Pivot Templators’ Challenges and Training: Insights from a Survey Study with Subtitlers and Subtitler Trainers Desafíos y formación para los creadores de plantillas dinámicas: hallazgos de una encuesta con subtituladores y formadores en subtitulado Défis et formation pour les créateurs de modèles dynamiques : résultats d’une enquête avec des sous-titreurs et des formateurs en sous-titrage Desafios e treinamento para criadores de modelos dinâmicos: resultados de uma pesquisa com legendadores e treinadores de legendagem Hanna Pięta Abstract Assistant Professor, Universidade nova de Lisboa, Centre for English, Given the increase of non-English audiovisual content, translating through piv- Translation and Anglo-Portuguese ot templates is increasingly common. Yet, pivot templates have attracted scant Studies (cetaps), Lisboa, Portugal. scholarly attention. Several factors remain unclear. Among the first factors are the hannapieta@fcsh.unl.pt questions of who creates pivot templates, from what languages are they translated, https://orcid. into what languages are they translated, and what audiovisual products and chan- org/0000-0002-5229-1941 nels are they for? Secondly, what are the challenges involved in the creation of templates? The third factor would be whether there is any training available for Susana Valdez pivot template creators, and whether this training is effective. To provide more 1 Assistant Professor, Leiden University clarity, we distributed a questionnaire and elicited replies from 100 pivot templa- Centre for Linguistics, Leiden, The tors and 75 subtitler trainers based in Europe. The results indicate that most pivot Netherlands. templators translate into English as L2, i.e., having learned it as a second language. Because of this, diverse difficulties arise. Training in pivot template making seems rare, especially compared to training in subtitling in general. The training that is offered comes mainly from the industry and not academia. This is problem- atic for various reasons: (a) the former does not typically cover issues related to subtitler ethics or the sustainability of the profession, and (b) translator training and professional codes of conduct eye L2 translation with suspicion. Drawing on Research by Hanna Pięta was funded the respondents’ insights, we argue that much can be gained from teaching non- by Fundação para a Ciência e a Tecno- logia under Projects uidp/04097/2020 English-language native speakers how to create English-language templates. and uidb/04097/2020. Research by Es- ter Torres-Simón was funded by mcin/ Keywords: Pivot templates in English, subtitler training, L2 translation, pivot aei/10.13039/501100011033 / feder, templators, online questionnaire, avt ue under Project pid2021-122897NB- I00, and research by Rita Menezes was Resumen funded by Fundação para a Ciência e a Tecnologia under Grant 2020.08087.bd El aumento de contenido audiovisual en lenguas distintas al inglés está extendien- and uidp/00114/2020. do la traducción mediante plantillas dinámicas. Sin embargo, estas han atraído Received: 2022-03-30 / Accepted: 2022-08-18 / Published: 2023-05-XX https://doi.org/10.17533/udea.ikala.v28n2a02 Editors: Jorge Díaz-Cintas, Professor of Translation Studies, University College London, United Kingdom; Lydia Hayes, Lecturer, School of Modern Languages, University of Bristol, United Kingdom. Copyright, Universidad de Antioquia, 2022. This is an open access article, distributed in compliance with the terms of the Creative Commons license by-nc-sa 4.0 International. Íkala, Revista de Lenguaje y Cultura Medellín, Colombia, Vol. 28 Issue 2 (May-August, 2023), pp. 1-21, ISSN 0123-3432 www.udea.edu.co/ikala Íkala Hanna Pięta, Susana Valdez, Ester Torres-Simón, and Rita Menezes s.valdez@hum.leidenuniv.nl escasa atención académica. Aún no está claro: en primer lugar, quién crea https://orcid. plantillas pivote, desde y hacia qué idiomas y para qué productos y canales au- org/0000-0001-5461-2078 diovisuales se crean. En segundo lugar, qué desafíos implican sus creaciones y si hay capacitación disponible para pivot templators (también conocidos Ester Torres-Simón como creadores de plantillas) y si es útil. Para cerrar esta brecha, informamos Assistant Professor, Universitat sobre los resultados de un cuestionario que involucró 100 plantillas dinámi- Autònoma de Barcelona (uab), cas y 75 capacitadores de subtituladores en Europa. Los resultados indican Barcelona, Spain. que la mayoría de los creadores de plantillas dinámicas traducen al inglés Ester.Torres@uab.cat como segunda lengua, lo que plantea variados desafíos. La formación en https://orcid. creación de plantillas dinámicas parece rara, especialmente en comparación con la formación general en subtitulaje. La formación que se ofrece provie- org/0000-0002-8109-4357 ne principalmente de la industria y no de la academia. Esto es problemático porque: a) el primero generalmente no cubre temas relacionados con la ética Rita Menezes del subtitulador o la sostenibilidad de la profesión, y b) la capacitación de Researcher, University of Lisbon, traductores y los códigos de conducta profesionales ven con recelo la traduc- Centre for English Studies, Portugal. ción en una segunda lengua. Basándonos en las opiniones de los encuestados, rita@ritamenezes.com.pt argumentamos que se puede ganar mucho enseñando a los hablantes nativos https://orcid. que no hablan inglés cómo crear plantillas en inglés. org/0000-0002-7918-6412 Palabras clave: plantillas dinámicas en inglés, formación de subtituladores, traducción en segunda lengua, creadores de plantillas dinámicas, cuestiona- rios en línea, tav Résumé La croissance du contenu audiovisuel dans des langues autres que l’anglais 2 favorise la traduction à l’aide de modèles dynamiques. Toutefois, ces der- niers n’ont guère attiré l’attention des universitaires. On ne sait toujours pas : premièrement, qui crée les modèles pivots, à partir de quelles langues et vers quelles langues, et pour quels produits et canaux audiovisuels ils sont créés. Deuxièmement, quels sont les défis liés à leur création et si une forma- tion est disponible pour les créateurs de pivot templates (également connus sous le nom de créateurs de modèles dynamiques pour le sous-titrage) et si elle est utile. Pour combler cette lacune, nous présentons les résultats d’un questionnaire auquel ont répondu 100 créateurs de modèles dynamiques et 75 formateurs de sous-titreurs en Europe. Les résultats indiquent que la majorité des créateurs de modèles dynamiques traduisent en anglais comme deuxième langue, ce qui pose un certain nombre de problèmes. La formation pour la création de modèles dynamiques semble rare, surtout par rapport à la formation générale en sous-titrage. Les formations proposées proviennent principalement de l’industrie plutôt que du monde universitaire. Cela pose problèmes car : a) la première ne couvre généralement pas les questions liées à l’éthique du sous-titrage ou à la durabilité de la profession, et b) la formation des traducteurs et les codes de conduite professionnels considèrent la traduc- tion en langue seconde avec méfiance. Sur la base des opinions des personnes enquêtées, nous soutenons qu’il y a beaucoup à gagner à enseigner aux locu- teurs natifs non anglophones comment créer des modèles en anglais. Mots clef : modèles dynamiques anglais, formation des sous-titreurs, traduc- tion vers une seconde langue, créateurs de modèles dynamiques, enquête en ligne, tav Medellín, Colombia, Vol. 28 Issue 2 (May-August, 2023), pp. 1-21, ISSN 0123-3432 www.udea.edu.co/ikala Íkala Pivot Templators’ Challenges and Training: Insights from a Survey Study with Subtitlers and Subtitler Trainers Resumo O crescimento do conteúdo audiovisual em outros idiomas além do inglês está difundindo a tradução usando padrões pivô. No entanto, estes têm atraído pouca atenção da academia. Ainda não está claro: primeiro, quem cria os padrões pivô, de e para quais idiomas e para quais produtos e canais audiovisuais eles são criados. Em segundo lugar, que desafios estão envolvidos em sua criação e se o treinamento está disponível para os pivot templates (também conhecidos como criadores de modelos) e se ele é útil. Para preencher esta lacuna, relatamos os resultados de um questionário envolvendo 100 padrões pivô e 75 treinadores de legendagem na Europa. Os resultados indicam que a maioria dos criadores de modelos dinâmicos traduzem para o inglês como um segundo idioma, o que representa uma série de desafios. A formação na criação de padrões pivô parece rara, especialmente em comparação com o treinamento geral de legendagem. O treinamento que é oferecido vem principalmente da indústria e não da