Judging Translation On Social Media: A Pragmatic Look At Youtube Comment Section

Reception: A buzzword in Translation Studies

Recent years saw the establishment of Web 2.0-mediated translation as a legitimate research area in Translation Studies. One of the staple ideas of the disciplines’ technological turn is associated with the changes in the inherent characteristics of the recipient. As envisioned by Cronin (2010), the emergence of the interactive web has redefined the audience:

It is no longer a question of the translator, for example, projecting a target-oriented model of translation on to an audience but the audience producing their own self-representation as a target audience. Such a shift makes problematic traditional distinctions which generally presuppose active translation agents and passive or unknowable translation recipients (p. 4-5).

The participatory web has engendered the category of a user-translator (O’Hagan, 2009) and the practice of user-generated (Desjardins, 2017) or social-media-driven translation (Hebenstreit, 2019). In recent years, audience studies have increasingly come to the fore, and reception has become a buzzword in Translation Studies (Gambier, 2019, p. 52). Studying reception means to investigate how (audiovisual) products are “processed, consumed, absorbed, accepted, appreciated, understood and remembered by the viewers, under specific contextual / socio-cultural conditions [...]” (Ibid;, p. 57-58).

As part of a research project on the nature of YouTube-mediated audiovisual user-generated translation (UGT) into Russian (Krasnopeyeva, 2018), this study examines the characteristics of viewer feedback given at the site of content consumption. It discusses the potential of YouTube comment section as a practical tool both for a user-translator, and a researcher. We present a short description of the broader context of the study, including our definition of a user-translator in the online field of YouTube, and the affordances leveraged by the said translators. The reported case study takes a look at the intentions behind viewer comments left on a popular Russian-language voiceover rendition of a PewDiePie video (2019).

YouTube as an online translation space

YouTube is an online platform which combines the features of a video-sharing service and an online social network. With a Google account, any user can watch videos on YouTube, rate (like) videos, and subscribe to channels. To have a public presence on the platform and upload videos, comment on other users’ videos, or make playlists, a user can start a channel. With the site infrastructure, a user can also edit their videos, and add descriptions and captions in different languages. The comment section feature enables users to discuss videos and communicate with the author of the video, as well as with other viewers.

Being a socio-technical environment, the medium of YouTube enables and fosters the following UGT practices (Krasnopeyeva, 2018, p. 53):

• Institutionalized YouTube community translation (including interlingual subtitling supported by YouTube infrastructure).

• UGT-focused channels.

In the latter case, a UGT-focused channel can be defined as a YouTube channel which uploads translated and revoiced versions of video content (including interlingual free commentary, voiceover or dubbing), originally published on YouTube in a different language. As a result, the original and the revoiced version of the same video simultaneously exist on the platform but lead different, yet connected lives.

YouTube can be viewed as an online field where distinction is formed by the struggle for the attention capital (Levina & Arriaga, 2014). Attention in the form of view and subscriber counts, likes and comments, can be eventually turned into social recognition and profit, which means the key stakes in the social game are also defined by this pursuit of popularity. Therefore, by creating a channel, translators, alongside other creators, are joining the YouTube social game – a network of relations connecting video uploaders and video viewers.

Based on the interview data and analysis of metatexts, such as blog entries and descriptions, Krasnopeyeva (2018) defines the following broad macro-strategies aimed at gaining attention capital and to move up the social hierarchy: 1) Choosing dubbing over subtitling. 2) Monitoring YouTube trends and choosing popular content as source texts. 3) Creating and maintaining the translator’s own narrative based on the borrowed visuals by turning dubbed videos into remixes by incorporating metatranslations (paratexts) in the form of snippets with translator commentary / adding (personal) original conversational videos, audio and other meta-translations to the content stream. 4) Surveying the tastes of the audience by communicating with the viewers via YouTube infrastructure and satellite communities.

This study deals with viewer–creator interaction in the video comment section. Therefore, it explores the fourth attention-accumulating principle which is grounded in the affordances offered by YouTube multimodality (Benson, 2017) and also entwined with the prosumer expertise of user-translators and viewers.

