You are on page 1of 296
Copyright 2010 by the Latin American Studies Association International Standard Serial Number US ISSN 0023-8791

Copyright 2010 by the Latin American Studies Association International Standard Serial Number US ISSN 0023-8791

PP5331.indb5331.indb

11

Serial Number US ISSN 0023-8791 P P5331.indb 5331.indb 1 1 LATIN AMERICAN RESEARCH REVIEW volume 45

LATIN AMERICAN RESEARCH REVIEW

volume 45

number 3

2010

editor’s foreword

3

Indigenous Language Usage and Maintenance Patterns among Indigenous People in the Era of Neoliberal Multiculturalism in Mexico and Guatemala hirotoshi yoshioka

5

Our Indians in Our America: Anti-Imperialist Imperialism and the Construction of Brazilian Modernity tracy devine guzmán

35

Reconstructing Indigenous Ethnicities:

The Arapium and Jaraqui Peoples of the Lower Amazon, Brazil omaira bolaños

63

Electoral Revolution or Democratic Alternation? maría victoria murillo, virginia oliveros, and milan vaishnav

87

The Eagle and the Serpent on the Screen:

The State as Spectacle in Mexican Cinema daniel chávez

115

Precursors to Femicide: Guatemalan Women in a Vortex of Violence david carey jr. and m. gabriela torres

142

research reports and notes

Household Shocks, Child Labor, and Child Schooling: Evidence from Guatemala william f. vásquez and alok k. bohara

165

Instituciones políticas y políticas públicas en la Federación brasileña fabiano santos and cristiane batista

187

en la Federación brasileña fabiano santos and cristiane batista 187 1 10/15/10 0/15/10 1 1:52:56 :52:56

110/15/100/15/10

11:52:56:52:56 PPMM

en la Federación brasileña fabiano santos and cristiane batista 187 1 10/15/10 0/15/10 1 1:52:56 :52:56
P P5331.indb 5331.indb 2 2 review essays Mobilizing and Negotiating Meanings: Studies in the Dynamics

PP5331.indb5331.indb

22

P P5331.indb 5331.indb 2 2 review essays Mobilizing and Negotiating Meanings: Studies in the Dynamics of

review essays

Mobilizing and Negotiating Meanings:

Studies in the Dynamics of Colonial-Imperial Transformations in Art, Science, Law, Historiography, and Identities in the Ibero-American World, 1500–1800 sara castro-klarén

213

Shifting Realities in Special Period Cuba archibald r. m. ritter

229

The Transformation of Venezuela harold a. trinkunas

239

Bolivia in the Era of Evo Morales jeffery r. webber

248

The Cold War and Its Aftermath in the Americas:

The Search for a Synthetic Interpretation of U.S. Policy robert a. pastor and tom long

261

Translated Abstracts

274

Notes on the Contributors

278

Forthcoming in Volume 46, Number 1

282

278 Forthcoming in Volume 46, Number 1 282 Editorial and production services for LARR are provided

Editorial and production services for LARR are provided by the University of Pittsburgh Press.

For information about the Latin American Studies Association (LASA), LARR, subscription information, and instructions to authors, please see inside back page.

information, and instructions to authors, please see inside back page. 1 10/15/10 0/15/10 1 1:52:57 :52:57

110/15/100/15/10

11:52:57:52:57 PPMM

EDITOR’S FOREWORD Open Access to LARR for Latin America and the Caribbean Philip Oxhorn McGill

EDITOR’S FOREWORD Open Access to LARR for Latin America and the Caribbean

Open Access to LARR for Latin America and the Caribbean Philip Oxhorn McGill University On behalf

Philip Oxhorn

McGill University

On behalf of Latin American Research Review’s Editorial Committee and the Executive Committee of the Latin American Studies Association, I am pleased to announce that, beginning with this issue, anyone residing in Latin America or the Caribbean will be able to access all current and past LARR issues free of charge through the LASA and LARR Web site. To the best of our knowledge, we are the first among our peer journals to offer this service. The impetus behind this initiative, which was developed in close col- laboration with LASA Executive Director Milagros Pereyra-Rojas and the open-access consultant Enrique Mu, came from the 2009 LASA Congress in Rio de Janeiro. The LARR-sponsored panel “Publishing Your Research” generated a lively discussion of the value of open-access publications to promoting the free exchange of knowledge. This is particularly true for Latin America and the Caribbean, given the scarcity of resources there. For example, subsequent research showed that only 4 percent of universi- ties in Latin America and the Caribbean had access to all LARR issues through their institutional membership to LASA. This very small percent- age is complemented by the access that 2.4 percent of these institutions had to back issues of LARR through Project MUSE and the 17.7 percent that offered their students and faculty the opportunity to purchase indi- vidual articles from back issues through their participation in JSTOR. In addition, of course, the slightly more than two thousand LASA members who resided in Latin America and the Caribbean in 2009 also enjoyed access to LARR. Nevertheless, it is clear that the vast majority of scholars in Latin America and the Caribbean simply are excluded from using the important research published in LARR. The ideal solution to this problem would be open access for all, al- lowing any interested person, anywhere in the world, the opportunity to download articles of interest free of charge. The problem, however, is cost: who will pay? Although LARR is in a unique situation compared with similar journals—as it is the flagship journal of LASA and individual and institutional LASA memberships pay for most of its subscriptions—

LASA memberships pay for most of its subscriptions— Latin American Research Review, Vol. 45, No. 3.

Latin American Research Review, Vol. 45, No. 3. © 2010 by the Latin American Studies Association.

PP5331.indb5331.indb

33

© 2010 by the Latin American Studies Association. P P5331.indb 5331.indb 3 3 1 10/15/10 0/15/10

110/15/100/15/10

11:52:57:52:57 PPMM

4 Latin American Research Review open access would still be cost prohibitive. This is because
4 Latin American Research Review open access would still be cost prohibitive. This is because

4 Latin American Research Review

open access would still be cost prohibitive. This is because LASA receives substantial revenues from royalties with Project MUSE, in particular, and JSTOR. These royalties would disappear with open access. There are other nonmonetary elements of risk as well. Open-access initiatives are still very new, and we know little about their potential implications, both positive and negative. Any such initiative by LARR is also risky, because LARR would be the first among its peer journals to do so. Restricting open access to residents of Latin America and the Caribbean should be viewed as a compromise and a strategic experiment; we are trying to address a real need at the same time that we are seeking to minimize risks and un- derstand the larger implications of open access for possible future LARR initiatives. Although we have tested the feasibility of open access and have attempted to ensure that it is consistent with the various copyright and indexing agreements LARR currently has, we will carefully monitor this going forward. The new open-access policy is also made possible by another new policy of the LASA Executive Committee: this issue will be the last that is automatically mailed to individuals and institutions located in Latin America and the Caribbean. For several years, we have been aware of the economic drain that mailing hard-copy issues of LARR to addresses in Latin America and the Caribbean entails for LASA. The average cost of printing and mailing LARR to that region is $80 per member, although individual membership fees for Latin American and Caribbean residents are between $27 and $53, depending on one’s income level. This means that LASA effectively loses between $27 and $53 dollars for each of the more than two thousand members in Latin America and the Caribbean. This is obviously a substantial amount of funds. The resulting savings not only will make open access feasible for all of the region but also will free up more money for the travel fund that subsidizes the participation at LASA congresses of people living in Latin America and the Caribbean, as well as other potential special initiatives. Of course, members can still request that hard copies of issues be mailed to them, provided that they pay the difference between their membership fees and the actual cost of printing and mailing the journal. Open access to scholarly literature is a dream that many people share. Although we are still far from reaching that dream, by providing open ac- cess to all of Latin America and the Caribbean, we have taken a big—and unprecedented—step in realizing it. As this first experiment unfolds, we will inform LASA’s membership of its progress. We also will continue to explore new ways to take advantage of information technology to make LARR accessible to as large an audience as possible at the same time that we remain committed to maintaining the highest editorial standards.

PP5331.indb5331.indb

44

to maintaining the highest editorial standards. P P5331.indb 5331.indb 4 4 1 10/15/10 0/15/10 1 1:52:57

110/15/100/15/10

11:52:57:52:57 PPMM

to maintaining the highest editorial standards. P P5331.indb 5331.indb 4 4 1 10/15/10 0/15/10 1 1:52:57
I N DIGE NOUS L A NGUAGE USAGE A N D M A I N

I N DIGE NOUS L A NGUAGE USAGE A N D M A I N T E NA NC E PAT T E R N S A MO NG I N DIGE NOUS PE OPL E I N T H E E R A OF N E OL I BE R AL M U LT IC U LT U R AL I S M IN MEXICO AND GUATEMALA

BE R AL M U LT IC U LT U R AL I S M IN

Hirotoshi Yoshioka

University of Texas at Austin

Abstract: In both Mexico and Guatemala, indigenous languages are at risk of extinc- tion. Because languages influence people’s ways of thinking and help them identify with particular ethnic groups, indigenous language loss can result in severe problems that extend well beyond the demise of these languages. Although current multicultural reformsofferindigenouspeopleunprecedentedopportunities,theseseeminglypositive changes may actually threaten indigenous languages and cultures. Using the latest demographic census data, I present how socioeconomic, demographic, and community factors negatively correlate with indigenous language usage. I contend that indig- enous language maintenance will become more difficult because neoliberal multi- culturalism endorses indigenous cultural rights without putting forth other necessary changes. Establishing effective language preservation strategies requires us to rec- ognize dangers hidden in the current multicultural agenda, to rigorously ask how we can destigmatize negative images attached to indigenous cultures, and to combat centuries-long oppression and discrimination against indigenous groups.

oppression and discrimination against indigenous groups. INTRODUCTION Throughout the world, hundreds of languages

INTRODUCTION

Throughout the world, hundreds of languages that are known to exist or to have existed are no longer spoken, and the pace of language loss has accelerated considerably during the past two centuries (Wurm 1991). Mex- ico and Guatemala are among the countries that face the danger of lan- guage loss. As a result of the oppression and subordination of indigenous groups in many parts of Latin America, many indigenous children no lon- ger learn indigenous languages and speak only the dominant language— Spanish, in most countries. Therefore, numerous indigenous languages in the region risk extinction (England 2003; Hale et al. 1992; Hawkins 2005). Indigenous people’s disadvantaged socioeconomic status and the pres- sure of assimilation into mestizo or Ladino society have been influen- tial on indigenous language loss. Hence, the pace of language loss may change as a result of drastic recent changes in the environment that sur- rounds indigenous people. During the 1990s, we witnessed the impres- sive mobilization of indigenous people in Latin America, and today, both

Latin American Research Review, Vol. 45, No. 3. © 2010 by the Latin American Studies Association.

PP5331.indb5331.indb

55

© 2010 by the Latin American Studies Association. P P5331.indb 5331.indb 5 5 1 10/15/10 0/15/10

110/15/100/15/10

11:52:58:52:58 PPMM

6 Latin American Research Review Mexico and Guatemala actively recognize indigenous cultural rights. Such changes
6 Latin American Research Review Mexico and Guatemala actively recognize indigenous cultural rights. Such changes

6 Latin American Research Review

Mexico and Guatemala actively recognize indigenous cultural rights. Such changes may lead us to believe that the preservation of indigenous languages will be successful, at least more so than in the past. Indeed, England (2003) notes that there has been some success in the promotion of Guatemalan Mayan languages. However, I argue that indigenous lan- guage maintenance will become more difficult; ironically, the increased difficulty partly derives from the recognition of indigenous cultural rights endorsed by what Hale (2002) calls neoliberal multiculturalism. In Latin America, the boundary between indigenous and nonindig- enous has never been static (Wolf 1986) or clear. As a result, what appears beneficial to indigenous people, including new indigenous cultural rights, may actually threaten their cultures. This point merits serious attention, because without a nuanced understanding of what neoliberal multi- culturalism really offers and endangers, it is impossible to design an ef- fective language maintenance strategy. Also, understanding the value of preserving indigenous languages is an essential component of a true and robust multicultural society, at least in Mexico and Guatemala.

