Home > Issue23 > Alonso article

Analysing the use and perception of Wikipedia in the professional context of translation

Elisa Alonso, Universidad Pablo de Olavide de Sevilla

ABSTRACT

This paper draws on the results of an online survey conducted among professionals of the translation industry (mostly translators) to explore, from a technological and sociological perspective, how they conduct their work, the needs they experience, and the tools and resources (human or human-driven) they resort to when translating. More specifically, this interpretative and descriptive work looks at how participants use Wikipedia and analyses their perceptions of this tool. The survey results suggest that respondents made extensive use of all sorts of technologies when translating, amongst which TM and MT/post-editing were not the most popular. They also resorted to human (or human-driven) resources (translator colleagues, experts, social networks, blogs, etc.) to meet their needs (general documentation, terminological/lexicographical, visual). Respondents had a good overall opinion of Wikipedia (usefulness, reliability and ease of use) and most of them reported using it when translating. However, some results suggest the existence of some kind of controversy or censorship with regard to the use of Wikipedia in professional contexts. A discussion relating the results of this survey to other studies with similar focuses (translation tools, the translation profession, Wikipedia) could help identify trends in the way translators interact with technology in the information society.

KEYWORDS

Wikipedia, translation tools, translator’s profession, sociology of translation, trans-humanisation, MT, TM.

1. Introduction

In the last fifty years, interaction between humans and technology has been analysed in several disciplines, including Psychology, Anthropology and Sociology. In the field of Translation Studies, as pointed out by Alonso and Calvo (forthcoming), much current literature analyses translation technologies “as subordinate parts of the translator's work, relying on approaches that we might consider predominantly human-centered, anthropocentric.” In contrast, this paper adopts Alonso and Calvo’s theoretical approach to what they call the “trans-humanisation of translation,” a proposal which takes a broader contextual aspect of Translation Studies (Buzelin 2005; Chesterman 2007; Wolf 2007, among others) as a starting point for a more detailed look at the now emerging idea of technological ‘extensions’ and their role in the translation process (Pym 2011; Alonso and Calvo 2012; Risku and Windhager 2013):

Our trans-human translation hypothesis (Alonso and Calvo 2012) refers to an extended cognitive, anthropological and social system or network which integrates human translators and technologies, whether specific to translation or not, and acknowledges the collective dimension of many translation workflows today. A technology-mediated approach envisages technologies in action and interaction with the human, fostering a plethora of instrumental developments, not only as isolated fragmentary tools utterly dominated by the human. The creative and learning dimension of technologies in both directions, from the user to the tool and vice versa, also plays a shaping role in this proposed construct (Alonso and Calvo, forthcoming).

In our research, we focus on the way translation professionals make use of translation tools (machine translation and translation memories), and particularly of general, non-translation-specific tools (such as the Internet and Wikipedia) and human-driven resources.

As a ground-breaking and interdisciplinary phenomenon, Wikipedia has attracted the attention of researchers from fields such as Computer Science, Library Science, Communication, Education, Economics, Management, Politics, Language, Law, etc. More precisely, within Translation Studies, McDonough Dolmaya (2012) conducted a survey focusing on how participants in crowdsourced translation initiatives (translators of Wikipedia in English) perceived translation, and Torres Domínguez (2012: 8) carried out a survey that showed, among many other things, that more than 70% of its respondents used Wikipedia as a terminological/lexicographical resource. However, not much attention has to date been paid to the use and perception of Wikipedia by professionals and a number of questions remain unanswered. How often do professionals use Wikipedia in the translation process? For what purposes? Only as an encyclopaedia? How often do translation professionals actively participate in Wikipedia?

This paper examines the results obtained from an online survey conducted among translation professionals. Following a qualitative and interpretative methodology, we attempt to shed light on the way professional translation is carried out in the Information Society. More specifically the survey explores how professionals use Wikipedia in the translation process and how they perceive it. We will first describe the method used in our research; then we will analyse the survey results (description of the sample profiles, use of tools and needs when translating, and perception and use of Wikipedia). After briefly discussing the results, we will present some suggestions for future areas of research.

2. Method

One of the goals of this survey was to determine the profile of the participants and gain a better understanding of their needs when translating and of their perception of Wikipedia. The method of our proposal is based on the works produced by Head and Eisenberg, Lessons Learned: How College Students Seek Information in the Digital Age (2009) and “How today's college students use Wikipedia for course related research” (2010), the reliability of which has been praised by Colin Robson (2011: 58), the acclaimed expert in Social Sciences research methodology. Head and Eisenberg’s works formed part of the PIL (Project Information Literacy), a series of studies developed from the Information School at the University of Washington and aimed at understanding “how college students find information and conduct research — their needs, strategies, and workarounds — for their course work and for addressing issues that arise in their everyday lives” (Head 2013: 472). Head and Eisenberg (2009, 2010) used a multi design approach with a first phase based on focus groups aimed at obtaining qualitative data that was then used to inform an online survey design. In the second phase, the online survey was distributed among an extensive sample of students in the U.S.A.

As stated above, we took this model and adapted it to the translation context, to our language (Spanish) and to a scale which was suitable for our study. We conducted two research group sessions, each one lasting approximately one hour, involving a total of five translation professionals. To recruit these people we searched the contacts of our LinkedIn network (which at that time totalled 315 contacts) and selected a total of 17 individuals that, as translation professionals, were suitable to our purpose and could contribute meaningfully to our focus group sessions. We contacted them individually through the email and invited them to take part in our sessions. After various email exchanges, aimed at checking their profiles, credentials and availability, a total of 5 individuals were selected. As professionals working in the translation sector, these individuals all met the criteria of our study. They also constituted a significant sample with varied profiles in terms of specialisation, work status, age, gender, experience and background.

