Translators’ Perspectives: The Construction of the Peruvian Indigenous Languages Act in Indigenous Languages

An urgent need is emerging in contemporary Latin America for the translation of legal texts from the languages of former European colonial powers into the many indigenous languages spoken across the region. This article addresses the issue in relation to the rise of legislation that requires States to uphold the principle of linguistic human rights. It takes as a case study the translation of the Peruvian Indigenous Languages Act (2011) from Spanish into five Amerindian languages, viewed as a postcolonial practice situated at the communicative interface between the State and the country’s indigenous populations. Our specific interest is the strategic behavior of the indigenous translators, as described by themselves, when communicating to their peoples the State norms contained in the Indigenous Languages Act. In order to analyze this behavior, we depart from text-analytical models and favor an approach based on the translators’ perceptions of their role and their rationales for the translation solutions adopted. The analysis combines theoretical strands from translation studies, legal studies and postcolonial studies so as 01.Meta 63.1 final.indd 16


Introduction
The Spanish conquest of the Americas, as was the case with the European imperialist enterprise worldwide, gave rise to extensive and persistent efforts to translate and interpret across the cultural and linguistic divides that opened up between native populations and colonial powers. In early periods, such efforts, aimed at communicating the values and principles of the colonizers to the native populations, focused particularly on translating Christian religious doctrine from Spanish into the Amerindian languages (Durston 2007, Hanks 2010. Today, while religious proselytism continues to provide the motivation for much translation practice, a new and urgent need is emerging for translation of legal texts into the many indigenous languages still spoken across the Latin American sub-continent. Since the late twentieth century, post-colonialist approaches to the academic study of cultural production have informed the field of Translation Studies, especially as far as translation of literary texts is concerned. Our study will broaden the debate by focusing on the translation of a piece of legislation, namely Peru's most recent Indigenous Languages Act (Ley n. 29735), from Spanish into the wide range of Amerindian languages spoken across the Andes and Amazonia. Our specific object of study is the strategic behavior of the indigenous translators, as described by themselves, when communicating the State norms contained in the Indigenous Languages Act to their peoples. To this end, we depart from text-analytical models and favor an approach based on the translators' perceptions of their role and on their rationales for the translation solutions they chose. We seek to combine the tools of translation analysis with strands from legal studies and postcolonial studies in an innovative way, so as to throw light on the translation of legal discourse from Spanish into the indigenous languages of Peru, as conducted by bilingual speakers situated on the cultural "inside." 1 To meet this aim, we will first explore issues relevant to the translation of legislation and comment on the problems that arise from legal translation in a postcolonial, multicultural setting such as Peru. We will then provide an analysis of five translations of the Indigenous Languages Act based on retrospective Think Aloud Protocols (TAPs) conducted with the translators. Finally, we will discuss the findings derived from our analysis. In order to provide relevant context, we commence by outlining the background to the Indigenous Languages Act and the process whereby it came to be translated.
The Peruvian Indigenous Languages Act 2 , as it is commonly known, passed in July 2011, has as its object "to regulate the use, preservation, development, recuperation, promotion and dissemination of the country´s indigenous languages," as expressed in its long title in Spanish (Ley que regula el uso, preservación, desarrollo, recuperación, fomento y difusión de las lenguas originarias del Perú). 3 Given its stated purpose, it was apt that, some three years after the passing of the Act, and as part of a range of actions taken to publicize the existence of the Act among the communities of speakers who constitute its intended beneficiaries, the Peruvian State, in the shape of the Ministry of Culture, commissioned its translation into a wide range of the languages. Unlike the source text, the target texts do not carry legislative weight and are not opposable to the Act in Spanish. Their value, rather, is communicative and symbolic. The texts were published, both in print form and orally recorded on CD, and distributed in a series of public events across the country organized by the Ministry. 4 As stated in the Presentation of the texts, their purpose is "to disseminate the contents of the Act, guarantee its implementation, and contribute to the positive recognition of linguistic and cultural diversity in [Peru]" (Ministerio de Cultura 2014: 3, our translation).
This ambitious initiative has to be considered against a highly complex sociolinguistic and cultural backdrop. As Translation Studies scholars concerned with the sociological siting of translation demonstrate, 5 translation between any language pair is influenced by the relative social-cultural position that each language occupies. In the Peruvian context, the difficulty entailed by the genetic and typological divergences between Spanish and the estimated forty-seven Amerindian languages of Peru is compounded by the fact that the translation process is being conducted between languages whose historical interrelationship carries a burden of social inequality and injustice. Thus, it is reasonable to assume that the translators of the Indigenous Languages Act faced challenges that stemmed from such cultural and linguistic asymmetries.
