heliana Mello

[ 181 ]

Table 6.4. Variation of 1st person singular subject-verb agreement in Helvecian according to age

Age

Number of agreement occurrences/number of occurrences

Percentage

Relative weight

752/1,154 787/920 1,017/1,060

65 86 96

.21 .51 .80

Over 60 40–60 20–40

Source: Lucchesi and Baxter 1995.

Table 6.5. Tendency toward subject-verb agreement in different BP varieties (%) Age groups Brazilian Portuguese Variety Standard urban BP Urban VBP Rural VBP Afro-Brazilian VBP

Older

Adult

Young

98 70 34 10

93 80 42 14

93 71 61 22

Source: Lucchesi 2006, 105.

rural communities but not in other VBP varieties is variable NP gender agreement. What seem to be the features common to speakers of all vernacular, less prestigious varieties of Brazilian Portuguese are social class, level of education, and rural provenance. These characteristics have also frequently been associated with people of mixed ancestry, thus the majority of Brazilians. The feature is, therefore, not an exclusive peculiarity of blacks or people from isolated Afro-Brazilian communities. As already highlighted, it is only natural that as social conditions improve and formal education increases in Brazil, a larger percentage of the young population tend to speak a vernacular with more standardized forms (Mello 1997; Lucchesi 2001). This is not to say that Brazilian Portuguese is moving closer to European Portuguese. Each one of these two varieties has its own standard, and as time passes, it is necessary to develop comparative quantitative studies that can help determine whether European and Brazilian Portuguese are converging or diverging. As far as lexis is concerned, words of Bantu and Kwa origin are part of the daily Brazilian vernacular, naming places, objects, and social practices across diatopies and sociolects. Phonological features, such as the tendency toward CV structure, abound as well. Morphosyntactic features,

[ 182 ]

chapter six

such as variable number marking in the NP and variable subject-verb concord, are also general VBP features found only with a higher frequency in the speech of Afro-Brazilian rural communities.

5. Conclusions The speech patterns in isolated Afro-Brazilian communities discussed in this chapter, focusing on the Helvécia dialect, differ from other VBP varieties only by their higher frequency. This suggests that Afro-Brazilian VBP must have the same origins as other VBP varieties and that these varieties may have evolved closer to standard BP because of schooling and population movements from rural to urban areas. This can be appreciated especially with the acquisition of inflectional morphology. The distinctness of rural Afro-Brazilian VBP does not point toward the existence of a particular language variety, creole or not, spoken exclusively by slaves during the colonial period. It is much more likely that from the sixteenth to the late eighteenth century there was great linguistic variability in Brazil. Many different languages were in contact and their speakers had to find common ways to communicate. As explained in section 2, during two centuries, Lingua Geral was the common lingua franca, especially on the coast and in southeastern Brazil. When the true settlement of Brazil started in the seventeenth century, with the increase of lucrative economic activities such as the sugarcane industry, large numbers of Africans were brought in. The number of Portuguese immigrants also increased, although they never equaled the numbers of Africans brought to Brazil, as shown in tables 6.2 and 6.3. In the Brazilian rural colonial society, contacts between whites and blacks were necessarily intense in order for the economic activities to be carried on. Just like other sugar colonies, Brazil had many small rural properties in addition to the big sugarcane plantations, where slaves worked and lived side by side with their masters. They produced food crops, cotton, and tobacco, which they sold to the large sugar plantations (Schwartz 1985). Although there was race segregation in colonial Brazil, segregation was not the predominant and uniform way of life. Closer contacts between Europeans and Africans also occurred during the mining period in the history of Brazil and the subsequent coffee plantation period. Population movements also increased concurrently with economic growth, such as with the addition of the mining-related industries. These movements might have contributed to the emergence of general VBP, which inevitably bore the marks of second-language acquisition and leveling. Yet another factor that had important linguistic consequences was intermarriages between white males and black females, which contrib-

Heliana Mello

[ 183 ]

uted a large mulatto component to the Brazilian population. The mulattoes assumed specialized functions in sugar mills, enjoying higher social positions than the bozal and the ladino slaves and playing an important role in the spread of Portuguese. The Portuguese language has maintained its high social and economically dominant status in the history of Brazil. It has never been abandoned in favor of another language, despite the importance of other languages in particular situations. The socioeconomic structure of Brazil has enticed native speakers of indigenous languages, other European languages, and African languages to shift to Portuguese, giving rise to new varieties. The shift has thus led to the contemporary situation of VBP, in which no extensively distinct sociolects have been found, not even in isolated AfroBrazilian communities, as shown in this chapter. Notes 1. The majority population in Brazil is of mixed ancestry; the low working class predominates in urban areas and peasants in rural areas. 2. From the late fifteenth century onward, undesirables were banished from Portugal and sent to the colonies overseas, hence the name degredados ‘banished from motherland’ (Costa 1998). 3. The prohibition of Língua Brasílica affected official business in a more direct way. It is very unlikely that the issuance of a law would change a population’s means of communication immediately. Clearly, other factors ultimately, with time, led to the replacement of Língua Brasílica as the lingua franca in the whole of Brazil. 4. For an in-depth appraisal of linguistic contact throughout Brazilian history, see Mello et al. 2011. 5. In the same area there are other communities, Rio de Contas and Cinzento, that have also been studied by Professor Dante Lucchesi, of the Federal University of Bahia, and his research team. 6. This issue was raised recently by Raso and Mello (2012) in their description of BP prosodic patterning and information structure.

References Baxter, Alan. 2004. The development of variable NP plural agreement in a restructured African variety of Portuguese. In Creoles, contact, and language change: Linguistics and social implications, ed. Geneviève Escure and Armin Schwegler, 99–128. Amsterdam: John Benjamins. Castro, Yeda Pessoa de. 2001. Falares africanos na Bahia: Um vocabulário afrobrasileiro. Rio de Janeiro: Academia Brasileira de Letras. Costa, Emília Viotti da. 1998. Primeiros povoadores do Brasil: O problema dos degredados. Revista Textos de História 6:77–100. Ferreira, Carlota. 1984–85. Remanescentes de um falar crioulo (Helvécia-Bahia). Revista Lusitana 5:21–41. Freire, Gilberto.1977. Obra escolhida. Rio de Janeiro: Nova Fronteira.

[ 184 ]

chapter six

Guy, Gregory. 1989. On the nature and origins of popular Brazilian Portuguese. In Estudos sobre el Español de América y linguística Afroamericana, 227–245. Proceedings of the 45th International Americanist Conference, Instituto Caro y Cuervo, Bogotá. Holm, John. 1992. Popular Brazilian Portuguese: A semi-creole. In Actas do Colóquio sobre crioulos de base lexical portuguesa, ed. Ernesto D’andrade and Alain Kihm, 37–66. Lisbon: Colibri. Instituto Brasileiro de Geografia e Estatística. 2000. Brasil: 500 anos de povoamento. Available at www.ibge.gov.br/ (accessed October 15, 2007). Lopes, Nei. 2003. Novo dicionário banto do Brasil. Rio de Janeiro: Pallas. Lucchesi, Dante. 2001. As duas grandes vertentes da história sociolingüística do Brasil. Documentação de Estudos em Linguística Teórica e Aplicada 17, no. 1: 97–130. ———. 2005a. Grandes territórios desconhecidos. In Revista Lingüística, ALFAL 14:191–222. ———. 2005b. Processos de variação e mudança induzidos pelo contato entre línguas. In Actas do XX Encontro da Associação Portuguesa de Linguística, ed. Inês Duarte and Isabel Leiria, 157–169. Lisbon: Colibri Artes Gráficas. ———. 2006. Parâmetros sociolingüísticos do português brasileiro. Revista da ABRALIN 5:83–112. Lucchesi, Dante, and Alan Baxter. 1995 A variação na concordância verbal com a 1ª pessoa do singular como parte do processo de descrioulização do dialeto de Helvécia-Ba. Paper read at the 47th Reunião Anual da SBPC, São Luís-Ma, Brazil. Mattos e Silva, Rosa Virgínia. 1991. Diversidade e unidade: A aventura lingüística do Português. In Curso de História da Língua Portuguesa: Leituras complementares, ed. Ivo Castro, 11–141. Lisboa, Portugal: Universidade Aberta. ———. 1998. Idéias para a história do português brasileiro: Fragmentos para uma composição posterior. In Para a história do português brasileiro, vol. 1, Primeiras idéias, ed. Ataliba Castilho, 21–52. São Paulo: Humanitas. Mattoso, Katia.1989. To be a slave in Brazil: 1550–1888. New Brunswick, NJ: Rutgers University Press. Megenney, William.1970. An ethnolinguistic study of West African influences in Bahian Portuguese. Ann Arbor, MI: University Microfilms International. Mello, Heliana. 1997. The genesis and development of Brazilian vernacular Portuguese. PhD diss., City University of New York. ———. 2002. Português padrão, português não-padrão e a hipótese do contato lingüístico. In Para a história do português brasileiro: Novos estudos, vol. 3, ed. Tânia Maria Alkmim, 341–358. São Paulo: Humanitas. Mello, Heliana, Cléo Altenhofen, and Tommaso Raso, eds. 2011. Os contatos linguísticos no Brasil. Belo Horizonte: Editora UFMG. Mussa, Alberto. 1991. O papel das línguas africanas na história do português do Brasil. Master’s thesis, Universidade Federal do Rio de Janeiro. Naro, Anthony, and Maria Marta Pereira Scherre. 2007. Origens do português brasileiro. São Paulo: Parábola. Pereira, Dulce. n.d. Crioulos de base portuguesa: História da língua portuguesa em linha. Available at http://cvc.instituto-camoes.pt/hlp/geografia/crioulosde baseport.html (accessed May 28, 2010).

Heliana Mello

[ 185 ]

Queiroz, Sônia. 1998. Pé preto no barro branco: A língua dos negros de Tabatinga. Belo Horizonte: Editora UFMG. Raso, Tommaso, and Heliana Mello, eds. 2012. C-ORAL-BRASIL I: Corpus de referência do português Brasileiro falado informal. Belo Horizonte: Editora UFMG. Rodrigues, Aryon. 1994. Línguas brasileiras: Para o conhecimento das línguas indígenas. São Paulo: Loyola. Rodrigues, José H. 1985. A vitória da língua portuguesa no Brasil colonial. In História viva, ed. J. H. Rodrigues, 11–48. São Paulo: Global. Santos, Maria E. G., and Pablo M. Camargo. 2008. Comunidades quilombas de Minas Gerais no século XXI: História e resistência. Belo Horizonte: Autêntica Editora. Scherre, Maria M. P., and Anthony J. Naro. 1993. Duas dimensões do paralelismo formal na concordância de número no português popular do Brasil. Documentação de Estudos em Linguística Teórica e Aplicada 9:1–14. ———. 2001. Sobre as origens estruturais do português brasileiro: Crioulização ou mudança natural? Papia 11: 41–50. Schwartz, Stuart. 1985. Sugar plantations in the formation of Brazilian society: Bahia, 1550–1835. New York: Cambridge University Press. Silva, Maria B. Nizza da, ed. 1986. Nova história da expansão Portuguesa: O império Luso-Brasileiro (1750–1822). Vol. 8. Lisboa, Portugal: Editorial Estampa. Silva Neto, Serafim da. 1986. Introdução ao estudo da língua portuguesa no Brasil. 2nd ed. Rio de Janeiro: Presença. Villalta, Luiz Carlos. 1997. O que se fala e o que se lê: Língua, instrução e leitura. In História da vida privada no Brasil: Cotidiano e vida privada na América portuguesa, ed. Laura de Mello Souza, 332–385. São Paulo: Companhia das Letras. Vogt, Carlos, and Peter Fry. 1996. Cafundó: A África no Brasil—língua e sociedade. São Paulo: Letras; Campinas: Editora Unicamp. Wehling, Arno, and Maria José C. de Wehling. 1994. Formação do Brasil Colonial. Rio de Janeiro: Nova Fronteira.

7 * Brazilian Portuguese and the Ecology of (Post-)Colonial Brazil j. clancy clements

1. Introduction It has been clear for a long time that Brazilian Portuguese and European Portuguese have diverged from each other substantially and continue to do so. This raises a number of questions, two of which I concentrate on here: whether these two language varieties have become two separate languages and, if so, why the divergence has been taking place. My approach to these questions is framed within a population-genetics model, of the kind proposed by Croft (2000) and Mufwene (2001). From a population-genetics perspective, a language is defined based on the concept of a speech community. Even though Vernacular Brazilian Portuguese (VBP) has evolved away from Vernacular European Portuguese (VEP) and seems to be changing its typological status in some ways, I argue that it is impossible to view the two varieties as two distinct languages. The question of why VBP has been diverging from VEP is a complex one that needs to be studied in detail. Here I focus on one possible reason for the divergence: the strong demographic presence of blacks and pardos ‘African-European mixed-raced people’ from early on in the colonization of Brazil until the present day. I show this by drawing on Schwartz’s (1985) documentation of slavery in Brazil from the sixteenth to the eighteenth century, as well as census records of three periods from the late nineteenth to the mid-twentieth century. The picture that emerges from the demographic information supports the assumption that the large number of Africans and their descendants played an integral role in the introduction of new features into VBP and in the propagation of these features among the population at large.

2. Language as Species Defining an entity as varied and complex as a language is difficult. If one gives precedence to purely linguistic criteria, crucial sociohistorical aspects of language are disregarded. While there is as yet no perfect

J. Clancy Clements

[ 187 ]

manner to define any given language, some recent approaches give more weight to sociohistorical than to linguistic criteria. In this study of the ecology of VBP, I explore the usefulness of the sociohistorical criteria in defining it and compare them to linguistic criteria. To do this, I follow Croft’s (2000) and Mufwene’s (2001) model, approaching VBP from a population-genetics perspective of language change, a perspective for which the distinction between sibling languages and polytypic languages is important. Specifically, the notion of a polytypic language will be useful in speaking of the divergence between VBP and VEP. The model developed by Croft (2000) and Mufwene (2001) views language change within real-time situations in which sociohistorical aspects of change make up the ecology of a language. In his treatment of the question of species and how it relates to language, Croft (2000, 13–20) compares an essentialist perspective to a population-genetics perspective: “In the essentialist view of a species, each species has immutable essential structural properties that identify it. . . . That is, the essentialist view is that a species instantiates an abstract type.” Croft notes that a major problem with this view is that there are reproductively isolated populations (known as sibling species) that cannot be distinguished structurally according to the essentialist definition. Similarly, there are also populations (known as polytypic species) that are structurally very differently from one another in terms of the essentialist definition, yet they still reproduce among themselves. In the population-genetics theory of species, “a species consists of a population of interbreeding individuals who are reproductively isolated from other populations” (13). According to this view, there is no species type, defined by an abstract structural trait, but rather a fundamental property that the individuals of a given population share, namely, that they are reproductively isolated. Croft goes on to compare a classification based on an abstract trait versus one based on a property of a population. Corresponding to the essentialist view of a species is taxonomic classification, based on similarities and differences with regard to structural traits. Phylogenetic classification, in contrast, corresponds to the population-genetic view of a species. The latter type of classification, he explains, is historical: “A proper phylogeny requires the differentiation of traits based on their history. If two taxa share a trait, it could be a retained trait from the parent population (a symplesiomorphy), or it could be a shared innovation of the two taxa (a synapomorphy). Only a shared, innovated trait can justify grouping the two taxa together phylogenetically” (2000, 15). If we apply the population-genetic definition of species to the problem of defining languages, we can say that languages are related phylogenetically, that is, in terms of their historical classification. To illustrate this

[ 188 ]

chapter seven

point, Croft offers examples of sibling and polytypic languages, which are analogous to sibling and polytypic species in biology. Two languages such as Hindi and Urdu, for example, are sibling languages in that they are so closely related phylogenetically that many consider them dialects of the same language; yet, they are perceived by at least one major group of speakers to be distinct, most ostensibly because of religious and cultural differences: Urdu speakers are Muslims and Hindi speakers are Hindus. Accordingly, Urdu uses Persian and Arabic as its sources for neologisms, whereas Hindi draws neologisms from Sanskrit. The definition of language in such a case depends on cultural, as well as on lexical and other linguistic differences. Croft uses the case of Chinese to illustrate a polytypic language. The dialects of Chinese (e.g., Cantonese and Mandarin, among others) are often mutually unintelligible; yet, they share the same writing system and political unity, factors suggesting that the different varieties are a single language. In this respect Wardhaugh (2002, 31) points out that speakers of Cantonese and Mandarin will say that they speak different dialects of the same language. However, if one speaker knows only Cantonese and the other only Mandarin, they will not be able to talk to each other; but if they both are literate, they will be able to communicate through their shared writing system. The insistence by speakers of these two varieties that they are dialects of Chinese is grounded not only in the shared writing system but, just as importantly, also in the shared history and strong tradition of cultural, social, and political unity, not to mention the language varieties’ typological commonality. Concepts involving culture, society, and politics then turn out to be essential parts of their definition as Chinese. There are various possible examples of polytypic languages among the Romance languages. Sicilian and Milanese, for instance, are considered two dialects of Italian. However, if one speaker knows only Milanese and another only Sicilian, they will have difficulty communicating with each other. A possible example of a polytypic language within the Iberian Romance family is Portuguese, whose vernacular varieties in Portugal and Brazil are quite different on various levels. In this chapter, I focus on the vernacular varieties of these two languages.1 In phonology, there are several salient differences between the canonical varieties of VBP and VEP. Posttonic /e/ in VBP raises to [i] but does not raise in this way in most varieties of VEP (cf. /dente/ VBP [ʤén-ʧi] with VEP [dén-tə]). In VBP, the stops /t/ and /d/ are palatalized as [ʧ] and [ʤ] preceding front vowels in most varieties but are not found at all in VEP. This is illustrated in (1), where the pronunciation of the sentence

J. Clancy Clements

[ 189 ]

equivalent to English ‘I give you the tooth’ in VBP contains three palatalized consonants but in VEP has none. (1) a.

Te

dou

2sg.obj.fam

give-1sg

[ʧi-dóu-ʤén-ʧi] (VBP)

b.

Dou

give-1sg

te

2sg.obj.fam

[dóu-təu-dén-tə] (VEP)

o

dente.

the.masc

tooth

o

dente.

the.masc

tooth

In VBP, we find variable agreement of number (2a) and gender (2b) in the noun phrase and the use of 3sg subject pronouns as object pronouns, shown in (3a), which is related to the reduced use of reflexives and passive. The standard way of expressing ‘he killed himself’ is shown in (3b). These features are rare or absent in VEP. (2) a.

Ficou

remained-3sg.preterit

uns

some

quarto

four

mês

month

lá.

(number)

there

‘S/he remained there about four months.’ (Mello et al. 1998, 102) b.

aquele

that-masc

coisa

thing-fem

estufado

stewed-masc

(gender)

‘that stewed thing’ (Mello et al. 1998, 102) (3) a.

Ele he

Matou

kill-3sg.past

ele he

mesmo.

self-masc

‘He killed himself.’ (Mello et al. 1998, 103) b.

Ele

he

Se

3sg.refl

‘He killed himself.’

matou. killed

Regarding subject pronoun use in a null-subject language, one finds a greater presence of overt subject pronouns in VBP (Mello et al. 1998;

chapter seven

[ 190 ]

Guy 1989; Naro and Scherre 1993); but in VBP, in general the order of clitic pronouns is preverbal with finite verbs, whereas in VEP clitics are immediately postverbal in declarative main clauses. VBP also allows null objects more commonly than does VEP (Schwenter and Silva 2002, 2003; Schwenter 2006). With regard to the verb phrase, Mello and colleagues mention for colloquial VBP the variable use of the subjunctive and a strongly reduced verbal morphology, especially in rural VBP, such that 2sg, 1pl, and 3pl forms are often reduced to the 3sg form (the default). The paradigm for VBP is shown in (4) for the indicative present and preterit tense forms of the verb falar ‘speak, talk’ (Mello et al. 1998, 124). (4) Present Tense

singular

1

eu

falo

2

você

fala

3

ele/ela

Preterit Tense

a gente fala

plural nós

fala ~ (falamo)

vocês

fala ~ (falam)

fala

eles/elas

fala ~ (falam)

1

eu

falei

nós

falou ~ (falamo)

2

você

falou

vocês

falou ~ (falaro)

3

ele/ela

a gente falou

falou

eles/elas

falou ~ (falaro)

Three other phenomena worthy of mention are double negation with a sentence-final negator (5), lack of Subject-Verb inversion in questions (6), and the use of em to express both location and goal (5). Regarding this last phenomenon, the locative {em} appears in a contracted form with the feminine-singular definite article a, yielding na. Eight traits in table 7.1 (numbers 2, 3, 4, 5, 6, 9, 12, and 13) are attested in VBP but in VEP are either rare or absent from all varieties. Double negation with sentence-final não (trait 10) is pervasive in VBP and attested with much less frequency in VEP. A further development of this pattern in VBP is sentence-final negation with não, as in canto não ‘I don’t sing’, which is not found in VEP. (5)

Não

neg

vou

go-1sg

na

in-the.fem

roça

não.

neg

‘I’m not going to the _______.’ (Mello et al. 1998, 105)

J. Clancy Clements

(6)

Você

you

mora

live

[ 191 ]

onde?

where

‘Where do you live?’ (Mello et al. 1998, 107)

Sentence-final negation with não in VBP has been attributed to the influence of African languages (Güldemann and Hagemeijer 2006; Schwegler 1996).2 Trait 1 is found in at least one dialect of VEP (southern Alentejano), and trait 11 is common in VBP and in VEP but is pragmatically more constrained in the latter. Finally, the presence of traits 7 and 8 also involve different pragmatic and semantic constraints between VBP and VEP, both for subject pronoun expression and object pronoun deletion. Table 7.1 summarizes the traits so far discussed, with a comparison of their presence in VBP and VEP. Given all the aforementioned differences, there can be some difficulty in communication between VBP and VEP speakers who have not been exposed to each other’s variety. These differences notwithstanding, there is no question that the two varieties are perceived by speakers in Brazil and Portugal as a single language. In my view, this perception of unity is attributable to the close cultural and social ties between Brazil and Portugal and the fact that both varieties share a common written register. Returning to the list of traits, I suggest that the pervasive or common presence of many of them in VBP may be explained by the fact that the traits are interconnected to a certain extent. I consider the decrease in use of the clitic reflexive pronouns to be linked to the development of another strategy to express the same idea. Moreover, the decrease in agreement in NPs and the strong reduction of affix use in the conjugation system, which itself is a decrease in agreement, coupled with the increased use of subject pronouns in a null-subject language, suggest that VBP seems to be in the initial stages of a typological transition to a morphologically sparser, overt-subject-pronoun language. The question that has rarely been addressed, if at all, is why this apparent typological transition is happening in VBP but not in VEP. One answer involves ecology. In the next two sections, I attempt to respond to this question with a sociohistorical answer that takes into consideration the demographics of colonial and postcolonial Brazil, the languages of different speech communities that were present, and how well they were represented in terms of their respective speakers throughout Brazil’s history.

Table 7.1. Traits of Vernacular Brazilian Portuguese (VBP) compared to Vernacular European Portuguese (VEP) Language Trait

VBP

1

Posttonic /e/ → [i]

2

/t/, /d/ palatalization

3 4

Variable number agreement Variable gender agreement

5

Subject pronoun used as object pronoun Reduced use of reflexive pronouns

Found in all varieties Attested dialectally of VBP Not found Found in many VBPa dialects, particularly in the southeast of the country; considered part of the Standard in Brazil Pervasive in VBP Rarec Exists in some dialects, Not found mainly rural AfroBrazilian; also found in VBP, at low frequency, in predicative constructions Common in VBP Not found

6 7

Greater subject pronoun expression

8 9

Common in VBP, especially with nonanaphoric se Pervasive (VBP is changing to an overtsubject pronoun language)

Greater use of null objects Strongly reduced verbal affixes in the conjugation system 10 Double negation não canto não 11 Less subject-verb inversion in Wh-questions

Greater presence Common in VBP; can be radical especially in the rural sector Pervasive in VBP (canto não is common) Common in VBP

12 Locative em used to express goal role 13 IO DO order with no marking of IO (e.g. ele deu João farinha ‘he gave John flour’)

Common in VBP Found in VBP, in the rural sector

VEP

Not found Not found (VEPb is a null-subject language; overt subject pronouns are found in pragmatically supportive contexts) Lesser presence Not foundd Attested (canto não is not found) Found in pragmatically supportive contexts Rare if present at all Not found

Standard Brazilian Portuguese. Standard European Portuguese. c, d Naro and Scherre found minor evidence of this in some varieties of VEP. They state: “The non-agreeing forms [in VEP] are statistically rare, but we encountered no difficulty in locating and checking a reasonable number of them in only a few days of listening to tape recordings” (1999, 241). a

b

J. Clancy Clements

[ 193 ]

3. Demographic Perspective and Speech Communities in Brazil In this section, we will try to understand the extent of the increasing presence of Africans and their descendants in colonial and postcolonial Brazil. First we look at slaves in Brazil’s sugar production and then in mining. In section 4, we follow this up with an examination of census data from the late nineteenth century and the first half of the twentieth, to see how blacks and pardos were distributed in the five regions of Brazil: Norte, Nordeste, Centro-Oeste, Leste, and Sul.

3.1. demographics in the sugar plantations The slave trade in Portugal began very early on. Ramos Tinhorão points out that throughout medieval times slavery was known and used by the Portuguese not only in war but also in commerce ([1988] 1997, 45, 437n90). He reports that from the mid-fourteenth century there were establishments in the Rua Nova of Lisbon where captured Moors and other Africans (some of whom were even brought from Seville, the entrepôt for Castile) were bought and sold.3 In the second half of the fifteenth century, about nine hundred African slaves were brought to Portugal yearly, such that by 1505 10% of Lisbon was made up of sub-Saharan Africans and 15% of the labor force in rural areas was African (101, 113; Schwartz 1985, 6). During the fifteenth century, Portugal introduced sugar production on its island colonies (Azores, Madeira, Cape Verde, etc.), and there is evidence that by 1520 Brazilian sugar was selling on the market in Antwerp. In the sixteenth century, increasing numbers of African slaves who spoke Kwa and Bantu languages were brought to Brazil to work on the sugarcane plantations, predominantly in Pernambuco, an eastern state of the northeastern region, where Recife is currently the capital, and in Bahia, an eastern state in the Leste region, where Salvador is the capital today (see map 7.1). In the first half of the sixteenth century, about eleven thousand to twelve thousand slaves were being brought yearly to Portuguese Brazil (as opposed to Dutch Brazil) from Angola alone (Schwartz 1985, 184). We know that by 1540 the sugar agriculture was well established in Brazil (16). Sugar mills (engenhos) were built in Porto Seguro, Illheus, and Bahia. Pernambuco and Bahia became the strongest producers of sugar in the sixteenth century. Around 1580, “together, these two captaincies [of Bahia and Pernambuco] probably accounted for some three-quarters of Brazil’s total sugar production” (19). The focus on sugarcane promoted the growth of slavery in Brazil. African slaves were strongly preferred over the indigenous Brazil population

[ 194 ]

chapter seven

Map 7.1. Regions of Brazil with key cities. Derek Roff assisted in the preparation of this map.

because the latter resisted farm labor and was also decimated by disease (e.g., smallpox and measles). Between 1570 and 1630, the “Bahian plantations crystallized into the distinctive social structure that characterized the area for the next two hundred years” (Schwartz 1985, 61). The distinctive social structure referred to by Schwartz here is big concentrations of Africans relative to Europeans. On the whole, the estimated total number of slaves taken to Brazil fluctuates between 3,325,000 and 3,646,000. Regarding this estimate, Curtin notes that even if it was high, “the discovery of new evidence or more accurate statistical inference is hardly likely to reduce it far below three million” (1969, 49). The significance of the estimate for Brazil can be appreciated when it is compared with the estimated total number of slaves taken to Spanish America, which Curtin puts at around 1,552,000, that is, roughly half the number thought to have been taken to Brazil (46).

J. Clancy Clements

[ 195 ]

Table 7.2. Racial/ethnic designation of parents and godparents for 234 baptisms, Engenho Sergipe, 1595–1608. The totals for each of the known racial/ethnic designations are also given as percentages. White

Indian

African

Negro/Crioulo

Mulatto

Unknown*

Father Mother Godfather Godmother

61 43 132 59

42 54 9 21

27 33 6 8

6 8 7 7

0 3 0 7

98 93 70 114

Total Total as %

295 55%

28 5%

10 2%

126 24%

74 14%

Source: Schwartz 1985, 61. *Schwartz (1985, 61n) notes that the 98 unknowns under ‘father’ are the result of illegitimate unions for which no father was recorded at the baptisms. The other cases are due to failure to report the information or lacunae in the documents.

Although we can assume that the presence of the plantations in Pernambuco and Bahia led to large concentrations of Africans, Schwartz provides reliable data from the Engenho Sergipe (engenho ‘sugar mill’), in the form of baptismal records, suggesting that the concentration of the African population was also in smaller units. I have reproduced Schwartz’s table 3.3 as table 7.2. The chapel records contain 234 complete adolescent baptisms, of which 171 are from slave mothers (73%). Schwartz states that the racial origins of the people recorded in the baptismal records constitute a rough gauge of the ethnic proportion of the population at large in this area at that time: whites make up about 32% of the mothers’ population, Indians 40%, and Afro-Brazilians 28% (Schwartz 1985, 61). To summarize what has been said so far, from the sixteenth to the eighteenth century, on the one hand, the states of Pernambuco and Bahia were the key centers of the plantation sugar production, and because of that large concentrations of Africans lived in these areas. On the other hand, there were also smaller plantations, such as the Engenho Sergipe, where African and Indian slaves lived and adapted to the Portuguese culture and religion. This latter type of plantation can be described as societés d’habitation ‘homesteads’, settings where the racial mixture was more balanced and contacts between acculturated slaves and the Portuguese were most likely relatively close. But even as the acculturated slaves and the Portuguese came to interact more within society, and even as there were manumissions, there continued to be a regular influx of new slaves to meet the labor needs on the big plantations.

[ 196 ]

chapter seven

3.2. slave demographics and mining At the end of the seventeenth century, large-scale agriculture in Brazil was limited to sugarcane. That changed. Between 1693 and 1695, large gold deposits were discovered in the interior of Brazil (see Minas Gerais on map 7.1), about two hundred miles inland from the Rio de Janeiro coast, and by 1715, more than eighty thousand blacks and twenty thousand whites were working in the mining areas. Not surprisingly, a struggle emerged in the first quarter of the eighteenth century between interests in mining and those in agriculture. One of the main complaints on the part of the planters involved slaves: in their view, the best ones were being sent to the mines and the plantations received slaves of poorer quality. Consequently, among other things, the second half of the eighteenth century witnessed violent swings in sugar production, leading to its gradual decline. By the advent of the nineteenth century, there were two types of agricultural operations in plantation-heavy areas such as Bahia: the established plantations that depended on the use of large numbers of slaves and the newly formed subsistence farming operations in the Recôncavo area (in Bahia), in which around 70% of the slaveowners held fewer than five slaves each and only twenty-five owners possessed more than twenty slaves. There were few sugar mills in the area; only three slaveowners held more than fifty slaves. In 1819, Bahia’s black slave population was only 31% of the total population (147,000 of 500,000 inhabitants), although the black population was certainly larger, because it included freed blacks. In Salvador, where there was a greater concentration of plantations at that time, around 50% (25,000 of 50,000 inhabitants) were slaves (Schwartz 1985, 440). Thus, from the beginning of the seventeenth century to around 1825, the development of mining led to another major concentration of black slaves, this time farther south in what came to be called Minas Gerais ‘general mines’. Over three centuries, then, the three major concentrations of Africans and their descendants were Pernambuco, Bahia, and Minas Gerais. During the nineteenth century, there was a shift in agriculture toward smaller operations, such as subsistence farming, among other things (e.g., domestic work). With this background, we now examine census data from three periods spanning seventy years. What becomes clear is that the changes in VBP discussed in the first section of this study are not unexpected given the demographic distribution the black population in Brazil from 1520 to 1940.

J. Clancy Clements

[ 197 ]

4. Census Data from 1872, 1890, and 1940 The history of the sugar plantation and mining areas is reflected fairly well in the census data from 1872, 1890, and 1940, although it is not clear how race was determined. African-European mixed-raced people (pardos ‘mulattoes’) are not distinguished from blacks, except in the 1940 census. Therefore, we will not make the distinction here. The overall population of Brazil in 1872, 1890, and 1940 is seen in table 7.3. Relative to the general population of Brazil, the black and pardo population decreases 21 percentage points (pp) (from 62% to 41% of the total Brazilian population) over the 68 years between 1872 and 1940. This represents a decline in the black/pardo demographic of .238 per year during the first eighteen years and .34 per year during the next fifty years. That is, the decline speeds up slightly.5 If we look at the black and pardo population by region, shown in table 7.4, we see that in 1940 at the historical center of major populations Table 7.3. Black and pardo and general population in 1872, 1890, and 1940 in Brazil 1872 1890 1940

Black and pardo population

Total population

6,022,000 (62% of total) 7,597,000 (56% of total) 13,822,000 (41% of total)

9,716,000 13,639,000 33,920,000

Table 7.4. Numbers (rounded) and percentage of black and pardo population in Brazil by region

Nordeste Norte Centro Oeste Leste Sul Totals

1872

1890

1940

Decrease (in percentage points)

2,086,000 (65%) 229,000 (73%) 118,000a

2,211,000 (63%) 306,000 (66%) 151,000a

4,807,000 (49%) 818,000 (61%) 238,000 (38%)

16 12 —

2,918,000 (63%) 671,000 (38%) 6,022,000 (62%)

4,117,000 (64%) 812,000 (29%) 7,597,000 (56%)

6,706,000 (48%) 1,253,000 (10%) 13,822,000 (41%)

16 28 21b

Some statistics were not available. The numbers for Mato Grosso are not available for 1872 and 1890. For this reason, the decrease for the Centro-Oeste region cannot be calculated. However, the totals for Centro-Oeste do contain the numbers from Goiás, which are reflected in the totals. The 21% decrease takes the numbers from Goiás into consideration. However, if we do not take the Centro-Oeste numbers into account, the average decrease is 18%. a

b

[ 198 ]

chapter seven

of African descent in Nordeste (Pernambuco) and Leste (Bahia and Minas Gerais), the percentages are, as might be expected, well above the national average of 41%. As to where the population of African descent is concentrated, the raw numbers and percentages are given in tables 7.5, 7.6, 7.7, 7.8, and 7.9 for each region, by state, to the extent that they were available. In the Nordeste region (table 7.5), Alagoas, in the southeastern corner, experienced the greatest decrease (31 pp), and Ceará (9 pp) and Rio Grande do Table 7.5. Numbers (rounded) and percentages of the black and pardo population in the Nordeste region.

1872 Maranhão Piauí Ceará Rio Grande do Norte Pernambuco Alagoas Paraíba Totals

256,000 (70%) 192,000 (72%) 447,000 (56%) 150,000 (56%) 550,000 (59%) 259,000 (74%) 232,000 (62%) 2,086,000 (65%)

1890 295,000 (68%) 159,000 (75%) 453,000 (63%) 132,000 (56%)

1940

Decrease (in percentage points)

665,000 (53%) 446,000 (55%) 986,000 (47%) 434,000 (56%)

17 17 9 0

606,000 (59%) 1,220,000 (45%) 323,000 (69%) 410,000 (43%) 243,000 (53%) 656,000 (46%) 2,211,000 (63%) 4,807,000 (49%)

14 31 16 16

Table 7.6. Numbers (rounded) and percentages of the black and pardo population in the Norte region.

Amazonas Pará Totals

1872

1890

1940

Decrease (in percentage points)

46,000 (81%) 183,000 (66%) 229,000 (73%)

106,000 (72%) 200,000 (61%) 306,000 (66%)

297,000 (68%) 521,000 (55%) 818,000 (61%)

13 11 12

Table 7.7. Numbers (rounded) and percentages of the black and pardo population in the Centro-Oeste region.

Mato Grosso Goiás Totals

1872

1890

— 118,000 (74%) 118,000

— 151,000 (66%) 151,000

1940 209,000 (49%) 229,000 (28%) 238,000 (38%)

Decrease (in percentage points) — 46

J. Clancy Clements

[ 199 ]

Table 7.8. Numbers (rounded) and percentages of the black and pardo population in the Leste region

1872 Minas Gerais Sergipe Bahía Espíritu  Santo Rio de  Janeiro Totals

1,209,000 (57%) 126,000 (54%) 1,048,000 (76%) 56,000 (68%) 479,000 (58%) 2,918,000 (63%)

1890

1940

1,891,000 (59%) 2,602,000 (39%) 219,000 (70%) 288,000 (53%) 1,428,000 (74%) 2,790,000 (71%) 79,000 (58%) 288,000 (38%) 500,000 (57%)

Decrease (in percentage points) 18 1 5 30

738,000 (40%)

28

4,117,000 (64%) 6,706,000 (48%)

16

Table 7.9. Numbers (rounded) and percentages of the black and pardo population in the Sul region

Paraná Rio Grande  do Sul Santa  Catarina São Paulo Totals

1940

Decrease (in percentage points)

1872

1890

57,000 (45%) 176,000 (39%)

 90, 000 (36%) 168,000 (30%)

152,000 (12%) 174,000 (11%)

33 28

34,000 (21%)

 43,000 (15%)

65,000 (6%)

15

404,000 (48%) 671,000 (37%)

511,000 (37%) 812,000 (29%)

862,000 (12%) 1,253,000 (10%)

36 28

Norte (0 pp) in the northeast of the region showed the least decreases. The decreases in the remaining regions cluster around the 16 pp average decrease in African and African-descendant population for this region. Given the three historical centers of concentration of blacks and pardos, it is interesting to find that their demographic in the Norte region (table 7.6) is so elevated and that the decrease is less than in the Nordeste region.5 In the Centro-Oeste region (table 7.7), the numbers for Mato Grosso were not available in the census. Note, however, that the high percentage of the black and pardo population in Goiás from 1872 to 1940 is among the highest in the country at that time and that the decrease, at 46 pp, is the highest recorded in the whole country for the period in question. For Leste (table 7.8), it is noteworthy that there is an increase of the black and pardo population between 1872 and 1890 in Minas Gerais

[ 200 ]

chapter seven

(2 pp) and in Sergipe (16 pp), and a minimal 2 pp decrease in Bahía during the same time. Also of note is the 20 pp drop from 1890 to 1940 in Minas Gerais. If we compare this to the negligible overall decrease from 1872 to 1940 of 1 pp and 5 pp in Sergipe and Bahía, it seems that there was a significant demographic redistribution in this region. Finally, in the Sul region (table 7.9) we see the sharpest overall decrease of any region, with 28 pp, and large decreases in three of four states, where the decrease in the black and pardo population is between 28 pp and 36 pp. The importance of this demographic information is that, with the exception of the Sul and Centro-Oeste regions, and independently of the almost exclusive monotonic decreases in every region, the blacks and pardos make up over 60% of the overall population in 1872 and 1890 and 48% or more of the population in 1940. I believe that the strong demographic presence of blacks and pardos, who (or whose ancestors) initially were brought to Brazil speaking languages other than Portuguese, is one key factor in the complex linguistic changes that VBP is currently undergoing. This is taken up in the next section.

5. Discussion and Conclusions I began this chapter by asking whether VBP and VEP should be considered two different languages, given the substantial divergence from VEP that VBP has experienced. Reviewing some key features of VBP, including those discussed by Mello and colleagues (1998), thirteen traits were identified that set VBP clearly apart from VEP (see table 7.1). Eight of these traits are either rare or nonexistent in VEP. For traits 7, 8, and 11, the difference between VBP and VEP is that VBP speakers seem to be moving toward conventionalizing the overt expression of subject pronouns (trait 7) and doing away with Subject-Verb inversion in Wh-questions (trait 11); speakers of VEP are decidedly not doing this. The licensing of null objects (trait 8) is constrained by different factors in VBP and VEP; trait 1 (/e/ → [i] in posttonic syllables) is pervasive in VBP but found only dialectally in VEP. Finally, negation with sentence-final não is common in VBP but nonexistent in any variety of VEP. A related feature, double negation (não canto não ‘neg I sing neg’), is pervasive in VBP but not common in VEP, although it is attested. The order of objects in double object constructions with two full noun phrases (e.g., ele deu João farinha ‘he gave John flour’) is impossible in any variety of VEP but is found in rural varieties of VBP. Given these differences, along with many others, a case could arguably be made for considering VBP and VEP linguistically different languages.

J. Clancy Clements

[ 201 ]

However, from a population-genetics perspective, the actual question is not whether there is substantial linguistic divergence between VBP and VEP, but rather whether each variety belongs to its own separate speech community. I suggested that VBP and VEP constitute a polytypic language. That is, although the two varieties differ substantially in terms of their respective linguistic features and continue to diverge, they share the same literature, and they continue to nurture common social, cultural, and historical ties. With few exceptions, they share the same conventions in the written language, as well. The other question I posed was why VBP and VEP have been diverging. In a populations-genetics model, languages are considered analogous to species, either plantish (Croft 2000) or parasitic (Mufwene 2001), and languages-as-species are made up of lexical and structural linguistic features. With the influx of such a large number of Africans into Brazil, from the sixteenth century onward, the linguistic feature pool available to speakers and learners of Portuguese was decidedly influenced by linguistic features from the African languages of the slaves. In sections 3 and 4, I discussed the vast numbers of slaves transported to Brazil from the sixteenth to the end of the eighteenth century and surveyed the distribution of blacks and pardos in three censuses, from 1872, 1890, and 1940. One outstanding fact that emerges from this overview is that the presence of blacks and pardos has been and continues to be strong. Based on the census data, blacks and pardos made up well over half of the total population up to the beginning of the twentieth century and the Norte, Nordeste, and Leste regions have had an especially dominant presence of blacks and pardos. Consequently, African-language features were more strongly represented and there was additional innovation through language learning. Silva (2004, 99–108) argues along the same lines, stating that the Africans and their descendants were the principal propagators of VBP. I consider it significant that large-scale agricultural operations gave way to smaller operations after the shift to mining in the eighteenth and nineteenth centuries. We then see diversified agricultural operations that, in turn, led to more contact between blacks and pardos, on the one hand, and between both of those groups and whites, on the other. By the time these small-scale and subsistence agricultural operations developed, many blacks and pardos would have spoken a mildly to substantially innovative variety of Portuguese that would arguably have propagated within their immediate speech communities. Over time and with increased mixing between these populations, these propagated innovations would have experienced further propagation, especially given the strong presence of blacks and pardos in all parts of Brazil until the beginning of the twentieth

[ 202 ]

chapter seven

century, particularly in the Norte, Nordeste, and Leste regions, throughout the first half of the twentieth century. The restructuring of VBP relative to VEP is doubtless exceedingly complex. But one key aspect of the propagation of the innovations in VBP has been the presence of a diverse and heterogeneous linguistic feature pool, which in turn is due to the strong presence of Africans and their descendants over the past three and a half centuries. Naro and Scherre’s position is not that different from the one developed in this study. They state, “Modern Brazilian Portuguese is the natural result of the inherent drift over time in the language brought from Portugal, undoubtedly magnified in Brazil by the profusion of contact among adults, speakers of languages of very diverse origins, and by the nativization of this language by the communities formed by these speakers. Our proposal is summarized thus: a confluence of motivations, with some stronger in the beginning of the historical process and others more relevant in subsequent historical stages” (2007, 69, my translation). The difference between Naro and Scherre’s view and mine is one of degree, not of kind. More specifically, with the exception of trait 10 in table 7.1 (sentence-final negation), all traits in that list can be characterized within Naro and Scherre’s statement. Specifically, the other traits arguably result from a complex interaction of innovation and propagation of linguistic features among blacks, pardos, and indigenous language speakers, followed by the further propagation of these features to the population at large as the different races began to interact more starting at the end of the eighteenth century, with the decline of large-scale plantation agriculture. Both the innovative traits and their propagation can be explained, at least in part, by taking into account the nature of the demographic distribution. There is a great need for more in-depth studies on exactly how the innovations of the blacks and pardos subsequently propagated throughout the black and pardo population and, subsequently in the rest of the Brazilian population, thereby leading to the divergence between VBP and VEP. The contribution of this chapter can hopefully serve as a small first step. Notes My thanks to Judy Bieber and Margo Milleret for advice on various aspects of Brazilian history, to Ana Luís and Patricía Amaral for their input on the discussion of Vernacular European Portuguese traits, and to Alan Baxter for his input on the traits of Vernacular Brazilian Portuguese. Finally, my sincere thanks to Salikoko Mufwene and Bruce Mannheim for their comments on a former version of this study. 1. For a detailed examination of the features of VBP, see Lucchesi et al. 2009. 2. Güldemann and Hagemeijer (2006) cite many examples of sentence-final negation from a rich variety of African languages. As examples, I include one from a

J. Clancy Clements

[ 203 ]

Mbundu language and one from a Kongo language, two Bantu language clusters that can be argued to have had an impact in Brazil. (1)

Ha-tú-vi-mbandà-kó. (Umbundu) NEG:1pl-8-doctor-NEG ‘We are not doctors.’

(2)

Ka n-zébia lâri ko. (Laadi) NEG 1sg-know Laadi NEG ‘I don’t understand Laadi.’

3. Tinhorão also mentions the record of a nun from the Chelas Convent who bought a female Moor as a slave in 1368. Thus, the Church also participated in the slave trade from early on. 4. In the 2000 Brazilian census, blacks and pardos made up 45% of the overall population, 75,900,000 of 169,799,000), an increase of 4 pp in the sixty years from 1940 to 2000. That is, the black and pardo population seems to have stabilized, showing only a slight increase. This is probably due to various factors, including improvements in socioeconomic conditions and access to the better health facilities, lifestyle changes, and, particularly for the pardo population, an increase in racial integration in the general population. However, this is an empirical question and would need to be studied. 5. For the purposes of this study, the reasons for the redistribution of population will not be addressed. Of primary interest here is the extent to which the black and pardo demographic is represented in each region.

References Croft, William. 2000. Explaining language change: An evolutionary approach. London: Longman. Curtin, Philip. 1969. The Atlantic slave trade: A census. Madison: University of Wisconsin Press. Güldemann, Tom, and Tjerk Hagemeijer. 2006. Negation in the Gulf of Guinea creoles: Typological and historical perspectives. Paper presented at the annual meeting of the Associção de Crioulos de Base Lexical Portuguesa e Espanhola, University of Coimbra, June 26–28. Guy, Gregory. 1989. On the nature and origins of popular Brazilian Portuguese. In Estudios sobre el español de América y lingüística afroamericana: Ponencias presentadas en el 45 Congreso Internacional de Americanistas, 227–245. Bogotá: Instituto de Caro y Cuervo. Lucchesi, Dante, Alan Baxter, and Ilza Ribeiro, eds. 2009. O Português Afro-Brasileiro. Salvador, Brazil: EDUFBA. Mello, Heliana R. de, Alan N. Baxter, John Holm, and William Megenney. 1998. O português vernáculo do Brasil. In América negra: Panorámica actual del los estudios lingüísticos sobre variedades hispanas, portuguesas, y criollas, ed. Matthias Perl and Armin Schwegler, 71–137. Madrid: Iberoamericana. Mufwene, Salikoko. 2001. The ecology of language evolution. Cambridge: Cambridge University Press. Naro, Anthony, and Marta Scherre. 1993. Sobre as origens do português popular do Brasil. Revista Delta 9:437–454.