academia. Isto é problemático porque: a) o primeiro geralmente não cobre questões relacionadas à ética da legendagem ou à sustentabilidade da profissão, e b) o treinamento de tradutores e códigos de conduta profissionais vêem a tradução para o segundo idioma com desconfiança. Com base nas opiniões dos entrevistados, argumentamos que há muito a ganhar com o ensino de falantes nativos de outros idiomas que não o inglês como criar padrões em inglês. Palavras chave: padrões pivô em inglês, formação de legendadores, tradução para uma segunda língua, criadores de padrões pivô, questionário online, tav 3 Medellín, Colombia, Vol. 28 Issue 2 (May-August, 2023), pp. 1-21, ISSN 0123-3432 www.udea.edu.co/ikala Íkala Hanna Pięta, Susana Valdez, Ester Torres-Simón, and Rita Menezes Introduction Theoretical Framework In this article, we discuss the use of templates, In this section, we outline relevant insights gained which are subtitle files “timed to film shots and from earlier studies on pivot translation, on pivot audio that [are] used as a basis to originate other templates, as well as on the relation between subtitle versions” (avt Masterclass, 2021, p. 4). these two practices and translation directionality, Templates “usually serve as a basis for interlin- English as a lingua franca, and subtitler train- gual subtitling in multilingual workflows where ing. In so doing, we highlight gaps in the current the content is localised into multiple languages.” understanding of these pivot practices. (p. 4). We are concerned with a particular group of templates, which audiovisual translation Pivot Translation and Pivot Templates (avt) researchers term pivot templates (i.e., tem- Pivot templates are a particular instance of pivot plates produced in a language different from the translation (also known as indirect translation; language of the original content and the final see Pięta, 2021). Pivot translation is a long-stand- subtitles). When referring to the professional prac- ing and widespread practice in many domains titioner who creates (pivot) templates, we use the (e.g., Bible translation, interpreting, video game term templator, coined by Díaz-Cintas (2021). localization). However, as a subfield of research, The rationale for this study is that, although pivot translation is recent, rarely looks beyond the the use of pivot templates is a widespread prac- 20th century, and has focused mainly on literature tice (Valdez et al., 2023), several aspects remain printed in book form (Pięta, 2021). unclear. First, who creates pivot templates, which 4 In the avt industry, templates started to be com- audiovisual products are created using this type of monly used with the arrival of the DVD in the templates, and which languages work as source, 1990s (O’Hagan, 2007, p. 162) in a paradigm stepping-stone, and target languages in pivot- where non-English content was rarely distributed template workflows. Second, what challenges on a global scale and when subtitles needed to be are involved in creating pivot templates. Third, produced fast (to avoid piracy) and cheaply (to whether there is training available for templators maximise profit). To efficiently manage subtitling and what type of training is considered help- projects where English content was simultane- ful to produce a fit-for-purpose pivot template. ously translated into multiple languages, templates This lack of research-based knowledge is prob- were created by native English speakers to ensure lematic given the increase in original content that the dialogues were reproduced accurately. production in non-English-languages (Netflix Q4 2021 Earnings Interview, 2022) and the resulting Pivot templates became a standard go-to solution demand for professionals with training and experi- in the 2010s when subscription video-on-demand ence in translating into English as a pivot language (SVoD) platforms forced a new paradigm in the (see, for instance, Bolaños-García-Escribano et al., distribution of audiovisual content (for instance, 2021, p. 5). Hayes (2021) discussed how SVoD platforms drove the increase in demand for dubbing into To address these issues, we report on the results English). It was then that, in the global media land- of a dedicated questionnaire aimed at templators scape, English language productions began to give and subtitler trainers; but, first, we analyse key way to non-English productions, and non-Eng- takeaways from related research and outline the lish originals became more common (for instance, methods used in questionnaire design and data Netflix (2022) reported non-English original con- analysis. tent as a higher growth area since 2015). Medellín, Colombia, Vol. 28 Issue 2 (May-August, 2023), pp. 1-21, ISSN 0123-3432 www.udea.edu.co/ikala Íkala Pivot Templators’ Challenges and Training: Insights from a Survey Study with Subtitlers and Subtitler Trainers Stakeholders’ opinions about pivot templates (Georgakopoulou, 2012, p. 81). Pivot templates vary. Language service providers tend to empha- are also seen as a threat to translation jobs and eth- sise their positives, as pivot templates help ics. For example, subtitlers translating from pivot minimise the technological requirements for sub- can be said to take away jobs from colleagues who titlers (Georgakopoulou, 2019, p. 139) and are already disadvantaged because of their work- streamline outsourcing (Kapsaskis, 2011). In ing languages (DuPlessis, 2020). many pivot workflows, subtitlers only need to know how to do linguistic transposition, as tim- Despite these controversies, pivot templates are ing is done by templators. unlikely to disappear. One reason for this is the streaming giants’ strategies to increase their non- Moreover, subtitlers do not need to understand English content (Rodríguez, 2017). the language of the non-English content, as they will be working from the English pivot. This Pivot Template Creation and Directionality allows minimisation of costs (translating from and Creating pivot templates is often combined with into English is typically cheaper than translating another commonly disapproved practice, namely between non-English languages; see Vermeulen L2 translation (or translation into a non-mother 2011, p. 122) while enlarging the pool of avail- tongue), whereby translators work out of their able translators (more subtitlers are working to L1 (or native language) and produce translations or from English than in language combinations in their “first foreign” language (L2; Whyatt, that do not involve English; Georgakopoulou, 2019, p. 80). L2 translation has traditionally been 2019, p. 138). Locked templates can be used as an considered substandard and cognitively more additional risk management strategy. When using demanding (Pokorn, 2005). However, this view locked templates, subtitlers are not allowed to 5 has not been consistently supported nor rebutted adjust timing and segmentation (e.g., no in- and by research. Despite an increasing body of dedi- out-time adjustments, no merging or splitting), cated studies (for an overview, see Ferreira et al., proving useful when updates to the video (and, 2016), we still do not know to what extent and consequently, the subtitles) are expected. how exactly L2 translation differs from L1 trans- lation — both in terms of the challenges involved In contrast, translators approach pivot workflows and the quality — although some progress has with mixed feelings. Some recognize the bene- been made, particularly with regard to written fits: more job offers, an opportunity to broaden translation. For example, Whyatt (2019) shows cultural horizons, less heavy-lifting (as timing that the effect of directionality (i.e., whether and some detective work about culture-specific translators work into or out of their L2) might be elements might have been done for the final subti- modulated by text type, translator experience, and tler, who can now focus on other relevant issues), the degree of familiarity with certain text types. and more freedom because viewers do not under- stand the non-English content either (Costa, The fact that translators (need to) translate into 2020; Oziemblewska & Szarkowska, 2020; their L2 has been acknowledged in different sub- Valdez et al., 2023). Others see pivot templates fields, and many training methods have been as a threat to translation quality. A mistake in studied (Campbell, 1998). However, in avt, sys- the English pivot may be replicated downstream; tematic research is rare. A telling example is that ambiguities and nuances in the non-English orig- mainstream handbooks and encyclopaedias of inal may be filtered (Casas-Tost & Bustins, 2021; audiovisual translation do not include