Anthony Pym
An interview by Anthony Pym with Rei Miyata, Assistant Professor at the Graduate School of Engineering, Nagoya University, Japan, and author of Controlled Document Authoring in a Machine Translation Age (2021): routledge.com/Controlled-Document-Authoring-in-a-Machine-Translation-Age/Miyata/p/book/9780367500207
updated 2 years ago
Four experts have been invited to respond:
Regina Torres-Quiazon (Multicultural Centre for Women’s Heath, Melbourne)
Erika González (RMIT, President of AUSIT)
Leanne Ringwood (Australian Department of Health)
Sharon O’Brien (Dublin City University)
Moderator: Rachel Macreadie, University of Melbourne.
They answer the following questions:
1. Was there a shift in official communication practices during the pandemic?
2. Did the communication needs present a challenge to policies on language and cultural diversity? Were those challenges met successfully?
3. Have there been any significant changes on the level of multilingual communication policy?
4. How should the work of certified translators and interpreters connect with the work of community organisations and mediators?
This event was organized by the joint doctoral program in Effective Translation Policy at KU Leuven and the University of Melbourne.
youtube.com/playlist?list=PLyQOhfgRQAKaWLOXFPi63ZEW52A9R6Nc3
youtube.com/playlist?list=PLyQOhfgRQAKaWLOXFPi63ZEW52A9R6Nc3
youtube.com/playlist?list=PLyQOhfgRQAKaWLOXFPi63ZEW52A9R6Nc3
youtube.com/playlist?list=PLyQOhfgRQAKaWLOXFPi63ZEW52A9R6Nc3
youtube.com/playlist?list=PLyQOhfgRQAKaWLOXFPi63ZEW52A9R6Nc3
youtube.com/playlist?list=PLyQOhfgRQAKaWLOXFPi63ZEW52A9R6Nc3
youtube.com/playlist?list=PLyQOhfgRQAKaWLOXFPi63ZEW52A9R6Nc3
youtube.com/playlist?list=PLyQOhfgRQAKaWLOXFPi63ZEW52A9R6Nc3
youtube.com/playlist?list=PLyQOhfgRQAKaWLOXFPi63ZEW52A9R6Nc3
youtube.com/playlist?list=PLyQOhfgRQAKaWLOXFPi63ZEW52A9R6Nc3
youtube.com/playlist?list=PLyQOhfgRQAKaWLOXFPi63ZEW52A9R6Nc3
How do we set standards for translator-training programs?
What technologies should be taught to translators?
When and how should interpreting be taught?
How can we evaluate translator-training programs?
See the whole course here:
youtube.com/playlist?list=PLyQOhfgRQAKbbcqQMKQ229uY72VldVtke
The interview with Rei Miyata is here:
youtu.be/eVfrtJtEfnI
More information at:
trans-int.org/index.php/transint/article/view/1325
Wine Teseur works on international NGOs, not intergovernmental ones.
More information: univdb.rikkyo.ac.jp/view?l=en&u=100001800&k
More translation scholars: youtube.com/watch?v=991CCfgPsD0&list=PL5DB4C984336E8BE7
More information at masaruyamada.xsrv.jp
Other translation scholars here: youtube.com/watch?v=991CCfgPsD0&list=PL5DB4C984336E8BE7
Recorded March 7, 2018
Tutorial activity:
In your language groups (as much as possible), please translate the following text for one of the purposes listed below (the purpose will be assigned to your group). You are free to seek additional information and consult parallel texts -- you might have to check what an obituary looks like in your language, who is in the photo, or how much your readers might know. Your translation should support the photograph given with the purpose.
Anna Chennault was one of the most influential powerbrokers of the twentieth century. Also known as Chen Xiangmei, she was a stylish fixture on Washington DC's political circuit - but also an unofficial diplomat who skilfully navigated 20th-century politics. She met John F Kennedy, Richard Nixon, Henry Kissinger, Deng Xiaoping, and Chiang Kai-shek. All were reportedly impressed by the woman dubbed by the Washington Post as the "legendary steel butterfly".
Purpose A: An obituary in a major broadsheet newspaper in your language and in one of its countries.
Purpose B: A description underneath a photo in a book of photographs of politicians in the twentieth century
Purpose C: A current history book for young teenagers in your language and in one of its countries
(Each purpose has a different photo with it.)
Recorded January 28, 2021
Anyone is free to use the lectures and activities presented here.
Students who successfully complete this course will demonstrate:
➢ Critical awareness of the ways translators transform texts.
➢ Engagement with the main arguments concerning the ideals of translation.
➢ Critical reflection on the student's own translation practice.
➢ Basic skills in post-editing, translation memories, and subtitling.
Recorded February 4, 2021
Recorded January 28, 2021
The lecture is followed up by an online quiz.
Students are asked to watch it instead of listening to their teacher drone on.
Recorded June 3, 2020.