The focus groups sessions followed the sequence described by Robson (2011: 280–284) and consisted of a fluent and semi-structured exchange of views among participants, conducted in a relaxed environment. Firstly, participants were asked about their background, specialisations, experience, languages, etc. Then, each of them described their way of approaching a translation brief: the tools they used, how they organised their work, the needs they had during the translation process, their relationships with other relevant agents (project managers, clients, other translators), etc. Finally, participants were asked whether they used Wikipedia, for what purposes and what perception they had of it.

As stated above, the qualitative information we obtained from these focus group sessions was used to adapt the model survey produced by Head and Eisenberg (2010) to our purpose. Additionally, in order to design our research plan, we carried out a thorough literature review and the main advances in the field of translation technologies were considered in a number of ways when adapting Head and Eisenberg’s (2010) survey to the context of professional translation. The following works and authors were helpful at different stages: to word questions about the subjects’ profile (years of experience, job, work situation, etc.) and their perceptions, or to document the tools, resources and translators’ needs to be considered in the survey. We found out that, in the field of the automation of translation, there has been a long debate around the way humans and machines should interact (Kay 1980; Melby 2006; García 2010). Moreover, the categorisation of systems ranging from 100% machine translation to 100% human translation has not always been clear and a number of authors have dealt with this issue at different times (Hutchins and Somers 1992; Trujillo 1999). Over the last decades, some researchers have analysed automated translation systems and their impact on translation, emphasizing on computer-assisted tools (Bowker 2002; Somers 2003; Olohan 2011), on machine translation and post-editing (Balling et al. 2012; Pym 2012; Alonso and de la Cova 2014, forthcoming) or on their training (Torres del Rey 2003; Calvo 2010), while others have been pioneers in foreseeing the effects that generic tools and connectivity (Internet) would have on the translation activity (O'Hagan 1996; Austermühl 2001; Cronin 2010). At an empirical level, there is a relatively long tradition of conducting surveys (based on ethnographic research, questionnaires, focus groups, or a combination of these techniques) with the purpose of documenting the impact of technologies on translation. Some authors have already studied the perceptions of translators towards machine translation and post-editing, outlining their low uptake among freelance translators (Fulford 2002), the professionals’ mixed feelings towards this technology (Guerberof Arenas 2013) or both things (Specia and Torres 2012). Deepening on translation memory (TM) systems, the survey conducted by Lagoudaki (2006) documented users’ evaluation of different TM tools; Dillon and Fraser (2006) stressed that TM tools were perceived more positively among younger translators than among experienced ones; the studies based on eye-tracking carried out by O’Brien (2006 and 2008) investigated translators’ cognitive interaction with this technology; and LeBlanc (2013), who conducted an ethnographic study in professional settings, outlined TM systems’ advantages and disadvantages from a translator’s perspective. Interestingly, organisations of the translation industry (LISA 2002; TAUS 2010; Kelly and Stewart 2011) and university research projects (eCoLoRe 2003) have also conducted their own surveys documenting the increasing presence of translation technologies in the professional context. Particularly relevant to this paper were the results of the survey conducted by Torres Dominguez (2012) that recently reported the first evidence of the presence of Wikipedia in a well-assorted translators’ toolbox, as well as the studies carried out by Fulford and Granell-Zafra (2005) and Désilets et al. (2009) that showed a widespread adoption of general-purpose software in the translators’ desktop.

During the pilot phase of our work, a draft survey was designed and distributed to a panel of judges made up of three experts (peer researchers with extensive experience in survey-based research). After incorporating their improvements, the survey, already in the online format, was distributed to another four translation professionals, together with a robustness test (aimed at determining the suitability of the survey: extension, difficulty, easy/difficult to understand, boring/enjoyable).
Once the pilot phase was completed, the final online version of the survey — designed with LimeSurvey software — was widely distributed among translators distribution lists, networks of professional translators, translation platforms (ProZ, etc.), translation vendors, freelance translators, university distribution lists, and translation-related social networks (Facebook, LinkedIn and Twitter).

The survey was written in Spanish and contained one introductory page outlining the filiation of the researcher and the general rationale for the study. The survey was given the general title of “The Translator in the Information Society” and, in the introduction, no mention was made directly to Wikipedia in order to avoid biasing the respondents’ answers in any manner.

The online survey was available from 14 October to 2 December 2013. The core of the survey consisted of 23 questions. The first five focused on the respondents’ professional profile (e.g. job, work situation, specialisations and years of experience). Only 3 questions in the survey were compulsory (question number 3, about years of experience; number 18, about the respondents' age; and 19, about their gender). However, the level of completion was significant and 302 subjects of a total of 412 respondents completed the survey in full. Questions 5 to 8 sought to find out how familiar respondents were with the Internet and computers in general, the needs they usually experienced when translating and which tools they used. The next nine questions were aimed at determining whether respondents used Wikipedia, how, for what purposes and how often. Our goal was also to elicit information on the respondents’ perceptions towards Wikipedia and on the possible existence of any kind of censorship or self-censorship pertaining to Wikipedia. The remaining five questions sought to define the respondents’ level of collaboration with Wikipedia and their personal profile (e.g. age, gender, country and languages). The last question was designed as an open field, where respondents could freely express their opinions about any topic mentioned in the questionnaire. A total of 65 respondents filled out that final field, commenting on Wikipedia, their work, the survey or simply wishing the researcher good luck.

In our analysis of the survey's results, we used IBM SPSS as a statistical tool for calculating frequencies and Microsoft Excel for data handling and presentation.

3. Preliminary findings

In this section we will examine our survey’s findings with regard to the respondents’ profiles, the needs and tools they dealt with when translating, and their perception and use of Wikipedia. As stated above, our work is interpretative by nature and it is not intended to be statistically representative of the translation industry as a whole. However, the results obtained may throw light on certain trends and, hopefully, stimulate debate.

3.1. Respondents profiles

A total of 75.1% of the respondents in this survey were female and only 24.9% were male. With regard to age, 63.2% of individuals in our sample were between 26 and 45 years old.
Since the survey was conducted in Spanish and distributed from a Spanish university, most respondents (51.4%) lived in Spain (see Figure 1) and reported having Spanish as their first mother tongue (62.9%) or their second mother tongue (40%). However, respondents from a total of 32 countries, with a total of 35 languages, were represented.

