The Teaching of Spanish as a Second

Language in an Indigenous Bilingual

Intercultural Curriculum

Rainer Enrique Hamel

Department of Anthropology, Universidad Auto´noma Metropolitana,

Mexico City, Mexico

Norbert Francis

College of Education, Northern Arizona University, Flagstaff,

Arizona, USA

This paper reports on the implementation of an ambitious bilingual instructional programme in the P’urhepecha-speaking region of Michoaca´n state in Mexico, the Meseta Tarasca. A curriculum of indigenous language preservation and cultural affirmation, overturning the previous Spanish-only programme, has been developed by a group of indigenous teachers in two P’urhepecha elementary schools, ‘Miguel Hidalgo’ of San Isidro and ‘Benito Jua´rez’ in the neighbouring village of Uringuitiro. Today, the P’urhepecha language is the nucleus of the curriculum. With the previous curriculum largely discredited, the bilingual teachers embarked on a project that would both provide instruction to children in a language they understand, and contribute to the preservation of their indigenous language, which in these communities, in all cases, is children’s first language (L1). Being cognizant of the importance of learninSpanish as a second language, a major current planning and curriculumdesign priority is to find a way to integrate Spanish language instruction into the academic subject areas in accordance with current models of content-based second language teaching. doi: 10.2167lcc313.0

Keywords: bilingual curriculum, Common Underlying Proficiency, indigenous

language bilingualism, literacy, Mexico, Spanish as a second language

Introduction

In 1995 the P’urhepecha teachers of two bilingual elementary schools in

Michoaca´n, in the central Highlands of Mexico, introduced radical changes

to the previous curriculum which had been based on the fast transition to

Spanish and submersion L2 Spanish instruction. From that school year

onwards, they have been teaching all subject matter including literacy and

mathematics in P’urhepecha, the children’s first language. In this paper we

take up some of the special circumstances that educators need to take into

account that may be different from those in which the social imbalance

between the languages of the bilingual community is less pronounced. In the

communities in which the study has been carried out, the overwhelming

majority of children entering first grade in ‘Miguel Hidalgo’ and ‘Benito

Jua´rez’ bilingual elementary schools are monolingual speakers of the

171

0790-83180602 171-18 $20.000 # 2006 R.E. Hamel & N. Francis

LANGUAGE, CULTURE AND CURRICULUM Vol. 19, No. 2, 2006

indigenous language. Outside of school, in fact, P’urhepecha dominates all

interpersonal communicative language use domains; and access to Spanishlanguage

television programming is significantly more limited than in other

rural communities in Central Mexico. The new bilingual programmes face a

new challenge, born of their initial success in attracting and retaining signifi-

cantly larger numbers of students, an approximately 60% increase in total

enrolment (Hamel & Iba´n˜ ez Caselli, 2000), in comparison to previous years

when perhaps a kind of early selection was imposed based on children’s

ability to benefit academically from instruction exclusively in Spanish. What

kind of programme design and distribution of languages across the curriculum

will serve the triple objectives of indigenous language developmentrevitalisation

and a significant improvement both in general academic achievement and

in the second language learning of Spanish? This question is posed most

immediately for children whose primary or sole contact with the national

language is in school, those who perhaps are more likely to stay in school

now thanks to the current linguistically inclusionary approach.

In the broad international discussion on bilingualism and school language

policy, two rationales could be advanced for the inclusion of a vernacular

that is children’s mother tongueprimary language (MTL1): (1) strong ethnolinguistic

loyalty on the part of a significant portion of the speech community

which supports an active project of language preservation or revitalisation, and

(2) if the indigenous language (IL) is the only language that children understand,

its exclusion from the curriculum represents a potentially serious

obstacle to academic achievement for many children, in particular in the case

of literacy learning. Or inverting the terms: the inclusion of the IL has the potential

of significantly upgrading children’s academic achievement, including the

skills of reading and writing.

Logically, and often in practice, the rationales of language preservationrevitalisation

(1) and the linguisticpedagogical (2) are separate. The latter may

come to be an important ingredient in facilitating initial access to academic discourse

and literacy in the absence of a broad community-wide revitalisation

project (although acquiescence would be a minimum condition). The former

rationale would also be sufficient even in the context of child bilingualism in

which the great majority of the first grade population is comprised of fluent

speakers of the national language (NL), as is the case, for example, in most

IL bilingual programmes in Canada and the United States. In the participating

Michoaca´n communities of the present study, by all accounts, the two rationales

coincide and mutually reinforce each other (Alonso et al., 2001; Hamel &

Iba´n˜ez Caselli, 2000; Silva Castello´ n, 2004), making for a privileged site for

observation of a bilingual instructional programme in which the IL forms an

integral part of the core curriculum. Extensive ethnographic description has

in fact confirmed this programmatic feature (Bernabe´, 2006; Hamel, in press),

placing this schoolcommunity project on a short list of current pilot experiments

in IL bilingualism of this type in Latin America.

