SciELO - Scientific Electronic Library Online

vol.4 edição especial índice de autoresíndice de assuntospesquisa de artigos
Home Pagelista alfabética de periódicos  


versão impressa ISSN 0104-9313

Mana v.4 Rio de Janeiro  2008


The war of the alphabets: indigenous peoples between the oral and the written


A guerra dos alfabetos: os povos indígenas na fronteira entre o oral e o escrito



Bruna Franchetto

Professor at the Programa de Pós-Graduação em Antropologia Social (PPGAS/ MN/ UFRJ). E-mail:

Translated by Luiz Costa
Translation from Mana, Rio de Janeiro, v.14, n.1, p.31-59, Apr. 2008.




Based on ethnographic fragments from linguistic research among the Taurepáng, Macuxi, Wapichana and Kuikuro, conducted in distinct times, regions and situations, this article analyzes the conflict occurring between orality and writing when the orthographization of an indigenous language transforms and crystallizes sounds and speech on sheets of paper. This is an open arena where different representations and agents of writing emerge, interact and clash: missionaries, researchers, agents of the state, indigenous teachers, indigenous preachers, the indigenous community itself and so on. Here I approach writing more as a metaphor or emblem than a simple technology of correspondences between codes. The article provides, then, an interpretation of the meaning of writing that may help us to understand, among other things, the reasons behind the successes and failures of bilingual education, literacy projects and the introduction of schooling. The Indians also observe, often with considerable perplexity, the wars or dances of the letters, where orthographization consecrates whatever it includes and condemns what it excludes – the vital organs of a language.

Key words: Ethnology, Writing, Indigenous Education, Linguistic Policies, Indigenous Languages


A partir de fragmentos etnográficos de experiências de pesquisa lingüística entre os Taurepáng, Macuxi, Wapichana e Kuikuro, em tempos, regiões e situações distintas, este artigo trata do confronto entre oralidade e escrita, quando a "ortografização" de uma língua indígena, falada por uma sociedade de tradição oral, transforma e cristaliza sons e ditos em folhas de papel. Em uma arena aberta, representações e agentes da escrita surgem, interagem, chocam-se: missionários, pesquisadores, homens do Estado, professores indígenas, pastores indígenas, índios etc. A escrita, neste contexto, é então abordada mais como metáfora ou emblema do que uma simples tecnologia de correspondências entre códigos. Trata-se, então, de uma interpretação do sentido da escrita que pode ajudar a entender, entre outras coisas, as razões de acertos e fracassos da "educação bilíngüe", do letramento e da escolarização. Os índios ainda observam, não poucas vezes perplexos, as guerras ou as danças das letras, enquanto a "ortografização" consagra o que ela permite e condena o que ela exclui, órgãos vitais de uma língua.

Palavras-chave: Etnologia, Escrita, Educação Indígena, Políticas Lingüísticas, Línguas Indígenas



Among the most significant experiences in the history of the encounter of Indigenous populations with colonizers is the discovery, the diffusion, the acquisition and the impact of writing, along with its inevitable corollaries: literacy, alphabetization and schooling. In the hands of "civilizing" agents, these are simultaneously delicate and powerful instruments which enable experiences that effect significant changes in Indigenous societies. There has been little reflection on this theme in the context of the history of the Indigenous peoples of Brazil, and we have heard little of what the Indians themselves have to say about it.

This essay approaches the confrontation between orality and literacy at the moment in which writing is introduced into a society that is based on oral tradition, and it investigates the system of concomitant and contradictory representations of writing that emerge, interact and confront themselves through the various actors on the scene. In what follows I take writing to be a metaphor or an emblem, directing writing beyond its apparently immediate nature as a technique for the transformation of codes. I thus analyze the representations that various actors in relation to one another elaborate on the meaning of writing, and how these representations concern the image that each one makes of the meaning(s) and the underlying political dynamics of orthographic metaphors. I thus aim to develop an interpretation of the "meaning of writing" that, among other things, can help to explain the reasons behind the relative failure of the projects of so-called "bilingual schooling".

I seek to investigate aspects of the transformation of an oral language into a written one through the vantage point of a privileged position for observing the intersection and the clash of ideologies and practices that constitute the field of so-called "Indigenous education" and its articulations within a wider arena. This position is that of someone occupying the métier of a linguist who studies Amerindian languages and who creates and manipulates alphabets and writing norms, ranging from the transcription of field notes to the resources for literacy education. I will not, therefore, detain myself on the rituals of literacy education nor on the varied and complex consequences of the introduction of writing through schooling in Amerindian societies, but rather on certain connotations of writing, thereby distinguishing functional aspects from the range of non-functional and strongly ideological connotations that accompany it.

There is, at present, a diffuse consensus concerning the healing power of educational programmes in Indigenous languages that has emerged from the disasters of an inefficient practice of guaranteeing "universal" rights to Indigenous populations which condemns Indigenous languages to extinction. The actors in this process, however, do not agree on the nature of these disasters nor on the results that they expect from the cure. The different orthographies proposed and the varied representations on the meaning of letters or graphemes become mirrors for the ideologies in conflict.

The history of any orthographical system is characterized by changes and adaptations. Any new writing system is constituted and reformulated through factors that are not only "technical" or "scientific", but also political, active or reactive. A veritable war of the alphabets is taking place in Brazil – a war that has been going on for some time, but that has grown increasingly tense and violent, and whose combatants are small armies of missionaries, members of government and non-government organizations, linguists and advisors. In the throes of this war, the Indians sometimes ally themselves with one group or the other, moving forward or retreating, negotiating.


Among the Schooled Indians of Roraima

In the end of the 1980's I was in Roraima, a state in the far north of Brazil, carrying out linguistic research in the villages of the Taurepang, a Carib-speaking ethnic group. The rare presence of a linguistic in the region drew the attention of the Indians and the government and non-government agencies (State Secretary of Education, Precinct of the Ministry of Education, Diocese, Indigenist entities). The model of "bilingual education" in Indigenous schools had just reached Roraima, and it was taken to be a sign of modernity, of progress, of the willingness of government and missionary institutions to accept new principles in the management of the Ministry of Education.

In my capacity as a linguist, I was invited to advise on educational programmes and projects for the Indigenous population of Roraima. In fact, it was more than an invitation: the linguist was constantly called upon to apply his specialized knowledge so as to evaluate proposals, to solve problems, to hold courses, and to elaborate literacy education resources. During three years, I thus visited not only Taurepang villages, but also the malocas1 of the Macuxi, also a Carib-speaking group, and the Wapishana, an Arawak-speaking group situated in the region between the savannah and the western hills of Roraima, to talk about the school, orthographies, and literacy in the Indigenous language2. A school existed at all of the malocas, and it was in the school that most meetings were held. Any consideration of bilingual education necessarily stemmed out of a discussion concerning a "new" school, in which the oral and written use of the Indigenous languages and literacy education in the Indigenous language would be present, along with the use of the official language, Portuguese.

In the midst of missionaries, government agents and advisors, what was the stance of the Indians? What representations do the Macuxi and Wapishana hold on school education, being, as they were, in the threshold of an abrupt passage from stigmatization and the crisis of their native tongues before an avalanche of proposals on how to use these languages in projects of bilingual education? A vacuum which nurtured expectations and contradictory discourses was formed. Words such as "acculturation" and linguistic and cultural "salvage" were constantly employed in Indigenous speech.

I came across a wide spectrum of positions. At one extreme was an explicit aversion to all type of intervention in formal schooling, as the conservative radicalism of an old Taurepang chief at the Maloca de Sorocaima makes clear:

I do not allow schools in my maloca. I can educate my children myself [...] many people have gone to start schools there, but I have always said that that is not the education that matters. I can teach how to plant, to sell products and to buy good things... I see that the children who go to school are becoming rude. They only want to play ball and hit others. (speech of chief Mário Flores Macário, 1986, Diocese of Roraima, Boletim n.11, no pages).

