Introduction

Learning emerged as a coping mechanism during months-long lockdowns of 2020. The COVID-19 crisis has acted as a catalyst in the growth of the remote learning industry. It has also led to a “silent explosion” of informal online learning in online settings. Consequently, the role of social networking sites in this matter has increased (Käpplinger and Lichte, 2020: 784). In these unprecedented times, the translation and interpreting (T&I) professional community has also come together to start a wide variety of formal and informal online professional development initiatives. However, for the interpreting professionals, this was not just a coping mechanism, but a collective effort to battle the new realities of work, as the pandemic triggered an increased demand in various forms of distance interpreting (DI), and remote interpreting (RI) in particular.

Despite being one of the oldest occupations known to mankind, interpreting owes its status as a profession to technology (Ball, 2021). Since the 1920s, a conference interpreter working in the booth has acted as an invisible link in the communication channel, a built-in part in the mechanism of simultaneous interpreting (Chernov, 1978). Satellite and Internet technology of the 1990s brought about a variety of DI scenarios, including videoconference and telephonic interpreting. Since then technical standards have improved, and the undeniable benefits of DI, such as reducing travel costs and increasing availability of interpreters, have led to its gradual adoption in business and community settings, including legal and medical interpreting (Ball, 2021; Braun, 2019). In conference settings, till recent times professional interpreters rarely worked with online platforms due to a number of risks associated with the technical side of the process, as well as a lack of health and safety regulations for RSI. Nevertheless, experimental studies on video remote interpreting (VRI) were conducted as early as 1976[1].

In 2016, Common Sense Advisory (CSA Research) introduced the term interpreting delivery platforms to describe applications offering interpreter service “at the push of a button”, including over-the-phone interpreting (OPI), VRI, RSI, and machine interpreting (MI) platforms[2]. However, till 2020, virtual interpreting technology (VIT) has existed as “a hype solution without a problem”[3].

In professional discussions, the recent rapid adoption of VIT has been compared to the early years of Translation Memory (TM) technology: eventually, non-believers either learned the technology or were replaced by tech-savvier colleagues who were ready to face the challenge[4]. Interestingly, the results of the early 2000s studies of TM application sound very similar to survey results of today – novice translators generally have a more positive view of TM irrespective of actual use, and perceived IT proficiency of translators plays a key role in translators’ perceptions of the benefits of TM (see, e.g. (Dillon and Fraser, 2006) and (Corpas Pastor, Gaber, 2020)). However, unlike the translator community, the interpreter community did not have the luxury of the gradual adoption of computer-assisted/virtual interpreting tools. The pandemic hit at the time when “(t)he interpreting profession was on the edge of a cliff waiting for the right breeze to come along so we could safely test our new technological wings and then quickly fly back to safety”[5].

The number of challenges the profession has faced in the COVID-19 crisis is unprecedented. “We entered the second half of the worst year in the history of our profession, and we did so full of uncertainty,” interpreter Tony Rosado shares in his blog[6]. The pandemic has triggered a massive shift for virtual event systems and consequently for RSI platforms and VIT has now been proclaimed “the solution to the problem”[7]. “Whatever happens in the long run, immediate needs mean selecting a solution now,” Hélène Pielmeier, senior analyst, CSA Research, commented on the boom of RSI capabilities[8]. The onset of the COVID-19 crisis made some members of the interpreting community homebound and without income. For individual interpreters, working in extremis meant finding the right equipment to work from home, adapting their working conditions to the new reality of lockdown, battling technostress and technology-induced fatigue, avoiding injuries, and educating clients on the new realities of remote work[9]. Therefore some were forced to dive deep into work here and now using remote solutions at hand. These topics inevitably dominated the discussions among the members of thematic/professional communities across a variety of online social media (OSM) platforms.

This paper contributes to the discussion on the adoption of DI practices from the perspective of translation discourse studies (Novikova, Naumova, 2021), aided by the input from OSM and user-generated content (UGC) studies. Since this is a companion study to a project examining video-mediated interpreting into Russian, the aim of the paper is of a purely descriptive nature. It seeks to describe key themes, including issues and practices, discussed in relation to an emergency shift to technology-mediated interpreting.

The reported study was conducted from January 2021 to August 2021. It draws on a DIY corpus of UGC generated over the course of 2020 by the members of the relevant T&I-related Facebook* groups[10]. It employs corpus techniques to describe patterns in the English language OSM discourse on DI, and applies content analysis to reflect on the emergent tendencies. Although a comparable Russian sub-corpus was also compiled, this paper focuses on the English language part of the sample.

