Notes about the emerging role of Machine Translation Post Editing from FIT2011 presentation

One of the most discussed and attended events for translators during the FIT XIX World Congress were Rosana Wolochwianski’s presentation on Machine Translation Post Editing. Since I believe that Machine Translation is here to stay, whether we like it or not, I wanted to hear more about Post Editing. I found her presentation very enlightening and particularly liked the interviews with current post editors. Here is a summary of what she said.

Translation is changing in the information age. Electronic technologies now allow multilingual and semiautomatic text generation with translation memory systems, content management systems, terminology management systems and machine translation systems. The development has led to cheaper, more accessible, faster translations that are easy to update and highly consistent. This means that translators can produce more in less time.

There are two different uses of machine translated text; translation aimed at assimilation by the user and translation aimed at dissemination (spreading) by the user. For many companies, translations aimed at assimilation (for their own consumption) does not have to have such high quality, but translation aimed for publishing should still be reviewed and corrected by professionals. Professional translators usually always strive for optimum quality. These are called “traditional” translators in this presentation. However, current end users might not always strive for optimum quality.

So what is involved in machine translation post editing?

Rosana divided post editing work into three stages:

-          The pre-editing phase

-          The MT-tool selection phase

-          The MT post editing phase

The pre-editing phase is a necessary stage for successful use of machine translation. This is when you remove typos and spelling mistakes, unnecessary hard returns or format issues from the text. Non-translatable items are also tagged.

To select the right MT-tool, you need to first perform a good analysis of the project details. Is there a reliable and large TM, does the target language have strong syntax requirements, or would a hybrid tool be appropriate, such as a tool trained for a specific client?

There are different types of MT post editing. Complete post editing is when high quality is required; the only aim is to improve speed. Then there is minimal post editing, which oftentimes generates resistance. Finally there is rapid post editing, when you only remove blatant and significant errors and is only for understandability. The two latter types are the ones that “traditional” translators have a hard time stomaching, since we only strive for the best linguistic rendering. ‘’

Pros and Cons of Machine Translation Post Editing vs. Translation

+ Time gain+ Nothing is skipped or repeated+ No typos or spelling mistakes introduced

+ No “blank page effect”

+ Large volumes of work available

- Recurring errors- Typos or spelling mistakes in original are not recognized- Non-translatable elements get translated

– Dull or non-user friendly interfaces

– Constant exposure to flawed language

– Lower quality environment/less creativity

– Unrealistic expectations

– Pressure to lower rates

 

What is needed from a post-editor?

- Long-term commitment to improve both the editor’s skills and the machines skills.

- Innovative responses to MT-errors

- Creative problem-solving

- Good keyboarding skills

- Certain degree of tolerance

- Ability to draw clear boundaries between improvements and corrections (just as a regular editor)

 

What does a post-editor need?

 

- Be able to volunteer for the task and not be forced into the role

- Be given time to learn

- Be heard when they refer a problem

- Receive proper training

- Be paid according to the time and effort applied

- Alternate with other tasks, in order to keep sharp

 

The machine translation post editor is a growing profession. In order for the post editor to thrive it is important that the post-editor is seen as a valuable part of the machine translation process, more as a co-developer, who can provide important information and innovation and improve performance by providing constant feedback, suggesting improvements and developing new solutions.

If you enjoyed this post, make sure you subscribe to my RSS feed!
-- For quick access to a few recent posts:
Screenshot 2014 06 22 13.46.50 150x150 Notes about the emerging role of Machine Translation Post Editing from FIT2011 presentation

Swedish-US business relations

I just wanted to share this updated publication that might be of interest to my US customers… Continue reading the rest of this article...

27 Comments

  1. My two cents. I think it is important to point out that all statistical MT is trained with human parallel data, so in fact, the MT output generated is a “computed” translation based on human translation. MT is quite human, after all, and needs the work of translator to improve its quality.
    The pressure to use MT comes mainly from software developers (rather than agencies)and it is due to different reasons (depending on the company), but lowering the cost is an important factor and this has an impact on the translators’ income (or agencies).
    In my opinion, the problem does not reside so much in the use of MT as part of the localization process with high quality output and discounts, but the use of poorly trained MT with low quality output and with the expectations to obtain very high discounts. This causes a problem similar to that of poor translation memories where certain discounts are agreed but the actually editing of the work takes longer than the agreed discount.