In Sorocaima, where all of the community is faithful to the Seventh-Day Adventist Church, there were no schools; the Taurepang language was alive, even in the almost obsessive and daily Adventist masses.

At the other extreme, a small group of Macuxi and Wapishana had been searching, for some years, for a way to revitalize their languages, well aware that these were undergoing a process of annihilation; their introduction in schools, strategically capitalizing on the vacuum created by the "fashion" of bilingual education, could have a miraculously salvaging function. In the Macuxi and, especially, Wapishana malocas that I visited in 1988, I was faced with a scene that is typical of linguistic loss and generational rupture. The eldest people, many of which were monolingual, made full use of the Indigenous language; their bilingual children communicated with their parents in the maternal language and with their children in Portuguese; the latter, although they could understand their grandparents, made exclusive use of Portuguese. Among members of the intermediate generation, I regularly heard phrases such as "I understand everything, but to speak, the language is hard, it gets rolled up, I can't say my thoughts", or "I only talk gibberish when they [their sons] do not understand what I am ordering, when I get mad".3

Ever since the first schools were founded by the Indian Protection Service in the 1920's and 1930's, as well as in the Catholic boarding schools and throughout the diffusion of state and municipal schools in the interior regions of Roraima from the 1950's onwards, schooling was the principal vehicle for linguistic and cultural repression. I was able to observe, in the obsessive disciplinatory rituals and in the folkloric pantomimes of official commemorations, that the access of Amerindians to school education only produced semi-literate people destined to occupy the tasks of a submissive and exploited labour force.

The Macuxi and Wapishana had elaborated an articulate discourse on the values of formal education and writing as desirable technologies for overcoming a chronically inferior position. At the same time, they spoke of the values of ethnic identity, emblematized in alterity and in linguistic diversity – a discourse which may have been born under the wings of Catholic missionaries, but which was beginning to follow an autonomous path. This discourse was most forcefully articulated by a group of political leaders, of which the younger, egressed from the education centres of the Diocese and relying on the experience and wisdom of some elder leaders, sought to experience methods and solutions that were independent of the official and missionary orthodoxies.

Between the poles represented in the discourse of the Taurepang chief from Sorocaima and the proposal of the young Wapishana leaders – both of which, in different ways, valued alterity – there were a majority of non-believers, among whom were those who had internalized stigmas (and fears), or who no longer had any belief that change could make the school produce truly literate people, or at least not those who would be literate in the official language. There were also those who maintained the "caboclo" and "civilized"4 rigidly separate, not accepting any proposal that sought to introduce Indigenous languages into the school.

My work with the Wapishana of Malacaxeta, in Roraima, was my initiation as a "linguistic advisor". It is this experience that I will now focus on, since it seems to me to encapsulate the theme of this article: the clash of alphabets.

An orthographic arena: ideologies in conflict among the Wapishana

At the start of 1987, a small group of Wapishana5 teachers from the school of the Maloca de Malacaxeta village, a few kilometres from Boa Vista, the capital of the state of Roraima, sought me out with an almost dramatic appeal for me to act as advisor to a project of revising the Wapishana orthographies and elaborating material for literacy education or for teaching the language in school. The teachers felt pressured from every angle: from their knowledge of the agonizing process which their maternal tongue was undergoing; from the incentive of linguistic rescue formulated by the Catholic missionaries and the recently created Nucleus of Indigenous Education of the Government Secretary of Education of what was then the Territory of Roraima; from the general distrust and criticism of individual residents of the maloca and by sectors of the government as to the necessity and viability of the project.

This small group of Wapishana, who had already begun to discuss the conflicting orthographies for their language, expressed their doubts and a considerable dilemma. Once the dialectical variations of the Wapishana language were homogenized, each one of the orthographies appeared to be a definitive object, a uniform written code. The Indians wanted to choose – or, better yet, to develop – a new script, that would mark their distance from the Protestant and Catholic missionaries, who were the proponents of the existing orthographies and who situated themselves on opposing sides. This new orthography would then be presented to the State as an emblem of the unity of the Brazilian Wapishana, neutralizing the religious divisions. Once made official, it would effect an utopian act of rescue and salvage, giving rise to the teaching of Wapishana as a second language in the schools of the malocas where it was no longer used by the schooled generations and those undergoing formal education.


The Writing of the Evangelists

Wapishana began its existence as a written language in the 1950's, at the headquarters of the Unevangelized Field Mission of Guyana (UFM). The writing system developed by the Evangelists then crossed the border into the Wapishana communities of Brazil by missionaries and Indians in their travels or in journeys as agents of religious proselytism. The "Evangelical" script then began to circulate not only through those segments directly tied to the Mission, but also among the few Indians who had been converted to the Catholic faith and who had become literate in the maternal tongue during their trips to the Evangelists of Guyana.

When I arrived in Roraima, it was these individuals who had been literate for longer, and who had greater knowledge of writing, and they transmitted the use of writing, in its "Evangelist" version, to others; it was also they who regularly visited the posts of the Evangelical Mission and who attended meetings with the missionaries, seeking to perfect their knowledge of writing and reading and to have excess to educational material in the Wapishana language. This material, mostly produced in Guyana, included teaching texts, manuals, history books, hymn books and Evangelical texts. There was nothing in Brazil to compare to the quantity and quality of the written production of "Evangelical" origin. The orthographical system developed by the UFM had been established based on criteria and processes that are widespread in all fundamentalist Evangelical missions, pioneers in the orthographical transformation of oral traditions and in literacy education among Indigenous populations. A similar pattern emerges in all of the work done by these missions, including their fieldwork style, linguistic investigation, the establishment of scripts, their philosophy and techniques for literacy education and in the printed material for schools and churches.

Polyglotism and the power of inter-linguistic translation are important characteristics of the Pentecostalist Evangelical vision. Alongside bilingual education and the reduction of oral languages to written ones, they constitute the same apparatus for integrating and assimilating the Indigenous populations, crystallizing the "civilizing" vocation of Western writing and, paradoxically, adding legitimacy to Indigenous languages. In those places with an Evangelical presence, they have "preserved", in their fashion, the use of Indigenous languages and are, as the Macuxi of the Napoleão maloca told me, "the only civilized people who are not ashamed to speak gibberish".6 To what do we owe this peculiarity of the Evangelical missions, which make the "scientific" study of Indigenous languages, and their transformation into written ones that can later be used in educational programmes, into a fundamental aspect of their job of converting and of their "civilizing" activity? We can highlight two definitive elements in this missionary activity: the nature of the evangelizing task and its vocation for civilizing.

To Evangelize is, literally, to spread the Gospel to all the peoples of the world; to make the "word of God" accessible to all men, regardless of their culture, their social system or language, thereby reifying it for all eternity as the universal truth in Christian holy scripture. In the Pentecostal view, the ability to dominate the equivalences between words and expressions in different languages is the most important gift of the Christian who is illuminated by the Holy Spirit. After all, if Jesus Christ preached that all of the people of the world should open themselves to the light of the Gospel, the pagan Amerindians cannot remain "innocent" and must have the opportunity to know the Holy Scriptures. Leaflets that circulate among the missionaries often contain the phrase, "If God is interested in me, why does he not speak my language?". In order to make the Gospel accessible to all peoples, it is thus necessary to know the language of everyone, to write unwritten languages, to translate the Holy Texts, to educate the individuals who are undergoing conversion and those who have already been converted so that they may read and reproduce the word of God. These are the necessary stages of Evangelical missionary work.