Our analysis explores the perceptions of five of them (speakers of Ashaninka, Shipibo, Aymara, Chanka Quechua, and Ancash Quechua) as to these challenges and their descriptions of the strategies that they applied in order to negotiate the transfer of meaning. Before we present the findings derived from our analysis, we will contextualize the situation in Peru within relevant literature and explain the methodology that we adopted. the construction of the peruvian indigenous languages act 163

Translating legal texts in postcolonial, multicultural settings
The translation of legal texts, embedded as they are in specific traditions of Law, raises particular problems in multicultural settings. Šaršević (1997: 14) observes: "The difficulty of a legal translation depends primarily on the [degree of] affinity of the legal systems and only subsidiarily on the [degree of] affinity of the source and target languages." Although she appears to relativize the importance of structural differences between languages, it can be argued that these create serious difficulties when the source language has a long-written tradition and well-established legal text-generic and discursive patterns, whereas the target language does not. She uses Switzerland as an example of how "translation operations are greatly simplified" (Šaršević 1997: 15) once a nomenclature has been agreed upon. True though this may be, in the case of Peru, as in other postcolonial contexts, a conceptual asymmetry between the source and target cultures (Spanish-speaking and indigenous, respectively) has to be added to the absence of equivalents in the indigenous languages for the specialized lexicon in the hegemonic language: there is effectively a non-equivalence of systems.
In his epistemic approach to the comparison of common law and civil law in the European context, Legrand (1996) remarks on the crucial role that "legal mentality" plays in the understanding of the law, thereby highlighting the difficulties that stem from the conceptual asymmetry mentioned above: The essential key for an appreciation of a legal culture lies in an unravelling of the cognitive structure that characterises that culture. The aim must be to try to define the frame of perception and understanding of a legal community so as to explicate how a community thinks about the law and why it thinks about the law in the way it does. (Legrand 1996:60) Glanert (2014: 22) elaborates further on the issue: "Within 'comparative law and comparative legal studies' it appears that one must admit that certain deep structures are untranslatable, while significant partial understanding is preferable to no understanding at all." That is, although the language of the law can be translated, the cognitive structures that it embodies, upon which understanding hinges, may not be. This is relevant to the translation of the Peruvian Indigenous Languages Act (a civil-law normative) into the languages of the Amerindian peoples, whose legal culture is rooted in customary law and, therefore, underpinned by different cognitive structures. Additionally, by creating a communicative interface that will enable the native peoples to reach an understanding, however partial, of the principles enshrined in the Act (which directly concern their linguistic rights), is, arguably, an effective way to raise or increase awareness of their entitlements among them.
In relation to the need to translate legislation, Cao (2007: 101) identifies two types of situation in which this is the case: the first one, "in bilingual or multilingual jurisdictions where two or more languages are the official legal languages," and the second, "in any monolingual country where its laws are translated into a foreign language or languages for information purposes." 6 In the first case, the purpose of translation is normative, "to publish the law in the language or languages of the citizens so that the law can be enforced" (Cao 2007: 103). In the second case, "translations are used for informative rather than normative purpose. The translated text does not have any legal force, and the original law and the translated text are not equal" (Cao 2007: 103).
Neither of these types of situation quite coincides with that of Peru. In this case, firstly, we have a multilingual jurisdiction where all languages are de jure official "in the zones where they predominate," as expressed in the Indigenous Languages Act. Furthermore, the cases of jurisdictions such as Canada, Switzerland and Hong Kong, which Cao (2007: 101) uses as examples, presume that the different language versions of the statutes are equally authentic and opposable, which is not the case in Peru: as mentioned above, while the status of the Peruvian Indigenous Languages Act in Spanish is one of legislative power, its translations do not have equivalent status, not having undergone the required authentication process for this to be the case. Cao (2007: 10;citing Šaršević 1997 andCorreia 2003) describes such a process as one by virtue of which the target texts "are not mere translations of law, but the law itself." 7 Thus, the translations of the Indigenous Languages Act into the indigenous languages of Peru would not be, in functionalist terms, "instrumental" (Nord 2005), as their purpose, or Skopos, is not the same as that of the source text.
It could be argued that the situation in Peru corresponds more closely to the second scenario described by Cao. Insofar as the translations of the Indigenous Languages Act do not have the legal status of the source text, there is a function shift which renders an illocutionary act as a locutionary act (Austin 1962): a shift from enacting law to communicating information. However, since Peru is a constitutionally embodied multilingual state and the indigenous languages are deemed official, albeit in a circumscribed way, an analogy with such a scenario, where the law is being translated into a "foreign" language for informative purposes, would be paradoxical.
