This paper examines the actual situation of Aymara language in Tarapacá (Region I), the traditional territory inhabited by these indigenous people in Chile. Its aim is to analyse the usage, knowledge and characteristics of the gradual displacement of Aymara language by Spanish, the dominant local language. To do so, the current levels of competence, conditions and contexts of usage and learning, both of the Aymara and Spanish languages are described. Finally, the sociolinguistic dynamics of continuity and change are discussed.

Keywords: Aymara language, linguistic vitality, displacement, northern Chile

Linguistic and sociolinguistic issues have occupied in Chile¹ a secondary place among the studies concerning indigenous people. This is especially true regarding the Aymara, which has received much less attention than the Mapuzungun, the Mapuche people’s native language (the most important indigenous group with regards to population and political relevance). This is due, partly, to the special requirements associated to this kind of investigation, but also to the distinct emphasis given to ethnic conflicts and political demands of native communities and organizations. Hence, analysis of the internal dynamics of indigenous cultures is overshadowed, despite the fact that it might become of great importance for both the accomplishment of the indigenous movement’s political objectives and indigenous policies. Academic literature on the sociolinguistic situation of indigenous languages in Chile has been limited, especially concerning the Aymara language. Most of the studies tackle the description of its internal structures, although during the last two decades sociolinguistic and historical dimensions have increasingly been incorporated.¹ During the same period, several sociolinguistic polls have been carried out among the regional Aymara population. The National Social and Economic Survey (CASEN survey) includes also, since 2000, a special sub-sample concerning indigenous people and a question on native languages. This has allowed generating data on very important matters, such as linguistic vitality and language use by the Aymara people.

The available information is also discontinuous over time. The first two sociolinguistic studies on the Aymara language in Chile were carried out in 1986 and 1990, respectively. Two new researches took place more than a decade later, in 2002 and 2005. Comparing data has also been difficult due to the fact that these studies have covered different geographical and social areas. Most importantly, historical and contextual factors have not been sufficiently evaluated, partly because of the methodology employed (social surveys) and the supporting institutions’ requirements, but also due to the predominance of a synchronic and internalist view within the linguistic theories. The present investigation aims, thus, to contribute to the development of a theoretical and methodological approach that could overcome these two limitations, assuming the historicity of language and the importance of extra-linguistic social factors in such transformations.

Our subject’s relevance goes beyond academia; it has obvious political, educational and cultural implications. The recognition of indigenous peoples’ rights has led to the fact that native languages have ceased to be considered, at least officially, as source of underdevelopment and an obstacle to the integration of indigenous people in Chile. They are commencing to be valued as an expression of the enrichment of our linguistic and cultural wealth, leading to the implementation of educational policies aimed at their acquisition and conservation, the Bilingual Intercultural Education (BIE). However, without a basic knowledge of the situation of vernacular languages and their trends of change it is impossible to design and implement effective measures of cultural protection and revitalisation of such languages. Hence the importance of reliably establishing the actual situation of the Aymara language and its trends of change.

The implementation of BIE has taken place after the language has been subject to an intense and prolonged pressure for its abandonment, especially from public agencies. The impact of these processes on native languages has been profound, leading namely to consider their possible disappearance in the near future. The data here presented does not endorse such a radical position, but shows a strong tendency towards the abandonment of the Aymara language in favour of Spanish. It also allows us to have a preliminary diagnosis on the intensity and extension of the native language’s displacement. Considering the scope of our subject, we have focussed our description on four main aspects: competence levels in Aymara and in Spanish, the acquisition and usage frequency of both languages and the degree of displacement due to the Spanish language.

Although the data and analysis here presented refer solely to Chile, and specifically to the Tarapacá and Arica-Parinacota regions, where we find the bulk of the Aymara population, it must be remembered that Aymara is spoken in three different countries: Bolivia, Perú and Chile. This tri-national condition is an influential factor on the national and regional linguistic dynamics, given the cross-border contacts and the migratory movements of Aymara population, mostly Bolivian, to the Chilean Andean zone. According to existing information, it seems that what distinguishes the Chilean situation from the others’ is the intensity of the Aymara’s displacement by the Spanish language and, in consequence, bilingualism’s decline and Spanish monolingualism’s advance.

¹Originally published in spanish in: Estudios Filológicos N°42, Universidad Austral de Valdivia, Chile, pp. 123-140. Translation by Pedro Araya

Data was obtained through an inquiry realised between 2002 and 2003 for the Program for Bilingual Intercultural Education (PBIE) of the Chilean Ministry for Education. The survey included the Aymara, Atacameño and Mapuche peoples. The interpretation here presented is new and the sole responsibility of the authors. We will focus only on the Aymara population from the General Lagos and Putre communes, in the Parinacota province, and the Huara, Camiñe, Colchane, Pica and Pozo Almonte communes, in the Iquique province. Due to the inquiry’s requirements, neither the Camarones commune, nor the Lluta and Azapa valleys from the Arica commune were included, despite having a significant population of Aymara origins.

Information obtained through a survey conducted for a representative sample of the existing rural population in the seven communes before mentioned, according to the 2002 census. The whole population, indigenous and non-indigenous, was included. This was due to the fact that, as we were dealing with an aleatory sample, both segments should have the same probability of being selected at the time of the drawing and according to their parameters. Only individuals above 18 years old were considered, as they were better able to account for their sociolinguistic acquis, determine with greater precision their linguistic competence and provide, in general, better quality information. Likewise, as the interview system –based on the person from the drawn home that received the interviewer or had disposition to answer the questions– did not secure a proportion of cases per genre in line with the parameters, a male-female ratio was established, using the existing masculinity index in each commune, considering only particular homes.