Figure 1. Responses to the question: ‘Country of residence’

A total of 99.2% of respondents reported having one foreign language, 80.7% had two, 45.8% had three and only 19.1% said they had four. English was the first foreign language of 67.6% of respondents, followed by Spanish (14.3%) and French (10.2%); French was the most popular second foreign language (41.6%), followed by English (19.6%), German (13.9%), Spanish (7.8%), Italian (5.7%) and Portuguese (4.4%).

In terms of job roles, 91.7% of the respondents reported working as translators. Table 1 shows the range of responses available to participants. More than one option was allowed.

Table 1. Responses to the question: ‘What do you do?’

Interestingly, only 21.4% of respondents who worked as translators gave it as their only task. The rest of them reported combining their activity as translators with other occupations (50.5% as reviewers, 30.2% as teachers, 18% as interpreters, 15% as audiovisual experts — subtitling, audiodescription, transcription —, 12.4% as managers, etc.). This suggests that professionals commonly referred to as translators typically carry out different tasks in their jobs.

As for their work situation, most respondents reported working as freelancers (64.3%), with most of them (70.6%) working solely as freelancers, i.e., they did not combine this occupation with any other. Table 2 shows participants’ answers regarding their occupation; in this case too, it was possible to choose more than one option.

Table 2. Responses to the question: ‘Which of the following statements most accurately describes your current work situation?’

Question number 3, where subjects were asked about their professional experience, was compulsory in the survey. According to the results, indicated in Table 3, a total of 69.2% of the respondents reported having more than 5 years of professional experience. We can therefore safely affirm that the majority of subjects in our sample had significant experience in the translation industry.

Table 3. Responses to the question: ‘How many years of professional experience do you have in the translation industry?’

To finish with the analysis of the respondents' profiles, the subjects reported working Almost always or Often (Figure 2) in the following specialisations: technical (46.1%), legal-economic/sworn (41%), marketing (33.1%), IT (29.7%), tourism (28%), scientific (26.2%), art and humanities (24.2%), social sciences (23.4%), biomedical (21.2%), literary (19.5%), audiovisual (17%) and videogames (8.3%) (see Figure 2).

Figure 2. Responses to the question ‘To which of the following categories, fields or specialisations are your services usually related?’

3.2. Use of tools and needs when translating

Respondents were asked how much their way of working had changed due to the use of Internet and other IT developments. In general, a total of 70.6% of the respondents considered that their way of working had changed Completely or A lot.

Since we acknowledge the ambiguity of the question, we refined the results according to the respondents’ years of experience. As shown in Figure 3, the more years of experience respondents had, the greater the change they perceived in their way of working.

Figure 3. Respondents' years of experience and the extent to which their way of working has changed

Respondents were asked how often they used Internet tools and resources when translating. As expected, the results suggest that the Internet is present in every translation process almost unfailingly regardless of the respondents’ job, since 94.5% of them said that they used it Almost always or Often.

The next question in this section was aimed at exploring the respondents' needs when translating (Figure 4). Some results, as expected, highlight the existence of terminological or lexicographical needs during the translation process (item ‘Find the meaning of a term’ ranked one; item ‘Find the use of a term or sentence in context’ ranked three; and item ‘Find previous translations of a term or sentence’ ranked five). However, others suggest new trends. For example, the importance respondents attached to general documentation needs (item ‘Find information about a topic’ ranked two and item ‘Find reliable sources for a topic’ ranked four), or to the visual dimension of words (item ‘View images associated to a term or idea’ ranked seven), appeared slightly ahead of translation needs that might have been expected to take greater priority, such as the ‘Use of translation memories’. In our view, it is also worth noticing that respondents also considered needs related to human resources and social interaction, i.e. item ‘Negotiate your translation decisions with clients, agencies or colleagues’ (ranked ten) and item ‘Consult with experts’ (ranked eleven). Less frequent needs were related to the use of library resources, mobile devices and creation of corpora.

Figure 4. Responses to the question: ‘How often do you experience the following needs when translating?'

Finally, respondents were asked about their use of certain tools or resources when translating (Figure 5). Subjects reported using – ‘Almost always’ or ‘Often’ – a wide range of resources (from greater to lesser frequency): Google (85.7%), bilingual and monolingual online dictionaries (82.8%), aligned or monolingual online corpora (62.4%), terminology databases (59.7%), image-search engines (55.4%), Wikipedia (53.6%), translation memory systems (53.2%), etc. As in the previous question, these results underline the importance given by respondents to documentation, terminological/lexicographical resources and visual resources.

Interestingly, a remarkable 39.1% of the participants reported using translation memory systems (TM) Always, which suggest a certain loyalty to this technology, i.e. those who work with TM tended to use it always.
In contrast to the importance currently attached to machine translation (MT) both by the industry and by academics, respondents reported low use of this technology. Only 14.4% used it ‘Almost always’ or ‘Often.’

As in the previous question, respondents gave consideration to human or social resources and reported resorting ‘Almost always’ or ‘Often’ to 'Translator colleagues' (44.6%), 'Subject experts' (34.3%), 'Translators' forums and mailing lists' (36.7%), 'Blogs about other topics' (19.9%) and 'Translation blogs' (19.8%) (see Figure 5).

Figure 5. Responses to the question: ‘How often do you tend to use the following tools or resources when translating?'

To conclude this section, according to the results above, we would like to point out the importance given to Wikipedia by respondents, since 83.2% of them said they used it ‘Almost always,’ ‘Often’ or ‘Sometimes when translating’; only 5.1% of respondents answered that they ‘Never used it,’ thus indicating that 94.9% of them used Wikipedia to some extent when translating.