Given that the paradox just mentioned stems from the status of Peru as a postcolonial country, 8 the intricacies of legal translation need to be considered, for the purpose of the present study, in the context of postcolonial theories of translation. Ever since (and including) Niranjana's seminal work Siting Translation was published in 1992, such theories, rooted in social anthropology, have often materialized in critiques that consider the tension that results from the power asymmetry between the colonizer and the colonized, and put forth suggestions as to how to counter the hegemonic ambitions of the former. In the post-colonialist literature, when texts are used to illustrate the violence that translation exerts on the cultures of the indigenous populations of former colonies, they tend to be literary or religious. Many of the studies relate to the colonization periods relevant in each case or the aftermath thereof, thereby providing historical perspectives and insights. The collections of essays edited by Bassnett and Trivedi (1999) and by Hermans (2006Hermans ( , 2014 on the subject of translation in postcolonial contexts bear witness to these trends. Our study departs from such trends in a number of ways: firstly, it adopts an empirical approach; secondly, it neither draws evidence from nor theorizes on the translation of literary or religious texts; and thirdly, it focuses on a contemporary dataset. Nonetheless, some of the notions proposed by scholars operating within a post-colonialist (and, more often than not, post-structuralist) framework, namely, interventionism and subversion, are relevant in the context at hand. Niranjana (1992: 173) claims that interventionist strategies, which are tantamount to a translation of resistance, can counter the colonial discourse. She was referring to translations of an Indian poem into English (the language of the Empire), but, as our analysis will show, intervention can also work in the opposite direction (from a majority, "world" language, into indigenous, minority, languages) and with respect to instantiations of non-literary genres (in our study, legal texts). Admittedly, this is, by necessity, a different kind of interventionism: one that appropriates the language of the hegemonic discourse, either by trying to replicate it in a lower-status language or by making it relevant to the shared knowledge and conventions of the target audience. The latter strategy is illustrated, for example, by the translation of the Universal Declaration of Human Rights from Spanish into Tzeltal, a Maya language spoken in Chiapas, Mexico, by couching it in a local genre called mantalil, a set of advice and principles used for guiding human behavior progressively, from childhood to adulthood (Pitarch 2001;Howard, Andrade Ciudad et al.).
The above relates to another, much-used, concept in postcolonial theory: subversion. The tenets that are put forth in relevant literature typically relate to translation from the languages of the colonized to those of their colonizers. Spivak (1993Spivak ( /2012, for instance, highlights what she calls the immorality of less-powerful socio-cultural identities being eradicated in translation. Bhabha (1994), on the other hand, argues from a theoretical point of view that the cultural hybridity that evolves from the complex relationships of contact between the colonizers and the colonized can afford the latter the opportunity to subvert the discourse of the powerful.
Bhabha's stance on the affordances of hybridity, questioned though it has been, is relevant to our study in that it is applicable to translation from the languages of the historical metropoles into the indigenous languages of the colonized territories, such as India, the Americas, Africa and Australia. However (and, perhaps, paradoxically), Rao's (2006: 89-91) critique from the perspective of a non-colonial (rather than post-colonial) translation, which, in his words, is "a radically foreign performance" and, therefore, "is not a mere 'rearticulation,' 'revaluation,' 'reversal,' or 'reenunciation' of the original" (Rao 2006: 90) is also relevant. He claims that "the original [is] a radical immanence indifferent to the (colonial) world and therefore untranslatable into it" (Rao 2006: 89). In other words: the source text is inherent to the political and cultural systems that rule (post-) colonial societies, but it is alien to their indigenous populations, which makes translation into their languages a futile endeavor, if not an impossibility.
The notion of untranslatability is, of course, contingent on how "translation" is defined or understood; yet the data analysis that follows leads us to suggest that the respondents found themselves as "foreign performers" of a text that was both "immanent" by dint of its very nature (being as it is part of the Peruvian State's legislation) and "indifferent" to the conceptualization of laws and regulations by their indigenous linguistic communities.
The abovementioned debates have to be contextualized within the notion of "committed translation," a term used by Simon (1996) in relation to feminism that has become associated with practices that visibilize collectivities that are oppressed or discriminated against. The methodology that we chose for eliciting data (see the following section) regarding strategic behavior allows us to ascertain whether the decisions that each translator made in order to relay the Indigenous Languages Act into the indigenous languages covered in this paper were either the result of a deliberate act of subversion of the discourse and genre patterns of the source text, as well as of the conventions of the dominant language (Spanish), or the result of idiosyncratic (Toury 1995) decisions to communicate a piece of the State's legislation to people who are on the periphery of decision-making processes, in terms to which they could relate.

