RSS feed

L2 translation in Brazil: Results of a survey

Rebecca Atkinson, Federal University of Rio de Janeiro

ABSTRACT

This study reports on an online survey conducted in Brazil with the aim of identifying the prevalence of L2 translation in the country and some of the characteristics and practices of the professionals engaged in it, focusing on translation into English. 522 valid responses were received from professionals living in Brazil (95%) and abroad (5%). 78% of the respondents whose L1 was Portuguese reported translating into other languages: 53% only into English and 11% into English and another language. 60% of the 326 respondents who translated into English did so 50% or more of the time. L2 translation of all types was reported, including audiovisual and literary. Only 27% of the respondents who did L2 translation into English always had their translations revised by a third party, and just 21% worked with a native English proofreader. 52% of the respondents who did L2 translations into English had done no training in Portuguese-English translation. High social media usage and interest in short L2 translation courses were identified, suggesting that video e-learning could be exploited for L2 translation training. Trainers and course designers should consider altering the proportion and content of practical training offered in L2 translation to reflect market demands.

KEYWORDS

L2 translation, questionnaire, survey, Brazil, English, translator training.

1. Introduction

Translation into a non-native or non-dominant language — known also by the terms ‘inverse translation’, ‘A-B (or C) translation’ and ‘L2 (or L3) translation’ — entered the translation studies agenda in the 1990s, a time when major shifts on the global geopolitical stage prompted a new wave of globalisation and thus unprecedented demand for the translation of technical and specialised texts. Translation into the dominant language (L1 translation) had traditionally been taken for granted in works of translation scholarship (Apfelthaler 2020) and had largely been imposed as a quality assurance by professional entities and institutions (Beeby Lonsdale 2009; ATA 2011). Indeed, despite active discussions about directionality in the translation literature, it appears that the L1 translation norm still largely holds true even in the higher education setting of Spain (Horcas-Rufián and Kelly 2019), where reflections on and research of L2 translation and its teaching have long been pursued. In Brazil, L2 translation has been studied from an empirical-experimental perspective, but rarely addressed from a market or translator training perspective (exceptions being Ferreira and Schweiter [2017] and Atkinson [2021]), and as a consequence it is also underrepresented in university education (Costa 2018). Whether this underrepresentation is reflected in translators’ everyday practice is the object of the study reported in this paper.

2. Overview of research into L2 translation

Generally speaking, the study of L2 translation has been pursued using an empirical-experimental approach. In this kind of research, translation is treated as a process that takes place inside the professional’s brain and methods of scientific enquiry are adopted to find out what goes on inside this ‘black box’. The aim is to design repeatable and replicable experiments in which specific variables are controlled in a bid to reveal selected aspects of cognition brought into play during translation (Alves 2002). When it comes to directionality, the main objects of interest have been the relative effort involved in translating texts into and out of the L1 (e.g. Ferreira 2010; Pavlović and Jensen 2009) and whether/how variables such as experience or text type influence translators’ performance and the translation product (e.g. Whyatt 2019; Dubĕda and Obdržálková 2021).

Alongside studies of this nature, surveys have also been conducted to gauge the prevalence of L2 translation in specific contexts and even some details of translators’ L2 translation activities. In a Germany newly unified after the fall of the Berlin Wall, Schmitt (1990, cited in Wimmer 2011) undertook a survey with 622 professional translators, finding that a high proportion of the respondents produced L2 and even L3 translations, and that there was more demand for translations out of German than into it. At around the same time, research in Finland (Betke 1987, cited in McAlester 1991) revealed that demand for translation out of Finnish tended to be greater than into it, but that the majority of the country’s translators had Finnish as their L1, leading to the conclusion that L2 translation was commonplace in this context.

More recent surveys have confirmed the inference that in countries whose languages are of limited diffusion, L2 translation tends to be the norm. In Croatia, for example, Pavlović (2007) devised a questionnaire which was answered by 61 translators and interpreters whose L1 was Croatian and L2 was English. Of this total, only two reported never doing L2 translation, both of whom were subtitlers. Meanwhile, 73% reported translating over half of the time into English, and a surprising 32% stated that over 80% of their workload was of this nature. When asked about their preference in terms of directionality, 33% reported preferring translating into English and 23% were equally comfortable doing L1 or L2 translation. Nonetheless, when they were asked directly about their beliefs with regard to L2 translation, 42% agreed with the view that translating into an L1 was the only way to assure a natural, accurate, effective translation product.

Also in Eastern Europe, Whyatt and Kościuczuk (2013) investigated L2 translation in Poland, finding a similar state of affairs. The vast majority (91%) of the respondents of their survey reported the existence of demand for L2 translation in the Polish market, and 42% stated that the lion’s share of their work was based on Polish source texts. As for their attitudes, 78% reported feeling positive about L2 translation, with the most experienced translators being more likely to express this view. The authors also found an association between preference for L2 translation and working in the fields of business, economics and e-commerce. One differential of this study was the fact that it surveyed only translators, whereas the study in Croatia also included interpreters.

While it might be expected for Finnish, Polish or Croatian translators to find themselves in a position of having to perform L2 translations, it is perhaps more surprising to find a similar phenomenon in a country like Spain. Investigating L2 translation there, Roiss (2001) found that 84.4% of the 330 translators who took part in the study did L2 translation and that some even translated more out of Spanish than into it. Thirteen years later, in a survey involving 500 translators, Gallego-Hernández (2014) found that 41% reported translating out of Spanish always or often, although curiously French was the most common L2 among these respondents, followed by the global lingua-franca, English. In the most recent study to date to come out of Spain, over 75% of the 232 respondents engaged in L2 or L3 translation, with 20% doing so at least half of the time (Horcas-Rufián 2022).