    Reply
  2. Out of purely academic curiosity, let me ask a few questions:

    If CAT tools are acceptable for helping translators do their job, how does that substantially differ from machine translation?

    If you’re willing to edit a human-translated text, why should you be against editing machine-translated text?

    Isn’t a translator a medium in helping someone communicate with someone else? In this light, you could see whatever means of transforming text in one language into a text in another language. The process how this is done should be secondary.

    Don’t you think the market will regulate the extent and value of machine translation? If clients find MT + post-editing more cumbersome, more expensive and lower-quality than human translation, there’s not much of a threat to the profession.

    Isn’t there always the possibility to decline agencies’ post-editing job offers? Go for the clients directly.

    I totally understand that when one’s livelihood is at stake, the arguments tend to become acrimonious. Then again, if we see translation as a business rather than art, we just adapt in ways that help us earn more. There will always be changes in the marketplace that are beyond our powers to stop.

    Thank you for your views.

    Reply
    • Rats, forgot to delete the sentence in the middle of the fourth paragraph:

      In this light, you could see whatever means of transforming text in one language into a text in another language.

      Sorry for my forgetfulness. My thought went in another direction but that incomplete sentence remained.

      Reply
  3. Well, just to wrap up some ideas from this post, I think everything boils down to common sense. There was a movement in 19th Century England called “luddism”. Luddites hated machines and were strongly against the Industrial Revolution. The even led riots destroying mechanizes looms. In the light of history, we must accept they did not succeed. We cannot hide the sun with our hands, economy and technology have evolved to a point with no return. No individual or Association effort can prevent MT from existing, but they can spread the word and the knowledge to make our colleagues aware of what is going on, thus protecting them. Our profession is flexible enough, and MT is already usable but not that efficient yet, and won’t probably be much better in the near future. So, that said, there will still be room for translators to work as most of us do today, adopting technologies like CAT tools and the like, but no MT. But if some translators want to give it a try, it is better to know what MT is and how it works, so they can benefit from it, and not become victims of it. Knowledge and debate protect us, they do not go against us. No translator in his senses should accept a job that any client around the corner preprocessed on Google. But someone who works for direct clients might want to have his own MT engine to produce more in less time. And there are large translation companies that are working seriously, investing in MT for vertical-industry, high volume projects, investing in engineering, in training for post editors, etc. So, if a translator wants to get involved in a project of that kind, doing it proactively, in order to find it more rewarding, both economically and for his own satisfaction, is an intelligent approach. If post editors make their voices heard, MT software developers might be able to improve their products and make them work more smoothly, for the benefit of all those involved. On the other hand, if a translator starts getting offers for MTPE, the more information he has, he will be in a better position to tell the wheat from the chaff, and reject those projects that look unfeasible or counterproductive, and company owners will know that the translators’ community is educated enough to know what projects are feasible and which are not, so they won’t impose MT in any scenario. Personally, I am not an active posteditor at the moment, simply because I still have plenty of satisfactory translation work and I enjoy it. However, I’m sure I would be well prepared to take PE jobs and make them work satisfactorily for me, if it were the case. Those are my 2 cents! Best, Rosana (rosana@rwtraducciones.com.ar)

    Reply
    • «There was a movement in 19th Century England called “luddism”. Luddites hated machines and were strongly against the Industrial Revolution. The even led riots destroying mechanizes looms. In the light of history, we must accept they did not succeed.»

      Whenever colleagues question post-editing, this is the first argument: that they are against technology. We all know very well that having a strong technological leg these days is as important (sometimes even more important) than having a strong academic leg.

      «We cannot hide the sun with our hands, economy and technology have evolved to a point with no return.»