a dedi- Oziemblewska & Szarkowska, 2020). If tem- cated entry on English as L2 in avt. Therefore, it plates are locked, fine-timing and easy adaptation remains unclear which challenges templators face to local guidelines (and norms) are hindered when working from their L1 into English. Medellín, Colombia, Vol. 28 Issue 2 (May-August, 2023), pp. 1-21, ISSN 0123-3432 www.udea.edu.co/ikala Íkala Hanna Pięta, Susana Valdez, Ester Torres-Simón, and Rita Menezes Pivot Translation and English The few studies that exist are product-oriented and focused on issues of quality (Artegiani & Pivot templates as well as pivot translation in Kapsaskis, 2014; DuPlessis, 2020; Oziemblewska general are intricately linked to the hegemonic & Szarkowska, 2020; Vermeulen, 2011). Much status of English as a lingua franca. Research sug- has been said about the negative consequences gests that pivot translation can reinforce and of this practice; but who creates pivot templates, undermine the hegemonic position of English what challenges templators face, and what type (Ringmar, 2007; Tesseur & Crack, 2020). On of training is available, if any all remain unclear. the one hand, English pivot templates enhance These are the issues that we want to explore. worldwide access to content originally created in languages other than English, thus working as an Method empowering device for the cultures where these languages are spoken. On the other hand, English We designed a comprehensive online question- pivot templates can act as anglocentric filters, fur- naire in SurveyMonkey to elicit templators’ and ther exacerbating the power imbalance between trainers’ experiences and expectations when cre- English and other languages and cultures (Valdez ating pivot templates and teaching about how to et al., 2023). With English acting as a main create pivot templates. The questionnaire, con- default middleman, “there is a danger of cultural ducted between January and March 2021, was homogenization, whereby consumer preferences distributed online among audiovisual transla- are anglicised, and English mediating is preferred tors and trainers in specialised groups on social to direct translation from more peripheral lan- media, through European professional translation guages” (Pięta, 2021; cf. Ringmar, 2007). and audiovisual translation associations, universi- 6 ties in Europe, and personal networks. Before its Pivot Translation in Training dissemination, the questionnaire was pilot tested, and the feedback was implemented. The most Given the conflicting opinions about pivot work- common error types related to survey research, flows, it is unclear to what extent creating pivot such as coverage, sampling, and non-response, translation is part of subtitler training. To the best were also carefully considered during the ques- of our knowledge, only one study tried to partly tionnaire design and data collection. address this issue. Torres-Simón et al. (2021) looked at the responses to their survey addressed To conduct this study, we abided by gdpr to translator trainers, at mainstream training text- (European Union General Data Protection books, at references to pivot translation in the Regulation) procedures. In addition, we followed European Masters in Translation (emt) compe- the ethical standards of research set by our uni- tences, and at the syllabi of emt programmes. versities, and we obtained informed consent from Their results suggest that there might be a dis- the respondents. In particular, before agreeing to association between what is publicly on display participate, respondents were asked to read a writ- and what happens in translation classrooms. emt ten text describing the study’s general goals, how competences, official emt syllabi, and published data is collected and stored, how confidential- textbooks are practically silent on pivot trans- ity is managed, and the respondents’ rights. We lation; but many trainers surveyed in this study also explicitly asked respondents to confirm that teach how to translate from a translation and, to they had read and understood the information a lesser degree, with a further translation in mind. given. Participants expressly agreed for the data collected to be stored and used in research in an To recap, pivot translation in other domains has anonymised form. If consent was denied, respon- been attracting growing scholarly attention, but dents were sent to the final page of the survey and studies on pivot templates are recent and rare. their information was not collected. Medellín, Colombia, Vol. 28 Issue 2 (May-August, 2023), pp. 1-21, ISSN 0123-3432 www.udea.edu.co/ikala Íkala Pivot Templators’ Challenges and Training: Insights from a Survey Study with Subtitlers and Subtitler Trainers Data Collection and Analysis Results and Discussion The questionnaire consisted of two main sections This section presents the questionnaire results to with fifty-six questions in total.1 The first part was discuss (i) the profile of templators and template aimed at subtitlers and the second at subtitling use, (ii) the challenges involved in creating pivot trainers. Each main section was subdivided into templates, and (iii) the availability and character- two additional sub-sections. The first contained istics of training considered helpful to produce a questions related to the profile of the respondents, fit-for-purpose pivot template. To do so, we have and the second questions were about respondents’ analysed 100 templators’ responses to 19 open and experience in working from and/or for pivot closed questions on their profile (demograph- templates. In this article, we focus on answers pro- ics questions) and professional practice (see vided by the templators and subtitling trainers. Templators). To understand if and how the cre- ation of pivot templates is taught, the profile of Different types of questions were employed: open these trainers, and the motivations behind not and closed questions, including multiple-choice, teaching how to create pivot templates, we anal- verification box, and Likert-scale type questions. ysed seventy-five subtitling trainers’ responses to The Likert-scale type questions included reserve- 19 open and closed questions (see Trainers). coded statements so that respondents did not “fall into the habit of agreeing or disagreeing” Templators’ Profile (Mellinger & Hanson, 2017, p. 32), and the order of the statements was randomized to avoid pri- Based on the 100 answers, the results show that macy bias and recency bias. our templators are mostly professional subtitlers (83), with seven working as volunteers and fan- 7 In the analysis phase, the data from the closed subbers and ten working as both professionals and questions were processed in Excel using descrip- volunteers. Their age ranges from 18 years to 55 tive statistical analysis. For the open questions, or above, but most are under 44 (61). the data were exported to atlas.ti. Here, the data were coded and organised around recur- Regarding qualifications, the data shows that ring themes (thematic analysis) using inductive most respondents (63) have qualifications in sub- coding. In other words, the extracted codes were titling or qualifications that included modules on based on the actual language used by the respon- subtitling, with 25 holding a master’s degree, 12 dents — their words, phrases, and terms (Saldaña, a bachelor’s degree, another 12 a course or work- 2016, p. 105). As for the presentation of the data, shop certificate, five a postgraduate degree, and when reporting the open answers, the respon- two a PhD. These results seem to be in keeping dents’ statement beliefs are quoted verbatim, with the most recent reports on the training of including any typos. avt. As Bolaños-García-Escribano et al. (2021, p. 2) point out, in the last 20 years, undergrad- Three hundred eighty subtitlers from 32 European uate and postgraduate avt modules have been countries answered our questionnaire. Of these, created and introduced in new and existing trans- 100 indicated that they knowingly create pivot lator training programmes in significant numbers. templates used for further translation into a third This has been corroborated to some extent by language. The next section reports on these 100 the survey conducted by the avt working group templators. of the European Master’s in Translation in 2020 (Valdez et al., forthcoming). Nevertheless, there 1 The complete questionnaire is available at https://doi. are still reports of a perceived distance between org/10.6084/m9.figshare.19368896.v1 what language service providers (lsp) need from Medellín, Colombia, Vol. 28 Issue 2 (May-August, 2023), pp. 1-21, ISSN 