In neighbouring Portugal, although no comparable data have been gathered, Ferreira-Alves (2012) did conduct a study of L2 translation among 28 translation companies operating in the country via the quantitative analysis of responses to a closed-ended questionnaire. One of the main findings was that the language into which more texts in Portuguese were translated was English. Also, and somewhat paradoxically, the respondents’ discourse revealed attitudes which associated translation quality with L1 translation, while in practice they regularly employed Portuguese L1 translators to translate into other languages.

Also in Europe, the OPTIMALE study (Garcés and Toudic 2013), coordinated by the Directorate-General for Education, Youth, Sport and Culture of the European Commission, aimed to ascertain the status of Europe’s translation companies. When it came to directionality, almost 40% of them regarded L2 translation competence to be important or essential. Another study of an even broader scope — this time, aiming to obtain a sample of translators from across the world — was “Translation into a non-native language,” organised by the International Association of Professional Translators and Interpreters (Piróth 2015). The survey received responses from translators and interpreters working in 80 (predominantly European) countries and 60 languages. Fewer than half (321, or 41.9%) of the 767 translators and interpreters did not perform L2 translation. Of the 446 who did, just 186 (41.7%) reported restricting this practice to an area in which they had expertise; the majority translated texts from all areas of knowledge into an L2. As for their L2 translation workload, the weighted mean of the responses of all the countries was 20.2%, with the weighted mean for Portuguese being only slightly higher.

Although this survey was quantitative, there was room for the respondents to leave comments on each of the questions. Four of these were made by people working in the Brazilian market. Apparently, the first two — on public translation — are from the same person. In Brazil, this service is provided by translators who pass a public competition and is regulated by legislation that draws no distinction between L1 and L2 translation. As such, as the translator in question reports, once you pass the competition, “the law says that I MUST translate in both directions” (Piróth 2015: 11). What could be seen as problematic ends up having its advantages in the “translation of convoluted and florid Brazilian legal opinions into English. It takes a native speaker of Brazilian Portuguese to sort out the mess” (Piróth 2015: 64). In other words, in official legal translation, where the accurate transposition of the informational content is paramount and the translation is explicit — meaning that the translator neither has to nor should create the illusion that the translated text was written directly in the target culture — then the fact that the translator’s L1 is the source language could actually be a boon.

The other two comments on translation in Brazil (which could also be by the same person) represent the other side of the coin: strong criticism of the quality of translation into English (L2) in Brazil. The first is written by someone whose L1 is English:

Innumerable Brazilians, native Portuguese speakers, claim to be able to translate into English, their non-native language. The results are usually atrocious. The clients don’t know how awful the translation is. [...] I know only one colleague [...] who has the translating skills to do so with excellent results; however, details in her translations into English give away the fact that it is not her native language, so I edit her work (Piróth 2015: 28; emphasis added).

These are generalisations that set great stock not just on L1 translation prowess, but also on having an L2 translation reviewed by a native speaker of the target language to eliminate any trace of the source language, which the writer clearly feels should be avoided at all costs. It is worth noting here that it could be in the interests of such ‘native’ speakers to perpetuate the native/non-native dichotomy and defend the purity of English written by an L1 speaker over a form of global or international English, precisely to perpetuate their role as arbiters of the language (Pokorn 2005: 9).

The fourth comment is along the same lines: “In Brazil, standards are very low. Clients do not usually care if a translator is native or not in the target language” (Piróth 2015: 28). Poor-quality L2 translation is enabled by clients’ blasé attitude toward the native speakership (viz, quality credentials) of the translators they hire. It is a value judgement that brings to mind a booklet published by the American Translators Association, Translation Getting it Right, one of whose sections is entitled “Professional translators work into their native language”:

If you want your catalogue translated into German and Russian, the work will be done by a native German speaker and a native Russian speaker. By the same token, native English speakers translate from foreign languages into English. [...] A translator who flouts this basic rule is likely to be ignorant of other important quality issues as well (ATA 2011: 16).

The ATA does, however, allow for the occasional exception: “Sometimes a linguist with special subject-matter expertise may agree to work into a foreign language. In this case, the translation must be carefully edited—and not just glanced through—by a language-sensitive native speaker before it goes to press” (ATA 2011: 16).

In Brazil, as mentioned above, there is scant literature on L2 translation outside the empirical-experimental approach. Any data on its prevalence are to be found tangentially in studies with other objectives. For instance, in an investigation of ten degree courses on offer at Brazilian federal universities, Costa (2018) found that just four offered L2 translation as part of their syllabus. At two of these, there were far fewer teaching hours of L2 than L1 translation, but at two the practical modules were distributed equally between L1 and L2 translation. However, these are, Costa (2018: 400) writes, the exception: “only a minority offer practice in inverse translation.” Meanwhile, in a study of translation within the Office of the President of Brazil, Noce (2017) found that L2 translation was performed by the vast majority (89%) of the respondents. The text types they tended to translate most out of Portuguese were correspondence and other instrumental texts, but some civil servants were also responsible for translating websites, newspaper articles, speeches, technical documents, videos, reports and other texts into English, Spanish and other languages. However, despite the unquestionable importance and responsibility attached to such communication, these same individuals were not required to have any translation qualifications and received no offer of on-the-job translation training, just some general language courses. Finally, in the most broad-based survey of the translation profession in Brazil, conducted by the Brazilian Association of Translators and Interpreters, Abrates (2015), the question of directionality was not addressed.