I argue that, in Mexico and Guatemala, preserving indigenous lan-

guages is an important component of indigenous mobilization and is fundamental to actually protecting their cultural rights. Therefore, a fur- ther understanding of factors, patterns, and mechanisms that relate to language loss is essential for combating socioeconomic difficulties and injustices that indigenous people have faced. However, to my knowledge, no individual-level statistical analysis of this topic exits. Using the latest demographic census data from Mexico and Guatemala, I examine the correlation between indigenous language usage among self-identified in- digenous people and various socioeconomic and demographic factors. In addition, I analyze whether indigenous language usage among the chil- dren of indigenous language speakers differs because of their parents’ socioeconomic backgrounds. In this work, I aim to demonstrate why opportunities that emerged un- der multicultural reforms may threaten indigenous languages in Mexico and Guatemala rather than promote them. On the basis of empirical evi- dence, I highlight the importance and difficulties of indigenous language preservation, why effective language maintenance strategies require us to seriously take into account various socioeconomic and structural fac- tors, and an understanding of the hidden dangers of multiculturalism for indigenous cultures.

INDIGENOUS LANGUAGE USAGE AS A METHOD OF COUNTING INDIGENOUS PEOPLE

LANGUAGE USAGE AS A METHOD OF COUNTING INDIGENOUS PEOPLE It is diffi cult to determine the

It is difficult to determine the number of indigenous people with any

single measure. No one measure has satisfied all researchers as a method

PP5331.indb5331.indb

66

measure has satisfi ed all researchers as a method P P5331.indb 5331.indb 6 6 1 10/15/10

110/15/100/15/10

11:52:58:52:58 PPMM

INDIGENOUS LANGUAGE USAGE AND MAINTENANCE 7 of counting indigenous people (Ramirez 2006). The methods of
INDIGENOUS LANGUAGE USAGE AND MAINTENANCE 7 of counting indigenous people (Ramirez 2006). The methods of

INDIGENOUS LANGUAGE USAGE AND MAINTENANCE

7

of counting indigenous people (Ramirez 2006). The methods of counting indigenous people differ from one country to another in Latin America. Some countries use the ability to speak an indigenous language, whereas others take into account an individual’s self-identity or whether a per- son wears traditional clothes (Hawkins 2005; Layton and Patrinos 2006; Patrinos 1994). In Latin America, researchers estimate the number of in- digenous people by one or a combination of the following measures: in- digenous language usage, self-identification, and residence in indigenous territories or an area in which indigenous people are geographically con- centrated (Gonzalez 1994; Layton and Patrinos 2006). The self-identification and geographic measures have been criticized on several grounds. For example, the self-identification method may un- derestimate the number of indigenous people because discrimination and prejudice may lead individuals to deny any affiliation with this group (Gonzalez 1994). A person’s self-identification can also change over time as his or her ethnicity goes through numerous processes of creation, re- creation, and redefinition (Beck and Mijeski 2000; Gonzalez 1994). For ex- ample, during the 1980s in the United States, there were people who had formerly dismissed their Native American heritage but are now starting to claim it (Nagel 1995; Snipp 1989). Furthermore, identifying second- or third-generation heritages is impossible for orphans or children of sin- gle mothers who are unsure of their father’s ethnic identity (Bell 1996). Similarly, although the geographical location measure avoids the previous problems related to the self-identification measure, one problem with this method is the classification of nonindigenous people living in indigenous areas as indigenous and vice versa. Language usage has been considered almost an invariable factor in determining whether a person identifies with one group over another (Sagarin and Moneymaker 1979). Yet it can also overestimate or un- derestimate the number of indigenous people. For example, Paraguay gave up the language measure because of a large number of nonindig- enous people speaking Guarani (Gonzalez 1994). In contrast, in Mexico, classification by indigenous language usage significantly reduces the number of the indigenous population. That is, if an individual speaks only Spanish, he or she is considered mestizo even if all of his or her ancestors were indigenous (Kampwirth 2004). Despite these limitations, indigenous language usage has been regarded as one of the most impor- tant measures in counting indigenous people in Latin America. The lan- guage usage measure is particularly useful in the region where there is a continuum of racial and ethnic categories, at least more than in the United States where the category “black” supposedly includes anyone with a known “drop of black blood” (Wade 1997, 14). The category of the mixed- blood (mestizo in Mexico and Ladino in Guatemala) symbolizes such a continuum.

PP5331.indb5331.indb

77

Ladino in Guatemala) symbolizes such a continuum. P P5331.indb 5331.indb 7 7 1 10/15/10 0/15/10 1

110/15/100/15/10

11:52:58:52:58 PPMM

Ladino in Guatemala) symbolizes such a continuum. P P5331.indb 5331.indb 7 7 1 10/15/10 0/15/10 1
8 Latin American Research Review Indeed, in many parts of Latin America, including Mexico, where

8 Latin American Research Review

Indeed, in many parts of Latin America, including Mexico, where the ideology of mestizaje and assimilation took precedence (Mallon 1992), in- digenous language usage is the best measure to count indigenous people, because the vast majority of mestizos do have indigenous ancestors. How- ever, as a result of the decline in the proportion of indigenous language speakers among indigenous people, we might need to reconsider the ef- fectiveness of the language usage measure in counting indigenous people. To understand why the proportion of indigenous language speakers has decreased and why the pace of language loss has accelerated over the past few decades (England 2003), it is essential to examine the socioeconomic and political positions of indigenous groups and the circumstances that surround them.

indigenous groups and the circumstances that surround them. INDIGENOUS PEOPLE AND LANGUAGES IN MEXICO AND GUATEMALA

INDIGENOUS PEOPLE AND LANGUAGES IN MEXICO AND GUATEMALA

According to the 2000 Mexican census, 7.13 percent of Mexico’s popula- tion is indigenous with respect to the language measure. There are fifty- six indigenous languages in Mexico, and 90 percent of indigenous lan- guage speakers reside in the states of Oaxaca, Yucatán, Puebla, Veracruz, and Chiapas (Gonzalez 1994). The 2002 Guatemalan census indicates that about 42 percent of Guatemalans are indigenous, a proportion surpassed in only Bolivia (Layton and Patrinos 2006). The Xinca and the twenty-one Mayan linguistic groups are regarded as indigenous in Guatemala. Despite the large difference in the proportion of indigenous people in the two countries, indigenous people in both Mexico and Guatemala face many more socioeconomic difficulties than nonindigenous people. This is also the case in other parts of Latin America. Indigenous peoples’ disadvantaged socioeconomic status in Mexico and Guatemala has con- tinued for centuries, since the onset of the colonial period. The division of classes, largely defined by ethnicity, implies that the region has been multicultural for a long time (Sieder 2002). However, the problems that indigenous people faced have largely been ignored, and ethnic differ- ences have not reflected the region’s politics or legal and administrative arrangements (Hall, Layton, and Shapiro 2006). Rather, the state regarded such problems as a class-based issue. Previous studies have shown two main factors that have influenced indigenous language usage. First, indigenous people’s disadvantaged socioeconomic status has discouraged indigenous language usage. In Guatemala, indigenous languages are often associated with the “nega- tive” values of the “traditional”: ignorance, lack of education, and poverty. In contrast, people associate the dominant language—Spanish—with the “positive” values of the “modern” (England 2003). Second, government authorities have affected indigenous language usage in the two countries. For example, regarding language diversity as an obstacle to achieve unity,

PP5331.indb5331.indb

88

language diversity as an obstacle to achieve unity, P P5331.indb 5331.indb 8 8 1 10/15/10 0/15/10

110/15/100/15/10

11:52:58:52:58 PPMM

language diversity as an obstacle to achieve unity, P P5331.indb 5331.indb 8 8 1 10/15/10 0/15/10
INDIGENOUS LANGUAGE USAGE AND MAINTENANCE 9 in 1911, the Mexican government put the Law for
INDIGENOUS LANGUAGE USAGE AND MAINTENANCE 9 in 1911, the Mexican government put the Law for

INDIGENOUS LANGUAGE USAGE AND MAINTENANCE

9

in 1911, the Mexican government put the Law for Rudimentary Instruc- tion in force, which prohibited indigenous children from using their ma- ternal languages in school (Bravo Ahuja 1992). Historically, bilingual edu- cation, which has been asymmetrical and unidirectional, has discouraged indigenous people from learning indigenous languages in both Mexico and Guatemala (Patthey-Chavez 1994; Richards and Richards 1997; Sali- nas Pedraza 1997). Since the 1990s, several events, including the awarding of the 1992 No- bel Prize to Rigoberta Menchú and the Zapatista revolt against the inau- guration of the North American Free Trade Agreement in 1994, dramati- cally changed the environment that surrounds indigenous people. And the quincentenary celebrations led the world to pay unprecedented atten- tion to indigenous populations in Latin America. Not only did indigenous populations attract more attention, they are pressing for rights of theirs that states did not recognize for a long time. For example, in 1992, the Mexican government started to acknowledge the existence of indigenous communities in the national legislation through the reform of Article 4 of the Mexican Constitution for the first time since the end of Spanish colonialism (Kampwirth 2004; Salinas Pedraza 1997). Similarly, the Gua- temalan government signed the 1995 Accord on the Identity and Rights of Indigenous Peoples, promising to take measures to recognize and com- pensate indigenous people (England 2003; Jonas 2000). 1 Despite these remarkable changes, inequalities between indigenous and nonindigenous populations have persisted in Latin America dur- ing the Indigenous Peoples’ Decade from 1994 to 2004 (Hall et al. 2006). Hale (2002) has attempted to explain this paradoxical phenomenon using the concept of neoliberal multiculturalism. The author argues that multi- culturalism took place in Latin America “in the general context of neolib- eral political and economic reforms” (Hale 2002, 493). And multicultural reforms affirm new rights without resolving socioeconomic inequalities (Hamel 1994). Because indigenous people in both Mexico and Guatemala belong to the poorest group, despite the recognition of indigenous cul- tural rights, they continue to face socioeconomic hardships. Therefore, a chance to achieve upward socioeconomic mobility for most indigenous people is small because of the high level of inequality in income and ac- cess to needed services that segregates citizens by their social class (Rob- erts 2005). However, the state’s attitude toward indigenous people and the disad- vantaged socioeconomic status of indigenous groups alone are insufficient to explain the accelerated rate of language loss in Mexico and Guatemala, because the relationship between indigenous and nonindigenous groups

the relationship between indigenous and nonindigenous groups 1. However, the accords, including the idea of either

1. However, the accords, including the idea of either officialization or co-officialization of

indigenous Mayan languages, were turned down in May 1999 (England 2003; Jonas 2000).

PP5331.indb5331.indb

99

turned down in May 1999 (England 2003; Jonas 2000). P P5331.indb 5331.indb 9 9 1 10/15/10

110/15/100/15/10

11:52:59:52:59 PPMM

10 Latin American Research Review has always been unequal. Besides, unlike in the past, bilingual

10 Latin American Research Review

has always been unequal. Besides, unlike in the past, bilingual education today in the two countries emphasizes the importance of ethnic diversity (Patthey-Chavez 1994; Richards and Richards 1997). Therefore, a close ex- amination of both micro- and macrofactors is essential to understand the accelerated language loss rate. It is also necessary to explain why multi- cultural reforms can threaten indigenous languages.

multi- cultural reforms can threaten indigenous languages. NEOLIBERAL MULTICULTURALISM AND INDIGENOUS LANGUAGE USAGE

NEOLIBERAL MULTICULTURALISM AND INDIGENOUS LANGUAGE USAGE

One reason many indigenous people have maintained their languages until today despite the strong pressure of castellanización and great socio- economic discrepancies between indigenous and nonindigenous groups is that there were few opportunities for indigenous people. As a result, little incentive existed for them to learn Spanish (Garzon 1998a; Richards 2003). This is especially the case in Guatemala, where bilingual education started much later than in Mexico (Hall et al. 2006) and lacked an official discourse of mestizaje (Hale 2002), which led indigenous groups to con- sider Spanish the outsider’s language. Multicultural reforms have drastically changed some of these condi- tions. I contend that the recent accelerated rate of language loss in both Mexico and Guatemala is attributable to new opportunities that emerged under multicultural reforms and related consequences, especially rural- urban migration and both quantitative and qualitative changes in con- tact between indigenous and nonindigenous groups. For example, today’s broader coverage of bilingual education has helped indigenous children perform better in school, and more indigenous people hold professional positions today. We must welcome such new opportunities and socio- economic advancement of indigenous people. However, it is also essential to realize that under today’s neoliberal multiculturalism, unless accom- panied by necessary structural changes, these opportunities can also be perilous for indigenous cultures. Indigenous children do not learn indigenous languages today because many parents believe that it is more useful to teach their children Spanish rather than their native languages. Such a belief often derives from their own traumatic experiences as a result of their inability to speak Spanish well at school (Brown 1998; England 1996). In other words, Spanish is con- sidered a power symbol of the mainstream society (Hill and Hill 1980). Therefore, indigenous children learn Spanish in the context of a serious sociocultural inequality (Lewin 1986). Although today’s multicultural re- forms certainly help some indigenous people overcome hardships and be- come included in mainstream society, the changes that seem apparently beneficial to indigenous people can be detrimental to their cultures. For instance, with respect to Guatemalan Mayan groups, Garzon (1998a) states that an indigenous population’s integration into mestizo or Ladino society

