There is a lot of talk suggesting a growing demand for almost instant translations lately. Translators are being advised to adapt their skills and keep up with evolving technologies in order to provide an increasingly speedy service. It is not very clear, however, where this demand actually comes from, or who will benefit from this new approach. So what’s behind the hype?
First of all: I am not against technology in translation. I use a CAT tool to help me provide consistent translations. Fuzzy matches are indeed welcome time savers as I don’t have to look up how I have translated a similar sentence earlier. So I am all for the development of tools to improve my work.
There is a difference, however, between using a tool to provide better translations and post-editing an automatically generated pre-translation. What is the purpose behind the latter: is it to provide a better translation or to provide a quicker translation? And is a quicker translation a better translation for the client?
Back to basics: what exactly is a translation?
There is a clear consensus that a translation is more than a transfer of words from one language into the framework of another. Translation is not only about vocabulary, grammar and syntax (“the mechanics”) but also about voice, purpose and style (“the spirit”). There are cultural aspects, target readerships and customer specifics to take into account.
When translating a text, we carefully assess the register and writing style of the source text in order to create a translation that reads like a genuine text in the target language while reflecting the style and evoking the same effect as the original. We make a myriad of decisions and choices based on our knowledge of and feel for a language – something a machine is not yet able to do, and probably never will. Translation is an intricate creative process, not an off-the-shelf commodity.
Do we really need to reinvent the wheel?
A machine translation transfers words, sentences and structures from one language into another using ever more clever algorithms. The mechanics may be there, but the spirit is missing. Therefore it is, strictly speaking, not even a translation. Of course, machine translations have their place. For someone who just wants to quickly get the gist of things, they can be useful indeed. But with regard to the process of a true translation, the key question is: why replace something that works perfectly well – the skills and expertise of professional translators – with something that cannot reproduce the same quality and thus needs editing?
Especially at a time where we still have to educate potential clients and the wider world that a translation involves more than just copying words from one language into another, I fear that we are not doing ourselves and our profession a favour if we jump onto the bandwagon of producing ever quicker auto-suggested post-edited machine translations.
Who says quicker is better, anyway?
Good-quality products take time, so how can it be different for services? No one would throw some pre-formulated sentences at copywriters to produce advertising copy quicker. No one would suggest a few diseases to doctors upfront in order to speed up the process of diagnosis. Why not? Because prompts can actually hinder creative and analytical processes instead of speeding them up, and the end result may not be desirable for the client. It seems to me that the only ones profiting from this new trend in translation are organisations that supply technologies that are not helping translators to do a better job but taking the actual job out of their capable hands, albeit with an inferior result.
The future of translation is in our hands
We should ask ourselves what we as professionals actually want to produce and what we want to be seen as. Do we want to produce translations or become machine editors? Do we want to be seen as enablers of good communication, or do we want to give in to pressures of an industry that dismantles the core element of our work and become mere sidekicks?
I think we should embrace technologies we can use as a tool to improve – even speed up – our work because our clients deserve the best possible service. Speed alone, however, is not the solution. Slowing down and taking great care is usually what improves a translation. So perhaps we should focus more on honing our translation skills than learning yet another software programme. And until machines can actually match the quality of the translations we humans already provide, we should take charge of the beautiful craft that is our business.