User-centeredness of UGT

Whereas examining audiences is a difficult task, ongoing mediatization and digitalization processes are making it an even bigger research challenge, thanks to a number of factors. Among them are rapid changes in the modes of consumption and technology (cf. the notion of prosumption ), ever-shifting viewing habits, and audience composition (Di Giovanni & Gambier, 2018, p. VII).

User experience research and its interfaces with Translation Studies resulted in the conceptualization of the User-Centered Translation (UCT) model/paradigm (Suojanen, Koskinen, & Tuominen, 2015) which explicitly emphasizes the central role of the user, and offers a range of tools to empower the translator through users. As the UCT paradigm was designed for application in professional settings, the authors note that “a full discussion of UGT would require a book of its own” (Suojanen et al;, 2015, p. 34), but conclude that UGT is UGC “in its most extreme application” (Ibid., p. 7).

The notion of user-centeredness as an inherent characteristic of UGT is supported by YouTube UGT studies, as well. Platform affordances and the socio-technical characteristics of the YouTube online field predispose the formation of close translation agent–recipient relations. These, in turn, account for the strict scrutiny in terms of perceived quality of translation. Non-professional translation is often considered a driving force in the conceptualization of quality in translation spaces (Orrego-Carmona, 2019, p. 2). However, reception and quality assessment in multimodal social media environments remains under-researched and calls for comprehensive context-dependent studies.

Case study background and context

In this study, we focus on viewer feedback given at the site of consumption and explore the communicative intention in users’ reviews. We examine the comment section of a popular video published by a PewDiePie fan-made UGT-focused channel.

With its 103 million subscribers, today PewDiePie remains the most popular English-language YouTube channel in the world. Swedish 30-year-old YouTuber Felix Kjellberg (often dubbed king and emperor of YouTube) produces videos that cumulatively have been viewed billions of times. In 2016, he was included in Time’s list of the world’s 100 most influential people (Parker, 2016). PewDiePie is a media persona, an influencer, and a global brand. Genre-wise his content ranges from Let’s Play (or video-game narration) to vlogs, humorous sketches, and commentary. Therefore, those users who render PewDiePie videos into their native language partly owe their future success to the world-wide popularity of the original.

For this case, we have chosen TheRainbowFox (2019) which is the most popular of many channels featuring translated content originally published by PewDiePie. TheRainbowFox is a UGT-focused channel and a PewDiePie fan community with over 517 thousand subscribers and 681 videos that cumulatively have over 116 million views and over 120 thousand comments. The channel is run by a user under the name of Lis working together with a group of YouTube volunteers. They collectively perform the following tasks: select and edit channel content; do the translation themselves in full or in part, or use crowdsourcing or third party services; dub or revoice the videos; communicate with the audience, other YouTube creators and advertisers.

In terms of the characteristics of the translation per se , the user-translators employ transcreation, or localization, rather than faithful translation of the source text. On the one hand, in many cases, their language choices are aimed at recreating naturally sounding speech, which is a traditional challenge in audiovisual translation (AVT). On the other hand, a variety of translators’ creative strategies are neither dependent on the source, nor do they help overcoming the issue of prefabricated orality. Many of the employed linguistic strategies are norm-defying. We may argue that highly informal conversational mode of delivery, excessive use of invectives (often not present in the source text), discourse markers, vernacular and diminutive forms are, in a way, intricately entwined with the pursuit of popularity in the online field. They help the translators build their own brand on YouTube as legitimate creators.

Data collection

We query a 6,266 token corpus of comments taken from a video called “Поздравление” ( ‘Pozdravleniye’, [Congratulations]) (TheRainbowFox, 2019), which was published by TheRainbowFox UGT channel on April 6, 2019. It is a version of PewDiePie’s “Congratulations” video (PewDiePie, 2019) dubbed into Russian. At the end of the video, the user-translators also include a short clip featuring the dubber’s commentary or a metatraslation.