To this must be added the civilizing task. All of the missions explicitly define themselves as "agents of change" – a profound change imposed not only by the values transmitted to a few through Evangelical texts, but also by the mere presence of the missionaries in the community that is being converted. The "American way of life" is inevitably proposed as a model for living: manners, aesthetics, hygiene, nuclear family, white race, technology and so forth. This message could be glimpsed in the rhetorical claims contained in the leaflets that were distributed in the SIL courses administered in Brazil, which were significantly called "Linguistics and Missionology" courses: 

 [...] we must consider it uncertain, yet ever more clear, that either North American civilization allies its strength with other civilizations – including that of the less privileged peoples – or no civilization will last for very long (Kahn 1986, n/p).

To hear the Indians say "He speaks our tongue, he is one of ours" is the crowning glory of modern missionary work. As one of the masterminds behind SIL ideology observes, the missionary should be able to converse about the most intimate details of the faith of his flock in order to convince them to accept principles that conflict with their history and culture.

Learning a language is more than the simple mechanical capacity to reproduce acoustic signals as if one were to try and sell merchandise or to find a way out. It is a process through which we make vital contacts with a new community, a new way of life and a new thought system. To achieve this to the best of our ability is a basic requisite of missionary work  (Nida 1957:8).

The conjunction between linguistics and the mission is thus established and the writing Indigenous languages becomes the focus of the whole system. Arriving at a scientifically correct orthography presupposes a long period of field research, involving a description and analysis of the language in all its levels so that it can become an instrument for Evangelization. The new script prints the word of God and becomes the necessary starting point for literacy education programmes, which are always established in conformity to the philosophy of a particular government. This is the double conquest of civilization.

Years of research are carried out by missionaries in the search for a definitive orthography. The "adequate" script is always psychophonemic or simply phonemic: the symbols and the graphemes always correspond to the phonological units (phonemes). A second criteria concerns the necessary adaptation of writing to the national language; the choice of graphemes is thus limited by those that already exist in the alphabet of the national language. In other words, the Indigenous language is an instrument for achieving a type of literacy that acts as a baseline for another objective: learning the dominant language, taught in stages, first in its oral form and then in writing.7

A knowledge of phonetics and phonology – acquired in linguistics and missionology courses – is a minimum prerequisite for this task. Furthermore, the elaboration of an orthography must take into account a number of factors that are labelled sociolinguistic: dialects, religious divisions, politics, pre-existing orthographies, age groups. Finally, the success of an orthography depends on how it is accepted by some in the group, who should come to use it with little difficulty until it becomes a part of their culture, as defined by the SIL. A script is, therefore, a pure conversion – adequate, if not perfect – of the actual phonological units of the spoken language. At this stage the task appears to be technical, so to speak, a mere conversion of codes based on scientific criteria. A whole process which involves translators and translation consultants then baptizes the new script (Stoll 1982; Barros 1993). The alliance of linguists and translators puts writing at the service of another task, making it a strongly ideological, yet subtle, instrument for cultural and social change.

The Evangelists had given the Wapishana an orthography that had already been tried and tested, but in which the latter saw a range of problems. In the course of seminaries which I had been called to organize and conduct, the Evangelical orthography was discussed and, gradually, its architecture and background came to be understood. In spite of its efficient and scientific veneer, certain insurmountable hindrances stood before its adoption: its foreign accent and its identification with a missionary segment that was inimical to the Catholics who worked with a large part of the Brazilian Wapishana.

The first problem stemmed from the fact that the orthography had been modelled in the English language, which is the official language of Guyana, an ex-British colony. There was concern to avoid "nationalist" criticisms, which could come from the educational agencies of the local government (Roraima) 8; the script should therefore become more akin to Portuguese, so as to strengthen the proposal under consideration. The second problem paradoxically stemmed from the most scientifically sound and convincing aspect of the Evangelical script: its phonemic "logic", through which each grapheme or letter represents a phoneme, a distinctive unit of the structure that organized the acoustic matter of the language.

Phonetic writing is a historical conquest of modern linguistics and it is geared towards the introduction of writing into an exclusively oral tradition, since it is based on the application of phonological knowledge towards the establishment of an alphabet and the other orthographical norms of a language. In other words, a phonetic script is thought of as natural, since it bases itself on the internal, unconscious linguistic knowledge of the speaker – a knowledge that is not only phonological, but integrally grammatical. The process that produces a phonemic writing nonetheless implies a considerable degree of abstraction and presupposes the inevitable intervention of a linguist. Once a phonemic script is consolidated, its success in promoting literacy is seen to be a consequence of its "naturalness", since it is accepted by the literate, native speakers of the language.

The apparently inexplicable rejection of purely phonemic script by the Wapishana (and by many other Indigenous peoples) can be seen as the expression of a tension between two "natures": on the one hand, the already mentioned association between graphemes and phonemes; on the other, the orthographical conventions of Portuguese which are perceived to be "natural" since it is a prestigious language for which writing is an integral part of its strength. We can thus explain the desire and the need to adapt the written norms of the Indigenous language, invented elsewhere and given to the Indigenous people, to the written norms of the language of the whites. The Wapishana, nonetheless, had not only to deal with the script of the Evangelists, but also with that of the Catholic missionaries.9


The Writing of the Catholics

Between the experiences registered in the very first encounter with Indigenous languages in the early days of the colony and the recent philosophy of bilingual education, the policies and practices of Catholic missions were characterized by a long period dominated by the annihilation of native linguistic diversity. In the context of a revision of missionary work carried out by the Catholic Church within the last years, the Catholics began to be concerned with respecting the implementation of a new perspective of bilingual education. Within the Diocese of Roraima (Mission of the Consolata), this was sometimes expressed in the words of the priests, other times in the speeches of the Indians themselves:

 [...] learning Portuguese is a good thing, in order to understand the whites and to not be fooled by them; but there is no way that we can forget out Macuxi language. We must defend what is ours and value all that our parents taught us. Only then can we better our lives and defend our rights. The Macuxi language is, for us, a weapon that we can use to better communicate between ourselves and which, furthermore, the whites do not understand [...] (Roraima Indígena, 83:15).

The book Waparadan, presented as a sort of guide for learning the Wapishana language, had been the first essay published by the "Catholic" press, and it had been elaborated with help of laymen advisors to the mission with superficial knowledge of anthropology and linguistics. The essay was not so much the result of a linguistic study for alphabetization in the Indigenous language, but rather an object whose symbolic effectiveness was meant to lie in the incentive for rescue – or, better yet – the consolidation of a change in missionary work. No sooner was the book printed and published, problems with the new script began to be manifested.

There are a series of linguistic mistakes in the script that was created. It is basically a mixture of imprecise phonetic registers, extreme adaptation to written Portuguese, and mistakes due to a hurried and superficial study of the structure of the language. Had the Wapishana not been exposed to the "Evangelical" script, they would no doubt have failed to notice other, even greater problems with the "Catholic" script. The Indians were well-aware of these problems, and could criticize them one by one. It was clear that the Catholics were out of step with the Evangelists in what concerned knowledge of modern linguistic techniques: "Catholic" script and grammar resulted from a process of simplification or transfiguration of the Indigenous tongue, through a scheme that was half-way between prescriptive grammar and classic scholastic categories – the same schemes that, centuries before, had guided Jesuit linguistics.

Missionaries began to produce written material in Macuxi and, to a lesser extent, in Wapishana. Along with the inevitable hymn books and translation of texts aimed at conversion, a considerable amount of work was dedicated to the editing of books of Indigenous "stories", accompanied by commentaries on the value and meaning attributed to this type of preservation and publication of oral memory:

We have tried to write differently from the writing contained in the Gospels [...] When the whites arrived in Indigenous lands they said that the stories of the Indians were lies, worthless foolishness. The Indians thus came to believe the words of the whites [...] This is why (the stories) vanished [...] The stories of the Indians came to be called "myths" [...] (Anna Maimu, Waparadan, 1981a, n/p).