[ 204 ]

chapter seven

———. 1999. Variable concord in Portuguese: the situation in Brazil and Portugal. In Language change and language contact in pidgins and creoles, ed. John McWhorter, 235–255. Amsterdam: Benjamins. ———. 2007. Origens do português Brasileiro. São Paulo: Parábola. Schwartz, Stuart B. 1985. Sugar plantations and the formation of Brazilian society. Cambridge: Cambridge University Press. Schwegler, Armin. 1996. La doble negación dominicana y la génesis del español caribeño. Hispanic Linguistics 8:247–315. Schwenter, Scott A. 2006. Null subjects across America. In Selected papers of the 8th Hispanic Linguistics Symposium, ed. Carol Klee and Timothy Face, 23–36. Somerville, MA: Cascadilla Press. Schwenter, Scott A., and Gláucia Silva. 2002. Overt vs. null direct objects in spoken Brazilian Portuguese: A semantic/pragmatic account. Hispania 85:577–586. ———. 2003. Anaphoric direct objects in spoken Brazilian Portuguese: Semantics and pragmatics. Revista Internacional de Lingüística Iberoamericana 2:109–133. Silva, Rose Virgínia Mattos e. 2004. Ensaios para uma sócio-história do Português Brasileiro. São Paulo: Parábola. Tinhorão, José Ramos. [1988] 1997. Os negros em Portugal: Uma presença silenciosa. Lisbon: Caminho. Ronald Wardhaugh. 2002. An introduction to sociolinguistics. Oxford: Blackwell.

8 * Maya and Spanish in Yucatán: An Example of Continuity and Change barbara pfeiler

1. Introduction They came to the coast of Yucatan to a province called [. . .] Maya after the language of Yucatan Mayathan, that means language of Maya.

Diego de Landa, Relación de las cosas de Yucatán, 1566

Maya, the name given to the language by its speakers, is spoken on the Yucatán Peninsula, in the states of Yucatán, Campeche, and Quintana Roo, Mexico, and in northern Belize.1 Although stigmatized since the Spanish conquest, Maya was used by missionaries and plantation owners in their efforts to achieve social and religious domination. The fact that “the colonial period was a time in which Maya spread as the predominant language of Yucatán rather than a time of decline for the region’s native language” (Lentz 2009, 135) shows that, unlike the case in other regions of Mexico, the Maya-speaking society was able to preserve spaces in which to use the native language during colonial rule. This made it the de facto language of the Yucatán Peninsula, and it continues to be spoken today. The Spanish conquest established Spanish as the language of power, but only a small proportion of the population in Yucatán actually used it. No concerted attempts were made by the colonial administration to promote the use of Spanish within the indigenous population. Indeed, soon after the conquest, Spanish lost its importance as a mark of differentiation between indigenous and nonindigenous people. Maya continued to be the mother tongue of the indigenous population and, based on a 1813 Spanish Crown report, Maya was spoken by the entire population, not just “Indians, mestizos and pardos,” but also by the Spanish and members of the “lower castes” (Gabbert 2004, 21). According to Nancy Farriss, it is possible to detect parallel processes of Hispanization and Mayanization when discussing language purity and diffusion: More than a lingua franca, Maya was the primary language of all the colony’s native-born inhabitants of every caste. . . . In the old days, even

[ 206 ]

chapter eight

(or perhaps especially) among the upper-class Yucatecans of impeccably Spanish blood lines, Maya was the first language. . . . Maya was in a very real sense, then, the creoles’ mother tongue, the language with which they continued to feel more comfortable as adults and used by preference “not only among the Indians but also at home with their own children, giving as their reason that it is easier to pronounce” (Farriss 1984, 112, quoting from a document of “visitas pastorales”).

Because Maya was the main language used in sermons, for general Christian evangelization, and in written texts with Latin signs in Maya, the native language was well rooted in daily life. In comparison with other parts of the Viceroyalty of New Spain by the end of the eighteenth century, most of the population had no need to learn Spanish (Gabbert 2004, 21–22). The early Franciscan friars had an intellectual curiosity about the indigenous population and their traditions, and the documents they produced constitute the first studies of the Maya language. Christian thought was translated from Latin and Spanish into Maya by the use of Maya reducido, the language of the colony (Hanks 2010, xiv). During the sixteenth century, Diego de Landa collected valuable data on linguistic contact between the Spanish and the Maya, and the volumes of the Chilam Balam2 were written, covering a large chapter of colonial Mayan literature. In addition, more than one hundred Mayan authors created a corpus representative of the colonial-era lexicon. Usage of Maya as a legal language was discontinued in the nineteenth century and increased importance was given to reading and writing in Spanish. This engendered an aversion toward Maya culture among the Spanish. During the Porfiriato (1876–1911), a national aim was the establishment of a mestizo Mexico with a strong, united identity. In Yucatán, however, the so-called indigenous policy took the form of incorporating the Maya population, a large proportion of the approximately 250,000 inhabitants living on the peninsula in 1895, into socioeconomic and urban development. The Yucatán Peninsula was not truly integrated into the rest of Mexico until local railroads were joined with the national rail network in the late 1950s. This isolation is one of the most important factors affecting the formation and use of Yucatán Spanish, a variant differing from that of central Mexico. Beginning in the 1970s, the federal government began to modernize the region through highway projects, tourist industry development at Cancún, and oil extraction off the Campeche coast. Modernization in the twentieth century made the spaces of modern life preferably nonindigenous, and Mayan inhabitants retreated to the closed

Barbara Pfeiler

[ 207 ]

areas of community life. Since the 1940s, the importance of the Maya language in public life has steadily decreased. This has been in response to far-reaching regional transformations such as agricultural and tourism development, internal rural-to-urban migrations associated with urban growth, federal policies aimed at proletarization and nationalization of the peasant population, federal education policy,3 linguistic centralism, and an over-arching strategy of Hispanizing rural populations (Pfeiler and Zámišová 2006, 287).

2. Distribution of Maya and Spanish Speakers in Yucatán Maya has a variable presence in the language mosaic of the Yucatán Peninsula. In western and southern Campeche state, Spanish predominates, a consequence of heavy immigration from central Mexico, while Maya is spoken in the rest of the state, alongside other Maya languages such as Ch’ol and K’iche’. The latter are the result of the immigration of Maya speakers from Chiapas state and Guatemala. Three subregions exist in Quintana Roo state: the Caribbean coast, where Spanish and English predominate; the center, where Maya is the primary language; and the south, where Spanish prevails (Pfeiler and Zámišová 2006, 282). The distribution of Maya speakers in Yucatán state differs from that within Quintana Roo and Campeche. Particularly striking are the high ratios of bilingualism and the fact that Maya is spoken in all of the state’s 106 municipalities. The state has been divided into eighteen subregions,4 which are then grouped into seven regions (Metropolitan, Coastal, CattleRanching, Henequen-Growing, South, Corn-Growing, and West), each representing a certain degree of homogeneity relative to territorial occupation, demographic dynamic, and agricultural/livestock specialization. See map 8.1. Most Maya speakers in Yucatán live in the southeastern portion of the state, near the border with Quintana Roo. The highest percentages are in the southern region (5a, 5b) and in the corn-growing east (6a) and west transitional (7c) subregions. The citrus-growing-south subregion (5a) is characterized by the use of intensive agriculture. The southern cone area has high natural biodiversity and no single defining economic activity. In contrast, the corn-growing-east subregion is dominated by corn cultivation, which is closely linked to the traditional Mayan lifestyle. The zones with the highest concentration of Maya speakers are also those in which indigenous traditions are more conspicuous. The exception is the west transitional subregion, which is near Mérida and is in transition from henequen production to corn and citrus production. There are surprisingly high numbers of Maya speakers in the Metro-

[ 208 ]

chapter eight

Map 8.1. Maya speakers in Yucatan state by socioeconomic region. My warmest thanks to Ana García de Fuentes for her help in producing this map.

politan subregion (1a), consisting of Mérida and its surroundings, and the west-suspended subregion (7b), in which there is no important economic activity. Rural depopulation since the 1950s has resulted in the high percentage of Maya speakers in the capital compared to the rest of the state, but they constitute a minority within the overall metropolitan region. The cattle-ranching subregions (3a, 3b, 3c, and 3d) cover the largest area but are proportionally less densely populated. Only the cattle-ranchingeast subregion has a moderate percentage of Maya speakers, while the remaining three subregions have low levels. Finally, the henequen-growing subregions (4a, 4b) and the coastal zones (2a, 2b, 2c) have relatively fewer Maya speakers.5 However, their combined population represents a small portion of the state’s overall population and therefore has only a minimal effect on the overall statewide proportion of speakers.

Barbara Pfeiler

[ 209 ]

This overview highlights the role that the geography of the state’s socioeconomic regions plays in determining language use. However, changes have occurred during the past thirty years. For example, regions that were much more conservative in terms of culture and languages, such as the corn-growing-center subregion (6b), currently have a medium-to-low percentage of Maya speakers (Pfeiler 1999). This panorama looks a bit different when viewed in terms of numbers of monolingual speakers and their proportions relative to the populations of their respective regions. From this perspective, the highest monolingual Maya population is in the corn-growing region, followed by the south, the west, and the cattleranching regions (fig. 8.1). The most developed zone is that influenced by Mérida, the former hub of the henequen industry and currently the center of Western-style industrial, commercial, and cultural development in the state. Usage of Spanish parallels this greater degree of development, but without complete elimination of the native language. The massive social changes experienced by the henequen-growing zone beginning in the 1960s are also evident in the limited number of Maya speakers it contains. The city of Mérida and the coast have the lowest numbers of Maya speakers, fewer than 20 percent, whereas the south and corn-growing zones have large proportions of bilingual speakers, over 64 percent of the population. Both zones also contain a high percentage of monolingual Maya speakers. If these are combined with bilingual speakers, Maya speakers represent over three-quarters of their populations. The state’s cattle-ranching and henequen-growing zones have percentages between these two extremes.

Corn-growing South West

Maya monolinguals

Cale-ranching

Bilinguals Spanish monolinguals

Henequen-growing Coastal Metropolitan 0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Figure 8.1. Monolingual Maya and bilingual Maya/Spanish speakers in the Yucatan state by socioeconomic region

chapter eight

[ 210 ]

From the highly urbanized areas of Mérida and its surroundings to the Quintana Roo border, the gradual increase in the proportion of Maya speakers is indicative of a transition toward the indigenous world in which usage and customs reflect greater maintenance of Mayan historical and cultural heritage. Census data from successive decades have shown a continuous reduction in the number of monolingual Maya speakers with an accompanying decrease in the number of bilingual (Maya-Spanish) speakers, especially over the past thirty years (fig. 8.2) (INEGI 2006). Decreases in the number of Maya speakers since 1940 are due to cultural changes in Yucatán associated with emigrations of Mayans to urban centers beginning in the mid-twentieth century. These weakened the traditional organization of Maya culture by removing emphasis on the milpa agricultural system as the main economic activity and by allowing the Maya to acquire more nonindigenous technical skills (Baños 2000; Gaskins 2003; Mossbrucker 1992). Rural and urban lifestyles have tended to merge, with campesinos also doubling as day laborers, workers, small-scale merchants, and employees. Domestic family units in this region of Mexico are modern in that they exhibit diversification in the distribution of work and have the fundamental characteristic of ubiquitous multitasking by each family member. Family members go to work during the week and then come together again on weekends. Access to information is no longer restricted to cities, since television, radio, and fiber-optic cable (i.e., Internet access, etc.) now cover the entire peninsula. The disappearance of traditional Maya housing, in which open space was essential to daily life, is lead-

80% 70% 60% Persons: Maya monolinguals 5+ years

50% 40%

Persons: Bilinguals 5+ years

30%

Persons: Spanish monolinguals 5+ years

20% 10% 0% 1930

1940

1970

1990

2000

2010

Figure 8.2. Percentage of bilingual (Maya-Spanish) and monolingual (Maya or Spanish) speakers in Yucatan state

Barbara Pfeiler

[ 211 ]

ing to a social change, and television now congregates the family in the interior space. The current processes of emigration and urbanization in Yucatán endanger the survival of the Maya language. This development is accelerated by the tendency of young bilingual people to use Spanish to communicate among themselves. The social domains that tended to be separated by exclusive use of Maya or Spanish are increasingly covered by Spanish only. In an increasingly larger number of families, Maya is no longer transmitted to children as the mother tongue, and bilingualism is thus transitioning to Spanish monolingualism. Results have included loss of fluency in the Maya language and what Maya speakers call a mixed variety (Pfeiler 1998, 131–134). Maya speakers currently distinguish two varieties: Hach Maya ‘true, pure and old Maya’ and Xe’ek’ (literally ‘mix’ or ‘jumble’) ‘modern, mixed, and Spanish-influenced’. Hach Maya borrows no vocabulary and preserves words that are no longer regularly used.6 It is said that Hach Maya is spoken exclusively by the elderly in the eastern part of Yucatán. Xe’ek’ is characterized by a number of Spanish nouns and verbs, which are adapted to the structure of Maya. It is the most widely spoken of the two varieties (Pfeiler 1998, 2012). The differences between Hach Maya and Xe’ek’ Maya reflect the discrepancy in linguistic identity experienced by the Maya of Yucatán. Hach Maya is perceived as “good” Maya (Armstrong-Fumero 2009, 362). Even though Hach Maya is no longer spoken by the younger generations, it is positively valued for being a strong potential source of identity. For instance, there are frequent and grandiose references to the pre-Hispanic Maya culture. The Xe’ek’ variety is perceived as modern, mixed, and Hispanized. Comparison of the endemic modern variety with the supradialectal, idealized one results in the community’s own linguistic image. This idealization of the older variety and disparagement of the modern one creates an intriguing dynamic in which the “true” Maya evokes a certain nostalgia, but one lacking any intention to recover its original status. This local Yucatecan ideological formation is not unique in Mexico; it also has been registered in the Nahuatl speaking society of the Malinche Volcano region (Hill and Hill 1986). In the next section, I focus on the different types of transfers found in the speech of bilinguals whose mother tongue is Maya and who use it in colloquial speech and Spanish in official situations.7 As Lipski (2004, 2) stated, Spanish-Maya bilingualism in Yucatán encompasses a wide range of Spanish language proficiency levels. Bilinguals who speak Spanish fluently exhibit little or no Maya influence in their speech. In contrast, bilinguals who are more adept with Maya use phonetic and grammatical

[ 212 ]

chapter eight

features of Maya, as well as constructions that may not be direct calques from Maya but are clearly characteristic of the interlanguage stage. The influence of Maya on the Spanish spoken in urban areas of Yucatán has been thoroughly studied, so I focus here on bilinguals more fluent in Maya, although I also highlight some characteristics of the Yucatecan Spanish spoken in urban areas (see section 3.2). The data used here are from the spontaneous speech of bilingual speakers living in one of three towns with differing degrees of contact between Maya and Spanish: Yalcobá, in the cattle-ranching area (map 8.1, 3a); Cantamayec, in the west transitional region (map 8.1, 7c); and Tekantó, in the henequen-growing-suspended region (map 8.1, 4b), where recent socioeconomic changes have forced inhabitants to learn Spanish as a second language.8

3. Cross-Linguistic Influences between Maya and Spanish 3.1 borrowings from spanish into maya The Maya language has coexisted with Spanish since the initial arrival of Europeans more than 450 years ago, so Maya has had ample opportunity to integrate Spanish vocabulary. Spanish borrowings have been adapted to the grammatical structure of Maya, allowing the language to adjust to and subsist in the new sociohistoric context. As in other cases of MayaSpanish contact (Brody 1987), Spanish conjunctions and discursive particles are the elements most frequently borrowed into Maya. Some Spanish nouns have been borrowed to refer to modern aspects of life, such as fráanses ‘baguette’ (< Spanish pan francés), commercial names such as coca ‘Coke’, le pepsi-o’ ‘the Pepsi’ or le champuhó’ ‘the shampoo’. Older Spanish loans are still used by elderly Maya speakers, for instance, pozuelo ‘pitcher’ and bijiga ‘balloon’.9 Some Spanish loans used as synonyms of Maya terms have led to the loss of the latter, thus reducing the extended functional range of Maya elements. For instance, aanio (< Spanish año) ‘year’ is used instead of ha’ab; kosiina (< Spanish cocina) ‘kitchen’ instead of k’oben; and flores ‘flowers’ instead of lool. Nouns, such as the equivalents of ‘flower’ and ‘grapes’ are borrowed in their plural form (flor-es, uva-s), but with the meaning of singular, as in U flor-es-il a pool? ‘Is the flower from your head?’. When used in the plural, these nouns take the Maya plural suffix -o’ob: le parselarios-o’ob’ (< Spanish los parselarios) ‘the lot owners’. This occurs because in Maya nouns referring to number are neutral (Lucy 1992, 46–48), and the Spanish plural suffix is considered a root morpheme.

Barbara Pfeiler

[ 213 ]

Maya does not have agreement for number. However, the Spanish numerals from one to five have substituted for the corresponding elements in Maya, producing agreement while eliminating the obligatory numeral classifier:10 Ts’a kaxtik cuatro ok(o)m-o’ob ‘you just found the four pilasters’ is said instead of Ts’a kaxtik kan p’éel ok(o)m ‘you just found the four-inanimate pilasters’.11 Spanish nouns and adjectives in Maya can bear a suffix for deixis (e.g., te’ sa(a)bado-a’ ‘this Saturday’), for negation (e.g., the enclitic -i’ in ma’ hach suave-i’ ‘it is not easy’), or for location (e.g., tarde-i’ ‘in the afternoon’). The -e’ topicalizer is used with Spanish nouns, but also with adjectives, numerals, adverbs, prepositions, and interrogative particles: in hermano-e’ ‘my brother’, cada in cada-e’ ‘every time or frequently’; desde antes-e’ ‘since before’. Diminutive and derivational suffixes are not productively used with any adjective or noun (Stolz 1998). In our data, Maya adjectives with the meaning of ‘small’ take the Spanish diminutive -ito:12 chichan-ito ‘small’ or mehen-ito ‘small, young’. Few Maya words take the Spanish augmentative suffix -ote: p’urux-ote, meaning gordote ‘very fat man’. When used as stative predicates, Spanish nouns and adjectives can take the Maya absolutive person marker: desde chich(a)nene’ vaquero-en ‘I’ve been a cowboy (vaquero) since I was a child’; h(j)ubilado-en ‘I’m retired’. Inchoative verbs are derived from Spanish nouns and adjectives. Examples are: (1) túun (= táan u) huevo-tal

asp=3erg egg-inch-inc ‘it is egging/laying eggs’ túun tarde-tal-e’

asp=3erg late-inch-top ‘when it is getting late’

Since the late seventeenth century, Spanish verbs have been morphologically adapted to Maya verbal conjugation (Stolz 1998, 166). Spanish verbs are borrowed by direct insertion of infinitive-like stems (Wichmann 2004) and used as transitive and intransitive predicates.13 In Maya active and inactive intransitives in the imperfective aspect have a verbal structure similar to that of possessive nouns, while perfective inactives have a structure analogous to that of predicative nouns:14 When adapted to the intransitive verb classes, Spanish verbs appear as active intransitives (2, 3), which in Maya correspond to derived verbs,

[ 214 ]

chapter eight

with the antipassive stem derived from a noun (“verbal nouns” in Bricker et al. 1998, 362) or from a root transitive.15 (2) Active imperfective

tin ts’íib, tin grábar

asp=1erg write asp=1erg record ‘I’m writing, I’m recording’ (3) Active perfective

le kamion boltear-nah te’elo’

det truck tip over-perf dist ‘the truck tipped over there’

When Spanish intransitives appear in a syntactic subordinated position, they follow the structure of the donor language and appear without any aspect marking (4, 5): (4) Active imperfective

leti’ in radio pero ma’ u tokar

3pro 1poss radio conj neg 3erg play ‘this is my radio but it doesn’t work’ (5) Active imperfective he’ a grabar-e’

asp 2erg record-deic3 ‘will you record’

In these examples, borrowing from Spanish is direct with a noun phrase such as (5) ‘certainly it is your recording’. The borrowings in (6) and (7) are even more like nouns: (6) lela’ ma’alob u desarollar beeora prox well 3erg grow now

‘this is growing well now/this has a good development now’ (7) wa mina’an poder a taal-e’

part nothing can 2erg come-top

‘if you cannot come (if there is no possibility of your coming)’

The fact that the Spanish infinitive-like stems can be used as nouns as well as intransitive verbs substantiates the categorial flexibility in Maya proposed by Lois and Vapnarksy (2006).

Barbara Pfeiler

[ 215 ]

Spanish infinitive-like stems in the context of transitive predicates are always recognizable as such, because of the applicative -t. For example: (8)

Transitive imperfective k-in lasar-t-ik wakax

impf-1erg rope-app-inc cow ‘I rope the cow’ (9)

Transitive imperfective negative ma’ tin aguantar-t-ik in xot kih

neg asp=1erg stand-app-inc 1erg cut henequen ‘I can’t stand to cut henequen’

The applicative is also used with transitive verbs in passive (10), but also with reflexive verbs (11). (10) Imperfective passive

ba’ax k-u tratar-t-a’al

what impf-3erg tratar-app-pass ‘what is it about’

(11) Imperfective reflexive

k-in apurar(-t-i)k-inba

impf-1erg apurar(-app)-inc-1refl ‘I’m in a hurry’

Spanish verbs with e- and i- conjugations are primarily borrowed as transitives, such as in (12): (12) Transitive imperfective ma’ a creer-t-ik-e’ex-e’

neg 2erg believe-app-inc-2pl-top ‘you don’t believe it’

Transitive verbs can be derived from Spanish adjectives by adding the suffix -t (13), or by use of the suffix -k(í)in-s (14), which in Maya is used to derive causative transitive verbs from inchoative, affect, and positional stems (Bricker et al. 1998). (13) Transitive derived verb with -t

k’abéet u responsable-t(-ik)-uba

necessary 3erg responsible- app-(inc)-3refl ‘he has to take responsibility’

[ 216 ]

chapter eight

(14) Transitive derived verb with -kins sáam a ka’ah sucio-kins a k’ab

recently 2erg twice dirty-caus 2poss hand ‘you recently got your hand(s) dirty again’

Most of these examples come from Tekantó, where the Maya population has been in constant contact with Spanish since the colonial era; consequently, the Maya spoken there contains a large variety and a high frequency of Spanish borrowings, aside from involving occasional codemixing, as in (15): (15) cada unoe’ debe ser u kaláant-ik u pak’al

each-top must be 3erg protect-inc 3poss crop ‘each person must take care of his crops’

The analysis of data from the three towns investigated for this chapter shows a continuum in the usage and frequency of borrowings, reflecting variation in the duration of contact between the languages. Maya speakers who have had only sporadic exposure to Spanish use Spanish nouns and certain adjectives and adverbs, but far fewer verbs. Their verb-borrowing strategy is the same as in the speech of fluent bilinguals; namely, Spanish infinitive-like stems are borrowed into Maya; bare infinitive forms take the inflection of intransitive verbs; and infinitivelike stems with the applicative -t take the inflection of transitive verbs.

3.2 transfers from maya into spanish The Spanish spoken in Yucatán has been characterized as unique among the varieties spoken in Mexico. This uniqueness is due to “Yucatecan peninsular insularity” (Alvar 1969, 169) and the influence of the Maya on the Spanish.16 Yucatán Spanish includes Maya words for fauna, flora, food, and local insults, among others, as demonstrated by Alvar (1969), Amaro Gamboa (1984, 1987), Barrera Vásquez (1980), Lope Blanch (1971, 1987b), and Nykl (1930). These are frequently used for comical effect in regional literature and facetious poetry. Maya terms are also commonly used in reference to particular body parts and serve to fill spaces left by taboo words in Spanish. Indeed, certain loanwords, calques, and hybrid lexemes are almost more common in the colloquial speech of Spanish monolinguals than in the Spanish of bilinguals.17 This includes familiar usage in upper and middle socioeconomic classes, not just in lower classes. Some of the most interesting aspects of Maya’s influence on Spanish

Barbara Pfeiler

[ 217 ]

are phonological, both segmental and prosodic (Nykl 1930, 453; Michnowicz 2006). The glottal stop “imparts a picturesque tone to the Mayan pronunciation of Spanish and its glottalized or emphatic consonants” (i.e., consonantes heridas) (Lope Blanch 1987a, 84). The glottal stop in conjunction with stress can cause vowel lengthening, imparting a specific rhythm to the (Spanish) utterance (García Fajardo 1984, 29). The incidence of its appearance in “loanwords” and Spanish words, the substitution or phonetic “mutation” of stress by this glottal stop, and the presence of vowel lengthening suggest an interpretation of glottalization as the union between the two linguistic systems (Colazo Simon 2007, 191–194). In addition to the insertion of the glottal stop, other factors contribute to the distinct sentence intonation and rhythm of Yucatán Spanish, for example, the influence of the Maya syllabic pattern and consequent morphophonological reduction in Yucatán Spanish. The reduction of the final consonant-vowel combination in words such as /reboso/ > /rebós/ ‘shawl’ or /gripa/ > /grip/ ‘influenza’ manifests a tendency to adapt certain words to the CVC syllable structure of Maya (Langer-Blaha 1986, 16–19). Compared to Spanish monolinguals employing the Maya vocabulary, Maya monolinguals and bilinguals with only limited exposure to Spanish transfer Maya nouns referring to their cultural habits when they speak Spanish. Examples include the elements of a palm-roofed house, agricultural implements, and terms used in sacred or traditional ceremonies. These nouns can be suffixed with the Spanish plural -es (e.g., ooch-es ‘opossums’) or diminutive -ito (e.g., muluch-ito ‘a small pile’, from Maya múul tuunich ‘rock pile’; chuchul-ito ‘dry, old, wrinkled’, from Maya chuchul ‘dried’). Maya adjectives and verbs can nominalize in Spanish by adding the derivative agent noun suffixes -ero and -on(a), as in ch’uhuk-ero ‘the candy vendor’, (from Maya ch’uhuk ‘sweet’) and but’-ona ‘heavy eater’ (from Maya but’ ‘fill up, stuff’). The derivative noun suffix -ero seems to be productive since it also occurs in nouns that do not denote an agent, as in sak’-era ‘itch’, from the adjective sáak’ ‘itchy’. Bilingual speakers also commonly use Maya particles, such as chen ‘only’, han ‘quickly’, hahan ‘to do brusquely or quickly’, sen ‘constantly’, and ka’ah ‘and, that, when’. Other loanwords include the quantifier láah ‘all’ and adjectives such as mehen ‘small, young’, chan ‘little, small’, and senkech ‘much’. The list of hybrid compositions documented by Amaro Gamboa (1984) can be expanded with two types of compounds, that is, noun + noun and adjective + noun. Even Maya ideophones have been documented: huruts or hirin huruts (hilin huluts), an action noun meaning ‘(somebody who) enters in a rush or quickly’. See table 8.1. Spanish monolinguals borrow Maya verbal roots primarily using the

chapter eight

[ 218 ]

Table 8.1. Hybrid compounds in modern Maya tisera-che’ plancha-ook way-lechona sit (g) rillo pibi-pollo

‘specific poles in a palm-roofed house’ (Spanish tijeras ‘scissors’ + Maya che’ ‘tree, pole, wood’) ‘flat foot’ (Spanish plancha ‘iron’ + Maya ook ‘foot’) ‘ghost’ (Maya wáaya + Spanish lechona ‘young sow’) ‘cricket’ (Maya síit’ ‘the jump’ + Spanish grillo ‘cricket’) ‘chicken cooked in an earth oven (pollo pibil)’ (Maya píib-i(l) ‘underground cooked’ + Spanish pollo ‘chicken’)b

Way m.a.: This is an ancient Maya name from Nahualli meaning ‘a family member of the necromancers or witches who is an animal and who, through a pact with the devil, transforms fantastically; and anything bad that happens to that animal also happens to the related witch’ (“un familiar que tienen los nigrománticos, bruxos o hechiceros, que es algún animal, que por pacto que hace con el demonio, se convierten fantásticamente; y el mal que sucede al tal animal, sucede también al bruxo cuyo familiar es” [Barrera Vásquez 1980, 156]).

a

In urban Yucatán Spanish this composition has the same meaning but is used in reverse syntactic order, i.e., pollo pibil.

b

DO-strategy (Amaro Gamboa 1987). Some examples (uncommon among fluent bilinguals and Spanish monolinguals) include hacer k’el ‘to toast’, from k’éel ‘toast’ and hacer wek’ ‘to crash’, from the Maya transitive root ‘dash to pieces’. In opposition to Spanish monolinguals, bilinguals transfer Maya roots and integrate them into Spanish using the Spanish conjugation pattern of derived verbs, with the ending -ear to form transitive, intransitive, and reflexive verbs, as in (16, 17, 18):18 (16) Transitive verb

siempre ch’op-ean los ojos

‘they always pick at the eyes (temporarily blind)’ (17) Intransitive verb

los señores iban a xoch’-ear

‘the men were going to lookout for’ (18) Reflexive verb

El melón desde hach chicos se ts’uk-ean ‘the melon(s) rot when very small’

In addition, Spanish verbs can be derived from Yucatec Maya nouns: pok’ol-ear ‘to become infested’ (pok’ol ‘flying insect’); or chechon-ear ‘to cry’ (chéech ‘whiner’). Finally, bilinguals maintain the Maya syntactic order of placing the adverb before the verb as in lo ki’ hacen k’el ‘it is well toasted’.

Barbara Pfeiler

[ 219 ]

Certain semantic borrowings have been documented among Spanish monolinguals as well as bilinguals. Sometimes this is in the form of semantic extension of existing Spanish lexemes. For instance, in Yucatán Spanish the Spanish verb buscar ‘look for, search’ is used both in its traditional meaning and as a synonym of encontrar ‘find’, corresponding to the Maya verb kaxan, which covers both concepts. Just the opposite development can be observed in some other items, whose meanings become more specific, as in the case of the Spanish term herman-ito ‘little brother’ used to differentiate the youngest sibling from all others, in contradistinction with the Maya borrowing (x)t’uup used for the last child, an analogue of íitsin ‘youngest brother’ (Lope Blanch 1971, 5; 1987b, 88; Luxa 1990, 76).

3.3 summary According to this analysis, lexical, phonological, and morphosyntactic transfers from Maya to Spanish by bilinguals with differing proficiency in Spanish differ in content and form from the linguistic borrowings of Spanish monolinguals. The Yucatán Spanish of monolinguals is characterized by a strong Maya influence in its vocabulary and phonology. Maya verbal or nominal roots are used as verbs in Spanish with the light verb hacer ‘to do’, or in some cases with the Spanish infinitive -ar. Bilinguals, in contrast, transfer nouns and verbs from Maya into Spanish, primarily as an adaptive response to their lack of Spanish proficiency. Particularly relevant aspects include phonological transfers and adaptation of Maya nouns and verbs to Spanish verbal grammar in which the -ea conjugation dominates. When speaking Maya, bilinguals integrate Spanish nouns that have no corresponding Maya terms, but they also substitute common Maya nouns and verbs for Spanish equivalents. In several contexts, nouns are borrowed from Spanish in their plural form as stems that combine with a Maya plural suffix. Spanish diminutive suffixes are used with certain adjectives and nouns. Of special interest is the borrowing of Spanish infinitive-like stems for use in Maya as verbs or nouns. Spanish verbs are integrated into the Maya verb grammar, differentiating between intransitive and transitive verbs. The transitive verb class is marked by the Maya applicative -t, while the bare infinitive-like stem can be used in noun and intransitive verb contexts. The question arises whether these infinitive forms are incorporated as indeterminate roots or as noun roots. In Maya, a transitive verb formed with this applicative indicates it is a derivative, whether from an intransitive verb or a noun. The fact that infinitive-like forms are also used as nouns confirms Maya’s high categorial flexibility.

[ 220 ]

chapter eight

4. Conclusions The current sociolinguistic situation in the Yucatán state is clearly linked to globalization processes that have been in place since the Spanish colonial era. This centuries-long process is manifest in linguistic mutual transfers between the two languages in contact. The types of features transferred by Maya or Spanish bilingual speakers in rural areas vary according to the language they are speaking, and the rate of transfers depends on their command of the language. Previous studies of Maya and Spanish in Yucatán have documented a language usage continuum ranging from the displacement of Maya by Spanish to a reaffirmation of Maya language and culture. The latter case is associated with the maintenance and even promotion of Maya. The displacement of Maya is most acute in urban areas, in contrast to rural areas, highlighting the socioeconomic morass between these two spheres in the state. Traditional rural areas are also experiencing this contraction in the use of Maya, although it is less pronounced and slower. As the transport and communications infrastructure expands, these rural communities come into more intense and sustained contact with the urban sphere, dominated by Spanish-speakers, in stark contrast with the extremely limited contact they had just thirty years ago. Since the creation of the Instituto para el Desarollo de la Cultura Maya del Estado de Yucatán (INDEMAYA, or Institute for development of Mayan culture in the Yucatán state) in 2001, a counterforce has emerged, aimed at helping indigenous Maya speakers reaffirm their culture and language. In addition to passing the Ley General de los Derechos Lingüísticos de los Pueblos Indígenas (Federal law for the linguistic rights of indigenous populations) in 2003, the INDEMAYA has motivated increasing numbers of Maya speakers to develop linguistic strategies that meet the goals of the institute and the law, whose overall objective (as stated in the law) is “the recognition and protection of the linguistic, individual and collective rights of indigenous communities, and the promotion and development of indigenous languages (DOF, 13 March, 2003).” Hypothetically, the governmental support and the language rights law could provide the conditions for an eventual resurgence of Maya into normal, nonextraordinary, habitual usage in all societal domains. Despite these positive initiatives, societal pressures still drive Maya-speaking parents to teach their children Spanish as a mother tongue, even though the parents speak Maya among themselves. To a certain extent, this is to be expected since, for example, one of the basic requirements for qualified employment or higher education is fluency in Spanish; no such requirement exists for Maya.

Barbara Pfeiler

[ 221 ]

Currently, it appears that social dynamics are more powerful than governmental initiatives when it comes to determining the extent and frequency of language use in Yucatán state. Use of both languages may be required by law, but as long as Maya lacks the prestige that Spanish enjoys and bilinguals are discouraged from using colloquial Maya at work, no linguistic political strategy will be successful for Yucatán’s Maya community or for Yucatecan society in general. Notes 1. The recent classification of the indigenous languages of Mexico published by the Instituto Nacional de Lenguas Indígenas (National institute of indigenous languages) lists the following forms for Maya: maaya [ma:ja]; maaya t’aan [ma:ja t?a:n], maayáa [ma:ja:] (INALI 2007). 2. “To transcribe a past and establish a present, that is the essence of the colonial literature contained in the book of the Chilam Balam.” Original: “transcribir un pasado y asentar un presente, tal es la esencia de la literatura colonial contenida en los libros de Chilam Balam” (Ligorred 1986, 250). 3. Bilingual education has been implemented in Yucatán since 1950. Instead of attaining its intended aim of creating harmonious, egalitarian bilingualism and multiculturalism, the practice has inscribed a new set of norms and hierarchies. 4. The division into subregions is based on ecology as suggested by biotic variation. Biotic regions that are difficult to associate with specific economic activities have been given a geographically descriptive designation (García de Fuentes and Córdoba 2010). 5. This is due to emigration to Mérida, to Quintana Roo state (mostly to Cancún), and to the United States. 6. The example of ‘sit down’ stands out in descriptions of this variety: kulen (modern form) versus xéektabah (old form). 7. Linguistic transfer is a common phenomenon in language contact and one that can lead to the emergence of new language varieties. This process involves nonstructural mechanisms such as adjustment to a new culture and social integration, which imply social or individual bilingualism and the modification of attitudes and identity construction. 8. Field data were collected between 1980 and 2007 in Cantamayec and Yalcobá by the author and in Tekantó by Cruz Bojórquez (2003). 9. Bijiga is an adaptation of vejiga, which referred to a balloon in older urban Spanish. 10. Only one example was found in our data in which a Spanish noun is used with the Maya numeral classifier: ka p’éel sesión ‘two-inanimate meetings’. 11. Briceño Chel (2001, 374) recorded examples such as: wak-túul máak-o’ob ‘six persons’ in which the Maya plural nominal is used in a numeral and classifier construction, a novel use context for this element. For Maya Mopán, see Verbeeck (1998). 12. The diminutive form in Maya is the lexeme chan. 13. “Intransitive verbs are inflected by cross-reference marking for their S-arguments, transitive verbs are inflected by cross-reference marking for their

[ 222 ]

chapter eight

A-arguments and their O-arguments. Intransitive verbs are further distinguished on the basis of other inflectional properties into active, inactive, inchoative and positional intransitives” (Bohnemeyer 1998, 155). 14. The structural similarities between verb and noun inflections have been thoroughly discussed by Robertson (1992), who questions the lexical status of the putative verbs and suggests they could be a type of noun. 15. Example: kol (root transitive) used as a transitive verb, táan in kol-ik ‘I’m clearing the field’; as an antipassive stem (= active intransitive verb), táan in kool ‘I’m clearing (the field)’; and as a noun: (in) kool ‘(my) cornfield’. Bricker et al. (1998, 270) claim that táan is a noun meaning ‘front, face, presence’. This lexeme is identical to that used as a progressive or durative aspect auxiliary, suggesting that the literal translation of táan in kool could possibly be ‘my milpa is present’. 16. As recently as the 1970s, Yucatán children of prosperous households were reared by Maya women and surrounded by Maya servants. 17. Traditionally this lexicon of hybrid lexemes is referred to as “Yucatecisms” or Mayan voices. 18. Fluent bilinguals and Spanish monolinguals prefer the a-conjugation class with Maya borrowings, such as wiix-ar ‘urinate’.

References Alvar, Manuel. 1969. Nuevas notas sobre el español de Yucatán. Iberoromania 1:159–189. Amaro Gamboa, Jesús. 1984. Hibridismo en el habla de Yucatán. Revista de la Universidad de Yucatán 26, no. 149: 128–150. ———. 1987. Vocabulario de el uayeísmo en la cultura de Yucatán. Mérida, Mexico: Universidad Autónoma de Yucatán. Armstrong-Fumero, Fernando. 2009. Old jokes and new multiculturalisms: Continuity and change in vernacular discourse on the Yucatec Maya language. American Anthropologist 111:360–372. Baños, Othon. 2000. La península de Yucatán en la ruta de la modernidad (1970– 1995). Revista Mexicana del Caribe 5:164–190. Barrera Vásquez, Alfredo. 1980. Estudios lingüísticos. Mérida, Mexico: Fondo Editorial de Yucatán. Bohnemeyer, Jürgen. 1998. Time relations in discourse: Evidence from a comparative approach to Yukatek Maya. PhD diss., Tilburg University. Briceño Chel, Fidencio. 2001. Lengua e identidad entre los Mayas de la península de Yucatán. Los investigadores de la Cultura Maya 10:370–379. Bricker, Victoria, Eleuterio Po’ot Yah, and Ofelia Azul de Po’ot. 1998. A dictionary of the Mayan language as spoken in Hocabá, Yucatán. Salt Lake City: University of Utah Press. Brody, Jill. 1987. Particles borrowed from Spanish as discourse markers in Mayan languages. Anthropological Linguistics 29:507–521. Colazo Simon, Antonia. 2007. Les phénomènes glottaux en situation de contact linguistique: Espagnol-Maya du Yucatán, Mexique. PhD diss., Sorbonne Nouvelle– Paris III. Cruz Bojórquez, María M. 2003. Interferencias lingüísticas del maya y del español entre los bilingües de Tekantó, Yucatán. Undergraduate thesis, Universidad Autónoma de Yucatán.

Barbara Pfeiler

[ 223 ]

Diario Oficial de la Federación (DOF). 13 March 2003. Ley general de los derechos lingüísticos de los pueblos indígenas. Farriss, Nancy M. 1984. Maya society under colonial rule. Princeton, NJ: Princeton University Press. Gabbert, Wolfgang. 2004. Becoming Maya: Ethnicity and social inequality in Yucatán since 1500. Tucson: University of Arizona Press. García de Fuentes, Ana, and Juan Córdoba. 2010. Regionalización socioproductiva y biodiversidad. In Biodiversidad y desarollo humano en Yucatán, comp. Rafael Durán and Martha E. Méndez, 63–70. Mérida, Mexico: CICY, PPD-FMAM, CONABIO, SEDUMA. García Fajardo, Josefina. 1984. Fonética del español de Valladolid, Yucatán. Mexico: Universidad Nacional Autónoma de México. Gaskins, Suzanne. 2003. From corn to cash: Change and continuity within Maya families. Ethos 31:248–273. Hanks, William, F. 2010. Converting words: Maya in the age of the cross. Berkeley: University of California Press. Hill, Jane, H., and Kenneth C. Hill. 1986. Speaking Mexicano: Dynamics of syncretic language in central Mexico. Tucson: University of Arizona Press. Instituto Nacional de Estadística, Geografia e Informática (INEGI). 2006. II conteo de población y vivienda 2005. Aguascalientes. Instituto Nacional de Lenguas Indígenas (INALI). 2007. Catálogo de las lenguas indígenas nacionales: Variantes lingüísticas de México con sus autodenominaciones y referencias geoestadísticas. Available at www.inali.gob.mx/catalogo2007. Langer-Blaha, Barbara (Pfeiler). 1986. Soziophonologische Interferenzphänomene des yukatekischen Maya in der spanischen Sprache Yukatans, Mexiko. Neue Romania 4:11–35. Lentz, Mark. 2009. Los intérpretes generales de Yucatán: Hombres entre dos mundos. Estudios de Cultura Maya 33:135–158. Ligorred, Francesc. 1986. Los primeros contactos lingüísticos de los españoles en Yucatán. In Los Mayas de los tiempos tardíos, ed. Miguel Rivera and Andrés Ciudad, 241–253. Madrid: Sociedad Española de Estudios Mayas, Instituto de Cooperación Iberoamericana. Lipski, John M. 2004. El español de América en contacto con otras lenguas. Available at www.personal.psu.edu/jml34/contacts.pdf. Lois, Ximena, and Valentina Vapnarsky, eds. 2006. Lexical categories and root classes in Amerindian languages. Frankfurt: Peter Lang Verlag. Lope Blanch, Juan M. 1971. El léxico de la zona Maya en el marco de la dialectología mexicana. Mexico: Universidad Autónoma de México. ———. 1987a. Estudios sobre el español de Yucatán. Mexico: Universidad Nacional Autónoma de México. ———. 1987b. Fisonomía del español de América: Unidad y diversidad. In Actas del I Congreso internacional sobre el español de América, ed. Humberto López Morales and Maria Vaquero, 59–78. San Juan, Puerto Rico: Academia Puertorriqueña de la Lengua Española. Lucy, John. 1992. Grammatical categories and cognition: A case study of the linguistic relativity hypothesis. Cambridge: Cambridge University Press. Luxa, Gabriele. 1990. Kultur- und Sprachkontakt in Mexiko am Beispiel des Spanischen in Yucatán: Konflikt und Harmonie in einer aussergewöhnlichen Sprachensituation. PhD diss., Goethe-Universität–Frankfurt am Main.

[ 224 ]

chapter eight

Michnowicz, James C. 2006. Linguistic and social variables in Yucatan Spanish. PhD diss., Pennsylvania State University. Mossbrucker, Harald. 1992. “Etnia,” “cultura” y el impacto de la migración entre los Mayas de Yucatán. América Indígena 14:187–214. Nykl, Alois R. 1930. Notes on the Spanish of Yucatan, Vera Cruz, and Tlaxcala. Modern Philology 27:451–460. Pfeiler, Barbara. 1998. El xe’ek’ y la hach Maya: Cambio y futuro del Maya ante la modernidad cultural en Yucatán. In Convergencia e individualidad: Las lenguas Mayas entre hispanización e indigenismo, ed. Andreas Koechert and Thomas Stolz, 125–140. Colección Americana 7. Hannover: Verlag für Ethnologie. ———. 1999. Situación sociolingüística. In Atlas de procesos territoriales de Yucatán, ed. Ana García de Fuentes, Juán Córdoba, and Pablo Chico, 269–299. Mexico: Universidad de Yucatán/ PROESA, Proyección Cartográfica. ———. 2012. Voces Mayas: Ethnografische und soziolinguistische Aufzeichnungen zur Zweisprachigkeit in Yukatan. Graz: ACPUB Academic Publishers. Pfeiler, Barbara, and Lenka Zámišová. 2006. Bilingual education: Strategy for language maintenance or shift of Yucatec Maya? In Mexican indigenous languages at the dawn of the twenty-first century, ed. Margarita Hidalgo, 281–301. Berlin: Mouton de Gruyter. Robertson, John, S. 1992. The history of tense/aspect/mood/voice in the Mayan verbal complex. Austin: University of Texas Press. Stolz, Christel. 1998. Hispanicisation in modern Yucatec Maya: Grammatical borrowing. In Convergencia e individualidad: Las lenguas mayas entre hispanización e indigenismo, ed. Andreas Koechert and Thomas Stolz, 165–194. Hannover: Verlag für Ethnologie. Verbeeck, Lieve. 1998. Borrado de la memoria, grabado en la lengua: Interferencias españolas e inglesas en el Mopán de Belice. In Convergencia e individualidad: Las lenguas mayas entre hispanización e indigenismo, ed. Andreas Koechert and Thomas Stolz, 141–163. Hannover: Verlag für Ethnologie. Wichmann, Søren. 2004. Structural patterns of verb borrowing. Paper presented at the workshop on loan word typology, May 1–2, 2004, Leipzig.

9 * Standard Colonial Quechua alan durston

1. Introduction There is a common perception that Spanish colonials simply exploited existing situations in their use of certain Amerindian languages as lingua francas or lenguas generales (e.g., Ostler 2005, 367–368). The vast geographical spread of Quechua throughout much of the former Inca Empire would be the most obvious case in point (Quechua was known in much of Spanish South America simply as la lengua general). However, a different picture has been emerging for some time from the research on colonial Quechua. This picture is more in tune with what scholars of nineteenthand twentieth-century European colonialism in Africa and Asia have been suggesting, that colonial powers crafted their own vehicular varieties through complex processes of selection, codification, and standardization rather than working with what was “already there” (e.g., Fabian 1986; Cohn 1996, 16–56). Even colonial understandings of what was “already there” are seen to have been determined to a large degree by European preconceptions concerning how language varieties relate to each other and to social groups (Irvine and Gal 2000). Certainly, any statement such as “The Spanish continued the use of Quechua as a lingua franca in the territories of the former Inca Empire” would be acceptable only with a long list of caveats and qualifiers, beginning with issues of dialectology (e.g., “What kind of Quechua?”). This chapter examines how Spanish colonialism used and reinvented Quechua during the first half of the colonial period in what is now Peru.1 My main concern here is the development of a written standard of missionary origin, which I will refer to as Standard Colonial Quechua (SCQ). I will also attempt to define more precisely the lingua franca roles that SCQ had under Spanish colonial rule.