PP5331.indb5331.indb

1100

integration into mestizo or Ladino society P P5331.indb 5331.indb 1 10 0 1 10/15/10 0/15/10 1

110/15/100/15/10

11:52:59:52:59 PPMM

integration into mestizo or Ladino society P P5331.indb 5331.indb 1 10 0 1 10/15/10 0/15/10 1
INDIGENOUS LANGUAGE USAGE AND MAINTENANCE 11 has often resulted in the internalization of negative images
INDIGENOUS LANGUAGE USAGE AND MAINTENANCE 11 has often resulted in the internalization of negative images

INDIGENOUS LANGUAGE USAGE AND MAINTENANCE

11

has often resulted in the internalization of negative images attached to in- digenous groups among indigenous people themselves. Higher educational attainment also leads many indigenous people to leave their communities for larger cities. Rural-urban migration dis- courages indigenous language use in two ways, which are related to each other. First, the proportion of indigenous language speakers is much lower in urban areas than in rural areas, which reinforces people’s learning of Spanish and offers fewer occasions to practice indigenous languages. Therefore, indigenous people who leave their communities permanently “may feel little need to assert their ethnic identity” (Garzon 1998a, 198). Second, those who migrate to urban areas usually have higher socioeconomic status than people who remain in rural areas. Because we usually interact with others of similar socioeconomic backgrounds, in- digenous language speakers in urban areas tend to have more frequent contact with mestizos or Ladinos. This point merits close attention, be- cause it reflects a qualitative change in contact between indigenous and nonindigenous groups, as this contact often takes place without a clear hierarchy. I argue that indigenous people who are in frequent contact with mes- tizos or Ladinos in a more egalitarian manner face a greater risk of los- ing indigenous languages because, even though the state argues for the importance of ethnic diversity and respect, indigenous people continue to face hostility and derision from the dominant mestizo or Ladino group (Garzon 1998a), both explicitly and implicitly. As a result, some indige- nous people prefer to abandon their ethnic identity and assimilate into the mainstream mestizo or Ladino culture. Because they tend to speak Spanish well and have other means, such as formal education, it is easier for them to abandon their indigenous languages (Garzon 1998a). The new type of interethnic relationship may also induce a tension between indigenous and nonindigenous groups. The increased contact among ethnic groups leads not only indigenous but also nonindigenous people to rethink ethnic hierarchies and relations in a society. For eco- nomically less advantaged nonindigenous groups, who are most likely to interact with indigenous groups in their daily life, such interactions may lead to the perception of abandonment among nonindigenous groups. 2 An additional problem of multicultural reforms today are the calls for equality in the existence of severe socioeconomic inequality, which also makes the protection of indigenous cultures more difficult. The rejection of the constitutional reforms in Guatemala reflects such a difficulty (Mon- tejo 2002).

in Guatemala refl ects such a diffi culty (Mon- tejo 2002). 2. According to the latest

2. According to the latest census data from Mexico and Guatemala, poor, nonindigenous

households are more likely to live in municipalities with a greater proportion of indigenous households. The results of those statistical models are available on request.

PP5331.indb5331.indb

1111

statistical models are available on request. P P5331.indb 5331.indb 1 11 1 1 10/15/10 0/15/10 1

110/15/100/15/10

11:52:59:52:59 PPMM

12 Latin American Research Review Without resolving fundamental social problems and rectifying past injustices,

12 Latin American Research Review

Without resolving fundamental social problems and rectifying past injustices, bilingual education and the emerging opportunities not only discourage indigenous people from practicing their languages but also lead to stronger oppositions toward indigenous language preservation both from nonindigenous and from indigenous groups. It is my conten- tion that, rather than protecting indigenous rights and cultures, multi- culturalist reforms lead to a vicious cycle of language loss. As I discuss here, indigenous language loss can result in severe problems that extend well beyond the demise of languages, including the disappearance of in- digenous cultures, thus endangering even the most basic premise of multi- culturalism. Hence, preserving and promoting indigenous languages and understanding how socioeconomic and demographic factors relate to in- digenous language usage patterns are very important.

to in- digenous language usage patterns are very important. DATA AND METHODS To examine indigenous language

DATA AND METHODS

To examine indigenous language usage patterns in Mexico and Gua- temala, I used the latest nationally representative demographic census data of Mexico (2000) and Guatemala (2002). The Instituto Nacional de Estadística y Geografía (INEGI) in Mexico and the Instituto Nacional de Estadística (INE) in Guatemala, respectively, collected the census data. The Mexican census data are a 10 percent sample, and I used the sam- pling weight provided in the data set; the Guatemalan census data are 100 percent data (i.e., the data set contains all households interviewed). The censuses counted indigenous people by indigenous language usage and by respondents’ self-identification. The Guatemalan census data pre- sent information about a person’s maternal language, and if respondents spoke more than one language, the census asked them to name up to two nonmaternal languages. In contrast, the Mexican census asks directly whether a respondent speaks an indigenous language and whether a per- son speaks Spanish. The current analysis has two parts. First, I examine the correlation be- tween socioeconomic and community characteristics and indigenous lan- guage usage among self-identified indigenous people. The purpose of the first part is to examine whether, among those who identify themselves as indigenous, indigenous language usage is related to socioeconomic background. In the second part, I analyze how indigenous language us- age among children of indigenous language speakers differs by their par- ents’ and households’ socioeconomic status and whether the household head’s spouse speaks an indigenous language. In the second part, I limit my sample to children between the ages of six and eighteen whose house- hold head speaks an indigenous language and for which the spouse of the household head was present in the household at the time of the census. I

PP5331.indb5331.indb

1122

in the household at the time of the census. I P P5331.indb 5331.indb 1 12 2

110/15/100/15/10

11:53:00:53:00 PPMM

in the household at the time of the census. I P P5331.indb 5331.indb 1 12 2
INDIGENOUS LANGUAGE USAGE AND MAINTENANCE 13 exclude those children whose household heads were not married
INDIGENOUS LANGUAGE USAGE AND MAINTENANCE 13 exclude those children whose household heads were not married

INDIGENOUS LANGUAGE USAGE AND MAINTENANCE

13

exclude those children whose household heads were not married because the most recent census data in these countries do not provide any infor- mation on a person’s ex-spouse. The unit of analysis in the first part is individuals who identify them- selves as indigenous, and in the second part, it is the children of indig- enous language speakers. In both parts of the analysis, I use multinomial logistic regression models in which I regress the use of indigenous lan- guages (0 if a respondent does not speak an indigenous language, 1 if he or she speaks only indigenous languages, and 2 if the respondent speaks both Spanish and indigenous languages) on three explanatory variables (household asset index, individual’s level of education, and migration sta- tus) and several sociodemographic factors, including respondent’s sex, urban-rural status, and the proportion of people in a municipality who speak indigenous languages. As the dependent variable indicates, I divide indigenous language speakers into two groups (i.e., monolingual and bi- lingual with Spanish) to closely examine whether independent variables considered in this study relate differently to indigenous language use on the basis of whether people use indigenous languages as their only lan- guage. I used principal component analysis to construct the asset index for households based on households’ access to or ownership of several resources, such as electricity, running water, and primary cooking fuel. I used this asset index as a proxy for household wealth. Because the Gua- temalan census does not provide any information on income, to make the study of indigenous language usage comparable between Mexico and Guatemala, taking into account household wealth it is the most appropri- ate option for this study. 3 I clustered the data set according to municipalities in which respon- dents lived to obtain robust standard errors, because a person’s place of residence may influence indigenous language usage. As noted, I also took into account households’ migration. I defined migration in this study as whether a person’s current residence of state (Mexico) or department (Gua- temala) differed between the time the census was taken and five (Mexico) or six (Guatemala) years prior to it. Therefore, I could not measure tempo- rary migratory movements such as seasonal migration using the census data. Hence, with the current data, I could not estimate the correlation be- tween temporary migratory movements that took place during five or six years before the census was taken and indigenous language usage, and it was necessary to take into account this limitation when explaining the correlation between this variable and indigenous language usage.

between this variable and indigenous language usage. 3. A list of the variables used to calculate

3. A list of the variables used to calculate the asset index is available on request. For more

information on the construction and validity of the asset index and principal component analysis, see Filmer and Pritchett (2001).

PP5331.indb5331.indb

1133

analysis, see Filmer and Pritchett (2001). P P5331.indb 5331.indb 1 13 3 1 10/15/10 0/15/10 1

110/15/100/15/10

11:53:00:53:00 PPMM

14 Latin American Research Review A major limitation of this study is that, although I

14 Latin American Research Review

A major limitation of this study is that, although I argue that we must

not see indigenous groups as one group, I do not differentiate between indigenous groups in statistical models. I decided to aggregate indigenous groups in this study, because the number of speakers for most indigenous languages in the two countries is not large, and therefore estimating

statistical models is not feasible. It is possible that treating indigenous languages as one language in this study overestimates the risk of lan- guage loss for those languages with a large number of speakers, such as K’iche’ in Guatemala but underestimates the risk for languages that fewer people speak. Therefore, the results of the study need to be viewed with caution.

the results of the study need to be viewed with caution. In this study, I test

In this study, I test the following four hypotheses:

H 1 : Indigenous people with a higher level of education, higher wealth index, and who live in urban areas are less likely to speak indigenous languages especially as their only languages. People with a higher socioeconomic sta- tus and those who live in urban areas are more likely to interact with non- indigenous people, and they need to speak Spanish more often than people living in rural areas. The interaction with nonindigenous people may also lead indigenous people to think that speaking Spanish is more important and useful than speaking indigenous languages. Furthermore, indigenous people in urban areas encounter few opportunities to learn and practice indigenous languages.

H 2 : The children of household heads who have recently migrated are less likely to speak indigenous languages as their only languages. Indigenous people are more likely to engage in agricultural work, and those who en- gage in agricultural work are, in general, less likely to migrate permanently. Therefore, migrant households have more contacts with nonindigenous language speakers and the children of migrant households tend to learn in- digenous languages less, which I argue is especially the case if households migrate when a child is very young.

H 3 : The children of indigenous language speakers with a higher socioeco- nomic status are less likely to speak the indigenous language. Relating to the first two hypotheses, people with a higher level of socioeconomic back- ground tend to live in urban areas and have nonagricultural occupations. Moreover, parents with a higher level of education may prefer to speak with their children in Spanish rather than in indigenous languages. Children living in urban areas are more likely to receive Spanish-only instruction at their schools, thus limiting their need to speak the indigenous language.

H 4 : The degree of correlation between a household’s socioeconomic back- ground and a child’s indigenous language usage is lower when a model takes into account the household head’s spouse’s indigenous language us- age. Because people tend to marry a person of a similar socioeconomic and cultural background, it is probable that indigenous language speakers with a high socioeconomic status are less likely to marry indigenous language

PP5331.indb5331.indb

1144

are less likely to marry indigenous language P P5331.indb 5331.indb 1 14 4 1 10/15/10 0/15/10

110/15/100/15/10

11:53:00:53:00 PPMM

are less likely to marry indigenous language P P5331.indb 5331.indb 1 14 4 1 10/15/10 0/15/10
INDIGENOUS LANGUAGE USAGE AND MAINTENANCE 15 speakers if there is a negative correlation between socioeconomic

INDIGENOUS LANGUAGE USAGE AND MAINTENANCE

15

speakers if there is a negative correlation between socioeconomic status and indigenous language usage.

between socioeconomic status and indigenous language usage. INDIGENOUS LANGUAGE USAGE AMONG SELF - IDENTIFIED INDIGENOUS

INDIGENOUS LANGUAGE USAGE AMONG SELF-IDENTIFIED INDIGENOUS PEOPLE

Table 1 presents descriptive statistics for Mexican and Guatemalan people who identify themselves as indigenous. More than half of Mexican indigenous people live in small communities with fewer than 2,500 inhab-

itants (50.9 percent). They also tend to concentrate in municipalities where at least half the population speaks an indigenous language (43.58 percent). At the same time, about 31 percent of people live in municipalities where fewer than 5 percent of inhabitants speak indigenous languages, which is in part due to a small proportion of indigenous people in Mexico. About

58 percent of self-identified indigenous people speak an indigenous lan-

guage in Mexico. In my sample, about 11 percent of self-identified indig- enous people in Mexico speak only indigenous languages. The asset in- dex indicates economic hardships that Mexican indigenous people face. Almost 75 percent of Mexican indigenous people belong to the poorest

40 percent of the Mexican national population.

