Brazilian Literature and Anthropology are still unknown among many readers and researchers of foreign communities, mainly concerning the elements related to the nationalist social formation and to the conception of identity. Therefore, it is important to focus on the process of exposing sociocultural and political aspects of Brazil, in order to encounter similarities/differences and to demystify prejudices, which have been ingrained in the ideologies since the Colonization Era. Considering these elements, the mediation among languages through a translational process¹ is widely necessary and assigns to the translator the responsibility for creating the ideal recognition of theories and proposals, as well as the capability to represent cultural, political and ideological manifestations.

With this in mind, in our research, we intend to promote a study about the translation into English of a work written by one of the most important contemporary Brazilian novelists and anthropologists, namely Darcy Ribeiro. His theoretical and literary formulations represent, in a South American social environment, relevant educational, political and sociological comprehensions of the problems Brazil has been facing through centuries.

Thus, analysing his works constitutes not only the recognition of his vast and complex conceptual approach but also represents the contact and the identification of Brazilian people with their principles of nationality, which can be understood considering its historical and ethnological aspects. Darcy Ribeiro utilizes a complex lexical-semantic field which expresses peculiar factors of Black, Indian and White communities, and characterizes the style of this Anthropologist, which makes the translation of his works a transcultural challenge.

Consequently, by using a disciplinary approach proposed by Camargo (2005, 2007), which has, as its main theoretical support, Corpus-Based Translation Studies (BAKER 1993 1995 1999; OLOHAN 2001), we verify a corpus composed of Darcy Ribeiro’s work: Maíra (1978), as well as its translation into English: Maíra (1985), performed by Goodland and Colchie. We also consider Faulstich's and Barbosa’s theories (2002), for the concepts of brazilianisms, variation and loans in Terminology, involving their social, ideological and historical elements, as well as Barbosa’s (2006) lexical studies.

Based on the prominence of these translators during the process of presenting Darcy Ribeiro’s Brazil to the Target Audience, we study their style and amplify the discussion by comparing it with other translators’ main behaviour present in Translational English Corpus (TEC), which is composed exclusively of translated texts into English from a number of European and non-European languages.

Our main hypothesis is that when we cross-reference all the data related to translational process for brazilianisms in Darcy Ribeiro’s work with TEC, by using the resources of corpora investigation, we identify a configuration for the use of loans and variations for this terminological lexicon. This allows us to conclude that these elements of language can promote a stylistic pattern, which is a response of cultural factors and of their influence on the development of Source and Target Audiences as social groups.

In order to develop our study, we consider Baker’s (1995) definition of corpus, which, according to her works, represents “[…] any collection of running texts (as opposed to examples/sentences), held in electronic form and analysable automatically or semi-automatically (rather than manually)” (BAKER 1995: 226).

We used two types of corpora, a parallel one (which is our main corpus) and a corpus of texts translated only into English (TEC). Parallel corpora are defined by Baker (1993: 238) as "corpora of source texts and their translations". The parallel corpus we verify in this work consists of a main corpus with source texts (STs) in Portuguese extracted from two of Darcy Ribeiro’s works, and a corpus with their target texts (TTs) in English.

According to Baker (1992), studying corpus helps to establish a style, a kind of "fingerprint" that is expressed in a variety of linguistic and non-linguistic features. Regarding the translation process, style can be related to an investigation that focuses on ways of expression that are typical of a particular translator, aiming at selecting his characteristic use of language and its individual profile of linguistic habits, comparing them to other translators. In other words, style may involve describing translated texts and linguistic preferences in translation, such as the incidence of loanwords as well as characteristics of variation.

Olohan (2001) points out that the conception of style provides a wide range of linguistic patterns across translators and texts. The researcher confirms the idea that corpus analysis assists in the recognition of the features and choices, which may be linked to a textual genre (e.g. novels vs. theoretical works), to the use of specific structures, to translators' individual ‘styles' or even to editorial intervention.

Regarding the variation in Terminology, Faulstich (2002) observes the language as an unstable sociocultural element which reveals that "terms (…) are entities subject to variation and to change and that communications between members of society are able to generate interactional concepts for the same term or to generate different terms for the same concept" (FAULSTICH 2002: 70).

Esteves (2010), in her work Um estudo sobre a equivalência conceitual entre termos do português do Brasil e do inglês: aspectos lexicais e semânticos², follows Faulstich’s theoretical framework (2002). The author notices that the concept of terminological variation bounds the definition of term as well as the comprehension of its use in different linguistic systems.

Thus, terms should be described as:

Under the researcher’s perception, a term has different specialized language functions. Faulstich’s analysis (2002, 2004) as well as Esteves’ investigations (2010) support the idea that terms assume specific functions"according to the context of use"; and that, in similar conditions of use, "variants will be considered similarly" (FAULSTICH 2002: 75).

