Dr Ana Frankenberg-Garcia (University of Surry) gave a really fascinating keynote presentation on how researchers are using corpora data to look and perform comparisons between texts that are not translated with texts that are translated.
This was new field to me and it was clear how corpora of texts can provide a powerful insight to how text based languages are typically expressed, especially from texts that were not translated. How words are used and sentences are designed on a large scale is mind blowing, especially when you hear Dr Frankenberg-Garcia explain her work consisting of 8,000 sentences or 100,000s words. The real joy in using corpora data, in my view, occurred when Dr Frankenberg-Garcia looked between languages to see where differences or similarities occurred. Simple examples such as ‘nowadays’ and describing where this word was typically located in English sentences with the Portuguese counterpart (initial, middle or end) or what words were collocated with the description ‘brown’ in English and Portuguese. In Portuguese it was brown dwarf (the star) or brown leather and it was not brown rice or brown hair but wholemeal rice and chestnut hair (much more creative use of language). On a basic word level there is much that can be learned and discussed.
Moving on from looking at texts that were not translated to corpora of translated texts, here researchers can look at trends to explain why a translated text does not look or feel like a non-translated text. Dr Frankenberg-Garcia gave examples of where words were under-represented or over represented, such as the word ‘that’ was used more frequently in translated in texts and obviously a higher level of borrowing. What Dr Frankenberg-Garcia also found was translators would try to maintain similar sentence structures to the original, trying not split or join sentences.
Throughout this presentation it really struck me how little research existed on translation to or from British Sign Language (BSL) or any other sign language in the world. The door is wide open to further our understanding with how translation work is being performed. How do translators work and what are the strategies used to generate translations from or to BSL. To me this is such a valuable field, especially because translators have the time to reflect and create parallel sentences or concepts. What this looks like in reality has yet to be openly discussed.
Admittedly, to our disadvantage we don’t have a software tool that can automatically code video data of BSL so researchers can analyse how signs are articulated, collocated with other signs or how sentences are structured (translated or not). This still has to be done manually and can be a real disadvantage to students
The topic of the next presentation was Google translate among translation studies students from Saadia Elamin (Prince Sultan University, Saudi Arabia). With unrestricted access to an online translation tool students can and do turn to online resources that can help them perform translation tasks, such as homework tasks. Without proper inspection and care, students can over-depend on tools like Google translate and produce translations that are contextually incorrect. Banning the use of online translation tools in the classroom was advocated, however, Saadia did acknowledge how online technology if used with an awareness of what is contextually correct can be beneficial and support learning.
Other presentations critiqued the quality produced by machine translation, which is being used to facilitate and increase productivity of translation output. My impression was that despite some level of progress in machine translation, improvements to how machine translations develop their output, the quality was still missing and unreliable. Trained human translator input or vetting was needed to restore the quality to the intended level.
Great day so far… later this afternoon the conference will look at interpreting related topics. My closing thought would be to encourage anyone interested in conducting research on translating to or from BSL to come forward with your ideas – it is needed!
Proximity Interpreting 