0123-3432 www.udea.edu.co/ikala Íkala Hanna Pięta, Susana Valdez, Ester Torres-Simón, and Rita Menezes subtitlers and the competences of (recent) gradu- Portugal (14), Spain (13), the United Kingdom ates (Nikolić & Bywood, 2021, p. 61). (nine), Germany (six), and France (five). In terms of experience in subtitling, most partici- Regarding the source and target working lan- pants (53) had more than ten years of experience: guages that they usually work from and into, these 25 participants with more than 20 years, 10 with templators usually work from and into English 16-20 years, and 18 with 11–15 years. Twenty-four (94 working from, and 43 into), Spanish (31 and participants had between 6 and 10 years of experi- 15), French (29 and five), German (18 and nine), ence, 19 had between 1 and 5 years, and four had Italian (12 and four), and Dutch (Netherlands) less than 1 year. (10 and 16). The data, therefore, suggests that English is the main source and target language In terms of distribution per country, our templa- independently of the templators’ native language. tors are based in 24 European countries. Figure 1 shows the distribution by country. The six most Figure 2 shows the usual source and target lan- represented countries are the Netherlands (16), guages with more than one respondent. The languages (source or target) with one respondent Figure 1 Distribution of Respondents per Country 8 Figure 2 Usual Source and Target Languages of Respondents Medellín, Colombia, Vol. 28 Issue 2 (May-August, 2023), pp. 1-21, ISSN 0123-3432 www.udea.edu.co/ikala Íkala Pivot Templators’ Challenges and Training: Insights from a Survey Study with Subtitlers and Subtitler Trainers were Serbian, Bosnian, Finnish, Latvian, Yiddish, template). However, in other cases, non-English Macedonian, Korean, Basque, and Ukrainian. pivot templates might be used as a linguistic aid too (i.e., subtitlers might use them to cut English as a Pivot Languages and Ultimate Source go-between, for instance between closely related lan- Languages guages, such as Portuguese, Spanish, and Italian). To identify the pivot languages and further Concerning the ultimate source languages (usl), the understand the predominance of hyper-central respondents were asked to indicate the languages of English or central languages in template creation, the audiovisual products they typically work from we asked the respondents to indicate the tem- when creating a pivot template. We wanted to iden- plate languages they usually create with a further tify the languages of the original audiovisual content translation in mind. As shown in Table 1, most when translators are asked to create templates to respondents (64) reported that they typically cre- understand if there were signs of the increased lin- ate templates in English. This is not surprising guistic diversity reported in the literature. Among given the traditional status of the English lan- the respondent templators, as seen in Table 2, the guage as the lingua franca worldwide and in the most common usl is Spanish (23), followed by av sector. These numbers also suggest that our English (18), German (14), and French (13). In respondents create pivot templates in English, other words, even though hyper-central English even when English is not their native language. (Heilbron, 2010) is not the most common usl, it still occupies an important second place. It also high- Concerning other languages, a smaller number of lights the role of English as a pivot language. It is also templators answered that they create templates in Spanish (13), French (eight), German (four), Dutch (four), Italian (two), Portuguese (two), 9 Table 2 Ultimate Source Language (With More Than and Russian (two), among other languages. One Respondent) We did not ask our respondents for what purpose their non-English templates are used. However, Language of the Number of based on earlier research (Nikolić, 2015), we Audiovisual Products Respondents presume that some are used only as a technical sup- Spanish 23 port (i.e., subtitlers downstream only make use of English 18 the predefined timing and not the language in the German 14 French 13 Table 1 Language of Pivot Templates (With More Swedish 6 than One Respondent) Japanese 4 Language of Number of Dutch 4 Pivot Templates Respondents Portuguese 4 English 64 Korean 3 Spanish 13 Chinese 3 French 8 Norwegian 3 Polish 3 German 4 Russian 3 Dutch 4 Czech 2 Italian 2 Catalan 2 Portuguese 2 Romanian 2 Russian 2 Slovenian 2 Medellín, Colombia, Vol. 28 Issue 2 (May-August, 2023), pp. 1-21, ISSN 0123-3432 www.udea.edu.co/ikala Íkala Hanna Pięta, Susana Valdez, Ester Torres-Simón, and Rita Menezes interesting to see Spanish, German, and French (i.e., Table 4 Channels or Platforms Where the Subtitles central, globally powerful languages) occupying such Created Based on Templates Are Distributed prevalence in this list of usls. Channels or Platforms Number of Respondents avt Products and Their Channels streaming services 47 We were also interested in the type of avt prod- cable tv 30 ucts that templators usually work with. When open/public tv 27 asked “Which avt products do you usually cre- cinema 26 ate a template for?” respondents indicated that in most cases, they create templates for films (tv websites 20 or cinema), series, sitcoms, and/or trailers (66; dvd 16 see Table 3). This is closely followed by docu- YouTube 13 mentaries, docudramas, and/or fly-on-the-wall I am not informed/ I don’t know 11 docudramas (53). In a smaller proportion, our templators create templates for corporate vid- (30), open/public tv (27), cinema (26), and web- eos, promotional videos, specialised videos and/ sites (20). dvd (16) and YouTube (13) gather the or commercials (27), news, political speeches, lowest number of responses. press conferences, and/or interviews (15), car- toons and/or anime (14), and video clips (music, Challenges and Expectations Creating YouTubers) (four). Pivot Templates 10 The respondents were also asked about the chan- In response to the question “Tell us what is your nels or platforms where the subtitles based on biggest challenge, in terms of quality, when their templates are distributed, and 47 indicated creating templates?”, answered by 81 templa- streaming services (see Table 4). This confirms tors, three broad themes emerged from the our expectation that, for our respondents, the use analysis: “Linguistics and Cultural Aspects” (45), of pivot language templates is most common on “Professional Ecosystem” (39), and “Temporal streaming platforms. This is followed by cable tv and Spatial features” (32). Importantly, the only pivot language mentioned in the dataset was Table 3 avt Products for Which Pivot Templates Are English. Usually Created for Particularly revealing is how the respondents Number of expressed their challenges. Throughout the data- avt Products Respondents set, templators conveyed concerns regarding Films (tv or cinema), series, sitcoms or 66 what the subsequent translator might need from trailers the pivot template. Commenting on this, one of the Documentaries, docudramas or fly-on- 53 respondents wrote that their main concern was the-wall docudramas “knowing that other translators will be working Corporate videos, promotional videos, 27 from it and that I need to lay the groundwork.” specialised videos and/or commercials Thus, when creating English-language templates, News, political speeches, press 15 templators attempt to anticipate the needs and conferences, and/or interviews expectations of translators working into other Cartoons and/or anime 14 languages. Templators then base their deci- sion-making process on these expectations of Video clips (music, YouTubers) 4 others. Curiously, even though pivot templates Medellín, Colombia, Vol. 28 Issue 2 (May-August, 2023), pp. 1-21, ISSN 0123-3432 www.udea.edu.co/ikala Íkala Pivot Templators’ Challenges and Training: Insights from a Survey Study with Subtitlers and Subtitler Trainers are sometimes reported to have dual audiences main recommended strategies when subtitling. (English-language viewers and subsequent trans- For instance, as Díaz-Cintas and Remael (2021, lators), these templators only referred to the pp. 146–147) write, “subtitles are rarely a verbatim translators’ expectations and not the viewers and detailed rendering of the spoken text, and they (Torres-Simón et al. 2022). need not be.” But this does not seem helpful when creating pivot templates, as Artegiani and Kapsaskis As for the most common challenges, 45 respon- (2014, p. 433) have found from their analysis of dents expressed concerns related to “Linguistics French, Greek, and Spanish subtitles of episodes of and Cultural