Theoretical background, material and methods

Scholars of a variety of disciplinary backgrounds have recognized the importance of OSM. Given their tremendous popularity, OSM give social sciences a new type of data for examining social behaviour. The Internet technology, including OSM and UGC platforms, such as YouTube, Twitter, and VK have given rise to new communities of practice.

Long before the advent of OSM platforms, social networks have been instrumental in the professional development of translators and interpreters (Garbovsky, Kostikova, 2015). As Robinson notes, from a translator’s internal perspective, networking is part of the activity that aims at the production of text (Robinson, 2012: 146). Therefore social networks, in general, are among the factors that control and influence the activity of translation, determining the “success” of translation (Ibid.). McDonough defines a translation network as a network in which the actors “share a common interest in translation or a translation-related profession (e.g. interpreting or terminology)” (McDonough, 2007: 794). Translators being involved in a variety of social networks (including professional ones) might be of little concern to translation users, but for translators, involvement in the profession is of utmost importance (Robinson, 2012: 26).

From email distribution lists in the early Internet years to the social media of today, translators have embraced multimedia communication channels to share resources and solve problems. As online communication becomes a component in a translator’s set of tools, the cognitive process of translation acquires a social dimension (Folaron, Buzelin, 2007: 634). When exposed to OSM environments, the interpreting profession acquires a different set of social tools as well: connected via OSM groups, language professionals discuss issues ranging from training and note-taking to research and networking (Downie, 2016: VII). The interpreting community is moving “from being a rare and rarefied elite to becoming a mutually supportive group of professionals with contacts all over the world” (Ibid.). For translation and interpreting scholars, discussion forums (McDonough, 2007; Novikova, Naumova, 2021), distribution lists, and blogs (McDonough Dolmaya, 2011; McDonough Dolmaya, 2018; Dam, 2013) have been a source of insight into the profession: research into the online networking practices has illustrated how translators socialize, collectively discuss, and solve problems; it also laid the groundwork for the future process-oriented studies and shed the light on the meaningful changes in the profession (McDonough Dolmaya, 2011: 81).

OSM content can serve as an economic and realistic alternative to Think Aloud Protocols and video recordings (Folaron, Buzelin, 2007; Desjardins, 2017). Therefore, patterns in the OSM discourse of the interpreting community may potentially reveal various directions for further investigation. OSM research is meant to “bridge” the gap between theorists and practitioners (Desjardins, 2017: 101).

At the moment, OSM are home to a large variety of specific T&I-oriented groups, from small private spaces to groups gathering tens of thousands of people interested in this field. The author of this paper has been a member of some of the relevant T&I groups for a number of years. Table 1 presents examples of Russian and English-language OSM groups that can be found using search terms interpreting, translation, interpreter, translator, terp, перевод, переводчик. Most of the groups do not specify the country of origin, unless it is included in the title or description box.

Table 1. Examples of T&I thematic OSM groups (January 2021)

Таблица 1. Примеры переводческих тематических сообществ в социальных сетях (январь 2021 года)

This case study was conducted from January to August 2021, and started with netnographic observation of a variety of the groups related to T&I practice and training. Among relevant groups reviewed, only 14 met the inclusion criteria, i.e. having an active status from 1 January 2020 to 1 January 2021, and the presence of relevant search terms in the posts and comments.

The nature of OSM communities requires to embrace ethical considerations of UGC research. As in many other fields, public groups represent just a small portion of active professional T&I online communities. The majority of the groups of interest in our case are marked as private, which means that a user is required to ask an administrator’s permission to join the group to be able to post and comment. Although some of the private groups include tens of thousands of users, it is important to consider their members’ views on the privacy of UGC. Traditionally used in social research, the principle of informed consent becomes problematic when applied to Internet research (Markham, Buchanan, 2017: 204). In the type of study that we are undertaking, it is almost impossible to obtain consent from all the authors of UGC generated in a certain timeframe. Advocating for a case-based approach, McKee and Porter introduce a pragmatic framework that includes a number of dimensions to consider in online research (McKee, Porter, 2009: 97). Among them are “public vs private, data ID, degree of interaction, topic sensitivity, and subject vulnerability” (Ibid.). AoIR 2019 principles also list data anonymization as one of the possible solutions for risks mitigation in Internet-based research[11].