What are, in fact, the myths and the stories of the Indians? [...] They describe the life of the ancient people [...] The myths are a weapon that the Indians use to defend themselves against the whites who want them to disappear or to become civilized [...] (Anna Maimu, Waparadan, 1981b, s/p).

In the Macuxi tales, the Jaguar represents danger, the threat of violence, the strongest: the "civilized oppressor" who wants to eat the Tortoise-Indian, with his lands, tradition, language, everything. The Jaguars represent the Violent Ones against whom the Tortoise-Indians must fight, with cunning and wisdom [...] The Jaguars of the New Dictatorship [...] The Tortoise-Indians only have one strategy to defeat the powerful Jaguars: to unite and to throw at them the stones of the new laws that have recently been approved in the Constitution [...] (Igreja a Caminho,1988:5).

The Wapishana pondered over the "Indigenous literature" books, some of which were produced by the Catholics, others by the Evangelists, and others still through the initiative of the State. The existence of texts written in Indigenous languages would attest to the originality of this "literature", most of it being in Macuxi. The Macuxi, too, had to deal with a Catholic and an Evangelical script. In spite of a discourse that proclaimed self-determination and proposed "education for freedom", contrasting itself from the integrationalist discourse, the Catholic missions' organization of books of Indigenous stories was equivalent to the output of the Evangelists. They were both, in the end, the same type of "literature". Both appropriated a knowledge (language, narratives) which they then drastically re-elaborated before returning it, devoid of its character, to those who originally produced it, now with a reinterpretation that imposed upon it an incontestable authority. Stoll's (1982: 256) criticism of the output of the Summer Institute of Linguistics can be extended to all of this "Indigenous literature": an abyss separates the sophistication of the Indigenous intellectual systems from the poverty characteristic of most of the material aimed at those who could read.

What Kahn says applies, in the final analysis, to all of it:

Before the authority that writing assumes for the Indian, this language, constructed and adapted, can come to be a new language, the language of new times. This is the place of alphabetization in the maternal tongue in the work of the missionary, which serves to give legitimacy to that procedure. What is taught in the texts extrapolates the conversion of sounds into symbols (writing) and creates texts that, once "adapted" to the symbolic universe of the group, actually create a new language, a new formula through which they can experience and live their lives [..] The "language spoken in schools" thus creates a new category, a new pattern of communication. Only the professionals of the language of God, the agents of civilization and "enlightenment", can create this new pattern. The ideology of a Western, Christian society in search of salvation is transmitted through the person of the missionary-teacher (Kahn 1986, n/p).

Writing was this new "language", a simultaneously religious, social and political means for conversion, diffused by the West and imposed upon others so as to fulfil, at whatever cost, its civilizing mission, in the process levelling and limiting the expression of forms of orality.

This encounter between oral cultures undergoing disaggregation and the universe of writing can be studied through the concepts of "restricted code" and "elaborate code" proposed by sociolinguists. What we have here is a type of inversion of the contexts found among the marginal segments of large cities: for Indigenous societies with the institutionalization of a restricted code of a monitored written expression what is lost is the elaborate code of verbal arts and oral tradition. In this passage from orality to writing there is a clear contrast in the ways in which, on the one hand, Indigenous stories and traditional narratives are treated and, on the other, how the stories of Christian texts are treated. The first undergo a process of reduction that results in impoverishment; the second, on the contrary, are the objects of a faithful translation, with all of the care of exegesis and the transposition of synthatic and semantic equivalencies. The result contradicts and demystifies the rhetoric of the chorus "writing in the service of salvage". Literate Indians are quick to compare their myths, condensed and trivialized – a folklore composed of small fictions – and the grand myths of the whites, consecrated in true books. In sum, the first result in the restricted code of so-called Indigenous literature and myth, with common-sense negative connotations; the second result in the elaborate code of a "true story-history", carefully distinct from the literary and mythical genres.

Through bilingual education, the Wapishana were beginning to skim the surface of Evangelist and Catholic rhetoric and, at the same time and along with the other actors on the scene, were appropriating the ambivalent double-speak of civilization and salvage.


The Wapishana want to Write

In spite of all their explicitly ideological force, neither the script of the Evangelist missionaries nor that of the Catholic missionaries left the Wapishana satisfied. Let us re-approach the matter through the written diary of a Wapishana teacher, who was then the principal of the school of Malacaxeta. His recollections begin with his time in a Catholic boarding school:

That was when I began to feel the first difficulties with writing, because we had a book called Wapishana Primer [produced by the Evangelists of Guyana] in which everything was written, but, to me, it was all wrong and I could not understand anything, and now how was I to go about teaching since I had to teach Wapishana [...] That is when I had the idea to write as I hear the sounds of the words [...] even though I did not know why, nor how to go about it, I continued to write as I heard the sounds and I set the book aside and went on writing what I thought was better [...] Every time I wrote, new doubts emerged [...] With no way out, I was forced to use the book which, for me, was all wrong, and on the other hand it was worth it because the book was ready to teach and learn a language. That was when we began to mount the skeleton of written Wapishana. We changed many things in our writing until we made a book called Wapishana Pardan, or Wapardan, meaning "our words" or "our speech". From then on we began to think that the writing in our language was not at all correct and we began to research what the correct writing might be [...]

This same teacher, commenting on these recollections, told me something that I heard from many other Indians: "I always had the impressions that our languages are hard, maybe even impossible to be written properly". The ups and downs of their passage to one or another orthography and the oscillations in the search for what was "right" through religious and national disputes had left them with a strong feeling of inferiority which was re-enforced by the representation of Portuguese as a language naturally gifted with writing.

I quickly realized that the Indians considered the problem of "orthographic fidelity" – the functional adequacy of orthography seen to be a phonemic transcription of a language – as another mark of the "Evangelical script of Guyana", rather than a quality that was independent of context. Cardon reports a similar situation in the history of the "alphabetization" of the languages of Africa, revealing how the introduction of writing accompanies the entrance of autochthonous peoples into the colonial world:

More than any pedagogic or technical consideration on the utility of the various systems, it is the political and religious affiliations instituted by each system that counts [...] Luganda is the first language for millions of speakers in Uganda and the second language for at least a further 3 million; some time in the second half of the 19th century the first contact with Arabs from the coast, who spoke Arabic or Swahili, led to the Islamization of the country and to the adoption of Arab or Arabic-Swahili writing, along with religion. Yet in 1879 came Evangelization and the simultaneous but independent introduction of two similar (albeit distinct) orthographical systems based on Latin, one by English protestant missionaries and the other by and French Catholics. The two systems, which corresponded to the two different religious and political affiliations, were superimposed onto the conflictuous situation of the country, divided into those who were loyal to the king and favourable to the Catholic missions and the anti-royalists, who had been educated by Protestants. The choice of script thus immediately made explicit the type of education and political positions of those who wrote. There was thus a need to find a system of compromise that, in unifying both systems, guaranteed the anonymity of those who wrote. Two meetings (in 1944 and 1947) allowed a unified script to be chosen, but the resistances remained for a long time, sometimes leading to serious conflict [...] (Cardona 1981:125, my translation).

In administering (among other things) reference to the Evangelists, the influence of Catholic missionaries and the pressures of the state, the Wapishana of Malacaxeta wanted, in fact, a new script that would distance itself, inasmuch as possible, from the existing proposals, leading both to an experiment of critical confrontation and to a sort of bricolage of graphemes. In a series of meetings led by the so-called "masters of the language" – some of the last speakers of Wapishana – the different scripts were analyzed, leading to a process of "discovering" the structures of the language. I was present for various orthographical rehearsals before they decided upon an orthography which, even if it could not be definitive, was at least the result of a collective discussion, and which could generate the official teaching texts promised to the authorities.