This is given along with translations of Benjamin by Rendall and Zohn (some of the students have already met the text in their German class). It is an optional reading designed to stimulate thought.
Recorded April 17, 2018
Tutorial activity:
This is a complicated activity. You are invited to work in pairs so that you don’t feel lonely and lost. You have a lot of time for the task, so there is no need to panic.
First, find an online video of a speech in your favorite L2. The speech should be of more than three minutes, ideally without interruption. It should be of someone speaking naturally, not reading, and not speaking too fast.
1. Practice respeaking ("shadowing") the first minute in the SL (i.e. reproducing the text in the start language, reducing it where necessary). Do not spend more than a few minutes on this. Once you think you have the hang of it, record your respeaking of the first minute.
2. Practice a simultaneous rendition of that same minute in the TL. Do not prepare for this beforehand by writing down a translation of the speech, but you may look up a few key terms in order to prepare. When you are ready, record a version of your simultaneous interpreting.
3. For the next two minutes of the speech (i.e. not including the minute you just did), practice consecutive interpreting then record that.
4. For the first minute of the speech, obtain a full written MT translation. For a YouTube speech, you may be able to export a transcript or subtitles, based on a transcript of the original speech, if there is one available. If not, you might like to use the speech-to-text tool in Google Translate.
So here is what you upload:
1. For the first two sentences of all the above (and only those sentences), write out the simultaneous, consecutive, and MT versions.
2. In two or three sentences, compare them, particularly with respect to compression and information loss. Is it true that simultaneous interpreting is most like subtitling?
Recorded March 7, 2018
Tutorial activity:
In your language groups (as much as possible), please translate the following text for one of the purposes listed below (the purpose will be assigned to your group). You are free to seek additional information and consult parallel texts -- you might have to check what an obituary looks like in your language, who is in the photo, or how much your readers might know. Your translation should support the photograph given with the purpose.
Anna Chennault was one of the most influential powerbrokers of the twentieth century. Also known as Chen Xiangmei, she was a stylish fixture on Washington DC's political circuit - but also an unofficial diplomat who skilfully navigated 20th-century politics. She met John F Kennedy, Richard Nixon, Henry Kissinger, Deng Xiaoping, and Chiang Kai-shek. All were reportedly impressed by the woman dubbed by the Washington Post as the "legendary steel butterfly".
Purpose A: An obituary in a major broadsheet newspaper in your language and in one of its countries.
Purpose B: A description underneath a photo in a book of photographs of politicians in the twentieth century
Purpose C: A current history book for young teenagers in your language and in one of its countries
(Each purpose has a different photo with it.)
Recorded March 25, 2020.
Looking at dictation, revision processes, quality control, and export formats, all with respect to a set of subtitles.
Students should be invited to explore further features for themselves, using the many online support materials.
Recorded July 23, 2020.
A few optional and critical comments on the industry discourse on translation memory suites
Recorded May, 21, 2021. A previous version from 2018 is here: youtu.be/Fs3woAarbwo
Tutorial activity:
Working in twos or threes, where possible:
1. Select a poem in a language other than English.
2. Develop a project to translate the poem into English in such a way to attract admiratio. The project should include the ST, a medium for the TT, parameters for the TT language and translation solutions, any accompanying materials, and a strategy for reaching an audience. (The translation can be in any medium you like.)
3. Present part of the project to the class.
Recorded May 20, 2022
Done with considerable trepidation by someone who is only an occasional interpreter.
Recorded September 17, 2020
The descriptors of the styles come from Mossop, "The workplace procedures of professional translators" (2000), working from Chandler (1993).
Tutorial experiment:
1. Pair up with a partner who has the same L2 (where possible).
2. Recording your screen activity and without any help from machine translation, translate Text A into your favorite LOTE at normal speed (this should not take more than 20 minutes).
3. Understand the categories for analyzing translator styles.
4. Analyse your screen recording into numbers of minutes for:
1) Pre-drafting (everything before you type the first word),
2) Post-drafting (everything after you type the last word of the translation for the first time)
3) Documentation (everything outside of Word)
4) Drafting (the rest).
5. Record these numbers on the template.
6. Recording your screen activity and without any help from machine translation, translate Text B at a speed that is 30% less than the translation you completed Text A, then do the same translator-style analysis.
7. Compare your translator styles (as four percentages) for your translations of TEXT A and TEXT B.
8. Depending on the available time, get your partner to revise your Text A and Text B translations using Track Changes. The quality of your translation will be very roughly expressed as the number of changes your partner makes to the text.
Recorded May 9, 2021.
To prepare for the tutorial, make sure you can use a free screen recorder. If you have a Mac, check that you know how to do screen recordings with QuickTime and play them back at x5 speed. If you have a PC and you do not already have a screen recorder, please download BB Flashback and make sure you know how to do a screen recording with it.