      Opposing to something does not mean hiding realities. We all profit from new technologies. Some colleagues are certainly using MT and later post-edit. I remember a colleague once said he had finished a translation in which he had used Trados + Dragon + Systran. Another colleague said: “hey, and when did you translate?” :D :D :D The important thing is that his final product was good for his client, and his investment in the software gave him the possibility to correct better and/or to have free time to… go and play tennis. As you can read, there was no greedy agency involved. Nobody is against post—editing tools per se. We all welcome new technologies. (We = translators). The undeniable fact is that technology is giving more profits to agencies (CAT tool discounts/post-editing) than to translators.

      «MT is already usable but not that efficient yet, and won’t probably be much better in the near future.»

      Well… It will much depend on whether our colleagues feel that the report is ok (“long term commitment to improve the machine skills”) or whether they think, like Katherine and myself among others (hopefully thousands), that translators should not improve MT software as it is completely counterproductive.

      «But if some translators want to give it a try, it is better to know what MT is and how it works, so they can benefit from it, and not become victims of it.»

      Translators must know about everything going on in the profession. There is no doubt about that. Translators must know about MT, but first and foremost to decide if they have no ethical objections to it (in view of the threat post-editing represents for the profession). ATA’s Code of Ethics says: “to ask for and offer due recognition of our work”. If I post-edit a machine, am I not going against this? Cleary, a translation has been done, but a translator has NOT been assigned the translation job. A machine has been in charge. Is that due recognition of our work?
      The reports I have read so far do not give an objective analysis, or the “blank page effect” would not be included. I can understand that a software vendor will include such a weird (can’t find a better adjective) argument, but an objective report prepared to enlighten fellow colleagues should not. Why aren’t the points I mention (regarding professional status of the profession) included as cons? That’s key. That’s the underlying concept we (translators) should be discussing. All the reports I have read so far are biased. I totally agree that we have to know as much as possible about post-editing, but we should make an objective analysis about the real threat for the profession. All reports mention “the human part” in a very vague way, just to include the idea (“oh, yes, we do care”), but later say things like there will be no blank page effect (!) and translators will feed the machine (!). Again, let’s remember that when we correct a machine, there was a translator that was not assigned the job. When you propose MT to your client you are destroying the efforts of hundreds, thousands of translators along the years to get social professional recognition (my main concern above). ATA’s Code of Ethics says: “to ask for and offer due recognition of our work”. Where are those ideas in the reports?

      «Knowledge and debate protect us, they do not go against us.»

      That’s why we are discussing.

      «But someone who works for direct clients might want to have his own MT engine to produce more in less time.»

      Of course! Like we use double or triple monitors to improve our work, CAT tools, voice recognition… No agency involved, right? Is that what you mean. I totally agree.

      «And there are large translation companies that are working seriously, investing in MT for vertical-industry, high volume projects, investing in engineering, in training for post editors, etc.»

      Of course! Are they investing because they are Mother Theresa and want to help software vendors raise their earnings, or are they aware of the much larger margins they will get if they succeed in convincing translators that post-editing is a niche?

      «So, if a translator wants to get involved in a project of that kind, doing it proactively, in order to find it more rewarding, both economically and for his own satisfaction, is an intelligent approach.»

      More rewarding economically? Can you explain how? More rewarding for his/her own satisfaction? Can you explain how? Even if post-editors were paid full translation fees, my concerns posted above are still valid.

      Thank you, Rosana.

      Reply
      • Aurora, I’m glad to see you moved from “In fact, why are we talking about post-editing? What is the need? ” to “debate is good, that’s why we are discussing”. In my FIT presentation, I did analyze the human factor, the psychological implications, etc. I will write an article soon about it and include that information, so you have no right to tell I neglected that. I also devoted a COMPLETE slide to “Gaining professional recongition” towards the end, not just a bullet in a summary, so once again you are debating with me on the basis of an extract somene else produced, without having been present in my presentation, and I find that almost offensive, because you are putting on my shoulders all your feelings against PE, and slandering my presentation which you never heard, you have a biased view of my work, which was certainly balanced, I also included the PE’s complaints.
        We all know agencies work for a profit. You agree with MT when it is used by translators (for a profit), not when it is used by agencies. We are working in a free market and we cannot tell people who should use what, and every individual is free to accept or reject a job. And “more rewarding economically” means freelance translators who can make a higher hourly rate postediting than translating, because they develop a high speed (over 1,000 words) and are required to deliever lower quality work, that also happens sometimes, I have their testimonies, and I won’t nag them for it. I did not invent the “blank page effect” comment, post editors told me about it in their interviews, they are quoted with name and surname in my work, so do not libel without having the necessary knowledge, my work is serious. It was nice to listen to your views, but I cannot understand how you dare discuss with someone on the basis of a presentation you never attended, and put comments or omissions in their mouth, I told you from the beginning that I spoke for 60 minutes on this topic, and you were not there. I cannot publish my whole presentation online for you to see it, you will have to trust my word or someone who attended. And talking about freedom, we are not obliged to agree, but I won’t accept any more ungrounded claims. Rosana