Both researchers list a series of norms which may guide the theories about terminological variation:

a) Dissociation between terminological structure and homogeneity or unity or monorreferentiality, associating terminological structure to ordered heterogeneity notion;

b) Abandonment of the categorical isomorphism between term and concept meanings;

c) Acceptance that the terminology is a fact of language and that it accommodates variable elements;

d) Acceptance that the terminology varies and this variation may indicate a change in course;

e) Terminology analysis of linguistic co-texts and discursive contexts of written language and oral language (FAULSTICH 2002: 76).

Regarding the translational process, these factors support the establishment of a correlation between possible changes in analytical perspectives from one language to another, through the identification of functions that take variants within the linguistic and social communities. Faulstich (2002: 76) observes that terms are closely related to the function they play in a social and cultural system, and their performances start as a pragmatic and empirical entity which determines the possible "variation mechanism".

Faulstich (2002) postulates that some variants:

The author believes that the occurrence of such variants is by pairs of loanwords and vernacular forms and she adds that:

Such set of variants confirms the propositions of our study, since in the course of the analysis we could find a lot of loaned terms, and, thus, we observed the interference of Portuguese in English language as well as in the translation correspondents which formulate the anthropological Terminology related to brazilianisms. We could also verify that Portuguese presents autonomy to be shown during the composition of Social Sciences Terminology, as well as Indigenous and African languages, and also some dialects used in different Brazilian regions, for example: grupiara, cabaça, cauim, jurupari, pajé, tuxauá, caboclo, cabra, capanga and paçoca.

In the translational process for texts which have brazilianisms as their basis, it is possible to understand that different stages occur and that the terminological items adequate themselves to numerous types of loans and variations. Therefore, in the translation context, elements such as discourse, regionalism, geography and time are important aspects to be considered during the professional activity.

In order to develop this study, we created a parallel corpus, constituted by the literary work: Maíra, written by Darcy Ribeiro and originally published in Portuguese, in 1978, and its respective translation into English, Maíra, performed by Goodland & Colchie, published in 1985.

We also utilize the TEC, which is composed of translated texts into English from a number of European and non-European languages. Moreover, TEC contains the largest collection of texts worldwide, with around ten million words, which was also the first fully computerized. It is divided in the categories of fiction, biography, newspapers and in-flight magazines. For our investigation, the fictional corpus is essential, whose translated texts are entirely incorporated. The stored translations are done by native speaker translators, and most of these translated texts were published after 1983. Among the texts selected to compound the corpus related to Literature, we focus on works translated from Brazilian Portuguese into English, such as: 1) Turbulence by Chico Buarque, translated by Peter Bush; 2) Discovering the World by Clarice Lispector, translated by Giovanni Pontiero; 3) The Hour of the Star by Clarice Lispector, translated by Giovanni Pontiero; and 4) Whatever happened to Dulce Veiga? by Caio Fernando Abreu, translated by Adria Frizzi.

With regard to the tools used for this research, on one hand, we used the program WordSmith Tools - version 4.0 (SCOTT 2004), and its applications WordList, KeyWords and Concord. We obtained a list of possible terms in the novel written by Darcy Ribeiro in Portuguese. Then the wordlist in English was also observed, and, thus, two lists were prepared: the first in Portuguese, and the other with the corresponding terms in English.

On the other hand, in TEC, data identification proceeded with the help of a software provided by TEC itself, which is called TEC Tools. We selected the corpora by utilizing the TT search tool (select sub-corpus). Then, we created a wordlist of each translated work and selected the possible cases of loans based on the observation of words in their context. In the end, we looked for loans in the respective co-texts, using the text extraction tool (Extract).

In order to verify the parallel corpus composed by Darcy Ribeiro’s literary work and by its translation into English, we intended to analyse how Goodland and Colchie develop their translational stylistic patterns. To reach this objective, we firstly generated the frequency list (Table 1), which was extracted from the ST, using the WordList. We also generated concordance lines based on the most frequent words in Portuguese and in English (Table 2). In the tables below, we find ten of the most frequent words in the ST and in the target text (TT).

Table 1:

N	Word	Freq.
1	DEUS	345
2	SENHOR	232
3	GENTE	216
4	MUNDO	213
5	MAÍRA	193
6	CASA	186
7	HOMENS	173
8	ALDEIA	156
9	TEMPO	154
10	MULHERES	153

Table 2:

N	Word	Freq.
1	GOD	325
2	PEOPLE	255
3	FATHER	239
4	MEN	226
5	HOUSE	215
6	WORLD	200
7	WOMEN	195
8	VILLAGE	160
9	INDIANS	131
10	GROUND	128

Regarding the most frequent words, we detected that the most occurring words in ST, for example, casa (186) and aldeia (156), also have a high level of repetition for their correspondents in TT, namely: "house" (215) and "village" (160).