The demographic characteristics of Guatemalan indigenous people are similar to those of Mexican indigenous people. Only about 37 percent of indigenous people live in urban areas, as defined by the Guatemalan gov- ernment. In addition, similar to their Mexican counterparts, Guatemalan indigenous people also face tough economic situations. About 64 percent of indigenous people account for the poorest 40 percent of the Guatema- lan population. The proportion of indigenous people who are in the poor- est 40 percent group is lower in Guatemala in part because the proportion of indigenous people is much greater in Guatemala than in Mexico. A major difference between Mexican and Guatemalan self-identified indigenous people is that Guatemala’s indigenous people are more concen- trated in communities where most people speak an indigenous language. Almost 83 percent of self-identified indigenous people in Guatemala live in municipalities where at least 50 percent of people speak indigenous lan- guages. In addition, almost 40 percent of Guatemalan indigenous people have received less than primary education. Because the age distribution in the two countries is quite similar, we can infer that indigenous people in Guatemala receive fewer years of formal education than Mexican in- digenous people. Finally, self-identified indigenous people in Guatemala (80 percent) are more likely to speak indigenous languages than are their Mexican counterparts, and a much greater percentage of indigenous peo- ple in Guatemala, compared to Mexico, speak only indigenous languages (28.63 percent).

PP5331.indb5331.indb

1155

speak only indigenous languages (28.63 percent). P P5331.indb 5331.indb 1 15 5 1 10/15/10 0/15/10 1

110/15/100/15/10

11:53:01:53:01 PPMM

speak only indigenous languages (28.63 percent). P P5331.indb 5331.indb 1 15 5 1 10/15/10 0/15/10 1
Table 1 Percentage Distribution of Self-Identifi ed Indigenous People Aged 6 and Older by Selected

Table 1 Percentage Distribution of Self-Identified Indigenous People Aged 6 and Older by Selected Characteristics, Mexico (2000) and Guatemala (2002)

Selected Characteristics, Mexico (2000) and Guatemala (2002)   Variable Percentage   Mexico Guatemala
 

Variable

Percentage

 

Mexico

Guatemala

Lives in urban area

36.80

Locality size Less than 2,500

50.90

2,500–14,999

19.82

More than 15,000

29.27

Percentage of indigenous language speakers in municipality Less than 5%

31.36

1.29

5.00–19.99%

11.64

8.63

20.00–49.99%

13.43

7.22

More than 50.00%

43.58

82.87

 

Female

50.41

51.42

Age

15

and younger

30.26

29.77

15–29

30.57

33.68

 

30–44

19.49

18.45

45–59

11.38

10.83

60

and older

8.30

7.27

Indigenous language Do not speak indigenous language Speak only indigenous language Speak both indigenous language and Spanish

41.87

19.95

10.85

28.63

47.28

51.42

Education None or less than primary Primary Secondary Postsecondary

18.16

39.98

56.97

50.80

21.91

8.40

2.96

0.82

Migrated

3.93

2.22

Asset index

 

40% lowest

74.69

63.82

40% middle

18.06

30.58

20% highest

7.24

5.60

40% middle 18.06 30.58 20% highest 7.24 5.60 Note : N = 971,074 (Mexico); N =

Note: N = 971,074 (Mexico); N = 2,927,568 (Guatemala). Mexican data are weighted. Data from 2000 Mexican Census; 2002 Guatemalan Census.

PP5331.indb5331.indb

1166

from 2000 Mexican Census; 2002 Guatemalan Census. P P5331.indb 5331.indb 1 16 6 1 10/15/10 0/15/10

110/15/100/15/10

11:53:01:53:01 PPMM

INDIGENOUS LANGUAGE USAGE AND MAINTENANCE 17 Table 2 presents the results from multinomial logistic regression
INDIGENOUS LANGUAGE USAGE AND MAINTENANCE 17 Table 2 presents the results from multinomial logistic regression

INDIGENOUS LANGUAGE USAGE AND MAINTENANCE

17

Table 2 presents the results from multinomial logistic regression mod- els. Model 1 includes individual and household characteristics. The model shows that most individual characteristics are significantly correlated with the use of indigenous languages in both Mexico and Guatemala, except for respondents’ ages among Guatemalans who speak only indigenous lan- guages. For example, people age fifteen and older are significantly more likely to speak indigenous languages than individuals who are from the ages of six to fourteen in both Mexico and Guatemala. However, such a difference does not exist between those younger than fifteen years and people between the ages of fifteen and twenty-nine among Mexican in- digenous people who speak only indigenous languages. In addition, a person’s level of education is negatively correlated to indigenous language usage, which is especially true among those who speak only indigenous languages in both countries. For example, in Mex- ico, the probability of speaking only indigenous languages among people with postsecondary education is only about 0.4 percent of that among people with less than primary education. Similarly, in Guatemala, people with a higher level of education are significantly less likely to speak an indigenous language, taking into account other individual characteristics. Although the negative correlation between the level of education and in- digenous language usage also exists among bilingual people, the differ- ence is not as substantive as among those people who speak only indig- enous languages. At the same time, the fact that the level of education is, in general, negatively correlated with the use of indigenous languages reflects the difficult task of preserving indigenous languages because in both Mexico and Guatemala, children tend to receive more years of educa- tion today than in the past. In both countries, members of a household with a higher asset index are also significantly less likely to speak indigenous languages either as their only languages or when they also speak Spanish. The negative correlation between a household’s economic status and its members’ indigenous lan- guage use is especially strong among those who speak only indigenous languages. For example, the probability of speaking only indigenous lan- guages among those from the richest 20 percent of households is less than 1 percent of that among people from the poorest 40 percent of households in Mexico. Similarly, in Guatemala, richer household members are espe- cially less likely to speak only indigenous languages than are those in poor households. The model also shows that, though women are significantly more likely to be indigenous monolingual speakers, they are less likely to be bilin- gual than men in both Mexico and Guatemala. This is probably because men face more occasions, such as jobs, that require them to speak both Spanish and indigenous languages. Because indigenous women are more likely to have less extensive networks, they are more likely than men to

PP5331.indb5331.indb

1177

networks, they are more likely than men to P P5331.indb 5331.indb 1 17 7 1 10/15/10

110/15/100/15/10

11:53:01:53:01 PPMM

networks, they are more likely than men to P P5331.indb 5331.indb 1 17 7 1 10/15/10

and Spanish

Indigenous

***

***

***

*

0.858

4.882

1.783

0.925

0.841

0.762

Model 2

***

***

***

Indigenous

**

*

Only

0.687

1.235

0.352

11.968

0.478

1.116

Guatemala

and Spanish

Indigenous

***

***

1.643

Table 2 Relative Risk Ratios from Multinomial Logistic Regression Predicting the Indigenous Language Usage

0.911

Model 1

***

Indigenous

Only

1.199

1.031

and Spanish

***

***

***

***

***

***

Indigenous

4.807

0.864

0.659

0.966

10.679

55.199

Model 2

***

***

***

***

***

***

Indigenous

**

Only

305.604

0.365

17.865

0.445

1.524

1.341

6.613

Mexico

and Spanish

Indigenous

***

***

0.841

2.184

Model 1

Indigenous

***

Only

1.047

1.451

Percentage of indigenous language speakers

Lives in urban area

in municipality (Less than 5%)

More than 50.00%

More than 15,000

Age (15 and younger)

Locality size (Less than 2,500)

20.00–49.99%

2,500–14,999

5.00–19.99%

Female

15–29

20.00–49.99% 2,500–14,999 5.00–19.99% Female 15–29 P P5331.indb 5331.indb 1 18 8 2.811 1 10/15/10 0/15/10 1

PP5331.indb5331.indb

1188

2,500–14,999 5.00–19.99% Female 15–29 P P5331.indb 5331.indb 1 18 8 2.811 1 10/15/10 0/15/10 1 1:53:01
2,500–14,999 5.00–19.99% Female 15–29 P P5331.indb 5331.indb 1 18 8 2.811 1 10/15/10 0/15/10 1 1:53:01
2.811
2.811

110/15/100/15/10

11:53:01:53:01 PPMM

***

10,864.81

–2,422,133.4

2,927,568

***

10,864.81

–2,609,128.8

2,927,568

***

10,864.81

–558,441.2

971,074

***

10,434.06

–731,758.5

971,074

*p < 0.05; **p < 0.01; ***p < 0.001

than primary)

Log-likelihood

Postsecondary

Older than 60

Education (None or less

(40% lowest)

Asset index

20% highest

40% middle

Secondary

Migrated

Constant

Primary

30–44

45–59

N

χ 2

1.311 *** 2.669 *** 2.345 *** 4.483 *** 0.953 2.001 *** 1.158 ** 2.392 ***
1.311
***
2.669
***
2.345
***
4.483
***
0.953
2.001
***
1.158
**
2.392
***
1.650
***
3.008
***
3.085
***
5.112
***
1.021
2.150
***
1.312
***
2.692
***
1.780
***
2.810
***
3.934
***
5.203
***
1.105
2.044
***
1.557
***
2.709
***
0.192
***
1.087
*
0.143
***
0.891
**
0.187
***
0.990
0.171
***
0.935
0.007
***
0.573
***
0.005
***
0.451
***
0.060
***
0.649
***
0.056
***
0.598
***
0.004
***
0.777
**
0.002
***
0.452
***
0.076
***
0.534
***
0.078
***
0.529
***
0.336
***
0.857
1.662
**
2.370
***
0.458
***
0.930
1.744
***
2.149
***
0.015
***
0.110
***
0.119
***
0.349
***
0.195
***
0.528
***
0.273
***
0.620
***
0.006
***
0.061
***
0.163
***
0.349
***
0.076
***
0.270
***
0.207
***
0.460
***
1.357
**
1.323
***
0.033
***
0.120
***
6.316
***
2.760
***
0.943
0.853

PP5331.indb5331.indb

1199

0.120 *** 6.316 *** 2.760 *** 0.943 0.853 P P5331.indb 5331.indb 1 19 9 1 10/15/10

110/15/100/15/10

11:53:02:53:02 PPMM

0.120 *** 6.316 *** 2.760 *** 0.943 0.853 P P5331.indb 5331.indb 1 19 9 1 10/15/10
20 Latin American Research Review speak only indigenous languages. Finally, in this model, I found
20 Latin American Research Review speak only indigenous languages. Finally, in this model, I found

20 Latin American Research Review

speak only indigenous languages. Finally, in this model, I found that mi- gration experience during the past five (Mexico) or six (Guatemala) years before the census was taken is significantly negatively correlated with indigenous language use among indigenous monolingual speakers in both countries. In Model 2, in addition to the individual characteristics considered in Model 1, I included two community characteristics: urban-rural status and an indigenous language prevalence rate at the municipality level. Taking into account the community characteristics, generational differ- ences in indigenous language usage are greater in Mexico than in Gua- temala for both indigenous language monolingual speakers and those who also speak Spanish. For example, although people age fifteen and older in Mexico are 1.3 times (age fifteen to twenty-nine) to 4 times (age forty-five to fifty-nine) more likely to speak only indigenous languages than are children from the ages of six to fourteen, people ages fifteen and older in Guatemala are only about 1.1 (age fifteen to twenty-nine) to 1.6 (people age sixty and older) times more likely to speak only indigenous languages than are children between the ages of six and fourteen. Hence, the model indicates that Mexican indigenous people are abandoning in- digenous languages at a more rapid pace than are Guatemalan indig- enous people. In Mexico, respondents living in larger towns and cities are significantly less likely to speak only indigenous languages than those in smaller com- munities. For example, people living in municipalities with at least 15,000 people are only about 37 percent as likely as those living in communities with fewer than 2,500 people to speak only indigenous languages. Simi- larly, in Guatemala, those living in urban areas are only about 48 percent as likely as those living in rural areas to speak only indigenous languages. Among people who speak both indigenous languages and Spanish, the difference in the probability of speaking indigenous languages by the locality size also exists, although it is not as large. Although the model does not show any significant difference in indigenous language usage between people who reside in communities with fewer than 2,500 people and those with at least 15,000 people among those who speak both Span- ish and indigenous languages, this is because people who are less likely to speak indigenous languages (e.g., those with higher levels of educa- tion and economically more advantaged) tend to concentrate in larger municipalities. 4 In both Mexico and Guatemala, people living in municipalities with a higher percentage of indigenous language speakers are significantly more likely to speak indigenous languages. The correlation is especially strong

indigenous languages. The correlation is especially strong 4. Indeed, without considering other factors, people living

4. Indeed, without considering other factors, people living in larger communities are significantly less likely to speak indigenous languages.