In a preliminary attempt to address the issue, Atkinson (2021) mined the translator databases of both Abrates and Sintra (the union of translators and interpreters) and an online survey on a private Facebook group for translatorsto find out what language pairs Brazilian translators were offering. She found that all three sources presented the same profile, with by far the highest proportion of translators offering services from English into Portuguese, followed by Portuguese into English, with translation into/out of Spanish being offered by considerably fewer translators. The data from the Abrates (2015) survey six years earlier also indicate a predominance of English (78%) as the respondents’ main working language after Portuguese. It would appear, then, that even though Brazil shares a continent with countries whose populations are mostly Spanish speakers, the global influence of English makes this its most important language aside from Portuguese.

This study aims to fill a sizeable knowledge gap about L2 translation in Brazil, focusing particularly on translation into English. The expectation is that the production of empirical data on this subject will raise awareness among the country’s translation scholars and professors as to the realities translators have to deal with and encourage syllabus designers to rethink the practical training they offer in terms of directionality, especially at public universities, whose students “come from extremely divergent secondary education contexts” (Silva 2020: 248). In broader terms, the hope is to bring to light the reality of L2 translation in a non-European developing country where private language schools are generally the means by which people are able to acquire the level of second language competence needed for translation activity, making it prohibitive to large swathes of the population, and contrast their experiences with those of translators from more central countries where surveys of this nature have been conducted. It is the first stage of exploratory doctoral research situated in the subfield of translator studies (Chesterman, 2009) in which I will go on to investigate the training, working practices, professional interactions and attitudes of Brazilian freelance translators working into English (L2).

3. Methodology

The questionnaire Panorama da tradução e versão no Brasil (Panorama of L1 and L2 translation in Brazil) was developed using Google Forms with the aim of identifying the prevalence of L2 translation and some of the characteristics of the translators engaged in this activity in the Brazilian market. It was divided into the following sections: (a) education/training and professional activities; (b) translation from Portuguese into English (restricted to those who stated they translated into English); (c) training in L2 translation; (d) translation tools; and (e) sociodemographic information. As Brazil is a big country, with a population of over 200 million, and as translation in the country is largely unregulated, making accurate estimates of the number and characteristics of the translators operating in the market impossible, the decision was made to gather as much contextual data as possible without making the questionnaire too long, as it would be helpful for profiling the market, however imperfectly. The questions all had checkbox or multiple choice answers, including an “other” option, and Likert-type scales were used for the frequency and opinion questions. The questions were designed in this way to facilitate data aggregation and ensure anonymity. Initially, the questionnaire was tested with two translators to evaluate the time it took to fill out1 and identify any ambiguities or technical glitches. It was then sent out to the members of Sintra and underwent some slight alterations based on feedback received from some of the respondents. After this, it was released to the members of Abrates and posted on three translation groups on Facebook and one email group. In all, the questionnaire was available online for one month between February and March 2021.

The data were collected on a Google Sheets spreadsheet. First, duplicate responses were eliminated, then the responses given in the “other” option were reviewed and either fitted into the pre-existing categories, grouped into new categories or kept as “other.” The data were then analysed using measures of central tendency (mean and median) and expressed in absolute numbers and percentages.

4. Results and Discussion

The survey received 527 responses, five of which were duplicates, leaving 522 valid responses. The respondents were mostly female (78%, 406) and white (82%, 430)2,3. Most (95%, 496) lived in Brazil, primarily in the south-eastern region (73%, 361), home to the two states from which the largest number of responses came: São Paulo (37%, 185) and Rio de Janeiro (31%, 152). Answers were received from all but five of Brazil’s 26 states and the federal district, covering all five of its macro-regions. Half of the responses were from individuals aged between 26 and 45 (26-35 years, 29%; 36-45 years, 22%).

The vast majority (472, 90%) of the respondents provided services as self-employed translators (even if they had other sources of income) and only 22 (4%) were employed by a translation company or agency. The former figure is consistent with the findings of the Abrates (2015) survey, where 86% of the respondents were self-employed. Over half (286, 55%) of the respondents provided translation services for agencies, from both Brazil (196, 38%) and elsewhere (196, 38%). A surprising 74% (386) were (also) commissioned directly to do translation assignments without the mediation of third parties. A closer look at the data showed that this figure ranged from 65% to 84%, with the more experienced translators (15 years or more) being the most likely to have their own clients. Online platforms (mostly but not exclusively specialised in translation) were a source of work for 112 (21%) of the respondents, as were online groups (e.g. Facebook), albeit for fewer (52, 10%), while sourcing work from colleagues was a reality for only 53 (10%). This scenario suggests that the Brazilian translation market is following the trends seen elsewhere, insofar as it is increasingly dominated by translation agencies, probably in response to the changing profile of the market, with its larger volumes, shorter turnover times, and increased professional specialisation and dependency on specialised systems, making it necessary for projects to be divided amongst teams of translators and for professionals who are not necessarily translators (project managers) to operate the interfaces between the translators at the ‘wordface’, the paying customer, and the other professionals and systems involved in the production process (Risku et al. 2019; Olohan and Davatti 2015; LeBlanc 2013).

As for experience, 30% of the respondents had been working in translation for up to five years, 27% were ‘veterans’ and the remaining 43% had between 5 and 20 years’ experience (Figure 1). In terms of education, when the categories were represented on a scale from 1 (secondary school graduate) to 6 (doctorate), the median level of education was found to be postgraduate diploma (38%). Almost two thirds of the respondents (334, 64%) had done some kind of translation course outside a university setting, and two thirds (351, 67%) had a higher education qualification in translation or language. However, 77 (15%) had no translation qualification or training inside or outside a university setting, over half of whom (41) had translation as their main source of income, showing that training in translation is not a prerequisite for working in the Brazilian translation market.


