Of tomatoes, translators and the importance of context

Car flashing headlights
Sometimes technology doesn’t make life easy for translators. Sometimes translators don’t make life easy for themselves either. I was recently trying to explain to a fellow translator why I always insist on receiving the actual formatted source text before starting work, but I think he thought I was just being awkward and a little bit old fashioned. I mean seriously, who, in this day and age, needs to see what a text will look like when it’s printed? He didn’t say as much but I could detect this young hipster’s incredulity at my “old-world” approach.


Now it might well seem old-fashioned and a little pernickity, but the fact that the files to be translated were in XML format and translated using a TM tool meant that, without the formatted source text, the translator was effectively translating blind. The same applies to any text you translate using translation memory tools.

Example of text in XML

A typical text in XML format

When you’re dealing with a text like this, it’s all too easy to miss out on the kind of contextual information that would otherwise help you to make appropriate translation decisions. If you can’t tell whether a segment is a caption for an image, a chapter heading or even an ordinary paragraph sentence how do you know what’s the best way to translate it? If you see the word Garderobe in the text, but can’t see the photograph that goes with it of a sign on a wall with the same word, how do you know not to translate it? You need to know what came before the sentence, what comes after it and what is around it before you can properly understand and translate it.


I always thought it was obvious that the context of a sentence is important when translating but apparently not everyone shares this view. But how do you explain a fairly nebulous and intangible idea like context?

Car headlight controls
After a couple of rounds of emails, which didn’t seem to be getting us anywhere, I started thinking of how people communicate with each other by flashing their car headlights. This simple act of flicking a switch several times usually means “go ahead, you can pull out in front of me“. When the recipient of your good deed thanks you by flashing their hazard lights, you flash your headlights again to say “you’re welcome” – same sign, same interaction but slightly different context.
Other times you might flash your lights at someone who is driving too slowly, driving on the wrong side of the road, or at someone who cuts in front of you in fast-moving traffic and the meaning is similarly clear. Some people even warn other drivers about a speed camera by flashing their lights. The number and frequency of flashes may change but the same basic “utterance” has a number of different meanings depending on the context.


To complicate matters, this simple language (which has been likened in sophistication to the communication skills of insects) is also culture-dependent. In France, apparently, flashing your lights means “stay where you are, I am not letting you pull out in front of me“. But you can still understand what each flash means because, assuming you’re paying attention, you are aware of the context and what is happening around you.


Although it’s probably a silly example that has nothing to do with translation per se, it does sum up the idea of context. Pretty much anyone can translate a sentence, but not everyone can come up with the right translation at the right time. To paraphrase the late Miles Kington, knowledge is knowing that a tomato is a fruit; wisdom is knowing not to put it in a fruit salad. Let’s hope my translator friend figures it out too.


The Rube Goldberg Approach to Translation

As a general rule I don’t have too much of an issue with doing test translations for new clients or agencies. Some really good agencies even pay translators for these tests. The logic behind test translations is pretty sound because it doesn’t matter what you have on your CV, the proof of the pudding is in the eating and the only way to gauge how good you are as a translator is to see a sample of your work.

Recently a translation agency made contact with me and, after the usual introductions, asked if I would like to work with them on technical translation projects. After much to-ing and fro-ing we were both satisfied that we could work together. I outlined my experience and specialisations, they described the type of projects they get and we both agreed that there was great potential for future work. Hell, we might even get matching tattoos. Rates and terms were agreed without so much as a whimper. Everything was going well until the issue of test translations cropped up.

The epitome of inefficiency - a Rube Goldberg Machine

The epitome of efficiency?

This agency had a pretty weird approach to test translations. Instead of a test translation at the start of our working relationship, which is the customary thing to do, they announced that they ask translators to complete short test translations for each project. The translator who produces the best test translation is then awarded the particular project. I read and re-read the email several times to make sure I understood what they were proposing, hoping that I had misunderstood. But no, when they received a job from a client, rather than contacting a translator who they had already screened, tested and knew was suitable, they would start a mini screening process before work could even start on the translation. I can imagine this being pitched to customers as a way of ensuring that “only the best translators are used” for their projects, and it’s probably well intentioned, but does anyone seriously think this is good practice?

Aside from the startling inefficiency of this Rube Goldberg approach to operations management, which disadvantages both customers and translators alike, things became even more interesting when I queried whether translators would be paid for producing numerous test translations over an extended period of time. The hitherto brisk and prompt exchange of emails suddenly ground to a halt. I never received a response and can only assume that it confirms my worst suspicions.



Hold your horses

Buried under a mountain of paper

No problem, it'll all be ready by morning (Image via Global Nerdy

I received an enquiry from a potential client asking whether I could verbally translate 400,000 words of specialised texts into English, on-site and over the course of 8 days. Now it’s possible that in my old age (I’m approaching another birthday with alarming speed) I’m slowing down ever so slightly but this equates to 50,000 words per day. The most I’ve ever managed to turn around in a day was around 18,000 words and that took a solid 20 hours to do with a good translation memory and it was for information purposes.  Surely nobody could do 50,000 words in a single day?