The previous Spanish-only model proved to be especially problematic for

children with little or no contact with the language of literacy and instructional

discourse outside of school. We could say that this posed a persistent and

deep contradiction between the official objective of elementary schooling,

172 Language, Culture and Curriculum

literacy, and a necessary objective, learning Spanish as a second language,

resulting in neither being attained satisfactorily (Hamel, 1988). Despite official

educational policy, in place for the last two decades, that would seem to

address this contradiction, little tangible progress is in evidence in rural

schools that serve children who are beginning learners of Spanish as a

second language.1

In the 1980s language policy across Latin America began to gradually shift

toward establishing new relationships between the central educational authorities

and indigenous peoples. Recognition of indigenous languages as a

part of the nation’s cultural patrimony was extended to school language

policy that granted communities the right to incorporate their languages into

the academic programme of public schooling, including explicit endorsements

of developmental bilingual instruction based on the general principles of intercultural

curriculum design (DGEI, 1990). While implementation has clearly

lagged far behind, official policies have opened the way for a number of important

experimental programmes, research projects, and most importantly, a small

number of community-based initiatives spearheaded usually by a young

generation of indigenous teachers fully proficient in the community language

and introduced to the scientific literature on bilingualism and bilingual

education during their studies in the various regional campuses of the

Universidad Nacional Pedago´gica.

Returning to the international discussion on bilingualism and school

language policy, the now well-known debates touch on even more fundamental

problems into which the Michoaca´n project inserts itself in a very self-conscious

way. Among the different perspectives specifically related to indigenous

languages and other vernaculars, we could frame one set of central questions

as follows. Three distinct hypotheses would make different predictions regarding

the inclusion of the IL (in some significant proportion) into the academic

curriculum, and specifically as a medium of literacy teaching:

(1) That it would come to represent an unnecessary obstacle to general

academic achievement, learning to read and write, and, pointedly, to

full acquisition of the nationalsecond language (NLL2).

(2) Neither obstacle nor expediting factor – that exclusive NLL2 medium

instruction poses no significant disadvantage to monolingual MTL1

speakers vis-a`-vis the three learning domains (general academic achievement,

literacy, and L2 learning of the national language).

(3) That all other extraneous considerations held constant, the inclusion of the

indigenous language represents a facilitative factor in at least one or two of

the above learning domains, and no disadvantage to any; and that learning

is expedited as a function of child monolingual speakers’ extracurricular

access to the NLL2 (i.e. in the case of Mexico, the benefit of MTL1

inclusion would be potentially most important for students with the

least contact with Spanish outside of school).

The Michoaca´n project forms part of a research current that has set out to

find supporting evidence for hypothesis (3), fundamentally an updated and

modern version of the UNESCO (1953) proposal on the use of vernacular

languages in school (Crawford, 2000; Hovens, 2002; Tabors & Snow, 2001);

Teaching Spanish as a Second Language 173

for Latin America: Escobar (2004), Hamel (2000) and Lo´pez (1998, 2001).

Proponents of both strong and weak versions of hypothesis (1) congregate

around a series of politically driven proposals, most notably associated with

the English-only movement in the United States (Porter, 1990). The clearest

exposition of the ‘no-difference’ hypothesis (2) is Wagner (1998), although

perhaps presenting a slightly qualified version.

Application of the Concept of the Common Underlying Proficiency

to Indigenous Language Bilingualism

A central theoretical framework that guides the analysis of findings from San

Isidro and Uringuitiro is the Common Underlying Proficiency (CUP) model of

Cummins (2000, in its most recent presentation).2 The next section will propose

a series of permutations of the well-known double-iceberg figure for the

purpose of highlighting the relationship between linguistic competence in L1

andor L2 and the development of cognitiveacademic abilities that are

‘shared’ in a purported ‘common’, or ‘central’, domain, but that nevertheless

require for their construction and consolidation a linguistic medium.

Crucially (as the above hypothesis (3) implies), the linguistic competencies

that correspond to this medium must be sufficiently developed so as to

ensure a minimum threshold level of comprehension and sufficiently developed

expressive capability on the part of preliterate child learners.

A previous study by Francis (2000, 2004) also sought to apply Cummins’

framework to the special sociolinguistic circumstances of indigenous language

bilingualism, although quite different from those of the P’urhepecha schools.

In an assessment of bilinguals’ ability to access abilities and skills from the

CUP, in this case learned primarily through Spanish, and stored in the

common ‘underlying’, or central, domains (where the icebergs ‘overlap’), children

were presented with academic language tasks in their indigenous

language (Nahuatl, from Tlaxcala state). Findings showed that access to these

(‘non-language-bound’) proficiencies was relatively unfettered, one could say

surprisingly so (although not completely) given the sharp sociolinguistic and

material imbalances that would conceivably favour performance in Spanish

in an overwhelming and one-sidedly dominating way. Performance in

Spanish did appear to be somewhat superior, as would be expected given

that it was the language in which children practiced their literacy skills.

Nevertheless, performance in Nahuatl on these same skills, even though they

were rarely practiced, showed significant upward tendencies of improvement

across the grades. One conclusion that presented itself for further inquiry was

that the Cummins model – specifically the autonomy (from L1 and L2) of

conceptual structures, metacognitive and discourse organising abilities, information

processing mechanisms related to literacy, etc. – offers a uniquely

useful way of analysing bilingual proficiency. And that the unfavourable distribution

of resources and the overbounding social imbalances of all kinds that

might militate against developing literacy-related abilities in an indigenous

language do not significantly obstruct or short-circuit the availability of CUPdomain

abilities.

174 Language, Culture and Curriculum

In a number of ways the Michoaca´n bilingual programme presents a more

interesting context – than that of Tlaxcala and of most other pilot bilingual

programmes in Mexico – for assessing the applicability of Cummins’ model:

(1) locally and regionally, the indigenous language is more secure (e.g. virtually

universal, and full native-speaker competence among children in San Isidro