Research methodology
We approach the study of the translations of the Peruvian Indigenous Languages Act via retrospective Think Aloud Protocols (TAPs) with the translators and/or reviewers of the translations into the following: two varieties of Quechua (Chanka as spoken in the department of Ayacucho and neighboring departments, and Ancash as spoken in the department of Ancash); Aymara as spoken in the southern highland department of Puno; Shipibo as spoken in the Amazonian department of Ucayali; and Ashaninka as spoken in the Amazonian region of the department of Junín. All the respondents combine their activities in the field of translation with other professional activities: the Chanka Quechua speaker is a civil servant who works at the Ministry of Culture; the Ancash Quechua speaker is a TV presenter and a digital language activist; the Aymara speaker is a university lecturer specialized in linguistics; the Shipibo speaker is a nurse and a radio presenter; and the Ashaninka speaker is a cultural activist and commentator.
In conducting the TAPs, we invited translators to read aloud the source text, then to read the corresponding translation, proceeding Article by Article through the text of the Act. They were asked to comment on the difficulties they had encountered in the translation process, the solutions they came up with, and how they came by these solutions. The use of TAPs proved the best method to reveal the respondents' perceptions of the strategies that had been applied (the focus of this article) and to elicit reflection. Furthermore, they were a very useful tool in the case of translation into languages of which we had limited practical knowledge. 9 However, while we sought to keep verbal prompts and reactions to a minimum as in conventional TAP procedures, this was an unfamiliar style of interaction for the speakers and we tended toward the more interactive style of TAP argued for by some scholars (van den Haak et al. 2003). Each TAP lasted an average of three hours and in two cases we had to reconvene with the translators, as one session was not sufficient to cover the full text. All of them were subsequently transcribed. In addition to the TAPs conducted with the translators/revisers, during a separate phase of the research we conducted interviews with them, as well as with end users, speakers and readers of the target languages in question, to see how effective and comprehensible the resulting translations were in their eyes. The results of this phase of research will be elaborated on in future publications.

Translating Peru's Indigenous Languages Act in practice
In this section we will provide an outline of the process followed for translating the Indigenous Languages Act from the point of commission. This is followed by an analysis of the translation strategies adopted, as described by the TAP respondents.

Translators' working methods and the nature of the task
The Ministry of Culture, through its Indigenous Languages Division, commissioned the translations of the Peruvian Indigenous Languages Act. 10 The latter set a submission deadline of around a month from delivery of the source text to the translators. The translators had completed the basic three-week training program in translation and interpreting in indigenous languages coordinated by the Division (known as the Curso Básico). The translation into each language was led by one person and revised by another, who, according to the TAP respondents, often took an active role in the negotiation of meaning, thus acting as a second translator. The two parties were frequently based in different locations, in which case the discussions between them took place by mobile phone rather than face to face.
The Indigenous Languages Division provided a basic intra-lingual glossary of legal terminology and facilitated contact between the translators and legal and other experts, who assisted by explaining the technical content of the Act and clarifying linguistic matters. The Aymara and Chanka Quechua respondents explained how this process worked in the following terms: I have not seen the word "aru" used in this sense in any other translation, but we discussed the issue with professor X.
[…] In this case we had to consult the lawyer at the Ministry of Culture himself: what does it mean "direct and inverse translation that guarantee" …? (Aymara TAP: 3 November 2014; translated by the authors) 11 We have a translator who has quite a lot of experience when it comes to translating, reading and so on from a communicative approach, rather than one that seeks to relay exactly something like a calque of things, so we consulted him… (Chanka Quechua TAP: 17 November 2014; translated by the authors) 12 For the recoding of the Spanish source text, and, in order to reach an agreement as to its expression in the target language, the translators often relied on the knowledge and the intuition of other native speakers of the relevant indigenous language. The words of the Shipibo respondent are illuminating in this respect: In order to translate this Act we consulted various people, to reach an agreement, for example my mother, my sisters, my brothers-in-law, cousins, who are bilingual teachers. (…) We received various contributions to the translation, but in the end we were the ones to decide what to write. (Shipibo TAP: 10 December 2014; translated by the authors) 13 A representative of the commissioning (Ministry of Culture) team informed us that, at the beginning of the translation process, a meeting was held with the translators to explain that the aim was not to produce a legally binding text, but rather to communicate the contents of the Act in the indigenous languages to their speakers, in order to spread understanding of their rights under the law (personal communication, email, September 2017). Nonetheless, as our TAP interviews suggest, there was some variation in practice as to how the translators interpreted the purpose of the task. The Ashaninka, Ancash Quechua, Chanka Quechua and Shipibo translators, in line with a communicative aim, simplified the expression and lowered the register. The Aymara reviser took a different approach, rising to what he saw as the challenge of reproducing the specialised register of the source text in the target text.