        Reply
        • I would like to close off this discussion with a small disclaimer. My notes were a summary of what I took with me from Rosanas presentation. The presentation included a lot more nuances, especially comments from other translators. Post editing is here, whether we want it or not, but fortunately we do not have to participate if we do not want to, and there are many areas of translation where post editing is not used. On the other hand there are other areas where machine translation and post editing is beneficial for the purpose of the translation. Rosanas presentation was taking up many more aspects than I was able to include in my blog post.

          Reply
  4. «Aurora’s focus seems to be that some agencies may use MT where a human translator would be better, and then ask a translator/post-editor to fix the translation at a vastly reduced rate, when the translator really should have done the translation in the first place, and received a higher rate.»
    It’s no surprise that one of the first agencies in Argentina (may be the first one) to organize an event presenting post-editing (and selling it as “the” niche) is an agency in Rosario very well known for the lousy rates they pay. They clearly said that post-editors (formerly known as professional translators) will be asked to submit a report together with the translation.

    «I would suggest that agencies that do this should be avoided by anyone like Aurora who values the human/intellectual input part of translation, and wants to provide quality work.»
    I would suggest all professional translators avoided them (the objective is to increase their already LARGE margines), and all associations warned the community of professional translators, based on the dangerous underlying concept (destruction of a professional status we have been fighting for for many years).

    «But I stand by my original comment that there are some jobs where translators aren’t being used currently, where post-editing could be a way for translators to get their hands on additional work that they wouldn’t otherwise have access to.»
    I have read this argument more than once, and am totally against it. First: if a certain company hates to spend on professional translators and resorts to Google or the manager’s secretary’s cousin, that company simply does not want to pay a single cent. They will never do! Also, if a company (currently unware of this possibility) finds out that it can start saving money, it will immediately share with peer companies in its sector. In difficult economic times as we are living, I guess a HR guy or Procurement Dep. guy who brings the good news (guess what I learnt from a peer in the World Congress of XX?), will be adored by his/her company’s Management for the potential cost reduction that could be achieved. Do we want this?

    «I generally wouldn’t advocate MT and I won’t be out there marketing my post-editing skills over my translation services, but I think putting your head in the sand and refusing to discuss MT is not going to make it go away.»
    I agree, but let’s not be the ones who are promoting this.

    «to educate users and clients on the need for human involvement in MT.»
    In this moment, most of our clients are only using Google translate to understand an email of a cool Hungarian guy they met last summer. They have no clue about things we do know (our tools). I am not planning to teach my clients what MT is. Why?

    «Also, I don’t agree with translators improving MT software, either! That seems completely counteractive.»
    It’s what they expect (it’s in the presentation and it’s what this Argentine agency explained).

    Nice to “meet” you, Katherine.

    Reply
  5. “A French anthropologist called Marc Auge [2000] points out that it is at the moment we develop our writing abilities, that we discover the subtleties of reading. We can all agree that this is certainly true. When we learn how to read, we do not get stuck by the differences between an “s,” a “c,” or a “z,” between “v” and “b,” we just go on reading. But it is when we intend to write, to produce, that we start to doubt, “which is the correct letter to use here?”, and we become aware of the subtleties of language.
    I think there is a possible analogy with translation work here. We learn to translate, by translating. It is by deciding creatively, each and every time, by making mistakes once and again, that we become well-seasoned translators, we acquire that subtlety that makes us good translators. So, if a novel translator enters the industry as an editor of material which has been preprocessed by an automatic program, will he really be able to acquire that subtlety? Isn’t it possible that the first time he notices a strange expression he changes it, the second time he thinks “this sounds familiar, I saw it somewhere else,” and the third time he already assumes “this is the way it is usually said”? What kind of translators will get formed in such a process? What will the threshold of quality be in the future? Is Creativity at Stake?”