Figure 1. Years of experience

In terms of working languages, only data on the respondents’ L1 and L2 were obtained. Portuguese was the L1 of 92% of the respondents, followed by Spanish (3%), English (2%), Italian, German, French and Chinese (1% or less each). As for the respondents’ L2, English was the most common (76%), followed by Portuguese and Spanish (7% each), French (5%), German (2%), Italian, Japanese, Russian, Chinese, Danish, Dutch and Brazilian sign language (1% or less each).

For the analysis of L2 translation, 42 of the 522 responses were excluded as they were from people whose L1 was not Portuguese, leaving 480. Of this total, 373 (78%) reported that they translated from Portuguese to another language, with 278 (53%) translating only into English and another 60 (11%) into English and another language, i.e. L3 translation. Similar levels of L2 translation have been found elsewhere: 78% in Poland (Whyatt and Kościuczuk 2013), 75% in Spain (Horcas-Rufián 2022) and 58% in the IAPTI international study (Piróth 2015). The data from Croatia (Pavlović 2007) are not comparable since they also include interpreters.

It might be expected that L2 translation would be done mostly by more experienced translators, but this hypothesis was not borne out by the data. Although those with more experience did in fact do more L2 translation, it was common among translators of all levels of experience (Figure 2).


Figure 2. L2 translation by years of experience

The data in this figure and others are divided into ‘translators’ and ‘part-timers’. This was done in an attempt to take account of the fact that three quarters (392) of the respondents had other sources of income apart from translation. In order to separate out those respondents who depended primarily on activities other than translation (e.g. interpreters) or did not self-identify as a translator, the answers to two questions were used: “What is your profession?” and “Is translation your main source of income?”. Those who self-identified as a translator (210) or a translator plus something else (162) and/or said translation was their main source of income (309) were classified as translators, summing 397; the remaining 125 were classified as part-timers, even though they may have engaged regularly in translation activity.

Although L2 translation was so common across the board, fewer than half of the respondents had ever done a course or course module that focused on its specific challenges (Figure 3). This stands in stark contrast with the situation in Poland, where universities offer “extensive training in translation from Polish into English” (Whyatt and Kościuczuk 2013: 75). It is also a far cry from the situation in Spain, where Horcas-Rufián and Kelly (2019) identified mandatory L2 translation modules in 21 of the 26 undergraduate programmes in the country and around half of the master’s courses, even if Horcas-Rufián (2020) still considers this insufficient in view of recent regulatory changes. It should be noted that the Spanish master’s programmes tend to be geared towards the market and therefore differ from the master’s programmes in Brazil, which are research-oriented; the Brazilian equivalent would therefore be the postgraduate diploma courses, which were the median qualification of the respondents of this survey.


N.B. These data represent all the respondents to the survey, including those whose L1 was not Portuguese.
Figure 3. Training in translation from Portuguese to another language

To find out whether there was an association between doing L2 translation and having some training in it, data on the 480 respondents whose L1 was Portuguese were investigated. This showed that 52% of the respondents who did L2 translation had no training in the activity, but that this percentage was even higher (72%) among those who only did L1 translation (Figure 4). To find out whether there was an association between training in and doing L2 translation, a statistical analysis was conducted. This revealed that the odds ratio of “doing L2 translation” to “having some training in L2 translation” was 2.4 (95% confidence interval: 1.49, 3.84), which means that in 95% of cases, a respondent who did L2 translation was 2.4 times more likely to have done some L2 translation training than one who did not do L2 translation, and a respondent who had some L2 translation training was 2.4 times more likely to do L2 translation than one who had no L2 translation training. It should be noted that no conclusions can be drawn as to causality.


Figure 4. Association between doing L2 translation and having trained in L2 translation

In an attempt to find out whether there is pent-up demand among Brazilian translators for more L2 translation training, the respondents were asked whether they would be interested in doing training in translation out of Portuguese, to which 412 (79%) replied that they would. An analysis of the data from these 412 respondents showed a strong preference for online (72%) versus in-person (21%) training. The factors that most appealed to the respondents were cost — with 246 (60%) saying they would do a course in L2 translation if it was not too expensive — and specialisation — with 177 (43%) saying they would do a course that was geared towards a specific type of text or genre. This suggests that short, market-oriented courses would best meet the translators’ own perceived needs (Figure 5).


N.B. Percentages based on the 412 respondents who expressed an interest in doing L2 training in the future.
Figure 5. L2 training preferences

Turning now to translation into English, of the 338 respondents who reported doing so, 12 had English as their L1 and were therefore excluded from the analysis, leaving 326 respondents who translated into English as an L2. Of this total, 34% (110) reported translating into English always or almost always, and 26% (88) did so half or more than half of the time. In other words, 60% of the respondents who translated into English (L2) did so at least half of the time (Figure 6). A comparison with other countries shows a contradictory picture. In Poland, 42% of the respondents to Whyatt and Kościuczuk’s (2013) survey reported doing more L2 than L1 translation, and in Spain, Gallego Hernández (2014) found that 56% of his respondents translated into an L2 often, very often or always. However, a more recent survey in Spain found that just 24% of the respondents who did L2 translation did so more than 50% of the time (Horcas-Rufián 2022). This figure is closer to that found in the IAPTI international survey (Piróth 2015), although in this survey the low L2 translation figure — just 16% — could be attributed to the large number of respondents from the UK and the rest of Europe, home to people from the world over and therefore more likely to be able to source translators whose L1 matches the target language.