2. Background The evidence for understanding both the situation of Quechua on the eve of the conquest and its colonial transformation is limited and dif-

[ 226 ]

chapter nine

ficult to interpret. The colonial-period literature in and about Quechua is much smaller and less diverse than the equivalent sources for Nahuatl, the only other colonial lengua general that came close to Quechua in importance. We have a handful of grammars and dictionaries, a missionary or pastoral literature, and a patchy assortment of nonpastoral texts ranging from plays to personal correspondence.2 The overwhelming bulk of the literature is of clerical origin—the extensive indigenous-language notarial literature that survives from dozens of Mesoamerican localities probably never existed in the Andes (Mannheim 1991, 143–144; Durston 2008). With very few exceptions, the literature employs the Quechua of the southern Peruvian highlands; the varieties of central and northern Peru, Ecuador, and Bolivia are barely represented, if at all. In the absence of a large and diverse vernacular literature, the colonial dialectology of Quechua is very difficult to reconstruct. To add to the difficulties, Spanish descriptions of linguistic situations as well as policy statements are often obscure. The glottonyms used tend to be shifters, for example, la lengua general or la lengua materna; so it may be unclear what particular language variety or even family they refer to (Mannheim 1991, 34; Durston 2007b, 46).3 As a result, there is some disagreement concerning how the Quechua literature related to spoken varieties. Over the past fifty years, the linguistic history of the Central Andes has been completely revised by interpreting the colonial written evidence in the light of the contemporary dialectology of Quechua. Since the dialectological work of Gary Parker and Alfredo Torero in the 1960s, Quechua has been recognized by most linguists as a family with two main branches, branch “I” (to use Torero’s nomenclature) being spoken in the central Peruvian highlands and branch “II” in the north and the south, extending into Ecuador and Bolivia.4 For the purposes of this chapter, I will distinguish between “Central Quechua” (corresponding to Torero’s branch I) and “Southern Peruvian Quechua,” with reference to the varieties of the southern Peruvian highlands. There is very limited intelligibility between these two groups. While Southern Peruvian Quechua can be considered a single language, there are also marked differences between its varieties, especially at phonological and lexical levels (Mannheim 1991, 4–16). I will occasionally make passing references to the Bolivian varieties, which are quite close to Southern Peruvian Quechua (together they make up Torero’s subbranch IIC), though separated from it by a large Aymara-speaking area. A first consequence of noting the profound differences between Quechua “dialects” is that they must have spread widely and diversified centuries before the Inca expansion, probably from origin points in the central Peruvian highlands and/or coast (Torero 1964, 475; 1975, 243;

Alan Durston

[ 227 ]

Cerrón-Palomino 1987, 323–336). The Inca Empire certainly contributed to the further spread of Quechua through resettlement policies (Mannheim 1991, 45–47), but distinguishing between Inca-introduced and preexisting varieties is problematic.5 Under the Inca Empire (and previously), varieties of Quechua, Aymara, and the extinct Puquina family functioned as lingua francas in different regions (cf. Mannheim 1991, 31–52). Whether the Incas employed an empirewide lingua franca, and what that lingua franca may have been, remain open questions. In this regard it is important to note two points. One is that the Incas did not necessarily regard Quechua as their “mother tongue” (if they had such a concept), since Aymara and Puquina were also spoken in Cuzco and probably predated Quechua (Cerrón-Palomino 1999). The other is that the empire’s main record-keeping device, the knotted-string record or quipu (khipu), seems to have been nonlinguistic. Nonetheless, Alfredo Torero, Rodolfo Cerrón-Palomino, and others have pointed to evidence that the Incas favored the use of a now-extinct form of Quechua that may have come from the central Peruvian coast rather than the Quechua of Cuzco for imperial administration (Torero 1974, 96, 132–133; Cerrón-Palomino 1987, 327–328; 1989, 1995; Taylor 2000, 36–38).

3. The Colonial Quechua Literature The grammar and dictionary published in 1560 by the Dominican friar Domingo de Santo Tomás, which contain the oldest known Quechua texts (a prayer and a short sermon), arguably record the Quechua spoken in the vicinity of Lima (Santo Tomás [1560] 1951, [1560] 1995; CerrónPalomino 1995). Santo Tomás commented on the great variability in the way Quechua was spoken throughout the former Inca Empire, but he regarded this variability as the fate of all languages and did not attempt to distinguish correct varieties, making no particular claims for the one he recorded (Santo Tomás [1560] 1995, 9, 18). This variety was relatively close to Southern Peruvian Quechua while also sharing some important features with Central Quechua. Its relation to the Inca lingua franca proposed by Torero and others remains unclear, because it shared some of its distinctive phonological characteristics while lacking others— the differences may reflect coastal Quechua dialectology, of which practically nothing is known, since Quechua disappeared from Peru’s coastal areas during the colonial period (Cerrón Palomino 1995; Durston 2007b, 189–190). The next-oldest set of Christian texts employ a different form of Quechua and are associated with a new approach to the issue of linguistic diversity. In 1582 the Third Lima Council commissioned a team of

[ 228 ]

chapter nine

translators (all Spanish or mestizo priests) to translate from Spanish into Quechua and Aymara an extensive collection of catechisms, sermons, and sacramental texts. The resulting trilingual corpus was printed as a three-volume set in Lima in 1584 and 1585 (Third Lima Council [1584, 1585] 1985). The importance of these books for the history of Christian discourse in Quechua and of the Quechua literature in general can hardly be exaggerated. The Quechua and Aymara translations of the most important texts were made obligatory for Christian instruction in these languages, and even the smallest deviations were heavily penalized. Further translations of these texts into other varieties of Quechua and Aymara were thus banned. The Quechua of the Third Lima Council corpus shows clear signs of a process of codification and uniformization at the grammatical, lexical, and orthographic levels. To aid priests in the use of these texts, a single-volume grammar and dictionary was printed in 1586 and went through several reprints and editions, completely superseding Santo Tomás’s works, which were all but forgotten (Grammatica y vocabolario [1586] 1603). Linguistic codification went hand-in-hand with the creation of a fixed Christian terminology for Quechua that relied heavily on Spanish loanwords, although some Quechua neologisms were used. The uniformity that was required of Christian belief and practice in the Counter-Reformation context, especially among neophytes, was to be achieved first at the level of language (Mannheim 1991, 66–67; Cerrón-Palomino 1992, 213–219; 1997; Durston 2007b, 86–104, 191– 197, 214–217, and passim). The Third Lima Council’s translation team set the standard for Quechua writing for much of the next century (cf. Cerrón-Palomino 1992, 217– 218). Their Quechua was the model, a kind of Platonic ideal, that defined SCQ, which in practice, of course, was not all that neat and homogenous (see below). But how did this Quechua relate to existing, spoken forms of Quechua? The translators declared that while they considered the Quechua of the former Inca capital, Cuzco, to be the most “perfect,” they had aimed for a “middle road” that would allow for intelligibility on the coast, in what are today Central Quechua speaking areas, and up north to Quito (Third Lima Council [1584, 1585] 1985, 167). This suggests an effort to produce a koiné, but in fact the conciliar corpus contains no distinctively Central or coastal forms. SCQ was a standardized and watered-down form of the variety of Southern Peruvian Quechua spoken in and around sixteenth-century Cuzco (Cerrón-Palomino 1992, 213–219). Lexical items that were unique to the area were eschewed in favor of pan-Quechua synonyms, and most idioms and tropes were avoided (Itier 2001, 71; Durston 2007b, 191–194). The highly influential Quechua orthography established in the Third Lima Council works also

Alan Durston

[ 229 ]

aimed at simplicity: it was strongly hypo-differentiating, failing to represent a number of important oppositions (see Mannheim 1991, chap. 6, on questions of colonial Quechua phonology and orthography; also Durston 2007b, 194–197). It appears that more than enhancing intelligibility among indigenous audiences that spoke other varieties of Quechua, SCQ was intended to be easily learned by Spanish priests. Cerrón-Palomino thus seems justified in referring to what he calls Quechua general as a “construct,” the result of a conscious effort of linguistic planning and normalización (1992, 217). At the same time it must be emphasized that SCQ was not a pidgin—it retained the full range of grammatical categories present in its base variety. Arguments that SCQ reflects Spanish influence at the grammatical level, or even a failure to understand Quechua grammar, seem unfounded (Durston 2007b, 181–185). Why the change in base varieties from Domingo de Santo Tomás’s time to the Third Lima Council? Torero points to two separate processes: (a) the demographic collapse of the indigenous population of the central coast, and (b) the rise of the silver mining economy (Torero 1974, 188; 1995, 14–15). This last point requires some explanation. During the decade or so prior to the Third Lima Council, viceroy Francisco de Toledo created the conditions for a dramatic rise in silver production at Potosí, in modern Bolivia. Toledo introduced the mercury amalgamation refining process, using mercury from Huancavelica (in southern Peru), and set up a vast labor draft or mita system to feed the mines of both Potosí and Huancavelica (cf. Bakewell 1984). The area where Potosí and the important colonial administrative center of Chuquisaca (Sucre) are located has a large Quechua-speaking population today, and the local varieties are very close to those of southern Peru. Huancavelica is located on the northern edge of the modern Southern Peruvian Quechua area. Torero’s argument, which was never developed in any detail, is that the growing economic importance of areas where varieties of the “IIC” subbranch were spoken resulted in a specialization of the colonial vernacular project in these varieties, although this applies more clearly to Bolivian Quechua than it does to Southern Peruvian Quechua. More recently, César Itier has argued that forces associated with the mining boom were actually transforming the linguistic landscape of the southern Peruvian highlands (Itier 2000, 2001). Processes of commercial integration, labor mobilization, and urban growth led to the convergence and spread of Southern Peruvian Quechua varieties, which functioned as a lingua franca throughout much of southern and central Peru. This lingua franca developed among the indigenous population of the colonial cities as well as among temporary urban residents from different regions, in particular mita draftees. According to this argument, the Quechua of

[ 230 ]

chapter nine

the Third Lima Council was a written codification of the rising Southern Peruvian Quechua lingua franca, and thus not as much of a construct as Cerrón-Palomino suggests (Itier 2000, 49; 2001, 68n19). In fact, Itier sees the phenomenon I am referring to as SCQ as a set of written reflections of varieties of the lingua franca spoken in different areas of southern and central Peru (2000, 50; 2001, 69). Itier also suggests that Southern Peruvian Quechua was expanding as a vernacular, replacing other varieties of Quechua and other language families (notably Aymara) throughout a large region. In particular, he argues that the boundary between Central Quechua and Southern Peruvian Quechua, which today runs just to the north of Huancavelica, was originally located farther south and east (Itier 2001, 64–67). While some convergence and spread of Southern Peruvian Quechua does seem likely, I see a less direct relation between written Quechua and spoken varieties. For one thing, conciliar Quechua appears too early to reflect forces that were not in play on a significant scale until the 1570s. I would also argue that variation in Quechua writing during the decades after the Third Lima Council reflects varying degrees of approximation to the conciliar model, rather than different forms of an oral lingua franca. Put differently, I see the relative homogeneity of the written Quechua of the late sixteenth century and much of the seventeenth century as a product of colonial standardization rather than a reflection of convergence and lingua franca spread among the indigenous population. However, more work is needed on a difficult and patchy body of data, and the reality may lie somewhere in between the top-down model and Itier’s bottom-up one. The church’s fixation on the Quechua of Cuzco is best explained by contemporary European linguistic ideology. Sixteenth-century European languages were undergoing their own processes of codification and general ausbau in competition with Latin and with one another. The most correct and expressive varieties, those that had the greatest potential to become languages of literature and learning, were generally thought to be those spoken in the urban centers of power, wealth, and learning (Burke 2004). In the eyes of the Spanish, Cuzco was the only real “city” of the Andean world, and its Quechua was thus the most “urbane” (Durston 2007b, 112). There was a historical corollary to this argument: the centers where languages were spoken most correctly were also thought to be their points of origin. If Quechua was the language of the Incas, its purest and most correct form had to be sought in their home and capital (Cerrón-Palomino 1995, li; Durston 2007b, 112–113). Over a period of more than sixty years following the Third Lima Coun-

Alan Durston

[ 231 ]

cil, numerous new pastoral texts in Quechua were published, usually with the intent of supplementing rather than challenging the conciliar corpus. They include hymns, sermons, and texts for the administration of the sacraments (Durston 2007b, 137–171 and passim). Two main movements away from the conciliar form of SCQ are visible in a literature that by and large reproduces the norm very faithfully. A sacramental manual published in 1631 by the secular cleric Juan Pérez Bocanegra often seems to represent the irregularities and idiosyncrasies of the spoken Quechua of Cuzco (Itier 2000, 50–55; Durston 2007b, 200–201). However, Pérez Bocanegra must have been wary of the problems this might create for priests who were not from Cuzco and often supplemented lexical “Cuzcoisms” with SCQ synonyms (Pérez Bocanegra 1631, 82, 83, 88, 140, 200, 202; cf. Durston 2007b, 201, 344n24). A second, far more widespread tendency developed in the archdiocese of Lima, an overwhelmingly Central Quechua–speaking area. As is discussed in more detail in the next section, local clergymen publicly denounced the reliance on SCQ (which they regarded simply as the Quechua of Cuzco) for Christian instruction in their territory. However, their own Quechua writings diverge very little from SCQ, rarely going beyond the insertion of the occasional Central Quechua word stem (Cerrón-Palomino 1992, 223–225; Torero 1995; Durston 2002; 2007b, 127–130, 202–230).6 Dialectological issues aside, there were also efforts to reform the Third Lima Council orthography by introducing additional notations to distinguish the uvular from the velar stop and the modified (glottalized and aspirate) from the plain stop series (modified stops probably existed only in some varieties of Southern Peruvian Quechua, including that of Cuzco) (Mannheim 1991, 136–137; Durston 2007b, 199–200, 203–205). The clerical literature aside, there is also a small corpus of Quechua texts of indigenous authorship, including personal correspondence, legal petitions, and notarial documents. A dozen cases (documents or small groups of documents) of “mundane” Quechua writing dating mostly from the seventeenth century have been discovered, including six documents written on behalf of women; two are letters that could conceivably have been penned by the women who speak through them (Itier 1991, 1992; Taylor 2000; Durston 2003, 2008). It has been argued that the corpus reflects a significant development of Quechua literacy among the indigenous elite for much of the seventeenth century, although it is clear that it was not widely used for purposes of institutional record-keeping, as happened with Nahuatl and other Mesoamerican languages (Itier 1991, 1992; Durston 2008). As Itier has suggested, the Third Lima Council texts must have played a key role in the spread of “mundane” Quechua literacy (Itier

[ 232 ]

chapter nine

1992, 1–2). Indeed, the mundane corpus follows the conciliar standard quite closely, although with some divergences, especially in texts from Central Quechua–speaking areas. These divergences can be put down to dialectal interference in writers who had not mastered SCQ (Durston 2008, 56). Much the same can be said for the Huarochirí Manuscript, a book-length account of the religious traditions of the Huarochirí province written by an Indian ca. 1600 (Salomon and Urioste 1991; Taylor [n.d. (1987)] 1999). While the Huarochirí Manuscript stands in a class of its own in many respects, the Quechua in which it is written is essentially SCQ, with some Central Quechua lexical items mixed in (Taylor 2000, 40–42; Durston 2007a). Even writers from southern Peru can be seen making a conscious effort to reproduce the standard, which inevitably differed from their native varieties. A letter written in the Huamanga (Ayacucho) area in 1660 shows the scribe crossing out a nonstandard form of a common suffix to replace it with the SCQ form (Durston 2003, 217). The evolution of SCQ as a medium of written communication is not easily traced beyond the middle of the seventeenth century. The last datable mundane Quechua texts are from the 1670s. Very few new pastoral and linguistic works were published between 1650 and the nineteenth century, and most of them derived from the pre-1650 works and thus reproduced the conciliar norms. For example, the Sixth Lima Council of 1772 sponsored a reedition of the Third Council’s sermons and also produced a new Quechua catechism that closely followed Third Council norms at every level (Sixth Lima Council 1772; cf. Rivet and CréquiMontfort 1951, 170–172). There are some important exceptions, however. In 1725 Bishop Luis Francisco Romero of Quito published a pastoral letter and a catechism in the Quechua of Quito, and a grammar of the same variety appeared in 1753 (Rivet and Créqui-Montfort 1951, 141– 147, 156–158). In 1739, Juan Antonio Dávila Morales, priest of Yotala in the archdiocese of La Plata, published a catechism that includes texts in a form of Quechua similar to that spoken today in the Sucre (Chuquisaca) area (Dávila Morales 1739). All three works were printed in Lima. Eighteenth-century devotional manuscripts from the southern Peruvian highlands also show deviation from SCQ norms. A manuscript from Cuzco reflects a series of wide-ranging sound changes that affected Cuzco Quechua beginning in the late seventeenth century (Quaderno de directorio espiritual n.d.; cf. Mannheim 1991, 208–225 on the sound changes). A manuscript from Huamanga is the oldest text I know of that records the modern “Ayacucho-Chanca” variety of Southern Peruvian Quechua (Untitled Devotionary n.d.). The colonial Quechua plays are also key testimonies to the evolution

Alan Durston

[ 233 ]

of Quechua writing after the mid-seventeenth century, but they present numerous problems of analysis because in most cases we know nothing for certain about their authorship or even when they were written. We do know that all or most of them were written by priests in or near Cuzco for an elite indigenous or Spanish American audience (or both) (Mannheim 1991, 72–73; Itier 1995, 1999, 2006). Several of the manuscripts can be placed at different stages in the processes of sound change that transformed Cuzco Quechua during the eighteenth century (Mannheim 1991, 147–152). Without having carried out a systematic study of the plays, I have the impression that the famous drama Ollantay (the saga of an Inca general who rebelled against the emperor) employs a more distinctively Cuzco lexicon than the other plays, which are distinguished from Ollantay both by their explicitly Christian themes and by greater proximity to SCQ and are probably older.7 In sum, this variegated literature from the mid and late colonial periods points to the demise of SCQ as the dominant, even the only written form of Quechua. Quechua texts began to reflect spoken varieties, although some priests continued to study SCQ and new Christian texts were written in it. The best illustration of the disappearance of SCQ as a widespread written medium comes from the indigenous-language fliers of the independence wars (1810–24). Both royalist and pro-independence forces printed fliers in Quechua, and more than a dozen of them have survived. Some announced invading armies, requesting the support of the local population. In other instances, Quechua speakers are informed of and sometimes invited to join processes of political liberalization undertaken by Spanish American congresses or by the Spanish resistance in Cadiz. The earliest broadsides are from Buenos Aires and are aimed at Bolivia, which the Buenos Aires independence leaders made several attempts to conquer. José de San Martín made use of Quechua propaganda during his invasion of Peru, and even Bernardo O’Higgins, the leader of the Chilean movement, issued a Quechua flier expressing the benefits of independence. The language of these texts is very diverse. There are differences in the dialects and orthographies employed, and in fact no two documents are alike. The Argentinian broadsides are in Bolivian forms of Quechua, as might be expected. Two broadsides addressed by the Spanish general José de Canterac to pro-independence guerillas in the central Peruvian highlands are the oldest known texts in Central Quechua. Other manifestos employ a literary Cuzco Quechua that is reminiscent of the language of Ollantay (cf. Itier 1995, 102–103). The unity that characterized Quechua writing in the seventeenth century was long gone and apparently not even remembered.8

[ 234 ]

chapter nine

4. Contexts of Use Did SCQ ever become a spoken language of everyday use? Did the Spanish seek to impose it as such? Did they rely on it as an administrative medium? The answer to each of these questions seems to be no. SCQ lived in certain well-defined circuits that were primarily ecclesiastical in nature. For several decades it was a key component of the training received by the bulk of the Peruvian clergy, since applicants for Indian parishes had to demonstrate that they knew Quechua, which in practice meant SCQ. This could be done by taking a course or via an oral examination. Beyond the hundreds or thousands who were exposed to some form of SCQ instruction or examination, the language competence system generated a number of salaried posts for career experts in SCQ: two higher-education chairs in Lima, an unknown number of instructors working for the Jesuits and the mendicant orders, and a host of diocesan examiners. A Quechua chair was established at the University of San Marcos in Lima by Viceroy Toledo in the late 1570s, and at that time a chair already existed at the Lima cathedral. Originally, the university chair was given wide-ranging powers and the entire Peruvian clergy were to be taught and examined by him if they hoped to be assigned a parish, or even to be ordained. Soon, a more decentralized examination system developed that functioned at the diocesan level, but the university chair continued to lecture and approve (and presumably also flunk) candidates for parishes in the archdiocese of Lima.9 The Third Lima Council volumes and the accompanying grammar and dictionary were key instruments for both teaching and examination, and the language competence system seems to have been largely directed at teaching future parish priests how to read and use the council’s SCQ texts. Thus, even the growing number of locally born clergymen who already spoke a variety of Quechua required instruction and examination (Durston 2007b, 114–122). It has to be emphasized, however, that SCQ training was limited not just to the clergy, but to a specific sector within it, albeit a very large one—those who needed to earn a living as priests of Indian parishes. No such training was required of the clerical elite (bishops, canons, diocesan judges and administrators, etc.), very few of whom knew any Quechua. Lay officials appear to have received no Quechua training at all. Instead, both lay and ecclesiastical justice and administration relied on interpreters or lenguas, preferably local Spanish ones rather than Indians (Taylor 2000, 38–39). Some of these interpreters were permanently attached to a specific court or official and interpreted between Spanish and Quechua (39–40), but when an official encountered the need to communicate with speakers of one of the lenguas particulares, which generally happened dur-

Alan Durston

[ 235 ]

ing a tour of inspection, an interpreter would be appointed on the spot (Durston 2007b, 126). One of the main justifications for the imposition of Spanish among the indigenous population was that this would allow them to communicate directly with those who administered justice, who, unlike parish priests, were under no obligation to learn the languages of their charges (131). There is a marked contrast here with the practices of later colonial powers, such as the British in India. British civil servants were trained in the use of select Indian languages (previously codified through linguistic analysis) at specially created colleges so that they could administer justice and collect taxes directly (Cohn 1996, 16–56; Trautmann 2006). Instruction in the use of vernacular “languages of command” extended even to daily interactions with native servants, for which special phrase books were created (Cohn 1996, 39–44). The instruction received by the Spanish clergy in Peru seems to have been restricted to catechetical and sacramental uses; the grammars and dictionaries provided no guidance at all on how to handle everyday situations. Increasingly, of course, the Peruvian parish clergy were locally born and already spoke some form of Quechua, so they required no special instruction on how to order breakfast. The same could not be said for more delicate tasks, such as administering baptism or explaining the Eucharist in Quechua. As for the possible spread of SCQ as a vernacular among the indigenous population, viceroys and archbishops did call for the adoption of la lengua general or la lengua del inga in order to facilitate religious instruction.10 It seems, however, that these calls referred to Quechua in general and were aimed at speakers of other language groups such as Aymara, Puquina, and Culli. Although the meaning of terms like lengua general and lengua del inga was slippery, they were mostly general designations for Quechua. There is no evidence of a policy of making speakers of Central Quechua, for example, adopt SCQ or any form of Southern Peruvian Quechua. Few lay administrators or even clergy of the upper ranks had much of an understanding of the differences between Quechua varieties. In any case, those varieties were not regarded as distinct languages even by the Quechuists. Nor have I found evidence of real efforts or concrete plans to implement even a generic “Quechua-only” policy. Economic and labor regimes that required periodic trips to colonial centers must have contributed to the convergence and spread of varieties of Quechua, as Itier has suggested. Indeed, Mannheim has identified homogenization—the spread of both Quechua and Spanish at the expense of other language families—as the dominant linguistic process of the colonial period (Mannheim 1991, 35, 80). In 1588 Viceroy Fernando de Torres y Portugal (also known by the title of Conde del Villar)

[ 236 ]

chapter nine

noted in a report to the king that even in (unspecified) areas where the women were ignorant of la lengua general, the men did speak it because of their work in the mines and other labor service obligations.11 Around that same time, the Jesuit Blas Valera wrote that many “common Indians” who visited the cities (he mentions Lima, Cuzco, Chuquisaca, and Potosí) were learning what he called la lengua del Cuzco because of the need to converse with Indians from other areas and were taking it back to their homelands (cited in Itier 2000, 48). As is often the case, it is difficult to determine the geographic scope of these claims and whether they refer to specific varieties or to any form of Quechua.12 Colonial systems of exploitation probably encouraged the use of a number of varieties of Quechua as lingua francas and, more gradually, their spread as vernaculars within different areas of varying size, the largest of which would have been the Southern Peruvian Quechua area.13 Linguistic homogenization must also have been encouraged by migratory phenomena that the colonial regime did not control and made every effort to prevent. Beginning in the late sixteenth century, Indians were abandoning their communities and jurisdictions of origin en masse in order to escape state-managed forms of exploitation. Some went to the cities, but most established themselves as forasteros (outsiders) in neighboring areas where they would have the advantage of anonymity (Wightman 1990). Clearly, homogenization was a product of these anonymous economic and demographic forces more than of colonial linguistic policies. The pastoral uses of SCQ seem to have resulted not in language shift but in situations of diglossia, especially in the highlands of the archdiocese of Lima, which correspond very closely to the modern Central Quechua–speaking area. It appears that in some respects SCQ stood in relation to Central Quechua varieties as Latin did to Romance languages in medieval and early modern Europe. For the majority of the population, SCQ would have been a liturgical language in which they memorized prayers and catechisms and heard sermons and in which they developed a limited and passive competence. Members of the elite would often have developed a more active competence, especially the assistants of the parish priest who carried out much of the religious instruction and had access to the pastoral volumes containing SCQ texts. SCQ also provided the most accessible form of alphabetic literacy, which Indians used primarily for personal correspondence and, less frequently, petitioning Spanish authorities (see above). It was always abundantly clear to the clergy that their reliance on SCQ created numerous problems for Christian conversion even in Quechuaspeaking areas. These problems were particularly evident in the archdiocese of Lima, and debates about the viability of SCQ raged among the

Alan Durston

[ 237 ]

Lima Quechuists throughout the first half of the seventeenth century, and probably earlier (Torero 1995; Durston 2007b, 126–130). In the late 1640s, Fernando de Avendaño, a canon at the Lima cathedral, accused adherents to SCQ of deliberately using a language in which they knew they would not be understood, for reasons that he did not disclose (Avendaño n.d. [1649]). Around the same time, Diego de Molina, a Franciscan based in Huánuco, argued that many of his parishioners had developed serious misconceptions of basic Christian texts and doctrines because they did not understand the términos cuzquenses into which they had been translated (Romero 1928, 85). Oddly, the Quechua sermons that Avendaño and Molina themselves wrote are some of the best examples of SCQ, aside from the insertion of a small number of Central Quechua word stems (Avendaño n.d. [1649]; Molina n.d. [1649]; cf. Durston 2007b, 202–203). The same issues came up in the arena of clerical language training. Avendaño’s sharp criticisms of the proponents of SCQ were implicitly directed against Juan Rojo Mejía y Ocón, a secular priest from Cuzco who at the time held the university Quechua chair and had attacked the “depraved opinion” that SCQ was not fully intelligible in the archdiocese of Lima (Rojo Mejía y Ocón 1648; cf. Torero 1995, 18). From 1614 to the late 1630s, however, the university chair had been held by Alonso de Huerta, a secular priest from Huánuco (a Central Quechua–speaking area) who was on the pro–Central Quechua side of the debate. Huerta approved applicants for parishes on the basis of their knowledge of Central Quechua and published a new grammar that allegedly provided instruction in it. Nonetheless, Huerta’s grammar is essentially an SCQ manual with only minimal and seemingly random coverage of Central Quechua forms (Huerta [1616] 1993; cf. Durston 2007b, 117–118, 126–127). One of his language-competence certificates mentions the use of the Third Lima Council texts to examine an applicant, while another states that an applicant who was already fluent in Central Quechua was “perfecting” his knowledge by following Huerta’s lectures at the university, “so as to be able to administer the sacraments.”14 SCQ was openly criticized in the archdiocese of Lima, but there appear to have been no viable alternatives to it from the perspective of the clergy. Given that the expression of Christian doctrine in an indigenous language was such a delicate matter, the use of Central Quechua would have required a new process of formal codification. SCQ, and the standard Christian terminology that went with it, had the seal of approval of a provincial council. Anyone venturing to write Christian texts in a different form of Quechua would have had to develop a new set of translation conventions that would inevitably come under attack in the absence of

[ 238 ]

chapter nine

comparable institutional backing. SCQ served key functions in the reproduction of religious orthodoxy and of colonial relations in general. Restricting Christian discourse in Quechua to a single, highly codified variety greatly enhanced the ability of the church to control it, a key issue in a Counter-Reformation colony. The systematic training that the clergy received in SCQ, added to the fact that its terms for Christian categories and institutions were mostly Spanish loanwords, meant that parish priests had a home advantage even if they were not native Quechua speakers. In other words, SCQ served to shore up clerical authority. When a secular priest in Huánuco was accused by his Indian parishioners of not knowing Quechua, as parish priests often were, he replied that he knew it better than they did and had a certificate from the university chair in Lima to prove it.15 How, then, did SCQ lose its monopoly over Christian discourse and written communication in Quechua? As mentioned above, the publication of new pastoral and linguistic works all but stopped around 1650, and it may not be a coincidence that the debates about the use of SCQ in the archdiocese of Lima culminated in the late 1640s. The clerical language competence system, certainly the most important conduit for the cultivation and transmission of SCQ, was also stalling in the second half of the seventeenth century, and it seems not to have functioned at all throughout the eighteenth century, in spite of the occasional calls that it be resurrected (Durston 2007b, 172, 177). The underlying forces behind these developments remain unclear. The intensification of calls for the imposition of Spanish among the indigenous population probably had a role, but it could also be argued that the conversion impulse itself was waning, or at least taking on less centralized and institutionally controlled forms (Estenssoro Fuchs 2003, 243–244; Durston 2007b, 171–176). As SCQ ceased to be widely cultivated by the clergy, its use as a medium of mundane written communication among the indigenous elite also seems to have waned, perhaps because they were no longer exposed to SCQ pastoral imprints. The written traditions that developed in the eighteenth century represent local, vernacular Quechuas. Barring the discovery of new eighteenth-century Quechua genres, these traditions seem to have been limited to devotional and, in Cuzco at least, “literary” uses, specifically drama.

5. Conclusions The term lingua franca evokes a medium that enables exchange between language communities. SCQ does not fit this definition very well; in many respects it was designed as an instrument of control and restriction rather

Alan Durston

[ 239 ]

than of wide communication. It could be considered the lingua franca of Christian conversion in Peru, but conversion was to be a unidirectional process—the church was certainly not hoping to enter into a debate with indigenous neophytes that would be facilitated by SCQ. Indigenous parishioners were only to “speak” SCQ in the form of oral performances of memorized texts. Of course, processes of religious appropriation and hybridization were always active, but for all we know these processes involved the rejection of a linguistic medium that was designed to prevent them. A perhaps unintended offshoot of the promotion of SCQ in pastoral contexts was that it became a general medium of written expression and communication for a period of about a century. However, written SCQ did not acquire significant administrative or legal roles, nor was there a policy that Spanish judges and officials learn SCQ or any other form of Quechua so as to communicate directly with their charges. Varieties of Quechua certainly did function as lingua francas for communication among Indians from different areas who converged in various colonial centers as labor draftees or migrants, as Itier has suggested. However, I see little evidence that there was such as thing as a pan-Andean or even pan-Peruvian Quechua lingua franca. Instead, there would have been regionally prominent varieties, or sets of closely related varieties. Spanish was rapidly becoming the lingua franca of the Andes. I would argue that the degree to which the colonial regime effectively promoted Quechua beyond strictly pastoral contexts has been overstated. Spanish rule relied heavily on indigenous elites who often had some competence in Spanish, and there was a growing and increasingly ubiquitous population of bilingual Spanish Americans and mestizos who could be relied on to facilitate communication between outsiders and locals. In contrast to unstable and mobile hinterland areas like the interior of Brazil (see Couto, chap. 3. this volume), a focal area of early modern Spanish colonialism like Peru was not fertile terrain for the development of a fullfledged indigenous lingua franca.16 Notes 1. Only a tiny fraction of the colonial-period Quechua literature comes from modern Ecuador and Bolivia, and very little is known about the linguistic history of these areas. 2. Mannheim (1991) surveys much of the known literature (see especially his chap. 6). On the pastoral literature, see Durston (2007a). On Quechua mundane writings, see Itier (1991, 1992), Taylor (2000), and Durston (2003, 2008). On the colonial Quechua plays, see Itier (1995, 1999, 2006). 3. The term lengua general in particular has caused problems. In much of the recent literature on colonial Quechua, it has been used (and read in the primary

[ 240 ]

chapter nine

sources) as a term for a particular lingua franca variety of Quechua, when it functioned most often as a generic label for Quechua. 4. The early, groundbreaking pieces on Quechua dialectology were by Parker (1963) and Torero (1964). Further key works that explore general Quechua dialectology and historical linguistics include those by Torero (1974, 2002), Cerrón-Palomino (1987, 1992), Mannheim (1991), and Adelaar and Muysken (2004). 5. For a summary of debates concerning the origins of Ecuadorian Quechua, see Cerrón-Palomino (1987, 343–345). 6. For an interpretation of the variation in the post–Third Lima Council pastoral literature as a reflection of variation in a lingua franca spoken by the indigenous population, see Itier (2001, 69, 71). 7. According to an argument put forward by Itier, Ollantay was written shortly after the execution of Tupac Amaru II in 1781 (Itier 2006). El robo de Proserpina and El hijo pródigo are both attributed to Juan Espinoza Medrano and would thus have been written in the mid-to-late seventeenth century (Itier 1999). 8. Very little research has been done on these broadsides; they are briefly discussed by Itier (1995, 102–104). Several are reproduced by Paul Rivet and Georges de Créqui-Montfort in their Quechua-Aymara bibliography (1951, 231–288 and passim, 497–499). 9. By the early seventeenth century, there were five dioceses in what is now Peru. From north to south, they were Trujillo, Lima (the metropolitan seat), Huamanga (modern Ayacucho), Cuzco, and Arequipa. 10. Viceroy Francisco de Toledo’s (1575) decrees for the newly created pueblos de indios ordered that the lengua del inga be spoken by all on the grounds that it would facilitate catechesis (Ordenanzas generales para la vida común en los pueblos de indios, 1575, in Lohmann Villena and Sarabia Viejo 1986–87, 2:251). In 1588 Viceroy Fernando de Torres y Portugal backed this policy in a letter to the king, although he used the term lengua general (Conde del Villar, April 25, 1588, letter to the king, in Lissón Chaves 1943–46, 3–15:496). Similar demands were repeated by the archdiocese of Lima during the first half of the seventeenth century (e.g., Lima Synods 1987, 44). 11. Conde del Villar, April 25, 1588, letter to the king, in Lissón Chaves (1943–46, 3:15, 496). 12. Even a term like la lengua del Cuzco should not necessarily be read as “the Quechua of Cuzco,” since it was assumed that Cuzco was where all Quechua came from and that varieties that differed from Cuzco Quechua were simply corrupt variants. 13. I see little evidence for the use of a Southern Peruvian Quechua lingua franca in Lima and parts of the central highlands, as Itier has suggested (2001, 69). 14. Several of Huerta’s certificates have survived in Lima’s archives, e.g., Curatos I/26a f. 34v (1616), Ordenaciones II/31 (1616), Ordenaciones IV/17 (1627), all in Archivo Histórico Arzobispal de Lima, Lima, Peru; and Registro 13 f. 501v (1628), in Archivo de San Francisco de Lima, Lima, Peru. Cf. Durston 2007a, 121, 127, 334n35, 336n53. 15. A 1651 suit against Agustín de Aller, Capítulos XV/5, in Archivo Histórico Arzobispal de Lima, Lima, Peru. 16. Although it appeared too late to be taken into consideration in this chapter, I would like to draw the reader’s attention to an important edited volume that addresses many of the issues discussed here: Heggarty and Pearce’s History and Language in the Andes (2011).

Alan Durston

[ 241 ]

References Adelaar, Willem, and Pieter Muysken. 2004. The languages of the Andes. Cambridge: Cambridge University Press. Avendaño, Fernando de. n.d. [1649]. Sermones de los misterios de nuestra santa fe catolica en lengua castellana y la general del inca. Lima: Jorge López de Herrera. Bakewell, Peter. 1984. Miners of the red mountain: Indian labor in Potosí, 1545–1650. Albuquerque: University of New Mexico Press. Burke, Peter. 2004. Languages and communities in early modern Europe. Cambridge: Cambridge University Press. Cerrón-Palomino, Rodolfo. 1987. Lingüística quechua. Cuzco, Peru: Centro de Estudios Regionales Andinos “Bartolomé de Las Casas.” ———. 1992. Diversidad y unificación léxica en el mundo andino. In El quechua en debate: Ideología, normalización y enseñanza, ed. Juan Carlos Godenzzi, 205–235. Cuzco, Peru: Centro de Estudios Regionales Andinos “Bartolomé de Las Casas.” ———. 1995. Estudio introductorio. In Santo Tomás [1560] 1995, vii–lxvi. ———. 1997. Las primeras traducciones al quechua y al aymara: Un caso de elaboración y desarrollo estilísticos. Boletín del Instituto Riva Agüero 24:81–102. ———. 1999. Tras las huellas del aimara cuzqueño. Revista Andina 17:137–161. Cohn, Bernard. 1996. Colonialism and its forms of knowledge: The British in India. Princeton, NJ: Princeton University Press. Dávila Morales, Juan Antonio. 1739. Practica de la doctrina christiana: Obra utilissima para los curas, y confessores de indios, y de rusticos. Lima: Francisco Sobrino. Durston, Alan. 2002. El Aptaycachana de Juan de Castromonte—un manual sacramental quechua para la sierra central del Perú (ca. 1650). Bulletin de l’Institut Français d’Etudes Andines 31:219–292. ———. 2003. La escritura del quechua por indígenas en el siglo XVII—nuevas evidencias en el Archivo Arzobispal de Lima (estudio preliminar y edición de textos). Revista Andina 37:207–236. ———. 2007a. Notes on the authorship of the Huarochirí manuscript. Colonial Latin American Review 16:227–241. ———. 2007b. Pastoral Quechua: The history of Christian translation in colonial Peru, 1550–1650. Notre Dame, IN: University of Notre Dame Press. ———. 2008. Native-language literacy in colonial Peru: The question of mundane Quechua writing revisited. Hispanic American Historical Review 88:41–70. Estenssoro Fuchs, Juan Carlos. 2003. Del paganismo a la santidad. La incorporación de los indios del Perú al cristianismo. Lima, Peru: Pontificia Universidad Católical del Perú and Instituto Francés de Estudios Andinos. Fabian, Johannes. 1986. Language and colonial power: The appropriation of Swahili in the former Belgian Congo, 1880–1938. Cambridge: Cambridge University Press. Grammatica y vocabolario en la lengua general del Peru llamada quichua, y en la lengua española: El mas copioso y elegante que hasta agora se ha impresso. [1586] 1603. Seville: Clemente Hidalgo. Heggarty, Paul, and Adrian Pearce, eds. 2011. History and language in the Andes. New York: Palgrave Macmillan. Huerta, Alonso de. [1616] 1993. Arte breve de la lengua quechua. Ed. Ruth Moya and Eduardo Villacís. Quito, Ecuador: Proyecto Educación Bilingüe Intercultural y Corporación Editora Nacional. Irvine, Judith, and Susan Gal. 2000. Language ideology and linguistic differentiation.

[ 242 ]

chapter nine

In Regimes of language: Ideologies, polities, and identities, ed. Paul Kroskrity, 35–83. Santa Fe, NM: School of American Research Press. Itier, César. 1991. Lengua general y comunicación escrita: Cinco cartas en quechua de Cotahuasi, 1616. Revista Andina 9:65–107. ———. 1992. Un nuevo documento colonial escrito por indígenas en quechua general: La petición de los caciques de Uyupacha al obispo de Huamanga (hacia 1670). Lexis 16:1–21. ———. 1995. Quechua y cultura en el Cuzco del siglo XVIII: De la “lengua general” al “idioma del imperio de los incas.” In Del siglo de oro al siglo de las luces: Lenguaje y sociedad en los Andes del siglo XVIII, ed. César Itier, 89–111. Cuzco, Peru: Centro de Estudios Regionales Andinos “Bartolomé de Las Casas.” ———. 1999. Los problemas de edición, datación, autoría y filiación de El robo de Proserpina y sueño de Endimión, auto sacramental colonial en Quechua. In Edición y anotación de textos coloniales hispanoamericanos, ed. I. Arellano and J. A. Rodríguez Garrido, 213–231. Madrid: Iberoamericana; Frankfurt am Main: Vervuert. ———. 2000. Lengua general y quechua cuzqueño en los siglos XVI y XVII. In Desde afuera y desde adentro: Ensayos de etnografía e historia del Cuzco y Apurímac, ed. Luis Milliones, Hiroyasu Tomoeda, and Tatsuhiko Fujii, 47–59. Osaka: National Museum of Ethnology. ———. 2001. La propagation de la langue générale dans le sud du Pérou. In Le savoir, pouvoir des élites dans l’empire espagnol d’Amerique, ed. Bernard Lavallé and Alain Milhou, 63–74. Centre de Recherche sur l’Amérique Espagnole Coloniale, Travaux et Documents no. 3. Paris: Université de la Sorbonne Nouvelle Paris III. ———. 2006. Ollantay, Antonio Valdez y la rebelión de Thupa Amaru. Histórica 30: 65–97. Lima Synods. 1987. Sínodos de Lima de 1613 y 1636. Madrid: Centro de Estudios Históricos del Consejo Superior de Investigaciones Científicas; Salamanca: Instituto de Historia de la Teología Española de la Universidad Pontificia de Salamanca. Lissón Chaves, Emilio, ed. 1943–46. La iglesia de España en el Perú: Colección de documentos para la historia de la iglesia en el Perú, que se encuentran en varios archivos. 4 vols. Seville, Spain: Editorial Católica Española. Lohmann Villena, Guillermo, and María Justina Sarabia Viejo, eds. 1986–87. Francisco de Toledo: Disposiciones gubernativas para el virreinato del Perú 1569–1574. 2 vols. Seville, Spain: Consejo Superior de Investigaciones Científicas. Mannheim, Bruce. 1991. The language of the Inca since the European invasion. Austin: University of Texas Press. Molina, Diego de. n.d. Sermones de la quaresma en lengua quechua. Biblioteca Nacional del Peru, Lima, Peru, MS B203. Written ca. 1649. Ostler, Nicholas. 2005. Empires of the word: A language history of the world. New York: Harper Collins. Parker, Gary J. 1963. La clasificación genética de los dialectos quechuas. Revista del Museo Nacional 32:241–252. Pérez Bocanegra, Juan. 1631. Ritual formulario, e institucion de curas, para administrar a los naturales de este reyno, los santos sacramentos del baptismo, confirmacion, eucaristia, y viatico, penitencia, extremauncion, y matrimonio, con advertencias muy necessarias. Lima: Jerónimo de Contreras. Quaderno de directorio espiritual. n.d. Biblioteca Nacional del Peru, Lima, Peru, MS PR185a.

Alan Durston

[ 243 ]

Rivet, Paul, and Georges de Créqui-Montfort. 1951. Bibliographie des langues aymará et kichua. Vol. 1, 1540–1875. Paris: Institut d’Ethnologie. Rojo Mejía y Ocón, Juan. 1648. Arte de la lengua general de los indios del Peru. Lima: Jorge López de Herrera. Romero, Carlos. 1928. “Un Libro interesante.” Revista Histórica 9, no. 1: 51–87. Salomon, Frank, and George L. Urioste, eds. 1991. The Huarochirí manuscript: A testament of ancient and colonial Andean religion. Austin: University of Texas Press. Santo Tomás, Domingo de. [1560] 1951. Lexicón o vocabulario de la lengua general del Perú. Lima: Instituto de Historia. ———. [1560] 1995. Grammática o arte de la lengua general de los indios de los reynos del Perú por el maestro fray Domingo de Santo Tomás de la Orden de Santo Domingo. Ed. Rodolfo Cerrón-Palomino. Cuzco, Peru: Centro de Estudios Regionales Andinos “Bartolomé de Las Casas.” Sixth Lima Council. 1772. Original del concilio provincial castellano de Lima celebrado el año de 1772 y de su version latina. Archivo del Cabildo Metropolitano de Lima, Lima, Peru, MS volume. Taylor, Gérald, ed. 1999. Ritos y tradiciones de Huarochirí. 2nd rev. ed. Lima: Instituto Francés de Estudios Andinos, Banco Central de Reserva del Perú, and Universidad Particular Ricardo Palma. ———, ed. 2000. Camac, camay y camasca y otros ensayos sobre Huarochirí y Yauyos. Lima: Institut Français d’Etudes Andines and Centro de Estudios Regionales Andinos “Bartolomé de Las Casas.” Third Lima Council. [1584, 1585] 1985. Doctrina christiana y catecismo para instrucción de indios: Facsímil del texto trilingüe. Ed. Luciano Pereña. Madrid: Consejo Superior de Investigaciones Científicas. Torero, Alfredo. 1964. Los dialectos quechuas. Anales Científicos de la Universidad Agraria 2:446–478. ———. 1974. El quechua y la historia social andina. Lima: Universidad Ricardo Palma. ———. 1975. Lingüística e historia de la sociedad andina. In Lingüística e indigenismo moderno de América (Trabajos presentados al XXXIX Congreso Internacional de Americanistas), 221–259. Lima: Instituto de Estudios Peruanos. ———. 1995. Acerca de la lengua chinchaysuyo. In Del siglo de oro al siglo de las luces. Lenguaje y sociedad en los Andes del siglo XVIII, ed. César Itier, 13–31. Cuzco, Peru: Centro de Estudios Regionales Andinos “Bartolomé de Las Casas.” ———. 2002. Idiomas de los Andes: Lingüística e historia. Lima: Instituto Francés de Estudios Andinos and Editorial Horizonte. Trautmann, Thomas R. 2006. Languages and nations: The Dravidian proof in colonial Madras. Berkeley: University of California Press. Untitled Devotionary. n.d. Biblioteca de San Francisco de Ayacucho, Ayacucho, Peru, MS volume. Wightman, Ann. 1990. Indigenous migration and social change: The Forasteros of Cuzco, 1570–1720. Durham, NC: Duke University Press.

