Embedding translation technologies into curriculum Mehmet ahin Izmir
Embedding translation technologies into curriculum Mehmet Şahin Izmir University of Economics Department of Translation and Interpretation Translation Forum Russia 2019 August 25, 2019 St. Petersburg, Russia
The Knowledge Gap “Translation education has historically lagged far behind commercial advancements in machine translation, which has contributed to a gap between academia and industry. ” (p. 22) “the gap is actually widening with programs still lingering on very traditional subjects” “So what should be the focus in education? Is CAT tool knowledge still needed in the current landscape of upcoming cloud tools being more and more intuitive and easy-to-use? Or should the focus be more with learning broader skills to fit the new roles we’ll need in the industry, like anthropologists or cultural advisors? ” A Review of the TAUS Global Content Conference in Salt Lake City, UT (USA), 2019
Translation and Interpreting (T&I) education Translation technologies have started to attract the attention of an increasing number of professionals, scholars, as well as students. Each group approach translation technologies from different perspectives, according to their motivations. With an increase in discussions on singularity and transhumanism in the translation industry, it seems appropriate to rethink educational policies and curriculum regarding translation and interpreting (T&I) studies. T&I departments seem to be facing pressure, both implicitly and explicitly, from the market, which is more product-oriented; however, many would dispute the idea that the framework of T&I education should be limited by the demands of the market.
Translators and machine translation: knowledge and skills gaps in translator pedagogy [. . . ] stand-alone modules or courses on translation technologies are insufficient to prepare students for this type of work. Instead, various aspects of machine translation must be embedded in courses across the translation curriculum. Terminology management, controlled authoring, post-editing, and engine tuning can be addressed in translation practice courses so that they develop the necessarily linguistic skillset needed to work with these translation tools. Moreover, they allow the translator-cum-post-editor to take an active role in the use of MT and to employ MT in the service of the translation act itself. In doing so, university translation programmes can bridge knowledge and skills gaps related to machine translation such that translation graduates can successfully serve as post-edito rs. (Mellinger 2017, p. 11)
Understanding the economics of machine translation Machine translation is emerging as a mature technology and becoming a standard component of translation workflows for certain types of projects and customers. As such, new career paths are opening up (PEMT, engine development and tuning; etc. ). Greater dialogue is needed between practitioners, researchers, trainers and educators to identify and develop relevant skill sets and best practices. (Vashee 2013, p. 146)
EMT Competence Framework 2009 - 2017
In the translation industry, technological change has had an ever-increasing impact on the way translation services are performed, though human intelligence, knowledge and skills are still the key factors in delivering quality translations and the growing range of language services which translators and translation companies can provide. Market needs have also evolved, with the continuing expansion of English as a lingua franca creating new needs that can only be met by reversing the traditional “mother tongue” principle in some translation environments. Simultaneously, artificial intelligence and social media have considerably changed people’s relation to communication in general and translation in particular, with machine translation applications and other language tools now commonly available on desktop and mobile devices. This is gradually impacting the translation process and many translation markets, and has changed the perception of translation among the general public and among translation studies students and graduates. Technological and societal changes such as these need to be taken on board in academic translator training programmes, so that future graduates become aware of both the challenges and opportunities that they represent, and can adapt their skills and practices accordingly. (EMT 2017, p. 2)
Translation Technology Service Provision
Grup PACTE (Procés d'Adquisició de la Competència Traductora i Avaluació) 1. 2. 3. 4. 5. language competence cultural, world knowledge and thematic competence instrumental competence translation service provision competence translation problem solving competence (Grup PACTE, 2018)
Instrumental competence
Global scale
Digital platforms
Machine Translation Building, training, and testing MT systems Customized systems Post-editing MT output MT evaluation (human or automatic)
CAT Online resources Translation Memory Systems Terminology Management Systems Translation Management Systems Quality Assurance Corpora
Localization A typical enterprise localization project, for example, can involve the translation of three million words, stored in 10, 000 files to be translated into up to one hundred languages, all to be made available within a very short period of time (Schäler 2004). Content is often multimodal, it can come as text, graphics, audio, or video, and can be stored in a large variety of file formats. Content can be highly repetitive and is often leveraged from previous versions of the same core product. (Schäler, 2010)
Technology in T&I curriculum in Turkey Despite the abundance of undergraduate programs in T&I, it is interesting to see that there is no consistency among universities in regard to the weight that technology should have in the process of educating translators (Şahin, 2013). 85% of faculty members did not have any training on CAT tools (Balkul, 2015). Technology competence is essential in the translation market and yet translators complain about the lack of training in technology (Gümüş, 2013). There is currently no single graduate program/module in translation technologies in Turkey, whereas we witness a growing tendency of graduate studies in the field of localization, machine translation or computer-aided translation in Europe and elsewhere.