PP5331.indb5331.indb

2200

cantly less likely to speak indigenous languages. P P5331.indb 5331.indb 2 20 0 1 10/15/10 0/15/10

110/15/100/15/10

11:53:02:53:02 PPMM

INDIGENOUS LANGUAGE USAGE AND MAINTENANCE 21 and significant among those who speak only indigenous languages.
INDIGENOUS LANGUAGE USAGE AND MAINTENANCE 21 and significant among those who speak only indigenous languages.

INDIGENOUS LANGUAGE USAGE AND MAINTENANCE

21

and significant among those who speak only indigenous languages. For example, the probability of people living in municipalities where more than half of people are indigenous language speakers to speak only indig- enous languages is more than 305 times that of among those who live in municipalities where fewer than 5 percent of inhabitants speak indigenous languages. Similar patterns can also be found in Guatemala, although the difference in indigenous language usage by indigenous language preva- lence rates is much greater in Mexico than in Guatemala. This is because indigenous people with higher education and better economic status are in Mexico more likely to live in municipalities with fewer indigenous lan- guage speakers than are indigenous people with similar socioeconomic backgrounds in Guatemala. An interesting finding is the correlation between recent migration and indigenous language usage. Controlling for both individual and commu- nity characteristics, I found that recent migration is positively correlated to indigenous language usage in both Mexico and Guatemala. The differ- ence is statistically significant in both countries and is present regardless of whether people speak indigenous languages as their only languages or in addition to Spanish. In Mexico, the probability of speaking only indig- enous languages among people who migrated during the past five years is about 1.6 times higher than that among those who did not, and in Guate- mala, recent migrants are about 1.74 times more likely than nonmigrants to speak indigenous languages as their only languages. This is most likely because in Model 2, I control for both individual and community charac- teristics. The data sets show that, in both countries, people are more likely to migrate to larger communities with a lower proportion of indigenous language speakers. Overall, in both Mexico and Guatemala, I have found that people with higher socioeconomic status are less likely to speak indigenous lan- guages. The results are similar in both Mexico and Guatemala, indicating that among indigenous people in the two countries, whether indigenous people speak indigenous languages is significantly correlated with their socioeconomic status and the environment that surrounds them.

status and the environment that surrounds them. INDIGENOUS LANGUAGE SPEAKERS AND THEIR CHILDREN ’ S

INDIGENOUS LANGUAGE SPEAKERS AND THEIR CHILDRENS LANGUAGE USAGE

Table 3 presents descriptive statistics for both the Mexican and Guate- malan samples used in the analysis of language maintenance among chil- dren of indigenous language speakers. About 53 percent of children aged six to eighteen whose household heads speak an indigenous language can also speak it in Mexico, whereas in Guatemala, about 83 percent of chil- dren do the same. In Mexico, fewer than 10 percent of children of indig- enous language speakers speak only indigenous languages, whereas the percentage is much higher among Guatemalan children (31.16 percent).

PP5331.indb5331.indb

2211

higher among Guatemalan children (31.16 percent). P P5331.indb 5331.indb 2 21 1 1 10/15/10 0/15/10 1

110/15/100/15/10

11:53:02:53:02 PPMM

Table 3 Percentage Distribution of Children Aged 6–18 of Household Head Who Speaks Indigenous Language

Table 3 Percentage Distribution of Children Aged 6–18 of Household Head Who Speaks Indigenous Language by Selected Characteristics, Mexico (2000) and Guatemala (2002)

Selected Characteristics, Mexico (2000) and Guatemala (2002) Variable Percentage   Mexico Guatemala

Variable

Percentage

 

Mexico

Guatemala

Lives in urban area

32.14

Locality size Less than 2,500

56.76

2,500–14,999

20.49

More than 15,000

22.75

Percentage of indigenous language speaker in municipality Less than 5%

17.18

1.14

5.00–19.99%

16.22

5.99

20.00–49.99%

17.67

4.93

More than 50.00%

48.93

87.94

Age of household head

20–29

5.31

4.26

30–44

56.96

55.35

45–59

32.21

34.05

60 and older

5.52

6.33

Education of household head None or less than primary Primary Secondary Postsecondary

20.61

46.27

62.37

47.97

13.32

4.86

3.71

0.89

Migrant household

2.60

1.87

Household head’s language Speaks only indigenous language Speaks indigenous language and Spanish

7.59

26.24

92.41

73.76

Spouse’s language Spouse does not speak indigenous language Spouse speaks only indigenous language Spouse speaks indigenous language and Spanish

21.45

6.84

18.13

43.68

60.42

49.48

Child’s language

Child does not speak indigenous language Child speaks only indigenous language Child speaks indigenous language and Spanish

47.17

17.30

9.43

31.16

43.39

51.54

Education of household head’s spouse None or less than primary Primary

31.48

71.61

56.97

25.73

PP5331.indb5331.indb

2222

31.48 71.61 56.97 25.73 P P5331.indb 5331.indb 2 22 2 ( continued ) 1 10/15/10 0/15/10

(continued)

110/15/100/15/10

11:53:02:53:02 PPMM

31.48 71.61 56.97 25.73 P P5331.indb 5331.indb 2 22 2 ( continued ) 1 10/15/10 0/15/10
INDIGENOUS LANGUAGE USAGE AND MAINTENANCE 23 Table 3 ( continued ) Variable Percentage   Mexico

INDIGENOUS LANGUAGE USAGE AND MAINTENANCE

23

Table 3 (continued)

Variable

Percentage

 

Mexico

Guatemala

Secondary

9.97

2.36

Postsecondary

1.58

0.30

Asset index

40% lowest

85.62

68.24

40% middle

11.85

27.72

20% highest

2.53

4.04

Child’s age (mean)

11.64

11.96

Note: N = 290,615 (Mexico); N = 842,530 (Guatemala). Data from 2000 Mexican Census; 2002 Guatemalan Census. Mexican data are weighted.

Census; 2002 Guatemalan Census. Mexican data are weighted. This is in part because, though about 92

This is in part because, though about 92 percent of heads of children’s households in my Mexican sample speak both indigenous languages and Spanish, only about 74 percent of their Guatemalan counterparts do. Therefore, indigenous languages are more likely to be languages spoken in households in Guatemala than in Mexico. The table also shows that Guatemalan children tend to receive fewer years of education than their Mexican counterparts. A notable difference between the children in the two countries is that, whereas 21.5 percent of spouses of indigenous language speakers do not speak indigenous languages in Mexico, the vast majority of Guatema- lan indigenous speakers have spouses who do. Only about 7 percent of spouses of indigenous language speakers do not speak it themselves in Guatemala. This is in part because finding a person who speaks an indig- enous language is much more difficult in Mexico, where the proportion of indigenous people in the country is smaller. Note also that, as is the case among heads of households, spouses of indigenous language speakers in Guatemala are much more likely to be monolingual in indigenous lan- guages (43.68 percent) than their counterparts in Mexico. Table 4 shows the results from multinomial logistic regression mod- els predicting children’s probability of speaking indigenous languages either as their only languages or along with Spanish. Model 1 includes all variables except for spouse’s characteristics. The model indicates that, in both countries, children in households with high socioeconomic sta- tus are less likely to speak indigenous languages, which is true whether or not they speak these languages as their only languages. For example, household heads’ education is significantly, negatively correlated with in- digenous language usage among their children. In addition, the negative correlation is stronger among those children who speak only indigenous

PP5331.indb5331.indb

2233

among those children who speak only indigenous P P5331.indb 5331.indb 2 23 3 1 10/15/10 0/15/10

110/15/100/15/10

11:53:03:53:03 PPMM

among those children who speak only indigenous P P5331.indb 5331.indb 2 23 3 1 10/15/10 0/15/10

and Spanish

***

***

***

***

***

***

***

***

Indigenous

0.644

0.657

1.045

0.849

0.635

0.733

0.570

0.518

3.412

1.153

0.821

Model 2

Indigenous

***

***

***

***

***

***

***

***

Only

Table 4 Relative Risk Ratios from Multinomial Logistic Regression Indigenous Language Usage among Children of Indigenous Language Speakers

0.834

0.654

0.567

0.464

5.463

0.449

1.069

0.455

0.937

0.274

0.374

Guatemala

and Spanish

Indigenous

***

***

***

***

***

***

***

***

**

*

0.360

0.890

0.756

0.732

1.793

0.576

0.301

0.641

1.193

0.618

7.961

Model 1

Indigenous

***

***

***

***

***

***

***

***

***

Only

0.538

0.558

0.480

0.707

19.904

1.278

0.378

1.374

0.231

0.415

0.145

and Spanish

Indigenous

***

***

***

***

***

***

***

*

*

1.503

0.567

0.532

0.275

0.878

0.712

0.982

0.841

0.412

0.747

0.746

6.199

Model 2

***

***

***

***

***

***

***

***

Indigenous

*

*

Only

0.254

1.345

19.054

0.477

0.882

0.406

2.252

0.515

0.216

0.074

0.814

0.143

Mexico

and Spanish

Indigenous

***

***

***

***

***

***

***

***

***

**

2.302

0.665

0.852

0.369

0.324

0.708

0.429

1.078

11.943

0.541

0.831

0.612

Model 1

Indigenous

***

***

***

***

***

***

***

***

***

***

*

*

Only

0.380

4.042

58.365

0.630

0.206

0.263

0.045

0.599

0.072

0.425

0.420

2.072

Percentage of indigenous language speaker in municipality

Lives in urban area

Education (None or less than

More than 50.00%

More than 15,000

Locality size (Less than 2,500)

Postsecondary

(Less than 5%)

20.00–49.99%

60 and older

2,500–14,999

5.00–19.99%

primary)

Secondary

Primary

Female

(20–29)

30–44

45–59

Age

Secondary Primary Female (20–29) 30–44 45–59 Age P P5331.indb 5331.indb 2 24 4 1 10/15/10 0/15/10

PP5331.indb5331.indb

2244

Primary Female (20–29) 30–44 45–59 Age P P5331.indb 5331.indb 2 24 4 1 10/15/10 0/15/10 1
Primary Female (20–29) 30–44 45–59 Age P P5331.indb 5331.indb 2 24 4 1 10/15/10 0/15/10 1

110/15/100/15/10

11:53:03:53:03 PPMM

Primary Female (20–29) 30–44 45–59 Age P P5331.indb 5331.indb 2 24 4 1 10/15/10 0/15/10 1

***

***

***

***

***

***

***

***

**

*

***

0.843

0.330

0.566

1.585

148.677

0.797

0.065

1.055

0.528

0.975

32.763

0.574

–543,401.6

20,257.9

842,530

***

***

***

***

***

***

***

**

*

0.807

1.386

581.565

0.220

18.853

0.558

0.466

0.873

0.692

0.301

1.015

0.166

***

***

***

***

**

***

1.847

1.064

0.267

0.582

1.272

0.526

0.628

8,773.3

–642,712.1

842,530

***

***

***

***

***

0.236

0.066

0.095

18.579

1.519

0.746

0.901

***

***

***

***

***

***

***

***

**

*

***

0.840

0.560

0.827

2.230

0.680

0.665

0.428

1.049

0.954

0.799

87.177

18.016

0.115

12,980.7

–164,045.6

296,150

***

***

***

***

***

***

***

***

*

1.387

0.256

1.005

1752.722

0.784

0.202

0.725

0.515

0.157

0.125

4.136

0.167

49.910

***

***

***

***

**

***

0.422

1.055

0.409

2.312

0.795

0.315

1.159

0.961

6,091.3

–193,826.0

296,150

***

***

***

***

**

*

0.068

0.096

6.682

0.134

0.815

0.617

1.195

6.701

*p < 0.05; **p < 0.01; ***p < 0.001

Migrated Head speaks indigenous language and Spanish

Spouse’s language (Does not speak indigenous language) Speaks only indigenous language Speaks indigenous language and Spanish