The latter aimed to maintain the "impersonality" and complexity ("obscurity") of the source text in the target language, as he explained: […] using legal language is difficult, isn't it? It is a different language, isn't it? So, whoever has to translate a legal text must have a good understanding of the Law, otherwise it will be fatal. Because that piece of legislation is going to be applied by a lawyer, isn't it? And, if the translator mistranslates an article, [translates it] in a different way, it is as if it had been underwritten by the President of the Republic, so that translation can prevail and he is ultimately not to blame, is he? And, perhaps, it could lead to an erroneous legal interpretation [thereof]. (Aymara TAP: 3 November 2014; translated by the authors) 14 The other four respondents reworked the content to facilitate comprehension and so that the target audience could relate to it. In fact, the latter were often explicit that their aim was "to communicate the message," rather than reproduce the register of the source text. The Quechua Chanka translator elaborated on this point: It has to be borne in mind that we were translating specifically for the [indigenous] speakers; therefore, it is a group who will not have been necessarily exposed to this type of document. Since they are not familiar [therewith], the aim was to communicate [the meaning], over and above retaining formal aspects. (Chanka Quechua TAP: 17 November 2014; translated by the authors) 15 The Ashaninka, Ancash Quechua, Chanka Quechua and Shipibo translators chose to depart from the "direct and neutral" language of the Act and address their target audiences in terms more closely aligned with the communicative conventions of their peoples. This occasionally led, for example, to the use of first person plural pronouns to replace impersonal expressions. They thus effected a shift in the textproducer/text-receiver interaction: the State addressing its citizens in a detached, impersonal and general manner is replaced by the translator addressing his/her own people as a member of their linguistic and cultural community, as we will show below. Both approaches could be interpreted as "subversive," in that they entail an appropriation of an alien genre (one that is "immanent" to the source culture and "indifferent" to the target culture, to use Rao's terminology; see Section 2). This appropriation lies, in the case of the discursive strategy adopted by the Ashaninka, Ancash Quechua, Chanka Quechua and Shipibo respondents, in the intended effects of the abovementioned interaction shift and, in the case of the approach followed by the Aymara respondent, in the effort to introduce new textgeneric patterns to the target language by reproducing the formal aspects of the source text.

Translation strategies
We shall now present the translation strategies that were adopted to negotiate the meaning transfer, as described by the TAP respondents. They also referred to morpho-syntactic modifications (word order, suffixation) that were determined by the rules of the target language. These have been excluded from our analysis, since they are obligatory shifts and, as such, not the result of strategic behavior, the description of which is the object of our study. We have taxonomized their strategies in the table below; however, it should be noted that the respondents themselves did not necessarily use the labels included in our classification.
As can be seen from Table 1, the translators availed themselves of a range of strategies to overcome the challenges posed by the systemic and cultural asymmetries we mentioned above. Let us deal with each of them in turn.
Only the Ancash Quechua translator added COMMENTS in the form of annotations to justify lexical choices. These were rejected by the commissioners (probably because the strategy is at odds with the conventions of the legal genre), which led to some frustration on her part. She and the Ashaninka translator included DIALECTAL VARIATIONS to make their target texts more inclusive.
The Aymara respondent was the only one who explained that he had resorted to the use of ARCHAISM (thuxriña, a term coined in colonial times to denote an administrator, to translate instancia administrativa [Administrative body]), 16 CALQUE (for interculturalidad [interculturality], combining the Aymara particle purapa [inter] and kultura [the Spanish word for "culture" transcribed in Aymara) and rephonologisation, i.e. an accommodation of the spelling of Spanish terms to the Aymara phonological system (e.g. Pirü for Perú) for terms proper to institutional and legal discourse (e.g.

Congreso de la República [Congress of the Republic], Ministerio de Educación [Ministry of Education], Decreto Supremo [Presidential Decree]).