Aspects” (see Figure 3). Among these, The Sopranos based on pivot template files: 15 templators commented that one of their biggest challenges was related to text reduction or, as one At the same time, because they follow the convention- respondent wrote, “condensation - because I know, al and normative format of interlingual subtitles, the the next translator will miss that information.” This template files examined made use of standard strate- is also observed when our respondents refer to a lit- gies of text reduction, which were on many occasions blindly reproduced in the tls. (...) [T]he consequenc- eral “translation/transcription,” indicating that they es of text reduction choices in the template left visible “should be as literal as possible” or that they should traces in all three languages examined and sometimes “adapt as little as possible so the target language can had an adverse impact on translation quality. remain as faithful as possible to the original.” Templators also expressed concerns regard- This belief that a pivot template should avoid ing segmentation and line breaks (seven). Some text reduction and be as “literal” or as “faithful” templators mentioned that when segmenting, as possible is very much opposed to what is recom- templators have to consider that “it has to work for mended for subtitling without a further translation different languages,” while others reminded that 11 in mind. Text reduction is, in fact, one of the “the grammar might be very different in the third Figure 3 Biggest Challenges When Creating Pivot Templates — Theme “Linguistics and Cultural Aspects” (45) Medellín, Colombia, Vol. 28 Issue 2 (May-August, 2023), pp. 1-21, ISSN 0123-3432 www.udea.edu.co/ikala Íkala Hanna Pięta, Susana Valdez, Ester Torres-Simón, and Rita Menezes language and might need different partitions.” anticipate. As one templator wrote, “Accuracy and This suggests that, as a pivot language, English making sure to anticipate any queries the trans- might be challenging for templators, as it might lator may have in terms of slang, culture, etc. so lack certain grammatical categories or syntacti- that I can add the relevant notes to the template.” cal structures that the usl and the ultimate target Another commented, “To consider if the English language (i.e., the language of the final subtitles) is ‘clear enough’ to be understood by my colleague, might share. Besides these considerations, others and ensure that substitling [sic] quality criteria also showed that they were aware that some trans- are met, dealing with culture specific items.” These lators might not be allowed to change some of the replies seem to stress that what matters in tem- spatial or temporal features of the template, like plate creation is not only the quality of English. adjusting time codes or merging or splitting seg- Templators’ prior knowledge of other languages ments, and how negative this can be. As Nikolić and cultures matters as well, as it might help mit- (2015, pp. 199–200) points out, discussing the igate the Anglocentric filter and provide helpful use of locked templates: “[W]hat subtitling com- information to subtitlers downstream, catering to panies tend to ignore is the lexical, syntactic and the needs of their languages and cultures. cultural differences between languages.” The second most recurrent challenges, men- For other respondents (nine), cultural aspects, tioned 39 times, were related to the Professional such as cultural references, taboo, language vari- Ecosystem (see Figure 4). This theme came ation, and humour are particularly challenging to up in discussions of challenges pertaining to 12 Figure 4 Biggest challenges when creating pivot templates - Theme “Professional Ecosystem” (39) Medellín, Colombia, Vol. 28 Issue 2 (May-August, 2023), pp. 1-21, ISSN 0123-3432 www.udea.edu.co/ikala Íkala Pivot Templators’ Challenges and Training: Insights from a Survey Study with Subtitlers and Subtitler Trainers annotations (16), the materials provided with the lost in the English translation due to reduction, template creation task (15), the profession (six), for example, might not be easily flagged for the and communication (two). benefit of subtitlers downstream because the tem- plator lacks a dedicated space for this purpose. Concerning annotations, templators mainly referred to the challenge of creating complete annota- In the case of the materials provided by the cli- tions. “Knowing what to include in the annotations/ ent or project manager with the template creation comments that could help the translators when task, the principal reported challenge is related to they translate from the template,” as written by the audio quality of the source file (seven). Other one of the respondents, was recurrent among the challenges include the dialogue lists that are templators that took this opportunity to discuss either not provided (two) or contain errors (two). annotations. Others also referred to templators’ Guidelines also seem to be a source of the problem lack of “skill” when it comes to creating “signifi- since, in some cases, they are “incredibly strict” or cant notes” and to the need to have “vast linguistic they are not provided at all, leaving the templa- knowledge, such as the importance of formality, tor to her own devices: “I didn’t receive further gender sentence structure of languages you don’t speak” to “be helpful to other languages.” instructions, so I did the same as I use to do when I translate and subtitle.” Related to this is the challenge of creating anno- tations “that will fit all languages.” This related Respondents (six) also reported challenges related challenge was expressed by six respondents. to the profession itself (see Figure 5). Among Notably, two of our respondents referred to not these challenges, the most reported were tight being allowed to add annotations and how that deadlines. For example, one of the respondents 13 hinders the process. In the words of one of these stressed that the deadlines “do not leave much respondents, “it would be great if it were possible time for research.” Related to this is the challenge to leave some comments for translators to help of rates since templators feel they are not paid them with translation and understanding linguis- proportionally to the time they need to dedicate tic choices in templates.” Such lack of annotation to the template creation. Writing about this, one of field thus seems to hold back the templator-sub- the respondents commented that her challenge was titler communication in pivot workflows: what is Figure 5 Biggest Challenges When Creating Pivot Templates - Theme “Temporal and Spatial features” (32) Medellín, Colombia, Vol. 28 Issue 2 (May-August, 2023), pp. 1-21, ISSN 0123-3432 www.udea.edu.co/ikala Íkala Hanna Pięta, Susana Valdez, Ester Torres-Simón, and Rita Menezes “spending enough time leaving long notes when Some of the above comments seem to relate to the pay is not worth the time spent.” locked templates, and to “unnecessary compres- sions,” (i.e., instances when timing unnecessarily Less mentioned but equally important is the lack reduces the space allowed, for example, by gen- of communication with clients or project manag- erating one subtitle instead of two; Wexler, 2021, ers and with the original creator (two mentions). p. 12). Many ultimate target languages might need “Getting information,” as one respondent points more space to convey the meaning contained in out, is particularly challenging. And to another the English translation. Given the increase of templator, the challenge is “[n]ot having access to characters, the viewers of the subtitles in the ulti- the creator to verify obscure dialogue.” mate target language may also need more time to read the subtitles. If the templator unnecessar- The third most recurrent challenges, mentioned ily compresses the subtitle timing in the English 32 times, were related to Temporal and Spatial pivot template, they might steal precious time features (see Figure 5). Templators (15) reported from subsequent subtitles in other languages. synchronization and spotting as challenges. In this case, the challenge seems to be related to Templators were also asked to rate statements on “timing properly to allow proper cps (characters a 5-point Likert-type agreement scale regarding per second) and cpl (characters per line) for any their beliefs about template creation (based on language the template might be used with, espe- Oziemblewska and Szarkowska, 2020). cially those very different from my own.” On this topic, one templator commented about the pos- The majority of the templators agreed (answering sibility of subsequent