The study was carried out with a 5% for margin of sampling error; a weighting factor of 50% was applied to variance assumptions, and a confidence level of 2 Sigs, that is, of 95.5%. For the calculation of samples, a commonly used formula for finite universes was used. Although the whole of the sample was proportionally distributed between the communes, according to their proportion, unities were increased up to 30 in those were, when applying the survey, a lesser quota than this resulted. In this way, we secured a minimum number of cases in each commune. Concerning the sample base’s geographical location and the specific location of the entities considered in the sample, we used census and cartographic data from the National Institute of Statistics (NIS). Accordingly, the existing individuals in each commune were numbered, placing them in the minimum census aggregate corresponding to the sector, that is, a half block in urban or nuclear sectors, and houses or group of isolated houses in disperse residency sectors. Given that the 2002 census was not yet available with these levels of disaggregation, the inquiry was carried out through the 1992 census base of each region, using the REDATAM system. Finally, we also considered the fact that in rural areas population variations are generally weak and can be corrected on the field, adding new areas and houses.

The draw was done using a random selection of numbers’ special program. Both the cases corresponding per sample and the replacements were drawn. To identify them on the field, we established survey roadmaps accompanied by the corresponding cartography so as to locate and draw the homes (adding the variations that could currently exist since the 1992 census). Obtained information was codified and entered into a database for further analysis using the SPSS program (Statistical Package for Social Sciences). In the elaboration of information, we privileged a geographical dimension regarding the existing ecological levels in the regional rural area: the altiplano (Andean high plateau), the high valleys or the Andean foothills, and the low valleys and oases. These geographical areas determine different agricultural productive specializations, different connections with the regional society and divergent linguistic processes.

Population, ethnic affiliation and indigenous language speakers in the studied localities

The communities surveyed correspond to the Andean foothills and the altiplano areas, where the Aymara are majority, and the low valleys or oases, where the proportion of non-indigenous increases. As we shall see later, the distribution of indigenous cultural practices, among them language acquisition and use, is linked to the settlement area. According to the information assembled, the Aymara population is distributed throughout the area, despite important variations according to the geographical sector. In general, Aymara ethnic identification decreases through a longitudinal axis as we descend from the upper areas to the lower ones. It is clearly predominant in the altiplano (89.9%), which corresponds to the General Lagos and the Colchane communes, in their totality, and part of the Putre and Pica communes. In the Andean foothills area, which comprises the whole of the Camiña commune and a part of the Putre, Huara and Pozo Almonte communes, such identification experiences an important decrease but still remains predominant within the ensemble (58.5%). In the low valleys oases areas, which comprise the whole of the Pica commune and a part of the Huara commune, it comes down to a 29.1%, while in the pampa area, which includes a part of the Huara and Pozo Almonte communes, it reaches its lowest level: a 16.4%.

Given the survey’s characteristics and the size of the sample, we considered the self-declarations concerning competency, as it was not possible to realise direct measurements through an evaluation instrument applied by trained speakers. Despite this, the data obtained show a high internal consistency and are, in principle, coherent with previous surveys carried out in the same area.^2‒7 The majority of the population from the area studied (a 64.5%) declares lacking linguistic competency in any indigenous language. If we only consider the group of 181 persons that define themselves as Aymara, a 65.7% (119 cases) points out possessing competency (in various degrees) in their language. As predictable, these percentages experience important variations according to the geographical areas: while in the altiplano competency rises to a 95.2%, in the Andean foothills it decreases to a 61.2%, and in the low valleys, oases and the pampa areas it reaches only a 38.6%. As well as self-affiliation, Aymara language competency descends towards the lowest areas, a matter we shall comment later on (Table 1).

To know the level of active competency will allow to better appreciate the information recently presented. Of those who declared having competency in Aymara language, a little less than a third of them (30.9%) declares having a high linguistic production capacity; a 20.6%, meanwhile, affirms having a medium or even sufficient competency, i.e. they do not consider themselves conspicuous speakers, but neither do they believe possessing high limitations in their performance. Both sides add up to something more than half of the cases (a 51.5%). The rest are speakers who affirm being able to handle themselves actively, although with great difficulties and limitations, which in some cases could be severe, to the point of emitting only elementally (some lexica, greetings, circumstantial phrases, etc.). The latter represent something more than the half of those in this level of competency, if we consider that a 26.5% of the total estimate that they only dispose of a limited comprehension (they understand very little, they arrive to a reduced comprehension) of the indigenous language in the communication events in which they engage (Table 2).

Considering the characteristics of the sample, the proportions found allow to situate some benchmarks. Firstly, they are representative of the Aymara’s relevance in the communes surveyed, as the sample is proportional, with small adjustments, to the population of such communes. Secondly, as we are dealing with areas with a long and continuous indigenous residency,⁵ and where a major permanence of Aymara language and culture is recognized, it is in these that we can find the highest percentages of speakers. Thirdly, given that the urban centres concentrate most of the regional indigenous population (the urban indigenous reach a 78.5%), it could be that here reside a majority of speakers but hardly of competent speakers. We will return to this matter in the conclusions.