The Shipibo respondent pointed at OVER-TRANSLATION as a strategy that allowed him to relay his understanding of the text. For instance, he back-translated educación intercultural bilingüe (intercultural bilingual education) as en nuestro propio idioma, así como también se debe enseñar las tradiciones orales [in our own language, and also oral tradition must be taught].
SPECIFICATION was used by the Ancash Quechua respondent to translate medios de comunicación [communication (mass) media]: she specified the most frequently accessed media, as she questioned how many people would understand the collective term. CONCENTRATION, or summary, was utilised by three of the respondents to bypass the conceptual entanglement of the text. The Ashaninka and Aymara respondents did not allude to the need for simplifying the long, convoluted sentences that are typical of legal language: the former described her experience of the task as largely unproblematic and the latter, as mentioned above, approached the task as a way of replicating an alien genre in his native language.
All the TAP respondents with the exception of the Ashaninka speaker indicated that they translated some terms by DEFINITION ) were turned into declarative sentences. Similarly, abstract nominal and adjectival constructions were often turned into interrogative sentences. This was a strategy which was routinely applied, according to all respondents, to the rubrics of the Act: for instance, Definición de lenguas originarias (Definition of indigenous languages) would appear as "What are indigenous languages?" and Objeto de la Ley (Subject Matter of the Act) as "What is the Act for?" AMPLIFICATION and REDUCTION were used to compensate for semantic asymmetries between languages. Interestingly, the respondents who are native speakers of Andean languages (AQ, CQ, AY) explained that they had to use a sequence of terms to relay the meaning of a Spanish word or combination of words (e.g. criterio [criterion], mantener y desarrollar [maintain and develop], dignidad cultural [cultural dignity]), whereas the two who are native speakers of an Amazonian language (ASH, SH) stated that, on occasion, a single word sufficed to cover the meaning of several Spanish ones: for instance, only one word was used to relay the meanings of planificado y progresivo (planned and progressive), pueblos originarios, andinos y amazónicos (indigenous peoples, Andean and Amazonian), and preservación y recuperación (preservation and recuperation).
The Shipibo translator explained that he also used OMISSION for, ostensibly, similar semantic reasons: entidades públicas y privadas (public and private bodies) became "public entities" in his version of the Act. He also deleted comunidades campesinas o nativas (peasant or native communities) altogether. The Chanka Quechua respondent resorted to the same strategy, but, in her case, to avoid redundancy: recuperación y preservación (recuperation and preservation) was translated simply as "preservation," as the idea of revitalizing the indigenous languages was included further on.
Three of the respondents used a HYPERNYM. The Aymara and Chanka Quechua speakers used it to translate Presidente (President), respectively, as "authority" and "head." The Shipibo speaker applied the strategy to Estado (State), which he relayed as apu, a term used by Andean highlanders to refer to the mountains as sacred beings, subsequently adopted by some Amazonian peoples to refer to the leaders of their political organizations.
All the respondents reported that they used rephrasing, borrowed terms from Spanish and metonymy: REPHRASING was used when abstract terminology was used in the Spanish source text, such as mantenimiento y desarrollo (preservation and development), titularidad individual (individual entitlement), ejercicio de sus derechos (exercise of their rights), conocimientos tradicionales y la cosmovisión de los pueblos (ancestral knowledge and peoples' Weltanschauung), identidad (identity), and planificación (planning). The respondents' comments reveal that they resorted to paraphrases that were determined by their respective linguistic communities' cultural experience, sometimes using metaphors related to the natural world: e.g. lineamientos (ruling principles) as caminos que vas a abrir (paths that you are going to clear] in AQ; zonas de predominio (areas of predominance) as [lugares donde se ha] anidado ([places where one has] built a nest) and preservar (to preserve) as brotar (to sprout) in CQ.
BORROWING from Spanish was used across the board for what the translators described as "proper names," i.e. labels that are part of the institutional discourse or refer to State-led initiatives or categories (e.g. distrito, provincia o región [district, province or region], Patrimonio Cultural Inmaterial [intangible cultural heritage], ámbito público/privado [public/private sphere], política nacional [national policy]). It was also used for acronyms (e.g. CONCYTEC). Both the Aymara and Chanka Quechua respondents explained that on occasion they had used citative phrases that mean "the so-called" to bracket the borrowings. The fact that all respondents stated that "proper names" should not be translated, points at the instruction that they received as part of their training (the aforementioned Basic Course). That they interpreted this "rule of thumb" as being extendable to semantic fields that are alien to their peoples' sociocultural constructs was repeatedly attributed to their assumptions as to comprehensibility. The use of borrowing in the indigenous languages is widespread when there is no vocabulary to express exogenous concepts; if borrowings are well established, it is assumed that they will be generally understood. In other words, because most indigenous people will have come across the borrowed terms in Spanish through the media or other means, it was thought that it would be confusing if they were presented with translations for which no frame of reference exists.