translators being negatively agree or strongly agree) with the first three statements 14 influenced by the template saying, “If the transla- (see Figure 6). These are: “I believe that templates tor is not careful, they might rely too much on the should allow for time code adjustment and chang- template and not make the appropriate changes.” ing subtitles” (85 respondents); “I believe that a Templators also referred to shot changes and the good template should contain on-screen text” (66); maximum number of characters per line being and “I believe that a good template should provide technically challenging. annotations” (66). In a lower percentage, 49 of our Figure 6 Templators’ Beliefs Regarding Template Creation Medellín, Colombia, Vol. 28 Issue 2 (May-August, 2023), pp. 1-21, ISSN 0123-3432 www.udea.edu.co/ikala Íkala Pivot Templators’ Challenges and Training: Insights from a Survey Study with Subtitlers and Subtitler Trainers respondents also agreed and strongly agreed that Table 5 Helpful Training on How to Create Pivot templates should only be created by native speakers Templates of the source language. Helpful Training Strategies Number of The statements about how to deal with dialogue Checked by Respondents Responses elicited mixed responses. Forty-seven templators Specific guidelines on template 9 disagreed or strongly disagreed with the statement, creation “When creating a template, I believe it is best to Feedback on expectations and 6 transcribe the dialogue word for word, including advice from fellow translators hesitations and false starts.” Interestingly, when it comes to condensing the dialogue, opinions seem Training on English language 6 to be divided: 34 respondents indicated they agree templates with a focus on annotations or strongly agree, and another 34 indicated they disagree or strongly disagree. Academic training 2 Q&A with pms 2 Templator training Inhouse training 2 As one of our aims was to identify if templators Video training 1 receive specific training (in the broadest sense) on Expert training 1 how to create pivot templates and what type of training they have been offered, we asked not only if they “ever received help, training, guidance or easier than splitting them; not wasting time creating instructions on how to create a template, knowing unnecessary annotations (such as providing a link to 15 that your translation will be used by other transla- a website for proper nouns that translators can easily tors,” but also what type of support they received research themselves or will already know); taking care and to what extent it was helpful. when choosing how to segment your subtitles, e.g. thinking about what translators will be able to edit As expected, little over half of the respondents (52) out and what they will need to include and segment- ing your subtitles accordingly. indicated that they had never received help, train- ing, guidance, or instructions on creating a pivot For two participants, the type of academic train- template. Those who had received some form ing that they received, for instance, in their ma in of training focussed mainly on the most helpful Translation, was particularly helpful, while one type of training received (see Table 5). Nine felt templator made a point of writing that “academia that specific guidelines or instructions on creating unfortunately is not able to provide sufficient templates were helpful, while six highlighted that training for the realities of the market per se.” feedback on expectations and advice from fellow Q&As with project managers, inhouse training, translators was the most helpful. One templator video training, and expert training were also men- even wrote that “colleagues are the best!” tioned as helpful. Other templators expressed that the most helpful Even though most templators focussed on their type of training received was on English language most positive experiences, a small number took templates focusing on annotations. For example, the time to indicate what was least helpful. For one templator wrote: them, general training on subtitling and spot- I think the most important thing is to always bear in ting was described as usually short and superficial. mind the translator who will be using your file, in par- For instance, one templator wrote, “Mostly I have ticular in terms of thinking that merging subtitles is received general training on time cueing (not useful, Medellín, Colombia, Vol. 28 Issue 2 (May-August, 2023), pp. 1-21, ISSN 0123-3432 www.udea.edu.co/ikala Íkala Hanna Pięta, Susana Valdez, Ester Torres-Simón, and Rita Menezes Figure 7 Trainers’ Beliefs Regarding Competences and Task Difficulty of Pivot Templators as I already knew how to).” Online instructions or the data for the 70 trainers who do not teach tem- “on-line guides” were also considered the least help- plate creation. ful, alongside general subtitling guidelines. As shown in Figure 7, most respondents (30) believe 16 Trainers that creating a template with a third language in mind is different from creating a regular template; Seventy-five subtitling trainers from 24 European therefore subtitlers face additional changes when countries answered our questionnaire. Of these creating pivot templates. What is more, Figure 7 75 trainers, 17 trainers reported teaching how shows that, according to most respondents (30), to translate from pivot templates, and five train- translating subtitles via a pivot template is more ers indicated that they teach how to create a pivot complex than translating subtitles via a template in template with a translation into a third language the same language as the audio (answering agree or in mind. This was somewhat expected given strongly agree to the three statements). Torres-Simón et al.’s (2021) findings. Opinions seem to be divided regarding the skills This section draws on the answers of the needed to create pivot templates. While 14 either 70 respondents who indicated that they had had disagree or strongly disagree that “the skills needed experience in teaching subtitling but did not train to translate subtitles via a pivot language template subtitlers on how to create pivot templates. It also are similar to the skills needed for translating sub- compares these 70 repondents to the five respon- titles via a template in the same language as the dents who have trained pivot templators. audio”; 18 either agree or strongly agree with this statement. It is also important to acknowledge that Trainers’ Beliefs the number of respondents that indicated that they “Neither agree nor disagree” seems to be high We asked our trainers to rate a series of belief state- (above 15) and might suggest that they did not ments about pivot templates and training (from have a clear opinion on the matter or that the strongly disagree to strongly agree). We first discuss statements were not sufficiently clear. Medellín, Colombia, Vol. 28 Issue 2 (May-August, 2023), pp. 1-21, ISSN 0123-3432 www.udea.edu.co/ikala Íkala Pivot Templators’ Challenges and Training: Insights from a Survey Study with Subtitlers and Subtitler Trainers Figure 8 Template Trainers’ Beliefs Regarding Competences and Task Difficulty of Pivot Templators Figure 8 presents the view of pivot template train- of the curriculum (seven times), and (5) does not ers regarding competences and task difficulty. involve a specific set of unique skills (three times). The pivot language covered in their training was, in all cases except one, English. The exception was With 14 mentions, time constraints were the a pre-conference workshop on English-French- main reason pointed out by the subtitling train- Japanese translation. The main difference is that, ers that do not teach students how to create pivot 17 in this case, most trainers agreed that pivot tem- templates. In the words of one respondent, train- plate creation is more challenging when compared ers “lack […] teaching hours to reach the subject.” with templates in the same language as the audio. Data analysis also revealed that this “insufficient For most of our trainers (four), it is also clear that time to cover those aspects” was mainly associ- translating via a pivot template is more difficult ated with another factor: one trainer felt “there than translating via a non-pivot template. are more relevant issues to cover first,” and other respondents who mentioned time constraints All in all, their answers do not diverge signifi- echoed this feeling. cantly from answers given by trainers in the first From their answers, it seems that our trainers group. mostly teach the “basics” or a basic set of subti- Motivations Behind not Teaching Subtitlers tling skills. Addressing this