Translation technologies in Turkey Nationwide survey conducted in June 2015 271 participants 125 translation students 65 faculty members 81 translation professionals (Results published in Şahin 2016)
Some conclusions & implications from 2015 Survey In Turkey, the number of T&I faculty members competent in using or teaching translation technologies seems limited. Moreover, only a few scholars are actively involved in research on technology, or in gathering information about the needs and developments in the translation market. This suggests a rather unhealthy interaction between the academia and the market, in which the latter tends to take the leading role. T&I departments with a weaker focus on translation technologies may be inclined to take a more market-driven approach to teaching CAT tools and MT.
Developments in MT put the future of translation profession into risk.
Only about 40% of the translation faculty members and professionals believed that having a translation degree is necessary for the PE tasks, a view that was shared by 64% of the translation students. Experience in the translation profession was seen as an asset for PE tasks. Special training for PE was foreseen by only 21% of the professionals and about 28% of the faculty members.
How many courses on translation technologies should be in the curriculum?
Competence in translation technologies is a requirement for translators in order to survive in the translation market.
Computer-Assisted Translation Studies Compulsory course Third semester of the four-year T&I program Three hours per week in a computer laboratory (14 -week semester) Aims at increasing digital literacy & technological competence of students Word processors Online resources Translation memory systems Terminology management
CAT and Localization Elective course Eight semester Three hours per week in a computer laboratory Aims at further increasing technological competence In-depth discussions about history of machine translation Advanced features of translation memory and terminology management systems Localization, Corpora Ethical issues around translation technologies
Surveys - 2015 -2016 & 2016 -2017 Conducted at the end of the course titled Computer-Assisted Translation Studies Views about technology Future of MT Post-editing Revision The quality of Google Translate for English-Turkish language pair
Survey - March 2019 20 students in CAT and Localization class Completed intensive reading assignments Wrote reflection papers Completed post-editing tasks Completed localization tasks
Focusing only on practical aspects of CAT tools is not enough for becoming a successful translator.
Data protection and copyright issues are crucial for translation professionals in the digital age.
Based on my experience in two courses (one compulsory, one elective), I can improve my practical skills on my own.
I feel confident that I have the necessary theoretical and practical skills in translation technologies.
How to embed translation technologies into curriculum The lack of translation and interpreting scholars specializing in technology and low number of studies on this topic is evident. More institutional support is needed for research and self-improvement. More graduate students should be guided to work in this area in order to increase the number of instructors to teach technology courses and to conduct related research.
How to embed translation technologies into curriculum The second important step seems to be the coordinated action of T&I departments and translation companies, which would lead to curricular revisions particularly for technology courses and to a certain degree of standardization among universities. Interdisciplinary research studies involving translation and interpreting scholars, computational linguists, terminologists, and language engineers would produce promising and more trustworthy MT systems and other local CAT tools for Turkish language.
How to embed translation technologies into curriculum Machine translation has not yet exceeded human translation quality and it seems that this will not happen in the near future. The quality of machine translation will increase as higher volumes of texts will be processed to provide more material for the systems. MT quality was not imaginable a decade ago and it is not easy to guess the quality level for the decade ahead. In the foreseeable future, translators are expected to focus more on translation tasks that require more creativity and interpretation. Moreover, they will take additional roles such as post-editor, terminology specialist, localization specialist, etc.
How to embed translation technologies into curriculum As a technological turn in translation studies is becoming a more general topic of discussion, it is imperative to prepare prospective translators and interpreters for the needs and challenges of the new era. The priorities of individual curricula seem to be increasing students’ awareness of and competence in new technologies, developing their ability to use MT effectively and consciously in the translation process and to evaluate affordances and disadvantages of CAT tools. With these priorities in mind, it should be possible for T&I faculty members to persuade their students to consider CAT tools as a natural component of the translation process, rather than as a threat to their profession.
