Successfully reported this slideshow.
We use your LinkedIn profile and activity data to personalize ads and to show you more relevant ads. You can change your ad preferences anytime.

Are interpreters and subtitlers better respeakers?

137 views

Published on

Szarkowska A., Dutka Ł. (2016) „Are intrepreters and subtitlters better respeakers” na Languages & The Media 2016, Berlin, Germany, 2-4 November 2016.

Published in: Education

Are interpreters and subtitlers better respeakers?

  1. 1. Are interpreters and subtitlers better respeakers? Agnieszka Szarkowska & Łukasz Dutka Agnieszka Chmiel, Agnieszka Lijewska, Krzysztof Krejtz, Krzysztof Marasek, Danijel Koržinek & Łukasz Brocki
  2. 2. Recruitment • Interpreters • Translators • Controls • (Subtitlers) Respeaking training • 2-day workshops • Working memory tests • Proof-reading tests • Paraphrasing tests Respeaking tests • Intralingual • Interlingual Respeaking quality analysis • NER model • Human raters Data triangulation • Quality • EEG • Self-reported cognitive load • Eye-tracking • Proof-reading • Paraphrasing • Ear-voice span
  3. 3. Respeaking test
  4. 4. Tasks in the respeaking test Intralingual respeaking –Polish to Polish –Four 5-minute clips in randomised order Interlingual respeaking –English to Polish –Speech by President Obama in Warsaw
  5. 5. Speech rate and number of speakers Slow speech rate Fast speech rate One speaker Speech News Many speakers Entertainment show Political chat show
  6. 6. 58 participants Mean age: 27.48 years (SD 5.71) 51 women, 7 men 22 Interpreters 12 Controls 10 Subtitlers 23 Translators
  7. 7. Professional experience 1 3 5 7 9 11 13 1-2 years 3-4 years 4+ years Translators Interpreters
  8. 8. Measuring respeaking performance 1. NER model (NERstar) –Accuracy rate = NER value –Edition errors –Recognition errors –Reduction 2. Human rating on 1-5 scale
  9. 9. Rating  Overall quality  Fluency of delivery  Pronunciation  Punctuation  Spoken to written language  Accuracy of content
  10. 10. Are interpreters better respeakers?
  11. 11. Interpreters had highest NER value out of all the three groups 90 91 92 93 94 95 Interpreters Translators Controls
  12. 12. Interpreters had highest NER value in all clips 90 91 92 93 94 95 Slow one-speaker Fast one-speaker Slow many speakers Fast many speakers NER value Interpreters Translators Controls
  13. 13. Interpreters had lowest reduction rates 0 5 10 15 20 25 30 35 40 45 Interpreters Translators Controls Percentage of words omitted (original speech vs. respoken text)
  14. 14. Interpreters were rated highest in all categories 1 2 3 4 5 Fluency Pronunciation Punctuation Spoken to written Accuracy Overall Interpreters Translators Controls
  15. 15. Interpreters were rated best in interlingual respeaking 1 2 3 4 5 Interpreters Translators Controls
  16. 16. Are better respeakers?
  17. 17. Subtitlers had higher NER value than non-subtitlers 90 91 92 93 94 95 Subtitlers Non-subtitlers NER value
  18. 18. Subtitlers had an overall higher rating than non-subtitlers 1 2 3 4 5 Subtitlers Non-subtitlers Overall rating of respeaking quality
  19. 19. Subtitlers were rated higher in all intralingual tasks than non-subtitlers 1 2 3 4 5 Slow one-speaker Fast one-speaker Slow many speakers Fast many speakers Overall rating of respeaking quality Subtitlers Non-subtitlers
  20. 20. Subtitlers had highest scores in the proof-reading test 0 20 40 60 80 100 120 140 Subtitlers Interpreters Controls Proof-reading test score
  21. 21. Summary of all study results  Interpreters and subtitlers achieved better results as respeakers than other groups – NER – Rating  NER and rating showed similar results  Higher working memory capacity correlated positively with high respeaking quality  Interpreters had lower cognitive load during respeaking  Interpreters were best in interlingual respeaking  Translators/subtitlers gazed longest in the subtitle area and achieved highest scores on the proof-reading test
  22. 22. Conclusions  Interpreters as respeakers?  Subtitlers as respeaking proof-readers?  Respeaking as part of interpreting training  Subtitling as part of respeaking training  Train working memory capacity
  23. 23. Respeaking in Poland in 2016 Live events –Intralingual –Interlingual –Online On public television –One-off events –Respeaking tests
  24. 24. Challenges  Legislation to increase subtitling quotas on TV  Communicating with broadcasters  Introducing respeaking as an independent module at university  Educating proof-readers for respeaking  Educating the general public about the quality of live subtitling
  25. 25. Our papers on respeaking  Are interpreters better respeakers?  Respeaking crisis points  Ear-voice span and pauses in intra- and interlingual respeaking  Cognitive load in intralingual and interlingual respeaking  Tapping the linguistic competence in respeaking: comparing intralingual paraphrasing performed by interpreters, translators and bilinguals  Visual attention distribution in respeaking – an eye tracking study
  26. 26. Acknowledgements This study was supported with the grant no. 013/11/B/HS2/02762 from the National Science Centre for the years 2014-2017 KRZYSZTOF KREJTZ ŁUKASZ BROCKI AGNIESZKA LIJEWSKA AGNIESZKA CHMIEL KRZYSZTOF MARASEK DANIJEL KORZINEK Many thanks to Juan Martínez Pérez and Pablo Romero Fresco for their support in this study
  27. 27. www.facebook.com/RespeakingProject www.avt.ils.uw.edu.pl a.szarkowska@uw.edu.pl lukasz.dutka@uw.edu.pl

×