issue, another trainer How to Create Pivot Templates wrote, “In order to master template translation and create any kind of templates, I find it impor- We also asked the subtitling trainers that do not tant to first master subtitling from scratch. I would teach how to create pivot templates to describe, in need a larger module to teach creating templates, an open question, their motivations behind it, and which I do not, unfortunately, have.” These com- their answers were quite telling. The main reported ments might suggest that training sometimes aims motivations for not teaching template creation to equip students with skills that are needed for were (1) time limitations (mentioned 14 times), (2) established practices (localizing out of English), not a common practice (10 times), (3) pivot trans- and not so much for emerging practices that are lation should not happen (eight times), (4) not part already on the horizon (localizing into English). Medellín, Colombia, Vol. 28 Issue 2 (May-August, 2023), pp. 1-21, ISSN 0123-3432 www.udea.edu.co/ikala Íkala Hanna Pięta, Susana Valdez, Ester Torres-Simón, and Rita Menezes The second most mentioned motivation, referred subtitlers can “build upon knowledge of template to 10 times, was that pivot translation is not a well- creation in general,” indicating a lack of awareness established translation practice. “Doesn’t seem of the specific challenges of pivot template creation. like a particularly useful skill in our national mar- ket”, “it’s been a rare task”, or “not all players use Overall, the motivations for not teaching transla- templates” are comments that illustrate this point tors how to create pivot templates seem to clash of view. On language combinations, one of the with pivot templators perspectives not only on respondents wrote for instance, “As things stand, the challenges they face when creating pivot tem- there is no call for this in my language combination plates (see Challenges and expectations when (de>en). There is not enough German content creating pivot templates) but also on their opin- that is localised into enough other languages, so ion on what they consider to be helpful training pivot templates are very rarely used”. This last view, (see Templators’ training). As one templator com- in particular, seems to be countered by the data dis- mented, “The more training there is, the better, I cussed in “Templators’ profile”: German is the third never had a training which was not helpful.” most common usl for our respondents. Unfortunately, little was reported by the group The third most mentioned motivation to avoid that teaches template creation in a similar open teaching the creation of pivot templates, referred to question on reasons to teach, so a solid compari- eight times, was that pivot translation should not son cannot be established. Therefore, we can only be practised. From the analysis, this motivation speculate that differences will lie in the intrinsic seems to be stemmed mainly from the belief that and extrinsic motivations discussed above. translators should translate into their native lan- 18 guages only and from a language and culture they Conclusions are proficient in. Three of them, for example, com- mented the following: Our study partly overcomes limitations of ear- lier studies, but it has its own limitations. Despite Example 1: The normalization of templates in the rigorous design, four participants reported certain countries means stealing work from trans- difficulties understanding some questions, so lators who actually have the skills to translate from their answers were not included in the analysis. source languages other than English and who can do a better job (i.e., avoid translation mistakes, Moreover, our data focuses on respondents based understand the cultural references and choose in Europe, with a possible over-representation of for themselves whether or not to keep them, explain respondents from the Netherlands, Portugal, and them, etc., for instance). Subtitling from templates Spain. English pivot templates are a global phe- should be a last resort, not the default. nomenon; therefore, our findings only portray a Example 2: We should translate into our native lan- part of the complex picture of ongoing changes in guages only. Other languages cannot be a concern. the subtitling English-language industry. Still, the study offers valuable insights into avt practices Example 3: The importance of knowing the origi- nal language is pivotal to creating a quality end for English as a pivot language and opportunities product. to reflect on ways to move forward. The two least mentioned motivations were that We characterize, for the first time, templators (their pivot template creation is not part of the curricu- profile and professional practices) and their train- lum (seven times) and does not involve a specific ing. We highlight the challenges and expectations set of unique skills (three times). As for the latter, when creating pivot templates and trainers’ moti- some trainers expressed that they “don’t think it’s vations behind not teaching about how to create different from creating a general template” or that pivot templates. Our data suggest that creating Medellín, Colombia, Vol. 28 Issue 2 (May-August, 2023), pp. 1-21, ISSN 0123-3432 www.udea.edu.co/ikala Íkala Pivot Templators’ Challenges and Training: Insights from a Survey Study with Subtitlers and Subtitler Trainers pivot templates is a common practice. Of the 380 Against this background, and based on the subtitlers from 32 European countries, 100 cre- respondents’ insights, we make suggestions that ate pivot templates, 83 of which also translate academic trainers, industry, and researchers might from pivot templates. It is clear from the data that want to take on board. First, to address the prob- translating from templates is an almost ubiqui- lematic issue of text reduction in pivot templates, tous activity for subtitlers, especially for subtitlers English templates that serve dual purposes should working for SVoDs. Most of these templators be avoided. Instead, two sets of English subti- are seasoned professionals (53 participants had tles should be created. First, the English subtitles more than ten years of experience) with formal should be created for viewers and the pivot texts to qualifications in subtitling, and the majority of be translated further. Second, subtitlers working the respondents (63) have qualifications in sub- from pivot templates should be allowed to adapt titling or qualifications that included modules the spatial and temporal features since these con- on subtitling. For our respondents, English is the siderations can be culture and language-dependent. main source and target language, independent of Importantly, pivot templators should be informed the templators’ native language. Most respondents if the templates will be locked or not since this (64) reported that they typically create templates affects the way they time and segment the content. in English. Among the respondent templators, the most common usl is Spanish (23), followed by Third, to help pivot templators anticipate the English (18), German (14), and French (13). This needs of translators downstream, trainers in goes to show that although English is not the most academia should teach how to create help- common usl for pivot templators, it still occupies ful annotations, so as to make explicit what has an important second place. become implicit in the pivot translation. Pivot language translators reported difficulties in select- 19 Templators from our dataset face various chal- ing what to foreground. Thus, source culture lenges related to linguistic and cultural aspects, subtitlers might be more aware of the elements their professional ecosystem, and the temporal to highlight, but would need to be trained in L2 and spatial features of templates. However, our translation. Multilingual classrooms and cross- findings suggest that teaching the creation of pivot cultural remote collaborations might be a fertile templates is rare, particularly when compared to ground for such training. training in subtitling in general or training on how to translate from pivot templates. Of the Fourth, to address the issue of English lacking 75 subtitling trainers from 24 European countries, many of the grammatical categories that usl and only five teach how to create pivot templates. The utl might have, the use of non-English pivot pivot-template-making training offered comes templates should be studied more systematically, mainly from the industry and not academia. This as these templates might be a valuable alternative is problematic because (a) in-house