 The process of this discussion was the most interesting aspect of the "orthographical creation" of the Wapishana, regardless of its consequences (success or failure) in the service of a process of linguistic salvage. An example can better illustrate the conflict and the development of the orthographies, as well as the whirlwind of letters and alphabets into which the Wapishana were drawn. The table below compares the writing of some Wapishana words in the different orthographies, which will be commented shortly: that of the Evangelists, the Catholics, the Wapishana in their first autonomous attempt (Wap) and, finally, the solutions proposed in the official texts.






Official Texts





tikier, tikiez





sur, sir, syz, syyz





kibiero, kibieru





kuarara, koarara, kuazaz







We can see how the spelling in Wap, initially created autonomously by the Wapishana, differs from the others. First, it reveals a radical "Portuguesization", reflected in certain omissions and choices. If certain oral variants enjoy a greater prestige than others among those who speak them (and for those who do not), the same can be said of the written variants. As Cardona says:

The programming of orthographies for oral languages is faced with evident value judgements. In ex-colonies, the orthographies (and not just the languages) of the colonizers enjoy great influence [...] It is obvious that this homage to the prestige of the colonizers has certain disadvantages. Where more than one language is spoken in countries of different influences [...] a purely external division is introduced, which inhibits the unification of published matter [...] Whosoever learns to read and write cannot make use of an orthography that reflects his language in functional terms, but must account for a system that was developed elsewhere [...] (Cardona 1981:122, my translation).

The "foreign" graphemes |k|10 and <sh>, present in the spelling of Wapishana as spoken in Guyana, were eliminated and respectively substituted by |c| or |qu| and |ch|, thus incorporating the problems of Portuguese spelling (|c| and |qu| for the same sound/phoneme as in quina and cobra).11 The palatal fricative, which in the Guyanese side is represented by |sh|, could have been written with the letter |x|, but its rarity in written Portuguese was interpreted as a foreign borrowing, which made it a bad alternative. On the other hand, the digraph |ch| emerged as a good Brazilian correspondent to the "English" digraph |sh| by analogy with its visible form. The representation of the retroflex fricative, a sound that is peculiar to Wapishana, was the cause of a specific concern: the letter |z| was avoided, because it too was considered "weird", rare or marginal in the writing of the national language; in its place the digraph |rr| was suggested. Again, there was a concerted effort to remain faithful to phonetic intuition, since |rr| spells a fricative in Portuguese (velar or glottal, depending on the dialect), but at the expense of phonetic precision (the palato-alveolar place of articulation and the fricative retroflex in Wapishana).

The strong reference to the Portuguese alphabet thus made a closed central Wapishana vowel disappear completely; a single grapheme – |i| – was used to express two sounds with distinct values, since, in Wapishana, there is also another vowel, [i], which is closed but frontal. The inexistence of a sound in the national language thereby condemned a structural element of the Indigenous language.12 We can also note, in this orthography, a significant oscillation between a perception of the phonemic system and a sensibility for purely phonetic variations whose orthographic realization is strongly determined by a orientation towards the writing of Portuguese. In this way, the alternation between [u] and [o], the phonetical manifestations of a single phoneme (kibiaru, kibiaro), and the result of processes of paltalization, like that of the consonant following the vowel [i], with a concomitant change from [a] to [e]: kibaro, in the consistently phonemic spelling, becomes kibieru or kibiaru or kibiaro ([b] becomes [by] and [a] become [e], after the vowel [i]) .

Throughout the discussion on the different orthographies with the Wapishana of the school of Malacaxeta, the Indians evaluated the phonemic accuracy of the Evangelical spelling, as well as the impoverished simplicity of the Catholic spelling and the problems inherited from the latter in the first scripts autonomously produced by the Wapishana. The process was drawn out and punctuated by delicate analyses until the "new" spelling was arrived at. The insecurity before the need to accept a degree of distance from the parameters of Portuguese and to introduce "strange" symbols was gradually substituted by an emblematic identification between the distinctiveness of Wapishana as a "true language" (rather than "gibberish") and its different sounds: retroflex sounds, vowels articulated in a different part of the oral cavity, a range of fricatives. At this stage, |k|, |w|, |x|, and |z| found their way back into the Wapishana alphabet. How, for example, were they to spell that central vowel? The succession of its orthographical symbols is also significant. The |u| of the Evangelists is rejected because of its confusion with another Portuguese vowel, while the |i|, which is used in Macuxi for the same sound, is rejected for failing to mark an important ethnic distinction. In the end, another grapheme – |y| – is now accepted without much concern, despite being a symbol that had previously been considered "foreign".

The discussion was characterized by two concomitant, but contradictory, tendencies, at least for the rigours of a linguist: on the one hand, the development of a sensibility to the phonetic and phonological peculiarities of Wapishana; on the other, a rejection of the abstraction of the purely phonetic script of the Evangelists. They thus carefully registered the long, phonemic vowels (syz, syyz) and the glottal stop, another phoneme of Wapishana. At the same time, however, the "new" spelling left room for phonemic fidelity, thereby distinguishing it from the "logical" writing of the North American missionaries. It was thus decided to keep the [u] and [o] variants and the evident results of the processes of palatization.

Teaching material was elaborated with the new script, before I finally left Roraima and the Wapishana teachers on the eve of a risky experience, the outcome of which was unpredictable: to teach the native tongue of the parents and grandparents of students who spoke Portuguese and who had been alphabetized in Portuguese. Our working meetings had been a true study of the structures of the language and of the history of each of the available scripts; the new script was the product of successive and different evaluations of the nature of the written code. Once a process was underway in which a discussion of writing had become an axis for a reflection on the school, the crisis, the alternatives, power and linguistic diversity, it was no longer easy to predict the direction in which things were headed.

The Wapishana of Malacaxeta were living through a tension between the precariousness of autonomy and the bargaining that was necessary for the concession of official support. Meanwhile, the old "masters of the language" were not recognized as formal educators and, in an ambiance of mistrust, the teaching material produced by the Indians was seen to be a threat, since it eluded official government or missionary scrutiny. Finally, it was impossible to predict the difficulties and the equivocations of implementing the teaching of Wapishana as a second language, and even of alphabetization in the Indigenous language, in the complicated context of a language crisis.

With their "new" writing, the Wapishana were at a crossroads, stuck in a paradox. Some conclusions could be envisioned from past experiences; as for the future, plausible hypotheses depended on the way out of a junction filled with contradiction. Which crossroads and which paradox? The new script was yet another version of "civilized" writing and integration; on its own, it was also a vehicle for the new language which emerged from the reified word of orthographical technology. It thus added to all of the steps that, ever since the "invention" of writing, have transformed the traditional cultures of orality. In the historical encounter between orality and writing there are losses and acquisitions, both of which are definitive. We must still carry out a critical accompaniment of these transformations where they are in progress, where we can still witness the first phases of this encounter.

There is often a disdain for a critical perspective before the more or less immediate support for the diffuse ideology that conjugates a civilizing mission and the need for cultural and linguistic "salvage" through education and writing. In the case of the Wapishana language, we might ask what is the true meaning of this salvage, a word that is in the mouths of everyone: Indians, missionaries and agents of the State. Rescue, salvage; to preserve what language, which traditions? What is Wapishana after the long crisis that asphyxiated it and its transfigurations in various orthographies? What kind of rescue operation is the reification of "myths" into childlike stories? What are the (somewhat predictable) consequences of a project of teaching the Indigenous language as a second language in the disciplinary space of formal schooling?