    This is also a quote from my presentation, so please do not attack the messenger! :) I am not telling people whether they should embrace MTPE or not, that is an individual choice. I just did extensive research on the topic for 6 years now, and I’m presenting the reality I found. I analyze both pros and cons. I’m truly involved with the human aspect, I even travelled long distances to interview posteditors in different cities and listen to their voice, and then transmit it. All my work is independent and unfunded, I do it out of a personal and intellectual interest, I am just transmitting my findings and reflections. Thanks everyone for your interest. Rosana (rosana@rwtraducciones.com.ar)

    Reply
  6. Aurora’s focus seems to be that some agencies may use MT where a human translator would be better, and then ask a translator/post-editor to fix the translation at a vastly reduced rate, when the translator really should have done the translation in the first place, and received a higher rate.

    I would suggest that agencies that do this should be avoided by anyone like Aurora who values the human/intellectual input part of translation, and wants to provide quality work.

    But I stand by my original comment that there are some jobs where translators aren’t being used currently, where post-editing could be a way for translators to get their hands on additional work that they wouldn’t otherwise have access to. Yes, the rates will be lower than for translating, but if the MT is good (by MT standards), the job will require less work than a full translation, so the rate should be lower.

    If it requires a lot of work because it is poorly done, then the rate should be raised. Sometimes an hourly rate can deal with this issue.

    I generally wouldn’t advocate MT and I won’t be out there marketing my post-editing skills over my translation services, but I think putting your head in the sand and refusing to discuss MT is not going to make it go away. And in fact, it may encourage and spread its use, and the exploitation of translators, by failing to educate users and clients on the need for human involvement in MT.

    Also, I don’t agree with translators improving MT software, either! That seems completely counteractive.

    Rosana, thanks for the links to your articles, I look forward to reading them (hopefully they aren’t *too* depressing!).

    Reply
    • Hi Katherine,

      I think you have a clear and balanced view, there is not much I can add. In certain scenarios, for materials that would otherwise not be translated at all, either for lack of time or budget, if certain conditions are met, such as:

      feasibility studies or pilot projects run (NOT any project)
      fair payment and specific training for post editors
      appropriate pre-editing phase in place
      clean TMs and adequate dictionaries applied
      reasonable volumes expected per day
      post-editors rotation in-house positions
      friendly user interfaces
      bugs solved or specific language macros implemented to facilitate correction,
      etc.
      Some people (not necessarily current translators, but eventually people who prepare themselves specifically for this task), might find it acceptable as a new type of work. Instead of translating 300 words an hour, they will gladly post edit 1000 words an hour at a lower rate, making the same money per hour.

      As regards the post editor as a “co-developer”, my suggestion was that, as the job is less creative (and by the way, it is also less creative with ANY CAT tool), people involved in it might find themselves more attracted if they contribute ideas and suggestions as regards how to improve everyday bugs and problems, that might make them feel more involved and less like “an appendix of a machine”, more in control of the process. That has to do with the subjective aspects of the job (any job, after all). Rosana

      Reply
  7. Aurora, I also have deep love for my profession, and I am not just “in favor” of post editing, you are speaking without having listened to me or read my production. I just present the problem and analyze all aspects, with a deep worry about the human aspect of this trend. If you are interested, please read my articles and we can then discuss anything you want.

    Reply
    • Rosana: if I say that I know your position it’s because I have read your articles. I am more than willing to discuss my points mentioned above.