Since the close analysis of the group members’ characteristics and interactions was not the focus of the study, no personal names or any other kind of proper names needed to be collected in the process. Methods involved in the study do not allow identifying personal opinions of the authors; instead, they help to paint a general picture of the themes in the discourse. The study does not deal with sensitive data and is not likely to influence the public image or perception neither of the specific group, nor the professional community. Email addresses, links, dates, and names present in some advertisements posted in the groups were excluded from the sample later in the clean-up process. Therefore the potential risks of this case study may be considered low, and authors of the collected posts were not approached; instead, the administrators of private groups were contacted via a built-in messenger feature to disclose the aims of the study and asked for permission to use the anonymized posts and comments for research purposes. As a result, it wasn’t possible to include some of the relevant groups in the sample, among which were the major Russian-language T&I communities. When investigating groups that were marked as public, the administrators’ permission or group consent was not sought.

After the initial observation, the social network’s built-in search engine was used to retrieve UGC published in 2020. The following search terms were selected to identify the relevant posts and comments: remote interpreting, distance interpreting, RSI, VRI, consec, consecutive, zoom, устн*, послед*, удаленн* (later in the course of the study, search terms virtual, синхрон* and онлайн* were also found relevant, although they are not present in the results reported below).

The terminology used to describe VIT is not yet standardized. Braun uses distance interpreting as an umbrella term to refer to any type of technology-mediated interpreting (Braun, 2019). The established modalities of DI are audio-mediated and video-mediated. Among the configurations of DI are remote interpreting (interpreter is located remotely) and teleconference interpreting (clients at different sites, interpreter is either co-located or at a separate site). Both configurations have been used for simultaneous (RSI) and consecutive modes. Although the umbrella term distance interpreting is used in many academic papers and in AIIC 2019 Guidelines[12], the reviewed UGC reveals that this term is rarely found in the group discussions. Remote interpreting, referring to both configurations, was the search term that returned the most results.

The search results were reviewed manually, and the relevant posts and comments were collected as plain text files for further quantitative and qualitative analysis. Posts that consisted of only a photo, a video or a link were not considered. Cross-postings among different groups were counted as one post. All in all, 361 files were collected. This resulted in a small purpose-built corpus comprising 65,457 tokens in English (and 32,531 in Russian). This paper reports on the analysis of a sample which includes the UGC from the following English language groups: Interpreters, Translators & Language Professionals; Continuous Training (CPD) & Events for Translators and Interpreters; Translators and Interpreters (ProZ.com); Interpreter Technology Group; Interpreters’ Practice Group – Group by Sarah Tiemann; Interpreter trainers blundering through online teaching in a pandemic.

Collected files were imported into desktop-based corpus managers[13] Antconc, #Lancsbox, and Taguette, which was used for manual coding. Antconc and #Lancsbox were used for the generation of frequency and keyword lists, as well as examining collocations in the corpus. Coding was done based on the analysis of frequency lists and collocations found in the posts and comments. However, the more emergent themes and scenarios related to DI and its perception among interpreters were revealed in the course of content analysis. In this paper, the results of content analysis are presented as the backdrop for the case study.

Results and discussion

The first step of the analysis was generating word frequency lists, as well as keyword lists using #LancsBox, which includes the TreeTagger feature for lemmatization and part-of-speech tagging. BNC spoken and written English (combined) frequency list[14] was used as a reference. Table 2 presents the most frequent content words and keywords in the corpus. They are very domain-specific, which reflects the compilation criteria.

Table2. Top 20 most frequent words and keywords in the corpus

Таблица 2. Топ-20 частотных единиц и ключевых слов в корпусе

Among the top 100 keywords are DI types, including RSI, VRI, and OPI (over-the-phone interpreting). Another category is platform names, such as Zoom, Webex, Kudo and Voiceboxer. A prominent theme is interpreting technology (headphones, laptop, sound, platform, webcam, phone, USB); also present are the terms reflecting the setting (meeting, client) and the situation (pandemic, covid, masks). Verbs need, can, should often associated with sharing experience, giving advice and recommendations also stand out in the lists (you need... a second device, a screen behind you, a phone and a telephone line, a PC, a headphone with a mic, a camera, a switching box). Gratitude formulae and politeness markers are found at the top of the list sorted by keyness (thank, please).