Three of the respondents (all of them speakers of Andean languages: AY, AQ, CQ) stated that, in some cases, they had added a definition to the Close analysis of the TAP responses shows that there were inconsistencies in the application of strategies within the same target text. It must also be noted that, according to the back-translations of the target text that the respondents sometimes provided, and the rationales they proffered, the choices that they made appear to have resulted in involuntary (non-strategic) semantic distortions and losses: for instance, afectan derechos [concern rights] became "damage rights" and sea factible [be feasible], "be better" in AQ; and cualitativos [qualitative] became "those who speak better" and en el territorio peruano [in Peru's territory], "when we live in a city" in SH.
However, it is the respondents' strategic behavior as explained by themselves (the object of study here) that is particularly revealing. The Aymara respondent is the outlier: he was the only one who understood the task that he was charged with as an opportunity to use translation as a way to develop a new genre in his language, rather than as a mere transmission of content. The other respondents appear to have seen their role as that of intermediaries between the State and their peoples, privileged communicators by dint of their comprehension of the source text and the structures it embodies, on the one hand, and of the sociocultural systems of their peoples, on the other.
In the cases of the Chanka Quechua and Shipibo translations, the TAP respondents commented that a further criterion guiding their choice of lexicon in translation was whether or not it was likely to be understood by the younger generation. There was a tendency to avoid words that remained in currency among older people but had fallen into disuse among the young. The Shipibo translator, experienced in working in media, had a strong view in this respect: Because old people use very old terms that are no longer part of our everyday speech, especially that of the youth. Old people can use the terms among themselves, but young people don't speak like that anymore and however correct it may be, we have left that language behind and we choose not to use it because it is a very old-fashioned term. We look for another option, what it should be. (Shipibo TAP: 10 December 2014; translated by the authors) 17 Nonetheless, it could be argued that it is the older generations who would benefit the most from having the Act translated into their mother language, as the younger generations tend to have higher competence in Spanish and more access to information about current affairs, along with the attendant terminology and phraseology. This can be interpreted as a reflection of the indigenous language revitalization agenda, which weighs heavily on the minds of the State-qualified indigenous translators. 18 Although the strategies mentioned above illustrate their willingness to make the Indigenous Languages Act accessible, the most obvious example of how the Ashaninka, Ancash Quechua, Chanka Quechua and Shipibo respondents identified with their peoples is the aforementioned shift in text-producer/text-receiver interaction, which influenced their translational behavior. The Chanka Quechua respondent encapsulates the motivation behind the shift as follows: The objective was to communicate and get people to appropriate the Act for themselves, that is to say, this is an Act that defends our rights […]. That is why here the person speaking is a person on the inside, who says "our languages," "ours," so it was a strategic choice too. (Chanka Quechua TAP 2014; translated by the authors) 19 The use of the first-person plural removes the State (as "impersonal" source-text producer from the target text, where it is replaced by a speaking subject who aligns himself/herself with the text receivers. The Chanka Quechua translation reviser clarified this strategy as far as the translation into that language was concerned (personal communication, email, September 2017). She pointed out, on the one hand, that Quechua necessarily requires the speaking subject to be present in the utterance; this subject, in the case of a legislative Act, is the State. On the other hand, she explained that the use of the inclusive "we" in Quechua (ñuqanchik) rather than the exclusive "we" (ñuqayku), positions the State together with all indigenous peoples and refers to all their languages, the word siminchikkuna ("our languages") being in the inclusive plural.
The combination of this strategic behavior with other decisions as to how to relay the content (such as excluding lexical choices that refer to either highland dwellers or Amazonian peoples, depending on the language into which the Act was being translated, as illustrated above under "METONYMY") may make the translated version of the Act read (or sound) as if it was devised for a specific indigenous linguistic community, excluding all others. This would run counter to the State's professed agenda of promoting equality not only between the Spanish-speaking majority and the speakers of indigenous languages, but also among the speakers of all the Peruvian indigenous languages themselves, however large or small their linguistic communities are and regardless of differences in cultural status.