Child’s age Migration × urban Migration × locality size 2,500–14,999 Migration × locality size more than 15,000

Education of head’s spouse (None or less than primary) Primary Secondary Postsecondary

Log-likelihood

(40% lowest)

Asset index

20% highest

40% middle

Constant

N

χ 2

lowest) Asset index 20% highest 40% middle Constant N χ 2 P P5331.indb 5331.indb 2 25

PP5331.indb5331.indb

2255

index 20% highest 40% middle Constant N χ 2 P P5331.indb 5331.indb 2 25 5 1
index 20% highest 40% middle Constant N χ 2 P P5331.indb 5331.indb 2 25 5 1

110/15/100/15/10

11:53:03:53:03 PPMM

index 20% highest 40% middle Constant N χ 2 P P5331.indb 5331.indb 2 25 5 1
26 Latin American Research Review languages than it is among bilingual children. Similarly, the asset
26 Latin American Research Review languages than it is among bilingual children. Similarly, the asset

26 Latin American Research Review

languages than it is among bilingual children. Similarly, the asset index is also negatively correlated with the probability of children speaking in- digenous languages, and the negative correlation is stronger among in- digenous monolingual children. Besides, Model 1 indicates that the nega- tive correlation is especially strong among children in Mexico because the proportion of indigenous language speakers in Mexico is much smaller than in Guatemala. The table also presents that children living in urban areas are signifi- cantly less likely to speak indigenous languages in both Mexico and Gua- temala, and this is especially the case among children who speak only indigenous languages. This makes sense because children face more occa- sions that require them to speak in Spanish in urban areas. In addition, in the case of Mexico, those living in municipalities with a high proportion of indigenous language speakers are significantly more likely to speak indigenous languages, especially as their only languages. Note that this correlation is significant in Guatemala only for those living in municipali- ties where at least half of inhabitants are indigenous language speakers among Guatemalan children who are monolingual in indigenous lan- guages. Overall, the positive correlation between the indigenous language prevalence rate at a municipality level and children’s indigenous language use is stronger in Mexico than in Guatemala, which is most likely because indigenous people are more likely to concentrate in particular areas in Mexico than in Guatemala. Also, controlling for other factors, there is no significant correlation be- tween a child’s migration experience and indigenous language use among Mexican children. This is probably because I controlled for community characteristics, and migrants are more likely to reside in larger communi- ties with a lower proportion of indigenous language speakers. Indeed, the interaction terms between migration and the community size show that children of migrant households living in a community with at least fifteen thousand inhabitants are 6.7 times more likely to be monolingual in indigenous languages and 2.3 times more likely to be bilingual than other children. In the case of Guatemala, Model 1 shows that migration experience is significantly correlated with children’s indigenous language usage among bilingual children. Children of recently migrated households are more than 1.8 times more likely to speak both Spanish and indigenous languages than are those children who did not recently migrate. Besides, the interaction term between migration and urban-rural status indicates that, though children who recently migrated and live in urban areas are significantly less likely to speak an indigenous language as their only lan- guages, they are more likely to speak it along with Spanish. Finally, Model 1 also considers household heads’ use of Spanish. The model shows that when a head of household speaks Spanish in addition

PP5331.indb5331.indb

2266

a head of household speaks Spanish in addition P P5331.indb 5331.indb 2 26 6 1 10/15/10

110/15/100/15/10

11:53:03:53:03 PPMM

a head of household speaks Spanish in addition P P5331.indb 5331.indb 2 26 6 1 10/15/10
INDIGENOUS LANGUAGE USAGE AND MAINTENANCE 27 to the indigenous language, his or her children are
INDIGENOUS LANGUAGE USAGE AND MAINTENANCE 27 to the indigenous language, his or her children are

INDIGENOUS LANGUAGE USAGE AND MAINTENANCE

27

to the indigenous language, his or her children are much less likely to speak an indigenous language in both Mexico and Guatemala. Therefore, we can infer that when people speak Spanish, they are less likely to teach their children to speak indigenous languages, which indicates the diffi- culty of preserving indigenous languages among younger generations. In addition to the covariates included in Model 1, Model 2 includes spouses’ characteristics. Although most covariates found to be statisti- cally significant in Model 1 remain significant in this model, their sub- stantive correlations with indigenous language use are not as strong in this model as in Model 1. Therefore, the model indicates that a spouse’s characteristics are also important factors in predicting children’s indig- enous language use. Indeed, household heads’ spouses’ ability to speak an indigenous language is most strongly related to children’s indigenous language usage. In both countries, the probability of children speaking only an indigenous language is significantly greater (1,753 times more in Mexico and 582 times in Guatemala) than for children of households in which spouses of household heads do not speak indigenous languages. Similarly, among those children who speak both indigenous languages and Spanish, whether or not household heads’ spouses speak indigenous languages is an important factor in predicting children’s use of indige- nous languages. The fact that a spouse’s characteristics are most strongly correlated with children’s indigenous languages indicates a few things. First, because people tend to marry those from similar socioeconomic backgrounds, children of economically advantaged households are much less likely to speak indigenous languages. This trend may lead to a further differen- tiation in socioeconomic status between indigenous language speakers and nonspeakers. Furthermore, a very strong and significantly positive correlation between a spouse’s indigenous language use and that of their children shows that it is very difficult to maintain an indigenous language even when one of the parents speaks it. In most cases, both parents need to speak an indigenous language to ensure that their children also speak it. Because such cases are not very common today, especially in Mexico, to preserve indigenous languages, it is imperative that children encounter opportunities to learn and practice indigenous languages outside their households, such as in bilingual education.

outside their households, such as in bilingual education. CONCLUSIONS The current study has examined statistical

CONCLUSIONS

The current study has examined statistical correlations between vari- ous socioeconomic, demographic and community characteristics, and indigenous language usage among self-identified indigenous people in Mexico and Guatemala. The analysis supported three of my hypotheses (H 1 , H 3 , and H 4 ). Among those who self-identify as indigenous, those liv-

PP5331.indb5331.indb

2277

those who self-identify as indigenous, those liv- P P5331.indb 5331.indb 2 27 7 1 10/15/10 0/15/10

110/15/100/15/10

11:53:03:53:03 PPMM

28 Latin American Research Review ing in urban areas, with a higher socioeconomic status, are
28 Latin American Research Review ing in urban areas, with a higher socioeconomic status, are

28 Latin American Research Review

ing in urban areas, with a higher socioeconomic status, are significantly less likely to speak indigenous languages. Similarly, the children of in- digenous speakers with a higher socioeconomic status are less likely to speak indigenous languages in both countries. It is important to note that, in the first part of the analysis, my sample is limited to those people who identify themselves as indigenous. Hence, although there is no way to identify those who are indigenous but do not identify themselves as such, it is probable that the negative correlation between various socioeconomic factors and indigenous language usage is even stronger. Contrary to my second hypothesis, I have found that there is no nega- tive correlation between migration and indigenous language use. How- ever, migration is partially correlated with indigenous language usage in both Mexico and Guatemala: those in Mexico who live in large com- munities and have recently migrated are more likely to speak indigenous languages whether or not they are bilingual. Similarly, in Guatemala, chil- dren in migrant households are more likely to speak both Spanish and indigenous languages than are nonmigrant children. Because people who migrate to urban areas tend to be socioeconomically better off than those who remain in rural areas, we should be concerned that socioeconomic differences between indigenous language speakers and nonspeakers may increase and that indigenous languages may further be seen negatively. Indigenous people in the two countries have suffered from severe socioeconomic status and discrimination for centuries. This situation has continued until today. What distinguishes the present situation from the past one are various opportunities that have become available to indig- enous people but have not resolved fundamental structural problems. This change resulted in a continuing but different type of pressure of assimilation. In the past, the pressure to assimilate came from the state through the discourse of mestizaje and bilingual education. Today, the state in both countries officially encourages ethnic diversity and cul- tural rights. Indigenous children remain in school for a longer period of time in both Mexico and Guatemala, and some indigenous people have achieved socioeconomic upward mobility. At the same time, as this study indicates, such socioeconomic advancements also threaten indigenous languages. This is because, as Garzon (1998a) indicates, the diversity of settings in which the indigenous and nonindigenous interact today leads many young indigenous people to accept a nonindigenous attitude. Such an attitude usually entails negative images of indigenous cultures. Given the finding that socioeconomically advantaged indigenous peo- ple are more likely to abandon indigenous languages and possibly also cultures, it is probable that a further negative image will be attached to indigenous cultures and people. Hence, as Sonntag (2003) argues, the rights-based approach to support for linguistic diversity does not work because, rather than combat conditions it opposes—such as the pressure

PP5331.indb5331.indb

2288

conditions it opposes—such as the pressure P P5331.indb 5331.indb 2 28 8 1 10/15/10 0/15/10 1

110/15/100/15/10

11:53:04:53:04 PPMM

conditions it opposes—such as the pressure P P5331.indb 5331.indb 2 28 8 1 10/15/10 0/15/10 1
INDIGENOUS LANGUAGE USAGE AND MAINTENANCE 29 of mestizaje and assimilation in the case of Mexico
INDIGENOUS LANGUAGE USAGE AND MAINTENANCE 29 of mestizaje and assimilation in the case of Mexico

INDIGENOUS LANGUAGE USAGE AND MAINTENANCE

29

of mestizaje and assimilation in the case of Mexico and Guatemala—it can reproduce them. Consistent with both my quantitative findings and Garzon’s (1998a) ar- gument, in my fieldwork in the western highlands of Guatemala, I have noticed that people who have achieved socioeconomic success hold a more nuanced understanding of inequality between indigenous and nonindig- enous groups. That is, instead of finding that current indigenous people’s situations are mainly caused by discrepancies in educational attainment, which is often the case among older generations, educated young indig- enous people tend to acknowledge that the barrier is difficult to surpass even if they have acquired a high level of education. Understanding this subtle but extremely hard-to-break barrier often drives indigenous people to abandon their affiliation. Therefore, taking the steps to improve cur- rent situations that indigenous people face, as Montejo (2002) proposed with reference to the Guatemalan Mayan—that indigenous movements must put an end to the century of silence and that indigenous people must take pride in their cultures—is extremely difficult in both Guatemala and Mexico. Moreover, the decreasing proportion of indigenous language speak- ers in both Mexico and Guatemala means that young indigenous people today, especially those with higher socioeconomic status, are more likely to marry nonindigenous language speakers or bilingual people. The sta- tistical analysis has shown that marriages between indigenous language speakers and those who speak Spanish greatly reduce the probability that the children will speak indigenous languages. This finding reflects that, though subordinate groups tend to be bilingual, members of dominant groups do not learn languages of the subordinate groups (Garzon 1998b). The acquisition of Spanish is almost ensured among indigenous children

because it gives them easier access to opportunities outside of their com- munities (Aubague 1986). Therefore, although indigenous parents may believe that their children will naturally learn indigenous languages even

if Spanish is used in their households, that is usually not the case, at least

in today’s Mexico and Guatemala. This is why preserving indigenous lan- guages will be more difficult even in the presence of some successes, as reported in England (2003). Nevertheless, I contend that we must make the effort to preserve and

promote indigenous languages. This may sound paradoxical given the findings of the statistical analyses, and the preservation of indigenous language itself is a contested discourse. However, I argue that, to achieve

a multicultural society in Mexico and Guatemala, preserving indigenous

language is fundamental, as languages influence people’s thinking and help them identify with particular ethnic and linguistic groups. There- fore, languages reflect the worldviews of ethnic groups and are the prin- cipal means to transmit such worldviews and cultural practices (England

PP5331.indb5331.indb

2299

such worldviews and cultural practices (England P P5331.indb 5331.indb 2 29 9 1 10/15/10 0/15/10 1

110/15/100/15/10

11:53:04:53:04 PPMM

such worldviews and cultural practices (England P P5331.indb 5331.indb 2 29 9 1 10/15/10 0/15/10 1
30 Latin American Research Review 2003; Watson 2006; Wurm 1991). In fact, indigenous languages are
30 Latin American Research Review 2003; Watson 2006; Wurm 1991). In fact, indigenous languages are