10 * Linguistic Subjectivity in Ecologies of Amazonian Language Change christopher ball

1. Introduction Variability in cultural ideologies of subjectivity is linked to language, specifically to the ways in which ethnolinguistic differences are socially mobilized and territorialized. Cultural values concerning what it means to be a full user of language shape how societies evaluate group membership, and such evaluations inform the sets of practices and institutions that constitute local groups and foster members’ senses of belonging and community. Ideologies of the speaking subject shape local ideals about the phenomenology of language and the subjectivity of language users, and in interaction with historical particulars of contact, they influence how and what kinds of linguistic communities are formed. This process is entwined with the ecology of language evolution. Demographic features including numbers of speakers; their social, linguistic, and ethnic backgrounds; gender differences in linguistic ideology and practice; and spatial factors including where linguistic communities are formed, how far apart they are, and how they are connected as speech communities and as language communities can all be influenced by ideologies of linguistic subjectivity. In Latin America, this process may positively or negatively affect indigenous and Iberian linguistic varieties from the point of view of linguistic vitality. I compare different indigenous ideologies of linguistic subjectivity within Amazonia through examination of two multilingual language and culture areas; the Vaupés or Upper Rio Negro and the Upper Xingu. I show that their populations harbor differing cultural assumptions. Those in the Vaupés value the benefits of plurilingualism, whereas those in the Upper Xingu adhere to monolingualism.1 I argue that these ideological differences have affected the development of these speech communities in divergent ways. My analysis follows previous careful and insightful comparisons of these two multilingual Amazonian areas, notably those of Franchetto (2001) and Stenzel (2005). I hope to add to these studies by emphasizing that difference in cultural notions of how one can properly be a

Christopher Ball

[ 245 ]

complete user of language has implications for the full personhood of individuals and their integration into social networks, and that such difference has historically affected the relative linguistic vitality of indigenous languages in the two regions. My comparison of these two Amazonian regions, which had different cultural ideologies of subjectivity, different (socio)linguistic compositions, and different colonial histories, should remind us of the diversity of multilingualism in Amazonia. It also shows how cultural ideas about what makes a speaking subject influence the course of development of the ecologies of multilingual speech communities. My analysis is based on literature review and fieldwork and brings an empirical and ethnographic approach to bear on general questions about the nature of language contact and the linguistic legacy of Iberian imperialism in Latin America.2 I follow Mufwene’s (2001) emphasis on local historical particulars in considering how linguistic ecologies mold language change, and I add to this commitment by seriously considering cultural construals of both linguistic and social-psychological aspects of subjectivity among the list of ecological factors contributing to language evolution. Subjectivity or personhood is socioculturally constructed, historically variable, and always semiotically mediated (Keane 1997). Subjectivity in general encompasses anything that can be culturally and historically relevant to reflexive understandings about the nature of being human. My concern with subjectivity is purposefully focused on linguistic components, often discussed in terms of the speaking subject, and in this case concerns especially but not exclusively monolingual versus plurilingual ideologies, precisely because these are locally relevant to Amazonian societies. I include some related ways in which personhood is constructed, especially though kinship and exchange, to show how linguistic ideologies of subjectivity connect up with wider cultural patterns.3 Subjectivity is only one aspect of the complex process of contact-induced language change, but I argue that it, and cultural determinants generally, should have a more prominent place in our accounts of such processes, alongside formal linguistic, economic, and social (read power and/or hierarchy) factors. In this way, cultural ideologies of the speaking subject present an analytic category that can help to link particular histories of cultural systems and the evolution of the sociolinguistic topographies of linguistic communities in contact.

2. Ecology, Community, and Subjectivity The ecology of language is here used in the spirit of Haugen’s original (1972) formulation, which is meant to include psychological and social

[ 246 ]

chapter ten

factors in the analysis of language variety and change. Haugen uses “ecology” similarly to how anthropologists use “context.” Like a term of art, it risks being so broad as to be meaningless, but its use stands as an important reminder that we should not give too much prominence to any one of the several master tropes that have guided linguistics in different times and places. Haugen identifies three guiding metaphors of linguistic science that he warns can take over to the extent that they cloud holistic study. These are the biological metaphor, which sees language as alive and evolving; the structural metaphor, which sees language as a monolithic ordered object;4 and the instrumental metaphor, which sees language as a tool used for communication. We can identify all of these as reductive tendencies in various guises (see also Silverstein 1987). Haugen’s point is that any one of them taken as an exclusive base of study may blind us to sufficient recognition of what he refers to as the duality of language as both ergon, or product, and energeia, or activity. What we need to do, in other words, is to focus on both structure and use, on both form and function. Another aspect of Haugen’s theory that makes it particularly amenable to investigation of social and cultural variability in what constitutes the speaking subject is his concern with psychology as a component of the ecology of language. Haugen states of language that “part of its ecology is therefore psychological: its interaction with other languages in the mind of bi- and multilingual speakers” (1972, 325). I share Haugen’s focus on the subjective aspects of bi- and multilingualism, although as an anthropologist I investigate cultural ideas about what it means to be a mono-, bi-, or plurilingual person, whereas psycholinguistics might directly inquire into the cognitive mechanisms of being a mono-, bi-, or plurilingual speaker. My analysis is an order removed from this type of question. I argue that there are cultural differences, in terms of systematic yet plastic interpretive schema, in what it means for an individual to count as a full language-using person, or a fully and effectively languaged human being. This can be basically stated as equivalent to the claim that in some cultures it might suffice for people to be speakers of only one language to count as fully developed subjects, whereas in another, speaking two, three, or more languages may be a requisite feature of human subjectivity. Add to this that in some places and times, speaking more than one language might be considered at best suspicious, or worse, immoral, or that speakers often encounter complex context-dependent decisions about which languages one should speak or shun, and to what ends and in what degrees, and we can see the potential for permutations and colorings that may contribute to variety in the ecologies of multilingual communities.

Christopher Ball

[ 247 ]

This concern is at once psychological, social, cultural, and linguistic; and attention to this variability reveals connections to language change. In application to Amazonia, I argue that different normative conceptions of the speaking subject and personhood more generally in the social systems of the Vaupés and Upper Xingu have contributed much to the different linguistic ecologies and thus the different trajectories of linguistic communities in the two sites. Central to the argument I make here is that linguistic subjectivity has an effect on the formation of linguistic communities. Linguistic communities, defined in some way by communication through, and affiliations to, some aspects of language, can be approached analytically from two perspectives. The first perspective construes linguistic communities in terms of “language,” yielding the analytic concept of a “language community.” The second perspective sees linguistic communities in terms of “speech,” giving rise to the analytic category of “speech community” (Silverstein 1996; Irvine 2006). These are complementary perspectives on language, and both are connected to linguistic subjectivity. On the one hand, language communities can be studied from the perspective of how they are united by a common orientation to norms of the structure of language. This attention to code is the basis for the social maintenance of what perhaps first comes to mind regarding linguistic community: groups of people who speak the same language. This orientation to code includes but is not limited to norms of what linguists call prescriptive grammar, or ideal but not necessarily observed grammatical rules that are imposed in some way, often through specific societal institutions of power. The orientation to grammar also includes attention to and implementation of what linguists contrastively refer to as descriptive grammar, those automatic and, in the linguist’s sense truly definitional rather than superfluous, structural aspects of individual languages as abstract computational systems. Language communities are groups of people with shared orientation to grammatical norms of communication; they share a grammatical “language.” On the other hand, speech communities can be studied from the perspective of how they are united by a common orientation to norms of the use of language. Speech communities are groups of people who share an orientation to interactional norms of communication; they share a pragmatics. Sharing in, contesting, and enacting pragmatic principles, the factors around which speech communities are constructed, is a widely encompassing domain of semiotic social action and communication. Speech communities, as shown by the work of Gumperz (1968) and Hymes (1968), among others, are often multilingual, containing many codes. Whether plurilingual or monolingual, speakers in a speech com-

[ 248 ]

chapter ten

munity share knowledge of strategies and repertoires. They know how to interpret the indexical signs of language use, where, for instance, phonological, morphosyntactic, gestural, or register choices signal social information to knowing participants and observers. In multilingual speech communities, what language is being used in a given situation can itself be a social signal of identity, status, gender, and so forth, as shown by studies of code-switching. But often the kind of information that defines speech communities can be communicated independent of language or code choice in an interaction, as actors mobilize principles of politeness, avoidance, accommodation, insult, or generosity that may employ, but do not exclusively depend on, specific grammatical features of a given language. Members of a speech community, multilingual or otherwise, develop a shared ability to recognize how discursive positioning and social actions are accomplished among speakers of potentially multiple languages, and they may develop ways to send similar pragmatic messages using diverse codes. Norms of use sift out to define principles of interaction, either building upon recognized differences in code or in spite of such differences. Intensity of contact between speakers is an important factor in the constitution of speech communities, but this is relative, for members of a speech community may be separated from face-to-face interaction wholly or in part, instead relying upon, for instance, print-mediated or ritually sporadic engagements to reproduce belonging. This points to the important observation that in the construction of speech communities, orientation to shared cultural knowledge about how to conduct interactional business and the sense of community this orientation can generate lie not only in communication and practice, but also in ideology. Membership in communities then, whether conceived of in terms of speech communities or language communities, is gradient (Silverstein 1996). Degree of membership is linguistically negotiated and continually subject to evaluation through diverse semiotic means in the give-andtake of daily interaction through which actors integrate themselves into and separate themselves out from social groups. Many of the principles that actors employ in achieving gradient membership in communities can be analyzed as involving ideologies of subjectivity. How one is socialized to think of what it means to count as a speaking subject informs in what ways and to what degree one can claim allegiance to different communities or effectively keep oneself out of a community, as the case may be. Linguistic anthropology has focused on how language contributes to the construction of personhood and how language is in turn incorporated as a component of personhood in different ways in different places. This constitutes the study of multiplicity in the constitution of the speaking

Christopher Ball

[ 249 ]

subject. To the extent that subjects are aware of and engage with the cultural principles that define their own and others’ subjectivities, we can say that reflexivity helps to generate ideologies of subjectivity. Subjectivity or personhood can be thought of as a cultural construction of the contours and limits of the human.5 The speaking subject in particular may be thought of in terms of the connection between cultural notions of subjectivity and language as manifest in the practice and perception of speech. This category is often used to apply implicit cultural ideas of subjectivity to individual occurrences of discourse and its participants, to evaluate specific subjects’ identity through their language use.

3. Variation in Amazonian Linguistic Communities By the simple fact that the Amazon forest or river basin is designated as an ecosystem, we may be tempted to assume uniformity in human geography as well. It is a mistake, however, to think of Amazonia as socioculturally and linguistically uniform. Anthropology has traditionally focused attention on ethnological pattern in lowland South America, sometimes in connection with environmental and ecological factors (Meggers 1971) and sometimes in connection with mythology (Lévi-Strauss [1969] 1982), but for the most part studies have emphasized regional diversity as well. Linguistics has also focused on documentation of indigenous languages and determination of genetic relations. This research has found related yet widely dispersed languages from stocks such as Arawak and Carib that share core morphosyntactic and lexical attributes. It has also demonstrated wide-ranging variability in the grammatical and phonological characteristics of Amazonian languages (Dixon and Aikhenvald 1999; Aikhenvald 2012). My interest here is to contribute to the documentation of sociolinguistic diversity in Amazonia, with particular attention to the different ways linguistic communities are constructed, and to the ways this relates to differing cultural notions of subjectivity, or the speaking subject. Amazonian speech communities may be monolingual or resolutely multilingual, and some are increasingly bilingual in indigenous and colonial languages as a result of contact. There are commonalities between the Vaupés and Upper Xinguan systems in terms of subjectivity that reflect some general Amazonian tendencies, and these should be discussed before pointing out differences. Common features include a sociocentric notion of subjectivity, whereby there is a default tendency for members of these societies to conceive of the individual in terms of his or her place in a social network. This contrasts with familiar egocentric definitions of subjectivity based on the prior and autonomous distinction of the individual (Geertz 1983). There

[ 250 ]

chapter ten

Table 10.1. Some cultural differences between the Vaupés and the Upper Xingu Ideal linguistic Marriage and repertoire residence Intergroup relations Vaupés Plurilingual

Upper Xingu

Monolingual

Exogamic marriage and viri-/ patrilocal residence

Specific relationships of alliance determined by phratry membership, shared myth, communication characterized by accomodation and code-switching Endogamous Generalized exchange, marriage and periodic ritual comambilocal petition, tendency for residence favored relations between groups of same linguistic stock, shared myth, communication mediated by bilingual translators

Intragroup relations Sib or local longhouse or settlement group hierarchies, sibling rank, and river position

Membership in noble class determined by genealogy, differentiation, and ethnicization of dialectal differences within local groups

is also a general Amazonian emphasis on the fabrication of subjectivity through corporal manipulation, such as diet, body painting, naming, and various kinds of ritual transformation at different life stages (Viveiros de Castro 1977; Hugh-Jones 1979). Another key commonality in both the Vaupés and the Upper Xingu is the role of language as emblem of language group membership and as a “badge” of identity for individual speakers (Jackson 1974). However, what is particularly interesting in looking at these two areas in Amazonia is how language groups are socially integrated with counterpart units in terms of economic exchange and marriage, how they are culturally connected to other groups through shared ritual and myth, and how they are linguistically integrated with or separated from other groups in terms of mutual intelligibility, plurilingualism, and code-switching. In these respects the Vaupés and the Upper Xingu differ in many regards. The most apparent differences in the ways the two areas define linguistic subjectivity are listed in table 10.1.

4. Vaupés Social Structure and Sociolinguistic Landscape Language is the key to identity in the Vaupés. As Jackson says, “the one fact that will be known about an Indian before anything else is his language-aggregate membership” (1974, 53).6 Languages from the East-

Christopher Ball

[ 251 ]

ern Tukanoan, Arawak, and Nadahup families are spoken in the region.7 In the Vaupés, an ideology of patrilineal transmission of ethnolinguistic identity is grounded in the original generation of ancestors from a mythical cobra-canoe, whereby each ranked local descent group can claim to inhabit the spot where their founding ancestor originally descended from the canoe. Perhaps the best-known sociolinguistic feature of the Vaupés is the system of linguistic exogamy that connects the Eastern Tukanoan groups8 and some Arawak speaking groups but excludes Nadahup, in part based upon a regional Tukanoan ideology that considers Nadahup speakers to be unmarriageable as a social class (Aikhenvald 1999; Sorenson 1967; Jackson 1983). The system of alliance and trade of the Vaupés is composed of language groups organized in a network of linguistic exogamy and social hierarchy. In this system marriage restrictions operate on at least two levels: (1) between members of the same linguistic group, and (2) between members of the same phratry, a compound unit composed of language groups related to one another as metaphorical siblings. All males resident in a specific community are speakers and possessors of the patrilineally determined language of community identification. All married women resident in a specific community necessarily come from other communities where they inherited the use of and identification with a paternal language. This results in societal multilingualism and individual plurilingualism, since children in their natal community acquire the language of their father, the language of their mother, and possibly the languages of the other in-marrying women (Jackson 1983; Sorenson 1967). Because residence is virilocal, patrilineal descent has the function of, at least ideally, tying descendants of those who originally exited the cobracanoe with the site where they alighted. This reinforces the hierarchical distribution of sibs, as primary descent groups, along stretches of river. Language groups cover swatches of territory that are internally complicated and mythologically ordered. The mythological cobra-canoe and the generation of language groups (Hugh-Jones 1979) contributes to a Vaupés specific ideology of territoriality with an emphasis on situatedness and strong connections of particular languages to ancestral places (Chernela 1993). In this region especially, movement and displacement, whether under flight from slave raids, forced migration to mission centers, or current emigration to population centers, especially by indigenous women, is demographically and symbolically disruptive to the social order. Since Sorenson’s (1967) original description of the system, it has been recognized that both being an inheritor of a paternally identifying language from one’s birthplace and being a plurilingual individual are central components of Vaupés subjectivity. I elaborate below upon these two key

[ 252 ]

chapter ten

aspects of Vaupés ideologies of subjectivity in relation to language. First is the territorial aspect of an individual’s linguistic subjectivity, whereby her or his place of birth and residence ties her or him, through language affiliation, to a genealogical identity that defines group membership. This in turn affects individual identities and subjective notions of being in the right place versus being out of place, which are ethnographically documented as extremely important to Vaupés constructions of personhood (HughJones 1979; Chernela 1993; Lasmar 2005). Not surprisingly, this component of linguistic subjectivity has very different implications for men and for women, given that exogamy necessitates women’s increased mobility. The second aspect of Vaupés linguistic subjectivity is the expectation of individual plurilingualism. This encourages socialization to the command of multiple codes. However, while individuals may speak many languages, the one code that is emblematic of patrilineal descent remains the central component of a person’s repertoire. In this way the process of becoming a plurilingual subject is culturally shaped. It is not the case that anything goes; culturally proper ways of being a plurilingual subject include upbringing in the right kind of multilingual community and the dominance of the father’s language. Historical changes that brought forced language mixing in mission schools and increased shifts of linguistic communities to Portuguese and Tukano have challenged this culturally Vaupés way of becoming a plurilingual person. I discuss the histories of two Eastern Tukanoan language groups, the Arapaço and the Piratapuya, and I connect their linguistic trajectories past and present to the territorial and plurilingual faces of linguistic subjectivity. The territorial aspect of linguistic subjectivity played a role in the effects of forced dislocation of Arapaço and Piratapuya speakers from their homelands by colonial forces at different historical moments. Cultural restrictions on the construction of plurilingual subjects reflected a basic tendency to negatively evaluate the mixing of languages outside of exogamically ratified exchanges. However, this was exactly the kind of mixing that was inadvertently encouraged by relocation of children to mission schools. It fostered language change including the rise of Tukano as a regional lingua franca at the expense of smaller Tukanoan varieties. I connect these histories to the place of the plurilingual ideology of subjectivity within the multilingual system of alliance. Changes in linguistic demographics caused in part by the conditions of forced relocation and schooling have been, perhaps ironically, perpetuated by the cultural continuity of exogamic principles in both the Arapaço and Piratapuya cases. The gender divide in territorial linguistic subjectivity, in combination with the plurilingual ideology, has also had effects on various Vaupés

Christopher Ball

[ 253 ]

ethnolinguistic groups as indigenous women have increasingly moved to urban centers and married Portuguese-speaking Brazilian men. The resulting picture of language shift in the Vaupés system is consistent with Fleming’s recent conclusion that considering “continuing clan exogamy under conditions of widespread language shift from Eastern Tukanoan languages to Tukano[,] . . . differences between patrilineallytransmitted languages (i.e. patrilectal variation) are not a necessary condition of marriageability[. Rather, it is] patrilectal differentiation [that] is laminated upon clan difference, with genealogical reckoning remaining the essential basis of social group membership.” Furthermore, “gendered registers of language (are) a more potent site for the interactional figuration of personhood” in the regional language ecology (Fleming 2010, vii). On this view cultural factors such as kinship, gender, and subjectivity or personhood articulate with language in ways that defy simple connections between loss of language and loss of culture.

5. Arapaço Obsolescence In the mid-eighteenth century, Spanish and Portuguese colonial forces both sought to control the area surrounding the confluence of the Vaupés and Negro Rivers. The Arapaço, an Eastern Tukanoan–speaking group that occupied the middle Vaupés at the limit of Portuguese river navigation and control was one of the indigenous Vaupés groups first and most drastically affected by Iberian imperialism. Spanish and Portuguese colonial interest in the area conspired with a desire for labor, and indigenous populations were enslaved and recruited for forced labor for a period of nearly two centuries (Hugh-Jones 1993; Wright 2005). The first decades of slave raids and forced migration are referred to as the descimentos, as Indians were taken downriver from relatively protected headwaters to larger colonial settlements. In the 1850s, extractive rubber collection boomed and indigenous slave labor from the Vaupés region became a crucial part of the operation. Salesian missionary efforts also encouraged indigenous relocation with increasing intensity from the late 1700s onward (Chernela and Leed 2003). The resulting long-term conquest created a set of new ecological conditions for Vaupés language groups, and it also had a lasting impact on indigenous subjectivity in many ways. Besides physically challenging the integrity of indigenous groups through severe population decline and intensive assimilationist practices, relocation upset ties between language and place as well as the communicative integrity and group identities of especially minor Eastern Tukanoan languages. Arapaço people experienced the linguistic effects of forced migration

[ 254 ]

chapter ten

beginning with a large-scale relocation to a missionary center in 1790. Chernela and Leed report that “sibs of diverse language groups were moved to inhabit the same settlement. For example, the mission-created settlement at Ipanore contained the Arawak-speaking Tariano with members of several language groups of the Eastern Tukanoan family: Tukano and Piratapuya as well as Arapaço. The Arapaço eventually abandoned their language for the lingua franca of Tukano and for the official Portuguese tongue” (2003, 44). The loss of Arapaço and the adoption of Tukano and Portuguese are described by Chernela and Leed (2003) as results of dislocation to the missionary center. The authors analyze a mythic narrative that tells how the Arapaço founding ancestor, Unurato, traveled downriver and acquired attributes of the Whiteman including Portuguese language ability, promising a future return upriver in the form of a submarine loaded with commercial goods. They interpret his story as a recasting of Arapaço collective identity in the face of colonialism, whereby the group gives up constitutive aspects of its indigenous self, namely its place and its language, both tied to the original ascent of Unurato in the cobra-canoe, in exchange for relative cosmopolitanism and the Portuguese language. The dissociation of language from ancestral place correlates with loss of the code, but the Arapaço become mediators just as Unurato did. The regional emergence of Tukano as a lingua franca, replacing in part the earlier Amazonian Língua Geral, now known as Nheengatu, should be seen as in part a product of the colonial language ecology (Aikhenvald 2002; Stenzel 2005; Moore, chap. 4, this volume; Lee, chap. 5, this volume). Groups that adopt Tukano in exchange for a lesser Tukanoan or Arawak language such as Arapaço and Tariana may benefit in their role as mediator, which afforded Tukano a position of power in the colonial ecology. In Chernela and Leed’s analysis of the Arapaço Unurato myth, the culture hero is cast as precisely the sort of cultural broker that defined the Arapaço’s collective linguistic role as they emerged from the colonial encounter with whites. However, the Arapaço preserve basic aspects of the territorial and plurilingual ideologies of subjectivity. Even though Arapaço people now speak Tukano and Portuguese, they remain ethnically Arapaço. According to Chernela and Leed, “They continue to identify themselves as one exogamous descent group, putatively related through a founding ancestor. In remaining exogamous, but not linguistically distinct, the Arapaço are unique among Eastern Tukanoan peoples” (2003, 44). The Arapaço continue to culturally value the importance of a demonstrated ancestral and linguistic link to a particular place in the novel form of Unurato’s upriver return as a Portuguese speaker. They also persist in defining their group in terms of the exogamous marriage exchange

Christopher Ball

[ 255 ]

that served to define plurilingual subjectivity before Arapaço became a moribund code. This example shows how linguistic subjectivity is enmeshed with other social relations in a given cultural system. It further shows that the loss of an indigenous language and of the necessary difference between father and mother languages in families and communities has not prevented the retention of linguistic subjectivities embodied in valuation of plurilingualism, now in Portuguese and Tukano, nor of exogamy, now uniquely phratric rather than phratric and linguistic. Social transformation and language change, including language shift, do not necessarily entail the loss of culturally specific norms of subjectivity.

6. Piratapuya Piratapuya is an Eastern Tukanoan language traditionally spoken in communities along the Papuri River above its confluence with the Vaupés. The contemporary Piratapuya linguistic scene includes the traditional upriver Piratapuya territory on the Papuri, the nearby downriver regional school center in Iauaretê at the confluence of the Papuri and Vaupés Rivers, and the administrative center of indigenous and state politics further downriver at the city of São Gabriel da Cachoeira on the Negro River. The downriver movement of recent generations of Piratapuya people within the past one hundred years is connected to linguistic subjectivity in ethnohistorical narrative. I interviewed Piratapuya speakers about the causes for shifts in recent generations to increased Tukano and Portuguese use and the perceived concurrent decrease in Piratapuya language use. Indigenous narratives focused on Salesian mission schools and their forced relocation of children to centers such as Iauaretê. While Portuguese was the language of instruction and the only one allowed for communication in these sites, the narratives tend not to make a direct connection between this fact and indigenous language loss. Rather, a more complex account of how territorial displacement and changing conditions on the construction of plurilingual subjects emerges. Piratapuya speakers describe indigenous language mixing, outside of the prescribed exogamic and hierarchical constraints, as the reason for intergenerational language shift. Sorenson (1967) saw fit to terminologically distinguish between the multilingualism characteristic of the Vaupés region as a cultural and linguistic area and the “polylingualism” (i.e., plurilingualism) associated with individual participants in the system. If we move beyond the usual sense of “polylingualism” as merely involving competence that allows use of alternative codes in different contexts of interaction, and also include the values attached to plurilingual practice, the normative rules defin-

[ 256 ]

chapter ten

ing what constitutes plurilingualism, and the deeply embedded place of multilingualism in the mythology, kinship, and economy of the region, then we can get a fuller picture of how community structure influences individuals’ cultural behavior in the Vaupés. The Vaupés ideology of individual plurilingualism reiterates a key feature of group multilingualism. The separation of distinct languages as belonging to distinct groups is necessary to the maintenance of the system and is ideologically reinforced. Separation of codes within individual speakers is one correlate of this. There is both relatively little lexical borrowing between languages and little code switching by speakers (Aikhenvald 2002). In the Piratapuya context, lamentations over the loss of cultural property and knowledge include the destruction of traditional longhouses by Salesian missionaries and, crucially, the rupture of lines of transmission of cultural and linguistic knowledge resulting from forced relocation of children to Salesian mission schools in regional centers after the beginning of the twentieth century.9 This movement is correlated with ruptures in the patrilineal transmission of a language as property and as knowledge, for example, in the form of names of local sites and their associated mythology (Santos-Granero 1998). The spatial movement represents an inversion of the upriver trajectory of the ancestral anaconda-canoe that generated the Tukanoan people, depositing them in village sites in an order corresponding with their positions in the social hierarchy. The Salesian mission schools, where children from various language groups were brought in contact with one another, are considered to have been sites of linguistic mixing. Salesian missionaries attempted to forcibly impose Portuguese monolingualism on Indian students. However, because children who were relocated continued to use indigenous languages rather than totally shifting to Portuguese, they were confronted with new language choices in their covert communications with one another. They remained plurilingual, but contemporary Piratapuya people with whom I spoke blame the environment of language contact for fostering language mixing. Linguistic mixing is represented as a process of lexical substitution, where the most commonly spoken regional languages, Tukano and Portuguese, served as the dominant sources. Narratives that I recorded in the Vaupés depicted minority languages of the Tukanoan and Arawakan language families, which were represented by lower numbers of speakers in the schools, as having been corrupted by language contact in this context, especially since the contact was not regulated in terms of the traditional system of marriage and exchange. Language contact according to exogamic principles is fine, because it discouraged language mixing, but mixing in the Salesian settlements is seen as unstructured and desultory.

Christopher Ball

[ 257 ]

In line with this ideological way of thinking about language purity, language mixing in boarding schools at the hands of forced migration produced a diminished generation of Piratapuya speakers. One person described to me how his father, who left Salesian school a Tukano- rather than Piratapuya-dominant speaker, reasoned that he had better teach his children Tukano well rather than a version of Piratapuya corrupted by mixing in the school environment. Since he had not been raised in a Piratapuya place, he told his children he had lost a vital connection to the Piratapuya language. It may be beside the point how well in fact a speaker can use a given language, if the sense of disconnect and corruption that come from the combined circumstance of relocation and mixing conspire to render the language inert. This process is aided by cultural ideologies of subjectivity. Using the right and proper code in appropriate contexts is positively valued in the Vaupés, for example, accommodating to the patrilineal home language when visiting a longhouse or a village, but speakers are reluctant to use a particular language until they feel they have achieved a strong degree of mastery (Sorenson 1967). This is an indication that completeness plays an important part in the regional ideology of speakerhood, that not only is there a correlation between fullness of persona and fullness of repertoire, but that also completeness or integrity of the plurilingual persona correlates with the robustness of each constitutive element of a speaker’s linguistic repertoire. In addition, the ideology of individual speakerhood cannot be extracted from its cultural ecology, in which it is linked to kinship, both marriage practice and descent reckoning, as well as mythology and the social hierarchies predicated on the juncture of mythical and shallow genealogy. This last point is borne out when we look at differential patterns of language loss in contemporary Vaupés communities, where persistence of cultural patterns of kinship, including descent-based social hierarchy and exogamic marriage alliance and persistence of cultural ideologies of speakerhood, combine with shifting language alliances. Salesian schools and the communities in Iauaretê, for example, that have grown up around their legacies directly challenged that ideology. Testimony to this effect by indigenous Piratapuya forms a part of the narrative history of language and culture loss. Current efforts at revitalizing the Piratapuya language in Papuri River communities also uncover an apparent paradox of cultural continuity: that observation of traditional marriage rules in line with cultural ideologies of subjectivity may in fact expedite shifts away from Piratapuya language use. Recent work by Kristine Stenzel (2005) has shown that Piratapuya is less vital than its neighbor Wanano. However, this does not mean that the vitality of cultural ideologies of subjectivity and their

[ 258 ]

chapter ten

corresponding institutions and practice is also threatened. In fact it may be that cultural continuity of ideologies that support exogamy between language groups is a contributing factor to language loss in this case. This reminds us that simple equations of the loss of language with the loss of culture are often misguided. Stenzel states that while the Piratapuya still participate in virilocal exogamy with at least eight other ethnolinguistic groups, the “preservation of marriage alliances actually works against them as far as language maintenance is concerned, because both the Tariana and the Desana have shifted almost completely to the Tukano language. In other words, nearly all in-marrying wives of Waikhana (Piratapuya) men—93% of the total—though they come from three ethnically distinct groups, are speakers of Tukano” (2005, 24). The problem is that while Piratapuya people continue marriages with their traditional partners, those groups have increasingly shifted to Tukano, so that the Tukano language is imported into Piratapuya communities by wives who otherwise identify as members of non-Tukano-speaking groups such as the Tariana or the Desana. Perhaps ironically here, the cultural persistence of exogamic patterns that were traditionally based on linguistic difference and the cultural valuation of plurilingualism produce the circumstance of Piratapuya marriage alliances undermining the production of a traditional Piratapuya multilingual community by promoting a disproportionate influx of Tukano speech (Stenzel 2005). Central aspects of Vaupés ideologies of linguistic subjectivity remain strong, even to the possible detriment of the continuation of the language.

7. Territorial and Plurilingual Ideologies and Gender The territorial aspect of Vaupés linguistic subjectivity, because of its connection to both kinship and plurilingualism, especially the exogamy and virilocality that ensure multilingual communities through marriage, presents a clear gender divide. Men in the Vaupés are typically more staunchly emplaced than women. In contrast, the latter have been described as perpetually misplaced, as necessarily uprooted from their places of linguistic identification and transplanted to the places of their husbands. Chernela (1993) depicts the social position of Wanano women precisely as “mixed.” Women describe themselves as belonging to the category of su/sari masono ‘mixed one’, tied to natal settlements and through their patriline to their founding ancestors, while simultaneously tied to their new place and language of residence by marriage. In their roles as wives, women are wanderers; their belonging is grounded in another sib and another ancestral location, and female linguistic subjectivity is in part defined by this fact, in addition to the fact of multilingual practice required by viri-

Christopher Ball

[ 259 ]

local exogamy. Not only does this contribute to women’s own subjectivities, as wanderers and outsiders, consistent with the narratives and songs documented by Chernela (1993, 72–83); it also makes women figures of alterity in the social system. Women stand as perennial strangers, and their use of and association with a foreign language, from the point of view of patrilineal ties to place, underscore that role. Virilocal exogamy sets up the expectation of plurilingualism within the multilingual social system as it defines the place of women as subjectively other. Lasmar (2005) documents recent manifestations of Vaupés descent and dislocation downriver in the form of largely female migration to centers such as São Gabriel da Cachoeira. Her analysis looks at how women who move from the traditional “community” to “the city” modify traditional marriage patterns and ideologies as they experience language shift. Lasmar recognizes the importance of both the territorial and the plurilingual/exogamic aspects of Vaupés linguistic ideologies of subjectivity that we have considered. Since the traditional system favors territorial stability for men but requires spatial displacement for women, increased female movement to larger towns represents a kind of continuation of the cultural system. Women’s subjective sense of mixing as they move out of their natal community to find work and possibly marry Portuguesespeaking Brazilian men is accommodated by wider regional ideologies. While women may be forging new linguistic identities for themselves and their children, they do so to a large extent within the bounds of traditionally Vaupés language ideologies of subjectivity. What are the effects on language socialization for their children as they inherit a putatively Portuguese linguistic identity from their fathers and receive Portuguese language instruction in town schools? Lasmar attests that the ethnic identity of the children of mixed indigenous and Brazilian unions in locales such as São Gabriel da Cachoeira is still a matter of great controversy. Patrilineage seems to influence interpretations of the direction of mixing whereby indigenous mothers contribute less of their sib identity to a child than would an indigenous male who fathered a mixed child. Lasmar’s strikingly insightful analysis shows how women actually use the practice of city exogamy to reinforce social ties with kin in traditional communities by sending back money earned in wage labor and by investing effort in having children incorporated as members of their paternal sib, both by bringing their own fathers or other paternal relatives to town to reside with their nuclear family and by enacting sibspecific naming rituals to confirm mixed children’s sib membership. We might also state that while it moves around some traditional features of Vaupés linguistic subjectivity such as strictly patrilineal transmission of linguistic identity, women’s extended sphere of linguistic exogamy at the

[ 260 ]

chapter ten

same time works to preserve the background assumptions in these ideologies of subjectivity. That is, women’s enduring status as linguistically and ethnically mixed subjects shows that even with hiccups in transmission, essentially one’s language identity comes through connection to an upriver community, perhaps even if one has been raised as a Portuguesedominant speaker.

8. The Upper Xingu The Vaupés ideology of individual and community plurilingualism reinforces marriage practices that emphasize ethnolinguistic difference and integration. In the Upper Xingu social system, in contrast to the Vaupés one, there is a regional emphasis on intergroup exchange between endogamous groups accompanied by an ideology that attributes purity of ethnicity to individual monolingualism. Monolingualism and ethnic group endogamy are the norm in this federation of allied ethnolinguistic groups. There are languages from three major Amazonian stocks plus one language isolate spoken in the social network that defines the Upper Xingu. Languages from different families spoken by Upper Xinguan groups are mutually unintelligible. The Upper Xinguan ethnolinguistic groups are listed in table 10.2. Many languages are found side by side in the Upper Xingu, and a network of production and trade of local manufactures and an intense intergroup ritual calendar bring people from different groups together quite regularly. However, plurilingualism as a communicative property of individuals is downplayed, while isolation and ethnolinguistic boundaries are emphasized. The Upper Xingu has experienced contact with the rest of the Brazilian society only recently and was never rigorously missionized or subjected to slave raids and economic exploitation. The upper reaches of the Xingu River were long isolated by impassible rapids. Contact between Xinguans and whites occurred considerably later than native-white contact in other parts of Amazonia. Until the late 1800s, Upper Xinguans experienced only indirect contact with whites. German explorer Karl Von den Steinen Table 10.2. Upper Xinguan ethnolinguistic groups Carib

Arawak

Tupí

Language Isolate

Kuikuro Kalapalo Nahukuwa Matipu

Yawalapiti Wauja Mehinaku

Kamayurá Awetí

Trumai

Christopher Ball

[ 261 ]

“discovered” the Upper Xingu at the turn of the twentieth century. Epidemics began to severely reduce Xinguan populations from this time. Historical records show that Upper Xinguan groups visited the government post at Paranatinga throughout the 1920s (Pina de Barros 2003). In the early 1950s, the exploratory Roncador-Xingu expedition entered the still relatively isolated area of the headwaters of the Xingu River. The brothers Cláudio, Orlando, and Leonardo Villas Bôas opened the Xingu to Western observation, made rounds among the groups, and started to centralize state presence in the region. Their forays culminated in a massive government project of demarcation and the eventual founding of the Xingu Indigenous Park in 1961. What the Villas Bôas viewed as their salvation of the Xingu was an act of pacification, in the sense of a state-centered coercive intervention enforcing political and demographic reorganization (Ramos 1998). The park has since become the most singularly recognizable emblem in the political economy of indigenous identity in Brazil (Garfield 2004). The park’s institutional imposition of ordered stasis, what Menenzes de Bastos (1989) has referred to as “Pax Xinguensis,” amounts to a spatiotemporal regulation of the Xingu from the top down. The public face of the park promotes an image of indigenous fixity and purity and suggests this as its reason for being in the first place. The park’s founders were able to claim to be preserving a nearly lost indigenous heritage at the same time that they subdued the indigenous population. While one of the park’s main goals was to protect Indians from linguistic and cultural contact and corruption, it actually set up a fishbowl environment in which indigenous groups have been both observed and challenged to overcome its boundaries. Since its foundation there has been no shortage of white people visiting the park. It has been often remarked that the Xingu Park hosts the elite of Brazilian and international professional and artistic society, and these people have been at the forefront of indigenous language contact with Portuguese. Besides the doctors, scientists, anthropologists, politicians, environmentalists, and journalists who have visited and worked there, celebrities such as Sting and Leonardo DiCaprio have famously played tourist or taken up activist causes in the park. Conversely, priests, rubber tappers, army, police, and other agents have been relatively absent from the Upper Xinguan contact situation. Since the Villas Bôas, the contact that Xinguans have had with Portuguese has been in large part in relationships with educated professionals engaged in discourses of environmental protection, indigenous health care, cultural valorization, and the politics of pan-indigenous rights in Brazil. This language contact, mediated by specific Portuguese

[ 262 ]

chapter ten

genres, has influenced the form and function of indigenized Portuguese in the Upper Xingu. The brothers Villas Bôas affected contact with Portuguese in the Xingu in another way as well. Upon their arrival, certain individuals and groups positioned themselves as brokers vis-à-vis the Villas Bôas and the white society. The Villas Bôas often took third and fourth sons of chiefs under their wings, bringing them out of their natal villages on reconnaissance missions around the Xingu and, importantly, teaching them Portuguese. One example is a recent Wauja chief. He traveled with Orlando on the early rounds that sketched the nascent territorial contours of the park’s new regime and tells of how Orlando taught him Portuguese. Despite several local impediments to his ascension to chieftainship, he managed to function as the recognized receiver of whites and for many years was the official leader of the community as far as Brazilians and many members of the Wauja community were concerned. It is not uncommon for Upper Xinguan villages to have several ranked chiefs, but the paramount chief is supposed to be the eldest of a noble line and the most senior in ritual knowledge. The Villas Bôas’s intervention some fifty years ago introduced a complication in this situation, producing a series of villages with effectively dual chieftainships, where a Portuguese-speaking elite assumed as much power as, or more than, their monolingual elders. Another effect of the park’s institutionalization and the adoption of Portuguese as the vernacular on the language ecology is the increasing exodus of the park’s residents to the outside. The park is seen nationally to contain an ideal indigenous culture characterized by a pacifistic ethic of cooperation and collaborative ritual exchange. Crossing the limits of the Upper Xingu and the park has been seen by both park residents and outside observers as the condition that undoes this integrity. But today in the Upper Xingu: “The elders die and the children become increasingly fascinated with the white universe, they undertake trips to the city, they circulate in Brasília, São Paulo, Rio de Janeiro. . . . There are ‘educational’ projects and proposals, with schools and experiments in literacy, researchers and producers of all types of images continue to frequent the Park . . . and the Upper Xinguans continue with the same pride in being ‘true’ Indians, pacifiers and civilizers of a savage world, amazing in their inventions and power of subjection” (Franchetto 1992, 354). In this way, we should see that the formation of an Upper Xinguan indigenous identity as traditionally pure has in fact come out of the conditions of contact. Through Upper Xinguan ideologies of linguistic subjectivity, I explore how groups relate to one another and how Portuguese has affected internal and external relationships.

Christopher Ball

[ 263 ]

9. Monolingual Ideology and Intergroup Communication Previous research has characterized the Upper Xinguan monolingualism in terms of a “language barrier,” whereby communicative impediments outweigh linguistic identity (Basso 1973; see also Gregor 1977, 318). On this line of thinking, the purity or integrity of an individual’s linguistic affiliation with his or her local group correlates with the person’s degree of monolingualism. This translates in turn into the practice and ethic of group monolingualism. It remains the duty of a minority of individuals who happen to be plurilingual, to serve as translators at intergroup meetings. Such meetings are frequent, and even though many are ceremonially “formal” in many respects, translation is seen as a potential aid to intergroup social relations, almost never a necessity. As a result, those individuals who do enact a plurilingual subjectivity, a marked Upper Xinguan category, are not categorized into an explicit intermediary social role. Their identification as plurilingual subjects is not institutionalized through participation in speech acts of interpreting between monolingual speakers from different ethnolinguistic groups. One gets the impression that the interpreting is opportunistic and casual, a favor that bilingual persons may do more because they desire to see and interact with a visitor from their old family or local group than because they have a set social role as interpreter. This suggests that the monolingual ideology extends to cover or perhaps downplay any possible overt recognition or prestige falling upon individuals because of their non-monolingualism. They may be asked to interpret, but with the recognition that they will not be singled out as remarkable for this linguistic expertise. It also indicates that, compared with a Euro-American ideology of language that emphasizes the informational and communicative function of language, in the Upper Xingu relatively little importance is placed on the referential communicative function as definitional of language. Unimpeded referential clarity is not seen as necessary for the mediation of successful interactions that can foster meaningful social relations. I have seen members of different Upper Xinguan language groups engage in interactions whose function is merely phatic contact. Participants may interpret interactions in which they claimed to be unable to understand the literal meaning of what their interlocutors were saying as successful in accomplishing some specific social work such as accepting an invitation, giving thanks, coming to a compromise about payment, and so forth. Some space for exogamy and plurilingualism remains in the Upper Xingu. There are two general determinants that lead to exogamic unions. First is the tendency for exogamy to occur more among noble families

[ 264 ]

chapter ten

than among others. The political implications of this include the obvious benefits chiefs may accrue in alliances formed through the marriage of sons and daughters to chiefs of other groups. The Villas Bôas also encouraged noble marriages between groups in order to emphasize interdependence and decrease fighting between local groups. The other circumstance that may lead to exogamy is fallout from witchcraft accusations internal to a given community. The crises that follow in the wake of group-internal factional disputes may push persecuted individuals to flee and seek refuge with another group. Dislocation from one’s natal community is a marginal phenomenon in both cases; either one is married to another group by virtue of membership in an exclusive noble sphere, or one is forced into relocation under threat and accusation of being the lowest type of human, a witch. In my residence in the Wauja village, I knew bilingual young people who were children of exogamic unions produced through both of these channels. Territorial aspects of monolingual subjectivity include the important point made by Stenzel (2005) that whereas in the Vaupés, patrilineal inheritance of language as practice and emblem is paramount, in the Xingu, locality is more important. Even bilingual children of bilingual parents display dominance in one language, which establishes their linguistic identity within the village where they are raised. Thus, one young WaujaKamayura bilingual speaker I knew went to great pains to emphasize his Wauja identity but would quietly use the Kamayurá language when visitors from his father’s group were present. He purposively cultivated as passive a bilingual identity as possible. Doing so conformed to the Upper Xinguan notions of linguistic subjectivity, which include the monolingual individual, endogamic marriage (no ethnic mixing), and emphasis on subdued expressive display.

10. Yawalapiti Mulitlingualism and Obsolescence The Yawalapiti are the exception that proves the rule when it comes to endogamy and monolingualism in the Upper Xingu. The Yawalapiti village is currently home to a mix of primarily people speaking Kuikuru and Kamayura, but also Wauja- and Mehinaku-speaking people. It is decidedly the most multilingual Upper Xinguan community, and its leadership is the paragon of pluriligualism: the current Yawalapiti chief speaks Arawak, Carib, and Tupí varieties in addition to Portuguese. Yet, the Yawalapiti language is one of the most severely endangered of all Upper Xinguan languages, with only a handful of remaining elderly speakers. The Yawalapiti are regarded by other Upper Xinguan groups as having lost their language and, along with it, an essential claim to their identity.

Christopher Ball

[ 265 ]

The history of the formation of the Yawalapiti community is a particularly good example of how the Upper Xinguan ideology of monolingualism and endogamy have conspired with historical conditions to produce concurrently both a novel linguistic and speech community and indigenous language obsolescence. Actually the case involves the convergence of two different monolingual ideologies, that of the Upper Xinguans and that of the planners of the park, the Brazilian explorers who tended to equate one monoethnic nation-state with a bounded locality and a single unified language. The Villas Bôas found the Yawalapiti dispersed in the middle part of the twentieth century. Orlando Villas Bôas planned to reunite the Yawalapiti in one settlement, and he conspired with the young chief Paru to found the new community near the center of the park. But demographic demands, specifically the necessity for more people than the handful of remaining Yawalapiti speakers to form a functioning and sustainable community, allowed the new village to be initially populated also with the Kuikuru and Kamayurá allies of the Yawalapiti. Orlando Villas Bôas also arranged for Paru to marry a noble Kamayurá girl in line with the traditional Upper Xinguan cultural exception to ethnolinguistic endogamy whereby the nobility may enter into politically efficacious intergroup alliance (Menenzes de Bastos 1989). But the Yawalapiti language has consistently receded as a practical medium of communication in the community; it now stands as an ambiguous cultural emblem of what the Yawalapiti have fashioned themselves to be in the contemporary Xingu scene. They have developed a singular multilingual settlement, and multilingualism has become symbolic of their status as mediators both among the Xinguans and between them and others. The current Yawalapiti chief is the son of the Yawalapiti chief Paru, and his Kamayurá bride was raised bilingual in his mother’s Tupí language. I have seen him assume the responsibility of translating into Portuguese for monolingual Carib-speaking chiefs at important meetings between Upper Xinguan chiefs and Brazilian representatives of national health care and international NGOs, scientists on environmental review panels, and so forth. He is regularly the center of attention and the pivot of power for a roomful of indigenous and Brazilian elites. Brazilians, including powerful representatives of regional state politics, with whom he eloquently communicates in Portuguese, clearly perceive how other Xinguan chiefs delegate this role of representation to him.10 He and his siblings speak Arawak varieties with Wauja and Mehinaku speakers, reinforcing a special connection that groups from the same linguistic stock share in the Upper Xinguan system. So the Yawalapiti leadership cultivates a cosmopolitan linguistic identity, as the ethnolinguistic group is

[ 266 ]

chapter ten

increasingly viewed as less authentically or purely Yawalapiti by other Xinguans. This has correlates in material culture, where the Yawalapiti are known to traffic in store-bought metal, cloth, and plastic goods from town in place of any proprietary indigenous manufacture in the regional trading network. The original motivation on both sides for saving the Yawalapiti and the rationale that to do so required a unique settlement involved some notion of ethnic purity dependent upon a single locality and perseverance of the Yawalapiti language. Further, the goal of an ethnolinguistically homogeneous community, even though it was undercut by the linguistic mixing engendered by its construction, combined with the mutually beneficial arrangement wherein the Yawalapiti served as mediators between the rest of the Upper Xingu and the Brazilian government representatives at the park’s central post and led to the founding of the new village in direct proximity to the white center of power. This situation has continued until today; while the Yawalapiti village has moved, it remains the closest to the post, making the Yawalapiti the group most in touch with the pulse of intergroup and intercultural contact mediated by the institutions of government, medicine, and education and the frequent flow of outside traffic into the park. The monolingual ideology has affected locality in complex ways. For many of the in-marrying non-Yawalapiti people and their descendants, adherence to the principle of individual monolingualism has produced Kuikuro- and Kamayurá-speaking factional blocs within the village. Yet this has happened against the ideologically preferred equation of one village with one ancestral language, where the ideal here would seem to have been a monolingual Yawalapiti-speaking community. Instead, ideology and history have conspired to produce a uniquely multilingual community headed by a model of plurilingual subjectivity. And the Yawalapiti language has nearly ceased to be used.