On the other hand, a quite high percentage (87.2%) of those who declare competency in Aymara language affirm a solvent competency in Spanish (Castilian), while the cases of low competency represent only a 12.8% (comparable, then, to bilinguals where the Spanish-Castilian is in a subordinate position). This high competency visibly contrasts with the 48.5% of low competency declared for Aymara. Similarly, broad Spanish (Castilian) proficiency contrasts with the much lesser proportion (30.9%) of persons assuring to have an equal proficiency in jaqui aru. This is confirmed and accentuated if we consider that approximately a third of those who identify themselves as Aymara is monolingual in Spanish, while Aymara monolingualism would have disappeared. The interviewees also agree to say that bilingual children are currently a rarity. Cases of Aymara-Spanish bilingualism among children and students appear only in the higher geographical areas, such as the altiplano. We are, then, in the presence of social contexts where Aymara-Castilian bilingualism is present in just over one third of the total of cases (35.5%), that is, where Castillian monolingualism clearly prevails.

Aymara and Spanish (Castillian) acquisition

Among those who declare competency in Aymara language, there is quite a heterogeneous situation concerning first language acquisition, which could be Spanish, Aymara or both languages at a time. The majority (a 44.8%) has learned simultaneously both languages; a 31.3% learned first Spanish and then Aymara. The inverse situation (learning Aymara and then Spanish) only appears in a 23.9% of cases and corresponds, mainly, to older people. Following the language’s trend towards permanency in those areas located at higher altitude, the major number of persons who declare having the Aymara as their first language reside in the altiplano (83.9%), a percentage that also includes those who learn it simultaneously with the Spanish. The invers occurs in the low areas. The distribution of first language acquisition indicates a rapid process of replacement by Spanish, which is being learned massively and early in all of the indigenous homes in the region’s rural area. In fact, there are hardly any more cases where people communicate first in Aymara and then in Spanish. Contrary to what Grebe and Salas stated two decades ago, Aymara, when so, is currently acquired together with or after Spanish.

The Andean language does not seem to be an object of a conscious learning effort such as the Spanish, quite the contrary: parents, voluntarily or not, tend to limit its usage and, by habit or decision, communicate in Spanish with their children and other persons from the town. Aymara is acquired only if there are favourable conditions and constitutes a common means of communication in the immediate social network. Its inter-generational transmission is reinforced by belonging to a family of active speakers, by living or being in permanent contact with high-Andean localities, where such a language has historically experienced less pressure from Spanish, or by having possibilities of frequently interacting with older people, who employ it regularly at home and in their town. Intervention of parents and other family members is crucial: the great majority of the interviewees who dominate Aymara learned it from their parents. The parental network and the locality (the town) can have a favourable influence for maintaining and, eventually, broadening the linguistic skills, but the family’s action is indispensable, it is where language is really acquired.

In contrast, the Andean population’s trans-localisation due to migrations, its redistribution throughout the region, and an increasing urban residency, produce a fragmentation of familiar and child-rearing networks. The nuclear family frequently appears outside rural localities, while broader forms of family grouping reside throughout the region. In fact, while Aymara-speaking older people reside mainly in rural towns, the majority of children and teenagers study and live in urban areas (except during brief periods of the year, where they can return to their birthplaces). Thus, those favourable conditions for the use of the Aymara and, above all, for its acquisition by new generations, cease to exist or diminish. As a consequence, today Spanish is largely dominant in the social interaction inside the home, family networks and the place of residence.

The differentiated use of Aymara and Spanish according to the given context has led to believe that there exists an institutionalised diglossia or diglossic situation between the two languages. The results here presented indicate that this is not true. Bilingualism has receded, and the uses and functions of both languages are not separated from each other in a well-balanced or complementary way. Even in the spaces of intra-ethnic interaction, such as communication inside homes or children’s education, Spanish is being extensively used or is alternated with Aymara.

The uses of Aymara and Spanish

The existing data evidences the Aymara language’s critical situation. This diagnostic can be confirmed when considering the conditions and characteristics of the use of indigenous language and Spanish. A first aspect concerns the geographical and social segmentation of Aymara usage. An important group of persons –the majority with undoubtedly indigenous ancestors– not only do not speak Aymara but also do not frequently hear it. In accordance with what we have seen, this phenomenon is more pronounced in the lower geographical areas. In these areas native language is little used and also presents internal segmentations that convey a socially differentiated use of Aymara. In fact, if we focalise on the Andean foothills, the low valleys and oases, we will see a notorious social distance between immigrants with a higher Aymara proficiency and native residents with no competency in this language. These settlements often interact intensely, but they do so in Spanish (Castilian), the language known by everyone and that clearly prevails in the communication. The use of Aymara, meanwhile, is circumscribed to the spaces of interaction between immigrants, the only ones who can speak it. When these interlocutors are more disseminated, as in the dessert villages or the coastal towns, the use of Aymara is limited to the home (if there are internal conditions and habits for doing so), or to communicative instances with members of parental networks and from the birthplace. In the altiplano ad the Bolivian frontier, however, Aymara is employed in a number of interactions, despite Spanish being also prevalent as a common language and in the external relations with educational agents and authorities.