30 Latin American Research Review

2003; Watson 2006; Wurm 1991). In fact, indigenous languages are just one important component of indigenous cultures. Hence, language loss can also lead to more profound and problematic changes. For example, language loss can lead people—both nonindigenous and indigenous—to consider various indigenous cultures, such as their customs, religion, and languages, as one culture, which is not true: indigenous cultures are very diverse. Modiano (1988) has stated that, although indigenous people in Mexico are geographically concentrated, the indigenous people are so diverse in terms of their languages and culture that it is impossible to characterize the indigenous population as one ethnic group, except that they are overwhelmingly rural and poor. Although at the moment Mo- diano’s argument seems right because of the large number of indigenous languages that exist in the two countries, and because language is a very important aspect of many indigenous cultures and the most widespread symbol of self-identity (Fischer 2001), its loss may finally lead various indigenous cultures to become one culture, which might not be “indig- enous” at all anymore. Makoni and Pennycook (2007) have stated that languages are socially constructed and that definitions of language have material consequences for people. I argue that in societies in which boundaries among ethnic groups are not very clear, including Mexico and Guatemala, concepts of ethnicity and languages are closely related (Fishman 1989); therefore, such material consequences can have a significant impact on ethnic groups and their cultures. In fact, as Fishman (1989) claims, ethnicity is as modifi- able and manipulable as other human characteristics such as religion and ideology. The difficulty of defining indigenous groups in Latin America reflects this fact. This is why preserving indigenous languages is an ex- tremely important issue in Mexico and Guatemala. And the robustness of indigenous languages seems to correlate with that of indigenous cultures. Indeed, Lewin (1986) argued that the reduction in minority languages’ ca- pacity for social communication as a result of the use of Spanish in many communal spaces, which is the case for both Mexico and Guatemala, co- incides with the deterioration of cultural reproductive ability and that of communicative languages. Hence, as Hamel (1995) has contended, the survival of indigenous languages in Mexico and Guatemala is a decisive factor for the nature of the two countries: multicultural or homogenous. Today’s challenge to preserve indigenous languages is heightened because more and more people live in urban areas. Bilingual education needs to be implemented not only in rural areas where a large number of indigenous people reside but also in urban areas where many nonindig- enous people live. Nonindigenous people who send their children into bilingual education are most likely to resist bilingual education for at least two reasons. First, they will resist because of the low prestige of indig- enous languages and the stigma attached to indigenous cultures. Second,

PP5331.indb5331.indb

3300

stigma attached to indigenous cultures. Second, P P5331.indb 5331.indb 3 30 0 1 10/15/10 0/15/10 1

110/15/100/15/10

11:53:05:53:05 PPMM

stigma attached to indigenous cultures. Second, P P5331.indb 5331.indb 3 30 0 1 10/15/10 0/15/10 1
INDIGENOUS LANGUAGE USAGE AND MAINTENANCE 31 as discussed previously, nonindigenous people who are most likely
INDIGENOUS LANGUAGE USAGE AND MAINTENANCE 31 as discussed previously, nonindigenous people who are most likely

INDIGENOUS LANGUAGE USAGE AND MAINTENANCE

31

as discussed previously, nonindigenous people who are most likely to in- teract with indigenous people are those who are economically disadvan- taged and who are most likely to feel abandoned by the state because of neoliberal multicultural reforms and threatened by the rise of indigenous activism. If current multicultural reforms cannot destigmatize indigenous cultures and decrease tensions between ethnic groups, the reforms can- not reinforce indigenous language usage. Rather, such reforms can dis- courage people from using indigenous languages. England (2003) argues that linguistics alone cannot ensure language retention. The current study has indicated that granting cultural rights is not sufficient to ensure the maintenance of indigenous languages either. And cultural rights cannot be ensured unless other rights, such as socio- economic and political rights, are also actively granted. This is because today’s multicultural reforms reflect what Riding (1985) noted more than twenty years ago about Mexico, which I argue is also applicable to Gua- temala. That is, although the two countries seem to be proud of their in- digenous pasts, they are ashamed of their indigenous present. And I must add that not only are mestizos and Latinos ashamed of the indigenous present; some indigenous people are also ashamed. I contend that the goal of today’s indigenous language preservation must be to help people speak both indigenous languages and Spanish and to ensure that they are included in societies rather than that they speak only indigenous languages and are economically marginalized. There- fore, teaching only indigenous languages, which Haviland (1982) sug- gested almost three decades ago, is not a viable option today. Similarly, discussing Guatemalan Mayan indigenous groups, Brown (1996, 167) stated that, “since the Spanish invasion, the survival of [indigenous cul- tures] has often depended on the successful mastery of Spanish cultural elements, including language, as an addition to, but not a replacement for, [indigenous] culture.” Given the diversity of indigenous languages in both Mexico and Guatemala, the acquisition of Spanish is an important step for pan-ethnic movements such as pan-Mayan movements, which Montejo (2002) believes to be necessary to legitimize cultural projects of revitalization. The inclusion of indigenous people through destigmatization of in- digenous cultures, including their languages, cannot be achieved un- less serious problems such as the high level of inequality and tensions among ethnic groups are resolved. Today’s indigenous language pres- ervation programs also require support from nonindigenous groups— without their supports, language preservation programs will not work. Therefore, language preservation programs must be planned simultane- ously with improvement in both indigenous and nonindigenous peoples’ socioeconomic, political, and cultural status. Such a plan must include the reconstruction and disinvention of indigenous languages (Makoni and

PP5331.indb5331.indb

3311

disinvention of indigenous languages (Makoni and P P5331.indb 5331.indb 3 31 1 1 10/15/10 0/15/10 1

110/15/100/15/10

11:53:05:53:05 PPMM

disinvention of indigenous languages (Makoni and P P5331.indb 5331.indb 3 31 1 1 10/15/10 0/15/10 1
32 Latin American Research Review Pennycook 2007), which entail a rigorous analysis of processes through

32 Latin American Research Review

Pennycook 2007), which entail a rigorous analysis of processes through which today’s indigenous languages have been constructed and classified and have led to the current low status of indigenous languages. Hence, the establishing of effective and valid language preservation strategies also needs to take into account poverty and the difficulties that many indigenous people continue to face today and to rectify past injustices. Therefore, such strategies must seriously consider how indigenous cul- tural rights can actually be protected and promoted.

cul- tural rights can actually be protected and promoted. REFERENCES Aubague, L. 1986 “Desplazamiento o

REFERENCES

Aubague, L.

1986 “Desplazamiento o afianzamiento de las lenguas indígenas de Oaxaca.” In Etni-

cidad y pluralismo cultural: La dinámica étnica en Oaxaca, edited by A. M. Barabas and M. A. Bartolomé, 371–399. Mexico City: Instituto Nacional de Antropología e Historia. Beck, S. H., and Mijeski, K. J.

2000 “Indígena Self-Identity in Ecuador and the Rejection of Mestizaje.” Latin American Research Review 9 (1): 119–137.

Bell, C. S.

1996 “Data on Race, Ethnicity and Gender: Caveats for the User.” International Labour Review 135: 535–551.

Bravo Ahuja, G. R. D.

1992 “The Process of Bilingualism in a Multiethnic Context.” International Journal of the Sociology of Language 96 (1): 45–52.

Brown, R. M.

of the Sociology of Language 96 (1): 45–52. Brown, R. M. 1996 “The Mayan Language Loyalty

1996

“The Mayan Language Loyalty Movement in Guatemala.” In Maya Cultural Ac- tivism in Guatemala, edited by E. F. Fischer and R. M. Brown, 165–177. Austin:

University of Texas Press.

1998

“Mayan Language Revitalization in Guatemala.” In The Life of Our Language:

Kaqchikel Maya Maintenance, Shift, and Revitalization, edited by S. Garzon, R. M. Brown, J. B. Richards, and Wuqu’ Ajpub’, 155–170. Austin: University of Texas Press. England, N. C.

1996

“The Role of Language Standardization in Revitalization.” In Maya Cultural Ac- tivism in Guatemala, edited by E. F. Fischer and R. M. Brown, 178–194. Austin:

University of Texas Press.

2003

“Mayan Language Revival and Revitalization Politics: Linguists and Linguistic Ideologies.” American Anthropologist 105 (4): 733–743.

Filmer, D., and L. H. Pritchett

2001 “Estimating Wealth Effects without Expenditure Data—or Tears: An Application to Educational Enrollments in States of India.” Demography 38 (1): 115–132.

Fischer, E. F.

2001 Cultural Logics and Global Economics: Maya Identity in Thought and Practice. Austin:

University of Texas Press.

Fishman, Joshua A.

1989 Language and Ethnicity in Minority Sociolinguistic Perspective. Clevedon, U.K.:

Multilingual Matters.

Garzon, S.

1998a

“Conclusions.” In The Life of Our Language: Kaqchikel Maya Maintenance, Shift, and

1998b

Revitalization, edited by S. Garzon, R. M. Brown, J. B. Richards, and Wuqu’ Ajpub’, 188–199. Austin: University of Texas Press. “Indigenous Groups and Their Language Contact Relations.” In The Life of Our Language: Kaqchikel Maya Maintenance, Shift, and Revitalization, edited by S. Garzon,

PP5331.indb5331.indb

3322

Shift, and Revitalization , edited by S. Garzon, P P5331.indb 5331.indb 3 32 2 1 10/15/10

110/15/100/15/10

11:53:05:53:05 PPMM

INDIGENOUS LANGUAGE USAGE AND MAINTENANCE 33 R. M. Brown, J. B. Richards, and Wuqu’ Ajpub’,

INDIGENOUS LANGUAGE USAGE AND MAINTENANCE

33

R. M. Brown, J. B. Richards, and Wuqu’ Ajpub’, 9–43. Austin: University of Texas

Press.

Gonzalez, M. L.

1994 “How Many Indigenous People?” In Indigenous People and Poverty in Latin Amer-

ica: An Empirical Analysis, edited by G. Psacharapoulos and H. A. Patrinos, 21–39. Washington, D.C.: World Bank. Hale, C. R.

2002 “Does Multiculturalism Menace? Governance, Cultural Rights and the Politics of Identity in Guatemala.” Journal of Latin American Studies 34: 485–524.

Hale, K., M. Krauss, L. J. Watahomigie, A. Y. Yamamoto, C. Craig, L. V. M. Jeanne, and N. C. England

1992 “Endangered Languages.” Language 68 (1): 1–42.

Hall, G., H. M. Layton, and J. Shapiro

2006 “Introduction: The Indigenous People’s Decade in Latin America.” In Indigenous Peoples, Poverty and Human Development in Latin America: 1994–2004, edited by

G.

Hall and H. A. Patrinos, 1–24. Basingstoke, U.K.: Palgrave Macmillan.

Hamel, R. E.

1994 “Linguistic Rights for Amerindian Peoples in Latin America.” In Linguistic Hu-

man Rights: Overcoming Linguistic Discrimination, edited by T. Skutnabb-Kangas,

R. Phillipson, and M. Rannut, 289–303. Berlin: Mouton de Gruyter.

1995 “Indigenous Language Loss in Mexico: The Process of Language Displacement in

Verbal Interaction.” In The State of Minority Languages: International Perspectives on Survival and Decline, edited by W. Fase, K. Jaspaert, and S. Kroon, 153–172. Exton, PA: Swets and Zeitlinger. Haviland, J. B.

1982 “El problema de la educación bilingüe en el área tzotzil.” América Indígena 17 (1):

en el área tzotzil.” América Indígena 17 (1): 147–170. Hawkins, R. J. 2005 “Language Loss in

147–170.

Hawkins, R. J.

2005 “Language Loss in Guatemala: A Statistical Analysis of the 1994 Population Cen- sus.” Journal of Sociolinguistics 9 (1): 53–73.

Hill, J. H., and K. C. Hill

9 (1): 53–73. Hill, J. H., and K. C. Hill 1980 “Mixed Grammar, Purist Grammar, and

1980 “Mixed Grammar, Purist Grammar, and Language Attitudes in Modern Na- huatl.” Language in Society 9 (3): 321–348.

Jonas, S.

2000 Of Centaurs and Doves: Guatemala’s Peace Process. Boulder, CO: Westview Press.

Kampwirth, K.