11. Ideology of Etiquette One reason why I stated above that referential communication is not the only, nor the most important, functional benchmark for evaluating linguistic communication in the Upper Xingu is that the area is precisely not a language community, defined in terms of orientation to a grammatical norm. Instead, it is an integrated speech community, with a prominent and clear orientation to pragmatic principles of linguistically mediated interaction.11 One finding of my research with the Upper Xinguan Wauja has been that Wauja actors take the basic interactional strategies that are key to managing successful exchange relations locally and export them to interlocal, interethnic, and international spaces. Other Upper Xinguan

Christopher Ball

[ 267 ]

groups who share a basic orientation to interactional pragmatics successfully reciprocate with the Wauja in interlocal meetings, evidence that the Upper Xingu constitutes a truly integrated speech community, wherein speakers share similar principles of politeness and linguistic comportment in general, although they speak different languages. Analysis of interactions with non-Xinguans, however, shows interesting contrasts and indicates that the Wauja and other Upper Xinguans alike are transforming pragmatic tactics in the face of interactional failures with outsiders. One of the pragmatic principles that link Upper Xinguans as a speech community is a shared ideology of complaint and respect. Basically, notions of social hierarchy, combined with expectations of generosity in exchange, produce a dynamic interrelation between respect and complaint as speech acts that signal particular social relations. A typical case exemplifying this relation might be interaction between a sponsor of a ritual, perhaps a chief, and his contracted performers who will prepare costumes, obtain musical instruments, and effect performances on behalf of the sponsor and his kin, receiving a payment of food, tobacco, and other gifts in return. The chief is obliged to give freely, and the performers are licensed if not expected to complain that what has been given is insufficient. Precisely because this reply is expected, it is more often than not interpreted as ceremonial, although dissatisfaction may indeed be genuine and a complaint may be recognized as marking factual inadequacy. Whether or not any given act of complaint is real or superficial, complaint in general is a second-pair part to an offer, gift, promise, or dispensation of some kind. Ideologically, complaint is recognized as something a person in the position of a supplicant does to someone in the position of benefactor. In this way a complaint may be read as a second-order index of deference. Receiving a complaint, and better, receiving it in stride, is the mark of a noble persona, one who is in the position to distribute to others. In the Upper Xingu, complaint is indexical, in the sense that it is both creative and reflective of political power relations. In addition to the ideology of individual speaker monolingualism, speaker subjectivity is also defined in the Upper Xingu by linguistic comportment. Let me clarify that I do not wish to suggest that such associations are necessarily definitional of individuals’ subjectivity in the sense that we may formulaically state that an individual X is a complainer and thus a commoner. Rather, subjectivity is precisely relational, such that being able to linguistically inhabit the appropriate role position or artfully break expectations in a given interaction is what is at issue. Today’s complainer may be tomorrow’s generous sponsor of a feast, dance, or house-building, himself expecting to be accorded status by receiving complaint. However, in intergroup relations, the monolingual ideology of the Up-

[ 268 ]

chapter ten

per Xingu seems to restrict the potential for complaint to escalate into openly aggressive abusive verbal interchange. Unlike the Vaupés case, for example, where plurilingual members of different local groups gathered for ritual dancing and drinking may expect some amount of raucous argumentation to develop (Goldman 1979; Sorenson 1967), such unpleasantness is almost nonexistent at comparable Upper Xinguan meetings. Verbal exchanges are typically respectful of decorum, and even when one party describes some perceived deficit in a complaint, this speech act is typically more sober (literally, given Upper Xingu temperance) and more densely mediated. As a result, the very real tensions and potential problems that occur in intergroup exchanges in the Upper Xingu end up being worked out through indirect linguistic mechanisms such as rumor, gossip, roundabout accusations, and witchcraft, as well as formalized singing, dancing, name giving, and exchange. One cultural location where this tendency is increasingly tested and at times undone is in contemporary interethnic political meetings at the park’s Central Post or in nearby Brazilian towns, often involving Brazilian and other exogenous interlocutors. I have observed many such meetings in which complaint, still interpretable in terms of its role in Upper Xinguan cultural frames of complaint and respect, took on a potentially more divisive and negative interpretation. This happens partly because Brazilian and foreign agents tend to treat complaints directed at them as highly prejudicial and counter to the communicative goals of the meetings and the overall goals of their advocacy (Ball 2012). That they do so is partially a legacy of paternalistic power structures in indigenism and partly a result of different cultural notions of the role of complaint as a speech act in discourse. In addition, the emergence of Portuguese as a lingua franca in the Upper Xingu has facilitated a new channel for complaint, one where the pragmatic rules of Brazilian culture, in some ways more permissive of grandiosity and passionate appeal, may creep in along with grammar. The use of Portuguese at such meetings also allows direct confrontation between members of ethnic groups in a public forum. Young men are especially prone to engage in arguments in such forums, and one gets the sense in such encounters that tension rapidly builds up and requires the intervention of chiefs to be diffused. Note that in a traditionally monolingual setting, any complaints or accusations made about other ethnolinguistic group members would most likely be made to one’s chief or other noble representative and then secondarily transmitted and mollified in diplomatic translation. Portuguese as lingua franca levels the field, allowing new possibilities for open debate and new potential for ill will in intergroup dynamics, both between indigenous groups and between indigenous groups and outsiders.

Christopher Ball

[ 269 ]

12. Conclusion In the context of Iberian imperialism in Latin America, and even within subregions such as Amazonia, we should not expect a uniform whole but rather a diversity of indigenous responses to, and subsequent fashioning of, the conditions of linguistic contact and change, in line with areal linguistic and cultural diversity. I have compared the Brazilian Amazonian Vaupés and Upper Xingu multilingual areas in terms of their histories of contact and current issues of linguistic change, including language shift from indigenous languages to Portuguese, language shift from indigenous languages to emerging indigenous lingua francas, and the value of language as emblem. I have tried to show how underlying differences between the social systems of the two locations, in combination with cultural values of language in relation to kinship and different economic histories of contact have conspired to produce different ecologies of language evolution. Both regions are currently experiencing language endangerment or shift, but these are interestingly different in form and substance, and my purpose here is to show how these contemporary experiences connect with the particular cultural and historical backgrounds of the two regions. When considering language change under conditions of contact, the analyst seeks to discover principles that guide change in linguistic form. Attention to purely formal questions, however, has for some time been recognized as insufficient to fully grasp generalities of linguistic change and specific histories of language contact. The careful observer of such processes is consistently confronted with the interplay between historical contingency and cultural frames within which language is used in lived practice and within which it is ideologically evaluated as an emblematic sign. My main point has been to emphasize that different ideologies of the speaking subject, how personhood is culturally defined in different societies, affects the contours of linguistic communities and the patterns of linguistic change in contact. The contrasts between linguistic subjectivity in the Vaupés and the Upper Xingu are interesting in part because they show that a range of difference in cultural ideas about language and the social practices that accompany them exists within indigenous Amazonia. Furthermore, as I have tried to show, differences in linguistic subjectivity have interacted with colonial and postcolonial processes of language contact. The time scales of contact are also important in differentiating between the two areas. The Vaupés groups, both Tukanoan and Arawak speakers, saw contact with Spanish and Portuguese colonists beginning in the 1700s. Upper Xinguan groups, by contrast, saw contact with, first, German explorers in the late 1800s and Brazilian nationals in the mid 1900s. We might say

[ 270 ]

chapter ten

that language contact in these two cases in relation to Iberian imperialism also reflects a range of different historical experiences within Amazonia. It is thus also interesting to compare how the consequences of Iberian imperialism as a social historical process are potentially transitive, since the experiences of groups in regions with different time depths of contact, or the experiences of descendants of members of the same groups, are based on shifting and evolving sets of conditions. It has been my purpose to show that one such condition, linguistic subjectivity, is an important component of the ecology of any linguistic community and that it is also socially constructed and historically conditioned. However, linguistic subjectivity is also often a resilient cultural object, and normative ideas about how members of communities can be full speaking subjects may persist for some time, perhaps even outliving linguistic codes that succumb to endangerment and disappearance under the effects of colonialism. One might object that in spite of particular cultural differences of the sort I have described as linguistic subjectivity, processes of language endangerment ultimately lead to obsolescence after all, so what is the point? Let me reiterate that one of my main points is that elements of linguistic subjectivity are in the domain of pragmatics and higher-order evaluations and norms of observable speech behavior, and, as such cultural ideological elements, they can persist for some time after linguistic code shift has occurred. This chapter shows that the elements or features composing ecologies of language evolution or change may march at different paces. Even if an observer were to insist that it all equals language death in the end—which is not much of an explanatory conclusion—it is worthwhile to tease apart how the internal mechanisms work in particular cases so that we may generalize to global processes. Attention to cultural diversity in sociolinguistic contexts in comparison both within and between Amazonian linguistic areas such as the Vaupés or the Upper Xingu yields insights into trajectories of linguistic change in specific linguistic communities. This is part of the larger argument that we should approach the ecology of language with a particularist eye, rather than jumping to generalizations about types, predictive models, and overly powerful metaphors of unidirectional change or loss. Notes 1. I distinguish “multilingual” from “plurilingual” here in reference to social aggregates and individuals, respectively. So we may say that, because they each contain multiple languages, the Vaupés and the Upper Xingu are both multilingual social systems. But in the Vaupés, plurilingual individuals are ideologically and demographically the norm, whereas in the Upper Xingu, plurilingual individuals are ideologically disfavored and are a demographic minority.

Christopher Ball

[ 271 ]

2. Fieldwork was conducted over a period of several months in São Gabriel da Cachoeira in the Brazilian Upper Rio Negro and for more than a year in the Wauja community in the Xingu Indigenous Park. This research was funded variously by the Tinker Travel Fellowship for Research in Latin America from the University of Chicago, a Doolittle Harrison Travel Grant from the Department of Anthropology at the University of Chicago, a Fulbright-Hays Doctoral Dissertation Research Abroad Fellowship, and a Goodman Research Grant from the Department of Anthropology at Dartmouth College. Fieldwork in both areas was made possible through the cooperation of the Área de Linguística of the Museu Paraense Emílio Goeldi. 3. In my work more broadly, I deal with these issues in relation to language, inalienable possessions and exchange, and the role of difference in creating social relations (Ball 2007). 4. Salikoko Mufwene must be credited with independently instilling this point in me through repeated insistence in graduate seminar lectures that “language is not monolithic.” See also Mufwene 1992. 5. We should be careful not to assume a universal definition of the human aspect of subjectivity, however. Related projections of subjectivity commonly found in Amazonia include ancestral, animal, and spirit subjectivities, among others. 6. See Hugh-Jones 1979 for an alternative view of the centrality of language among the productive emblems of group identity. 7. The Nadahup family is also known as Maku, a term that has become disfavored by some linguists and, presumably, speakers. 8. The Eastern Tukanoan–speaking Cubeo are an exception (Goldman 1979). 9. The longhouse has traditionally been the most important center of village life in the Vaupés. Longhouses are also culturally dense objects, and social divisions such as those between male and female and young and old, and even cosmological principles of human relationships to the spirit world are reflected in their design and practices of habitation (Hugh-Jones 1979). The destruction of longhouses by colonists and the subsequent pressure to organize villages as collections of smaller houses disrupted cultural life at many levels. Recently Tukanoan groups have begun to reconstruct longhouses, in part as a way to reclaim traditional ways of life and in part to serve as emblems of indigenous resistance to colonialism. 10. The Yawalapiti chief’s assumption of this interpreter and spokesman role is not without controversy, and his attempts to speak for other Xinguan groups in agreements with Brazilian state and private interests have been targets of anger and resentment on the part of Xinguans. 11. The claim that the Upper Xingu forms a speech community overlaps with, but is not identical to, Beier, Michael, and Sherzer’s (2002) claim that it is a “discourse area,” in which diffusion of discourse patterns over the past fifty to one hundred years appears to have preceded formal linguistic diffusion in the sense of familiar effects of language contact.

References Aikhenvald, Alexandra. 1999. Areal Diffusion and Language Contact in the Içana Vaupés Basin, North-West Amazonia. In The Amazonian Languages, ed. R. M. Dixon and Alexandra Aikhenvald, 385–416. Cambridge: Cambridge University Press. ———. 2002. Language Contact in Amazonia. New York: Oxford University Press.

[ 272 ]

chapter ten

———. 2012. The Languages of the Amazon. Oxford: Oxford University Press. Ball, Christopher. 2007. Out of the Park: Trajectories of Wauja (Xingu Arawak) Language and Culture. PhD diss., University of Chicago. ———. 2012. Stop Loss: Developing Interethnic Relations in Brazil’s Xingu Indigenous Park. Journal of Latin American and Caribbean Anthropology 17, no. 3: 413–434. Basso, Ellen. 1973. The Kalapalo Indians of Central Brazil. New York: Holt, Reinhart, and Winston. Beier, Christine, Lev Michael, and Joel Sherzer. 2002. Discourse Forms and Process in Indigenous Lowland South America: An Areal-Typological Perspective. Annual Review of Anthropology 31:121–145. Chernela, Janet. 1993. The Wanano Indians of the Brazilian Amazon. Austin: University of Texas Press. Chernela, Janet, and Eric J. Leed. 2003. The Deficits of History: Terms of Violence in an Arapaco Myth Complex from the Northwest Amazon. In Language and Social Identity, ed. Richard Blot, 39–56. Westport, CT: Greenwood Press. Dixon, R. M., and Alexandra Aikhenvald. 1999. The Amazonian Languages. Cambridge: Cambridge University Press. Fleming, Luke. 2010. From Patrilects to Performatives: Linguistic Exogamy and Language Shift in the Northwest Amazon. PhD diss., University of Pennsylvania. Franchetto, Bruna. 1992. O Aparecimento dos Caraíba. In História dos Índios no Brasil, ed. Manuela Carneiro da Cunha, 339–356. São Paulo; Companhia das Letras. ———. 2001. Línguas e História no Alto Xingu. In Os Povos do Alto Xingu: História e Cultura, ed. Bruna Franchetto and Michael Heckenberger, 111–156. Rio de Janeiro: Editora da Universidade Federal do Rio do Janeiro. Garfield, Seth. 2004. A Nationalist Environment: Indians, Nature, and the Construction of the Xingu National Park in Brazil. Luso-Brazilian Review 41:139–167. Geertz, Clifford. 1983. Local Knowledge. New York: Basic Books. Goldman, Irving. 1979. The Cubeo: Indians of the Northwest Amazon. Urbana: University of Illinois Press. Gregor, Thomas. 1977. Mehinaku: The Drama of Daily Life in a Brazilian Village. Chicago: University of Chicago Press. Gumperz, John. 1968.The Speech Community. In The Encyclopedia of the Social Sciences, ed. D. Sills, 9:381–386. New York: Macmillan. Haugen, Einar. 1972. The Ecology of Language. Stanford, CA: Stanford University Press. Hugh-Jones, Christine. 1979. From the Milk River: Spatial and Temporal Processes in Northwest Amazonia. Cambridge: Cambridge University Press.– Hugh-Jones, Stephen. 1994. Shamans, Prophets, Priests, and Pastors. In Shamanism, History, and the State, ed. C. Humphrey and N. Thomas, 32–75. Ann Arbor: University of Michigan Press. Hymes, Dell. 1968. Linguistic Problems in Defining the Concept of “Tribe.” In Proceedings of the American Ethnological Society, 1967, ed. J Helm, 23–48. Seattle: University of Washington Press. Irvine, Judith. 2006. Speech and Language Community. In The Encyclopedia of Language and Linguistics, ed. Keith Brown, 689–698. 2nd ed. Amsterdam: Elsevier. Jackson, Jean. 1974. Language Identity of the Colombian Vaupés Indians. In Explorations in the Ethnography of Speaking, ed. Richard Bauman and Joel Sherzer, 50–64. New York: Cambridge University Press.

Christopher Ball

[ 273 ]

———. 1983. The Fish People: Linguistic Exogamy and Tukanoan Identity in Northwest Amazonia. Cambridge: Cambridge University Press.– Keane, Webb. 1997. From Fetishism to Sincerity: On Agency, the Speaking Subject, and Their Historicity in the Context of Religious Conversion. Comparative Studies in Society and History 39:674–693. Lasmar, Christiane. 2005. De Volta ao lago de Leite. São Paulo: Editora Unesp. Lévi-Strauss, Claude. [1969] 1983. The Raw and the Cooked: Mythologiques. Vol. 1. Chicago: University of Chicago Press. Meggers, Betty. 1971. Amazonia: Man and Culture in a Counterfeit Paradise. New York: Aldine. Menenzes de Bastos, Rafael José de. 1989. Exegeses yawalapití e kamayurá da criação do Parque Indígena do Xingu e a invenção da saga dos irmãos Villas Boas. Revista de Antropologia 30–32: 391–426. Mufwene, Salikoko S. 1992. Why Grammars Are Not Monolithic. In The Joy of Grammar: A Festschrift in Honor of James D. McCawley, ed. Diane Brentari, Gary Larson, and Lynn McLeod, 225–250. Amsterdam: John Benjamins. ———. 2001. The Ecology of Language Evolution. Cambridge: Cambridge University Press. Pina de Barros, Edir. 2003. Os Filhos do Sol. São Paulo: Editora da Universidade de São Paulo. Ramos, Alcida Rita. 1998. Indigenism: Ethnic Politics in Brazil. Madison: University of Wisconsin Press. Santos-Granero, Fernando. 1998. Writing History into the Landscape: Space, Myth, and Ritual in Contemporary Amazonia. American Ethnologist 25:128–148. Silverstein, Michael. 1987. The Three Faces of “Function”: Preliminaries to a Psychology of Language. In Social and Functional Approaches to Language and Thought, ed. M. Hickmann, 17–38. New York: Academic Press. ———. 1996. Encountering Language and Languages of Encounter in North American Ethnohistory. Journal of Linguistic Anthropology 6:126–144. Sorenson, Arthur. 1967. Multilingualism in the Northwest Amazon. American Anthropologist 69:670–84. Stenzel, Kristine. 2005. Multilingualism in the Northwest Amazon, Revisited. Proceedings of CIIL. Austin: University of Texas. Viveiros de Castro, Eduardo. 1977. Individuo e Sociedade no Alto Xingu: Os Yawalapíti. Master’s thesis, Universidade Federal do Rio de Janeiro, Museu Nacional. Wright, Robin M. 2005. História Indígena e do Indigenismo no Alto Rio Negro. Campinas, Brazil: Mercado de Letras.

11 * The Ecology of Language Evolution in Latin America: A Haitian Postscript toward a Postcolonial Sequel michel degraff

While reading the preceding chapters in this volume, on Iberian Imperialism and Language Evolution in Latin America, I kept trading two distinct hats on my bald head: one for the theoretical linguist interested in the cognitive aspects of language contact and language evolution, the other for the MIT professor challenged by social injustice in language policy and education in my native Haiti and other Creole-speaking communities. These communities, like many others in the world, including the United States, still suffer from insidious colonial and neocolonial imperialist prejudices and practices. By the time I finished those chapters, I realized that the two hats are fundamentally made of the same material. As a theoretical linguist, I was fascinated by the contributors’ insightful illustrations of the complexity of language contact in Latin America— complexity in sociohistorical, ecological, and linguistic-structural dimensions. As a Haitian and a Haitian Creole–speaking linguist, I was curious as to how language shift, language change, language endangerment, and (meta-)linguistic correlates of social hierarchies in Iberian America may help us better understand related phenomena in the Caribbean, and vice versa. I’ve used the phrases Latin America and Iberian America with some trepidation, as I realize that the chapters to which I am responding have focused exclusively on areas of Latin America that were colonized by the Spanish or the Portuguese, leaving aside Latin American territories that were or are still under the control of France. Now consider my native Haiti, where both French and a French-derived Creole are spoken; Haitian Creole is spoken by virtually everyone there, and French by a small minority, no more than 10 percent (Dejean 2006). Taking the Latin in Latin America in its linguistic genealogical sense, we can then consider Haiti at least as “Latin” as Mexico, Bolivia, Peru, Brazil, and so forth. Those who still subscribe to the classic dogma that Creoles derive from pidgins and therefore fall outside the scope of the comparative method and its associated Stammbaum (“family tree”) model for language change

Michel DeGraff

[ 275 ]

should confront these views head on with the extended argument that Haitian Creole is genealogically related to French and strictly within the scope of the comparative method (Weinreich 1958; Mufwene 2008; DeGraff 2009; DeGraff et al. 2013; Aboh and DeGraff forthcoming-a, forthcoming-b). It may seem ironic that, on linguistic genealogical grounds, Haiti, where everyone speaks at least one Romance language (namely Haitian Creole, or Kreyòl), is more “Latin” than some of the areas studied in other chapters of this volume. Those areas include communities that, by and large, speak Indigenous languages such as Maya, Quechua, and Nheengatu, which are not genealogically related to Latin.1 Furthermore, and unlike Haiti, communities can be found in the Andean highlands and in the Amazon where no Romance language is spoken as a native language (Hornberger and Coronel-Molina 2004; Godenzzi 2008; various chapters in this volume). Also of note is that Kreyòl, with more than 10 million speakers in Haiti alone, is the third-largest Romance language spoken in Latin America (Mathieu 2005). Lastly, Haiti bears the distinction of being the first nation in Latin America to rid itself of its European colonizers. There is thus some poetry in having this volume end with a postscript that treats Haiti as one important case study in our joint investigation of the consequences of European imperialism on language evolution in Latin America. In section 1, I raise basic questions about language contact and language change in Latin America. Then I turn in section 2 to the often insidious and too rarely discussed relationship between knowledge and power in studies of language evolution. In sections 3–7, I examine the politics of Creole studies (e.g., Creole Exceptionalism) and of language and education in Haiti; then I use Haiti as a case study to help us better apprehend related methodological issues vis-à-vis the rest of Latin America, as well as the effects therein of (neo)colonialism on language vitality and endangerment. Comparing Creole Exceptionalism to its counterparts with respect to the rest of Latin America may shed new light on some of the common sociohistorical roots of various myths about Creole and Indigenous languages and about their speakers. In section 7, I also compare language shift across the Caribbean and Iberian America, with a detour about the role of cultural subjectivities in language shift. Then in sections 8 and 9, I consider some of the book’s insights about the impact of cultural subjectivities on language evolution. These insights raise constructive, but still unanswered, questions about Creole formation in the Caribbean. Sections 10 and 11 turn to the future and consider lessons from the past that we have yet to apply in order to counteract some of the hierarchies that most affect social justice in Latin America. Sections 12 and 13 are an optimistic plea for a “sequel” where North-South collabo-

[ 276 ]

chapter eleven

ration among linguists and educators can improve social justice in Latin America and beyond. From my own experience as a Haitian linguist at MIT with both personal and academic ties to the Caribbean, I feel that linguists and educators from both the Global North and the Global South have a great deal to learn from, and contribute to, each other’s academic and political agendas.

1. Who? Where? Why? How? When? What? On the linguistic empirical and theoretical fronts, the data and observations throughout this book should convince us of the futility of applying a cookie-cutter approach to the results of language contact, including the languages that linguists have labeled Creoles. It seems to me that this book’s insights support the general approach that I have advocated in previous work (e.g., DeGraff 2009). Let’s assume, a priori, no more than what we know to date about universal constraints on language variation across the human species. Then, let’s further examine local historical particulars such as those investigated in this book, and let’s try to understand how they may have influenced the contingencies of specific language-contact situations. No matter the range of contingent ecological factors, the outcome of language acquisition across what may seem wildly differing circumstances must still fall within the range of grammatical structures that is bounded by our species’ biological endowment for language. The scientific challenge posed by the vast range of contingent ecological factors affecting the outcomes of language contact may be best illustrated by comparing the chapters with one another. But that seems to me a boring thing to do. Instead I’ll let my own hunches about languagecontact phenomena and my work on language and education in my native Haiti guide my meditations about the contributions to this volume. My basic working assumptions are “Cartesian-Uniformitarian” (DeGraff 2009), whereby language change is the result of the interaction between contingent historical factors and invariant biological constraints on language. This interaction is best apprehended by examining as diverse a range of historical factors as possible. On that score, John Lipski’s chapter on Spanish dialect diversification in Latin America insightfully illustrates the diversity of historical factors that influenced the changes affecting one set of speech varieties that together often pass as belonging to one language labeled “Spanish.” What Lipski suggests is that certain diachronic patterns in various local varieties of Spanish in the Americas are reflexes of analogous tendencies of Spanish in Europe. In Lipski’s scenario, it took at least a couple of centuries for the evolution of Latin American varieties of Spanish to become

Michel DeGraff

[ 277 ]

decoupled from structural changes affecting their cousins in Europe. Lipski posits that one key factor for this decoupling was the advent of big cities in Latin America, and they did not exist until the 1700s. Once Latin American cities surpassed a certain demographic threshold to start functioning as “big cities,” the specific linguistic norms for the corresponding urban varieties became stable and robust enough to achieve autonomy and acquire relative immunity to potential influence from the speech of newly arriving settlers and migrants from Europe. In a somewhat similar fashion, big cities also played a formative role in the role of Quechua as lingua franca in the colonial Andes (Durston, chap. 9, this volume). In the evolution of Hawaiian Creole as well, it’s been argued that it is in cities, and not on plantations, that Hawaiian Creole achieved its stable norms as an autonomous speech variety (Roberts 1998; Mufwene 2008). In contradistinction, it’s in rural areas that the most restructured varieties of Brazilian Portuguese are found, but that too is a reflex of historical contingencies, which in this case involve the geographic specifics of the language-contact situation in Brazil (cf. Mello, chap. 6; Clements, chap. 7, this volume). For Lipski, another crucial factor affecting the evolution of Spanish in the Americas is the constant contact between Spanish and various other languages. From a mentalist perspective, this “contact” takes place in the minds of those who speak Spanish as a second language or those who are native speakers of both Spanish and some other language. Spanish thus came in contact with Amerindian languages such as Quechua and Aymara and with the Niger-Congo languages of the enslaved laborers from Africa such as Yoruba and Kikongo in colonial Cuba (also see Clements’s and Mello’s chapters). Spanish also came into contact with other European languages spoken by non-Spanish settlers, such as Italian in Buenos Aires and Montevideo in the nineteenth century. In the Dominican Republic from the nineteenth century up to the present, Spanish has found itself in contact with Haitian Creole, another product of colonial contact. The details of the periodization and demographics of these overlapping layers of contact phenomena varied from locale to locale, thus rendering the overall historical scenario extremely complex. This complexity is reflected in an impressive range of “innovative hybrid patterns” found across Latin American Spanish varieties, as surveyed by Lipski. Compare, say, these three classes of contact-induced phenomena: (1) Quechua and Aymara influences on the morphological markers of evidentiality in Andean Spanish; (2) Italian influences on the “long-fall” pitch accent of Buenos Aires and Montevideo Spanish; and (3) Bantu influences on double negation markers (with both pre- and postpredicate no) in certain varieties of Dominican and Afro-Cuban Spanish.

[ 278 ]

chapter eleven

As it turns out, similar instances of pre- and postpredicate negation can also be found in both Haitian Creole and Haitian French, as in, respectively, Mwen pa kontan, non and Je suis pas content, non ‘I am not happy, no’. The Haitian Creole (HC) fact is duly noted by Lipski, who then considers the possibility that HC could have been a conduit for these doublenegation patterns into Dominican Spanish. (Bantu influences on Portuguese are likewise documented in Clements’s and Mello’s chapters.) It’s worth noting that such cases of contact-driven syntactic innovation are amenable to constructive theoretical analyses. In the case of Andean Spanish varieties with morphological markers of evidentiality, Sánchez (2004) proposes a “functional convergence hypothesis” whereby interpretable features in the complementizer domain are the loci for syntactic changes via the idiolects spoken by Spanish-Quechua bilinguals: the innovation arises when functional features in the two grammars (i.e., the I-languages that underlie the Spanish and Quechua idiolects of the bilinguals) assume common values. In Sánchez’s hypothesis, “interpretable features are the locus of permeability between grammars in the bilingual mind” (2004, 147). Related increments of complexity in historical and demographic patterns, with concomitant increments of complexity in patterns of structural innovation cum “hybridization,” obtain in the history of the American varieties of Portuguese, English, French and Dutch, alongside the multitude of Indigenous and contact-language varieties of the Americas, namely, Yucatec Maya, Carib, Arawak, Tupí-Guaraní and Jê languages, the contact varieties known as “Lengua/Língua Ge(ne)ral,” the Caribbean Creoles, and so on. Regarding Portuguese, see the aforementioned chapters by Clancy Clements and by Heliana Mello. As for Yucatec Maya, Barbara Pfeiler’s chapter is equally informative. The chapters by Christopher Ball, Alan Durston, Denny Moore, and M. Kittiya Lee provide useful information on the “Lengua/Língua Ge(ne)ral” varieties of the Andes and the Amazon. The latter four chapters remind us that language contact, as entailed by migrations, explorations, and conquests, prevailed in the Americas long before the arrival of the Europeans. Indeed, long before contact with the Europeans, population movements among the Maya, the Tupí-Guaraní, the Tukanoan people, and the Incas, among others, were already spreading their respective languages and producing new Indigenous contact varieties. The chapters on Amerindian languages are particularly informative in highlighting one aspect of history that is too rarely discussed: the roles of certain Indigenous peoples in the Americas (e.g., the aforementioned Tupí-Guaraní and Quechua speakers) as sophisticated agents of exploration and conquest, rather than merely as victims of European imperial-

Michel DeGraff

[ 279 ]

ism. It is in fact the routes already taken by these Indigenous expansionists that the European colonizers often followed in order to build their own Latin American empires. Then again, Ball’s chapter highlights fascinating insights about the various ways in which linguistic phenomena such as plurilingualism and language mixing, alongside certain attendant complex ideologies (e.g., about language “purity”), enter into Amerindian peoples’ definitions of personhood and their senses of community membership, social networks, and social classes. As argued by Ball, these “cultural ideologies of subjectivities” may have impacted the outcome of contact, both among certain Amerindian languages and between Amerindian and European languages. In particular, these ideologies have to do with beliefs and attitudes toward monolingualism versus plurilingualism, as well as the role of language in defining gender roles, social classes, intermarriage practices, and the relationships between geography and identity. In Amazonia, one such set of beliefs relates to whether full personhood is culturally defined as requiring competence in no more than one single language, as in the Upper Xingu area, or in more than one language, as in the Vaupés area. Upper Xingu “ideology . . . attributes purity of ethnicity to individual monolingualism. Monolingualism and ethnic group endogamy are the norm,” whereas the Vaupés manifests “connection to both kinship and plurilingualism, especially the exogamy and virilocality that ensure multilingual communities through marriage” (Ball, chap. 10, this volume, p. 258). Here too, the book constructively forces a reversal of our traditional views of Indigenous people as victims of “loss” that is “caused” by contact with “dominant” Europeans. According to Ball, language “loss” by certain Amerindian groups can be analyzed as part of a concerted response to (re)negotiate identity, power, and conquest, coupled with conscious strategies for maintaining one’s own culture, including its mating practices. In Amazonia, certain recurring patterns of language shift (with monolingualism or plurilingualism), as mediated by exogamy and migration, can be part and parcel of the very cultures that are being maintained. Indeed language mixing and language shift are experienced differently by, and mean different things to, an Amerindian whether it’s monolingualism or plurilingualism that is valued as a criterion for personhood. (In other sections I will have more to say about this and other insightful observations in Ball’s chapter). On the linguistics front, many of the resulting “innovative hybrid patterns” that are surveyed in this book could be taken to have added a certain degree of local complexity (i.e., complexity in certain domains) to the pool of grammatical structures available in the language-contact

[ 280 ]

chapter eleven

situation (for related discussions, see DeGraff 2001b, 250–259; 2009, 963n8; Aboh 2009; Aboh and DeGraff forthcoming-a, forthcoming-b). There are certainly other innovations (e.g., so called morphosyntactic “loss” and morphological “erosion” in inflectional paradigm) that may suggest a contact-induced decrease in complexity. But such decrements, it seems to me, are of a local nature as well and do not necessarily entail an overall decrease in complexity, contrary to still popular claims in Creole studies (see DeGraff 2001b, 2009; Aboh and DeGraff forthcoming-a, forthcoming-b, for overviews and counterarguments). I cannot think of any documented case of a natural language becoming simpler in all modules of grammar at once. At a more fundamental level, I have yet to see a rigorous measure of complexity that applies to an entire language—its lexicon, phonology, morphology, syntax, semantics, pragmatics, and so on (DeGraff 2001b, 265–273; Deutscher 2009). In this vein, Denny Moore’s chapter is a fascinating study of Nheengatu, the lingua franca that was seeded by contact between Tupínamba and other Amerindian varieties in the Amazon. In his documentation of contact-induced diachronic patterns, Moore describes structural innovations based on Indigenous morphemes: the emergence of ‘to have’, the emergence of nominal plural markers, the replacement of nominalizations by relative clauses, and so forth. He argues that these innovations are “constructions which do not exist in Portuguese and probably did not exist in Tupinambá.” Thus, if anything, these would signal an increase of complexity if complexity is taken simplistically to be a matter of bitcounting (DeGraff 2001b, 265–273). As for the cases of “loss,” Moore argues that “most [of the lost features] could be said to be relatively difficult to learn as an adult,” and he analyzes these diachronic patterns as the result of general processes of language acquisition. These patterns are similar to the effects of second-language acquisition in better-studied cases of contact-induced language change where “morphological erosion” reduces the size of certain morphological paradigms. These are cases of “local simplification” that have long been observed in the evolution of various languages, including Creole languages (Bunsen 1854; Meillet 1914; Weinreich 1958; I discuss related cases in a section titled “Local Simplicity” in DeGraff 2009, 957–958; also see Aboh and DeGraff forthcoming-a). To borrow and adapt a quip from John Lipski’s chapter, the “who’s,” “where’s,” “why’s” and “when’s” in the demography, geography, sociology, and history of language contact in the Americas are acutely varied and even more varied when we add the “how” questions that are suggested by Christopher Ball’s chapter. These questions are too diverse for a uniform answer. They challenge any hypothesis whereby the results of

Michel DeGraff

[ 281 ]

language contact would all fall into some prototypical mold and would all congregate toward the bottom of some still ill-defined and arbitrary hierarchy of grammatical complexity. The enlightening studies in this book thus invite us to look deeper into the fine details of linguistic, demographic, ethnographic, and historical patterns in order to develop a rich and nuanced narrative for language change in the Americas. My hunch is that the elaboration of empirically adequate narratives of contact-induced language change in the New World has been hampered by a propensity on the part of many observers, especially in the colonial period, to describe, or even prescribe, particular speech varieties of their choice and use them as tools for imperialism—a topic to which I now turn.

2. Colonial Power, Colonial Knowledge, and Linguistic Imperialism One of the common threads in narratives of language change in the Americas is the mutually sustaining relationship between, on the one hand, the production and transmission of knowledge about the ethnic groups brought into contact by colonization, enslavement, and forced migration and labor and, on the other hand, the production and transmission of hierarchies of power among the groups involved. One core domain for such production of colonial knowledge is constituted by the descriptions, analyses, and uses of diverse speech varieties in the colonial milieu, especially those spoken by the groups at the bottom and in the middle of these colonial hierarchies of power. These varieties have traditionally been described by scholars at the top of the hierarchies. Interestingly, these scholars were among those most invested in keeping the other groupings in a subordinate position, and their writings have promoted a series of structural shibboleths in order to create, then separate and control, subaltern categories of people. Scholarly and scholastic activities and beliefs relating to the speech varieties of those speakers in socioeconomically and politically subordinate positions have played a key role in the creation, negotiation, and transmission of power and in the distribution of symbolic and material capital, within and across these groups. Such relationships between, on the one hand, linguistic and metalinguistic knowledge and, on the other hand, competition for power and capital and the resulting hierarchies are best apprehended in the chapters that deal with either of the following issues: (1) the fate of Indigenous and African languages in Latin America; (2) the varying degrees of intellectual development that are ascribed to the nonwhite populations of the Americas. These populations’ intellectual development is often

[ 282 ]

chapter eleven

measured by their respective degrees of fluency in the relevant European languages—usually the ones spoken by those who put themselves in charge of ethnographic cum linguistic description, education, socioeconomic development, and the like. In other cases, such as in the colonial Caribbean, intellectual development is equated with some arbitrary measure of perceived linguistic complexity on the part of the nonwhite populations. As a matter of fact, these hierarchies of power and their correlates in perceived hierarchies of linguistic, cultural, and intellectual differences are still with us today, with often brutal consequences for social justice or, more precisely, the lack thereof. In the sixteenth century it was claimed by European colonists that the Amerindians were “so brute that they [did] not even have words” (Couto, chap. 3, this volume, p. 83), and now in the twenty-first century it is still claimed, even by linguists (e.g., McWhorter 2001), that Creole languages have the simplest grammars of all natural languages. (See DeGraff 2001a, 2005a for responses to these claims and for historical surveys of these and related views in Creole studies and their impact on Creole-speaking societies.) The next five sections (3–7) examine the politics of linguistics, language, and education in Haiti as a looking glass to help us analyze the ecology of language evolution in the rest of Latin America, especially the effects of colonialism and neocolonialism on the vitality of Indigenous languages. I argue that the interaction in Haiti between linguistics and education is not part of an isolated case of Creole Exceptionalism (in the sense of DeGraff 2005a) but that it reflects, in a rather spectacular fashion, local and global political struggles that are quite similar to those that have affected education and language evolution in the whole of Latin America, especially after the advent of European imperialism (cf. Devonish 2007; Roberts 2008; Migge et al. 2010). This comparison will provide us with empirical and epistemological tools to demystify various linguistics- and education-related myths about Indigenous languages and their speakers in the Americas and beyond (Martínez Cobo 1987).

3. Language (Mis)management, (Mis)education, and Poverty in Haiti My own native Haiti is the country where the raw manifestations of social injustice with linguistic correlates are perhaps the most blatant and crushing. In this (in)famous homeland of mine, the entire population of some 10 million speaks Haitian Creole (locally known as Kreyòl), but, in effect, French has been, for more than three centuries, the dominant language of administration, law, education, business, and so forth, especially

Michel DeGraff

[ 283 ]

in the highest echelons. One of the most common arguments invoked to favor French over Kreyòl is that French grammar is more developed and more adapted to the modern world than Kreyòl grammar, whose structures are claimed to be exceptionally simple. Like the grammars of other Creole languages, Kreyòl’s grammar has been considered a “handicap” for its speakers, thus unsuited for education and other formal domains that presuppose it. (See DeGraff 2001a, 2005a; Devonish 2007; Roberts 2008; Migge et al. 2010 for historical overviews of such claims.) Another popular argument invoked in favor of French over Kreyòl is that the former, unlike the latter, provides an expansive window to the world outside Haiti, and that Kreyòl isolates Haitians from the rest of the world. This ethnographic situation of French in relation to Kreyòl in Haiti seems somewhat similar to that of Spanish and Portuguese in relation to the Indigenous languages of Latin America discussed in previous chapters in this volume, but with the notable distinction that in Haiti the local language is spoken by virtually every Haitian. As such, it is Kreyòl, and not French, that offers a common window, and a common means of communication, to all Haitians in Haiti. This key characteristic of Kreyòl in Haiti—as a truly national language in practical terms—rules out certain arguments that have typically been offered as a common explanation for the language-based inequities that obtain throughout Latin America. The ethnographic status of Kreyòl as the sole national language in Haiti preempts the argument often invoked against Indigenous languages, namely, that they prevent communication across ethnic groups within national boundaries. In Haiti, it is striking that the domination of French persists even after Kreyòl was proclaimed an “official” language in the 1987 Haitian Constitution and is unquestionably the only language that “bonds” all Haitians together. In a related vein, the language was accorded an official orthography in 1979, and since the early 1980s the country’s official programs for education have prescribed the use of Kreyòl as the initial language of instruction in primary schools. But in practice most books and most exams are still in French, even though the language is spoken by less than 10 percent of the population. Students from communities where only Kreyòl is spoken (by far the most typical situation) have little chance to succeed in school and even less chance to make it to university. This linguistic cum educational apartheid seems reflected in the following statistics: out of ten children who start primary school, at most one will successfully complete secondary school (Groupe de Travail sur l’Education et la Formation 2010, 151). This apartheid and the concomitant failure of the school system are, in turn, among the factors that seem correlated with Haiti’s overwhelming poverty (Dejean 2006; Dejean and DeGraff 2013). Indeed, it has been

[ 284 ]

chapter eleven

convincingly argued that, by and large, countries that do not use their populations’ native languages as the generalized media of instruction are those with the worst records of academic achievement and the worst levels of national development (Walter 2008; Hebblethwaite 2012). The prominence of French in schools and exams presents one additional and major challenge to those students who come from Kreyòl-only speech communities. These students are typically excluded from the successful minority who make it to university—no more than 10 percent of those who enter the first grade (Groupe de Travail sur l’Education et la Formation 2010). This is a most brutal case of “élite closure” (MyersScotton 1993) where lack of fluency in French is a steep barrier to academic and socioeconomic advancement for the majority of the population. As for those who are born among the tiny proportion of families that speak both French and Kreyòl (less than 5% of the population), they will automatically acquire both languages as mother tongues, whereas another 5 percent will manage to learn French as a hard-won second language in school (Dejean 2006; Dejean and DeGraff 2013). It is these bilingual Haitians who, by and large, are likely to become successful professionals and dictate the future of the country’s governmental, academic, economic, and cultural institutions. It is thus that native(-like) fluency in French has become a jealously guarded birthright to elite membership.

4. Linguistic Correlates of Colonial Hierarchies of Power in the Americas In Iberian America as well, language barriers are key instruments for elite closure, and these barriers have roots in colonial hierarchies of power as described, for example, in Heliana Mello’s chapter on Portuguese in Brazil. There we read about the origins of elite closure and its linguistic and ethnographic correlates among the slaves: the Negros Ladinos ‘acculturated blacks’ who could speak Portuguese were favored over the Negros Boçais ‘bozal/stupid blacks’ who could not speak Portuguese. “The ladinos had higher status owing to their linguistic skills, which were an important asset and helped some move up in the colonial population structure” (Mello, chap. 6, this volume, p. 171). The perceived superiority of Ladinos over Boçais in Brazil is analogous to that of Creole blacks over Bossal blacks throughout much of the colonial New World, including the Caribbean. The perceived superiority of Creole was most famously quantified by Moreau de Saint-Méry, who assigned to the Creole slaves “worth [that] is always a quarter more than that of the Africans” (1797, 1:40). I return to this issue in section 8. It’s noteworthy that the linguistic markers of Ladinos versus Boçais

Michel DeGraff

[ 285 ]

were relatively independent of literacy levels: illiteracy was a general characteristic of the colonial population at that time, across racial groups. As noted by Mello, most of the early Portuguese settlers in colonial Brazil were themselves illiterate, with no formal education. So their fluency in any language was not a matter of schooling or an index of intelligence, but a mere reflex of where they had grown up and who they had the opportunity to interact with in their everyday lives. Here I am reminded of the Haitian saying Pale franse pa vle di lespri ‘Being able to speak French doesn’t mean that one is intelligent’. But, among the blacks in colonial Brazil, the Negros Ladinos (i.e., those who could speak Portuguese) were considered intellectually superior to the Negros Boçais even though their linguistic skills were generally due not to schooling or any intellectual prowess, but to contingent factors such as the length and circumstances of their exposure to Portuguese speakers. The Negros Boçais, like the Bossales in the Caribbean, were considered less intelligent simply because they were born in Africa, had recently arrived in the New World, and were less acculturated than the locally born and seasoned slaves. This ascription of superiority to Ladinos over Boçais is echoed by the superiority that is (self-)assigned to the very small percentage of contemporary Haitians who are fluent in both French and HC and have maintained an approximation of the stratification that was current during the colonial period, although the colonial stratification, unlike the postcolonial one, included the French-born colonial settlers and their Creole descendants at the top of the socioeconomic hierarchy. Furthermore, in Haiti one often hears the argument, among policymakers, intellectuals, educators, and parents, that those Haitians who do not speak French are, to put it politely, cognitively or socially handicapped (see Mathieu 2005, 2008, 2010; Zefi 2011; Saint-Fort 2011). What is often not taken into account is that fluency in any language, even for children, requires adequate exposure to that language, either via instruction or via immersion (i.e., extensive contact with fluent speakers of the language). These conditions do not obtain for the majority of Haitians in Haiti with respect to French; neither did they obtain, with respect to Portuguese, for the Negros Boçais in colonial Brazil.