It is also important to verify plausible changes in meaning that may be performed during the writing of TTs by translators, focusing on some aspects and neglecting others, thus, producing new texts. In this sense, by noticing the repeated use of words in the ST, we can evaluate, for example, terms like "village", used by Ribeiro 156 times in his literary work. In its common definition, one “village” is a land location that represents the smallest demographic unit found among human groups. This construct, therefore, relates to the notions of "town" and "city".

In target language (TL), “village” is represented, according to The Sterling Dictionary of Anthropology (DAS 1997: 141), as a collection of houses treated as a unit and allocated in order to compose a community in which all residents are integrated. The dictionary also states that the “village” proposition has begun to exist during Neological Period, although there have been records of houses like these since 8000 B.C., in Egypt.

According to the concept of style, it is compelling to verify that, in this example, Goodland and Colchie establish a sense of respect for Brazilian Indigenous “villages” which rebuild the ideas of these nuclei’s regularity and uniformity. Consequently, their TT refers to the “village” as the simplest type of settlement that is more or less dispersed.

In Brazil, it is important to consider that the term aldeia is more often applied to Indigenous settlements. It is linked to the context of "sedentary collector people" and to the purposes of "subsistence production". The internal organization is established similarly to a “village”, based on political and religious powers of a spiritual leader determined according to a hierarchy based on kinship lineage.

Hence, the translators deal with fields that combine anthropological writings transfigured into a literary text. Their style reveals some sociocultural aspects, concealing others. For example, when they translate the term povo, they frequently choose the word “people”. However, Goodland and Colchie use this term 39 times more in TL, which shows a new text organization that sums up, probably, other occurrences of terms such as povo, pessoas, povoações.

It is necessary to consider that the translational style encompasses the conduct of variation related to lexical choices for some terms found in ST. Faulstich (2001) indicates the existence of variability especially with regard to social and cultural issues, which led us to consider the possible occurrence of this phenomenon in the translation of the works that compound the main corpus

in this research. Therefore, we could observe that there were many different options assumed by Goodland and Colchie to the translational process for brazilianisms (Chart 1). In Chart 1, we present some examples of variation:

Chart 1:

Brazilianisms in Maíra ST	First option	Second option	Third option	Forth option
Feitiço	Spell	Sorcery	-------	------
Guri	Boy	Tot
Panema	Adversity	No Lucky
Patuá	Basket	Box of Woven Straw	Wealth of Leather Sack	Woven Grass Basket
Nhá	Miss	Dona	Mistress	Mrs.

In those examples, the word feitiço, according to the Dicionário de Ciências Sociais (1986), represents beliefs in supernatural attacks, which require the search for the attacker. There are, therefore, the concepts of "sorcery" and "witchcraft" that are linked with “guessing” and “premonition”.

By choosing to vary between the terms “spell” and “sorcery”, Goodland and Colchis understand the concept of “speech” which contains values that govern human acts, as shown by The Dictionary of Anthropology (1961), in which the meaning of “spell” is designated as a number of words recited or sung that some groups believe to be able to perform a magical wish. The effect of a "spell" is mandatory and, if properly presented, the expected effect comes naturally. An archaic Terminology may be used in the "spell" in order to make it unintelligible to the new generations and to cause more discomfort and intimidation.

The meaning found in this dictionary for the term “sorcery” refers to the cause or cure of diseases, used, generally, by people properly designated for action. However, within anthropological Terminology, the term "sorcery" is associated with "witchcraft" and reports tribal societies' hidden actions. The distinction is made by Anthropologists regarding the idea of malignancy involved in the process, which, according to Evans-Pritchard (1937), is configured in the ability to cause a disease, which is carried on by a possessor of (dark) "magic".

Translators also opt for associating loans to other explanatory terms. Therefore, the more intrinsically related to regionalisms and rituals a term is, the more translators link its loans to translational procedures.

In this context, we present Chart 2 with examples of loans tied to the use of general language terms in order to offer new themes to Anthropology, and, at the same time, to support this theoretical area with vocabularies recognized by readers, allowing them to select their own understanding for concepts and theories provided by STs and TTs (Chart 2):

Chart 2:

Brazilianismsin Maíra ST	First option	Second option	Third option	Fourth option
Ambir	Ambir	Ambir the old	------	------
Anhangá	Evil Spirit	Devil	Anhangá	------
Inharon	Inharon	Berserk	Madman	------
Jacui	Jacui	Living Flute	Piping Guan
Javari	Javelin	Lance Javari Palm	JavelinThrowingMatch	Javari

When Goodland and Colchie face terms that are distinctly “Brazilianized”, it is possible to observe that they construct a relational behaviour, promoting, at first, a loan for the term in SL and then articulating them to words that explain it or that make it more explicit, for example, in the case of Anhangá. This brazilianism is the name that Tupi Indians used to give to spirits who wander on Earth after death and who torment living beings. According to legends, these ghosts take the form they want, even though they prefer the figure of a fiery-eyed deer with a cross drawn on its forehead.