3.3. Perception and use of Wikipedia

In the following paragraphs we will analyse how respondents use and perceive Wikipedia: how they usually access Wikipedia, for what purposes, opinions about its quality, overall evaluation (useful, reliable, easy to use) and rating of its features, their level of collaboration with Wikipedia as a project, and their censorship towards Wikipedia.

First, we will focus on the way subjects tended to access Wikipedia. In line with the aforementioned results stressing the prevalence of Google above any other tool, the majority of respondents (70.6%) reported accessing Wikipedia Almost always or Often from search page results, with only 29.9% of them directly searching Wikipedia.

Respondents answered that the used Wikipedia for a variety of purposes (Figure 6), emphasising, as in previously mentioned results, the importance of documentation, terminological/lexicographical and visual aspects.

Stressing the multi-purpose dimension of Wikipedia in the translation process, participants from the focus group sessions also alluded to Wikipedia not only as an encyclopaedic resource, but as a tool with the ability to foster a myriad of other uses; for example, Wikipedia as a comprehensive repository of visual knowledge, a multilingual corpus or dictionary, and a global tool that enables consensus with clients and colleagues:

Localiser: […] For that purpose I too normally use Google Images more, but the good thing about Wikipedia is that if you search for something in English, I don’t know, if the source text is in English, you can see the image related to the text and that helps me, because I can see the image and the explanation. So that helps me to understand: “Ah yeah!, so that’s what it means.”
Interviewer: You just mentioned, [Translator-trainer], that you switch from an article in one language to a version in another language. Do you do that within the same article?
Translator-trainer: Yes, yes, yes, in Wikipedia, that’s what I do in Wikipedia, to see if what I’m searching for is really what I think I’m searching for. So, yes.
Project Manager: When you tell a client “It [a term] appears in Wikipedia,” they say “OK!;” they respect it.

The results related to the purpose ‘To find the meaning of cultural references,’ ranked two (Figure 6), deserve particular focus because, in our opinion, they highlight the cultural dimension attributed to Wikipedia by respondents. The high participation rate in this question, ranging between 85.4% and 89.0%, is also worthy of comment: it suggests that a high number of respondents did use Wikipedia. To illustrate this, the percentage of respondents that answered that they ‘Never’ used Wikipedia in the most popular purpose, 'To find information about a topic,’ was only 5.4%.

Figure 6. Responses to the question: ‘How often do you use Wikipedia for the following purposes?'

The next question in the survey was aimed at exploring some users’ concerns, remarks and observations about Wikipedia that had been compiled during the qualitative research (focus groups with a total of five translators) and the pilot phase of the survey (four translators), conducted prior to the launch of the final online version of the survey. More specifically, some individuals from the focus groups and the pilot phase had remarked on the quality of Wikipedia, and mentioned some sort of censorship or self-censorship related to the online encyclopaedia. For example, a senior translator and a translator-trainer from the focus groups, in spite of having said that they used Wikipedia, made the following comments:

Senior translator: When you see that university students quote Wikipedia in their little essays and so on, it's simply hair-raising. You mustn’t quote it. Wikipedia doesn't have the academic status to be a source of reference.
Translator-trainer: […] I don't even quote Wikipedia. I don't ever quote.

As shown in Figure 7, most respondents agreed (‘Strongly agree’ or ‘Somewhat agree’) with statements 1-5 about the quality of Wikipedia. This depicts a situation where respondents use Wikipedia to get a general idea, but they usually compare the information it provides with other sources; in addition, participants acknowledged that the quality of Wikipedia varies from one language to another, but they felt confident of their own ability to detect whether an article was good or bad, an original text or a translation. It is worth pointing out that this behaviour was also reported by some of the participants in the focus groups:

Translator-trainer: What I do with Wikipedia results is look elsewhere and find information about what I found in Wikipedia, to corroborate it.
In-house translator: Yes, it’s true that sometimes, not always, but sometimes in the Spanish version, when I’m trying to find a term that appears in the English version, it turns out that you can’t find it. This is because the translation is a summary, or they have simply written the essentials about a term or whatever.
Interviewer: In your case, [Project Manager], do you have any criteria for assessing the reliability of a [Wikipedia] article?
Project Manager: It’s absolutely intuitive. You can often see, just by intuition, that something is not written coherently.

In line with these statements, most respondents disagreed (‘Somewhat disagree’ or ‘Strongly disagree’) with statement 6, ‘I trust the information contained in Wikipedia and I do not usually corroborate it,’ which reinforces the results of statements 1–5.

Regarding the statements about censorship and self-censorship when using Wikipedia, most respondents said they disagreed (‘Somewhat disagree’ or ‘Strongly disagree’) with statements 7–10, i.e. they disagreed with the idea that people, translators or students should not use Wikipedia (Figure 7).

Figure 7. Responses to the question: ‘How strongly do you agree or disagree with the following statements?'

The next three questions in the survey were aimed at determining the respondents' overall evaluation of Wikipedia in terms of its usefulness when translating, its reliability and its ease of use.

As illustrated below, respondents’ perception was good, since they considered that Wikipedia was ‘Very useful’ (18.2%) or ‘Useful’ (61.4%) when translating (Figure 8); ‘Very reliable’ (57.5%) or ‘Reliable’ (40%) (Figure 9); and ‘Very easy to use’ (58%) or ‘Easy to use’ (40.1%) (Figure 10).

Figure 8. Responses to the question: 'How useful do you find Wikipedia when translating?'

Figure 9. Responses to the question: 'How reliable do you think Wikipedia is as a resource?'

Figure 10. Responses to the question:
'How easy do you think Wikipedia is to use?'