Identifying the attributes concerning the social organisation of Aymara usage does not suffice. A second aspect concerns the periodicity of its usage. According to the information obtained, the frequencies of Aymara use are generally low. Again, variations according to altitudinal levels are notorious. Only a little over a third of those who speak it (37.3%) do it on a daily basis, a large majority of altiplano residents, while the rest employ it occasionally (a 30.5%) and rarely or almost never (a 32.2%). The distribution of cases follows the tendency already mentioned of a decrease in the language’s vitality along the Andean altitudinal gradient. In the altiplano it is frequently used, due to the existence of favourable conditions; on the other extreme, the low valleys, oases and pampa, fewer persons know it and use it very little or occasionally. In the foothills area, although there are more speakers than in the low areas, its use is also occasional or infrequent. We must remember that, both in the foothill valleys and the low areas, Aymara use is circumscribed to Bolivian immigrants and Chileans from the altiplano, and there are no native speakers.

The structuration of Aymara usage acquires, thus, a distinctive physiognomy: in the high-Andean communities a concordance between social and linguistic aggregates tends to appear; on the other hand, in the Andean foothills, low valleys and desert oases, there is an ethnic and social segmentation between those who have emigrated from the altiplano, and have competency in Aymara, and those who are originally from these areas and no longer understand or speak it. There is a symbolic frontier between both groups that defines and reinforces social stereotypes. The “native” identify the immigrants as a clearly different group and separated from them, and refer to them using devaluating denominations: “paisanos” (peasants) or “indios” (Indians). Doing so, they refuse for themselves such an indigenous condition and, with it, any association with the Aymara language; in occasions, even at the level of an historical memory. Despite this, during the last years this group has experienced a phenomenon of ethnic re-identification (especially induced by State agencies such as the CONADI) that, apparently, has not eliminated, but diminished the social distance with the other group, which could carry a favourable change concerning the attitude towards the Aymara language.

Aymara usage would then be organised according to two principles: the lesser or greater extension of the social interaction’s circle and its intra- or extra-ethnic character. The native language can be expected to be spoken mainly in social interactions between indigenous and persons who are socially close; that is, in intra-ethnic and socially restricted domains. On the contrary, its use decreases exponentially as we enter spaces of inter-ethnic relation and/or where a greater number of participants do not have family or community relations between them, and where the presence of Spanish monolinguals is common. Thus, a greater proximity in the communicative interaction will plausibly lead the involved actors to opt for the Aymara. On the contrary, the more public and external the ambience, Spanish will be mostly used (Table 3).

According to this table, the spaces of social interaction mostly associated to the use of Aymara are the homes, the relations with family members or the agricultural activities between indigenous peasants. On the contrary, community meetings, communication in the educational sphere and, we could add, many of the market relations, tend to be carried out in Spanish. If we consider formal contexts of communicative interaction, the tendency increases and the use of Aymara becomes very low or inexistent. For example, the 78.2% of respondents indicate that vernacular language is not used (it is not heard or not used) at the community’s school. The same happens in other spaces of local interaction, particularly in the presence of State agents who, excepting a few Aymara officials, completely ignore such language: the school, local governments and public institutions. In all of these, there is no other communication code than Spanish.

The longer we move away from the high lands towards the Pacific coast and the urban centres, the possibilities of Aymara usage proportionally decrease. It is not that its use is unthinkable outside ethnic spaces or that there is necessarily an explicit repression, but rather that in the presence of non-Aymara and Spanish monolingual persons, the change of code turns in the practice to be something automatic. The switch to Spanish happens also in the interactions between Aymara in which a non-speaker participates or when the persons consider they have a better performance in Spanish. Aymara can be even used to inform, concord or comment affairs over which one does not want one or more of the presents to find out, but that is something circumstantial, brief and normally discrete. In these cases, code alternation is usual, Aymara being doubly disadvantaged regarding Spanish: its alternation is more frequent in the opposite direction and Spanish is the language known by everyone and does not present any social rejection. Thus, at the level of interactions, Aymara usage depends on a number of factors: a) those who establish the dialogue; b) certainty regarding the linguistic capacities of those interacting; c) a reciprocal expectancy or disposition to dialogue in the language and, d) the context or communicative situation. If they are not favourably present, Spanish is used.

Nevertheless, the situations of interaction between Aymara and Chilean authorities currently tolerate a relation of absence/presence more complex than the simple exclusion of yore. Let us take the example of the rituals practiced by Aymara officiants at the beginning of a political meeting, which have received the name of “rogativa” (rogation) or pawa (in Aymara) and that nowadays are part of the Chilean public institutions’ official protocol. These ceremonies have a double character, political and religious, where very heterogeneous elements come together, both in the performance and the symbols being used. The rituals are preferently realised in Spanish, although some indigenous expressions or short phrases are used. Its usage responds to a double function: expressive and symbolic; language itself is transformed into a symbol at the expense of its communicative capacity. The officiants consider that the correct execution of rites must be done in Aymara, but accept or justify the use of Spanish given the language has been “forgotten”. In the case of a bilingual speaker, we would be before a form of linguistic interference, but the Aymara monolingual in Spanish recur to these linguistic substitutions as a way of tracing an ethnic frontier regarding the non-indigenous Chileans and as a way of showing their competence and fidelity to their culture. This symbolic use of Aymara is not accompanied y an interest in learning the language, similarly to the increasingly frequent use of English terms among Spanish speakers in Chile, mostly among the younger generations, not generally related neither with a functional necessity nor with a competency in it.⁸ An additional difficulty is represented by the fact that, for a number of reasons, the Aymara lexicon has become confined to traditional cultural representations and practices, while “modern” objects and experiences are exclusively dominated by Spanish, which corresponds to a lexical impoverishment.