The data was gathered over two days in January 2020. At the time of data collection, out of 681 videos on TheRainbowFox channel, Pozdravleniye was 10th in terms of view count, and 34th in terms of the number of comments, with viewer engagement ratio of 0.4 percent. A number a YouTube creator should aspire to is 0.5 (Robertson, 2014), which happens to be the ratio of the PewDiePie’s original Congratulations. Pozdravleniye features 1,024 comments. Interestingly, the average number of comments on TheRaibowFox videos is 178.

On closer look, it is apparent that the dubbed version of Congratulations gained a lot of translation-related viewer feedback. We may assume that one of the reasons for that is lip-sync dubbing, which user-translators resorted to instead of the familiar voiceover technique. However, the choice of the video for analysis was not (only) motivated by this factor. First and foremost, it stands out on the channel thanks to the immense popularity and memetic character of its source – an upbeat synth-pop/hip-hop music video. The video has been viewed over 165 million times and embodies a culmination of a year-long battle between PewDiePie and T-Series for the title of YouTube’s most-subscribed channel. This is how Vox’s Aja Romano (2018, par. 9, 11) summarizes fans’ attention to this rivalry: “[...]In their zeal to keep PewDiePie on top, they [PewDiePie followers] have turned “subscribe to PewDiePie” into a massive internet meme, stretching across multiple social media platforms and even into the real world”.

Therefore, the source video holds a certain symbolic significance for the global fan community. All of these factors make Pozdravleniye’s comment section a unique space, where highly attached fans voice their emotionally charged views on YouTube UGT, judge the expertise of user-translators as cultural mediators, and compare the source and the target texts. We can hypothesize that a larger-scale study of such data could aid user-centered UGT research and help reveal the expectation of the target viewer or translation expectancy norm (Chesterman, 2016, p. 62-65).

Pragmatics of viewer response to UGT on YouTube

Word frequency and collocation lists provide lexical foci for further qualitative analysis of pragmatic intentions in the viewer–creator communication. The following semantic groups prevail in the data: translation ( перевод, озвучка, субтитры, голос, оригинал ); quality assessment – satisfaction, gratitude, praise / dissatisfaction ( лучше, очень, круто, лайк, шикарно, спасибо, топ, молодец, годно ); YouTube infrastructure/culture ( видео, Лис, Пьюдипай, Пьюдс, Пьюди, T-series ). Search terms озвуч* , перев* , ошиб* , искаж* , оригинал* were used to select 144 relevant comments/contexts featuring user engagement in the translation quality assessment. The statements reflecting on translation in general, e.g. users’ attitude towards UGT-focused channels, were excluded from the search results. Commenters communicate both positive and negative evaluations of the user-translators’ performance. They compare subtitled and dubbed versions of the original and discuss translation errors and misinterpretations of the source text. Table 01 shows the distribution of positive and negative feedback across the key topics found in the comment section, which will be discussed further. While comments can be organized in threads, in this case we do not focus on the conversations between users, but rather treat comments as separate statements, following the aims of this study.

Comments related to the AVT technique. One of the recurring topics in the commentary is the choice of an AVT technique. (See Table 02 . All examples preserve original spelling; swear words are redacted). Most of the commenters use the terms ‘дублировать’ [to dub], ‘озвучка’ [dubbing, voiceover], ‘переозвучка’ [revoicing], ‘кавер’ [cover] and ‘перевод’ [translation] interchangeably. Thus, most of the viewers do not perceive dubbing and voiceover as two different techniques, which speaks to their lack of expertise. However, from the evaluations we can infer that their preference lies with the dubbed Pozdravleniye rather than with regularly published revoiced videos.