More specifically, respondents were asked about how positive or negative a series of Wikipedia features were for translators (Table 4). Most features were considered clearly as positive: free, available on the Internet, covering many topics, free license, in many languages, allows linking concepts, states references, contains cultural information, images and updated information. However, the last two features are worthy of further analysis. With respect to feature 11, ‘It sometimes contains information that it is not possible to find anywhere else,’ we would draw attention to the relatively low participation rate in comparison with the other features. This may possibly reflect some kind of discomfort or non-conformity of respondents with the existence of such a feature. Feature 12, ‘Different people can take part in the drafting or translation of articles,’ referring to the collaborative nature of Wikipedia, was considered by most respondents as a feature that was Negative in some respects (29.7%), ‘Neither positive nor negative’ (28.9%) or ‘Positive in some respects’ (23.5%). These figures may suggest that respondents are not sure about exactly what this feature means. Another possibility is that respondents might have different views about collaborative authoring, some positive (maybe considering benefits sometimes attributed to Wikipedia, such as peer-reviews, neutral points of view or intersubjectivity), some negative (for example, lack of control over the content, the serendipitous nature of editing and revising Wikipedia, etc.) and some neither positive nor negative (it does not make any difference if an article is written by one or by many authors).

Table 4. Responses to the question: 'Please indicate how positive or negative you think the following Wikipedia features are for translators.'

Finally, respondents were asked how often they let other people know that they used Wikipedia in their work (Figure 11). This question was aimed at exploring the existence of any kind of censorship about Wikipedia. The fact that the response rate (302–325/412) was lower than in previous questions could be due to the fact that, as can be seen from Figure 5, the percentage of respondents that had reported ever using Wikipedia at some time was not the 100% but 94.9%. In any case, we acknowledge that the phrasing of the sentence could be tricky or that the question was somehow controversial. With this in mind, we included a final open question at the end of the survey: ‘If you wish, in this space you can freely express your opinion about any of the topics mentioned in the questionnaire.’ Interestingly, 6 out of the 65 respondents that filled out that open field referred to this question, pointing out either that they did not use Wikipedia or that they had never been in a situation to let others know that they used it.

Figure 11. Responses to the question: 'How often do you let other people know that you use Wikipedia in your work?' (302-325/412)

Notwithstanding the above, respondents answered that they let others know (Almost always or Often) that they used Wikipedia as follows (from greater to lesser frequency): family and friends (64.1%), translators and reviewers (47.4%), PMs and superiors (36.4%), and their clients (31.3%). We think these figures might indicate the existence of censorship or self-censorship with regard to Wikipedia, since respondents do not always let others know that they use it. These figures might also suggest that the more professional the environment, the lesser respondents are willing to acknowledge that they use Wikipedia. Self-censorship is particularly evident if we relate these figures to other results of the survey, Figures 8‑14, which show that most respondents use Wikipedia and have an overall good perception of it.

Respondents were asked whether they participated in Wikipedia in any manner. The results shown in Table 5 indicate that, in general, subjects in the sample group had a very low level of participation in Wikipedia. Over 90% of respondents reported that they ‘Never’ write, translate or edit translations of Wikipedia articles, or take part in decision-making or administration of Wikipedia. Only items 5, ‘I donate to Wikipedia,’ and 6, ‘I add information to Wikipedia articles or revise parts of them in my mother tongue,’ suggest slight participation by respondents.

Table 5. Responses to the question:
Do you collaborate with Wikipedia in any way?'

4. Discussion

Apart from offering an insight into the way translation professionals work and their needs, tools and resources when translating, the survey examined in this paper also explores how these professionals use and perceive Wikipedia. In previous sections we mentioned some of the survey’s shortcomings, such as its language/country bias and its limited scope. In no way, therefore, do our results claim to represent the whole industry. Notwithstanding, we think our survey has a qualitative value, with the added merit of documenting a reality that has to date been studied very little. With this descriptive approach in mind, we will try to identify certain trends.

As pointed out by Fulford and Granell-Zafra (2005), Lagoudaki (2006) and Désilets et al. (2009), the translator’s toolbox would include an increasing and varied assortment of tools. The most popular tools/resources used by respondents to meet the above described (documentation, terminological/lexicographical and visual) needs, would almost invariably include Google, online dictionaries/corpora, terminology databases and image-search engines in the translators’ toolbox. The importance attached by the respondents in our survey to human-driven resources, such as translator colleagues, subject experts, translators’ forums, mailing lists and blogs is particularly noteworthy. In contrast, the need to use translation memories, while certainly important, would not seem to be pressing at all; although our results also suggest that respondents who use TMs would remain loyal to that technology and would use it always. In any case, this seems to concur with the findings of the 2011 Common Sense Advisory report: “Translation memory (TM) is still not broadly adopted” (Kelly and Stewart 2011: 58–59). The low acceptance attained by machine translation in this survey could be explained by the nature of our sample (mostly freelance translators). In this sense, previous works (Specia and Torres 2012, Guerberof Arenas 2013, and TAUS 2010, among others) have noted translators’ resistance to post-editing. This observation, far from constituting a firm conclusion, is only aimed at opening a new avenue of investigation that could be of interest to the increasing number of analysts carrying out research into post-editing.
Given our special focus on Wikipedia, we would draw attention to the popularity of this online encyclopaedia, ranked sixth in our survey among the most used tools and resources when translating. However, most respondents reported accessing it from search engine results rather than by searching directly from the Wikipedia web page, thus suggesting that using Wikipedia during the translation process is not usually part of a pre-planned strategy.

From the survey’s results it can be deduced that respondents used Wikipedia to solve their most immediate needs, i.e. those related to documentation and to terminological, lexicographical and visual aspects. As in Head and Eisenberg (2010), where most respondents said they went to Wikipedia to obtain background information or an overview of a topic (‘big picture’), most of our respondents said they used Wikipedia ‘To find information about a topic.’ This behaviour was also reported by subjects taking part in our focus groups who talked about Wikipedia “to get an overall picture,” while also stressing that “Wikipedia is the Google of encyclopaedias.” In the same vein, as stated by Head and Eisenberg (2010), students’ usage patterns proved that they were aware of Wikipedia’s limitation: “Overall, college students use Wikipedia. But, they do so knowing its limitation. They use Wikipedia just as most of us do — because it is a quick way to get started and it has some, but not deep, credibility” (Head and Eisenberg 2010).