What, in short, is the meaning of this rescue operation once its processed as a superficial rhetoric by the "Indigenous education" programmes financed by the State? The Wapishana of Malacaxeta followed a double strategy, hoping to maintain a delicate equilibrium. They were quick to include themselves in the official programmes for publishing pedagogic texts – the first step in an ample programme of implementing bilingual education – so that they could publicly validate their new script and, thus, to announce a language reborn from the scorched earth of linguistic assimilation.


The Kuikuro in the dance of the letters

We still need an ethnography of the ongoing or final experiences of writing in Indigenous societies. I did not dedicate the same attention to the Wapishana, whom I met on few occasions, as I did to the Kuikuro,13 whose language and livelihood I have been studying for the past thirty years. I was the protagonist and the authority responsible for the genesis of what may today be called the writing of the Karib language of the Upper Xingu. Theirs is a very different, and apparently more tranquil, story than that of the Wapishana, since the Upper Xingu was, until very recently, an area from which missionaries were barred. Is this, then, a virgin terrain for the serene practice of applied linguistics and for a trauma-free discovery of writing by the Indians?

The experience of writing existed before I arrived as a linguist and a researcher in 1981 and helped to rudimentarily alphabetize a young man undergoing puberty seclusion.14 It continued to develop, with greater or lesser intensity, in-between each of my fieldwork periods. Among the first encounters of the Kuikuro with the written form of their words is, without a doubt, that of the spelling of their names in documents and medical files, which was later appropriated as signatures, and re-appeared in notes, sculpted onto trees or pieces of wood, painted in posts and doors. The names of the Kuikuro appear in a variety of different spellings, and each one of them implies some "deafness" on the part of the whites.

The sound of a vowel that is inexistent in Portuguese – the high central/frontal vowel that is represented by the [] symbol in the International Phonetic Alphabet (IPA) – has sometimes been written as u and sometimes as i, thereby denying its distinctivity by assimilating it either to one of the high vowels in Portuguese or, less often, with |y|, the symbol which has traditionally been employed for the same sound in the orthographies of Tupi-Guarani languages. Upon "discovering" this sound, the young Karib men undergoing alphabetization attributed it the |ü|, an invention, a bricolage of visual elements used to graphically nominate this sound and to thus distinguish it from other high vowels (i, u) present in their language or in Portuguese, but avoiding the distinctivity of the letter |y| used by the neighbouring Kamayurá, who had been compelled to follow the tradition of writing in Tupi-Guarani languages. As an example, we can see the different spellings for the name of a Kuikuro woman. The first spelling is the linguist's transcription, utilizing IPA, and is included here as a reference:

Msé       Musé    Misé     Mysé    Müsé

Different spellings of proper names remain osculant and concomitant, including the current spelling (Müsé) used by teachers and students of the village school. At present there is a conflict between the existence of a "norm", a "way of writing correctly" from the point of view of those who have undergone formal schooling, and "wrong" spellings which, nonetheless, remain in the official documents that function to identify individuals before the institutions of the State that supply health care, services and goods, upon which the Indians are increasingly more dependent.

There is yet another aspect to this congealment of the spelling of proper names. In the societies of the Upper Xingu, an individual receives two stocks of names, one from the maternal and one from the paternal side, and changes his names (to the plural) during each new stage of the life cycle: childhood, puberty, birth of the first child and the first grandson. Documents and files thus indelibly perpetuate only one phase of the identity of each person, a fact that depends on the moment in which a name was definitively inscribed in the logic of identification of the world of the whites. Even in a context such as that of the Upper Xingu, in which no church held baptisms, the addition and adoption of a name of the "caraíba" (the whites) is a means to overcome the discomfort that arises in the occasions in which these two logics confront each other. Even the ethnonymn "Kuikuro", through which this population has come to be known, contains in its spelling a dense history of meanings:

Kuhikuu     Cuicutl     Cuicuru       Cuicuro     Kuikuru        Kuikuro      Kuhikugu

At the end of the 19th century, the German ethnologist Karl von den Steinen recorded, among the various inhabitants of the banks of the Culuene river, the existence of the Guikuru or Puikuru or Cuicutl (Steinen 1940 [1894]). Steinen observed the difficulty in representing a distinctive and particular sound, which is nonetheless very common in the Karib languages of the Upper Xingu. The sound in question is an uvular tap which is sometimes heard as a voiced velar fricative or as a voiced velar occlusive, sometimes registered by the untrained ear as a voiced velar occlusive [ɡ] or an alveolar tap [ɾ]. There is still no specific symbol for the uvular tap in the International Phonetic Alphabet, which is why we have come to provisionally represent it with the symbol for voiced velar fricative []. The name that Steinen heard and attempted to record is, in fact, Kuhikuu, the local group that, at the time, inhabited the kuhikuu area. It is a contraction of kuhi ikuu, "stream of kuhi fish", near a lake that is rich in kuhi fish (Potamorraphis, fam. Belonidae). The Kuhikuu were the first village of a new local group (ótomo, in the Kuikuro language) that split from the other Karib local groups of the Upper Xingu, probably some time before the mid-nineteenth century. They were the founders of a people that the whites continue to call Kuikuro, but which still auto-designates itself Lahatuá ótomo, after the name of the village that was forcibly abandoned in 1954 following an outbreak of measles that decimated half of its population.

Steinen was intrigued by the phonetic quality of the of the uvular flap and was able to describe it with precision, even proposing to adopt a symbol from the Greek alphabet – the lambda – to transcribe it, or else to use the digraph |tl|. After Steinen, the peculiarity of this sound condemned it to be represented by the grapheme |r| and it was a toilsome and almost dramatic process that led Kuikuro teachers to recognize its specificity and to begin search for an "adequate" orthographic representation. As always, on the edge between recognizing a "specific reality" and the desire to avoid marks of excessive difference, in particular in what pertains to Portuguese as a normalizing reference, the young "owners of writing" finally chose |g|, a compromise between the register of its articulatory quality and a letter from the alphabet. The deformation of the name Kuhikugu ótomo – ancient and ancestral – was congealed in the collective name of their descendents and on the individual surname of each one of them: to the whites, they were Kuikuro.

A further example illustrates a different case, this time concerning a proper noun that is also a common noun ("pepper"):

Φomi    Fomi     Homi

In the Kuikuro language there is an alternation between two sounds, the bilabial fricative [Φ] and the glottal [h], and they are indicative of a generational and positional variation. The use of |f| seeks approximation with the first variant, used by elders and characteristic of the "beautiful speech" of formal traditional discourse. In the writing of those who have been schooled, the letter |h| is dominant, representing the variant of younger people and condemning the first variant to disappear from the norm that is to be imposed as correct.

The alphabet, memorized and recited in school, which appears in the first page of all learning material, is an almost untouchable, almost sacred object, listing its letters in perfect order; writing once again came to be seen as a constituent part of the language of the Caraíba.15 Once all of the possibilities of associating certain sounds of the Indigenous language with one or another of the existing letters have been exhausted, where does one place the letters, such as the digraphs and trigraphs?