5. “Creole Exceptionalism”—from the Caribbean to the Andes? As recently as June 2010, one very prominent Haitian intellectual and politician, historian and former president, Leslie Manigat, described Kreyòl as an “infirmité” (Mathieu 2010; Zefi 2011). When a foremost Haitian intellectual, speaking in French, calls a bona fide language an “infirmity,” he is illustrating both the depth of anti-Kreyòl ideology among Haitian

[ 286 ]

chapter eleven

élites and the aforementioned Kreyòl saying Pale franse pa vle di lespri ‘Being able to speak French doesn’t mean that one is intelligent’. Manigat’s statement also brings to mind the common assumption that Creole languages constitute an exceptional linguistic/cognitive handicap for their speakers because of their alleged utmost morphological simplicity. In earlier work (DeGraff 2001a, 2005a) I have surveyed and provided rebuttals to various versions of this assumption that came to light through the colonial and postcolonial periods. Documenting empirical and theoretical lapses in these claims, I argued that we still lack a rigorous set of criteria for assigning to Creole languages, as a class, an invariant set of structural templates characterized by utmost grammatical simplicity (DeGraff 2001a, 2001b, 2005a; Aboh and DeGraff forthcoming-a, forthcoming-b). As for the “infirmité” that has often been attributed to monolingual Kreyòl speakers in Haiti, it seems more accurately analyzed as a socioeconomic and political “infirmité” cum marginalization that is imposed on monolingual Kreyòl speakers by the exclusionary use of French on the part of intellectuals, politicians, administrators, educators, and so on (see Devonish 2007; Roberts 2008; Migge et al. 2010 for an overview of related issues in Creole-speaking communities). As for the Indigenous languages studied in this volume, for example, Maya, Quechua, Aymara, Tupí-Guaraní, and Tukano, linguists would be hard-pressed to claim that they too are all at the bottom of some stipulated hierarchy of grammatical complexity. Yet in these cases as well, there is an assumption that Amerindian Indigenous languages fall outside the structural range of “normal” languages—a view that Bloomfield (1925) eloquently argues against, in spite of his own exceptionalist attitudes regarding Creoles (DeGraff 2001a, 111n25). The Library of Congress codifies this exceptionalist view of Creole and Indigenous languages by including these local languages in their “PM” subclass “Hyperborean, Indian, and Artificial languages”! Fortunately, we find local educators/ activists, in the Caribbean and elsewhere, who view their native languages as normal languages and who are working for a future in which these languages will, at last, be integrated as media of instruction and communication at all levels, including higher education and public administration, and as instruments for socioeconomic advancement in their communities, on a par with European languages. Quechua and Aymara are cases in point. Like Kreyòl in Haiti, these languages are often perceived as exceptional linguistic “handicaps” by politicians, administrators, intellectuals, educators, and parents. Some Bolivian parents have even threatened to burn pedagogical materials in Indigenous languages after “Intercultural Bilingual Education” became Bolivian law in 1994. These materials were designed to promote Quechua

Michel DeGraff

[ 287 ]

and Aymara speakers’ linguistic and cultural rights as well as to improve pedagogical practice, but parents were concerned that teaching literacy in Indigenous languages would hamper their children’s learning of Spanish, increase discrimination against them, and lessen their chances for upward mobility (Parackahua Arancibia 2011). It is often believed that teaching literacy and content to Quechua-speaking children in Quechua will disadvantage them. Similar beliefs are prevalent about Kreyòl in Haiti. These beliefs persist despite decades of research and multiple UNESCO proclamations supporting instruction in the mother tongue. (See also King and Hornberger 2004; García 2004; Godenzzi 2008; McCarty et al. 2008, for related observations about parents’ rejection of bilingual education programs that include Indigenous languages alongside Spanish. Resistance to bilingual education is also prevalent in the United States, with the striking exception of elite parents who can afford private “international” schools, exchange programs abroad, etc., for their children.) The officialization of bilingual education in Andean countries and related post-1970 laws making Indigenous languages “official” or “national” (see Godenzzi 2008 for a survey) suggest that the ideological climate around these languages has improved since the times when they were considered “animal languages” (King 2004, 337). Yet, as in Haiti, there is a wide chasm in Iberian America between the rhetorics of bilingual education and the practice of exclusion via the use of dominant European languages. This chasm opposes the laws that declare Indigenous and other local languages “official” or “national” to the fact that these languages and their speakers are still being discriminated against on a daily basis (Godenzzi 2008; Dejean 2006; Saint-Fort 2011; Dejean and DeGraff 2013). Andean children born to Quechua-speaking families are among those most likely to fail in school or drop out of school (Laurie and Bonnett 2002; Godenzzi 2008). Thus an age-old colonial tradition is perpetuated, whereby Indigenous populations have, for the most part, been excluded from full participation in the school system and from other avenues of socioeconomic development. There certainly are examples of Indigenous individuals who have become successful professionals, such as Aymara-speaking Evo Morales, current president of Bolivia, but these are among a small minority. (I return to this issue in sections 10 and 11.)

6. Local Languages, Contact Languages, “Standard” Languages . . . Tools for Hegemony? There are also fascinating differences between the fates of Creoles in the Caribbean and the fates of Indigenous languages in Latin America, notwithstanding the sociohistorical parallels between the two geographical

[ 288 ]

chapter eleven

areas. We’ll delve into these differences to sort out the fundamental mechanics of power that have been at work. Such fundamentals should be factored into the design of more effective language policy. Compare, say, the history of Quechua in the Andes, Tupí-Guaraní in the Amazon, Maya in the Yucatán, and Kreyòl in Haiti. Unlike Kreyòl, these three Indigenous languages were vibrant languages and even served as lingua francas among certain Indigenous groups long before the arrival of the European colonists. In Haiti, the Amerindian population and their languages, by and large, quickly vanished through genocide and disease after the Spanish arrived with Columbus in 1492. Today Haiti’s population, unlike that of Iberian America, is almost completely of non-Indigenous stock; the people are primarily of African ancestry, with small percentages of people of European and Levantin ancestry and an even smaller percentage of people of Amerindian ancestry (see Fouchard 1972; Price-Mars 1956; Dubois 2011, for additional details and references). In what is now Iberian America, Quechua-, Tupínamba- and Mayaspeaking communities included many groups whose role and fame as rulers or conquerors predated that of the Europeans, as noted by Hildo do Couto, Alan Durston, M. Kittiya Lee, Denny Moore, and Barbara Pfeiler in this volume. Upon their arrival in the Andes, in the region that is now Brazil, and in the Yucatán, the Europeans initially enlisted these Indigenous languages as tools to advance their imperial expansion and control and to strengthen their own nascent power, including Christian conversion. Take Brazil in the sixteenth century as described by Couto: “What the Portuguese called Língua Brasílica or Língua Geral became so important in the first days of Portuguese colonization that Luís da Grã (1523–1609) imposed its study among the Jesuits of the present State of Pernambuco” (chap. 3, this volume, p. 83; see also Moore, chap. 4, this volume). Ditto regarding Maya in the Yucatán: “Although stigmatized since the Spanish conquest, the Maya language was used by missionaries and plantation owners in their efforts to achieve social and religious domination” (Pfeiler, chap. 8, this volume, p. 205). Similarly, Spanish colonial authorities in the Andes codified their own “standard” version of written Quechua, produced an official description for it, introduced it for study at the university level, and used it as a lingua franca in religious, administrative, and legal matters. They even made fluency in Quechua a requirement for holding clergy posts in Indian parishes (Durston, chap. 9, this volume, p. 234). European colonization may have helped instrumentalize and standardize Quechua and even spread it beyond its former boundaries within the Inca Empire, thus expanding its spread as a regional lingua franca, continuing the Inca language policy that was in force before the arrival of

Michel DeGraff

[ 289 ]

the Europeans. These observations are consistent with those made about Yucatec Maya, Língua Geral, Nheengatu, and other major Indigenous lingua francas in the chapters by Pfeiler, Couto, Lee, and Moore. Yet, as explained by Alan Durston (chap. 9, this volume), the Spanish missionaries’ very codification of Quechua at the onset of Spain’s colonial enterprise was a “reinvention” of the language. This codification did serve the colonizer’s geopolitical expansion: the written Standard Colonial Quechua (SCQ) that they produced through their linguistic “description” was not equivalent to any version of Quechua that already existed. On the contrary, their SCQ corpora, mostly catechisms and liturgical texts, was elaborated toward an “ideal” Quechua based on the attitudes, standards, and agenda of the Spanish. Their main priority was speedy indoctrination of Quechua speakers into a uniform set of a beliefs and practices. Such uniform doctrine was to rest partly on a uniform lingua franca that would be easy to acquire, especially for those having to learn it quickly as a second language, including the Spanish themselves. In Durston’s analysis, key domains of this codified lingua franca (e.g., liturgical terminology) were easiest to acquire for the Spanish since the terms therein were borrowed from Spanish. This “standard” Quechua, with key terms influenced by Spanish, “served to shore up clerical authority” and was “designed as an instrument of control and restriction rather than wide communication.” Indeed there was little communication in SCQ on the part of Indigenous parishioners, who “were only to ‘speak’ SCQ in the form of oral performances of memorized texts” (Durston, chap. 9, this volume, pp. 238–39). These texts in SCQ are an early example of the sort of prescription that is camouflaged as description; what was done with SCQ is similar to the cases of “description as prescription” documented by Bourdieu (1982) and is reminiscent of the contemporary use of European languages for “control and restriction rather than wide communication” in many Creole-speaking communities. In the specific case of Haiti, many a student is reduced to “speaking” French “in the form of oral performances of memorized texts.” Although the two originated from distinct ecologies, both written SCQ in the colonial Andes and French in postcolonial Haiti—in contrast to vernacular Quechua and Kreyòl, respectively—can be analyzed as tools to implement elite closure, giving a “home” advantage to the respective elites of these two countries. Durston (chap. 9, this volume, p. 230) mentions the possible contributions of colonial cities to the differential status of various Quechua varieties as potential lingua francas. He describes Spanish attitudes toward urban versus rural varieties, with preference given to varieties associated

[ 290 ]

chapter eleven

with “urban centers of power, wealth, and learning.” This observation adds fascinating socioeconomic and geographic dimensions to these early explicit efforts at language planning. It brings to mind John Lipski’s observations about the (unplanned) contributions of big cities to the divergence of Latin American varieties of Spanish away from their European counterparts. This pattern of evolution is also similar to the emergence of Hawaiian Creole in S. J. Roberts’s (1998) Creole-formation scenario whereby cities played a key role in the divergence and crystallization away from English. The history of SCQ as documented in this volume thus reveals important traces of subservience to power: as already mentioned, the written variety that is found in the early Spanish missionaries’ liturgical corpora did not exist prior to the Christian expansion in the region, but it reflects the European writers’ own preferences and it was codified to best fit Europe’s mission civilisatrice (Durston, chap. 9, this volume). Couto, Lee, and Moore report similar facts about the adoption of certain Indigenous contact languages (e.g., Língua Geral Paulista and Língua Geral Amazônica a.k.a. Nheengatu) as lingua francas by the Portuguese Jesuit fathers in sixteenth-century colonial Brazil. The Spanish and Portuguese colonists were all too aware of the geopolitical advantages of these Indigenous languages, instead of their native European languages, as lingua francas. Another factor in such choices may have been the Europeans’ belief that the Indigenous populations were cognitively unable to learn European languages (Couto, chap. 3, this volume, p. 83). From this perspective, the best option was for the Europeans themselves to learn the Indigenous contact languages so they could carry out their mission civilisatrice as promptly and broadly as possible. In any case, in the Yucatán, the Andes, and the Amazon of the sixteenth century, Indigenous languages like Maya, Quechua, and the Brasílica had functionality that Spanish and Portuguese didn’t. Witness, for example, the appellations Lengua General for Quechua (Durston, chap. 9, this volume) and Língua Geral Amazônica for the contact language that emerged out of the Brasílica as it became the lingua franca across Indigenous speech communities in the Amazon (Couto, Lee, and Moore, this volume). These languages became “emblem[s] of European occupation” (Couto, chap. 3, this volume, p. 91). Consider, again, Quechua, whose history is richly documented in the chapters by Durston and Lee. Notwithstanding any racial prejudice against Amerindians, what the Spanish who first arrived in the Andes quickly understood was that there was already a vast swath of territory occupied by Quechua speakers that they could reach and colonize through Quechua. The missionaries instrumentalized the language through the production of catechisms, sermons, and sacramental texts in SCQ (Durston, chap. 9,

Michel DeGraff

[ 291 ]

this volume). The colonial administrators used the few individuals who spoke both Quechua and Spanish to interface between them and the Indigenous populations they would “civilize.” Thus, the Spanish adapted to a particular power structure and its metalinguistic hierarchies, using the relevant language varieties to help build a profitable empire under their political control. So, how did the promotion of Quechua come to an end? The story, as documented by Durston (chap. 9, this volume) is complex. But one reason, among others, for the eventual fall of Quechua is that it became one overt symbol of nationalist anticolonial resistance, as in the rebellion led by Túpac Amaru II. Starting in the early eighteenth century, the Spanish started promoting Castilianization, and eventually they banned Quechua linguistic and cultural practices in the public sphere (García 2004; King and Hornberger 2006; Durston, chap. 9, this volume). A related argument can be made about the status of Tupí-Guaraní languages in the Amazon at the time of the Europeans’ arrival. The Europeans’ initial use of Tupí-derived contact languages (e.g., Língua Geral Paulista and Língua Geral Amazônica, a.k.a. Nheengatu) as lingua francas, before the ascendancy of Portuguese, has been taken to reflect the status of Tupí-Guaraní traders as conspicuous explorers, conquerors, and power brokers in the Amazon. Here it seems worth quoting M. Kittiya Lee at length: “The elevation of the Tupi-Guarani languages from the mother tongues of coastal Indians to the unofficial lingua franca of the colony and the subsequent proliferation of grammars and catechisms in the Brasílica underscore the ongoing involvement throughout the sixteenth and seventeenth centuries of native speakers of Tupi-Guarani with religious agents of European colonization” (Lee, chap. 5, this volume, p. 157). Couto offers a related observation: “From the language of a dominated people (the Tupinambás) it [Nheengatu] became an instrument of colonization through the missionaries, colonists, and others. In fact, it became an emblem of European occupation” (chap. 3, this volume, p. 91). These observations remind me of two related remarkable facts from the history of Indigenous languages in North America: (1) the first Bible ever published in the Americas, by John Eliot in 1663, was in the Algonquian language Wampanoag; (2) this first Wampanoag Bible contained sentences like the following: Pomantamwaheuhkon pauwau ‘Do not allow a “witch” to live’. That Wampanoag verse is a translation from Exodus 22:18 (King James Version): “Thou shalt not suffer a witch to live.” In Eliot’s Wampanoag translation, ‘witch’ is translated as pauwau, which nowadays is more familiar to us as the powwow ‘medicine man’ in Wampanoag culture. Thus, Eliot’s translation can be interpreted as an exhortation that the converted Wampanoag Christians go in earnest and kill

[ 292 ]

chapter eleven

the powwow, their Indigenous medicine men. In a related fashion, the colonial Spanish eventually reinterpreted the “morally neutral” Quechua word supay ‘spirit’—a key word in Quechua metaphysics—into the Spanish demonio ‘demon’ (Harrison 1989, 47–48, 136–137).2 Semantic drift has also affected the Haitian word Vodou, etymologically related to the Gbe word for ‘spirit’. For millions of adepts in Africa and the Americas, “Vodou” is an umbrella term for a family of African and Afro-Caribbean religions, but it has now been (mis)used in U.S. popular culture to refer to evil spirits, sorcery, spells, magic, hoaxes, frauds, and the like. Such uses are found in George H. W. Bush’s put-down of Ronald Reagan’s economic policy as “voodoo economics,” in titles of trade books such as Voodoo Science: The Road from Foolishness to Fraud, and even in children’s picture books like Dav Pilkey’s Ricky Ricotta’s Giant Robot vs. Voodoo Vultures of Venus (read in prekindergarten to my then four-year-old son Nuriel). What we see here—in North America, the Andes, and Haiti—is a global phenomenon in which religious concepts from outside Europe are demonized when translated into European languages. The virulent denigration of Haitian Vodou may well be ultimately related to the (perceived) role of Vodou in fomenting slave revolts in lateeighteenth-century colonial Haiti, in the military victory of the Africans against the French Napoleonic army at the turn of the nineteenth century, and in the creation of the first Black Republic in the Americas—an “unthinkable” event, given the world order at the time (Trouillot 1995). Already at the turn of the eighteenth century, Père Labat was warning against what he saw as the joy and lasciviousness and the potential for rebellions in the slaves’ Vodou dances (Labat 1722, 4:466–467). These warnings were repeated at the turn of the nineteenth century by Moreau de Saint-Méry, who thought that Vodou celebrations were a front for disgusting bacchanalian secret meetings where weaker souls could be corrupted and led to sinister ends (1797, 1:49–51). In sections 8 and 9 we look at Vodou in the ecology of Creole formation. The Indigenous languages and cultures of the Caribbean, in contrast to those in North and South America, were not used as instruments of expansion and control by European colonial authorities. In Haiti and most other locations in the Caribbean, the Amerindian languages of the precolonial Indigenous people vanished with the genocide of their speakers, shortly after Columbus’s arrival in 1492. Amerindian languages are, by and large, no longer spoken in the Greater Caribbean, although they still persist as minority and endangered languages in some Central American countries, such as Belize, among the descendants of the Garifuna people deported there from the Caribbean island of St. Vincent in the eighteenth

Michel DeGraff

[ 293 ]

century (Devonish 2007, 193–195, 233–241). But where there is a parallel between the Caribbean and the rest of Latin America is in the politically astute co-opting of the most widespread local languages as lingua francas and as (contested) tools for control.3 For example, the use of Indigenous languages as “lingua franca[s] [and] useful communicative and symbolic tool[s] in different ways for different groups,” as described by King and Hornberger (2006, 184) for Quechua, is similar to the use of Kreyòl during the Haitian Revolution (1791–1803) by both the black leaders fighting for liberty and the proslavery French emissaries of Napoleon Bonaparte. The black leaders, whose troops were made up mostly of African-born soldiers from different ethnic groups, co-opted Kreyòl and certain African languages in order to be understood by all their troops and constituencies. In a related vein, the proclamations distributed by Napoleon Bonaparte’s emissaries to the blacks during the Haitian Revolution were among the first official documents ever to be written in Kreyòl. These proclamations aimed at countering the antislavery and anticolonial spirit that was spreading throughout the colony. They were meant to be read in public squares in order to reach the greatest numbers of people. This use of Kreyòl was thus motivated by the need of the French to maintain their colonial authority over the disfranchised majority fighting for their freedom. Thus, both blacks and whites were using Kreyòl as an indispensable tool for hegemony—on a par with the use of certain Indigenous languages during Europe’s conquest of the Americas and Africa. For instance, Quechua was used in the eighteenth century for its communicative and symbolic power both by the Spanish colonizers and by rebellious Incas fighting against the Spanish (García 2004; King and Hornberger 2006; Durston, chap. 9, this volume). Similar uses of Indigenous languages are also found in the history of Europe’s occupation of Africa (see Samarin 1986 for the history of Lingala as a lingua franca in the Congo). I revisit the early uses of Kreyòl in colonial Haiti (known then as Saint-Domingue) in the section 7.

7. Linguistic Ideology, Language Shift, and Language Endangerment Another language-evolution parallel between the colonial Andes and colonial Haiti relates to the fact that through its expansion in and beyond the (former) Inca Empire, Quechua, even when it became dominated by Spanish, also dominated, then caused the extinction of, many languages with fewer speakers, such as the Jaqaru, Puquina, and Uro-Chipaya languages (King and Hornberger 2006, 185; also see Durston, chap. 9; and Couto, chap. 3, this volume). Similarly, Kreyòl in Saint-Domingue, al-

[ 294 ]

chapter eleven

though it was less prestigious than French, became by far the most widely spoken language, while the ancestral languages of the Africans (the least prestigious in the colonial milieu) eventually vanished from the ecology of language contact there. In fact, according to Mufwene (2008, chap. 11), Creole formation in the Caribbean and the Indian Ocean can be viewed as a concomitant of language shift among the enslaved Africans. Mufwene relates the speed of loss of ancestral languages to various factors, including the competing socioeconomic values of the languages in contact. According to him, these differing values are among the reasons why the loss of ancestral languages proceeded faster among the Africans in the Caribbean than among the Amerindians in the Andes and in the Amazon. In turn, the shift to Kreyòl on the part of Africans and their descendants in eighteenth- and nineteenth-century Haiti is similar to the shift in subSaharan Africa to urban vernaculars like Sango, Swahili, Lingala, and Kituba. As Mufwene reminds us, the shift to urban vernaculars, with the concomitant loss of the demographically or socioeconomically “smaller” languages in the contact situation, also occurred in the history of Europe, North America, and Australia (chap. 1, this volume, pp. 14–15). As a matter of fact, according to Christopher Ball and Hildo do Couto (chaps. 10 and 3, this volume), Tupí-Guaraní and its contact-language descendants and other major vernaculars such as Tukano can also be viewed as “threats” to smaller Amerindian languages in the Amazon. What is also documented, so insightfully, in Ball’s chapter—which deserves an expository detour—is that the adoption of Portuguese and Tukano by (former) speakers of, say, Arapaço and Piratapuya is embedded in a set of concerted practices aimed at accumulating power or at preserving key aspects of Arapaço and Piratapuya ethnicity and ancestral cultures. The latter include deeply rooted senses of identity and gender- and marriage-related practices, some of which are intricately related to communal attitudes and beliefs about language purity, language mixing, language transmission, and so on. In one such practice, the male Arapaço maintained and adapted his plurilingual exogamous cultural identity by shifting to plurilingualism in Tukano and Portuguese (through forced relocation) instead of plurilingualism in Arapaço and some other language (through exogamy). As for the Piratapuya, their stated reason for shifting to Tukano and Portuguese is a culturally rooted aversion to language “corruption” through language mixing in mission schools—language mixing outside of the traditional bounds of exogamy. In this case, “Indigenous language mixing, outside of the prescribed exogamic and hierarchical constraints, [is] the reason for intergenerational language shift.” One Piratapuya father felt that “he had better teach his children Tukano well than a version of Piratapuya corrupted by mixing in the school environment.” Piratapuya’s

Michel DeGraff

[ 295 ]

cultural subjectivity puts high value on the correct use of each language in a speaker’s plurilingual repertoire. This father was thus being culturally Piratapuya even as he was teaching Tukano, instead of Piratapuya, to his children. In such cases, “social transformation and language change, including language shift, do not necessarily entail the loss of culturally specific norms of subjectivity” (Ball, chap. 10, this volume, pp. 255–57). Ball also documents the critical role played by gender roles in these processes of social transformation and language change, with women helping to maintain both plurilingualism (at the individual level) and multilingualism (at the communal level) via virilocal exogamy (i.e., the practice of local men marrying female outsiders speaking nonlocal languages). These female outsiders perform their exogamous cultural practices in innovative ways when they are married to Portuguese in big cities and when they produce racially mixed children who inherit, from their fathers, Portuguese as their “mother” tongue, alongside a (colonial? quasi? neo?) Portuguese identity. Yet, cultural inheritance is not straightforward. These Amerindian women with Portuguese husbands do not strictly follow their cultures’ dictates: they contradict patrilineal patterns of cultural transmission somewhat “by investing effort in having [their] children incorporated as members of [the mothers’] paternal sib, both by bringing [the mothers’] own fathers or other paternal relatives to town to reside with their nuclear family and by enacting sib-specific naming rituals to confirm mixed children’s sib membership” (Ball, chap. 10, this volume, p. 259). These women are thus critical agents of cultural and linguistic pluralism in their households. This is yet another way in which cities play a decisive role in patterns of language change, this time through modification of Indigenous women’s gender-related cultural practices. These women are still exogamous, in keeping with their culture: they have moved out to big cities, where they marry outside their groups. Yet they make choices that contravene traditional patriarchal patterns by enabling their children to maintain their Indigenous identity through European-inflected “hiccups in transmission” (Ball, chap. 10, this volume, p. 260) that, in effect, produce plurilingual Portuguese-cum-Amerindian households, thus new ecologies for linguistic and cultural contact. To end this fascinating expository detour about the role of cultural subjectivities in language shift in the Amazon, let’s note that the role of exogamous women in language contact and language shift in the Amazon as described by Ball is somewhat similar to that of the mestizos (children of Amerindian women and European men) discussed by Couto (chap. 3, this volume); they were an important vector in the spread of Portuguese in colonial Brazil.

[ 296 ]

chapter eleven

We’ve now compared, on the one hand, the threats that major Indigenous languages have posed to smaller languages in the Andes and the Amazon to, on the other hand, those threats that Caribbean Creoles posed and carried out, with respect to a variety of Niger-Congo languages in the colonial Caribbean. The locally born (or “Creole”) descendants of the enslaved Niger-Congo speakers did eventually switch to the next available Creole or European language for the familiar ecological reasons of economic and political power (Mufwene 2008). In Guyana, it is a Creole language (in this case, Guyanese Creole, a.k.a. Creolese) that, after superseding its African ancestor languages, has now become a threat to local Indigenous languages such as the Arawakan language Lokono; in Belize, it’s Belizean Creole that poses a threat to Garifuna, although the latter is not truly “Indigenous” to Belize (see note 1), having been transported from St. Vincent (Devonish 2007, 193–195, 233–241). One notable difference between Creoles in the Caribbean and the Indigenous contact languages studied in this book is that the latter predated the arrival of the Europeans, whereas the former came into existence as the initial cohorts of Africans in the Caribbean colonies were progressively shifting to the locally available varieties of European languages. Another key difference has to do with the available evidence and the methodological approaches in studies of language evolution in colonial Iberian America and in the colonial Caribbean. This is the topic of section 8, where I take inspiration from Christopher Ball’s chapter as I delve into a timid and tentative exercise in linguistic anthropology to try to shed light on one understudied factor in language change in the Caribbean, namely, the role of linguistic ideologies in language shift.

8. Cultural Subjectivities and Language Shift among the “Incas” of Haiti One sort of research that is sorely missing in work on language evolution in the Caribbean is detailed ethnographic study on a par with Ball’s investigation of language shift by Amerindians in the Amazon. For example, research on linguistic ideologies among Africans in the colonial New World would shed light on the ways in which language shift and language change in the Caribbean may have been influenced by the Africans’ own cultural subjectivities. In lieu of full-fledged ethnographic studies of Africans in the colonial Caribbean, what we do have are extensive reports of Europeans’ attitudes vis-à-vis the Indigenous, African, and Creole languages spoken in the colonial Caribbean (see DeGraff 2001a, 2005a, for overviews and references). In the case of Haiti—itself an Amerindian name meaning “land

Michel DeGraff

[ 297 ]

of mountains”—European observers considered the Creole-speaking Africans as an improvement over those who spoke only African languages. One such observer is Saint-Méry (1797), who presents a hierarchy of Creole varieties based on their proximity to French: the best Creole, in his opinion, is the variety that is closest to French and furthest away from the influence of African languages. This is the Creole variety that creolists characterize as “acrolectal.” Saint-Méry saw the non-Creole people (i.e., those born outside the Caribbean, whether African- or European-born) at a linguistic disadvantage: the Creoles spoke native Creole varieties that were superior to the nonnative varieties, especially those spoken by the African-born slaves, the “Bossals.” According to him, “this [Creole] language . . . is often unintelligible when spoken by an old African; one speaks it all the more fluently if one learns it at a younger age” (1797, 1:64). This linguistic advantage, based on an accident of birth, conferred upon the numerical minority of Creole blacks an allure of superiority over the African-born majority. As Saint-Méry assessed it, quite arithmetically, “for all tasks, it is the Creole slaves that are preferred; their worth is always a quarter more than that of the Africans” (1797, 1:40). SaintMéry’s rationale for this comparative advantage reminds us of Mello’s description of the perceived superiority of the Ladino blacks over the Boçais blacks. In Saint-Méry’s terms, “Creole Blacks are born with physical and moral qualities that truly give them the right to be superior over Blacks that have been brought from Africa”; “domesticity has embellished the [black] species” (1:39). This allure of superiority may have played a key role in the relatively rapid spread of the linguistic features associated with the speech of Creole blacks (DeGraff 2002, 378–382). Saint-Méry created race- and ethnicity-based hierarchies that applied to both the mixed race (the mulattoes) and the African populations of Saint-Domingue. The etymology of mulatto (from mule) is related to the fact that the hybrid offspring of mixed-race couples was considered as defective as the sterile mule. On the ethnographic and linguistic front, Saint-Méry took the Congos to be docile, joyful, easiest to assimilate to the colonial milieu, good learners of Kreyòl, and well suited for domestic life and skilled trades, whereas the Aradas were considered the worst language learners among the Africans (1797, 1:29–32). Saint-Méry’s hierarchies also include spectacularly biased ethnographic details, as in his description of the Mondongues (1797, 1:33): hideous and feared by all and, worse yet, aficionados of human flesh, especially that of babies, including their own! Unfortunately there is no counterpoint from the African side about their own metalinguistic attitudes and ideologies and for their own fashioning of identities.4 As I look for some window—any window—on the

[ 298 ]

chapter eleven

cultural subjectivities of Africans in the colonial Americas, I will now try to extrapolate from popular sayings in contemporary HC, with the provisional assumption that the metalinguistic attitudes and beliefs that are revealed through these sayings may go back to colonial times or ancestral cultural practices. Better yet, I’ll examine early reports of metalinguistic attitudes among prominent blacks in the colonial Caribbean. Let’s start with the Haitian saying that contrasts Kreyòl as lang rasin ‘root/ancestral language’ with French as lang achte ‘purchased language’ (Valdman 1984, 82; DeGraff 2005a, 570). The Kreyòl word rasin (literally ‘roots’) is often used as a modifier for terms in the Vodou religious domain, as in lwa rasin ‘ancestral spirits’ and mizik rasin ‘ancestral rhythms’. Rasin endearingly evokes (imagined notions of) family heritage with “roots” that go back generations, all the way back to Ginen ‘Guinea’ (i.e., the mythical Africa of ancestral origins). The lwa rasin are seamlessly transmitted from parents to children through home and communal practices. In Vodou’s cosmology, the lwa rasin are inherited from the ancestors in Ginen or bequeathed by the spirits of eighteenth-century black leaders of the Haitian Revolution such as Jean-Jacques Dessalines (Dayan 1995). Dessalines was a former slave who fought and won against the French’s Napoleonic proslavery army in the battles that led to the independence of Haiti in 1804. It was the only nation created out of a slave revolt—and the first nation to break free of the Europeans’ colonial empires in Latin America. Ginen is a mythical symbol for those who practice Vodou. In the context of Amazonian cultural subjectivities, Ginen brings to mind the “upriver” original location of the founding ancestor Unurato, who looms so large in the Aparaço myth of creation (Ball, chap. 10, this volume, p. 254). In Vodou’s own mythology, the lwa rasin ‘ancestral spirits’, also known as lwa fanmi ‘family spirits’ or lwa eritaj ‘heritage spirits’, are considered benevolent and are trusted to give lasting strength and to promote individual and communal wholesomeness. In contrast, the lwa achte ‘purchased spirits’, also known as lwa djab ‘evil spirits’, are not part of one’s ancestral lineage. They are costly, foreign, and satanic “manmade” spirits that impose heavy demands on those who acquire them—often for malevolent purposes. The lwa achte may be efficacious for certain selfish purposes, but they are immoral and untrustworthy spirits that eventually undermine the individual and the community’s well-being. There are three other, even more popular, Haitian sayings that attribute positive communal and communicative values to Kreyòl as compared to the limits and disruptive power of French. Consider the saying Kreyòl pale, kreyòl konprann. Its literal, and perhaps original, meaning is ‘Creoles speak, Creoles understand’ which can be interpreted as “We

Michel DeGraff

[ 299 ]

Creoles understand one another.” Nowadays it is often taken to mean ‘Creole spoken, Creole understood’, that is, ‘the Creole language facilitates mutual understanding’. This latter interpretation is related to two other sayings: Sispann pale franse (literally, ‘Stop speaking French’) ‘stop obfuscating’ and the aforementioned Pale franse pa vle di lespri ‘Being able to speak French doesn’t mean that one is intelligent’. These four Kreyòl sayings illustrate linguistic ideologies related to the contemporary comparative values of Kreyòl versus French. They make one wonder about the enslaved Africans’ attitudes and beliefs about French as they were learning the latter as a second language and thus seeding the varieties identified today as Kreyòl—and as they were fashioning early forms of Vodou out of their ancestral African religious practices in negotiation with the dogmas of Christianity to which they were exposed. These negotiations between African religions and Christianity may have started in Africa as early as the fifteenth century, long before the Africans arrived in the Americas (Heywood 2002; Heywood and Thornton 2007). Both Kreyòl and Vodou helped create a community in Saint-Domingue, where there was none among the Creole and African-born blacks with diverse ethnic origins and ancestral languages. The Vodou blood-oath ceremony of Bwa Kayiman in 1791 has often been described as one of the most important symbolic overtures to the Haitian Revolution (e.g., Fouchard 1981; Fick 1990; Dayan 1995; Dubois 2011; BeauvoirDominique 2011; but see Geggus 1991, 2002). Kreyòl as lingua franca was certainly a key instrument for communication on the battlefronts on all sides, and necessarily so at a time when the multilingual African-born population was the majority in the colony. Yet there seems to have been a certain ambivalence vis-à-vis both Kreyòl as colonial lingua franca and the Creole blacks themselves even though—or perhaps because— the latter, especially the free Creoles, generally held the most social and economic power among the black population—a power that included a certain amount of control over the Bossal slaves (some of the free Creole blacks even owned slaves). This ambivalence around Kreyòl can be deduced from three sources, among others: (1) reports about the language attitudes of two of the most famous revolutionary leaders: FrançoisDominique Toussaint Louverture (1743–1803) and Jean-Jacques Dessalines (1758–1806), former slaves who spearheaded the war against the French (see, e.g., Jenson 2011 for a comprehensive summary); (2) Vodou songs warning about the ambivalent allegiances and religious beliefs of the Creole blacks (see, e.g., Beauvoir 2009); and (3) descriptions of class-based ideological, political, and military struggles opposing Creole and African-born blacks, including reports of persecution of Vodou practitioners by Creole leaders (see, e.g., Fouchard 1953; Fick

[ 300 ]

chapter eleven

1990; Thornton 1993; Trouillot 1995; Beauvoir and Dominique 2003; Ramsey 2011). Here we’ll take a brief look at reports about language attitudes on the part of revolutionary leaders in Saint-Domingue. Toussaint Louverture, for one, would sometimes berate those who addressed him in Kreyòl. Yet that’s the language that he himself would sometimes use to scold or persuade (e.g., Descourtilz 1809, 3:245–246, 251; Saint-Louis 2006, 160; Fick 1990, 116; Jenson 2011, 65). He is also “said to have spoken fluently the language of his ‘Arada’ (Ewe-Fon) father . . . and to have enjoyed speaking it with other slaves of his father’s ethnic group” (Geggus 2002, 16). Perhaps he perceived, as many still do in contemporary Haiti, that those in positions of power must speak, and be spoken to, in French in order for them to extract respect and maintain authority. On occasion, he would even affect aristocratic “old régime ideology” through, say, the use of Latin (Saint-Louis 2006, 160). Yet Toussaint himself was not perfectly fluent in French: he spoke what Haitians today would call a français marron ‘Brown French’ (Jenson 2011, 65), that is, Kreyòl-influenced French of the sort that is often ridiculed by middle- and upper-class Haitians. As for Dessalines, he was reported to berate those who spoke French to him and to exhort them to speak their lang rasin or, in Dessalines’s own words (according to Descourtilz 1809, 3:281), the langue à vous ‘your own language’. Although there are conflicting reports as to what languages Dessalines actually spoke (Dayan 1995, 21), some historians believe that he didn’t speak French (Geggus 2002, 293) and that he enjoyed “coarse Creole” (H. Trouillot 1962, 90). It has even been reported that Dessalines would kill those who would answer him in French instead of Kreyòl (Dayan 1995, 22) and that he would use differences in the pronunciation of certain Kreyòl words in order to distinguish between light-skinned Haitians with African ancestors, on the one hand, and, on the other hand, French settlers with similar complexion—only the latter were targeted for execution after Haiti’s independence in 1804, and Kreyòl was used as a source of shibboleths to single out the French (H. Trouillot 1962, 23). According to Descourtilz (1809, 3:281), Dessalines indeed hated the French language, especially after the arrival in 1802 of Napoleonic troops who came to try to suppress the revolution and reestablish slavery in the colony in revolt. In the new Haitian state’s first official proclamations in 1804, Dessalines and his secretary Boisrond-Tonnerre clearly expressed their government’s beliefs that the French colonists had used their “horrible language” against the blacks as an instrument of dehumanization and to hide France’s crimes against humanity (Daut 2009, 48; Jenson 2011, 134). This is what Dessalines called “the deceitful eloquence of the

Michel DeGraff

[ 301 ]

Proclamations of [French government] agents” (cited in Jenson 2011, 134; also see Casimir 2011, 34n15). There is another fascinating philological connection with one of the main themes of the volume—about the fate of Indigenous peoples and their languages in Latin America. For some time during the revolutionary war in Saint-Domingue, Dessalines called his forces “Army of Incas” and “children of the Sun” (Madiou 1847, 2:357, 421). In the Declaration of Independence dated January 1, 1804, and signed by generals of the “Indigenous Army,” Dessalines rejected the colonial French name Saint-Domingue and baptized the independent nation with the Amerindian name Haiti to commemorate the island’s native Amerindians; then in April 1804, Dessalines proclaimed that that he had “avenged America.” (See Jenson 2011 for sources, original passages, and their translation into English.) Yet it is in French that the new republic’s first proclamations were written, even though the Haitian Revolution had been led and won in the field mostly in Kreyòl. In sections 9 and 10 I return to some of the reasons for the use of French in these official proclamations of Dessalines.5

9. The Colonial Linguistic Market and Language Shift: Haiti Again as a Case Study of Normal Language Change Recall that Vodou may have played an important role in bonding the blacks together against the French in battle, as in the aforementioned Bwa Kayiman ceremony in 1791. This Haitian religion has its ancestral roots mostly in Africa but is infused with Christian elements (Hurbon 1988), somewhat on a par with indigenized varieties of Christianity in Africa from the fifteenth century onward (Heywood 2002; Heywood and Thornton 2007). Recall that the word Vodou itself is from the Gbe word vodũ ‘spirit.’ The Gbe-Kreyòl etymological and associated cultural links are additional reminders that “simple equations of the loss of language with the loss of culture are often misguided” (Ball, chap. 10, this volume, p. 258). Then we have the fact that Kreyòl emerged as lang rasin in the Caribbean, far away from Ginen, with massive input from French, and then supplanted the African ancestral languages, even as key concepts of West and Central African religious beliefs and rituals found new expression in Kreyòl (cf., e.g., Hurbon 1988; Heywood 2002; Montilus 2006; Heywood and Thornton 2007; Beauvoir 2009; Beauvoir-Dominique 2011). Such facts remind us that language loss, as in Amazonia, studied by Ball, can indeed happen without total loss of identity—or of the perceived ‘roots’ thereof. Indeed Kreyòl became the de facto linguistic badge of identity of the new Haitian nation from the nineteenth century onward: it became its new lang rasin with newly grown Creole ‘roots’.

[ 302 ]

chapter eleven

From such a perspective, the saying Franse se lang achte ‘French is a purchased language’ reveals a sobering view of Haiti’s linguistic market, in which French, like Spanish and Portuguese in Iberian America, has the greatest socioeconomic power—at the greatest cost. The comparison of French to a lwa achte suggests that it often comes at great sacrifice and is viewed as a factor of malevolence for the community. Indeed, the reality is that, for the vast majority of Haitians, French as lang achte is strictly outside the ancestral community culture. As such it is learned by the select few who can afford the better schools or who grow up in the very few homes where French is fluently spoken—no more than 10 percent according to certain estimates (Dejean 2006). The sayings lwa achte and lang achte seem to express a mistrust of entities (whether spiritual or linguistic) perceived as relatively “foreign” and an attachment to cultural items that can be perceived as related to ancestral values (such as the lwa rasin from Ginen and the Kreyòl language as lang rasin) and that can smoothly become part of a Haitian “socioculturally-constructed personhood,” in the spirit of Ball’s chapter. Notwithstanding these popular statements of attachment to ancestral Africa, the structures of Kreyòl itself, even as lang rasin, are not exclusively of African ancestral roots. Instead Kreyòl structures can be reasonably considered the normal outcome of language change affecting seventeenth- and eighteenth-century varieties of French as enslaved Africans with their own ancestral languages were shifting to French in the Caribbean (DeGraff 2002, 2005b, 2009). This scenario is similar to the history of early French as the outcome of language contact and language shift. Indeed, the seeds of early French were sown as Latin was being appropriated by speakers of Celtic and Germanic languages (Mufwene 2008, chap. 3). Among HC structures, we find contributions both from French (in greater part) and from African ancestral languages, alongside innovations (Fattier 1998; DeGraff 2002, 2005b, 2009; Aboh and DeGraff forthcoming-a, forthcoming-b). These language-change patterns are reminiscent of the sort documented in the history of, for instance, Nheengatu by Denny Moore and Latin American Spanish varieties by John Lipski. In terms of linguistic ideology, and as Max Beauvoir carefully explains (pers. comm., December 2011), what makes Kreyòl a lang rasin is the fact that it’s spoken by all Haitians as their mother tongue, whereas French as lang achte is learned only by a small minority, most of whom struggle to learn it in school as a relatively distant second language. Through the detailed studies by historians, sociologists, and anthropologists, we have learned a great deal about the origins and cultures of the various groups of Africans who were forcibly taken to the Americas and whose ancestral languages contributed to the emergence of Kreyòl

Michel DeGraff

[ 303 ]

as lang rasin via “hiccups in transmission” of French. “Hiccups in transmission” are also characteristic of the history of Spanish and Portuguese in the Americas (Ball, chap. 10, this volume, p. 260), as well as of the earlier emergence of these Romance languages from Latin in Europe: all these developments instantiate contact-induced language change. In the case of Saint-Domingue, we even have archival reports, including firsthand reports, about what black leaders such as François-Dominique Toussaint Louverture and Jean-Jacques Dessalines allegedly said and believed about Kreyòl and French in Saint-Domingue (Jenson 2011). My hope is that as these studies deepen, we will learn more about the sociocultural histories of the Africans’ attitudes and beliefs about geography, migration, language transmission, language mixing and language shift, language purity, intermarriage patterns, and so on—as we have learned in Ball’s chapter about the Eastern Tukanoan. Here are some of the questions we would love to ask of future ethnographic studies of the Africans in the colonial Caribbean: Were the enslaved Africans more like the Tukanoan peoples in the Vaupés or more like those in the Upper Xingu? Or were they more like the Arapaço or the Piratapuya? How did they come to terms with their own “dissociation of language from ancestral place”? How did the Africans’ diverse cultures each accommodate the pressures toward language mixing and language shift? In what ways did the biracial children of African women and European men become vectors for cultural and linguistic shift versus maintenance? Was this in any way similar to the situation of Amerindian exogamous women mating with Portuguese men in certain big Iberian American cities and renegotiating their gender roles vis-à-vis patriarchal patterns of cultural and linguistic affiliation?6 To what extent, if at all, did Africans from various ethnic groups wish for their children to maintain their ancestral languages and/or learn Kreyòl or French? How did they analyze the power and prestige being ascribed to French- and Kreyòl-speaking blacks (the “Ladinos” of the Caribbean)? Or were they purists who, like the Piratapuya speakers in the Amazon (Ball, chap. 10, this volume, pp. 256–57), did not want their children to speak varieties of their ancestral languages that had been “corrupted by mixing”? Or, given their respective cultural subjectivities, did (some of) these Africans want their children to be plurilingual with, at least, mastery of the language(s) that would give them power as cultural and linguistic brokers? How did these cultural factors filter various sorts of potential substrate influence (from, e.g., Fongbe or Kikongo) into the emerging Creole? For example, if women played any major role in language contact, then the comparison of sex ratios across ethnicities would constitute a related factor. Such studies may thus help us understand why certain African lan-

[ 304 ]

chapter eleven

guages had greater influence than others on the structures of Caribbean Creoles (see, e.g., Geggus 1989 for some relevant data and observations about gender demographics in Saint-Domingue). These are some of the many fascinating questions that the present volume invites us to ask about the colonial Caribbean and the ecology thereof. What we know, for now, is that by the late eighteenth century, Kreyòl in Haiti—as a language genealogically descended from French with various influences from Niger-Congo languages—had accumulated enough prestige and socioeconomic and political clout to become the generalized target of language shift. Thus it led to the eventual disappearance of the ancestral African languages and became the new nation’s sole lang rasin from then until today.

10. Managing Power and Managing Languages in the (Post)colonial Americas In the colonial Andes, the language-policy switch from the use of Indigenous languages to Spanish as lingua franca occurred in the seventeenth and eighteenth centuries, with concomitant efforts to suppress the rise of Indigenous languages (Durston, chap. 9, this volume; García 2004). The goal of “Castilianization” was to make the Indigenous languages “disappear” by teaching Spanish to the Indigenous peoples. Yet these efforts were ambivalent, since Indigenous laborers without formal education were more helpful to the colonists than their educated counterparts. However, after the rebellion of Túpac Amaru II in the late eighteenth century and his use of Quechua as an instrument and symbol of Indigenous nationalism and resistance against the colonial regime, the Spanish colonial administrator José Antonio de Areche found it necessary to promulgate decrees explicitly banning the Quechua language and culture from the public arena, further strengthening the Hispanization mandate of the Bourbon Reforms started earlier in the eighteenth century.7 At this point, Quechua was viewed more as a liability than an asset for Spanish domination. This ban lasted for some two hundred years, until the language reform of the latter half of the twentieth century (García 2004). In a similar fashion, by the mid-eighteenth century, Língua Geral Amazônica (LGA) was given second-class status when a Portuguese colonial law (the Regimento das Missões) “aimed at promoting Portuguese and eliminating LGA. One reason for this was to claim more territory by showing the presence of Portuguese speakers there” (Moore, chap. 4, this volume, p. 119). In the Yucatán, “usage of Maya as a legal language was discontinued in the nineteenth century and increased importance was given to reading and writing in Spanish” (Pfeiler, chap. 8, this volume, p. 206).

Michel DeGraff

[ 305 ]

These Indigenous lingua francas have survived the various sorts of bans imposed by the colonial Europeans and their postcolonial descendants. But not all Indigenous languages have survived with equal vitality. Quechua and Maya now fare much better than Nheengatu and most other Indigenous languages in Latin America. Today Nheengatu is spoken by a mere eight thousand people (Lewis 2009), while Quechua is spoken by some 8 to 12 million (Hornberger and Coronel-Molina 2004). With about the same number of speakers as Kreyòl in Haiti, Quechua is thus still very much alive, and it is the majority language in the rural Andes, although it must be noted that Quechua is actually composed of a family of languages with substantial differences (Durston, chap. 9, this volume). Along with a variety of Indigenous languages, Quechua has even become an “official” language (somewhat nominally), with various provisions, in the constitutions of Peru, Bolivia, Ecuador, and Colombia. So has Maya been recognized as a “national” language. In Brazil, it is now the law that “regular elementary education shall be given in the Portuguese language, the Indian communities also being ensured the use of their native languages and specific learning procedures” (Brazil’s 1988 constitution with amendments, online at www.servat.unibe.ch/icl/ br00000_.html). Yet, as in Haiti, the de facto status of these Indigenous languages contrasts with their de jure status: in practice, all these local languages have been relegated to second-class status. Thus, Indigenous parents and educators in Iberian America, like their Caribbean Creole-speaking counterparts, often think of their native languages as burdens to get rid of, seeking European languages as the only ones that will give them the tools they need for socioeconomic success. After national reforms in 1994 promoting bilingual education, rural Andean communities started receiving urban teachers who would come for a week or a month at a time to reside at the school and teach, returning home on weekends. One such teacher, appointed as the new director of a school during the education reform in Bolivia, was denied entry by parents because she had proposed to teach their children to read and write in their native Quechua in addition to Spanish. Only after the passage of time was she able to win them over (Parackahua Arancibia 2011). These parents’ refusal of bilingual programs and their desire for the European language as the exclusive medium of instruction reflect the anti-Indigenous discrimination they have experienced and wish to avoid for their children. This story and many others like it throughout the Americas bring to mind Pecola, the black girl in Toni Morrison’s story The Bluest Eye, who hates her blackness and wants blue eyes because of the signs she sees all around her that promote whiteness over blackness. Indeed, European languages throughout

[ 306 ]

chapter eleven

Latin America and the Caribbean continue to be invested with symbolic capital—linguistic “bluest eye”—that best serves the interest of the ruling class. Another case in point is that of Peruvian children, for whom literacy “is not only equated with learning how to read; it is equated with learning how to read in Spanish, and disassociated almost entirely from Indigenous languages” (García 2004, 359). This situation is a negation of the precolonial Inca world order, in which Quechua was the de facto official language. Now, parents’ insistence upon Spanish as the language of instruction makes sense to them in light of the fact that it is fluency in Spanish, not Quechua (and not Aymara, Maya, and so on), that opens up opportunities for higher education and better jobs, a point well articulated by Barbara Pfeiler (chap. 8, this volume). The comparison between Amerindian languages in Iberian America and Kreyòl in Haiti brings up an ironic twist: it may seem that the victorious Túpac Amaru of Haiti—namely, Jean-Jacques Dessalines, the first president of independent Haiti—chose to relegate Kreyòl to second-class status as he used French, instead of Kreyòl, for the newborn country’s first official declarations. However, he too, very much like Túpac Amaru and his Spanish adversaries, was using both language and metalinguistic knowledge for political purposes. In Dessalines’s resistance against European imperialism, his written proclamations in French were directed not toward his (mostly illiterate) Haitian compatriots, whom he verbally addressed in Kreyòl, but toward audiences overseas, including (1) France, which he was warning never again to try to take over Haiti, and (2) other foreign audiences to whom he wanted to promote the new country’s historical achievements against colonization and slavery, as well as Haiti’s viability as a trade partner and otherwise. (See more detailed arguments in Jenson 2011.) Yet Dessalines’s and subsequent Haitian governments’ exclusive use of French in the official business of the new nation and in its school system doubled as an instrument of class differentiation, helping to keep power in the hands of the few for control over the disfranchised majority (e.g., Hoffmann 1984, 57–63; DeGraff 2005a; Dejean 2006; Saint-Fort 2011; DeGraff 2013a, 2013b; Dejean and DeGraff 2013). Ever since then, speaking French—Haiti’s own linguistic “Bluest Eye” or a linguistic symptom of “bovarysme” in Price-Mars’s (1956, 136) terminology—has been an instrument for, and a reflection of, power even as it’s widely accepted that Pale franse pa vle di lespri (‘Being able to speak French doesn’t mean that one is intelligent’). The promotion of, and the production of knowledge about, certain languages and their speakers in the history of the Americas seem to have always been conditioned by the ways in which capital, goods, and other

Michel DeGraff

[ 307 ]

limited resources are allocated to competing groups. Those of us who are struggling to promote native-language instruction in our communities— whether in the Americas, Africa, Australia, or elsewhere—may stand a better chance of making inroads with our projects if we pay attention to “positive initiatives” that can effectively shift the allocation of desirable resources in order to match the promotion of the use of local languages in schools. This is a tall order!