To further explore the semantic categories present in the keyword and frequency lists, we extracted collocations from the corpus. Collocation graphs and networks are analytical tools that can potentially provide insight into meaningful lexical connections that reflect meaning relations in discourse that cannot be identified otherwise (Brezina, McEnery, Wattam, 2015: 164-165). They represent an abstraction based on the examples of word co-occurrence and therefore have to be interpreted using the concordance feature (Brezina, 2018: 69). This type of analysis captures complex discursive processes in a single image (Brezina, McEnery, Wattam, 2015). Unlike word maps in a thesaurus, collocation networks are reflecting associations in discourse, not a dictionary, as they are oriented towards syntagmatic relations (Brezina, 2018: 61). Among the limitations of this study is the size of the corpus. Although small corpora can be used in linguistic studies (Ghadessy, Henry, Roseberry, 2001), we approach the results with caution and as an exploratory aid that should be interpreted in combination with concordance and content analysis. In collocation networks presented below, mutual information score (MI) was used as an association measure. MI statistic is not dependent on the size of the corpus (Hunston, 2002) and has been traditionally applied in discourse analysis to highlight rare and unusual combinations that are not necessarily representative of the discourse (Brezina, McEnery, Wattam, 2015). Collocates closest to the node have the strongest association according to MI-score, a darker colour shows the frequency of co-occurrence for individual collocations (node in the centre and collocate) (Brezina, 2018).

Figure 1. On the left – Collocation graph of the search term interpreting. On the right – Collocation network of the search terms RSI, OPI, and VRI

Рисунок 1. Слева – визуализация коллокаций единицы interpreting. Справа – единиц RSI, OPI и VRI

Fig. 1 presents a collocation graph[15] for the search term interpreting and a collocation network for the terms VRI, OPI, and RSI. Concordance-based examination of the interpreting collocates opens up directions for further investigation. Contexts with the verb feel (e.g., 1, 2, 3) are of a particular interest for the analysis of attitudes to DI in the community. Among them are the uses of a set expression feelfreeto (…share, takealook, to suggest ideas) which engages the members in sharing of information, and makes them join other activities in the group. Another context is sharing emotions and feelings, usually related to professional contexts, such as to feel supported, happy, differently, happy, tired, useless.

(1) …Love these sessions, great way to share ideas and experiences. And also feelpart of a community. Here’s the link if you missed it – feelfree to send it to your clients before meetings!

(2) Feel free to make a post asking for individuals who might want to practice with you…

(3) The language-specific practice groups meet live to actually practice interpreting. Feelfree to make a post asking for individuals who might want to practice with you on a smaller scale rather than a bigger group.

Compared to onsite interpreting, distance solutions cause more stress and fatigue in various DI scenarios (Braun, 2019). As opposed to VRI and OPI solutions, RSI platforms have not gained much trust in pre-pandemic years. Due to reduced sensory output and other factors, RSI is often considered a more stressful activity than onsite interpreting. However as remote work has become a reality, it required adaptation and learning not only on the interpreters’ part but also on the part of clients and project managers. Collocates in the middle of the VRI–OPI–RSI network (Fig. 1) are shared by the three nodes. Interestingly, VRI and OPI share collocates lower and cent, which is indicative of the discussion on the changes in the interpreters’ rates during the pandemic crisis. Zoom, which is the name of the web conferencing platform offering an RSI solution, is associated with both RSI and VRI. Popular, experience, anyone, and everyone are the collocates that RSI does not share with the other two modes of interpreting. Concordance lines for these terms reflect the demand for RSI solutions during the lockdown period, the lack of RSI expertise among the groups’ members and their willingness to learn from their peers.

In the graphs (Fig. 1–3), we focus only on nouns, verbs, adjectives, and adverbs to see the main conceptual connections. However, grammatical words sometimes can give an even more nuanced picture. For example, the word how as a collocate of remote and interpreting is often present in the contexts where a user either shares his/her instrumental knowledge or asks for advice about a certain topic (usually a tech-related one) (4, 5).

(4) Can anyone who’s used the feature let me know… how to go about it?

(5) Colleagues and I ran a few tests in the recent weeks – it works fine but you have to leave some time at the beginning of the talk to explain to the participants how the function works.

The collocation network for the terms remote and distance in Fig. 2 shows how a distant collocate safe shares the conceptual space with remote (interpreting), and therefore underlines the ultimate reason for this emergency shift to DI technologies, which is a global effort to contain the pandemic.

Figure 2. Collocation network of the search terms remote and distance

Рисунок 2. Визуализация коллокаций единиц remote и distance

The adjective good is often used in the contexts discussing essential tech requirements of DI (goodsoundquality, goodInternetconnection), which is logical given tech is one of the key themes in the discourse. However its collocation network (Fig. 3) adds a new semantic dimension to the results of this thematic analysis (goodjob, goodluck). This is how B. Slaughter Olsen reflected on the process of transition in the interpreter community[16]:

(6) “Interpreters everywhere were initially asking: ‘Can I work from home?’ and ‘Should I work from home?’ But those questions gave way to ‘How do I work from home?’ because for the vast majority of interpreters it was the only possible way to work. It has been heartening to see so many interpreters tackle this challenge head-on. From webinars to YouTube videos and from online practice groups to technology demos, forward-thinking interpreters are helping their colleagues transition to this new reality that few foresaw. These initiatives are extremely valuable and should be encouraged and supported.”