The term origin is bound to the word añanga, a mythical animal, hunting and hunters’ protector. However, for White colonizer culture, Anhangás are "evil spirits", demons, often confused with anhangueras (Old Devils).

We also noticed the intense reuse of direct loans (Chart 3). In Chart 3, we present some examples of this procedure:

Chart 3:

Brazilianisms present at Maíra ST	Goodland and Colchie’s translation option
Anacã	Anacã
Avá	Avá
Bororo	Bororo
Caariara	Caariara
Curupira	Curupira

The translator presents the text as a place of social interaction, which they tends to explore, analyse, interpret and rewrite. The formation of the translational style, here, runs through the construction of other behaviours and skills related to the role of the translator as a social actor as well as a researcher, reader and analyst.

Following this premise, from the selected words, we carry out the observation of concordance, using the Concord tool. Thus, the terms, in contexts, can dialogue with the lexicon and the nuclei of adjacent words, favoring the reading and interpretation of ST and TT (Chart 4). The keywords of a work compound "extended" linguistic elements in order to show the correlation between words and terms. We present below, in Chart 4, two of ST keyword and the expressions formed by them (aldeia/s and índio/s):

Chart 4:

ALDEIA/S

Aldeia Mairum - Aldeia antiga - Aldeia dos índios mairum - Aldeia indígena - Pequenas aldeias - Aldeia natal - Pátio da aldeia - Caminho da aldeia

ÍNDIO/S

Puta de índio - Histórias dos índios - Índios mairuns - Índios bravios - Índios hostis - Índios Xaepês - Índios arredios - Língua dos índios - Ação de guerra dos índios - Deus de índios e de pretos - Menino-índio - Índio genérico - Ex-índio - Fundação Nacional do Índio - Índio convertido - Índios selvagens - Cemitério de índios - Índios bravos - Iparanã dos índios

We also checked the formulations offered by Goodland and Colchie with respect to the same keywords in TL (Chart 5). Below, we present Chart 5, containing extracted expressions of Maíra TT:

Chart 5:

VILLAGE/S

Dancing ground of the village - Mairun village - Village of Corrutela - Tuscan village - Canindejub of the village - Division of the village - People of the village - Genoese village - Village of wild Indians - Village of the maiun Indians - Indian village - Half-deserted village - Great village of vulture people

INDIAN/S

Indian Protection Service - Hostile Indians - Wild Indians - Mairun Indians - Indian’s dialect - Xaepês Indians - Indian girl - Untamed Indians - Old Indian woman - Language of the Indians - Indian Blood - Indian boy - Groups of Indians - Ex-Indians - Indian village - National Indian Foundation - Indian cemetery - Converted Indian - Generic Indian - Forest Indian - Real Indian - Indian population

As we have been presenting during our research, most of the terminological set related to Brazilian society undergoes some changes of meaning and of concepts during the translational process, which leads to a standardized language developed by translators' style.

We can conclude that Darcy Ribeiro brought to light various aspects of national culturality to be worked out as part of the anthropological language, creating new terms and referring to various elements of "Brazilianness" to build his analysis. These characteristics formed the brazilianisms nuclei observed in our research.

The WordSmith Tools software, through its tools and utilities, greatly facilitated the search in a large amount of data, which was obtained much more quickly and accurately than had it been obtained manually. The concordance lines served as support and clarified doubts regarding the Terminology, because they indicated the co-texts in which terms are inserted.

Using TEC and investigating its literary works favored the knowledge about corpora, as well as the exploration of translation options that may or may not be repeated when working with brazilianisms.

We noted the formulation of a translational style that supports variations and loans, linking them to the high frequency that occurs in both corpora. The terms were culturally relevant because they showed translational style features and allowed us to verify that even in works of different areas, there is a behavioural pattern that translators employ similarly.

As with WordSmith Tools, TEC Tools also represented an important analysis instrument, allowing us to handle a large amount of data and to verify elements of social and linguistic nature based on statistics. Therefore, we can comprehend not only the translational style shared by the translators of our corpora, but also the different possibilities and the creation of new terms in the anthropological universe.

Finally, the theoretical-practical approach developed by Baker and Olohan favored the analysis of style and the investigations about the social act of translating through the perception of linguistic and terminological phenomena. Besides, the Brazilian works in TEC associated values and principle of nationality to our main corpus, allowing the researchers to deepen their analysis and to intensify the knowledge about Brazil.