At the AUSIT Biennial Conference in November 2010, Tineke Van Beukering shared her thoughts and practical experience of post-editing machine translation output. It was a great session so I was delighted to attend another more indepth talk by Tineke on the same topic a few weeks ago.
Among other things, she covered:
- the role of the MT post-editor.
- the advantages of using MT + a skilled post-editor (i.e. it’s not just about the money).
- the differences between MT post-editing and checking.
- a typical MT workflow, including other emerging sources of work, e.g. MT optimisation.
- how to make a lucrative income stream out of MT post-editing.
- how and why she got into this line of work.
- what she finds most satisfying about it.
Machine translation: not a question of “good” or “bad”
Something that particularly interested me was Tineke’s discussion of her initial resistance to working with MT output, and how she came to terms with that.
She said it posed an ethical dilemma for her when she was first offered this work, because she didn’t want to contribute to practices which she believed might lead to the end of her profession. These initial concerns reflected many of the usual fears we hear bandied about when the topic of MT makes an appearance, and I was impressed with the way in which Tineke went about testing her fears, before coming to her current and ultimately more informed conclusions.
What constitutes a “real” translator?
I also got the impression that Tineke felt she had to strongly defend her decision to work with MT output to her peers. I sensed that she may have been at the receiving end of explicit or implicit criticism for her choice in the past. (To be fair, I didn’t get a chance to talk to Tineke about this so I could be way off base.)
It caught my attention because I’ve seen this kind of snobbery rear its ugly head in the profession before. And it’s a shame that some translators feel they can look down on peers who apply translation skills in a way that may not fit the traditional “word-in, word-out” model of translation.
To stay relevant, translators need to be able to apply translation-related skills to a range of communicative tasks. I see it as an extension of the poverty-trap mentality which Jill Sommers refers to here, when those who dare to earn a living and adapt to a changing profession are somehow deemed less of a “proper” translator for doing so.
Translation professionals vs. people who translate
I didn’t entirely agree with all the points Tineke made though. For example, she drew some comparisons between the use of translation technology today and the automation of tasks during the industrial revolution, which saw so many workers lose their jobs.
For me, this doesn’t sit well for a couple of reasons, primarily because it compares translators to workers who were often unskilled and who did low-level, repetitive manual labour. While many such workers were indeed replaced by technology during the industrial revolution, some jobs were retained, and many new ones were created for supervisors, managers, skilled technicians, and so on. In my view, professional translators are more akin to these managerial, skilled and/or specialised roles.
Tineke spoke too about the financial benefits of working in this emerging field. This reminded me of an ITI salary survey from 10 years’ ago (the most recent one, as far as I’m aware). Those who reported generating the highest income were referred to as “language professionals”, while all other categories of earners were “translators”. At the time I hypothesised that maybe this was because these professionals applied more to their work than their (not inconsiderable) translation skills, and viewed themselves more broadly than through the prism of translation only.
Crystal-ball gazing has never been my thing. But I’m now more convinced than ever that this is how the careers of professional translators will look in the future. There will always be work for those with highly developed translation skills, but very few of us will work with words in a way we recognise today. This is not something to be sad about, any more than we should feel sad that accountants now use Excel instead of abacuses. Because as our roles become increasingly sophisticated and complex, we evolve, and professionalise, and gain greater recognition.
Above all, we need to be careful with our judgements and our snobberies (and I include myself in that). We need to encourage more translators like Tineke to stand up and share what they’re experiencing at the wordface, without making them feel like their commitment to the profession will be called into question in the process. Because we can’t prepare for the future if we’re not even aware of what’s happening around us in the present.

{ 12 comments }
In his keynote speech at our company’s worldwide summit last year, Tex Texin expressed something similar: How machine translation isn’t something for translators to hate or fear. In fact, it presents the perfect opportunity for translators to improve the scope of their skills and services, and market themselves in more unique ways. He gave “transcreation” as an example. Recreating content in a less technical manner is something that machine translation can’t really handle at this point in time, and it’s an angle that makes professional translation shine.
Of course, there are more examples than that. But I thought he made a good point: Instead of fighting the tides of technological process, people would do well to embrace change and use it to sell themselves and their services more effectively.
Sounds like an interesting speech (and I hope they flew you somewhere nice for the summit!). Certainly the more people within the industry who talk about how machine translation can be useful, the better. It’s a shame it’s still rare to hear translators themselves speak up.