The orthographical decisions were taken in conversations between teachers and between teachers and the linguist-advisor. All of this occurred with some degree of anxiety, in a sort of calculus that would ponder the possible alternatives, the limits of possible interventions in the alphabet, the explanations of the linguist, and the emergence of a metalinguistic conscience that writing gradually confirmed. Thus since the |g| was already taken, the nasal velar was presented by the digraph |ng|; |g| and |ng| already being taken, the pre-nasalized voiced occlusive gave rise to the trigraph |nkg|,16 a fairly complex and difficult symbol, though less so for native than for non-native speakers. Below we see the successive and concomitant spellings of another proper noun and the terms for "maraca" and for a ritual that is known in the ethnographical literature as "tawarawanã":

O|i           Oni       Ogi       Ongi

ã|ke          anke     ange     ãke       ãge       angke

ntuhe         ntuhe    nduhe   duhe

And how is one to fix nasalization or vocalic lengthening, which results from fairly complex prosodic and phonological adjustments? The acoustic matter seemed to always escape its orthographic entrapment; a problem solved now was bound to turn up again later. All hesitation was interpreted not as a symptom of the hiatus between the oral and the written, but rather as incapacities of the "writers" or as being due to the nature of the Indigenous language.17 In little time the expression "our language is poor" was substituted by "our language is difficult", another prejudice that they absorbed in their contact with the whites, their language and their writing. They thus went from the attribution of "paucity" (simplicity, primitivism, etc.) to that of "difficulty", an apparently contradictory claim when uttered by a native speaker. After all, as I have argued elsewhere (Franchetto 1995, 2000, 2001), the "orthographicization" of the native language is probably less of a "conquest" – a process marked by feelings of discouragement and frustration which, if not for the imperatives of "bilingual education" that came from the exterior as a necessary step in formal education – and more likely a process that led to the abandonment of "writing in the native language". Significantly, the genesis and impact of writing approximates, in this way, the Wapishana and the Kuikuro, two completely distinct people in language, culture and history.



Throughout Brazil village schools proliferate, as do the courses aimed at preparing Indigenous teachers and the publication and circulation of didactic material (books, pamphlets), bilingual or in Indigenous languages, according to formal directions that present themselves as the definitive implementation of bilingual, intercultural, specific and differentiated education. Yet experiences similar to those of the Wapishana and the Kuikuro continue to exist, and they proliferate in the same proportion as the dissemination of practices and proposals for writing in Indigenous languages. Educational agents, government sponsored or otherwise, seek the standardization and "nationalization" of these scripts, motivated by pragmatic imperatives. They thereby ignore or annihilate dialectical differences and structural characteristics of languages, while at the same time they do not hesitate in producing and using various orthographies for the same language when control over projects, financing, souls and territories are at stake. The Indians either remain perplexed by the dance of the alphabets or get carried away by it.

The arena of ideologies in conflict, in which the Indians of the literate world are the actors or the victims, also includes the disagreements between the linguist advisors. It is understandable that many non-missionary linguists who work with Indigenous languages (the expression "Indigenous linguistics" has been coined for them) strive to erase the Evangelist missionary legacy, distancing themselves from it in many, and often contrasting, ways. There are those who accuse Americanist linguistics to be the bearer of a phonemic, grapho-centric dictatorship, even if it declares itself to be focused on the documentation of oral languages and proclaims the supremacy of orality as an object of scientific attention (Barros 1993). There are the proponents of "spontaneous writing", which remains unmonitored by criteria that presents itself as scientific, who underlie the historicity of writing and the role of the Indians as actors/users who navigate creatively from one system to the next, as various systems succeed one another or remain concomitant for a single language/ethnic group. Those who are party to "spontaneous writing" – "Write! Write any way, right or wrong, it doesn't matter" – are convinced that it is imperative and primordial to immerse oneself in writing as if it were a revigorating bath that makes it an instrument of an immediate and integral expression. "To do linguistics", applying academic or scientific knowledge in the genesis of a script, is condemned as an authoritarian and colonialist exercise.

At the other end are a handful of linguists who elaborate a severe critique of the domesticating operations enacted by "orthographicization" of Indigenous languages – nationalization, standardization, choice of vehicular languages or those for alphabetization, the erasure of inconvenient acoustic characteristics – in favour of a competent, but not naïve, application of scientific knowledge, relying on the involvement of the Indians as teachers and students of linguistic wisdom. The respect for Indigenous languages in so-called "educational" projects is here established in the construction of knowledge and conscience in both the linguist and the speaker, a position that is argued by, for example, Gomez-Imbert:

A "good" script for the linguist depends on a competent study of the language, on native participation in this study, on an understanding of what a script is, on the (joint) establishment of phonological script that avoids vehicular languages or those for alphabetization in detriment of minority, "weak" languages which are, at any rate, destined to disappear [...] The general practice is to simultaneously teach how to speak, read and write in Spanish (or Portuguese) to children who do not understand it. The result is frustration, self-commiseration, self-inferiorization, evasion and failure at school, reinforcement of interior and exterior stereotypes [...] It is my conviction as a linguist that if the orthographic system that children first learn for their first language establishes a coherent relationship between written code and implicit (internalized) knowledge in children in what concerns their own tongue, the task of learning to read and write would be much more simple [...] (Gomez-Imbert 1998, n/p).

Within this perspective, it is believed that certain effects of the technology of words brought by the whites should be made explicit and be redirected in order to make them into an object of conscious apprehension.18

The Xavante of Mato Grosso divide themselves between those who use the |t| and the |ts| of the Evangelical missionaries of the SIL and those who prefer the |d| and |dz| of the Salesians; the first are faithful to the principles of phonological writing; the second ignorant (or tolerant) of the phonetic realization of phonological units. A Xavante teacher, a candidate for a place in the first university course for Indians, once asked me: "What is behind this? I am here because I want to study more to have an answer".

Tonal languages always run the risk of losing their tones, acoustic elements that are as distinctive as the segments represented by letters, because alphabetic script does not tolerate the visual chaos created by the superimposition of "exotic" diacritics. "Orthographicization" thus becomes the sieve through which what it enables becomes established, but also for the condemnation of vital parts of a language. Certain linguists claim that the filter of writing is innocuous: the acoustic structures will remain in operation so long as an integral knowledge of the language is maintained. Yet, do we know enough to ignore the interference of the experience writing on orality at the moment of its inoculation?



Arquivo Indigenista da Diocese de Roraima. 1986. "Povos indígenas do nordeste de Roraima". Boletim n. 11.

BARROS, Maria Cândida D. M. 1993. Lingüística missionária: Summer Institute of Linguistics. Tese de doutorado em ciências sociais, Unicamp, Campinas.

CARDONA, Giorgio R. 1981. Antropologia della scrittura. Torino: Loescher Editore.

Centro de Informação da Diocese de Roraima. 1988. Igreja a Caminho, ano 4, n. 10.

Diocese de Roraima. 1983. Roraima Indígena, n. 2.

FRANCHETTO, Bruna. 1995. "O papel da educação escolar na domesticação das línguas indígenas pela escrita". Revista Brasileira de Estudos Pedagógicos, 75(179-181):409-421.

________. 2000. "Escrever línguas indígenas: Apropriação, domesticação, representações". Catálogo da Exposição Os Índios, Nós. Lisboa: Museu Nacional de Etnologia. pp. 44-50.

________. 2001. "Assessor, pesquisador: reflexões em torno de uma experiência em educação indígena" In: A. L. da Silva & M. K. L. Ferreira (orgs.), Práticas pedagógicas em escolas indígenas. São Paulo: Global. pp. 87-106.

KAHN, Marina. 1986. "As Missões de Fé e os cursos de missiologia". Trabalho apresentado no Grupo de Trabalho sobre Política de Pesquisa Lingüística, as Missões de Fé e as Línguas Indígenas, XV Reunião da Associação Brasileira de Antropologia, Curitiba, março de 1986. Ms.

GOMEZ-IMBERT, Elsa. 1998. "Writing the tukanoan languages: educational politics in the Vaupés area in Colombia and Brazil". Paper presented in 14th Congress of Anthropological and Ethnological Sciences. Workshop "Politics and Culture in the Northwest Amazon". Williamsburg, July 1998. Ms.

Informativo do Setor Indígena da Diocese de Roraima. 1981a. Anna Maimu, Waparadan, 15. s/p.