11. “Positive Initiatives” against Language-Based Discrimination in the Americas “Positive initiatives” in favor of Amerindian languages include the creation in 2001 of the Institute for Development of Mayan Culture in the Yucatán state and the passing in 2003 of the federal law for the linguistic rights for the Indigenous people (Pfeiler, chap. 8, this volume, p. 220). Similar initiatives have been reported by Godenzzi (2008) and Morales (2010) for Bolivia and Peru. These and related efforts all aim at the “recognition and protection of the linguistic, individual and collective rights of indigenous communities, and the promotion and development of indigenous languages” (the 2003 federal law for the linguistic rights of Indigenous populations, as quoted by Pfeiler, chap. 8, this volume, p. 220). Similar human-rights efforts, on the part of both the state and grassroots organizations, are taking place throughout Indigenous communities in Latin America. In Bolivia the most recent constitution (2009) makes all thirty-six Indigenous languages “official” alongside Spanish; it encourages the usage, protection, development, and study of Indigenous languages; and it makes it mandatory for government employees to demonstrate proficiency in Spanish and at least one Indigenous language (Morales 2010; Political Database of the Americas 2011). One effort to connect linguistics with grassroots activism in the rural Andes—one that I am personally familiar with through colleagues involved therein—is the Proyecto Yachay q’ipikuna, supporting Quechua as language of instruction in Bolivia’s and Peru’s rural highland (Chacón et al. 2011). The papers in Hinton and Hale (2001), King and Hornberger (2004), and Haboud and Ostler (2011) provide a comprehensive overview of similar projects throughout the world and their potential benefits and limitations. I now put on my activist-educator hat to describe some of my own work, in collaboration with colleagues at MIT and in Haiti, on Kreyòl as a language of instruction in Haiti toward facilitating technology-enhanced active learning in science and mathematics in high schools and universities. (In April 2013, this work received the support of the Haitian state through an agreement between our MIT-Haiti Initiative and Haiti’s Min-

[ 308 ]

chapter eleven

istry of Education.) One lesson I have learned in the course of this work is that laws from the state are often insufficient to stop language-based discrimination and to effectively promote local languages (DeGraff 2010; DeGraff and Ulysse 2011; DeGraff 2013a, 2013b). Consider, for example, two now familiar facts, which stand out in the history of language evolution in the Americas: (1) Kreyòl is the one single language that is spoken by all in Haiti; (2) constitutionally, Kreyòl has been an official language in Haiti, on a par with French, since 1987. Yet, in practice, French is still being promoted as superior to, or more useful than, Kreyòl in schools and universities, courts, public offices, the written press, and so on. In many schools, speaking Kreyòl in the classroom often triggers costly and often shameful penalties, physical or otherwise. This stigma, which is often enforced from the highest echelons of the schools’ administration, seems the toughest obstacle to projects like the MIT-Haiti Initiative that promote deep and active learning through the use of the one language that is most familiar to most Haitian students. Yet schools usually offer classes on Kreyòl composition, because the sixth-grade official exams include one exam on Kreyòl. But some of these same schools forbid the use of Kreyòl in their other classes! To date, the majority of the state’s official exams (to enter secondary school and university), the better paying and prestigious jobs, the court systems, and so on, still function, in effect, for the exclusive advantage of those who speak French—that is, those who, by privilege of birth, have grown up in homes where French is spoken, usually alongside Kreyòl, and those with enough luck or talent to learn French in school. Although the 1987 constitution requires that every law or decree and all other state communications be published in both French and Kreyòl, the vast majority of state documents are still published in French only. A recently proposed amendment to the 1987 constitution was written in French only, even though the constitution was promulgated in two official versions, French and Kreyòl. It’s only in July 2013 that for the first time the Parliament voted a law in Kreyòl only: the law to establish a Kreyòl Academy, as provided for by the 1987 constitution. So there are encouraging signs of progress ahead. But meanwhile the vast majority of the population (the monolingual Kreyòl speakers) cannot understand most of the laws that they must abide by. Our various projects on the promotion of mother tongues as languages of instruction—whether in the Yucatán, the Andes, the Greater Amazonia area, or the Caribbean—are not just linguistic educational projects. They are, above all, political and socioeconomic projects for social justice (Martínez Cobo 1987; McCarty et al. 2008). Indeed, in order for stigmatized local languages to be effectively promoted as languages of instruction in the classroom, there must be some clear and relatively di-

Michel DeGraff

[ 309 ]

rect socioeconomic advantages that result from the use of these native languages outside the classroom. If literacy in Indigenous or Creole languages does not help students pass state exams and obtain viable lifelong employment, then the use of these languages as media of instruction is a doomed proposition. To elaborate on this predicament and make it more concrete with an example that I know firsthand, consider the vast majority of families and communities in Haiti—those in which only Kreyòl is spoken. Those parents may well be aware that their children will learn better if they can be taught in their native Kreyòl. But they are certainly smart enough to observe that those who pass the state’s official exams and who accumulate the most political and economic capital usually come from the group of French-Kreyòl bilinguals. Therefore, Haitian parents, even those who speak Kreyòl only, want their children to learn (in) French even if that means that their children will end up learning, in most cases, only a much reduced version of French, which will constitute yet another language barrier on their way to state exams and better jobs (Groupe de Travail sur l’Education et la Formation 2010, 149). The few students who pass the state exams do not necessarily have any deep understanding of the corresponding disciplines: rote memorization is the most prevalent method in preparing for these exams. The very few from Kreyòl-only homes who, because of great talent, a lot of luck, and great sacrifices, manage to succeed at school and move up the class system fuel the hope that such an exceptional feat is likewise possible for the masses. The latter, in turn, waste an extraordinary proportion of their meager income on schools that fail their children: these schools “teach” in a language that most of the students and even teachers are not fluent in, a language that most Haitians are rarely, if ever, exposed to in their everyday lives outside of school. Similar challenges obtain in many other parts of the world where Indigenous or local languages are losing the battle with international European languages as languages of instruction and socioeconomic advancement (Mufwene 2008). Like Mufwene (2008), Pfeiler makes an observation that supports the aforementioned proposition that, for local languages to be accepted as valid languages of instruction, they must also function as efficient tools for socioeconomic advancement, and we have to make it possible to conceive of these languages as instruments and symbols of power as well: “Currently, it appears that social dynamics are more powerful than governmental initiatives when it comes to determining the extent and frequency of language use in Yucatán State. Use of both languages may be required by law, but as long as Yucatec Maya lacks the prestige that Spanish enjoys and bilinguals are discouraged from using colloquial Maya at work, no linguistic political

[ 310 ]

chapter eleven

strategy will be successful for Yucatán’s Maya community or for Yucatecan society in general” (Pfeiler, chap. 8, this volume, p. 221). Pfeiler’s comment bears on virtually all the local languages in the Americas, Africa, Australia, Papua New Guinea, New Zealand, and so on. To tackle this global challenge, let’s borrow some hints from the Spanish colonists who instrumentalized Quechua as the language of Christianity in the colonial Andes: they made fluency in Quechua a prerequisite to holding certain clerical posts in Indian parishes and they introduced Quechua courses at university. Perhaps most importantly, “the [Quechua] language competence system generated a number of salaried posts for career experts in [Standard Colonial Quechua].” These salaried posts included “instructors working for the Jesuits and the mendicant orders,” “diocesan examiners,” two highereducation posts, one of them being a university chair, initially with “wideranging powers [such that] the entire Peruvian clergy were to be taught and examined by him if they hoped to be assigned a parish, or even to be ordained” (Durston, chap. 9, this volume, p. 234). As noted by Godenzzi, “the first chair of Quechua was created in Peru in 1570, 2 years before a chair of Dutch and 4 years before a chair of English” (2008, 323) Similar measures, without the subservience to “mendicant orders,” were implemented in the twentieth century to protect French in Quebec from the “threat” of English. In Quebec, language-policy initiatives that enlisted legislative and socioeconomic measures were crucial to the government’s efforts to promote French (Mufwene 2008, chap. 11). Indeed, “it is an enhanced market value that will really revitalize a language in its vernacular function” (Mufwene 2008, 242). The successful promotion of Quechua in the colonial Andes and of French in contemporary Quebec demonstrates what can be accomplished through “positive initiatives” in favor of the local languages of Latin America, Africa, Australia—and anywhere else in the world where local languages are dominated by international, typically European, languages. These considerations apply most urgently to the Indigenous languages of Latin America, especially keeping in mind that the majority of them—unlike Yucatec Mayan, Quechua, and Kreyòl—are now moribund, as is made evident by the growing literature on language endangerment (see, e.g., Couto, chap. 3, this volume; Hinton and Hale 2001; Austin and Simpson 2007; Grenoble and Whaley 1998; Haboud and Ostler 2011).

12. Toward a Sequel: Local Languages for Education, Research, and Social Justice Barbara Pfeiler duly worries that “the current processes of emigration and urbanization in Yucatán endanger the survival of the Maya language.”

Michel DeGraff

[ 311 ]

She explains, “This development is accelerated by the tendency of young bilingual people to use Spanish to communicate among themselves. The social domains that tended to be separated by exclusive use of Maya or Spanish are increasingly covered by Spanish only. In an increasingly larger number of families Maya is no longer transmitted to children as the mother tongue, and bilingualism is thus transitioning to Spanish monolingualism” (chap. 8, this volume, p. 211). We can start addressing this concern by again looking at Haiti as our case study for language evolution in Latin America. Then we’ll see how the Haiti case bears on the rest of Latin America. The good (and bad) news from Haiti is that the scenario about the endangerment of Maya in the Yucatán is not applicable to HC. Since most families in Haiti speak Kreyòl only, it’s only Kreyòl that is transmitted as native language to most Haitian children. Schoolteachers, by and large, are not fluent in French, even though their main task, as they and the public perceive it, is to prepare their students to pass exams that, for the most part, are administered in French only. This is one of the main reasons why nine out of ten students do not make it through high school. At this time, unlike the Yucatán situation with respect to Spanish, there’s no nationwide tendency in Haiti for people, old or young, to communicate in French among themselves. Even as a bilingual child in Haiti (a statistical outlier by accident of birth), I myself generally did not use French to communicate with my peers; I only spoke Kreyòl to them, except in the classroom. As a child, I considered French to be reserved for school and for other occasions when parents or other authority figures were within earshot—even though those very figures usually spoke Kreyòl among themselves. If anything, the communicative domains that, in the past, tended to be the exclusive province of French (e.g., TV) now seem wide open to Kreyòl, and we’re finding more and more use of Kreyòl in churches, in newspapers, on the radio, in advertising, in public service announcements, in films, on the Internet, in textbooks, and so on. I myself have been involved, through the MIT-Haiti Initiative, in the production, evaluation, and dissemination of Kreyòl-based pedagogical materials for elementary through higher education, and we now even have high-quality technology-enhanced resources for active learning in science and mathematics in Kreyòl (DeGraff and Ulysse 2011; DeGraff and Driscoll 2011; DeGraff 2013a, 2013b).8 Better yet, such materials bring concrete proof that Haitian Creole, on a par with other languages, is structurally adequate for science, contrary to popular belief even among educators and linguists (DeGraff 2001a, 2005a). Furthermore, the use of Kreyòl seems an indispensable ingredient for deep learning in Haiti, from primary school to university, as documented in recent

[ 312 ]

chapter eleven

collaboration between university faculty at MIT and in Haiti (DeGraff 2013a, 2013b). My hunch and hope is that this trend in favor of Kreyòl is for the better. In October 2011, there was a conference in Port-au-Prince organized by the State University of Haiti to plan the Akademi ayisyen pou lang kreyòl ‘Haitian academy for the creole language’, a much-awaited follow-up to article 213 of the 1987 constitution, which prescribes the creation of a Haitian academy “to standardize the Creole language and enable it to develop scientifically and harmoniously.” This institution is to be conceived along the broad lines of the aforementioned Institute for Development of Mayan Culture in the Yucatán state. As previously mentioned, in July 2013 the Haitian Parliament voted (unanimously!) to establish this Kreyòl Academy. This unanimous vote is all the more noteworthy in light of the fact that the promulgation of the law has to date been blocked by the president of the republic because it was presented to him in Kreyòl only. A note of caution is in order about institutions, such as the French academy, whose priority is to impose “standards” from above without much regard to actual linguistic practice in the lower social strata: When it comes to “standardizing,” we should learn from the limitations of the colonial Spanish efforts to codify a written version of Quechua that few Indigenous people actually spoke; it served the interests of the powerful instead of the people. So we may advise against one task often assigned to language academies: that of creating standards that cater to elitist hierarchies of power and make the “standard” relatively inaccessible to vernacular users of the language. Luykx warns against the neocolonial “standard language” ideology in contemporary efforts to codify Quechua, because she sees “the beginnings of a new sociolinguistic elitism around Quechua” (2004, 152); see Dorian (1998) for a general argument against the sort of “Western ideologies” that undermine “small-language prospects.”9 Such prospects may fare better with language academies that shun prescription in favor of description and applications. Academies of this type would function as scientific centers that help the state, along with academic institutions and local communities, make more constructive uses of their countries’ linguistic assets. Such language academies would help survey, transcribe, preserve, and study the richness of local folk traditions (tales, fables, proverbs, songs, prayers, charades, puns, and the like) many of which have immense linguistic, intellectual, historical, scientific, and spiritual value. In the case of Haiti, this plea has long been argued for by Haitian Vodou priest Max Beauvoir, who reminds us that the origins of Greek philosophy are in oral traditions harking back to

Michel DeGraff

[ 313 ]

Homer and Socrates (see, e.g., Beauvoir 2009; Beauvoir is Chief Supreme of the National Confederation of Haitian Vodou). It is certainly helpful to have some relatively uniform conventions for written representations of any language (e.g., for ease of communication and digital processing), but standardization should not distract from other valuable priorities that function to promote the language for the benefit of those who use it the most. In Haiti, Kreyòl’s official orthography is to be applauded as one of the rare cases in which the “standard” was designed while taking into account the linguistic patterns and the pedagogical needs of the majority population—in this case, the “masses” of monolingual speakers of the local language instead of the bilingual elite (cf. Schieffelin and Doucet 1994). Besides “standardization,” the other goal legally assigned to the Haitian Creole Academy is to “develop” Kreyòl. Here too we may learn important caveats from those sixteenth- and seventeenth-century Indigenous and European leaders who had such a shrewd understanding of the relationship between language and power. In the early colonial period, the Spanish authorities did “develop” Quechua (or Standard Colonial Quechua [Durston, chap. 9, this volume]), and they made fluency in Quechua a ticket for certain well-paying jobs and for heightened social status in the Andes. The well-paying jobs were usually for the European (i.e., nonnative) speakers of Quechua. These economic and social benefits now elude most of those who speak local languages only. Thus, “developing” a local language will not serve speakers of that language in the absence of necessary structures for socioeconomic advancement, including an adequate school system with quality pedagogical and reading materials that speakers of local languages can access in their mother tongue (for the Haiti case, see Dejean 2006; DeGraff and Ulysse 2011; DeGraff and Driscoll 2011; DeGraff 2013a, 2013b; Dejean and DeGraff 2013). “Developing” local languages whose speakers have long been impoverished and stigmatized requires the use of all means necessary to invest the language with scientific, cultural, and socioeconomic capital. These means would include literary contests in the language (such as those organized by the journal Bon Nouvèl in Haiti), enforceable requirements that all businesses affecting local speakers be conducted in the relevant local languages, and so on. But, more than any academy, it is the state that must come up with measures to stop “linguistic apartheid” practices. Here too we can find inspiration from the Andes: in Peru, a law that was passed in 2006 penalizes “exclusion for linguistic reasons,” and in Bolivia the current government of Evo Morales mandates that all public servants learn an Indigenous language in addition to Spanish (Godenzzi 2008,

[ 314 ]

chapter eleven

323; Morales 2010). In Haiti, all public servants, like every other Haitian, already speak Kreyòl. All they need to do is to make use of it in every verbal or written interaction with the public they serve, the majority of whom Kreyòl only. There’s already a model for that in Haiti: When it was under the direction of Suze Mathieu, the National Bureau of Ethnology conducted all its business in Kreyòl. Suze Mathieu has made several pleas, so far with relatively little effect, that other state offices follow suit (Mathieu 2005, 2008). This resistance against Kreyòl may not be surprising when one remembers that, as recently as June 2010 at a public forum titled “National Reconstruction” at the state university, Leslie Manigat, an eminent historian and former president of Haiti, declared that Kreyòl was “not only a limitation, but an infirmity as well.” We linguists can encourage the promotion of local languages by collaborating with local educators and scholars in their ongoing efforts to promote more informed knowledge about the viability of local languages, both as objects of research and as indispensable tools for education and socioeconomic betterment. One way to collaborate, if collaboration is desired, is to help train additional linguists who are native speakers of local languages. Hinton and Hale (2001), Hornberger and Coronel-Molina (2004), and Haboud and Ostler (2011) provide overviews of such collaborative efforts, including “bilingual intercultural” teacher-training and university-level efforts to promote the knowledge and use of Indigenous languages. Such training is especially important in communities— such as in the Caribbean, Meso-America, the Andes, the Amazon, and so forth—where speakers themselves, for (neo)colonial reasons, are at best often unaware of the pedagogical and intellectual values of their local languages or, worse yet, have been persuaded to reject proposals that their native languages be used as media of instruction (King and Hornberger 2004; García 2004; Godenzzi 2008). Another avenue for collaboration is in the production of high-quality and freely accessible pedagogical materials in and about these languages that we linguists often study for our own intellectual and socioeconomic betterment. From my own perspective as a Creole speaker and a linguist (a perspective quite different from that of the majority of linguists writing about language evolution), such collaboration to promote education and research in and about local languages is a win-win proposition for at least the following seven reasons: (1) Linguists who are interested in Latin America and who are affiliated with established institutions in the Global North could ensure that their writings about local languages reach, and enter into dialogue with, the corresponding communities. With the latter’s consent (“free, prior, and informed consent” as outlined in United Nations 2009), these lin-

Michel DeGraff

[ 315 ]

guists could use their influence to help usher in language policies that are beneficial to local communities. In return, the linguists’ writings would gain validity as they became more grounded in the actual needs of the people who speak the languages under study. (2) As linguists, some of us can function as vectors of social justice because our know-how—rooted as it is (for the most part, and for better or for worse) in relatively prestigious Western traditions—can help invest local languages with the sort of intellectual and scientific capital that they have been denied in the past. Keeping in mind the rise in prestige of French and Italian through the writings of Descartes and Galileo, respectively, one can surmise that linguistic capital can be created through the production of literary, scientific, and educational materials of the highest quality for speakers of local languages. As linguists we could contribute some of our expertise to the production of such materials, especially in academic domains to which Indigenous communities have had too little access so far. Ideally, Indigenous educators themselves would produce such materials for their own communities—materials such as school textbooks, online open resources, higher-education academic materials, and so forth (see McCarty et al. 2008, 305–307 for related arguments). Fortunately, the development of materials by native speakers is already occurring to various extents, especially for the “bigger” local languages (Hornberger and Coronel-Molina 2004). (3) For those of us collaborating on the promotion of local languages as media of instruction in the Americas, Africa, Australia, and so forth, this collaboration can broaden and deepen our understanding of our objects of study and their larger context of use. For example, consider that many of these local languages have yet to be used at universities for academic disciplines such as science, technology, engineering and math (STEM) or in the humanities. Producing materials in local languages for STEM and for the humanities in higher education will require a fair amount of research and ingenuity as we work toward the optimal development of technical and scholarly terminology in these languages. This is the kind of work that would benefit from the advice of linguists with expertise in the corresponding languages. Furthermore, materials for higher education and research in local languages can double as concrete proof against the still popular belief that local languages such as Kreyòl in Haiti are an “infirmity” or cannot express complex abstract semantics because they allegedly “lack the more sophisticated features of languages backed by a rich and extended cultural past and a large, well-organized literate society” (Seuren 1998, 292–293; see DeGraff 2001a, 2005a, for overviews of such claims; also see note 8 to the current chapter for references to recent concrete proofs against such claims in the context of recent work

[ 316 ]

chapter eleven

for the production of resources for university-level science and math in Kreyòl). (4) From a practical standpoint, such materials will improve educational opportunities for communities where Creole or Indigenous languages are spoken by the majority—these are the communities that have been overlooked the most in terms of quality education material and other resources. Linguists’ engagement in the production of these materials will help implement “Education for All” that really includes all. The use of local languages in schools and in academia is not to be considered solely or primarily as a remedy for (potential) language shift toward a dominant language (cf. Luykx 2004 for important reasons why such a remedy may be problematic). Indeed, Kreyòl in Haiti is alive and well, since it is already spoken by all in Haiti, while French, the dominant language in the schools for the past two centuries, is spoken by no more than 10 percent of the population. So there is no current risk of language endangerment in the Haiti case. The goal envisaged here is fair access to education that is constructively rooted in local knowledge, culture, and language: we should aim to achieve education that is really accessible to all, in the language(s) spoken by all students, keeping in mind that instruction is best carried out in the language that students are most fluent in, as recognized by UNESCO a long time ago (UNESCO 1953). Only pedagogical methods that enlist languages in which students are fluent will prepare the students to actively participate in the construction of knowledge and turn them from passive consumers to proud producers (Luykx 1999). For the endangered Indigenous languages of Latin America and elsewhere, preventing language loss also requires reinforcement of the use of local languages at home and in communities, above and beyond any scholastic uses (Martínez Cobo 1987; Dorian 1998; Luykx 2004). (5) Nowadays education, especially higher education and education in STEM areas and in the law, seems an indispensable tool for development. Yet communities that function primarily in a local language are severely underrepresented in schools and, especially, in universities and research and legal institutions, even though scientists and lawyers are making great profits and advances through the exploitation of natural, cultural, and intellectual resources in Indigenous territories. Consider, say, the huge profits in the exploitation of Indigenous medicinal plants in “ethnobotany and bioprospecting” by the pharmaceutical industry (McManis 2007). Ironically, Indigenous communities in the Americas are among the ones that suffer the most from health disparities; in the United States, of all places, death from tuberculosis among Native Americans is 500 percent higher than in the general population (Indian Health Services 2011). The more Creole and Indigenous speakers can be active in access-

Michel DeGraff

[ 317 ]

ing, producing, and disseminating scientific knowledge—including their own Indigenous knowledge—in their own languages, the better off their communities and the rest of the world will be in the long run. Consider, for example, the potential benefit of traditional Indigenous knowledge and practices for protecting the environment, with “longstanding connections to and reverence for the land, traditions of sustainability, historical knowledge of the land . . . and expertise in natural resource and wildlife management” (Curry et al. 2011, 22; see McManis 2007 for an overview of related issues; Devonish 2007, 240, links such traditional knowledge to the survival of Indigenous languages). (6) Now consider what it would take to fully enforce both article 29 of the United Nations Declaration on the Rights of Indigenous Peoples, which grants Indigenous peoples the right to the “conservation and protection of the environment and the productive capacity of their lands or territories and resources,” and article 32, which requires states to “consult and cooperate in good faith with the Indigenous peoples concerned . . . in order to obtain their free, prior and informed consent prior to the approval of any project affecting their land or territories and other resources” (United Nations 2007, emphasis added). Such consent presupposes unfettered communication between parties and full access to the relevant information. Yet, communication with states and other powerful agents (especially multinational corporations) and access to, and production and dissemination of, information (especially scientific information) are hampered by the imposition of foreign or distant second languages as the media of communication and instruction. This is one of the main rationales for projects in which educators and linguists collaborate for the production of pedagogical materials in local languages. See article 14, declaring the rights of Indigenous peoples to “education in their own languages, in a manner appropriate to their cultural methods of teaching and learning” (United Nations 2007; cf. Martínez Cobo 1987). One must hope that these materials can, in turn, invest additional intellectual and social capital in the local languages and attract additional native speakers as collaborators (Hale 1965, 1972; Hinton and Hale 2001; England 2007). (7) If respectful collaboration based on “free, prior and informed consent” (United Nations 2009) can be sustained with speakers of local, and often understudied, languages, there is also the exciting prospect that novel knowledge and data from these speakers will enrich science and make way for joint discoveries and for new theories that would be unthinkable in the absence of such diversity of perspectives (Hale 1965, 1972, 1998). In my own field of Creole studies, I am grateful to, among others, scholars from the Caribbean and Africa who have extraordinarily deepened my understanding of Creole languages and their ecologies. If

[ 318 ]

chapter eleven

I may say so myself, I think (or hope) that, once I got through the arduous work of unlearning the exceptionalist dogma of Creole studies, my own insights as a Creole speaker and a linguist have, in turn, enriched Creole studies through, among other things, my analyses of Haitian Creole and my contributions to the dismantling of age-old Creole Exceptionalism tenets (see, e.g., DeGraff 2001a, 2005a, 2005b, 2009; DeGraff and Walicek 2005). From this perspective, we native speakers of Creole and Indigenous languages are not passive consumers of, or informants for, preexisting mainstream theories—in ways that further marginalize them as “others.” Instead, we prefer to become actively engaged as equal partners in the production of analyses and theories that are enriched by diverse contributions from both within and outside the local groups. (For one recent example in theoretical syntax and semantics, see Coon et al. 2011.) The interaction between linguistic theories and fieldwork on lesser-studied languages is indeed most constructive when the latter are studied by linguists from the corresponding Indigenous communities. This fact again points to the need for many more speakers of local languages to become linguists in their own rights and on their own terms—an approach long advocated by my dear late colleague Ken Hale (see, e.g., Hale 1965, 1972). Furthermore, these native-speaker linguists are in the best position to train additional linguists in their own communities (the African Linguistics School [www.als.rutgers.edu] is a recent addition to these efforts). And the more numerous native-speaker linguists become in their community, the more their community, including educators and parents, will become aware of the intellectual and socioeconomic values of their local languages.

13. Envoi This engagement—of linguists wearing two hats, one theoretical and one applied—seems to me a welcome avenue for mutually enriching NorthSouth collaboration with those communities that have given us linguists such fascinating data and insights to work on, communities whose own linguistic knowledge and behavior have made some of us “rich and famous” (hum, hum . . .). Of course, we need to remember that we all have our limitations, including ideological blind spots, and these can implicitly pollute work with local communities. One limitation is insuperable: “outsider linguists simply do not have the power to create a new generation of speakers”; “only community-based projects have any hope of success” (Speas 2011, 17, 25; also see Dorian 1998; Luykx 1999, 2004, for related caveats). The engagement advocated here will contribute to correcting what Rickford has described as “the present unequal partnership between

Michel DeGraff

[ 319 ]

researcher and researched [which] is widespread within linguistics” (1997, 161; also see McCarty et al. 2008, 305–307). Rickford’s invitation is echoed by article 31 of the UN Declaration on the Rights of Indigenous Peoples: “Indigenous peoples have the right to maintain, control, protect and develop their cultural heritage, traditional knowledge and traditional cultural expressions, as well as the manifestations of their sciences, technologies and cultures. . . . They also have the right to maintain, control, protect and develop their intellectual property over such cultural heritage, traditional knowledge, and traditional cultural expressions.” Arguments related to those in paragraphs 1–7 above have been made before, and more eloquently, in inspiring work by many linguists and activists before me, especially Hale (1965, 1972, 1998), Hinton and Hale (2001), and Haboud and Ostler (2011). For two recent examples of linguists’ collaborations with Indigenous communities, consider the work of Makepeace and co-workers (2010) on the Wampanoag case in Massachusetts (online information at http://wlrp.org/ and at http://web.mit.edu/norvin/ www/wopanaak.html) and Chacón and co-workers (2011) on Quechua in the Andes. The Wampanoag case is a particularly striking success story, as the language was unspoken for seven generations and has now been revived, acquiring in 2004 its first native speaker in a century: Mae Alice Baird, the daughter of Jessie Little Doe Fermino Baird, an MIT-trained Wampanoag linguist (Fermino 2000). Fermino Baird’s teacher at MIT was Ken Hale, who had a clear understanding of language-endangerment issues from the perspective of linguistics training, an understanding that he summarized as follows: “The future of American Indian linguistics will depend critically on how successful an effort there is to engage American Indians in the active study of their own languages—not as informants as in the past, but as linguists, philologists, lexicographers, creative writers, and the like. To put it another way, significant advances in the study of American Indian languages can be made, in my judgment, only when a significant portion of the field is in the hands of native speakers of the languages concerned” (1972, 87). Let’s wait for the sequel of this anthology, a sequel that, one must hope, will showcase a healthy share of chapters by Indigenous speaker-linguists writing about successful linguistic, educational, and socioeconomic development programs in their respective communities. (See Haboud and Ostler 2011 for inspiring examples of such programs and future prospects throughout the world.) I take it as yet another reflex of the socioeconomic matrix of language endangerment that this anthology lacks any contribution from Indigenous linguists from the very communities whose native languages are endangered (see note 1). Notwithstanding such limitation, this book provides

[ 320 ]

chapter eleven

a stimulating and enriching set of lessons for the understanding of the relationships among cultural, historical, socioeconomic, and languageevolution issues in the Americas. It invites those of us who are interested in language contact and its structural, cultural, and socioeconomic outcome to do much better in broadening and sharpening our lenses as we reexamine the complex ecologies for complex scenarios of language shift and language change. To the contributors to this anthology, Chapo ba! ‘Hats off!’—both of those hats. Notes I am grateful to Enoch Aboh, Max Beauvoir, Rachel Beauvoir-Dominique, Jean Casimir, Yves Dejean, Elena Geretti, Juan Carlos Godenzzi, Deborah Jenson, Susan Kalt, Kendall King, Suze Mathieu, Salikoko Mufwene, and Nuriel Vera-DeGraff for discussing with me various aspects of this chapter during its gestation, and for opening up many additional avenues for investigation. Special thanks go to Salikoko Mufwene, Susan Kalt, Juan Carlos Godenzzi, and Kendall King for thorough and most constructive comments on an early draft. Of course, no one but me can be blamed for any leftover errors. 1. I use the phrase Indigenous languages to refer to the languages of “Indigenous” populations in the geopolitical sense of being there “first,” whereby “Indigenous communities, peoples and nations are those which, having a historical continuity with pre-invasion and pre-colonial societies that developed on their territories, consider themselves distinct from other sectors of the societies now prevailing in those territories or parts of them” (Martínez Cobo 1987, 29). Population movements throughout human history often make it difficult to determine who exactly was there “first.” This difficulty creates a need in certain cases to think of degrees of Indigenousness, as suggested by Salikoko Mufwene (2001). In the cases at hand, this question is relatively clear-cut: the Amerindian peoples and their languages are (the most) Indigenous to Latin America, because they predate their European, African, and Creole counterparts. 2. I am grateful to Norvin Richards for discussing with me some of his work on Wampanoag and on Eliot’s translation. Thanks are also due to Susan Kalt for alerting me to the theological and ideological parallels between the European renditions of Wampanoag pauwau and Quechua supay. 3. Here and in subsequent sections, I use the phrase local languages to refer to languages such as Creole and Indigenous languages in the Americas, in opposition to European languages that have become “international” through imperial expansion. 4. But, with enough of a critical stance, one can collect helpful hints toward the development of such a database from reports by Labat (1722), Moreau de Saint-Méry (1797), Descourtilz (1809), Price-Mars (1956), Fouchard (1953, 1981), H. Trouillot (1962), Debien (1971), Geggus (1989, 1991, 2002), Thornton (1993), Dayan (1995), Heywood (2002), Trouillot (1995, 2002), Montilus (2006), Midy (2006), Klein (2006), Singler (1996), Heywood and Thornton (2007), Warner-Lewis (2003), Beauvoir and Dominique (2003), Beauvoir (2009), Beauvoir-Dominique (2011), and Jenson (2011). 5. These observations are all about men in power and in battle who had to affirm

Michel DeGraff

[ 321 ]

and project an identity of authority to their troops and to the early-nineteenthcentury world at large, a world mostly hostile to African slaves fighting for their freedom against European armies—at a time when race-based slavery was an engine of wealth for most of the Western world. So it’s not likely that these reports can shed much light on the cultural subjectivities of ordinary Africans and their descendants in the colonial Caribbean. Yet it can be surmised that the attitudes and beliefs of popular revolutionary leaders such as Toussaint and Dessalines did influence or reflect more general attitudes about Kreyòl and French among the general population. The latter was itself ethnically diverse, and there were certainly ideological differences among the various groups. We also need to stress that most of our available reports about blacks in the colonial era are filtered through the racist subjectivities of white authors such as Moreau de Saint-Méry and Michel-Etienne Descourtilz, who were deeply embedded and invested in the colonial world order. Even the famous proclamations by the Haitian Revolution leaders were written or transcribed by secretaries who were often raised or educated in France (Daut 2009; Jenson 2011). Ambivalent or negative stereotypes about Africa, many of them inherited from European and American scholars, can be found as far as in the writings of pro-Négritude Haitian scholar Jean Price-Mars (e.g., Price-Mars 1956, 41–44). So it is only indirectly and with great care that we can glean the ideologies of ordinary Africans and their Creole descendants in the colonial Americas. 6. In a sociologically fascinating passage, Moreau de Saint-Méry reports that African women preferred black men over white men because the blacks were better “physical agents for love,” hinting at, among other things, the fact that white males were often sexually coercive toward black females (Dayan 1995, 236). 7. Areche had also condemned Túpac Amaru II to a most brutal execution (he was hung, drawn, and quartered!). 8. For a sampling of these Kreyòl resources online, see, e.g., http://haiti.mit.edu, http://haiti.mit.edu/resources/, http://star.mit.edu/genetics, http://star.mit.edu/ biochem, http://haiti.mit.edu/workshops/august-2013-workshop/august-2013 -resources/. For a wider sample, beyond STEM areas, also see the resources available at http://www.educavision.com, www.editionsuniversitecaraibe.com, http://laction sociale.com/catalogue.php, www.editionskonbit.com, www.potomitan.info/vedrine/ vedrine1.php. 9. One anonymous reviewer mentions the case of the Euskara Batua variety of Basque, which was promoted by the Academy of the Basque Language, with some degree of success, as the written “standard” variety of the language.

References Aboh, Enoch. 2009. Competition and selection: That’s all! In Complex process in new languages, ed. Enoch Aboh and Norval Smith, 317–344. Amsterdam: John Benjamins. Aboh, Enoch, and Michel DeGraff. Forthcoming-a. A null theory of Creole formation. In The Oxford handbook of universal grammar, ed. Ian Roberts. Oxford: Oxford University Press. ———. Forthcoming-b. Some notes on nominal phrases in Haitian Creole and in Gùngbè: A trans-Atlantic Sprachbund perspective. In Language contact and language change: Grammatical structure encounters the fluidity of language, ed. Tor Afarli and Brit Mæhlum. Amsterdam: John Benjamins.

[ 322 ]

chapter eleven

Austin, Peter, and Andrew Simpson, eds. 2007. Endangered languages. Hamburg: Helmut Buske Verlag. Beauvoir, Max. 2009. De la littérature orale vodou haïtienne. Paper read at the Institut d’études et de recherches africaines, October 22, and at the Faculté de linguistique appliquée, October 29, Port-au-Prince, Haiti. Beauvoir, Rachel, and Didier Dominique. 2003. Savalou e. Montreal: Les Editions du CIDIHCA. Beauvoir-Dominique, Rachel. 2011. Significance of the Bois Caiman ceremony, SaintDomingue, 1791. Paper read at the Black in Latin America Conference, W. E. B. Du Bois Institute for African and African American Research, Harvard University, Cambridge, MA, January 29. Bloomfield, Leonard. 1925. Why a linguistics society? Language 1:1–5. Bourdieu, Pierre. 1982. Ce que parler veut dire: L’économie des échanges linguistiques. Paris: Fayard. English edition: Language and symbolic power, ed. and trans. Gino Raymond and Matthew Adamson (Cambridge, MA: Harvard University Press, 1991). Bunsen, Christian Charles Josias. 1854. Outlines of the philosophy of universal history applied to language and religion. Vol. 2. London: Longman, Brown, Green, and Longmans. Casimir, Jean. 2011. Haiti’s need for a Great South. The global south and world dis/ order, special issue of Global South 5, no. 1: 14–36. Chacón, Jaime Aráoz, Martin Castillo Collado, and Susan E. Kalt. 2011. Proyecto yachay q’ipikuna. Available at www.rcc.mass.edu/language/faculty/Sue_Kalt/ sue_project.asp. Coon, Jessica, Pedro Mateo Pedro, and Omer Preminger. 2011. The role of case in a-bar extraction asymmetries: Evidence from Mayan. Ms. available at http://ling .auf.net/lingBuzz/001401. Curry, Renee, Charissa Eichman, Amanda Staudt, Garrit Voggesser, and Myra Wilensky. 2011. Facing the storm: Indian tribes, climate-induced weather extremes, and the future for Indian country. Boulder, CO: National Wildlife Federation. Available at http://www.nwf.org/~/media/PDFs/Global-Warming/Reports/NWF_Tribal LandsExtremeWeather_FINAL.ashx. Daut, Marlene. 2009. Un-silencing the past: Boisrond-Tonnerre, Vastey, and the re-writing of the Haitian Revolution, 1805–1817. South Atlantic Review 74, no. 1: 35–64. Dayan, Joan. 1995. Haiti, history, and the gods. Berkeley: University of California Press. Debien, Gabriel. 1971. Les esclaves aux Antilles Françaises (XVII–XVIII siècles). BasseTerre, Guadeloupe: Société d’Histoire de la Guadeloupe. DeGraff, Michel. 2001a. Morphology in Creole genesis: Linguistics and ideology. In Ken Hale: A life in language, ed. Michael Kenstowicz, 53–121. Cambridge, MA: MIT Press. ———. 2001b. On the origins of Creole: A Cartesian critique of Neo-Darwinian linguistics. Linguistic Typology 5, nos. 2–3: 213–310. ———. 2002. Relexification: A re -evaluation. Anthropological Linguistics 44, no. 4: 321–414. ———. 2005a. Linguists’ most dangerous myth: The fallacy of Creole Exceptionalism. Language in Society 34, no. 4: 533–591.

Michel DeGraff

[ 323 ]

———. 2005b. Word order and morphology in “creolization” and beyond. In The Oxford Handbook of Comparative Syntax, ed. Guglielmo Cinque and Richard Kayne, 293–372. New York: Oxford University Press. ———. 2009. Language acquisition in creolization and, thus, language change: Some Cartesian-Uniformitarian boundary conditions. Language and Linguistic Compass 3–4:888–971. ———. 2010. Language barrier: Creole is the language of Haiti, and the education system needs to reflect that. Boston Globe, Opinion section, July 16. ———. 2013a. Men anpil, chay pa lou: An nou sèvi ak lang kreyòl la pou bon jan edikasyon ak rechèch ann Ayiti. In Akademi kreyòl ayisyen: Ki pwoblèm? Ki avantaj? Ki defi? Ak kolòk entènasyonal sou Akademi Kreyòl Aysiyen an, ed. Renault Govain and Komite pou tabli Akademi Kreyòl Ayisyen, 133–166. Port-au-Prince, Haiti: Editions de l’Université d’Etat d’Haïti. ———. 2013b. MIT-Haiti Initiative uses Haitian Creole to make learning truly active, constructive, and interactive. In Educational Technology Debate. Online article at https://edutechdebate.org/cultural-heritage-and-role-of-education/mit-haiti -initiative-uses-haitian-creole-to-make-learning-truly-active-constructive-and -interactive/. DeGraff, Michel, Robert Berwick, and Trevor Bass. 2013. Language acquisition and computational phylogenetics: Creole languages and family values. Ms., available at MIT Linguistics & Philosophy. DeGraff, Michel, and Megan Driscoll. 2011. MIT’s Michel DeGraff on expanding access to education in Haiti. Education Portal, online interview, at http://education -portal.com/articles/Expanding_Access_to_Education_Professor_Michel_DeGraff _Talks_About_His_Work_for_Haiti_at_MIT.html. DeGraff, Michel, and Katia Ulysse. 2011. Our word is our bond. Voices of Haiti, online interview at www.voicesfromhaiti.com/inner-views/michel-degraff-mit-linguistics -professor-our-word-is-our-bond/. DeGraff, Michel, and Don Walicek. 2005. Creole Exceptionalism and accidents of history. Sargasso, 1–34. Dejean, Yves. 2006. Yon lekòl tèt anba nan yon peyi tèt anba [An upside-down school in an upside-down country]. Port-au-Prince, Haiti: FOKAL. Dejean, Yves, and Michel DeGraff. 2013. Language mismanagement in Haiti. Ms., available at MIT Linguistics. Descourtilz, Michel Etienne. 1809. Voyages d’un naturaliste, et ses observations faites sur les trois règnes de la nature, dans plusieurs ports de mer français, en Espagne, au continent de l’Amérique septentrionale. . . . 3 vols. Paris: Dufart, Père. Deutscher, Guy. 2009. Overall complexity: A wild goose chase? In Language complexity as an evolving variable, ed. Geoffrey Sampson, David Gil, and Peter Trudgill, 243–251. Oxford: Oxford University Press. Devonish, Hubert. 2007. Language and liberation: Creole language politics in the Caribbean. 2nd ed. Kingston, Jamaica: Arawak. Dorian, Nancy. 1998. Western language ideologies and small-language prospects. In Grenoble and Whaley 1998, 3–21. Dubois, Laurent. 2011. Haiti: The aftershocks of history. New York: Metropolitan Books. England, Nora. 2007. The influence of Mayan-speaking linguists on the state of Mayan linguistics. In Austin and Simpson 2007, 93–112.

[ 324 ]

chapter eleven

Fattier, Dominique. 1998. Contribution à l’étude de la genèse d’un créole: L’atlas linguistique d’Haïti, cartes et commentaires. 6 vols. PhD diss., Université de Provence, France. Fermino, Jessie Little Doe. 2000. An introduction to Wampanoag grammar. Master’s thesis, MIT. Fick, Carolyn. 1990. The making of Haiti: The Saint Domingue revolution from below. Knoxville: University of Tennessee Press. Foster, Charles, and Albert Valdman, eds. 1984. Haiti—today and tomorrow: An interdisciplinary study. Lanham, MD: University Press of America. Fouchard, Jean. 1953. Les marrons du syllabaire: Quelques aspects du problème de l’instruction et de l’éducation des esclaves et affranchis de Saint-Domingue. Portau-Prince, Haiti: Imprimerie Henri Deschamps. ———. 1972. Langue et littérature des aborigènes d’Ayti. Paris: Editions de l’Ecole. ———. 1981. The Haitian maroons: Liberty or death. New York: E. W. Blyden Press. García, María Elena. 2004. Rethinking bilingual education in Peru: Intercultural politics, state policy, and Indigenous rights. Bilingual Education and Bilingualism 7, no. 5: 348–367. 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: 23–44. ———. 1991. Haitian Voodoo in the eighteenth century: Language, culture, resistance. Jahrbuch für Geschichte von Staat, Wirtschaft und Gesellschaft Lateinamerikas 28:21–51. ———. 2002. Haitian revolutionary studies. Bloomington: Indiana University Press. Godenzzi, Juan Carlos. 2008. Language policy and education in the Andes. In Encyclopedia of language and education. Vol. 1, Language policy and political issues in education, ed. Nancy H. Hornberger and Stephen May, 315–329. New York: Springer. Grenoble, Lenore A., and Lindsay J. Whaley, eds. 1998. Endangered languages: Language loss and community response. Cambridge: Cambridge University Press. Groupe de Travail sur l’Education et la Formation. 2010. Pour un pacte national pour l’éducation en Haïti. Port-au-Prince, Haiti: Bibliothèque Nationale. Haboud, Marleen, and Nicholas Ostler. 2011. Endangered languages: Voices and images. Bath, UK: Foundation for Endangered Languages. Hale, Kenneth. 1965. On the use of informants in field-work. Canadian Journal of Linguistics 10:108–119. ———. 1972. A new perspective on American Indian linguistics. In New perspectives on the pueblos, ed. Alfonso Ortiz, 87–110. Albuquerque: University of New Mexico Press. ———. 1998. On endangered languages and the importance of linguistic diversity. In Grenoble and. Whaley 1998, 192–216. Harrison, Regina. 1989. Signs, songs, and memory in the Andes: Translating Quechua language and culture. Austin: University of Texas Press. Hebblethwaite, Benjamin. 2012. French and underdevelopment, Haitian Creole and development: Educational language policy problems and solutions in Haiti. Journal of Pidgin and Creole Languages 27, no. 2: 255–302. Heywood, Linda. 2002. Central Africans and cultural transformations in the American diaspora. Cambridge: Cambridge University Press. Heywood, Linda, and John Thornton. 2007. Central Africans, Atlantic Creoles, and the foundation of the Americas, 1585–1660. Cambridge: Cambridge University Press.