That said, and considering the appearance of experiences such as the one recently discussed (language as a symbolic resource in political rituals), or that the State and the Andean indigenous leadership and intellectuality give opposite signals, we are not before an objective phenomenon of a relative decrease in the usage of Aymara regarding the functional spheres, where it can be employed as a means of communication, but also before the continuity of the language’s diminished cultural status for its speakers, who tend to consider it as “underdeveloped” and intrinsically limited with regards to Spanish. Obviously, such a limitation does not exist: this has more to do with the Aymara’s linguistic development, which lacks the supporting resources of languages such as the Spanish, written and codified since centuries, which allows it to be used as a means of communication and education in such dynamic spheres as the sciences, technology and economy. This does not happen with the indigenous languages, which until recently have been languages of an oral tradition and that, with the exception of the colonial grammars, have not been codified for their teaching and usage in formal spheres. In fact, in the bilingual intercultural education programs such languages are only incorporated as study assignment and not as tools to teach mathematics, sciences or other subjects, which ends by reaffirming their subordinate position against Spanish (Castilian).

The current inhabitants of the Aymara community areas of Tarapacá and Arica-Parinacota are mainly Spanish (Castilian) speakers. Only a 35.5% of the total of interviewees declared having competency in indigenous languages and, among those who recognize themselves as Aymara, just over the half of them (51.5%) show high or medium levels of active competency in their language. This means that a little less than a fifth (17.4%) of residents, indigenous and non-indigenous of the municipalities surveyed have a fluid usage of the jaqui aru. These data additionally correspond to the rural population (excepting Pica and Putre, as mentioned before), the area expected to be less affected by linguistic displacement.

Aymara’s vitality within the Andean space has been historically quite variable. The altiplano and the high central Andes –and with them, the Bolivian bordering regions– have been the great reservoirs for the conservation of this language, despite the rapid changes that these regions have experienced, particularly since the last decades. In the low valleys, the oases and the pampa, as with the regional urban centres, Spanish largely predominates. The high valleys and the foothills present an intermediate situation. In this area, some speakers remained until a few decades ago and it is the destiny of migrations from the altiplano and Bolivian Aymara bordering regions. Table 4 shows the declared usage of Aymara and Spanish (as a language of exclusive or prevalent use according to the declarants’ knowledge) by the ancestors of those interviewees who declare competency in their native language. Among the progenitors, a 57.2% spoke both languages equally well (the native language and an “Andean Spanish”),¹ while a 25.5% used predominantly the vernacular language. The latter are not necessarily monolingual in Aymara: they are rather bilingual, where Spanish occupies a subordinate position. Values in favour of the Aymara are slightly better if we consider the women (the interviewees’ mothers), which is consistent with the information that it is among women that we find a higher competency in Aymara.^3,9 Results also confirm the importance of the Chilean high-Andean and Bolivian provenance for the language’s vitality. In effect, a 63.7% of the informed progenitors were born in the altiplano and if we add those coming from Bolivia, the percentage increases up to a 82.4%. They concentrate the competency in the indigenous language: the 90.3% of cases where the prevalent use of Aymara among progenitors is declared and the 79.1% of those declared with equivalent use of both languages. If further detailed inquiries were to be realised regarding the cases of progenitors dominating the Aymara, on the part of those individuals born outside the high lands, probably most of them would have a high-Andean family background.

At the same time, information about the grandparents of the interviewees evidences results coherent with those observed on the progenitors with regards to bilingualism’s situation and tendencies. As expected, the percentage of those who predominantly use Aymara increases considerably (51.2%) and the one of those who employ both languages decreases (33.7%), while those who exclusively or predominantly use Spanish remains approximately in the same order of magnitude (15.1%). This group shows then a high level of Aymara vitality and a less frequency of Spanish, even when solidly rooted in some areas. Difference in competencies and usages of the languages when going from the group of grandparents to that of progenitors, lies in the increase of those who use both languages and the diminution and those who speak only or mainly Aymara. Furthermore, the relevance of these predecessors’ high-Andean and Bolivian provenance is confirmed: a 32.2% has Chilean high-Andean origins which, added to the Bolivian Aymara contributions, represents a 51.2% of cases.

Aymara language’s vitality in the low areas is explained by recent immigration of individuals and families from Aymara-phone areas, since there, language had been early displaced by Spanish, a process under way for more than a century.⁴ The new spatial distribution cannot be associated, then, to stable permanence of bilingual speakers in these areas. Currently, we have the paradoxical and problematic situation of having a progressively smaller proportion of speakers to total indigenous population, and who also are distributed throughout the region, even in urban areas. Aymara has lost its feature of being a language concentrated in the high Andean geography, to be spread along with its speakers throughout the region. However, with this it has become more exposed to the pressures coming from the dominant language, which confirms the relevance of migration processes to Aymara sociolinguistic dynamics. In fact, its irradiation is consistent with the redefinition of ethnic frontiers.