Tutorial experiment:
1. Pair up with a partner who has the same L2 (where possible).
2. Recording your screen activity and without any help from machine translation, translate Text A into your favorite LOTE at normal speed (this should not take more than 20 minutes).
3. Understand the categories for analyzing translator styles.
4. Analyse your screen recording into numbers of minutes for:
1) Pre-drafting (everything before you type the first word),
2) Post-drafting (everything after you type the last word of the translation for the first time)
3) Documentation (everything outside of Word)
4) Drafting (the rest).
5. Record these numbers on the template.
6. Recording your screen activity and without any help from machine translation, translate Text B at a speed that is 30% less than the translation you completed Text A, then do the same translator-style analysis.
7. Compare your translator styles (as four percentages) for your translations of TEXT A and TEXT B.
8. Depending on the available time, get your partner to revise your Text A and Text B translations using Track Changes. The quality of your translation will be very roughly expressed as the number of changes your partner makes to the text.
Recorded May 4, 2021.
Tutorial activity:
Working in groups of two, three or four:
Select a multinational company or agency (e.g. the World Bank, Coca-Cola, Apple, Microsoft) and see how its website is localized. Can you make clear distinctions between what is translated and what is localized? Does the company adopt a general approach of standardization or adaptation?
Now choose a national or local website of a company in the same field. How does it use localization and/or translation in its bilingual or multilingual website?
The hypothesis to test: The more local the website, the less it uses adaptation and the more it uses traditional translation.
Write up your answer in about half a page.
Recorded May 4, 2021
Recorded March 7, 2020
Recorded April 10, 2021.
Suggested activity:
Working in pairs, as far as possible: For your LOTE, please give one example of each of the six main solution types (Copying Words, Copying Structure, Perspective Change, Density Change, Cultural Correspondence, Text Tailoring) found in translations of Alice in Wonderland in your LOTE.
Please add a back-translation of your examples so we know what is going on.
Recorded March 24, 2020.
Students have great fun subtitling, so always do it.
Here we were using the really old software Aegisub. You should probably use something more recent like Zoo Subs or Ooona.
For the activities, we usually use clips from Big Bang Theory, but it makes sense for students to locate and propose their own favorite clips. Songs are also an easy way to start.
Recorded March 25, 2020 (the software will have changed since then)
In this introductory course, we just use a simple online TM suite. There is then a postgraduate course that teaches the more advanced suites.
Sometimes we teach MateCat and Wordfast Anywhere (smart students can pick up both quickly), then get students to compare and evaluate the two.
Tutorial activities:
1. Register in Matecat (Links to an external site.) (nothing bad happens).
2. Enter Matecat and set up a project named COVID. Indicate your languages (select "English US" as the SL - not any special kind of English). Under "TM and glossary" select "Create resource" then create a new TM called "COVID". Make sure this TM is marked "Active" when you translate.
3. Upload COVID C / Analyze / Translate.
4. Working in pairs or threes, translate COVID C into your favorite LOTE. (Since the problems will be technical rather than linguistic, you do not all need to share the same TL.) You can use an MT feed: click on the cog to find where to turn the MT feed off and on.
5. Export your translation of COVID C (Go to the download arrow at the top.)
6. Now upload COVID D and translate it into your favorite LOTE, using your “COVID” TM and an MT feed if available (You may have to activate your "COVID" TM - go to the cog to make sure the TM you created ("COVID") is active.
7. Export your translation of COVID D.
9. Export the file "Job TMX" (from under the same download arrow).
For groups who are going well
10. Open Wordfast Anywhere (Links to an external site.) and import the TMX file that you have just created in Matecat, To do this, go to "Wordfast Anywhere" (top left) / Setup TM & Glo / Translation Memories / Upload, and name it COVID. (Check that the language names in this new TM are the same as the ones you are going to translate between.)
11. Go to Project / New / put in the same language pairs as your project in Matecat. Your TM should come up immediately. Click on it to make it Active.
12. Go to your project. Click on the green + button and upload COVID E. Click Upload and Open
13. To start translating, double click in the first TT space. Your previous translation should appear. As you go through the segments, you should see proposals both from your TM and from the MT feed.
14. Export the translation (from under File) and the TMX file (from under Setup TM & Glo).
15. If you are bored and working on Chrome, try the Dictation tool in Matecat (under the cog / Advanced).
Evaluation activity
Working in twos or threes, please answer the following questions as briefly as possible: (This will probably have to be done after class)
For either WFA or Matecat, are your bi-texts being shared with the company that created the tool?
For either WFA or Matecat, how can you apply a spell-check to the segments while you post-edit?
For either WFA or Matecat, do your matches first come from the TM or from MT? Can you adjust this?
Does WFA have a better MT feed than Matecat?
What does TMX stand for?
Can the Dictation tool in Matecat work for languages other than English?
Which of the two TM tools do you prefer? Why?
Recorded March 11, 2018
Recorded August 27, 2020
Recorded February 14, 2021.
Recorded February 4, 2021.