Interestingly, the second-ranked use of Wikipedia was ‘To find the meaning of cultural references,’ which reinforces the cultural dimension of Wikipedia as a repository of human knowledge and relationships (Alonso 2012). The use of Wikipedia as a terminological or lexicographical resource was highlighted in our survey, as it was in that of Head and Eisenberg (2010), whose subjects reported using the online encyclopaedia to ‘Find the meaning of terms.’ We would also stress the perceived visual usefulness of Wikipedia for the respondents in our survey, who reported using it to ‘View the images associates with a term or idea.’

As stated above, respondents’ overall perception of Wikipedia was good or very good in terms of usefulness when translating, reliability and ease of use. Some of the survey’s results suggest a process in which respondents use Wikipedia to get a general idea, but usually compare the information it provides with other sources. The most positive features in Wikipedia according to respondents were that it is free (cost), is available on the Internet, covers many topics, has a free license, is available in many languages, makes it possible to link concepts, states references, contains cultural information, contains images and features updated information. In contrast, the results do not identify a clear position of respondents with respect to the collaborative authoring of Wikipedia. In the same vein, Head and Eisenberg (2010) concluded that college students use Wikipedia for a number of reasons (coverage, currency and comprehensibility).

Participants in our survey felt confident of their ability to evaluate the quality of articles (which, they considered, varied from one language to another) and to determine whether an article was an original text or a translation. This self-confidence to evaluate loosely-controlled resources was also observed by Désilets et al. (2009) during their ethnographic research. Professionals’ skill in evaluating the reliability of resources, together with the fact that the information obtained from Wikipedia is usually checked with other sources, seems to be one of the core factors that might differentiate professional translation from amateur translation.
With regard to the existence of censorship or self-censorship in Wikipedia usage, the results of our survey indicate that respondents did not agree with the idea that people, translators and students should not use Wikipedia. On the contrary, they would feel comfortable with the idea of not discouraging them from using Wikipedia. However, respondents showed some kind of discomfort with the idea of letting others know that they use Wikipedia, especially in contexts outside their family and friends, i.e. (from greater to lesser discomfort): clients, PMs and superiors, translators and reviewers.

In our view, this suggests that whereas Wikipedia is present in the translator’s toolbox and appreciated by translation professionals, there is still some kind of self-censorship or controversy concerning its use in professional contexts. This caution might have its origin in more traditional approaches to translation practice and in a certain academic reluctance to rely on loosely controlled resources.

Finally, the respondents in our survey reported having a very low level of participation in Wikipedia, i.e. they did not usually write, edit, revise, translate, manage or donate to Wikipedia.

In our opinion, it is worth comparing the profile of our respondents with those of the survey conducted by McDonough Dolmaya (2012) among Wikipedia volunteer translators. Whereas most individuals in that survey were male (84%), 35 years of age or younger (76%), had never worked as professional translators (68%), and only a low percentage of them (15%) were working in the language industry (McDonough Dolmaya 2012: 172–175), our sample was mostly made up of females (75.1%), older than 30 years (74%), working in the translation industry (as freelancers, in agencies, etc.), and with more than five years of professional experience (69.2%). In our view, the explanation for the fact that the participants in McDonough Dolmaya’s survey all worked as volunteer translators of Wikipedia, whereas in our survey participants had a very low level of collaboration with Wikipedia must be sought in the parameters defining our samples (gender, age and professional experience). Another consideration to be taken into account would be the emerging trend of translating Wikipedia articles as part of translators’ training (e.g. Cánovas and Samson 2008).

5. Closing remarks and further research

The following statements are presented as a means of opening up new avenues of research and are not intended to be considered as authoritative conclusions.

The conceptual framework on which we have built our proposal, the trans-humanisation of translation (Alonso and Calvo, forthcoming) — an approach that envisages translation workflows as extended, technology-mediated systems or networks with the ability of incorporating human translators and technologies in collective workflows — may provide a valid point of departure from which to further our understanding of translation activities today. However, more theoretical and empirical works addressing the fundamentals of trans-humanisation (the centrality of technological extensions incorporated by humans in the translation process, technology-enabled collaborative translation, and an instrumental rather than an artifactual approach to tools) would be needed in order to consolidate this theory in Translation Studies.

Probably very few people question the increasing use that translation professionals make of technology (both generic and translation-specific); but our results would also suggest that translation is still (to a large extent) a social activity where clients, experts, translator colleagues, forums, blogs, etc. all play a relevant role.

In the same vein, this survey’s results depict translation professionals as people who feel comfortable using Wikipedia (as consumers), but who do not always admit that they use it, especially to clients. This may illustrate the moving of Wikipedia from the periphery to the centre of the translation system, though some concerns remain and are manifested in the form of self-censorship. Interestingly, our respondents reported some rather unexpected uses of Wikipedia, emphasising the online encyclopaedia’s cultural and visual potential. This is a trend which, if confirmed in future research, could be of interest for CAT/MT developers, translation trainers and in a number of different disciplines (Translation Studies, Sociology, Information Sciences, etc.).

Nevertheless, we have found no evidence of any real interaction between Wikipedia and our respondents, who reported a very low level of participation in the online encyclopaedia (as producers). This may suggest that Wikipedia plays a role in the translation process, but whereas volunteers (mostly male and inexperienced as in McDonough Dolmaya’s 2012 study) might be producing Wikipedia articles, translation professionals (mostly female and experienced as in our survey) would be consuming these articles and somehow incorporating their content into the mainstream translation workflow.

Finally, we think that the present study — with its combined descriptive, qualitative and interpretative approach — provides a useful basis for further research, which could include experimental works analysing cognitive aspects of Wikipedia use (possibly incorporating eye-tracking and the recording of screen and brain activity); the development of innovative online documentation resources or tools specifically designed for translators (as in Oliver and Climent 2012 for instance); and new proposals for translator training syllabuses incorporating Wikipedia-based information and documentation skills.