According to available historical information, until the first half of the 19th century, the Aymara language extended throughout the Andean space, including the “Indian” communities located in the low valleys and the mining centres next to Iquique, where Aymara labourers worked.⁴ Aymara was the language that presented the broadest distribution and the biggest contingent of speakers in the whole region, given that the indigenous formed the bulk of the population. Since then, a process of profound changes associated to the conformation of a mining cycle to export saltpetre began. This was accompanied by the development of the Peruvian state bureaucracy and the arrival of mestizo population to the area, including the Andean valleys. The Aymara participation in the mining industry and associated activities led many inhabitants of the low and middle valleys to begin to define themselves as non-indigenous, within a greater process of internal social differentiation, redefinition of social identities and the abandonment of the native language. Complementarily, the relation with the contingent of immigrants who were monolingual in Spanish, in the mining camps and areas such as the valleys, subject to the direct influence of the saltpetre industry, the action of elementary schools and the Andean settlers’ own interest to “chileanise” themselves contributed to the Amerindian language’s constant regression.

The reduction of the number of inhabitants due to Spanish advancement within the population of the low areas led to the language’s vital space gradually being limited to the high lands. After the region was annexed to Chile, Aymara language was extensively used only in the high foothills and the altiplano. This reinforces the idea of the high lands as an area of indigenous or “Indians” (according to a neutral or a contemptuous emphasis, respectively), a condition linked to the knowledge and use of Aymara language.^3,5 It should be recalled that, at least since the end of the 19th century, the denomination “Indian” expresses a social and cultural inferior category (“underdeveloped”, “uncivilised”), which ends facilitating the abandonment of the language with which it is associated. The effects of this negative valuation continue to this day, despite the existence of a multicultural indigenous policy that recognises and values ethnic diversity, and despite the emergence in some sectors of Chilean society a favourable attitude towards indigenous cultures and languages.

During the 20th century, the opening of roads and general infrastructure improvement accelerated economic and political integration of those inland areas to the regional centres. Here, where its dominance was not yet generalised, public schools disseminated national identification and the use of Spanish (Castilian), as well as instrumental contents. Military service encouraged the adhesion to national values (and also castilianised together with providing literacy skills) among young male Aymara. The market’s action over the Aymara communities complemented that of the State, leading the peasants to further the commodification and monetarisation of their economies. And, with this, it incremented their dependency on sources of employment and external supplying (the low valleys and Bolivia), which also meant the establishment of fluid relations with commercial agents who were monolingual in Spanish. Since the 1930s, in the Aymara-phone areas sustainability agricultural problems emerge, a phenomenon that precipitates the departure of many individuals form the valleys and the Andean cordillera. Firstly, the progressive dismantlement of the saltpetre companies, with the dismissal of their working population, limited the income options and the product placement to the valley agricultural producers. At the same time, the population increase could not be supported by the high lands’ traditional livestock-based economy. Migrations to Andean and non-Andean desert and coastal areas occurred (Arica and Iquique, Pozo Almonte), a redistribution of the Aymara population in the region and an increase of spatial mobility. In recent years, the demand for education and access to public services has also been a strong stimulus to migrations from traditional indigenous areas.⁵

These processes had an overall effect, favouring cultural assimilation, the linguistic displacement of Aymara language and the increase in Spanish monolingualism. The description model based on the displacement of the Aymara language according to the ecological criterion of isolation (greater or smaller distance from urban centres, accessibility, presence or absence of schools and public services in priority settlements, etc.) proposed by Harmelink⁹ and resumed by Salas¹ is limited. It does not help us to understand the past nor to clarify the present. The geographical and spatial dimensions are not explanatory factors in themselves, unless they are associated with certain configurations of social relations and in different historical situations. Furthermore, this model is a very schematic representation of complex facts and tendencies. According to this model, it would impossible to understand the phenomenon of military conscription of the young Aymara –including those from isolated areas– and its castilianising impact, effective since the first half of the 20th century. Neither could we understand the fact that in areas such as Isluga and Cariquima, where the national educational system had not yet expanded, around 1940 and 1950, communities considered to be conservative in cultural and linguistic matters recruited “private teachers” from the valleys and Bolivia for basic Spanish literacy training, which obviously contributed to its dissemination. At the same time, in the village of Chapiquiña, neighbouring the hydroelectric power plant of the same name, and therefore where there were frequent relations with the personnel who were monolingual in Spanish, until 1986 there was still a number of Aymara elderly speakers whose families were originally from there.

More decisively, this model proposes no hypothesis that could allow to interpret, historically and sociologically, the linguistic consequences of the social, ethnic and cultural division that existed, already towards the end of the 19th century, within the Andean space, between the occidental valleys and the high lands. The ones farmers, the others livestock breeders; the ones did not consider themselves “Indians” (a description they reserve for those from the altiplano), while the others, if accepting it, they did so reluctantly; the ones were already Spanish speakers or increased their contingent, finally completing their Castilianisation, while those from the high lands maintained a much more conservative linguistic attitude. Furthermore, the isolation of some valleys was as severe as the altiplano’s, whose livestock-based economy was very connected to the mining activities of the desert, the low valleys or the ports. Linguistic differences, created in the long term, became quite real and entailed practical social consequences. The links between the shepherds from the high-Andean communities and the low lands, the valleys, the desert mining activities and the coast were, for the last century and a half, mediated by farmers, muleteers, merchants, petty civil servants and authorities from or settled in the valleys and the Andean foothills. The relations that nurtured this mediation (asymmetrical exchanges, peonage, disparities in prestige and authority, etc.), gave birth to a form of non-institutionalised class domination. They constituted a form of social and cultural contention for the ones (which contributes to their cultural and linguistic conservatism), as well as a means for outreaching the exterior for the others (towards economic and political powers, in the mining activities in the desert and the coast).⁵