Emerging issues Intellectual property rights Ethical issues Data protection Ergonomics Transhumanism Singularity
An insight from 1960 Why do we not train more translators? It is true that this is an urgent task for our schools and universities. It is also a very long-term and often thankless task, and one which has so far not generally been well tackled, except in a handful of institutions where the approach has been realistic, without premature seeking after literary effects. The advent of machine translation is no reason for the schools to relax their efforts. On the contrary the machines will require the services of a great number of linguists schooled in the best methods of human translation. (Delavenay 1960, p. 4)
Çok teşekkürler! Спасибо большое! Thank you very much! Mehmet Şahin Izmir University of Economics, Department of Translation and Interpretation rbsmsahin@gmail. com
References ● ● ● ● Balkul, H. İ. . (2015). "Türkiye'de Akademik Çeviri Eğitiminde Çeviri Teknolojilerinin Yerinin Sorgulanması: Müfredat Analizi Ve Öğretim Elemanlarının Konuya İlişkin Görüşleri Üzerinden Bir İnceleme. " Unpublished doctoral dissertation. Sakarya University. Delavenay, É. (1959). La machine à traduire. Paris: Presses Universitaires de France. Delavenay, É. & Delavenay, K. M. (1960). Bibliographie de la traduction automatique. The Hague: Mouton. European Commission. (2017). European Master’s in Translation Competence Framework 2017. Retrieved from https: //ec. europa. eu/info/sites/info/files/emt_competence_fwk_2017_en_web. pdf on August 15, 2019. Fırat, G. (2019). Commercial Translation and Professional Translation Practitioners In The Era Of Cognitive Capitalism: A Critical Analysis. Unpublished Master’s Thesis. Boğaziçi University, Istanbul. Gümüş, V. Y. (2013). Training for the Translation Market in Turkey: An Analysis of Curricula and Stakeholders. Doctoral Thesis. Tarragona: Universitat Rovira i Virgili. Retrieved from http: //www. tdx. cat/bitstream/handle/10803/127111/Volga_YILMAZ_GUMUS_Thesis. TDX. pdf? sequence=1 on August 15, 2019. Mellinger, C. D. (2017). Translators and machine translation: knowledge and skills gaps in translator pedagogy. The Interpreter and Translator Trainer , DOI: 10. 1080/1750399 X. 2017. 1359760 Group, P. A. C. T. E. , Hurtado Albir, A. , Galán-Mañas, A. , Kuznik, A. , Olalla-Soler, C. , Rodríguez-Inés, P. , & Romero, L. (2018). Competence levels in translation: working towards a European framework. The Interpreter and Translator Trainer , 12(2), 111 -131. Schäler, R. (2010). Localization and translation. Handbook of translation studies, 1, 209 -214. Şahin, M. (2013 a). "Technology in Translator Training: The Case of Turkey. " Hacettepe University Journal of Faculty of Letters 30 (2): 173 -190. Şahin, M. (2013 b). "Using MT post-editing for translator training. " Tralogy II. Paris: Institut de l’information scientifique et technique. Retrieved from http: //lodel. irevues. inist. fr/tralogy/index. php? id=255 on August 15, 2019. Şahin, M. (2015). "Çevirmen Adaylarının Gözünden İngilizce-Türkçe Bilgisayar Çevirisi ve Bilgisayar Destekli Çeviri: Google Deneyi (Machine Translation and Computer-Aided Translation for English-Turkish from the Viewpoint of Prospective Translators: The Google Experiment). " Hacettepe Üniversitesi Çeviribilim ve Uygulamaları Dergisi (Hacettepe University Journal of Translation Studies) (21): 43 -60. Şahin, M. (2016). Translation Technologies for a Less Commonly Translated Language: Promises and Challenges. Translatologia, 1(1), 22. TAUS (2019). A Review of the TAUS Global Content Conference in Salt Lake City, UT (USA), 2019. Retrieved from https: //www. taus. net/events/conferences/84 -taus -global-content-conference-exhibits-2019 on August 15, 2019. Vashee, K. (2013). Understanding the economics of machine translation. Translation Spaces 2 (1): 125 -149.
- Slides: 41