As a labour group, translators are generally pretty slow to adopt new technologies because of the way our profession is structured: as large numbers of us work in relative isolation, we don’t often see new work-related technologies in action, which makes it hard to get a concrete understanding of how they work and might fit into our existing workflow. We also lack strong professional role models to identify with – or for some reason choose professional role models who are curmudgeons and luddites
These factors are very powerful indicators of how quickly a technology diffuses through a group of people (cf. good old Moore & Benbasat. Not friends of mine, but the basis of some research I carried out during my masters degree many moons ago). Hmm, I think there’s another blog post in that…
Very nice column, Sarah. I think you did a good job of summarizing the current situation and pointing at the opportunities involved. And I think your conclusions have a lot more credibility when they come from a professional translation perspective, rather than from a purveyor of MT technology.
As to the industrial revolution metaphor, I think the picture of translator-as-craftsman may be useful. Before machines, even the best craftsmen had to spend long hours on the routine, boring stuff that was the precursor of the final “craft” steps where the real difference in skill became visible. The advent of machines brought about two significant changes. The first, the wide availability of cheaper, machine-produced items, is the one we like to camp on, but even for “mass-produced” goods, there is another evident change. Somebody has to design the items in the first place, and who better than the craftsman, now freed from the repetitive routine? Of course you are right … there are also managers, skilled technicians (another transformed craftsman), etc. that are needed in the new, higher productivity workflows. Most importantly, though, there will always be a market for the work of a skilled craftsman, even when the majority of the market is well-served by lower-cost alternatives.
Keep up the good work!
The translator-as-craftsman metaphor does indeed add another dimension to this context, thanks Bob.
Machine translation is absolutely a positive thing in my field.
I sometime receive a translation order with a machine translation of a Japanese patent attached as a PDF or Word file.
I think it is safe to assume that many clients would not be able to figure out which patents they in fact need to have translated (to have a real translation from a human translator) without MT.
On the other hand, “working with MT” to me sounds like a really horrible way to die. My heart goes out to people who have to do something like that for a living.
I shall keep them in my prayers.
Thanks for the example of how MT helps to contribute to your bottom line, Steve. But gosh – your heart goes out to your colleagues who work with MT? You consider it a horrible way to die and you’ll keep them in your prayers?! Sounds exactly like the kind of attitude I referred to in my post, don’t you think?
I think people who have this “attitude” are expressing a valid point.
That is why I said what I said.
Because I work with MT frequently, I personally feel that if I had to spend my life editing the debris created by machine translation of Japanese patents, I would be eventually be driven to suicide.
I don’t think that “editing MT” is what a translator should do with the rest of his or her life. That’s my personal opinion, not an “attitude”.
I agree, Steve. Every translator is entitled to their opinion, personal and professional.
But regardless of one’s personal opinion, it is unhelpful and unnecessary to share it in such a way as to denigrate the otherwise professional work of colleagues, especially in a public sphere. No-one’s opinion is so valid as to justify that. Yet some seem to think that theirs is, and this is the attitude I refer to in this post.
Thanks Sarah for you great post once again. I would not like to do into this deep discussion by other fellow translators here, But yes i would like to appreciate your effort to put an another good picture of Machine translators to all of us.
Thanks for your Great posting once again.
i think translation is the way to communicate two or more then two persons, its necessary for communication, thanks for sharing information
I have to admit, MT scares me to death, just like a seamstress was probably afraid of advanced and automated sewing technology back in the day. Still, I have clients that ask me about this from time to time. I had one that asked me just a few days ago if I would simply edit a translation that they put through Google translate. I told them yes, but only for the same price as if I had originally translated it because the timing was about the same. I’m running out of things to tell these people because this does seem to be the wave of the future and is changing our jobs entirely. Will we all become glorified editors of work that machines are doing that at one time we would have all balked at the idea of a machine doing? Andy Heath
Additionally, I often get into ethical debates with myself (and sometimes others) over what the true role of a translator is in some regards. For example, there are some things that a machine would never pick up but that I, as a human translator, would. Obviously I can tell my client that their employment application says, “Do you have at least 18 anuses?” when it was supposed to say, “Are you at least 18-years-old? But do I tell the client that their toll free number to get more information has no prompt for Spanish? Andy Heath