      Reply
  8. Hi Aurora,

    It would take too long to answer such broad questions in a post. If you are interested in my vision on the topic, I can recommend you 3 of my published articles on MT and automatization:

    Article: “What is really at stake with machine translation? An overview of its impact on the different stakeholders,” in The ATA Chronicle, A Publication of the American Translators Association, Vol. XXXVIII, N° 4, April 2009, pages 26-31. (similar content digitally in (http://www.ata-divisions.org/LTD/documents/newsletter/2008-3_LTDnewsletter.pdf).

    Article: “An Overview of the Impact of Machine Translation on the Professional Translation Community”, Language Tech News, Newsletter of the Language Technology Division (LTD) of the American Translators Association (ATA), Vol. 4, N° 2, October 2010, page 12 (http://www.ata-divisions.org/LTD/documents/newsletter/2010-2_LTDnewsletter.pdf).

    Article: “English>Spanish Translation in an MT environment”, in Multilingual Computing & Technology. September 2008 issue, pages 38-44. (I do not have a digital address, but if you are a subscriber, you might have access to previous content).

    Rosana

    Reply
    • Hola, Rosana. My questions are not broad at all, on the contrary, they are very simple and to the point. It’s ok if you don’t want to answer. Regarding your position on post-editing, I disagree with you 100% based on what I have expressed above. My position has to do with the immense love and deep respect I have for my profession.

      Reply
  9. Rosana Wolochwianski is a freelance translator. She does not represent an agency at all. Her work as a translator and researcher is completely freelance and her opinions are only based on her own experience, the interviews of post-editors and her extensive research on the topic.

    Reply
    • Hi all!
      The mere concept of post-editing should be rejected (condemned) by the whole community of translators (and by all associations defending us). I understand it will be accepted (and promoted!) by the relevant vendors and some agencies looking for larger margins, but not by translators. There are many (if not all) things to be questioned from the list of”pros” of post-editing (pros for us, translators, I mean). The “blank page effect” is a very weird concept. Show me one translator for whom there is such thing as the “blank page effect”. And… if 3 or 4 post-editors are fine with this lethal tool, that means nothing to us translators. I am tired of listening to Argentine translators saying that the 0.03 USD their Argentine agencies pay (sometimes less) are ok. These testimonials mean nothing to me. I don’t need to read them to know what post-editing is about or what it represents for our future. Among other red lights in this new “niche” that vendors and some agencies are inventing (to sell their software and to get better margins, respectively), this is my favourite: the editor will improve the machine skills. I’ll be back later with more comments. Regards,

      Reply
      • I would like to point out that I am not against post-editing per se (for instance, if used by my translators, and provided they offer good quality translations). The post-editing software would be another tool for them (them=translators). It will probably not work for me (I need to be the owner of the document’s impronta), but might work ok for others (same as some colleagues can’t imagine translating without using voice recognition tools, while others don’t like them or have never tried them). I reject the idea of agencies getting even higher margins. I reject the idea of talking with my clients about human or non-human translation. (!) That would make me go back years in my efforts to educate them. “We the translators are X, Y, Z” (praising our intellectual work). New message: “you know what, a machine can now replace us.” The underlying concept hurts our profession and emphasizes the lethal idea that translation is a commodity. And the obvious: if 100 post-editors help the hangman (by teaching him), very soon only 50 will be needed, then 10, etc… All the rethoric about post-editing should not confuse fellow translators. In fact, why are we talking about post-editing? What is the need? Or better, who is interested in this becoming a “normal” topic? Let’s say “no” to post-editing jobs. None of my clients (local or international / direct clients or agencies) have any idea about this. And this happens to the many colleagues I discuss with. In times when we are doing huge efforts to gain professional recognition, how can we be talking about post-editing “spontaneously”?

        Reply
        • I agree with Aurora. I’m not against anything that speeds up the work of the translator, as long as the end product is a quality translation. However, I would never accept jobs post-editing MT because it takes longer to fix a bad translation than to do a good one from scratch. I enjoy translating texts from scratch, and post-editing takes that away from the translator. I didn’t go to university to be correcting machine translation! I want to translate!
          Many agencies ask us not to charge for 100% matches in translations done with CAT (but they still want us to check these matches). I think post-editing will be the next step towards reducing the prices for translations. Let’s face it most agencies will use post-editing to pay extremely low rates to translators. Why on earth are some translators promoting/defending this is beyond me!