Acknowledgements

The author gratefully acknowledges the funding provided by the Universidad Pablo de Olavide to support her project Technology and Media Assisting the Translator. I am grateful also to Dr Jorge Díaz-Cintas and his collaborators for welcoming me during my research stay at Imperial College London and for their assistance with the distribution of the questionnaire. My gratitude also goes to Dr Head and Dr Eisenberg, for providing me with their questionnaire, to Dr Elisa Calvo, Dr Gabriel Cepeda and Dr Marián Morón (panel of experts), to Dr Manuel A. Vázquez-Medel and Dr. Adrián Fuentes (my supervisors), and to Andrew Taylor (my proofreader). Finally, I would like to thank pilot-phase participants and survey participants for their input and all those who contributed to the distribution of the survey.

Bibliography

• Alonso, Elisa (2012). “Análisis del discurso en Wikipedia desde la Teoría de los Polisistemas.” Paper presented at the meeting Tendencias actuales en los estudios sobre el discurso, 15 November 2012. Sevilla: Universidad de Sevilla. http://www.youtube.com/watch?v=Bs5yRjG7owMandfeature=youtu.be (consulted 01.03.2014).
• Alonso, Elisa and Elisa Calvo (2012). “The ultimate turn: from translator to trans-translator.” Michael Boyden (ed.) (2012) EST Symposium "Same Place, Different Times". Book of Abstracts. Vienna: John Benjamins, 5.
• ─ (forthcoming). “Developing a Blueprint for a Technology-mediated Approach to Translation Studies.” Meta: Journal des traducteurs / Meta: Translators' Journal.
• Alonso, Elisa and Elena de la Cova (2014, forthcoming). “Apuntes metodológicos para la aplicación de la socionarrativa a la evaluación de herramientas de traducción: «Érase una vez Google Translator Toolkit».” Tradumàtica 12.
• Austermühl, Frank (2001). Electronic Tools for Translators. Manchester: St. Jerome.
• Balling, Laura Winther, Michael Carl and Arnt Lykke Jakobsen (eds) (2012). Proceedings of the International Workshop on Expertise in Translation and Post-editing - Research and Application, ETP|CBS 2012, Center for Research and Innovation in Translation and Translation Technology, Copenhagen Business School. Copenhagen, 17–18 August 2012. http://bridge.cbs.dk/events/ETP-2012-Proceedings.pdf (consulted 01.03.2014).
• Bowker, Lynne (2002). Computer-aided Translation Technology: A Practical Introduction. Ottawa: University of Ottawa Press.
• Buzelin, Hélène (2005). “Unexpected Allies: How Latour's Network Theory Could Complement Bourdieusian Analysis in Translation Studies.” The Translator 11(2), 193–218.
• Calvo, Elisa (2010). Análisis curricular de los Estudios de Traducción e Interpretación en España: perspectiva del estudiantado. PhD dissertation, Granada: Universidad de Granada. digibug.ugr.es/bitstream/10481/3488/1/1852574x.pdf (consulted 30.10.2014).
• Cánovas, Marcos and Richard Samson (2008). “Dos ejemplos de aplicación de software libre en la docencia de la traducción.” Óscar Díaz Fouces and Marta García González (eds) (2008). Traducir (con) software libre. Granada: Comares, 193–211.
• Chesterman, Andrew (2007). “Bridge Concepts in Translation Sociology.” Michaela Wolf and Alexandra Fukari (eds) (2007) Constructing a Sociology of Translation. Amsterdam/Philadelphia: John Benjamins, 171–186.
• Cronin, Michael (2010). “The Translation Crowd.” Revista Tradumatica 8, 1–7. http://www.raco.cat/index.php/Tradumatica/article/view/225900(consulted 01.03.2014).
• Désilets, Alain, Christiane Melançon, Geneviève Patenaude and Louise Brunette (2009). “How Translators Use Tools and Resources to Resolve Translation Problems: An Ethnographic Study.” MT Summit XII - Workshop: Beyond Translation Memories: New Tools for Translators MT, Ottawa, 29 August 2009. http://www.mt-archive.info/MTS-2009-Desilets-2.pdf (consulted 01.03.2014).
• Dillon, Sarah and Janet Fraser (2006). “Translators and TM: an investigation of translators’ perceptions of translation memory adoption.” Machine Translation 20(2), 67–79.
• eCoLoRe (2003). Translation Memory Survey 2003. http://ecolore.leeds.ac.uk/xml/project/news.xml?lang=en (consulted 01.03.2014).
• Fulford, Heather (2002). “Freelance translators and machine translation: An investigation of perceptions, uptake, experience and training needs.” Sixth EAMT Workshop "Teaching machine translation", 14-15 November 2002, UMIST, Manchester, 117–122, http://www.mt-archive.info/EAMT-2002-Fulford.pdf (consulted 01.03.2014).
• Fulford, Heather and Joaquín Granell-Zafra (2005). “Translation and Technology: a Study of UK Freelance Translators.” The Journal of Specialised Translation 4, 2–17. http://www.jostrans.org/issue04/art_fulford_zafra.pdf (consulted 01.03.2014).
• García, Ignacio (2010). “The Proper Place of Professionals (and Non-Professionals and Machines) in Web Translation.” Revista Tradumàtica 8, 1–7. http://www.raco.cat/index.php/Tradumatica/article/view/225898/307309(consulted 01.03.2014).
• Guerberof Arenas, Ana (2013). “What do professional translators think about post-editing?” The Journal of Specialised Translation 19, 75–95.
• Head, Alison J. (2013). “Project Information Literacy: What Can Be Learned about the Information-Seeking Behaviour of Today’s College Students?” Association of College & Research Libraries (ACRL) 2013 Proceedings. http://www.ala.org/acrl/acrl/conferences/2013/papers (consulted 28.10.2014)