2004 Feminism and the Legacy of Revolution: Nicaragua, El Salvador, Chiapas. Athens: Ohio University Press.

Layton, H. M., and H. A. Patrinos

2006 “Estimating the Number of Indigenous Peoples in Latin America.” In Indigenous Peoples, Poverty and Human Development in Latin America: 1994–2004, edited by

G. Hall and H. A. Patrinos, 25–39. Basingstoke, U.K.: Palgrave Macmillan.

Lewin, P.

1986. “Conflicto sociocultural y conciencia lingüística en Oaxaca.” In Etnicidad y plu- ralismo cultural: La dinámica étnica en Oaxaca, edited by A. M. Barabas and M. A. Bartolomé, 331–369. Mexico City: Instituto Nacional de Antropología e Historia.

Makoni, S., and A. Pennycook

2007 “Disinventing and Reconstituting Languages.” In Disinventing and Reconstituting Languages, edited by S. Makani and A. Pennycook, 1–41. Clevedon, U.K.: Multi- lingual Matters.

Mallon, F. E.

1992 “Indian Communities, Political Cultures, and the State in Latin America, 1780– 1990.” Journal of Latin American Studies 24 (suppl.): 35–53.

Modiano, N.

1988 “Public Bilingual Education in Mexico.” In International Handbook of Bilingualism and Bilingual Education, edited by C. B. Paulston, 313–327. New York: Greenwood

PP5331.indb5331.indb

3333

Press.

Paulston, 313–327. New York: Greenwood P P5331.indb 5331.indb 3 33 3 Press. 1 10/15/10 0/15/10 1

110/15/100/15/10

11:53:05:53:05 PPMM

34 Latin American Research Review Montejo, V. 2002 “The Multiplicity of Mayan Voices: Mayan Leadership

34 Latin American Research Review

Montejo, V.

2002 “The Multiplicity of Mayan Voices: Mayan Leadership and the Politics of Self- Representation.” In Indigenous Movements, Self-Representation, and the State in Latin America, edited by K. B. Warren and J. E. Jackson, 123–148. Austin: University of Texas Press.

Nagel, J.

1995 “American Indian Ethnic Renewal: Politics and the Resurgence of Identity.” American Sociological Review 60 (6): 947–965.

Patrinos, H. A.

1994 “Methods and Data.” In Indigenous People and Poverty in Latin America: An Empiri- cal Analysis, edited by G. Psacharapoulos and H. A. Patrinos, 41–54. Washington, D.C.: World Bank.

Patthey-Chavez, G. G.

1994 “Language Policy and Planning in Mexico: Indigenous Language Policy.” Annual Review of Applied Linguistics 14: 200–219.

Ramirez, A.

2006 “Mexico.” In Indigenous Peoples, Poverty and Human Development in Latin America:

1994–2004, edited by G. Hall and H. A. Patrinos, 150–197. Basingstoke, U.K.: Pal- grave Macmillan. Richards, J. B., and M. Richards

1997 “Mayan Language Literacy in Guatemala: A Socio-Historical Overview.” In In-

digenous Literacies in the Americas: Language Planning from the Bottom Up, edited by N. H. Hornberger, 189–211. Berlin: Mouton de Gruyer. Richards, M.

2003 Atlas lingüistico de Guatemala. Guatemala City: Editorial Serviprensa.

Riding, A.

1985 Distant Neighbors: A Portrait of the Mexicans. New York: Alfred A. Knopf.

A Portrait of the Mexicans . New York: Alfred A. Knopf. Roberts, B. R. 2005 “Citizenship,

Roberts, B. R.

2005 “Citizenship, Rights and Social Policy.” In Rethinking Development in Latin Amer-

ica, edited by C. H. Wood and B. R. Roberts, 137–158. University Park: Penn State University Press. Sagarin, E., and J. Moneymaker

1979 “Language and Nationalist Separatist, and Secessionist Movements.” In Ethnic

Autonomy—Comparative Dynamics: The Americas, Europe and the Developing World, edited by R. L. Hall. New York: Pergamon Press. Salinas Pedraza, J.

1997 “Saving and Strengthening Indigenous Mexican Languages: The CELIAC Expe- rience.” In Indigenous Literacies in the Americas: Language Planning from the Bottom Up, edited by N. H. Hornberger, 172–187. Berlin: Mouton de Gruyer.

Sieder, R.

2002 “Introduction.” In Multiculturalism in Latin America: Indigenous Rights, Diversity and Democracy, edited by R. Sieder, 1–23. Basingstoke, U.K.: Palgrave Macmillan.

Snipp, C. M.

1–23. Basingstoke, U.K.: Palgrave Macmillan. Snipp, C. M. 1989 American Indians: The First of This Land.

1989 American Indians: The First of This Land. New York: Russell Sage Foundation.

Sonntag, S. K.

2003 The Local Politics of Global English: Case Studies in Linguistic Globalization. Lanham, MD: Lexington Books.

Wade, P.

1997 Race and Ethnicity in Latin America. London: Pluto Press.

Watson, K.

2006 “Language, Education and Ethnicity: Whose Rights Will Prevail in an Age of

Globalisation?” International Journal of Educational Development 27 (3): 252–265. Wolf, E. R.

1986 “The Vicissitudes of the Closed Corporate Peasant Community.” American Eth- nologist 13 (2): 325–329.

Wurm, S. A.

1991 “Language Death and Disappearance: Causes and Circumstances.” In Endangered Languages, edited by R. H. Robins and E. M. Uhlenbeck, 1–18. Oxford, U.K.: Berg.

PP5331.indb5331.indb

3344

and E. M. Uhlenbeck, 1–18. Oxford, U.K.: Berg. P P5331.indb 5331.indb 3 34 4 1 10/15/10

110/15/100/15/10

11:53:06:53:06 PPMM

OUR INDIANS IN OUR AMERICA Anti-Imperialist Imperialism and the Construction of Brazilian Modernity Tracy Devine

OUR INDIANS IN OUR AMERICA Anti-Imperialist Imperialism and the Construction of Brazilian Modernity

Imperialism and the Construction of Brazilian Modernity Tracy Devine Guzmán University of Miami Abstract:

Tracy Devine Guzmán

University of Miami

Abstract: Indigenous peoples have been used and imagined as guardians of the Brazilian frontier since at least the mid-nineteenth century. This association was central to the foundation of the Indian Protection Service (Serviço de Proteção aos Índios, or SPI) during the early 1900s and culminated with the Amazonian Vigilance System (Sistema de Vigelância da Amazônia, or SIVAM) at the turn of the millennium. Throughout the period, the abiding desire to establish defensive do- minion over disputed national territory subjected individuals and groups identified as “Indians” to the power of overlapping discourses of scientific progress, national security, and economic development. A trinity of Brazilian modernity, these goals interpellated native peoples primarily through the practice and rhetoric of education, which grounds their historical relationship with dominant national society. Draw- ing on SPI records, government documents, journalism, personal testimonies, and visual media, this article traces the impact of this modernist trinity on indigenist policy and in the lives of those who have been affected by its tutelary power. By transforming private indigenous spaces into public domain, Brazil’s politics of anti- imperialist imperialism propagated a colonialist, metonymic relationship between “our Indians” and “our America” into the twenty-first century.

and “our America” into the twenty-fi rst century. FORMER INTERNATIONAL RESERVE OF AMAZON FOREST The image

FORMER INTERNATIONAL RESERVE OF AMAZON FOREST

The image in figure 1 has been circulating as part of an Internet chain letter since the year 2000. Alleged to be an excerpt from a geography book used in U.S. middle schools, it outlines the Amazon on a map of Brazil and neighboring countries. The text explains in poor English that the “most important rainforest in the world” has been seized from the “il- literate” and “primitive” peoples of South America, appropriated for safe- keeping by the United States and the United Nations, and renamed the Former International Reserve of Amazon Forest (FINRAF). The supposed author, neoconqueror David Norman, justifies the confiscation of national

My thanks to Bianca Premo, Bill Smith, Edilene Payayá, Denise Ferreira da Silva, Jan Hoff- man French, George Yúdice, Juvenal Payayá, Marc Brudzinski, Kate Ramsey, Sônia Ronca- dor, faculty and students of the Faculdade 2 de Julho, the Atlantic Studies Working Group at the University of Miami, and the three anonymous LARR reviewers for their comments and suggestions on previous versions of this work.

Latin American Research Review, Vol. 45, No. 3. © 2010 by the Latin American Studies Association.

PP5331.indb5331.indb

3355

2010 by the Latin American Studies Association. P P5331.indb 5331.indb 3 35 5 1 10/15/10 0/15/10

110/15/100/15/10

11:53:06:53:06 PPMM

36 Latin American Research Review Figure 1 “An Introduction to Geography.” territory from the “irresponsible,

36 Latin American Research Review

36 Latin American Research Review Figure 1 “An Introduction to Geography.” territory from the “irresponsible, cruel

Figure 1 “An Introduction to Geography.”

territory from the “irresponsible, cruel and authoritary [sic] countries” of Latin America by urging protection of the “world’s lungs” and referring to the rainforest as patrimony “of all humanity.” 1 Titled “A Amazônia é nossa!” this anonymous e-mail message led Internet users throughout the Americas to believe that U.S. schoolchildren have indeed been study- ing the “necessary” takeover of Brazilian territory and its conversion to

1. The Normans conquered territories from Hastings to Syria between the eleventh and the early thirteenth centuries. The historian David Nicolle (1987, 4) associates them with military prowess, ruthlessness, “business sense,” and “appreciation for money.”

PP5331.indb5331.indb

3366

sense,” and “appreciation for money.” P P5331.indb 5331.indb 3 36 6 1 10/15/10 0/15/10 1 1:53:07

110/15/100/15/10

11:53:07:53:07 PPMM

OUR INDIANS IN OUR AMERICA 37 FINRAF. Between 2000 and 2005, the bogus e-mail persuaded
OUR INDIANS IN OUR AMERICA 37 FINRAF. Between 2000 and 2005, the bogus e-mail persuaded

OUR INDIANS IN OUR AMERICA

37

FINRAF. Between 2000 and 2005, the bogus e-mail persuaded so many people that the U.S. State Department (2005) and the Brazilian Embassy in Washington were compelled to post disclaimers on their respective Web sites attesting to the falseness of its allegations. 2 As have other perceived infringements of national sovereignty, spurious or otherwise, FINRAF draws on a telluric anti-imperialism that pervades dominant Brazilian culture and has associated the country’s interior with the heart and soul of Brazilianness since long before the declaration of the First Republic, in 1889. From nineteenth-century Indianist fiction (Treece 2000) to twentieth-century Amerindian symphonies (Béhague 2006) and political cartoons “commemorating” Brazil’s colonization at the turn of the millennium (Bundas 2000a, 2000b; Devine Guzmán 2005), the Amazon and its inhabitants appear repeatedly in national cultural production as a patriotic trope to represent the whole or imagined essence of the coun- try. In policy initiatives and the popular imaginary alike, patriotic re- sponses to looming violations of territorial sovereignty—like FINRAF— have triggered the preemptive takeover of “virgin” Amazonian lands that are deemed lacking of state occupation, control, and ownership. 3 But of course, the original “owners” of these territories have “occupied” and “controlled” them for thousands of years. Here I consider these Amazonian imaginaries, the ongoing processes of Brazilian modernity, and anti-imperialist articulations of national iden- tity in relation to those whose fate has been implicated in all three sets of discourse: “the Indians,” real and imagined. 4 I argue that the imperialist logic of rightful possession reflected in the bogus geography textbook is, paradoxically, analogous to the Brazilian state’s own historical rationale of interior occupation and border defense. 5 This discourse dates to the

and border defense. 5 This discourse dates to the 2. Ambassador Rubens Antônio Barbosa attributed the

2. Ambassador Rubens Antônio Barbosa attributed the fabrication to retired military

personnel and the anonymous Web site Brasil, Ame-o ou Deixe-o (a slogan associated with

the 1964–1985 dictatorship). Paulo Roberto de Almeida (minister-counselor of the Brazilian embassy in Washington) documented the circulation of the rumor in Brazil by university professors and members of Congress, in addition to countless outraged citizens. Having condemned the uncritical embrace of the message and its accompanying ultranationalist discourse by the same intellectual left that had once opposed the military dictatorship, Almeida was charged by both sides with support for U.S imperialism (see ht