training does for some language combinations. not typically cover issues related to templator eth- ics or the sustainability of the translator profession Fifth, software developers should optimise the anno- and (b) translator training and professional codes tation fields on their subtitling platforms. Templator of conduct are underpinned by a rationale that annotations should be easily accessed by all trans- prioritises translating into the subtitler’s native lators downstream, facilitating templator-subtitler language (Delabastita, 2008) and, based on our communication. They should not be approached as responses, the key worries of templators are not add-ons stashed away for sporadic use (e.g., for notes addressed in training. to self or for quality control only). Medellín, Colombia, Vol. 28 Issue 2 (May-August, 2023), pp. 1-21, ISSN 0123-3432 www.udea.edu.co/ikala Íkala Hanna Pięta, Susana Valdez, Ester Torres-Simón, and Rita Menezes All things considered, if the aim is to better Association. https://www.ata-divisions.org/AVD/ prepare avt students for current and future pivot-languages-in-subtitling/ challenges, much can be gained from teaching Ferreira, A., Schwieter, J. W., Gottardo, A., & Jones, J. (2016). non-English-language native speakers how to cre- Cognitive effort in direct and inverse translation ate English-language templates. performance: Insight from eye-tracking technology. Cadernos de Tradução, 36(3), 60–80. https://doi. org/10.5007/2175-7968.2016v36n3p60 Acknowledgements Georgakopoulou, P. (2012). Challenges for the audiovisual The authors would like to thank the participat- industry in the digital age: The ever-changing needs ing templators and trainers in this study as well as of subtitle production. JoSTrans: The Journal of Spe- cialised Translation, 17(a), 78–103. https://www. Chris Mellinger and Tom Hanson for their feed- jostrans.org/issue17/art_georgakopoulou.pdf back in the first stages of the survey development. Georgakopoulou, P. (2019). Template files: The holy grail of References subtitling. Journal of Audiovisual Translation, 2(2), 137–160. https://doi.org/10.47476/jat.v2i2.84 Artegiani, I., & Kapsaskis, D. (2014). Template files: Asset Hayes, L. (2021). Netflix disrupting dubbing: English dubs or anathema? A qualitative analysis of the subtitles and British accents. Journal of Audiovisual Trans- of The Sopranos. Perspectives, 22(3), 419–436. lation, 4(1), 1–26. https://doi.org/10.47476/jat. avt Masterclass (2021). Glossary. https://www.avtmas- v4i1.2021.148 terclass.com/courses/template-creation [accessed Heilbron, J. (2010). Structure and dynamics of the world April 2023] system of translation. Presented at the unesco, Bolaños-García-Escribano, A., Díaz-Cintas, J., & Massidda, International Symposium Translation and Cultural S. (2021). Latest advancements in audiovisual trans- Mediation, Paris, February 22–23, 2010. 20 lation education. The Interpreter and Translator Kapsaskis, D. (2011). Professional identity and training Trainer, 15(1), 1–12. https://doi.org/10.1080/175 0399X.2021.1880308 of translators in the context of globalisation: The example of subtitling. Jostrans - The Journal of Spe- Campbell, S. (1998). Translation into the second language. cialised Translation, 16, 162–184. https://www. Longman. jostrans.org/issue16/art_kapsaskis.pdf Casas-Tost, H., & Bustins, S. (2021). Role of pivot transla- Mellinger, C., & Hanson, T. (2017). Quantitative research tions in Asian film festivals in Catalonia. Journal of methods in translation and interpreting studies. Audiovisual Translation, 4(1), 96–113. https://doi. Routledge. org/10.47476/jat.v4i1.2021.85 Netflix. (2022). Morgan Stanley Technology, Media & Costa, D. (2020). Getting out of your comfort zone: Trans- Telecom Conference (transcript). Investor News lating subtitles from foreign audio and an English and Events, March 8th, 2022. https://s22.q4cdn. Template? Deep Focus 6 (Spring), 7–10. com/959853165/files/doc_downloads/2022/03/ Delabastita, D. (2008). Status, origin, features: Transla- Netflix_MORG007_030722a_.pdf tion and beyond. In A. Pym, M. Shlesinger, & D. Netflix Q4 2021 Earnings interview. (2022). [YouTube]. Simeoni (Eds.), Beyond descriptive translation studies Netflix Investor Relations. https://www.youtube. (pp. 233–247). John Benjamins. com/netflixir Díaz-Cintas, J. (2021). avt training and industry needs: Are we on the same page? Guest lecture, Translating Eu- Nikolić, K. (2015). The pros and cons of using tem- rope Workshop (online). plates in subtitling. In R. B. Piñero & J. D. Cintas (Eds.), Audiovisual translation in a global context: Díaz-Cintas, J., & Remael, A. (2021). Subtitling: Concepts mapping an ever-changing landscape (pp. 192– and practices. Routledge. 202). Palgrave Macmillan uk. https://doi. DuPlessis, D. (2020). Pivot languages in subtitling. Audio- org/10.1057/9781137552891_11 visual division part of the American Translators Nikolić, K., & Bywood, L. (2021). Audiovisual translation: The road ahead. Journal of Audiovisual Transla- Medellín, Colombia, Vol. 28 Issue 2 (May-August, 2023), pp. 1-21, ISSN 0123-3432 www.udea.edu.co/ikala Íkala Pivot Templators’ Challenges and Training: Insights from a Survey Study with Subtitlers and Subtitler Trainers tion, 4(1), 50–70. https://doi.org/10.47476/jat. Torres-Simón, E., Pięta, H., Bueno Maia, R., & Xavier, C. v4i1.2021.156 (2021). Indirect translation in translator training: Taking stock and looking ahead. The Interpreter and O’Hagan, M. (2007). Impact of dvd on translation: Translator Trainer, 15(2), 260–281. https://doi.org Language options as an essential add-on feature. /10.1080/1750399X.2020.1868173 Convergence, 13(2), 157–168. Torres-Simón, E. Valdez, S. Menezes R. & Pieta, H. (2022) Oziemblewska, M., & Szarkowska, A. (2020). The quality Are pivot templates to be trusted? An exploratory of templates in subtitling. A survey on current mar- multimethod approach to trust in template-centred ket practices and changing subtitler competences. workflow. Paper presented at est 2022 Conference, Perspectives, 30(3), 432–453. https://doi.org/10.10 OsloMet. 80/0907676X.2020.1791919 Valdez, S., Pięta, H., Torres-Simón, E., & Menezes, R. Pięta, H. (2021). Indirect translation. In Y. Gambier & L. (2023). Pivot language templates: What do they tell van Doorslaer (Eds.), Handbook of translation stud- us about language hierarchies in streaming service ies (pp. 113–119). John Benjamins. https://doi. platforms? Target, 35. https://portalrecerca.uab. org/10.1075/hts.5.ind2 cat/en/publications/subtitlers-beliefs-about-pivot- Pokorn, N. (2005). Challenging the traditional axioms: templates-what-do-they-tell-us-abo Translation into a non-mother tongue. John Ben- jamins. https://doi.org/10.1075/btl.62 Valdez, S., Secara, A., Perez E., & Bywood L. (forth- coming). Audiovisual translation and media Ringmar, M. (2007). Roundabout routes: Some remarks on accessibility training in the emt network. jat indirect translations. In F. Mus (Ed.), Selected papers - Journal of Audiovisual Translation. https://west- of the cetra Research Seminar in Translation Studies minsterresearch.westminster.ac.uk/item/w1v31/ 2006. cetra. http://www.arts.kuleuven.be/cetra/ audiovisual-translation-and-media-accessibility- papers/files/ringmar.pdf training-in-the-emt-network Rodríguez, A. (2017). Netflix says English won’t be its pri- mary viewing language for much longer. Quartz. Vermeulen, A. (2011). The impact of pivot translation on 21 the quality of subtitling. International Journal of https://qz.com/946017/netflix-nflx-says-english- Translation, 23(2), 119–134. wont-be-its-primary-viewing-language-for-much- longer-unveiling-a-new-hermes-translator-test/ Wexler, D. (2021). Top-ten principles of subtitle timing. Deep Focus, 9 (Fall), 11–14. Saldaña, J. (2016). The coding manual for qualitative re- searchers. sage Publications. Whyatt, B. (2019). In search of directionality effects in the translation process and in the end product. Transla- Tesseur, W., & Crack, A. (2020). ‘These are all outside tion, Cognition & Behavior, 2(1), 79–100. https:// words’: Translating development discourse in ngos’ doi.org/10.1075/tcb.00020.why projects in Kyrgyzstan and Malawi. Journal for Translation Studies in Africa, 1, 25–42. https://doi. org/10.38140/jtsa.1.4332 How to cite this article: Pięta, H., Valdez, S., Torres-Simón, E., & Menezes, R. (2023). Pivot templa- tors’ challenges and training: Insights from a survey study with subtitlers and subtitler trainers. Íkala, Revista de Lenguaje y Cultura, 28(2), 1-21. http://doi.org/10.17533/udea.ikala.v28n2a02 Medellín, Colombia, Vol. 28 Issue 2 (May-August, 2023), pp. 1-21, ISSN 0123-3432 www.udea.edu.co/ikala