Figure 6. Frequency of L2 translation into English among (a) translators and (b) part-timers

As for the types of texts translated into English (L2), the results shown in Figure 7 indicate a strong presence of academic texts of all kinds, translated by 189 (58%) of the total of 326. The volume of academic translation could be a reflection of the incentive to publish in (high-impact) international journals, which often yield a higher evaluation score for both the authors and the postgraduate programmes where they work, whose funding depends in part on the internationalisation of their research activities (CAPES 2020). Another particularly notable category was corporate and business (148, 45%), certainly driven by the globalised nature of business and the fact that many multinationals use English as a lingua franca. Some other outstanding categories were legal texts (83, 25%) and official documents (86, 26%). In Brazil, both are the bread-and-butter of sworn or public translators: translators who passed an official examination granting them the exclusive right to translate official documents for use by the country’s institutions. Thus, although the competition for public translators tests candidates’ translation competence into and out of Portuguese, the only situation in which translations are legally required to be done by a Brazilian public translator is when they are for Brazilian entities (i.e. into Portuguese)4. Nonetheless, more than a third (29, 35%) of the survey respondents who translated such documents into English were in fact sworn translators.


Figure 7. Types of text translated into English (L2)

Two categories with surprisingly high numbers were “audiovisual” (73, 22%) and “literary” (48, 15%). In the former case, it would be interesting to find out what kind of audiovisual material is translated into English and under what circumstances. As for the latter, the mere fact that there are Brazilian works of fiction being translated into English by Brazilians is a phenomenon worthy of more study. A deeper dig into the data shows that the traditional route for publishing abroad through a publishing house is being circumvented in these cases, because only 15 of the 48 respondents who reported translating literature into English provided services for a publisher. It would be interesting to discover whether this market share is essentially self-publication or whether there are other circumstances in which such translation practices occur. Another curiosity is how these translations are produced: whether they follow the traditional translation-proofreading model or whether they are collaborative endeavours, as found in China (Wang 2009), for example.

In the survey, the respondents were asked about how they reacted when they received offers of translation jobs into English (Figure 8). More than half (52%) reported that they accepted jobs of this kind (almost) whenever they received them. The factor that had the greatest influence on a positive decision was whether the source text was from an area with which the translator was familiar (54%). This chimes with what Whyatt and Kościuczuk (2013) found in Poland, where the subject matter of the L1 source text was also decisive for 54% of the respondents. Curiously, 36% of the respondents in the Brazilian survey reported accepting L2 translations into English because they enjoyed doing them. Although there are no studies investigating the relationship between directionality and satisfaction, there are data which indicate that the mental stimulation of the job is a source of satisfaction (Courtney and Phelan 2019). It is therefore conceivable that for some translators, this stimulus could come from the challenge of translating texts to an L2.


Figure 8: Reactions to offers of translation assignments into English (L2)

Another surprising result was that just 13% of the respondents reported basing their decision as to whether to accept L2 translation jobs into English on the fact that it paid better than translation into Portuguese (Figure 8). On the Brazilian market, translation out of Portuguese tends to command higher rates than the opposite direction, so it would be reasonable to expect this pay differential to be a decisive factor. However, the real figure could be higher, as 68 (38%) of the 180 respondents who answered that they “always or almost always” accepted L2 translation jobs when they were offered them did not select any other option, suggesting their answers could display evidence of the primacy effect, when a respondent shows bias for one of the first response options (Revilla and Ochoa 2014). This could not be avoided (e.g. by rotating the answer options) given that the survey was divulged online via a single URL.

The survey also sought to gather data on the use of proofreaders for translations into English. Perhaps surprisingly, it was found that the proofreading of L2 translations into English by a third party was relatively uncommon among the respondents: only 27% said that their translations were always proofread, while 52% reported that their translations were rarely or never proofread by someone else (Figure 9). Overall, the proofreading figures in this survey are considerably lower than what Whyatt and Kościuczuk (2013) found in Poland, where 61% of translators reported working with proofreaders at least sometimes.


Figure 9. Frequency with which L2 translations into English are proofread by another person

This item was followed by a question designed to find out more about the proofreaders with whom the translators worked. It was answered by 190 (58%) of the 326 respondents who reported doing L2 translations into English, with those who reported that they rarely or never worked with a proofreader accounting for all but 12 of the ‘lost’ answers. Just 58 (18%) paid for the services of a proofreader out of their own pocket and a further 57 (18%) counted on the services of someone who worked at the same company or agency (Figure 10a). Only 68 (21%) worked with proofreaders who were native speakers of English (Figure 10b) – considerably lower than the 47% found in Spain (Horcas-Rufián 2022).



Figures 10a Payment/employment and 10b linguistic background of proofreaders of L2 translations

As for the discrepancies between translators and part-timers, it was notable how the former seemed to have more recourse to native English speakers to proofread their work and largely eschewed Brazilian non-translators specialised in the subject matter (Figure 10b). This latter phenomenon could also have to do with the potential for translators operating in the market (and not ‘part-timers’) to be more likely to use CAT tools and therefore have access to translation memories from agencies and other sources, which would contain the genre-specific language that they might otherwise consult specialised professionals about. This was borne out partially by the figures. While 70% of all the survey respondents (522) reported using some specialised translation tool (mostly but not exclusively CAT tools), this figure was higher (80%) in the subgroup of respondents identified as translators who did L2 translation.

A question on machine translation (MT) usage was also asked of all the respondents (522), 61% of whom said they used it. When asked what uses they made of it, checking words or phrases in translations (47%) topped the list, followed by L2 translation (174, 33%) and L1 translation (170, 33%) (Figure 11). However, if we assume, conservatively, that none of the 110 respondents who reported doing L2 translation “always or almost always” did L1 translation, we are left with a subgroup of 412 who did L1 translation vis-a-vis 373 who did L2 translation. The MT usage of these groups was therefore 41% for L1 translation and 47% for L2 translation, indicating MT usage was proportionally higher for L2 translation.