In some cases, positive reviews of dubbing quality can be critical towards perceived translation quality (see Table 2 ). In the following example a commenter gives advice to the user-translator: “крутая озвучка но лучше бы ты посмотрел субтитры клипа оригинального ты бы перевод точнее сделал” [Cool voiceover, but you should have used the subtitles of the original music video; you could have made a more accurate translation]. Eight more commenters express their regret about the fact that user-translators attempted to dub a music video, while the original Congratulations already has built-in Russian subtitles. Out of 32 negative comments, 20 voice seemingly unmotivated critique of the translator. Some of them emphasize negative emotion and contain invectives (4 occurrences), negative evaluation (13), e.g . ‘плохой’ [bad]), ‘неправильный’ [wrong, incorrect] , ‘кривой’ [lousy, defective]. Some comments (7) highlight the issue of translation accuracy and translator’s fidelity.

Comments related to the quality of translation, and translation errors. Only five commenters provide a context or give an example of a translation error. However, there are cases when users offer their own translation solutions. For example, in the following comment a user suggests a word-for-word translation as a reference: “вот здесь должно быть : «как я понял чтобы победить одного шведского паренька вам надо миллиард азиатов». Перевёл не правильно” [Here’s what you should have said: you need a million Asians to take on one Swedish guy. And your translation is incorrect]. Translation solutions suggested by viewers often have poor grammar or misspelled words, or even criticize viable translation options. For example, one commenter wrote: “Дефомация? Такого слова вроде и нет… ” [Defomation? I reckon there’s no such word]. In the Russian version, ‘ диффамация’ [defamation] is a contextually relevant translation solution.

Four commenters note a misinterpreted joke. For example, “Шутка про: Ща устрою геноцид, чтоб называли Гит..... Не очень получилось так перевести. Нужно было пропеть ги...... И потом смена кадра. А то если не смотреть перевод текста, ничего не понятно” [A joke: I’ll start a genocide so that they call me Hit… Not the best option. You should’ve sung Hit… and then changed the scene. Otherwise, it’s not clear without the text]. In this example we can track the components of the viewer’s expertise: he/she not only reviews the translation but also suggests a solution, which factors into both the format and function of the target text. Moreover, he/she comments on the user-translator’s video editing skills and proposes changes to the video layout.

Only eight comments in this category combine critique and positive evaluation. In these cases, users seem to rate their overall experience from watching the video. For example, “Кривой перевод но а так топ” [Everything is top-quality, translation is bad though], or “Очень круто, но больше половины неправильно переведено” [Very cool, but half of the video is translated incorrectly]. In the contexts featuring озвуч *, перев *, ошиб * and искаж * search terms, users’ criticism is often leveled with the translator’s lexical choices. For every two subjective comments, there is one valid argument, whereas for every three examples of severe dissatisfaction there is one comment combining criticism with praise, expression of gratitude, support, or sympathy.

Overall users’ positive reviews are more common (48 comments) than negative ones. Commenters who are satisfied with translation quality often praise user-translators for the immense effort they have put into translating the song and producing the video. Examples in Table 2 show that the language of the comments is emotional. Viewers use various lexical and syntactic means of expressing emotion and positive semantic orientation, e.g. superlatives, invectives (in 3 positive comments), emphatic punctuation, capitalization, and emoji. In some cases, they speak directly to the user-translator and ask rhetorical questions.

Similarly to the negative evaluations of quality, positive ones are emotionally charged but do not have any substance. Most comments convey users’ attitude towards the video – thus, 19 out of 48 positive comments are expressions of praise and gratitude for making/publishing the translation. Some commenters specify the reason why they like the video. For example, 13 comments point out that user-translators have successfully attempted to write song lyrics (e.g. “перевел песню в рифму” [translated song lyrics rhyme]). Those users, who express their appreciation of an effort to dub a song, are more lenient towards translators’ strategies. In 15 positive comments, users note that the UGT video allowed them to enjoy Congratulations without subtitles. For example, “тупо респект!!! долго ждал когда кто-то сделает перевод прост субтитры скушно читать XD” [You have my respect! I’ve been waiting for someone to make a translation, subtitles bore me haha]. As in this statement, users often note that they anticipated the video, which speaks to their attachment to the subject, and interest in UGT in general ( “Да этого перевода я ждала с самого выхода этой песни [I’ve been waiting for this translation to come out since the song was released]”);