Concluding remarks
There is no doubt that translation played a crucial role in the colonization of what is now the Republic of Peru, as it did in the case of other colonized territories of the Americas. From the case discussed here, it might be expected that translation of the text of a legislative Act, which, by definition, represents the will of a State that occupies a hegemonic position with regard to its indigenous populations, would contribute to perpetuate that hegemony. In terms of the transmission of ideology that the exercise of hegemony entails, translating the Indigenous Languages Act could arguably be expected to be on a par with translating Christian religious texts, as occurred in earlier times and still occurs in the present day.
However, there are significant contextual factors here that serve to counter such an argument. The Indigenous Languages Act was brought before Congress by indigenous members of parliament and approved in response to demands from their grassroots organisations. The translators in question are indigenous people themselves, working avowedly in favour of their communities of origin. The Indigenous Languages Act is thus a manifestation of the will of the contemporary Peruvian State to be seen as upholding and promoting the linguistic and cultural legacy of the autochthonous population of the country. In this respect, this translation initiative can most rightly be seen as countering the abovementioned hegemony, in that it signifies an innovative attempt to translate a piece of State legislation on indigenous rights into the indigenous languages of the country.
In this paper, we have provided an overview of how the communicative interface between the State and the indigenous population is being managed through translation in 21 st century Peru, and, more specifically, an insight into how strategic decisions were instantiated in the translations of the Indigenous Languages Act into five Peruvian indigenous languages. The inclusion of two dialectal varieties of Quechua (Ancash and Chanka) contributed to prove that translation choices were not determined by the make-up of the languages in question, as the variation illustrated in the analysis shows.
Based on our evidence, the translators clearly perceived the source text as a piece of legislation that, supportive though it may be of the linguistic human rights of their peoples, relates to a sociocultural sphere that is alien to those that it seeks to protect (Rao 2006). However, the fact that it is enshrined in the national legislation and that, therefore, it is binding for all Peruvian citizens regardless of their ethnic or linguistic background, did not prevent the translators from making it "their own." It can be argued that the translators' "subversion" of the source text was possible because of the communicative and symbolic, rather than legislative, value of the translations. As the participants in the TAPs related, the target texts became a means to communicate information, to make a piece of legislation accessible and relevant to their peoples (as the Ashaninka, Ancash Quechua, Chanka Quechua and Shipibo translators made clear), or an attempt to reproduce a genre that is alien to the target audience in their own language (as the Aymara respondent asserted).
The lack of equal status between legislation written in Spanish and the translation thereof into the indigenous languages of Peru stands in stark contrast with the situation within national borders elsewhere described by Cao (2007: 103). This can be understood as a manifestation of the asymmetries that characterize intercultural relations in the Latin-American context. The fact that to date there is no system in Peru for authenticating translations of legal texts in indigenous languages (which, arguably, made the aforementioned shift in function acceptable) can be seen to be related to the socioeconomic situation in 21 st -century Peru, a developing country with low levels of education, especially among the indigenous population. An authentication procedure would require the validation of the target texts by a State body that would take the advice of qualified legal experts who have sufficient knowledge of the indigenous languages. These are few and far between, and can cover only a narrow range of languages between them. Additionally, there is at present no official organization in Peru whose remit includes authentication of legal translations into the indigenous languages.
Having said all that, it is noticeable from the TAP respondents' reflections that they were aware of the translation strategies of which they could avail themselves to negotiate the differences between the cognitive structures that characterize their indigenous cultures (Legrand 1996) and those that are immanent (Rao 2006) in the Peruvian rule of law; yet, the major stumbling blocks that they said they had encountered were related to cultural asymmetries between the constructs of the Spanishspeaking State and those of their own peoples. The difficulty in handling the lack of one-to-one lexical equivalents, which is commonplace between any pair of languages, pales in comparison with the difficulty that lies in conveying concepts such as "rights," "heritage" or "private" that are either alien to or conceptualized differently by indigenous communities, as the respondents indicated.
It could be argued that a limitation of the methodology that we chose is that the findings derived from the data analysis may conflict with those that could be elicited from contrastive text analysis. Yet this potential limitation is offset by the insights into translational behavior that the respondents' reflections afforded us; their perceptions were, after all, our object of analysis. The methodology also highlighted the need to determine how the target texts were perceived by their intended target audiences and whether or not the State-led translation initiative raised awareness of linguistic rights among indigenous communities and, just as importantly, among Spanish speakers. 20 These are, we believe, important issues that should be considered when reviewing translation policy in the broader context of language policy. The communicative and symbolic value of the translation of the Indigenous Languages Act into the Amerindian languages of Peru needs to be evaluated in the context of the State's objectives of effectively disseminating information to the beneficiaries of the Act and of raising awareness about language rights among the general population and, especially, the providers of public services.