Figure 11. Uses made of machine translation

The last two items for the respondents who translated into English were designed to elicit their beliefs about L2 translation. Asked whether they agreed with the statement that “translations done into a native language are better than translations done into a foreign language,” more agreed than disagreed (44% vs. 17%), which tends to confirm that belief in the superiority of L1 translation holds true, even among translators who regularly earn money from L2 translation. Interestingly, these results are consistent with the findings of Pavlović (2007), in Croatia, but diverge from those of Whyatt and Kościuczuk (2013), in Poland. They are also inconsistent with the findings of some empirical research, where it has been found that experience and text type could have as much or more of a bearing on translation quality than directionality (Whyatt 2019; Heeb 2020).

As for the second statement, that “L2 translation can be done by any translator,” more disagreed than agreed (49% vs. 17%), suggesting they felt that some specific skills and competences are brought into play when translating out of one’s L1. Some scholars have suggested that the competences required for successful L2 translation include genre literacy (Beeby Lonsdale 2003), instrumental competences (Kościuczuk 2016) and target readership awareness (Stewart 2013). Similarly, it cannot be assumed that a given translator’s confidence and ability to assess a translation product are symmetrical in both directions (Kiraly 2000), with experience playing a crucial role in this respect (Whyatt 2019; Dubĕda 2018).

5. Limitations

Although this study has the advantage of being the first on the subject to have been conducted in Brazil, it has some limitations. Firstly, the sample size was just over the minimum set for the study (of 500). It would have been preferable to have more responses, particularly in the middle experience brackets. Also, although answers were received from all regions of Brazil, the overwhelming majority came from the southeast. Furthermore, while the study revealed that 11% of the respondents translated into English and another language, no data were gathered on the respondents’ L3 languages, making a more detailed analysis of their language profile and L3 translation in Brazil impossible. Finally, the two belief statements were only presented to the respondents who translated into English, making it impossible to ascertain whether their views differed from the views of the other translators.

6. Concluding remarks

With its market orientation, this study aimed to paint a picture in broad brushstrokes of what kind of L2 translation is done in Brazil, how, by whom, and what these practitioners make of this activity, focusing especially on translation into English. The results about its prevalence were largely consistent with those found elsewhere, confirming a now widely accepted understanding that L2 translation is the norm rather than the exception in most countries where English is not the L1. However, there still seems to be a degree of contradiction between the reality of L2 translation and the beliefs held by the professionals engaged it. Clearly, the nuances of such beliefs cannot be picked up by an instrument of this nature.

I would suggest that one way to dig deeper into this and other aspects of L2 translation would be to engage in ethnomethodological or fully fledged ethnographic studies. These could be used to home in not only on the variegated views translators have about L2 translation, but on a host of other aspects of L2 translation in selected contexts, producing rich descriptions of situated, real-life practices and bringing forth the discourse of translators and other actors involved in L2 translation processes. There are already studies of this nature focusing on a variety of aspects of the translation process, from working practices in translation agencies (e.g. Kuznik and Verd 2010) to issues of ergonomics (e.g. Ehrensberger-Dow and Heeb 2016). Research that targeted the L2 translation process or how it contrasts with L1 translation could shine a light on CAT tool or machine translation usage, revision procedures, search engine usage and much more in specific contexts. The findings of such investigations could potentially link back to the findings of empirical investigations of L2 translation, indicating points of convergence and divergence and throwing up new research questions.

Another area where L2 translation scholarship could develop is in translator training. The findings of researchers investigating the translation process, whatever approach they use, could and should be harnessed to rethink classroom practices. New approaches could then be piloted in qualitative, quantitative or mixed-methods studies, hopefully having a positive impact on L2 translation pedagogy. On a broader level, university courses that do not yet offer the same number of hours of tuition in L1 and L2 translation should rethink this balance in light of the consistency of the data on the prevalence of L2 translation, at least in countries whose main language is not English. Indeed, if the data from this Brazilian survey are anything to go by, the practical disciplines should cover an equally wide range of text types and modes in both directions, including subtitling and literature. Also, given the limited recourse to proofreaders made by the respondents of this survey, one aim of L2 tuition should be to help novice translators gain autonomy, particularly through developing their strategic and instrumental subcompetences, to borrow the PACTE nomenclature (Hurtado Albir 2017).

Finally, for course curricula to remain relevant and meet real market needs, there have to be consistent efforts to engage actively with the market. In Brazil, according to the data from this study, that cannot be achieved simply by interacting with translation associations, since fewer than half (47%) of the respondents were members. The main channels of communication the respondents reported using were public and private translation/translator groups on social media platforms like Facebook and LinkedIn and messaging apps like WhatsApp and Telegram. For the results of translation scholarship to reach translators, different channels and forms of communication need to be exploited, like YouTube videos or slide presentations, whose links can be shared across social media5. Outreach of this kind can surely only enrich the work of translation scholars and practitioners and hopefully engage more stakeholders in meaningful dialogue, ultimately strengthening translation as a profession and translation studies as an autonomous field of research.

Acknowledgements

Thanks to Daniel Atkinson Oliveira for sharing his statistical expertise. This study was financed in part by Coordenação de Aperfeiçoamento de Pessoal de Nível Superior - Brasil (CAPES) Finance Code 001.