Tales from Portsmouth

Spinnaker Tower in Portsmouth (via Wikitravel)

Spinnaker Tower in Portsmouth (via Wikitravel)

I’ve just come back from a conference in sunny Portsmouth where I gave a keynote on the subject of “the translator as writer“. Of course you can never be completely happy with your performance at conferences but overall I think my presentation went down pretty well. It was quite unnerving though that it was recorded on video so hopefully I won’t find it too painful when I get to watch it back. My topic was “Are technical translators writing themselves out of existence” and at some point over the next few days I’ll write something about it here.

The conference itself was excellent as usual with a good variety of topics presented by practitioners and academics from all over the world. The Portsmouth translation conference is, I have to admit, my favourite not just because of the topics discussed or the friendly, relaxed atmosphere but because it draws practising translators as well as researchers. For me, conferences that are solely for academics can be quite dull so to have presentations from the word-face too is quite refreshing. Sometimes academics can lose sight of what translation is about and focus on obscure issues far removed from the actual process of translation.

Continue reading


Translation agencies turning the screw on freelancers

Getting blood from a stone

Should translators be at the mercy of agencies? (Image: Colin Anderson/

When I teach translation technologies to my students, I always make the point that we are not just concerned with the nuts and bolts of how the technology works but also with the sociological, commercial and financial effects the technology has on the profession. Acquiring skills in translation technologies, so the literature goes, helps translators improve themselves by adding new skills to their repertoire and this helps raise their self-image and raises the status of the profession. Tools like translation memories eliminate the mundane, repetitive tasks which are the less palatable part of a translator’s lot and allow us to concentrate on the creative, challenging and ultimately more satisfying aspects of translation. All true to a certain extent although from a translation and linguistic point of view, I’m still somewhat sceptical about the merits.

One of the key effects translation memories have had on the industry is that they have brought about a re-evaluation of payment practices and translation rates. This is well documented (for example, here, here and here). Although it is rather unfair that we lose money if we use translation memory tools, most of us have come to terms with this, but I recently received an email from a relatively new client informing me that “given the current economic climate”, many of their customers are “demanding even bigger discounts for fuzzy matches” and as a result, they would be imposing a new pricing structure. This new structure involved even bigger discounts for customers and even less money for translators. What made the email so interesting and to be honest, more annoying was that it was a diktat; there was no question of negotiation or compromise. This was the agency’s decision and as a translator, I would have no choice but to comply. That’s what they think because I am going to exercise my right to choose never to accept work from them again. Or, to refuse to use Trados on any of their translations. See how they like that!

I’ve seen on other forums and blogs that this isn’t the only agency to chance their arm at squeezing a few more drops of blood from translators using the global economic downturn as a convenient, yet cynical, smokescreen. We shouldn’t be surprised because there is a long and ignoble tradition of translators being penalised for investing in the latest technology, which rarely comes cheap incidentally. I can’t be the only person who thinks that the price you pay for a service should reflect the quality of the service. So if you choose a service that uses the latest technologies, you should expect to pay more for it. It’s simple market economics really: the service provider invests in new technology and then factors this into their fees to reflect the improved service and ultimately to recoup the cost of the investment.

Of course you could always argue that by investing in technology, the service provider gains more business and, depending on the technology, will have a higher work rate and this will offset the investment. This is most likely the case with translation. But with these increasingly grasping discount systems, translators are seeing any commercial benefits being eroded. The discounts are effectively negating the whole point in getting the technology. Can you imagine paying a doctor less for using a shiny new scalpel than if the doctor used a rusty old hacksaw? Or would you expect to pay less for a meal cooked in a modern, clean kitchen than you would for something cooked on a hot stone at the side of a busy road? Unlikely.

The more altruistic among us would say “Ah, but greater cost efficiency and less effort is only part of the story. The real benefit is an improved product for the client”. Such improvements might include greater consistency in translations, better safeguards of accuracy and fewer formatting errors. But the other benefits for the client and in particular, agencies, include the reduced costs as a result of discount schemes imposed for repetitions and fuzzy matches, faster turnaround times and, more worryingly, less dependence on a particular translator. You see, once upon a time a translator who worked regularly on projects for a particular client became, over time, an invaluable repository of useful information, expertise and know-how relating to that customer and their documentation. A regular translator would accumulate the kind of knowledge you simply couldn’t get elsewhere. And with all of this information safely stored in the translator’s head, agencies and clients had to keep using the same translator if they wanted to ensure the same level of quality, consistency and expertise. With translation memories the translator is no longer the guardian of this expertise – it is segmented, formatted and stored in translation memories which can be sent to any number of other translators if the original translator is unavailable or ceases to be economically viable. So by using translation memories, the translator not only loses money but also loses job security.

You’d almost be tempted to stop using translation memories altogether and start using a typewriter.

Discounts for a fuzzy what? Pish!

The only fuzzy thing around here is me...