          Reply
          • Thank you for your comments! I know this is a hot topic and that is part of the reason I found Rosanas presentation so interesting. I look forward to reading her articles too.

  10. Hi Aurora! I’d like to point out that I AM a freelance translator, not an agency. In my presentation, I presented both the prons and cons of MT PE. Tess included just a summary of a 60 minute-presentation, highlighting the aspects she found more interesting, but the original presentation analyzes many more nuances. Rosana

    Reply
    • Thank you Rosana and Gabriela for pointing this out and thank you Rosana for a very interesting presentation.

      Reply
  11. I see what Aurora is saying but I am not sure it is that straightforward. I don’t know how clients are using MT, and maybe it is a threat to translators in its current form, but I would suggest that being a post-editor can work out for individual translators and the profession as a whole, if done thoughtfully.
    For example, you may want to suggest to your client that the text be translated by a human rather than a machine, thereby “saving” translators from obsolescence. But in other cases, a machine translated text may be being used or distributed by a client in its ‘raw’ state, and as a post-editor, you may be able to convince a client to have a human translator (you) work on the text to improve the quality. This not only creates jobs for translators, but also improves the overall quality of translations in publication, helping the industry as a whole.

    I think this relates to what Rosana was saying about becoming a co-developer, so that you can have input with the client into what should be human translated, and what should be post-edited (and what requires neither).

    Tess, did Rosana explain at all how this level of input could be achieved? e.g. would you have to work with a client as they develop their own MT system, or go work for an agency as an employee? How could freelancers ensure they were fully involved in the process?

    Reply
  12. Thank you, Tess! It would be great to discuss a presentation like this one, but presented by a translator (and not an agency). We all know that post-editing is a threat for the profession. Whenever you post-edit a machine it’s because a translator has lost a translation assignment. Whenever you post-edit (and delivery translation + report) you are feeding the machine. Thus the “niche” they are trying to sell us, becomes a non-methaphorical niche… and game over to the profession. I wonder what other professionals would help accelerate their death. Big no to post-editing. Best,

    Reply
    • Tess, Aurora, Rosana, I’ve read most of your postings here. I don’t see any reason for such a virulent reaction against MT post editing. Aurora, why not admit that you are against MT post editing instead of saying sweeping generalizations? Statements such as “We all know that post-editing is a threat for the profession? Whenever you post-edit a machine it’s because a translator has lost a translation assignment. Whenever you post-edit (and delivery translation + report) you are feeding the machine ” reveal an emotional outburst and are bound to be challenged by those who don’t share it.

      Whenever I read tags such as “everyone knows” or “we all know” I read unsupported claims. It’s better to nuance our message for better effect and to qualify our statements without sounding so emotionally invested. This is supposed to be a dispassionate discussion over a thorny subject.

      Yes, I am a freelancer and I don’t like doing MT postediting.

      Also, our strong emotions sometimes make us insert typos and other inaccuracies in our message. Can machines be post-edited? Where’s the data to support the claim that a translator loses an assignment because the job has been processed by MT? More importantly, why the attacks on unnamed translation agencies on the basis of low rates? Many translation agencies are headed by former freelance translators; their margin rates have plummeted along with ours, based on my own interactions with clients. Why demonize them? Why assume there’s a conspiracy against us translators?

      I am glad Tess reported on the presentation. I wasn’t there, and having a colleague’s opinion on such a difficult subject is very helpful. I liked the box with the pros and cons, which painted a clear (if not complete, I can’t say) picture of MT post editing and translators. I don’t care about performing MT post editing and even I was impressed.

      Reply
  13. Hi Tess!

    I have a question, apart from this presentation, was there another one about post-editing but presented by a translator (which, of course, would point out no advantage whatsoever of post-editing)? Or was there room only for an agency? Thanks!

    Reply
    • No, I did not attend or see any presentation about post-editing held by a translator. However, in this presentation, the presenter had interviewed 3-4 people that worked with post editing of machine translation and it was interesting to hear their point of view (movie clips). The presenter, and my summary had also taken their comments into account.

      Reply

Submit a Comment

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>