Informativo do Setor Indígena da Diocese de Roraima. 1981b. Anna Maimu, Waparadan, 16. s/p.

NIDA, Eugene. 1957. Learning a foreign language. Cincinnati: Friendship Press.

RICARDO, Carlos A. (org.). 2000. Povos indígenas no Brasil, 1996-2000. São Paulo: Instituto Socioambiental.

STOLL, David. 1982. Fishers of men or founders of empire? The wycliffe bible translators in Latin America. London: Zed Press.

STEINEN, Karl von den. 1894. Unter den Naturvölkern Zentral-Brasiliens. Reiseschilderung und Ergebnisse der zweiten Schingú-Expedition 1887-1888. Berlin: Geographische Verlagsbuchhandlung von Dietrich Reimer. Tradução brasileira: SHADEN, Egon. 1940. "Karl von den Steinen. Entre os aborígenes do Brasil Central". Revista do Arquivo Municipal, pp. 34-68.



1 The term maloca, often rendered as ‘longhouse', is a part of the local Portuguese vocabulary, and refers to the villages of the Indigenous peoples of the Roraima savannah (Macuxi, Taurepang and Wapishana) – often no more than rural villages
2 The work was carried out in the malocas of Boca da Mata and Bananal (Taurepang, municipality of Boa Vista), Napoleão (Macuxi, municipality of Normandia), Taba Lascada and Malacaxeta (Wapishana, municipality of Bonfim).
3 After an initial stay in the mid-nineteenth century, and after the diffusion of the important Aleluia cult – a religious movement that originated in the Anglican missions of Guyana in the last quarter of the nineteenth century – the Unevengalized Field Mission entered Brazil in 1968 from Guyana, were it had been acting since 1950. From the 1940's onwards, the Baptist Mid-Mission, the Seventh-Day Adventists and the Pentecostals gained considerable influence. Currently, the influence of the MEVA (Missão Evangélica da Amazônia [Amazonian Evangelical Mission, or the Regular Baptist Church) and of the Assembly of God is widespread. Numerous Evangelical churches are present among the Indians of the Roraima savannah, but the missions involved in linguistic research and educational practices are MEVA (Missão Evangélica da Amazônia), Brazilian New Tribes Mission, the MICEB (Missão Cristã Evangélica do Brasil [ Christian Evangelical Mission of Brazil]) and the Unevangelized Field Mission in Guyana. They all form a sort of constellation, the gravitational centre of which is the SIL, acronym of the Summer Institute of Linguistica, at present re-baptized, in Brazil, as the Sociedade Internacional de Lingüística. MEVA, with its headquarters in Boa Vista, the capital of Roraima, is the point of reference for the Evangelists who work in the region; the SIL is the most powerful and developed institution in Brazil in what concerns logistical support for other missions, and in the techniques, scientific preparation and in the regular training of agents for conversion and linguistic work in Indigenous areas. It is, in fact, a complex network of missionaries, whose contributors and mentors live in the United States of America.
4 The terms "caboclo" and "civilized" refer, respectively, to the inhabitants of the malocas of the savannah (Macuxi, Taurepang, Wapishana), who are considered to be "acculturated", and non-Indians. The two terms make up a triad of categories along with a third term – "Indian" – used exclusively to refer to the Yanomami, "savages of the jungles" of the western mountains.
5 At the time, the Wapishana numbered 6,500 people in Brazil (Roraima) and 4,000 people in Guyana, according to the estimates of the Instituto Socioambiental (Ricardo 2000);
6 The term "gibberish" (Portuguese: gíria, literally: ‘slang'), which is how both the Indigenous and non-Indigenous peoples of Roraima refer to Indigenous languages, appears to have been diffused in the 1950's; and they connote shame and stigma. Their status as (true) "languages" is denied, this label being reserved for Portuguese.
7 From this angle, the Wapishana case was aberrant. We shall return to this shortly.
8 The Projeto Calha Norte (PCN, "Northern Corridor Project"), a Brazilian military programme for the revival of the international borders of Amazonia through its occupation by bases, garrisons and villages, was getting under way in the 1980's. A large part of the Amazonian border crossed Indigenous territories, dividing ethnicities into two or more countries. The political climate in the border regions – and Roraima is a border state – was saturated with nationalist feelings, when not outright hostility, in the definition of "friends and foes".
9 As of 1948, the Mission of the Consolata, with seats in Normandia, Surumu and Maturaca, succeeded the Benedictines, who had arrived in the Upper Rio Branco in the early twentieth century.
10 We have used the current graphic conventions to distinguish sound, in brackets; phoneme, between oblique lines; and grapheme, between vertical lines.
11 To make it clear that we are dealing with phonemes that are not exotic to the world of writing, we can follow the history of the value placed on <k> in Italian, as narrated by Cardona (1981:120): "[...] In the last decades, the use of <k> in Italian initially took on (ironic) connotations of modernity [...] after a certain date, <k> comes to assume negative, political connotations; after the 1972 film ‘The Amerikano' by Costa Gravas, in which the protagonist is a CIA agent, the <k> comes to connote imperialism, repression, violence [...]".
A concern with an adaptation to the national language and its consequences is characteristic of most of the experiences of Indigenous peoples with writing. According to Gomez-Imbert, writing of her experience among the Colombian Tukano (1998: n/d): "Establishing a practical and adequate written system to be used in bilingual education means facing technical problems with an ideological solution. A ‘good' script must be approximated to Portuguese or Spanish [...] this ignores the large structural differences that exist between the Romance languages and Indigenous languages, such as certain phonological and morphological properties that cannot be adequately represented by the conventions used for Portuguese or Spanish". 
13 The Kuikuro are one of the four local groups that speak a language belonging to the Karib family, and they are situated along the eastern headwaters of the Xingu River, in the Brazilian state of Mato Grosso. They currently number some 500 people living in four villages.
14 The experience of writing among the Kuikuro – an experience that is simultaneously the linguist's and the Indians' – is described and interpreted in a text published in the catalogue to an exhibition at the Ethnological Museum of Lisbon in December 2000 (Franchetto 2000).
15 This representation of writing underscores the meaning of the phrase "we do not want to mix the things of the Indians with the things of the Whites", which is often said by traditional leaders faced with the proposals of implementing bilingual education in village schools, implying a resistance to the "orthographicization" of Indigenous languages and to its use in the space of the school (Franchetto 2001).
16 We can see that the script of the Kuikuro, like that of the Wapishana and of many other languages whose speakers do not integrally abide by the principle "one phoneme/one grapheme" (a characteristic of Americanist linguistics as applied by Evangelical missionaries, as well as being an ideal norm), shows distinctive and subphonemic elements. These include those that result in phonological processes of assimilation and "resyllabification" (palatizations, vocalic harmony, voiceness and pre-nasalization of occlusives pre-nasalization of occlusives preceded by a nasal).
17 A typical example is the fluctuation in the writing of a single individual in what concerns the establishment of separations/spaces between words. There is here a constant conflict between consciousness of the hegemony of the "word" unit in writing, the "word" reality in one's own language, the junctions and phonological frontiers, and the inherent ambiguity in, for example, the clitic elements.
18 This is the practice in certain ongoing experiences. In the ethno-educational seminars that occurred among the Tukano of Colombia, during certain courses for Indigenous teachers in Brazil, such as those held in the Xingu Indigenous Park, or in higher education in the universities of Mato Grosso, a meta-language is created to explicitly analyze linguistic and cultural knowledge through reflections and collective activities. The Indians become conscious of the richness of their languages when they discover that they obey rules that can be appropriately formulated and worked through, and that they are not only a bunch of words (or sounds) as many of the whites would have the Indians believe.