References
  • Abrates (2015). O perfil profissional dos tradutores e intérpretes no Brasil. Powerpoint presentation of survey results. https://pt.slideshare.net/Abrates/o-perfil-profissional-dos-tradutores-e-interpretes-no-brasil-abrates-2015 (consulted 26.02.2022).
  • Alves, Fabio (2002). “O processo de tradução: delimitação de um objeto de estudo.” Cadernos de Tradução 2(10), 9–22.
  • American Translators Association, ATA (2011). Translation, Getting it Right: A guide to buying translation. http://www.vivantraduzioni.it/wp-content/uploads/AITI-Right-translation-guide-2014.pdf (consulted 21.03.2022).
  • Apfelthaler, Matthias (2020). “Directionality.” Mona Baker and Gabriela Saldanha (eds) (2020). Routledge Encyclopedia of Translation Studies (3rd ed.). London: Routledge, 152–155).
  • Atkinson, Rebecca (2021). “Translation into a Non-Native Language: Practice, Research and Training in Brazil and Around the World.” Belas Infiéis 10(1), 1–21.
  • Beeby Lonsdale, Allison (2003). “Genre Literacy and Contrastive Rhetoric in Teaching Inverse Translation.” Dorothy Kelly et al. (eds). La direccionalidad en traducción e interpretación: perspectivas teóricas, profesionales y didácticas, Granada: Atrio, 155–166.
  • Beeby Lonsdale, Allison (2009). “Directionality.” Mona Baker and Gabriela Saldanha (eds) (2009). Routledge Encyclopedia of Translation Studies (2nd ed.). London: Routledge, 84–88.
  • Betke, Paul (1987). Yritysten käännöstarpeista. Auktorisoitu translaattori kansainvälisen viestinnän asiantuntija. Turku.
  • CAPES (2020). Guia para Aceleração da Internacionalização Institucional: Pós-Graduação Stricto Sensu, versão 1. https://www.gov.br/capes/pt-br/centrais-de-conteudo/23122020_Guia_para_Acelerao_da_Internacionalizao_Institucional.pdf (consulted 27.02.2022).
  • Chesterman, Andrew (2009). “The name and nature of Translator Studies.” Hermes — Journal of Language and Communication Studies 42, 13–22.
  • Costa, Patricia Rodrigues (2018). A formação de tradutores em instituições de educação superior públicas brasileiras: uma análise documental. PhD thesis. Universidade Federal de Santa Catarina.
  • Courtney, Jennifer and Mary Phelan (2019). “Translators’ experiences of occupational stress and job satisfaction.” Translation & Interpreting 11(1), 100–113.
  • Dubĕda, Tomaš (2018). “La traduction vers une langue étrangère: compétences, attitudes, contexte social. Meta 63(2), 492–509.
  • Dubĕda, Tomaš and Vanda Obdržálková (2021). “Stylistic Competence in L2 Translation: stylometry and error analysis.” The Interpreter and Translator Trainer 15(2), 172–186.
  • Ehrensberger-Dow, Maureen and Andrea Hunziker Heeb (2016). "Investigating the ergonomics of a technologized translation workplace.” Ricardo Muñoz Martín (ed) (2016), Reembedding Translation Process Research. Amsterdam: John Benjamins, 69–88.
  • Fan, Weimiao and Zheng Yan (2010). “Factors affecting response rates of the web survey: A systematic review.” Computers in Human Behavior 26, 132–139.
  • Ferreira, Aline Alves (2010). Direcionalidade em tradução: uma investigação do processamento cognitivo de tradutores profissionais em tradução direta e inversa no par linguístico inglês-português. Master’s dissertation. Universidade Federal de Minas Gerais.
  • Ferreira, Aline and John Schweiter (2017). “Directionality in Translation.” John Schweiter and Aline Ferreira (eds) (2017). The Handbook of Translation and Cognition. Hoboken: John Wiley & Sons, 90–105.
  • Ferreira-Alves, Fernando (2012). “Translation Companies in Portugal.” Anglo Saxonica 3(3), 232–263.
  • Gallego-Hernández, Daniel (2014). “A vueltas con la traducción inversa especializada en el ámbito profesional: Un estudio basado en encuestas.” TRANS: Revista de Traductologia 18, 229–238.
  • Garcés, Carmen Valero and Daniel Toudic (2015). “Technological Innovation and Translation: Training Translators in the EU for the 21st Century.” Verbeia 0, 183-202.
  • Heeb, Andrea (2020). “Effects of directionality on professional translators’ processes, products and effort.” PhD thesis. University of Hildesheim.
  • Horcas-Rufián, Sara (2020). “Estudio exploratorio-descriptivo sobre la direccionalidad en traducción: perspectiva del traductor en formación.” TRANS: Revista de Traductologia 24, 335–352.
  • Horcas-Rufián, Sara (2022). “L2 translation practice in Spain: Report on a survey of professional translators.” Translation & Interpreting 14(1), 121–141.
  • Horcas-Rufián, Sara and Dorothy Kelly (2019). “Inverse (A-B/C) translation education in Spain under the EHEA.” Perspectives: Studies in Translation Theory and Practice 28(2), 300–319.
  • Hurtado Albir, Amparo (ed.). Researching Translation Competence by PACTE Group. Amsterdam: John Benjamins, 2017.
  • Kiraly, Donald (2000). “Translation into a non-mother tongue: From collaboration to competence.” Meta Grosman (ed.) (2000). Translation into Non-Mother Tongues: in Professional Practice and Training. Tübingen: Stauffenburg Verlag, 117–123. 
  • Kościuczuk, Tomasz (2016). “L2 Translation: to Teach or to Discourage?” Journal of Translator Education and Translation Studies 1, 4–22.
  • Kuznik, Anna and Joan Miquel Verd (2010). “Investigating Real Work Situations in Translation Agencies. Work Content and Its Components.” Hermes — Journal of Language and Communication Studies 44, 25–43.
  • LeBlanc, Matthieu (2013). “Translators on translation memory (TM): Results of an ethnographic study in three translation services and agencies.” Translation & Interpreting 5(2), 1–13.
  • McAlester, Gerard (1991). “Teaching Translation into a Foreign Language: Status, Scope and Aims.” Cay Dollerup and Anne Loddegaard (eds) (1992). Teaching Translation and Interpreting: Training, Talent and Experience. Amsterdam/Philadelphia: John Benjamins, 291–298.
  • Noce, Gisele (2017). Tradutores na Presidência da República Federativa do Brasil: perfis profissionais e política linguística institucional em 20142015. Master’s dissertation. Universidade de Brasilia.
  • Olohan, Maeve and Elena Davatti (2015). “Dynamics of Trusting in Translation Project Management: Leaps of Faith and Balancing Acts.” Journal of Contemporary Ethnography 1(26), 1–26.
  • Piróth, Attila (2015). Translation into a non-native language. IAPTI survey. https://www.iapti.org/files/surveys/2/IAPTI_non-native_report.pdf (consulted 20.03.2022).
  • Pavlović, Nataša (2007). “Directionality in translation and interpreting practice: Report on a questionnaire survey in Croatia.” FORUM Revue Internationale d’interprétation et de Traduction / International Journal of Interpretation and Translation 5(2), 79–99.
  • Pavlović, Nataša and Kristian Jensen (2009). “Eye tracking translation directionality.” Translation Research Projects 2, 93–109.
  • Pokorn, Nike K. (2005). Challenging the Traditional Axioms: Translation into a non-mother tongue. Amsterdam/Philadelphia: John Benjamins.
  • Revilla, Melanie and Carlos Ochoa (2014). “What are the Links in a Web Survey among Response Time, Quality and Auto-Evaluation of the Efforts Done?” Social Science Computer Review 33(1), 97–114.
  • Risku, Hanna, Jelena Milošević and Regina Rogl (2019). Changes in the Field of Translation Project Management: Findings of a Longitudinal Ethnographic Study. Working Paper. https://socotrans.univie.ac.at/fileadmin/user_upload/p_socotrans/ 20190613_Working_Paper_Changes_Risku_Milosevic_Rogl.pdf (consulted 20.03.2022).
  • Roiss, Silvia (2001). “El Mercado de la Traducción Inversa en España: un Estudio Estadístico.” Hermēneus. Revista de Traduccion e Interpretacion 3, 1–10.
  • Schmitt, Peter A. (1990). “Was übersetzen Übersetzer? Eine Umfrage.” Lebende Sprachen 35(3), 97–108.
  • Silva, Márcia Moura da (2020). “A importância do ensino da tradução inversa em curso de formação de tradutores.” Germana Henriques Pereira et al. (eds). Formação de Tradutores: Desafios da Sala de Aula, vol. 9. Campinas: Pontes, 243–262.
  • Stewart, Dominic (2013). “From Pro Loco to Pro Globo Translating into English for an International Readership.” The Interpreter and Translator Trainer 7(2), 217–34.
  • Wang, Baorong (2009). “The Issue of Direction of Translation in China: a Historical Overview.” Translation Journal 13(3).
  • Whyatt, Bogusława (2019). “In search of directionality effects in the translation process and in the end product.” Translation, Cognition & Behavior 2(1), 79–100.
  • Whyatt, Bogusława and Tomasz Kościuczuk (2013). “Translation into a non-native language: The double life of the native-speakership axiom.” mTm Translation Journal 5, 60–79.
  • Wimmer, Stefanie (2011). El Processo de la Traducción Especializada Inversa: Modelo, Validación Empírica y Aplicación Didáctica. PhD thesis. Universitat Autònoma de Barcelona.
Data availability statement