On the other hand, we must acknowledge that these linguistic changes could not have occurred without the own Aymara speakers’ decision to abandon the teaching and speaking of their language in their homes and communities. On the contrary, the prevalent view among current Aymara leaders and educators places the exclusive or almost exclusive responsibility for the loss of Aymara on external factors. This means to consider themselves as passive victims of the actions carried out on them. But, however inequitable the relation with the Chilean society was (and still is), this does not cancel the Andean indigenous’ reflexivity and capacity for action. Underlying this view an inexact representation of the power phenomenon, which always consists of a relation between persons (individual and collective) and, thus, where one part can annul the other part’s capacity for opposing or resisting only by eliminating the other person as such; that is, ending the relation and, thus, the conditions of possibilities for the exercise of power. This suggests that power (or, better said, the dominant pole in the power relation) gets stronger when it wins the support of the dominated.

In our case, the Aymara people went beyond simply accepting the integration process conducted by the Chilean State and its regional authorities. Throughout the 20th century, they solicited the establishment of public schools in their villages and towns, even creating them with resources of their own.¹⁰ Also, while it is true that the school played a decisive role in the introduction of Spanish and the weakening of Aymara, it is also true that this occurred within the educational space and not outside it, at least not in a constant or systematic way. Therefore, the abandonment of Aymara inside homes cannot be explained as a direct consequence of the school’s deed but, at the outset, as result of the speakers’ decision. There was a wide range of motivations to do so: to adequate to the dominant winds of change, genuine desire for integration, to search a way of leaving the vilifying condition of being an “Indian”, as a defense mechanism before prejudices and discriminatory acts, etc.

We must also mention the active role played by the communities when petitioning the government for the construction of roads and the provision of services (such as electricity) or when migrating to urban centres so that their children could complete a higher and better-quality schooling. Such demands were also formulated from a self-definition as farmers, peasants from the interior or people living in the frontier zones, among others, but not as Aymara people. In this way, not only in its content but also in its form, the Aymara revindications followed a similar integrationist pattern and invisibilized the indigenous condition which, until a decade and a half ago, remained an element of identification for local communities or in the relations between communities, but not within an inter-ethnic public space. In this context of rapid and intense changes, knowledge, tools and communication codes were increasingly those provided by the national culture and Spanish (Castilian) language.

The findings of our survey show that the majority (66.3%) of the inhabitants of rural communes (including the towns of Pica and Putre) belonging to the Arica-Parinacota and Tarapacá regions are Spanish speakers. If we consider only the persons who define themselves as Aymara (181 cases), a 65.7% (119 cases) affirm having some level of competency in their language. If we add the cases of person who do not ascribe themselves as Aymara but have some knowledge of the language (198 cases), the efficient speakers of the language represent a 35.2% of the bilingual group; namely, that just under one fifth (a 17.4%) of the total has some fluency in Aymara.

The distribution of Aymara knowledge and usage within the regional space is not homogeneous, showing important variations between socio-geographical areas. The number of Aymara speakers and the frequency of language usage decrease along an altitudinal gradient. In the altiplano, it is more frequently spoken due to more favourable social and linguistic conditions. On the other extreme are the low valleys, the oases and the pampa, where there are less persons who know the language and where it is very little or occasionally employed. In the Andean foothills area, we find an intermediate situation, for while there are more speakers than in the low areas, its usage is also occasional and rare.

There are also differences regarding the command of the language according to usage contexts. Aymara is usually spoken in private spheres and intra-ethnic spaces, while Spanish prevails in public and inter-ethnic spaces. It is still used inside homes, in relations with relatives or in traditional economic and religious activities, but most of the internal communicative interaction, in the school environment or in the relations with public officials, as well as the majority of the relations with the regional society are realised in Spanish (Castilian). Moreover, this description corresponds only to the highest areas, in the low areas Aymara remaining vital only among persons who arrived from the Chilean or Bolivian altiplano, where the usage of the language is even more restricted. The native inhabitants ceased to speak it many decades, and even many generations, ago.

The advancement of Spanish and the decline of Aymara cannot be explained only by the possibilities and frequencies of social interaction with the speakers of the dominant language. From a historical perspective, they are the result of social relations established between indigenous segments who have experienced very different processes of cultural and social change. The inhabitants of the lower areas, in the 19th century, began a process of assimilation and relative abandonment of their native culture and language, ceasing to consider themselves as “Indians”. Such condition they reserved it for the inhabitants of the higher areas. In this way, the distinction “Peruvian” or “Chilean” as opposed to “Indian” was re-established as social categorisation scheme within the same Aymara world, producing their division in two opposite groups regarding their collective identification. Alongside this, among the historical Aymara themselves a social and ethnic gap opens up from which unequal relations –and domination–, founded on affiliation to one or the other sociocultural category, occur. Hence, the differences according socio-geographical levels are very important to understand the distribution and the changes of the Aymara language in the region. In the lower areas, Aymara was early displaced by Spanish, thus the current presence of this language there is explained by the presence of individuals and families that have emigrated from Aymara-phone areas.^11‒17

At present, linguistic frontiers have been redefined in a contradictory way: although the Aymara have disseminated throughout the region and the pressures to hide and abandon the language have diminished, the dominant social conditions do not favour its preservation and repossession. To many Aymara, their culture and language is something of the past or a tradition to be preserved only in so far as it does not become an obstacle to integration. Aymara language may persist in spheres internal to family and community social life, but it is not functional to the Aymara community’s regionalisation, in the way it has occurred so far.