• Head, Alison J. and Michael B. Eisenberg (2009). Lessons Learned: How College Students Seek Information in the Digital Age. Washington: The Information School, University of Washington. http://projectinfolit.org/images/pdfs/pil_fall2009_finalv_yr1_12_2009v2.pdf (consulted 28.10.2014).
• Head, Alison J. and Michael B. Eisenberg (2010). “How today's college students use Wikipedia for course related research.” First Monday 15(3) http://journals.uic.edu/ojs/index.php/fm/article/view/2830/2476(consulted 01.03.2014).
• Hutchins, John and Harold Somers (1992). An Introduction to Machine Translation. London/San Diego: Academic Press.
• Kay, Martin (1980). The proper place of men and machines in language translation. Palo Alto, CA: Xerox Palo Alto Research Center. [Reprinted in: Machine Translation (1997) 12 (1-2), 3–23] http://www.mt-archive.info/70/Kay-1980.pdf(consulted 01.03.2014).
• Kelly, Natally and Robert G. Stewart (2011). The Language Services Market: 2011. An Annual Review of the Translation, Localization, and Interpreting Services Industry. Lowell: Common Sense Advisory.
• Lagoudaki, Elina (2006). “Translation Memories Survey 2006. User’s perceptions around TM use.” Translation and the Computer 28, ASLIB, 1–29 http://mt-archive.info/Aslib-2006-Lagoudaki.pdf (consulted 01.03.2014).
• LeBlanc, Matthieu (2013). “Translators on translation memory (TM). Results of an ethnographic study in three translation services and agencies.” The International Journal for Translation and Interpreting Research 5(2), 1–13 www.trans-int.org/index.php/transint/article/download/228/134 (consulted 01.03.2014).
• LISA (2002). LISA 2002 Translation Memory Survey: Translation Memory and Translation Memory Standards http://www.tradulex.org/Bibliography/LISATMsurveyX.pdf (consulted 01.03.2014).
• McDonough Dolmaya, Julie (2012). “Analyzing the Crowdsourcing Model and Its Impact on Public Perceptions of Translation.” The Translator 18, 167–191.
• Melby, Alan (2006). “MT+TM+QA: The Future is Ours.” Revista Tradumàtica 4, 1–6. http://www.fti.uab.es/tradumatica/revista/num4/articles/04/04.pdf(consulted 01.03.2014).
• O’Brien, Sharon (2006). “Eye Tracking and Translation Memory Matches.” Perspectives: Studies in Translatology 14(3), 185–205.
• ─ (2008). “Processing Fuzzy Matches in Translation Memory Tools: An Eye-Tracking Analysis”. Susanne Göpferich, Arnt Lykke Jakobsen and Inger Mees(eds)(2008). Looking at Eyes. Eye Tracking Studies of Reading and Translation Processing. Copenhagen: Samfundslitteratur, 79–102.
• O'Hagan, Minako (1996). The Coming Industry of Teletranslation. Clevendon: Multilingual Matters.
• Oliver, Antoni and Salvador Climent (2012). “Using Wikipedia to develop language resources: WordNet 3.0 in Catalan and Spanish.” Digithum 14, 68–76.
• Olohan, Maeve (2011). “Translators and translation technology: The dance of agency.” Translation Studies 4(3), 342–357.
• Pym, Anthony (2011). “What Technology Does to Translating.” The International Journal for Translation and Interpreting Research 3(1), 1–9. http://trans-int.org/index.php/transint/article/view/121/81 (consulted 01.03.2014).
• ─ (2012). “Translation Skill-sets in a Machine-translation”. 16th Symposium on Interpreting and Translation Teaching, Fu Jen Catholic University, Taiwan. http://usuaris.tinet.cat/apym/on-line/training/2012_competence_pym.pdf (consulted 01.03.2014).
• Risku, Hanna and Florian Windhager (2013). “Extended Translation: A Sociocognitive Research Agenda.” Target 25(1), 34–46.
• Robson, Colin (2011). Real World Research. Chichester: Wiley.
• Somers, Harold (ed) (2003). Computers and Translation. Amsterdam/Philadelphia: John Benjamins.
• Specia, Lucia and Ruth Torres (2012). “Post-Editing practice – a survey of translators' preferences and concerns.” Proceedings of the International Workshop on Expertise in Translation and Post-editing - Research and Application, ETP|CBS 2012, Center for Research and Innovation in Translation and Translation Technology, Copenhagen Business School, 34–35. Copenhagen, Denmark, 17–18 August 2012.
• TAUS (2010). “Postediting in Practice. A TAUS Report.” https://www.taus.net/reports/postediting-in-practice (consulted 01.03.2014).
• Torres del Rey, Jesús (2003). Nuevas tecnologías y enseñanza de la traducción: límites y posibilidades de los modelos de aplicación. PhD Thesis. Salamanca: Universidad de Salamanca. http://web.usal.es/~jtorres/investigacion/jtorresdelrey.pdf (consulted 01.03.2014).
• Torres Domínguez, Ruth (2012). “2012 Use Of Translation Technologies Survey.” Mozgorilla. http://mozgorilla.com/en/texnologii-en-en/translation-technologies-survey-results/ (consulted 01.03.2014).
• Trujillo, Arturo (1999). Translation Engines: Techniques for Machine Translation. London: Springer.
• Wolf, Michaela (2007). “Introduction: The Emergence of a Sociology Translation.” Michaela Wolf and Alexandra Fukari (eds) (2007) Constructing a Sociology of Translation. Amsterdam/Philadelphia: John Benjamins, 1–36.

Biography

Elisa Alonso holds an MA in Literature, Communication and Advertising from Universidad de Sevilla. Elisa has over 10 years of experience in the localisation industry and as a translators’ trainer. Currently, Elisa is a full-time researcher and lecturer at Universidad Pablo de Olavide (Spain), where she is working on her doctoral thesis about translation technologies.

E-mail: elialonso@upo.es.

Home | Past Issues | Keyword Search | Editorial Board | Events | About | Contact | Help