The data that support the findings of this study are openly available in GitHub at https://github.com/danielatk/panorama_traducao_brasil.

Biography

Rebecca Atkinson is doctoral candidate at the Interdisciplinary Postgraduate Programme in Applied Linguistics of the Federal University of Rio de Janeiro, studying L2 translation in the Brazilian market. She has a master’s in language studies from PUC-Rio, where she has also taught and supervised students of the postgraduate diploma course in translation. She is a freelance translator and erstwhile conference interpreter.

ORCID: 0000-0002-6269-3629

Email: rebecca@letras.ufrj.br

Notes

According to a systematic review by Fan and Yan (2010), for an online questionnaire to obtain a good response rate, it should not take any longer than 13 minutes to complete.

Note 2:
The question about skin colour was included because, for historical reasons, Brazil is a racially divided country where access to quality education and opportunities for second language acquisition are still largely the privilege of lighter-skinned people. In the 2010 census, just 47.5% of the population self-identified as “white” — far lower than the 82% of the respondents to this questionnaire.
Return to this point in the text

Note 3:
The sociodemographic questions were not mandatory, which is why each question has a different total number of responses.
Return to this point in the text

Note 4:
Since the survey was conducted, a law (no. 14.195, of 26 August 2021) has been passed that eases the legal requirements for individuals providing official translations. A civil suit has been filed against the new law by a state association of public translators and the constitutionality of the law has also been challenged in the Supreme Court by the national federation of public translators and interpreters (Fenatip). At the time of writing, final rulings on both legal challenges were pending.
Return to this point in the text

Note 5:
A slide presentation of the findings of this survey was prepared in Portuguese and posted online. The link was shared via the same channels by which the survey questionnaire was publicised and also by email to the respondents who gave their email address. It has been viewed by over one thousand people. Available at: https://pt.slideshare.net/Rebecca418680/panorama-da-traduo-e-verso-no-brasil-resultados-de-um-questionrio (consulted 12/12/22).
Return to this point in the text