A fundamental issue, not inquired, is the situation of Aymara language in cities, where the majority of the current indigenous population resides. Given the lack of surveys on this matter, we can only make some conjectures. The first element to take into consideration is migrations. Among the urban Aymara from rural areas and villages we find speakers of the vernacular language. However, social and linguistic spaces to which they have incorporated provide limited opportunities and restrictive conditions for its usage, which is basically reduced to ethnic contexts of interaction or to the domestic sphere. Although emigrants from Aymara-phone areas, including the Bolivian border ones, mostly head down to the low valleys (Azapa and Lluta, in particular) and the coastal cities, they find a resident population that for the most part ignores, is not aware or disregards the indigenous language. Thus, the usage of Aymara by the elderly should tend to decrease or minimise, their children and grandchildren probably being passive speakers and/or monolingual in Spanish (Castilian).^18‒22

At the same time, it is important to consider the relations that migrants from these areas maintain with their (birth) places and, in the Bolivian case, from these locations with Chile, where they find jobs and support from their “paisanos”. Together with definitive emigration, we find an ever-increasing spatial mobility of Aymara population from Bolivia, facilitated by the increasingly expeditious means of transportation and communication in the region. Migrants originally from Aymara-phone areas are exposed to the influence caused by their permanence and their contact with a population that mainly speaks only Spanish. But there are also the temporary returns to their communities or birthplaces, frequently boasting material achievements and a more diverse and prestigious cultural capital. This is exactly why it is of great importance to incorporate a subjective element into the analysis of linguistic displacement: the Bolivian Aymara emigrants in Chile are appreciated as persons comparatively more successful and the region to which they arrive is seen as part of a country that offers more possibilities, is more developed, with more progress and “civilisation”. Spanish (Castilian) forms part of this state of things, qualitatively different and better, unlike Aymara and its speakers becoming minorities in the region.

None.

None.

Author declares that there is no conflict of interest.

1. Salas Adalberto. Indigenous languages ​​of Chile. Ethnography. Contemporary indigenous societies and their ideology, Santiago: Editorial Andrés Bello, 257-295.
2. González Sergio. Chileanizing Tunupa. The Public School in the Tarapacá Andino 1880-1990. Santiago: DIBAM; 2002.
3. Gundermann Hans. How many speak the Aymara language in Chile? Revista de Lingüística Teórica y Aplicada. 1994;32:125‒139.
4. Gundermann Hans. About how Aymara learned Spanish (ending by forgetting Aymara). Estudios Atacameños. 1997;12:97‒113.
5. Gundermann Hans. Andean community, organization and social complexity in northern Chile. Doctoral Thesis, Center for Sociological Studies, El Colegio de México: Mexico; 2001.
6. Gundermann Hans, Rolf Foerster, Jorge Iván Vergara. Count indigenous people in Chile. Ethnic self-enrollment in the 1992 and 2002 census experience. Estudios Atacameños. 2005;31:91‒115.
7. INE. Censo 2002. Resultados Población y Vivienda. Santiago: INE. 2002.
8. Vergara Jorge Iván, Luis Godoy. Interculturality, education and indigenous languages. Global challenges and regional experiences. Culture, education and peace in the framework of subregional integration. Iquique: Ediciones Campus, Universidad Arturo Prat; 2003. p. 71‒100.       
9. Harmelink Brian. Linguistic investigations in the Chilean highlands. A study on the Chilean Aymara community. Arica: Instituto Lingüístico de Verano, Universidad de Tarapacá; 1985.
10. González Héctor. Characteristics of the city-field migration between the Aymaras of northern Chile. Arica: Andean Studies Workshop, Working Documents Series; 1996.
11. Adelaar Willem. The problem of languages ​​in danger. South America. In: R Hobins, E Uhlenbeck, B Garza, editors. Languages ​​in Danger. México: INAH; 2000.
12. Albó Xavier. Aymaras between Bolivia, Peru and Chile. Estudios Atacameños. 2000;19:43‒73.
13. Reyes R, J Rafael. Chilean aboriginal groups, their current situation and territorial distribution. Santiago: Universidad de Santiago de Chile; 1989.
14. Fasold Ralph. The sociolinguistics of society. Introduction to Sociolinguistics. Madrid: Visor Libros. 1996.
15. Ferguson Charles. Diglossia. Word XV. 1959;325‒340.
16. Fishman Joshua. Sociology of language. Madrid: Cátedra; 1988.
17. González Héctor. The Economic Insertion of Aymara migrants in the City. Work as Family Business and Cultural Reproduction. Proceedings of the Second Chilean Congress of Anthropology: Santiago; 1997. p. 567‒579.
18. Holmes Janet. An Introduction to Sociolinguistics. England: Longman Group. 1992.
19. Hymes Dell. Communicative Competence. In: U Ammon, N Dittmar, K Mathier, et al. Sociolinguistics. Berlin: De Gruyter; 1988. p. 219‒229.
20. Rommaine Suzanne. The language in society. Barcelona. Ariel Lingüística; 1996.
21. Vergara Jorge. Description of the sociolinguistic context in indigenous communities of Chile. Santiago: MINEDUC-Programa EIB/Orígenes; 2005.