Michel DeGraff

[ 325 ]

Hinton, Leanne, and Kenneth Hale. 2001. The green book of language revitalization in practice. San Diego: Academic Press. Hoffmann, Léon François. 1984. Francophilia and cultural nationalism in Haiti. In Foster and Valdman 1984, 57–76. Hornberger, Nancy, and Serafín Coronel-Molina. 2004. Quechua language shift, maintenance, and revitalization in the Andes: The case for language planning. International Journal of the Sociology of Language 167:9–67. Hurbon, Laënnec. 1988. Le barbare imaginaire. Paris: Cerf. Indian Health Services 2011. IHS Fact Sheets: Disparities. Rockville, MD: Federal Health Program for American Indians and Alaska Natives. Available at www.ihs .gov/newsroom/factsheets/disparities. Jenson, Deborah. 2011. Beyond the slave narrative: Politics, sex, and manuscripts in the Haitian revolution. Liverpool: Liverpool University Press. King, Kendall. 2004. Language policy and local planning in South America: New directions for enrichment bilingual education in the Andes. Bilingual Education and Bilingualism 7, no. 5: 334–347. King, Kendall, and Nancy Hornberger, eds. 2004. Quechua sociolinguistics. International Journal of Sociology of Language 167 (special issue). ———. 2006. Quechua as a lingua franca. Annual Review of Applied Linguistics 26:177–194. Klein, Thomas. 2006. Consciousness and linguistic agency in Creole: Evidence from Gullah and Geechee. Internet Journal for Cultural Sciences 16. Labat, R. P. 1722. Nouveau voyage aux isles de l’Amérique contenant l’histoire naturelle de ces pays, l’origine, les moeurs, la religion & le gouvernement des habitans anciens & modernes. . . . 8 vols. 1742 edition: Paris: Guillaume Cavelier père, libraire. Page references are to the 1742 edition. Laurie, Nira, and Alastair Bonnett. 2002. Adjusting to equity: The contradictions of neoliberalism and the search for racial equity in Peru. Antipode 34, no. 1: 28–53. Lewis, M. Paul, ed. 2009. Ethnologue: Languages of the world. 16th ed. Dallas, TX: SIL International. Online version: www.ethnologue.com/. Luykx, Aurolyn. 1999. The citizen factory: Schooling and cultural production in Bolivia. Albany: State University of New York Press. ———. 2004. The future of Quechua and the Quechua of the future: Language ideologies and language planning in Bolivia. In King and Hornberger 2004, 147–158. Madiou, Thomas, fils. 1847. Histoire d’Haïti. 8 vols. Port-au-Prince, Haiti: Imprimerie Jh. Courtois. Makepeace, Anne, Jennifer Weston, Mary Lampson, and Jessie Little Doe. 2010. Âs Nutayuneân: We still live here. Oley, PA: Bullfrog Films. Martínez Cobo, José. 1987. Study of the problem of discrimination against Indigenous populations. Vol. 5, Conclusions, proposals, and recommendations. Submitted by the Special Rapporteur of the Sub-Commission on Prevention of Discrimination and Protection of Minorities. New York: United Nations. Mathieu, Suze. 2005. Depi nan Ginen nèg renmen nèg: Kreyòl ak demokrasi ann Ayiti [Since Ginen, blacks love blacks: Creole and democracy in Haiti]. Pòtoprens, Haiti: Près nasyonal peyi d Ayiti. ———. 2008. Lang nou, kilti nou, idantite nou [Our language, our culture, our identity]. Interview by Nélio Joseph. Le Nouvelliste, July 25. ———. 2010. Yon lèt tou louvri pou Mesye Vernet Henry Rektè Inivèsite Leta Ayi-

[ 326 ]

chapter eleven

syen [An open letter for Mr. Vernet Henry, Rector of the State University of Haiti]. Le Nouvelliste, August 21–22. McCarty, Teresa L., Toe Skutnabb-Kangas, and Ole Henrik Magga. 2008. Education for speakers of endangered languages. In Spolsky and Hult 2008, 297–312. McManis, Charles. 2007. Biodiversity and the law: Intellectual property, biotechnology, and traditional knowledge. London: Earthscan. McWhorter, John. 2001. The world’s simplest grammars are creole grammars. Linguistic Typology 5, nos. 2–3: 125–166. Meillet, Antoine. 1914. Le problème de la parenté des langues. Scientia, Rivista di Scienza 15, no. 35: 3. Reprinted in Linguistique historique et linguistique générale, by Antoine Meillet, vol. 1 (Paris: Honoré Champion, 1958). Midy, Franklin. 2006. Les Congos à Saint-Domingue: De l’imaginaire au réel. Ethnologies 28, no. 1: 173–201. Migge, Bettina, Isabelle Léglise, and Angela Bartens. 2010. Creoles in education: An appraisal of current programs and projects. Amsterdam: John Benjamins. Montilus, Guérin. 2006. Vodun and social transformation in the African diasporic experience: The concept of personhood in Haitian Vodun Religion. In Haitian Vodou: Spirit, myth, and reality, ed. Patrick Bellegarde-Smith and Claudine Michel, 1–6. Bloomington: Indiana University Press. Morales Ayma, Evo. 2010. Ley No 070 Ley de 20 de Diciembre de 2010: Ley de la educación “Avelino Siñani—Elizardo Pérez.” In Gaceta Oficial del Estado Plurinacional de Bolivia. Sucre, Bolivia: Ministerio de la Presidencia. Available at www .gobernabilidad.org.bo/images/stories/documentos/leyes/070.pdf. Mufwene, Salikoko S. 2001. The ecology of language evolution. Cambridge: Cambridge University Press. ———. 2008. Language evolution: Contact, competition, and change. London: Continuum. Myers-Scotton, Carol. 1993. Elite closure as a powerful language strategy: The African case. International Journal of the Sociology of Language 103:149–163. Parackahua Arancibia, María Cristina. 2011. Resistencia de padres y docentes contra la Educación Intercultural Bilingüe (EIB) y cómo vencer. Presentation at the Harvard University Workshop “Indigenous Voices in the Andean Classroom,” Cambridge, MA, November 3. Political Database of the Americas. 2011. http://pdba.georgetown.edu/Constitutions/ Bolivia/bolivia09.html. Price-Mars, Jean. 1956. Formation ethnique, folklore et culture du peuple haïtien. 2nd ed. Port-au-Prince, Haiti: Imprimerie N.A. Théodore. Ramsey, Kate. 2011. The spirits and the law: Vodou and power in Haiti. Chicago: University of Chicago Press. Rickford, John. 1997. Unequal partnership: Sociolinguistics and the African American speech community. Language in Society 26, no. 2: 161–197. Roberts, Peter. 2008. Roots of Caribbean identity. Cambridge: Cambridge University Press. Roberts, Sarah Julianne. 1998. The role of diffusion in the genesis of Hawaiian Creole. Language 74, no. 1: 1–39. Saint-Fort, Hugues. 2011. Haïti: Questions de langues, langues en question. Port-auPrince, Haiti: Editions de l’Université d’Etat d’Haïti. Saint-Louis, Vertus. 2006. Le surgissement du terme “africain” pendant la révolution de Saint-Domingue. Ethnologies 28, no. 1: 147–171.

Michel DeGraff

[ 327 ]

Saint-Méry, Moreau de. 1797. Description topographique, physique, civile, politique et historique de la partie française de l’isle de Saint Domingue. 3 vols. Philadelphia: Published by the author. Samarin, William. 1986. Protestant missions and the history of Lingala. Journal of Religion in Africa 16, no. 2: 138–163. Sánchez, Liliana. 2004. Functional convergence in the tense, evidentiality, and aspectual systems of Quechua Spanish bilinguals. Bilingualism: Language and Cognition 7, no. 2: 147–162. Schieffelin, Bambi, and Rachelle Charlier Doucet. 1994. The “real” Haitian Creole: Ideology, metalinguistics, and orthographic choice. American Ethnologist 21, no. 1: 176–200. Seuren, Pieter. 1998. Western linguistics: An historical introduction. Oxford: Blackwell. Singler, John. 1996. Theories of creole genesis, socio-historical considerations, and the evaluation of the evidence: The case of Haitian Creole and the relexification hypothesis. Journal of Pidgin and Creole Languages 11, no. 2: 185–230. Speas, Margareth. 2011. Language ownership and language ideologies. Manuscript available online at http://people.umass.edu/pspeas/lgownership.pdf. Spolsky, Bernard, and Francis Hult, eds. 2008. Handbook of educational linguistics. Malden, MA: Blackwell. Thornton, John. 1993. “I am the subject of the King of Congo”: African political ideology and the Haitian Revolution. Journal of World History 4, no. 2: 181–214. Trouillot, Hénock. 1962. Les origines sociales de la littérature haïtienne. Port-au-Prince, Haiti: Imprimerie N.A. Théodore. Trouillot, Michel Rolph. 1995. Silencing the past: Power and the production of history. Boston: Beacon Press. ———. 2002. Culture on the edges: Caribbean creolization in historical context. In From the margins: Historical anthropology and its futures, ed. Brian Keith Axel, 189–210. Durham, NC: Duke University Press. UNESCO 1953. The use of vernacular languages in education. Monographs on Fundamental Education, 8. Paris: UNESCO. United Nations. 2007. United Nations declaration on the rights of Indigenous peoples. Available at www.un.org/esa/socdev/unpfii/documents/DRIPS_en.pdf. ———. 2009. The UN declaration on the rights of Indigenous peoples, treaties and the right to free, prior, and informed consent: The framework for a new mechanism for reparations, restitution, and redress. Available at www.un.org/esa/socdev/unpfii/ en/EGM_A42.html. Valdman, Albert. 1984. The linguistic situation of Haiti. In Foster and Valdman 1984, 77–99. Walter, Stephen. 2008. The language of instruction issue: Framing an empirical perspective. In Spolsky and Hult 2008, 129–146. Warner-Lewis, Maureen. 2003. Central Africa in the Caribbean: Transcending time, transforming cultures. Kingston, Jamaica: University of West Indies Press. Weinreich, Uriel. 1958. On the compatibility of genetic relationship and convergent development. Word 14: 374–379. Zefi, Lemèt. 2011. Edikasyon ak lang nan divès pwopozisyon plan rekonstriksyon d Ayiti [Education and language in various proposals for the reconstruction of Haiti]. In Entre refondation et reconstruction: Les problématiques de l’avenir postsismique d’Haïti, ed. Wilson Dorlus, 257–264. Port-au-Prince, Haiti: Editions de l’Université d’Etat d’Haïti.

contributors

christopher ball is Assistant Professor of Anthropology at the University of Notre Dame. He completed the Joint PhD in Linguistics and Anthropology at the University of Chicago. He conducts linguistic and ethnographic fieldwork with the Arawak-speaking Wauja people of the Upper Xingu of Brazil. He writes in the areas of language in culture, the political economy of language in society, and ritual performance. j. clancy clements is Professor of Linguistics and of Spanish and Portuguese at Indiana University–Bloomington. His main areas of interest are contact linguistics, especially pidgins and creoles, and functional syntax, with a focus on Iberian varieties of Romance languages. His most recent publications include The Linguistic Legacy of Spanish and Portuguese: Colonial Expansion and Language Change (2009), various edited volumes, and many articles on language-contact phenomena and the cognitive processes that account for them. hildo honório do couto is “Pesquisador Associado” in the Department of Linguistics at the University of Brasilia, Brazil. He is the author of O Crioulo português da Guiné-Bissau (1994), Introdução às línguas crioulas e pidgins (1996), Linguística, ecologia e ecolinguística (2009), and O tao da linguagem (2012). He was the editor of the Brazilian journal of creole languages Papia for twenty years. He is currently working on ecosystemic linguistics, the Brazilian branch of ecolinguistics that includes ecological discourse analysis. michel degraff is Associate Professor of Linguistics at the Massachusetts Institute of Technology. His research is primarily on the development and structures of creole languages, especially his native Haitian Creole. He is the author of several essays in which he argues that creole formation is a normal instance of language change, and he is the editor of Language Creation and Language Change: Creolization, Diachrony, and De-

[ 330 ]

Contributors

velopment (1999). Recently he embarked on an NSF-funded project using Haitian Creole and educational technology to enhance the active learning of math and science in Haitian universities. alan durston is Associate Professor of History at York University in Toronto. He has published Pastoral Quechua: The History of Christian Translation in Colonial Peru, 1550–1650 (2007). He is currently working on a book manuscript entitled The Politics of Quechua Letters: Language and Nation in Modern Peru, 1900–1950. m. kittiya lee is Assistant Professor of History at California State University–Los Angeles. She has published essays and articles on interethnic relations during the Portuguese colonization of Brazil. Currently she is working on a book manuscript about the historical development of indigenous lingua francas in colonial Brazil. john m. lipski is the Edwin Erle Sparks Professor of Spanish and Linguistics at Pennsylvania State University. His research includes the study of language contacts affecting the evolution of Spanish and Portuguese, including the contribution of the African diaspora and Ibero-Romance lexified creole languages, as well as contact with indigenous languages and languages of voluntary immigrants. He is also the acquisition editor for the Georgetown University Press Hispanic Linguistics series. His most recent books include El habla de los Congos de Panamá (2011), Varieties of Spanish in the United States (2008), and A History of Afro-Hispanic Language (2005). heliana mello is Associate Professor of English and Linguistics at Universidade Federal de Minas Gerais, in Brazil. Among her most recent publications are the books Os contatos linguísticos no Brasil (2011), CORAL-BRASIL: Corpus de referência do português falado informal (2012), and Pragmatics and Prosody (2012). She is currently coediting Spoken Corpora, which is to be published by John Benjamins in 2014. denny moore is the coordinator of the Área de Linguística of the Museu Paraense Emílio Goeldi, a research institute of the Brazilian Ministério de Ciência, Tecnologia e Inovação in Belém, Brazil. A former MacArthur fellow and current associate editor for the International Journal of American Linguistics, he has published several articles on the situation of the indigenous languages of Brazil and is heavily involved in language documentation and preservation in Amazonia. He coordinates a collec-

Contributors

[ 331 ]

tive study involving nine linguists, which focuses on Tupian languages and their prehistory. salikoko s. mufwene is the Frank J. McLoraine Distinguished Service Professor of Linguistics and the College at the University of Chicago. His books include The Ecology of Language Evolution (2001), Créoles, écologie sociale, évolution linguistique (2005), and Language Evolution: Contact, Competition, and Change (2008). He is the series editor of Cambridge Approaches to Language Contact. His current research is on various aspects of language evolution, including the effects of colonization and globalization. barbara pfeiler is Professor of Linguistics at the National Autonomous University of Mexico in Mérida, Yucatán. She has conducted sociolinguistic and ethnographic fieldwork in Yucatan since 1980. She has published in the areas of sociolinguistics, dialectology, and language acquisition of Mayan languages and is the author of Voces Mayas: Ethnografische und soziolinguistische Aufzeichnungen zur Zweisprachigkeit in Yukatan (2012) and coauthor, with Glen Ayres, of Los verbos mayas: la conjugación en el maya yucateco moderno (1997).

subject index

Academy, 308, 312–313, 321 adjective, 86, 88, 119, 126–127, 130, 135, 178, 213, 215–217, 219 adverb, 22, 68, 126–127, 132–136, 140, 178, 213, 216, 218 adverbial, 68, 132–136, 140 Africa, 2–5, 7–10, 13, 17–20, 23, 27–28, 30–33, 46, 56–59, 61–63, 67, 80, 111, 113, 115, 143, 145–146, 151, 154– 156, 160, 169–177, 179, 182–183, 191, 193–199, 201–202, 277, 281, 284–285, 288, 292–294, 296–304, 307, 310, 315, 317–318, 320–321 Afro-Hispanic, 56, 59–61 alignment system changes, 118 alliances, 110, 159–160, 257–258, 264 Amazon region, 85, 91, 112, 170; Amazon River, 84, 89, 95, 112–113, 119–120, 147, 155 Amazonia, 5, 13–14, 16–17, 24, 89–91, 98, 112–120, 131, 136, 143–146, 148, 156, 158–159, 161–162, 245, 247, 249–250, 254, 260, 269–271, 279, 298, 301, 308 Amazônica, the, 13, 16 American settlement, 41, 66 Amerigo Vespucci, 152, 163 Amerindian, 13, 77–78, 80, 82–83, 86, 90–92, 95, 97–100, 102–104, 169, 172, 174–175, 277–280, 282, 286, 288, 290, 292, 294–296, 301, 303, 306–307, 320 Andalusian, 8, 12, 39–43, 46, 65, 68 Andean Spanish, 48–52, 54–55, 68, 277–278

Andes, 47, 226, 239–240, 277–278, 285, 288–290, 292–294, 296, 304–305, 307–308, 310, 313–314, 319 Arapaço, 252–255, 294, 303 Arawak, 110, 113, 152–153, 163, 169, 249, 251, 254, 256, 260, 264–265, 269, 278, 296 Argentina, 8, 11, 27–28, 46, 49, 54–55, 57, 66–68, 100, 162 Arraia Grande, 159 Arte da Grammatica, 112 Arte da Língua Brasílica, 112 aspectual, 136 auxiliaries: colonial, 17, 24; verbal, 130, 136 Ayacucho, 232, 240 Aymara, 6, 24, 48, 50–52, 226–228, 230, 235, 240, 277, 286–287, 306 Bahia, 20, 81, 84–85, 100, 109–112, 149, 169–171, 173, 176, 183, 193–196, 198 bandeirantes, 84, 86, 172 Baniwa, 82, 89–90, 95, 97–98, 101, 109, 136 Bantu, 33, 59, 63, 171, 177, 181, 193, 203, 277–278 Baré, 82, 91, 95, 100, 109 Bay of All Saints, 143, 155, 162 Bay of Guanabara, 149, 152, 158, 163 Belém, 95, 112–113, 120, 122, 131, 136, 158 Berlin Treaty, 2, 5 Bertioga, 150 bilingual, 18, 22, 25, 47–48, 50, 52–53, 67, 80, 86, 97–98, 104, 116, 120,

[ 334 ]

Subject Index

bilingual (continued) 123, 136, 138–139, 207, 209–212, 216–222, 239, 249–250, 263–265, 278, 284, 286–287, 305, 309, 311, 313–314; bilingual education, 139, 221, 286–287, 305 bilingualism, 53, 138, 207, 211, 221, 311 black, 13, 18–20, 23, 56–60, 62, 81, 86, 111, 113, 120, 145, 171–172, 181–182, 193, 196–203, 284–285, 292–293, 297–301, 303, 305, 321 Boçais, 19, 171, 284–285, 297. See also Bossales; Bozal Bolivia, 40, 49, 51, 55, 60, 67, 226, 229, 233, 239, 286–287, 305, 307, 313 borrowing, 14, 22, 31, 41, 46, 68, 87, 114–117, 119, 122–214, 126, 128, 132, 135, 138, 154, 212, 214, 216, 219, 222, 256 Bossales, 285. See also Boçais; Bozal bovarysme, 306 Bozal, 10, 19, 57–58, 60, 62, 64, 171, 183, 284. See also Boçais; Bossales Brasílica, 6, 13, 16–17, 83–84, 89, 112–115, 117, 143–146, 153, 155, 157, 161, 170, 183, 288, 290–291. See also Língua Brasílica Brazil, 2–7, 9–10, 13–16, 18–20, 24, 26–30, 32–33, 55, 61, 67, 77, 80–85, 87–90, 92, 95, 97–105, 109–117, 119–120, 138–140, 143, 145–155, 157–158, 160–163, 169–183, 188, 191–197, 200–203, 239, 253, 259– 262, 265–266, 268–269, 271, 277, 284–285, 288, 290, 295, 305 Brazilian: government policies, 139; national census, 140 Brazilian Portuguese, 4, 9, 18–19, 61, 88, 90, 175, 177, 181, 192, 202, 277 brazilwood, 88, 109, 146–149, 151–154, 158, 160, 170, 172 Bretôa, 151 Buenos Aires, 11, 46, 53–56, 66, 68, 233, 277 Cabanagem, 89, 120 Caeté, 159 Caipira, 88

captives, 110–111. See also slaves Caramuru, 100, 109 Carib, 110, 113, 249, 260, 264, 278 Carijó, 158 Castilian, 7–9, 11–12, 26–27, 30, 40–41, 46, 65, 291, 304 Catalan, 8, 30, 65 catechism, 17, 112, 114, 116, 143, 157, 228, 232, 236, 289–291 Ceará, 112, 198 Chinook Jargon, 6, 122 Christopher Columbus, 2, 152, 163, 288, 292 Cocoliche, 11, 68 code-mixing, 31 colleges, 111, 235 colloquial, 24, 114, 190, 211, 216, 221, 309 colonialism, 239, 254, 270–271, 275, 282 colonization, 2, 5, 7–8, 12–13, 16, 18, 22, 24–25, 28–29, 31–32, 41–42, 80, 82–83, 91, 103, 113–114, 143, 146, 148, 157, 159–160, 281, 288, 291, 306 communion, 82, 104 community, 12, 48, 68–69, 77–79, 86, 91, 98–99, 111, 136, 145, 179, 201, 207, 211, 221, 245, 247–248, 251– 252, 256, 258–260, 262, 264–267, 270–271, 279, 298–299, 302, 310, 318 comparative approach, 7, 9, 32 comparative method, 275 complementizer, 122, 128, 135, 278 complements, 134 complexity, 9, 11, 122, 169–170, 179, 277–282, 286 compounds, 128, 217–218 Conciliar Quechua, 23, 230 conjunctions, 90, 119, 212 contact varieties, 278 coordination, 135 copula, 130, 140 creole (language), 4–6, 9–10, 16, 18–20, 23, 31–34, 40, 59, 62–64, 66–67, 69, 80–81, 85, 91, 98, 101, 104–105, 138, 171, 175, 177, 182, 206, 275–278, 280, 282–287, 289–290, 292, 294,

Subject Index

296–301, 303–305, 309, 311–314, 316–318, 320–321; Creole Exceptionalism, 275, 282, 285, 318 Criollo, 85 cross-linguistic influence, 173, 212. See also interference Cuba, 3, 5, 9–10, 19, 39, 57–64, 66, 68, 80, 277 cultural subjectivity, 295 Cuzco, 22, 53, 227–228, 230–233, 236–238, 240 dead language, 99–101 degredados, 18, 28, 170–171, 183. See also lançados demographic information, 200 demonstrative, 118, 126–127, 135 derivation, 22, 127–129, 213; derivational affix, 127–129 descending, 2, 13–14, 17, 33, 92 dialectology, 176, 226–227, 240 diglossia, 114, 236 disease, 111, 113, 119, 194, 288 diversity: linguistic, sociolinguistic, or cultural, 11, 103, 227, 245, 249, 269–270 DOBES, 139 Dominican Republic, 3, 5, 9–10, 19, 59, 61, 63–64, 67, 80, 277 Dominican Spanish, 59, 63–64, 278 drogas do sertão, 5 Dutch, 3–5, 10, 17, 20, 109, 146, 158, 193, 278, 310 ecolinguistics, 77–79 ecological factors, 5, 7, 9, 17–18, 24, 27, 29, 32, 245, 249, 276–277 economics, 138–139, 292; economic alternatives, 139 economy, 11, 21, 172, 229, 256, 261 ecosystem, 77–78, 99, 103–104, 249 Ecuador, 22, 32, 50, 60, 226, 239–240, 305 education: bilingual, 139, 221, 286–287, 305 elite closure, 284, 289 embedded clause, 16, 132, 134 embedded question, 134 enclave, 13, 65, 146

[ 335 ]

Endangered Languages Documentation Programme, 139 endangerment, 6–7, 13, 30, 34, 102, 139, 269–270, 275, 293, 310–311, 316, 319 endogamy, 260, 264–265, 279 engagement, 147, 149, 157, 161, 248, 316, 318 ethnic, 2, 11, 15, 26–27, 33, 46–48, 52, 57–58, 78, 81, 85, 89, 92–94, 97–98, 100–102, 110, 139, 143–147, 149–150, 154, 160–162, 195, 250, 254, 258–260, 264–266, 268, 279, 281, 283, 293–294, 297, 299–300, 303, 321 European Portuguese, 28. See also Vernacular European Portuguese evidential, 49, 51, 90, 119 evolution, 6–9, 11–12, 15–16, 18, 24, 26–28, 30–33, 40, 52, 67, 77, 91, 102, 109, 112, 114–115, 117–118, 138, 153, 161, 177, 232, 245, 269–270, 275–277, 280, 282, 290, 293, 296, 308, 311, 314, 320 exchange, 24, 28, 82, 98, 109, 146, 151, 160, 238, 245, 250, 252, 254, 256, 260, 262, 266–268, 271, 287 exogamy, 25–26, 98, 251–253, 255, 258–259, 263–264, 279, 294–295 expansion, 2, 5–6, 12–13, 15–17, 21, 30, 92, 98, 110, 117, 143–144, 157, 159–161, 226, 288–290, 292–293, 320 factor (as Portuguese feitor), 14, 17–18, 28, 32, 152, 154 factory, 32, 151–52 Fort St. Louis, 112 French academy, 312 functional convergence hypothesis, 278 Galician, 8, 65 Gê, 143, 155, 160 genitive, 135 glides, 117, 120, 125 Global North, 276, 314 Global South, 276 glottal stop, 117, 125, 217 gold rush, 5, 17

[ 336 ]

Subject Index

grammaticalization, 119, 129 Guarani, 16, 77, 81, 84–85, 92, 98, 100–101, 104, 109–110, 115, 117, 124, 133, 143–147, 149–163, 169–170, 291 Hach Maya, 114, 211 Haiti, 3–4, 30–31, 57–58, 63–64, 67, 80, 275–78, 282–289, 292–294, 296, 298–302, 304–309, 311–316, 318, 321 Haitian Creole, 63–64, 67, 275, 277– 278, 282, 311, 313, 318 Haitian Revolution, 57, 293, 298–99, 301, 321 Hakani, 139 Hawaiian Creole, 277, 290 Hevea cultivation, 5 Huancavelica, 229–230 I-language, 278 Iberoamerica, 3, 5 immigration, 21, 46, 53–56, 65–66, 68, 90, 120, 172, 174–175, 207 Inca Empire, 6, 22–23, 53, 227, 288, 293 independence, 18, 56, 119–120, 233, 298, 300–301 Indian slaves, 195 indigenous: group, 15, 17, 109–110, 113, 115–116, 139, 147, 162, 253, 261, 268, 288; language, 7, 13–18, 24–25, 28–31, 46–49, 52–53, 66, 102, 104, 112–113, 115, 119, 123, 138, 140, 169, 171, 183, 202, 220–221, 237, 245, 249, 255–256, 261, 265, 269, 275, 282–283, 286–288, 290–294, 296, 304–307, 310, 313–314, 316–318, 320; societies, 143, 152, 155; tribe, 111 inflection, 22, 49, 61, 127–129, 182, 216, 222, 280 inter-ethnic, 110 interference, 115–117, 138, 232. See also cross-linguistic influence interpreter, 14, 16–18, 23, 28–29, 32, 81, 91, 112, 150, 154–155, 170, 234–235, 263, 271. See also línguas interrogative word, 132

intransitive verbs, 118, 126, 214, 216, 221–222 Irish, 17, 146, 158 Jesuit, 29, 83, 86, 89, 104, 109, 111– 113, 115–117, 119, 143, 157, 159, 170, 234, 236, 288, 290, 310 João Lopes de Carvalho, 152, 154 JOCUM, 139 Jopará, 85, 100 just war, 111 King Manuel, 147–148 koinés, 39, 170, 228 Kreyòl, 64, 275, 282–89, 293–294, 297–316, 321 labiovelar stop, 125 lançados, 18. See also degredados language documentation, 139–140 language enclave, 13. See also enclave; language island language endangerment. See endangerment language family, 110, 143, 145–147, 155, 157–160, 162 language island, 13, 15, 77, 92, 95, 98–99, 146. See also enclave; language enclave; population island language manual, 112, 114 language speciation, 6 language vitality, 7, 275 Latin America, 3, 5–9, 11–12, 18, 26–28, 30, 39–44, 46–49, 52, 56, 59, 65–68, 245, 269, 271, 275–277, 279, 281– 283, 287, 290, 293, 298, 301–302, 305–307, 310–311, 314, 316, 320 lenguas, 23, 28, 221, 234. See also interpreter; línguas lexical, 14, 16, 18, 22, 31, 32, 46, 59, 61, 67, 87–88, 101, 116–117, 119, 121–122, 129, 138, 146, 170, 189, 201, 219, 222, 226, 228, 231–232, 249, 256 lexicon/lexica, 19, 54, 122, 152–154, 170, 177, 206, 222, 233, 280 LGA, 14, 16–17, 89–91, 100, 102, 114–120, 138, 145, 304. See also Língua Geral Amazônica

Subject Index

Lima: archdiocese of Lima, 231, 234, 236–238, 240; Lima Council, 227–232, 234, 237, 240 Língua Brasílica, 6, 13, 83–84, 89,112, 170, 183, 288. See also Brasílica; língua geral Brasílica lingua franca, 6, 11, 13, 15, 17, 22–25, 81, 83–85, 90–92, 94–95, 97, 101– 102, 104, 115, 117, 143–145, 157, 161, 169–170, 176, 182–183, 205, 227, 229–230, 236, 238–240, 252, 254, 268–269, 277, 280, 288–291, 293, 299, 304–305 Língua Geral, 6, 13–17, 24, 32, 77, 81, 83–85, 88–89, 97, 101–102, 104, 112–114, 116, 143–146, 254, 288, 290–291, 304 Língua Geral Amazônica, 13–15, 77, 85, 88–89, 101–102, 104, 108, 112–113, 144–145, 290–291, 304. See also Nheengatu língua geral Brasílica, 17, 143. See also Brasílica; Língua Brasílica Língua Geral Paulista, 14, 77, 85, 101–102, 104,108,114, 145, 290–291 linguas, 16. See also interpreter; lenguas línguas, 83, 93, 112, 139. See also interpreter; lenguas literacy, 231, 236, 262, 285, 287, 306, 309 liturgy, 16, 33, 114 live burial of children, 139 lower castes, 21, 205 Madeira River, 103, 121, 159 Manaus, 82, 91, 95, 120 Maranhão, 84, 88–89, 112–115, 120, 143, 159, 173, 176, 198 matrix clause, 129–31 Maya, 11, 21–22, 24, 27, 31, 110, 112, 114, 129, 138, 205–222, 275, 278, 286, 288–290, 304–307, 309–312 Media Lengua, 6, 14, 22, 31–32 mestizo, 14, 17, 21, 28–29, 47–48, 83, 85–86, 89, 109, 111–113, 115–116, 154, 172, 205–206, 228, 239, 295 Mexico, 3–4, 11, 21, 28, 39–40, 46, 56, 60, 66–68, 205–207, 210–211, 216, 221

[ 337 ]

Middle Rio Negro, 97, 100, 121 Minas Gerais, 5, 17, 20, 84–86, 99, 103, 145, 163, 173, 176, 196, 198–200 Mining, 23, 56, 160, 182, 193, 196–197, 201, 229 mission villages, 111 missionary/missionaries, 13, 15, 17, 21, 23, 29, 33, 84, 86–87, 89–91, 100, 111–113, 115–116, 139, 143, 157, 159, 205, 226, 253–254, 256, 288–291 Montevideo, 11, 46, 53–56, 66, 68–69, 277 mortality rates, 113 multiple VPs, 129 Munduruku, 121 Mura, 121 Museu do ĺndio, 139 Museu Goeldi/Museu Paraense Emílio Goeldi, 122, 136, 271 nasalization spreading, 124 nasals, 120, 122, 124 native language, 47, 91, 109–110, 112, 138–139, 172, 176, 205–206, 209, 275, 284, 286, 305, 309, 314, 316, 319 negation, 58, 61–64, 69, 103, 131, 177, 190–192, 200, 202, 213, 277–278, 306 negative purpose subordination, 133 negros boçais, 19, 171, 284–285 negros de reino, 19, 28 negros ladinos, 19, 28, 171, 284–285 New Spain, 2–4, 30–31, 206 Nheengaíba, 91, 115 Nheengatu, 6, 14–16, 85, 88–91, 95, 97–98, 100, 102, 104, 109, 118–123, 125–130, 132–136, 138–140, 144– 145, 170, 254, 275, 280, 289–291, 302, 305. See also LGA; Língua Geral Amazônica nominalization, 119, 280 nominal series, 118, 124, 126–127, 129 numerals, 90, 127, 178, 213 objects of transitive verbs, 118 oral voiced stops, 124

[ 338 ]

Subject Index

palatal nasal, 124 palatalization, 123, 192 Palenquero, 4, 19, 59–60, 63 pardos, 19–21, 28, 193, 197, 199–203, 205 particle, 49, 52, 63, 68, 90, 119, 126–127, 130–133, 136, 140, 212–213, 217 Pedro Álvares Cabral, 82, 143 Peninsular Portuguese, 4, 19 periodization, 9, 277 Pernambuco, 81, 83, 101, 109, 112, 149, 158–159, 162, 176, 193, 195–196, 198, 288 person hierarchy, 118 person-marking system, 118 Peru, 3, 23–24, 40, 49–50, 53, 84, 90, 104, 226–236, 239–240, 305–307, 310, 313 phonemic inventory, 117–118, 120, 123 pidgin, 11, 16, 18, 23, 28, 32, 47, 57–58, 61, 85, 91, 101, 104, 138, 229 plural marker, 90, 119, 280 polar questions, 132 polytypic languages, 187–188 population genetics, 186–187, 201 population island, 15. See also language island Portugal, 2–4, 19, 30, 111–114, 119, 143, 148, 170–172, 183, 188, 191, 193, 202, 235, 240 Portuguese. See Brazilian Portuguese; European Portuguese; Peninsular Portuguese; Vernacular Brazilian Portuguese; Vernacular European Portuguese Portuguese settlement, 18, 112, 155, 157, 162 postoralized nasals, 122, 124 postpositions, 118, 126–127, 136 Potosí, 229, 236 predicate adjective phrase, 126 predicate noun phrase, 130 prenasalized voiced stops, 124 ProDocLin/Projeto de Documentação de Línguas Indígenas, 139 pronouns, 42, 126–127, 189–192, 200 Quechua/Quichua, 6, 14, 22–24, 31–32, 41, 48, 50–53, 68, 83, 114–115, 138,

226–240, 275, 277–278, 286–293, 304–307, 310, 312–313, 319–320 Quilombo, 4, 10, 19, 176 Ramalho, 87, 109 reanalysis, 68 reduplication, 129 Regimento das Missões, 113, 119, 304 relative clause, 50, 119, 133–134, 140, 280 relativizer, 128, 133, 140 replication, 119, 122, 138 resettlement, 111, 115, 117, 227 Rubber Boom, 120, 138 São Gabriel da Cachoeira, 89–90, 97, 138, 255, 259, 271 São Luis, Maranhão, 112 São Vincente, 109 settlement, 2–5, 13–15, 18, 30–31, 33, 39–42, 47, 66, 109–110, 112–115, 119, 143, 146–148, 150, 155, 157–158, 162, 182, 250, 253–254, 256, 258, 265–266; Spanish settlement, 40–42, 47 sibling languages, 187–88 slave raids, 26, 110, 113, 251, 253, 260 slaves, 2, 4–5, 9–10, 17–20, 32, 56, 58, 63, 110–111, 113, 160, 170–172, 174–176, 182–183, 193–196, 201, 284–285, 292, 297, 299–300, 320. See also captives Solimões River, 98, 121 Spanish, 3–12, 14, 21–23, 26–28, 30–32, 39–44, 46–69, 83, 85, 97–98, 100, 113, 128, 136, 146, 148–149, 151, 157–159, 194, 205–207, 209–222, 226, 228–230, 233–236, 238–239, 253, 269, 276–278, 283, 287–293, 302–307, 309–313. See also Andean Spanish; Dominican Spanish Standard Colonial Quechua, 22, 114– 115, 289, 310, 313. See also Quechua/ Quichua standardization, 230, 313 stative verbs, 118, 126 steamship, 16, 120 stress, 52–54, 59, 61, 123, 125, 217, 321

Subject Index

subjectivity, 24–25, 245–255, 257–260, 262–264, 266–267, 269–271, 295 subordinate clauses, 119, 132 subordinating particle, 133, 136 subordination, 133, 135 sugar plantations, 58, 110–111, 182, 193 sugarcane cultivation, 3–5, 17, 30 Survival International, 139 SVO, 16, 90, 118 syllable pattern, 126 syntax, 16, 33, 46–47, 122, 129, 177, 280, 318 Tapuia, 81, 90–91, 109, 143 Tariana, 95, 97, 100, 134, 254, 258 topicalization, 131, 178 trade language, 115, 154 transfer, 41, 43, 48, 50–52, 54–55, 68, 112, 152, 211, 216–221 transitive verbs, 118, 126, 130, 134, 214–216, 219, 221–222 Treaty of Tordesillas, 2, 4, 148 Tukano, 25, 90, 95, 97–98, 100–102, 104, 113, 123, 251–258, 269, 271, 278, 286, 294–295, 303 Túpac Amaru II, 291, 304, 321 Tupi, 81–92, 100, 103–104, 109–110, 121, 149, 264–265 Tupi-Guarani, 16, 77, 81, 84–85, 92, 98, 101, 104, 109–110, 115, 117, 124, 133, 143–147, 149–151, 153–155, 157–163, 169–170, 291; Tupi-Guarani Indians, 117, 144–145, 155–157, 159–160, 162 Tupinaé, 16, 110, 143, 160 Tupinambá, 6, 13–15, 17, 24, 28, 81, 84–85, 89–91, 100, 104, 110, 113–118, 122–123, 127–129, 138, 143, 149–151, 153–155, 158–160, 162–163, 169, 280, 291 UNESCO Atlas of the World’s Languages in Danger, 139 unsung heroes of colonization, 18, 28

[ 339 ]

Upper Rio Negro, 14, 89–90, 95, 97–98, 100, 102, 105, 109, 118, 120–122, 136, 271 Upper Xingu, 14, 26, 93, 247, 249–50, 260–271, 279, 303 Uruguay, 11, 27, 55–57, 66–67, 100, 147 variant, 21, 69, 101, 103, 110, 120, 140, 145, 206, 240 VBP. See Vernacular Brazilian Portuguese VEP. See Vernacular European Portuguese verb object order, 135 verbal series, 118, 126, 129 vernacular, 3–6, 8–11, 14–15, 18–20, 22–24, 26–27, 29, 33, 40, 49, 59, 61, 80, 91, 98, 114, 144, 154, 171, 174, 176, 180–181, 188, 192, 202, 226, 229–230, 235–236, 238, 262, 289, 294, 310, 312 Vernacular Brazilian Portuguese, 4, 19–20, 61, 175, 177–183, 187–192, 200–202 Vernacular European Portuguese, 171–172, 175, 177, 181, 190–192, 200–202 voodoo economics, 292 Vulgar, the (brasilica vulgaris), 104, 108, 117, 144, Vulgar Latin, 23, 27, 33 Warekena, 95, 109 WH words, 134 word class, 126–127 word list, 17–18, 162 word order, 22, 48, 50–51, 118, 178 Xe’ek’ Maya, 114, 211 Xingu Park, 139, 261 Yucatán, 11, 21–22, 24, 27, 31, 138, 205–207, 210–212, 216–217, 219– 222, 288, 290, 304, 307–312 Yucatán Maya, 21, 24, 31, 138, 207

author index

Aboh, Enoch, 27, 275, 280, 286, 302 Aikhenvald, Alexandra, 134, 249, 251, 254, 256, 271 Anchieta, Joseph de, 83, 111–112, 154, 170 Armstrong-Fumero, Fernando, 211 Ball, Christopher, 14–15, 24–26, 29, 139, 158, 268, 271, 278–280, 294–296, 298, 301–303 Bang, Jørgen Christian, 78 Barrera-Vásquez, Alfredo, 216, 218 Barros, Maria Cândida D. M., 87, 115–117, 157, 261 Beauvoir, Max, 299–302, 312–313, 320 Beauvoir-Dominique, Rachel, 301, 320 Bettendorff, João Felipe, 114, 157–159, 163 Borges, Luiz, 122, 124 Bricker, Victoria, 214–215, 222 Cabral, Ana Suelly, 82, 98, 117, 143, 170–171 Casas, Bartolomé de las, 29, 31 Cerrón-Palomino, Rodolfo, 227–231, 240 Chaudenson, Robert, 8–10, 12, 19, 21, 28, 31–32, 80 Chernela, Janet, 251–254, 258–259 Clements, Clancy, 2, 9, 19–20, 28, 30, 69, 277–278 Coronel-Molina, Serafín, 275, 305, 314–315 Couto, Hildo Honório do, 6, 13–19, 29, 77, 79, 82, 88, 104, 116, 120,

144–146, 152, 157, 163, 170, 239, 282, 288–291, 293–295, 310 Cruz, Aline, 62, 109, 117, 122, 126, 131 Daniel, João, 115–117 Dayan, Joan, 298–300, 320–321 DeGraff , Michel, 18, 30, 32, 275–276, 280, 282–284, 286–287, 296–298, 302, 306, 308, 311–313, 315, 318 Dejean, Yves, 283–284, 287, 302, 306, 313 Descourtilz, Michel Etienne, 300, 320–321 Dessalines, Jean-Jacques, 298–301, 303, 306, 320 Devonish, Hubert, 282–283, 286, 292, 296, 317 Døør, Jørgen, 78 Dubois, Laurent, 288, 299 Durston, Alan, 22–24, 29, 114–115, 226– 232, 234–235, 237–240, 277–278, 288–291, 293, 304–305, 310, 313 Eakin, Marshall, 2, 31 Edelweiss, Frederico, 144–145 Emmerich, Charlotte, 92–94 Escobar, Anna María, 6–7, 48 Fayer, Joan, 29 Felix, Maria Invanete de Santana, 120, 122 Figueira, Luís, 112, 114, 157 Fouchard, Jean, 288, 299, 320 Franchetto, Bruna, 139, 262 Freire, José Bessa, 16, 91, 104, 113–114, 144–145, 157, 162, 171

[ 342 ]

Author Index

Gabbert, Wolfgang, 21, 205–206 García, María Elena, 61, 208, 217, 221, 287, 291, 293, 304, 306, 314 Geggus, David, 299–300, 304, 320 Godenzzi, Juan Carlos, 275, 287, 307, 310, 313–314 Haboud, Marleen, 307, 310, 314, 319 Hale, Kenneth, 307, 310, 314, 317–319 Heine, Bernd, 53, 119 Hemming, John, 109–111, 113, 115 Heywood, Linda, 299, 301, 320 Hinton, Leanne, 307, 310, 314, 317, 319 Holanda, Sérgio Buarque de, 83, 104 Hornberger, Nancy, 275, 287, 291, 293, 305, 307, 314–315 Hugh-Jones, Stephen, 250–251, 253, 271 Itier, César, 24, 228–231, 233, 235–236, 239–240 Jackson, Jean, 250–251 Jenson, Deborah, 299–301, 303, 306, 320–321 Kaufman, Terrence, 115–116, 118 King, Kendall, 30, 33, 287, 291, 293, 307, 314 Kuteva, Tania, 53, 119 Labov, William, 81 Langfur, Hal, 155, 160 Lasmar, Christiane, 252, 259 Lee, M. Kittiya, 6, 13–14, 16–18, 28–29, 32, 84, 89, 110, 112–115, 117, 170, 254, 278, 288–291 Leite, Yonne, 118, 157 Léry, Jean de, 149–153, 158, 160, 162–163 Lipski, John, 7–9, 11–12, 18, 20, 27, 39, 55, 57–58, 60–61, 63–64, 67–68, 211, 276–278, 280, 290, 302 Lois, Ximena, 214 Lope Blanch, Juan M., 216–217, 219 Lopes de Sousa, Pero, 82, 151–152, 155 Louverture, Toussaint, 299–300, 303 Lucchesi, Dante, 176–181, 183, 202 Luykx, Aurolyn, 312, 316, 318

Mal-Molinero, Clare, 8, 12 Mannheim, Bruce, 68, 162, 202, 226–229, 231–233, 235, 239–240 Martínez Cobo, José, 282, 308, 316–317, 320 Mathieu, Suze, 275, 285, 314 McCarty, Teresa L., 287, 308, 315, 319 Meader, Robert S., 100–101 Mello, Heliana, 9–10, 18–19, 28, 171, 177, 180–181, 183, 189–191, 200, 277–278, 284–285, 297 Monserrat, Ruth, 114, 117, 120 Monteiro, John, 143, 147, 156–158, 162 Moore, Denny, 6, 13–18, 29–30, 33, 85, 89, 104, 116, 129, 139, 144–145, 154, 157, 163, 170, 254, 278, 280, 288–290, 302, 304 Morales Ayma, Evo, 287, 307, 313–314 Morrison, Toni, 305 Moñino, Yves, 19, 30 Mufwene, Salikoko S., 4–5, 7–13, 18, 20–21, 27–28, 32–34, 40, 77, 79–80, 100, 104, 154, 162, 187, 201–202, 245, 271, 275, 277, 294, 296, 302, 309–310, 320 Naro, Anthony Julius, 19–20, 28, 179, 190, 192, 202 Negrão, Esmeralda, 9, 20 Noelli, Francisco Silva, 144, 147, 157, 161 Nunes, Diogo, 155 Ostler, Nicholas, 30, 307, 310, 314, 319 Parker, Gary, 226, 240 Pfeiler, Barbara, 10–11, 21–22, 27–29, 110, 112, 114, 129, 207, 209, 211, 278, 288–289, 304, 306–307, 309–310 Pigafetta, Antonio, 149–153, 162–163, 170 Pilkey, Dav, 292 Pompa, Cristina, 155, 157 Price-Mars, Jean, 288, 306, 320–321 Quijano, Aníbal, 7

Author Index

Roberts, Peter, 277, 282–283, 286, 290 Rodrigues, Aryon Dall’Igna, 82, 84–86, 89–90, 104, 109, 118, 120, 127–128, 139, 143, 145, 162–163, 169, 172 Saint-Fort, Hugues, 285, 287 Sampaio, Teodoro, 88, 159 Scherre, Maria Marta Pereira, 20, 179, 190, 192, 202 Schwartz, Stuart B., 19, 30, 39, 155, 182, 193–196 Shappeck, Marco, 14, 22, 31–32 Silva Neto, Serafim da, 171 Silverstein, Michael, 122, 246–248 Soares de Sousa, Gabriel, 145, 158 Staden, Hans, 149–154, 158, 162

[ 343 ]

Stenzel, Kristine, 254, 257–258, 264 Stolz, Christel, 213 Stradelli, Ermana, 116, 120, 122 Taylor, Gerald, 104, 122, 124, 227, 231–232, 234, 239 Thévet, André, 149–150, 154, 158, 162 Thomason, Sarah Grey, 115–116, 118 Torero, Alfredo, 226–227, 229, 231, 237, 240 Trampe, Wilhelm, 78 Vapnarsky, Valentina, 214 Viotti, Evani, 9, 20 Wichmann, Søren, 213