This quotation mirrors the results of the thematic analysis of the corpus. The network of share–thank–good in Fig. 3 paints a picture of a supportive collective, grateful to its members for their contribution and the overall wellbeing of the community (7–10).

Figure 3. Collocation network of the search terms thank*, good and share

Рисунок 3. Визуализация коллокаций единиц thank*, good и share

(7) Remote is always a topic where people have a lot to say. After looking at the evidence and spotting trends, I came up with my own, personal position // Thanks for the interesting reflections

(8) Would it be possible to have a chat with you as you seem to have a good deal of experience in remote interpreting and I would like to have more info on the technological solutions and your insight. Thanks

(9) …sorry I can’t be more helpful. I admire all of you freelancers. Best of luck

(10) …I checked the course. Good luck to all who wants to participate or may be even refresh amid the aftermaths of COVID19 and other weirdo global opportunities.

Table 3 presents the results of semi-open coding done for UGC featuring remote and distance (n=117) search terms. As it could be argued based on the collocation network and frequency list analysis, sharing experience and asking for advice about DI practices (especially in tech-related issues) are the broad categories that are the most prominent in the sample.

Table 3. Categories in semi-open coding

Таблица 3. Категории контент-анализа (открытое кодирование с акцентом на ряд ранее выделенных тем)

Interestingly, categories found in the UGC, as well as their distribution, largely resemble the list of themes in translation blogs reviewed in (McDonough Dolmaya, 2011): offering advice (on translation, languages or working as a translator), offering opinions or reflections on translation-related issues, sharing translation news, offering personal anecdotes or sharing daily experiences, sharing resources, seeking feedback from readers, sharing content not related to translation, reviewing articles, books and/or software, sharing translation-related humour, interviewing translators and others. All in all, this small sample indicates that interpreters share their experience and advice more often than they ask for the help of their fellow practitioners. Posts and comments featuring words of appreciation of the colleagues’ effort comprise a different dominant theme. This deep collegiality and peer support often results in the organisation of informal learning initiatives and peer training groups via OSM.

Expressing feelings and emotions is another side of communication in the communities. Quantitative tools employed in this study do not allow tracing the general attitude to the emergency shift to DI in 2020 but reflect the thematic content of the discussions. Therefore this category is probably the most difficult to study using the methodology described above and should be further explored qualitatively from a discourse analytic perspective: the sample features mentions of frustration, exhaustion, uncertainty and stress, which is reasonable in the light of adaptation processes in the midst of an on-going global pandemic.

The affordances of OSM give translation practitioners and students a platform to ask for help and share opportunities for informal professional development. OSM groups provide both instrumental and emotional peer support in the community. Krutka, Carpenter, and Trust (2016) study the role of OSM groups in the formation of teachers’ professional learning networks (PLN). Their model involves sharing, engaging, discovering, reflecting and experimenting as network activities, which has commonalities with the dynamics of the T&I groups. Modeling of the interpreters’ PLN via OSM research can be viewed as a potential direction of future research.

Conclusion

The reported case study demonstrates that UGC can become not only a source of practical information and a helpful hand from colleagues but also an important source of insight into the profession for translation theorists. Although the findings of this exploratory study are quite limited, they add to the discussion of the application of OSM data in translation and interpreting research, as well as to the study of DI practices. This case sheds some light on the key trends that emerged in the OSM spaces in response to the COVID-19 crisis, including the major themes in the groups’ discourse on DI, such as tech challenges (hardware, software/platforms, and interpreter’s home studio), informal professional development initiatives, and the importance of peer instrumental and emotional support in the translation and interpreting community. Other associations with the emergency shift to DI present in the groups’ discourse include the sense of community, compassion and gratitude.

A consistent longitudinal study of professional social media groups involving qualitative netnographic methods could allow tracing the development of professional OSM discourse; however, it should be carefully designed with respect to the ethics of UGC research. Preliminary examination of the Russian-language sub-corpus shows that it shares many features with the English-language sub-corpus, however thematically leans more towards the applied aspects of DI and instrumental help, which may reflect the linguocultural specificity of translation discourse and needs further discussion.