Isn’t it odd how you can go back to something you wrote almost 10 years ago and go “Wow, nothing’s changed. Ten years later and plus ça change, plus c’est la même chose.” Be it translation agencies that demand that translators drop their rates or dock a certain percentage of the already low rate they’re paying because their business is suffering; agencies that ask translators to accept a 90-day payment term or else accept a prompt payment surcharge for a 30-day payment term; agencies that offer customers the moon and the stars. But wait! If you order now, you make get an extra planet at no additional cost all for lowest price ever; translators who copy each other’s CVs and forget to change the wording; translators who speak seven languages fluently and specialize in every area imaginable in the hope that you don’t speak one of the languages you need and therefore have no clue what the translation actually says.
Ten years ago, machine translation (MT) was all the hype. Let me rephrase that, Google Translate (GT) was all the hype and MT was being sold as the greatest thing since sliced bread. “People” were and still are using GT to get a basic understanding of a text, which is perfectly fine, if that’s all you need. Google does a so-so job at picking and choosing the documents it gets its translations from. Today, automatic translations into English are no longer as hilarious as they used to be, but they still contain some pretty serious mistakes. Translators were starting to use GT to increase their productivity or offer languages they didn’t speak. While all the big translation agencies are still competing on price, translators are competing on price and volume: a once respectable output of 2,000 words a day is now synonymous with “a translator who takes a lot of coffee breaks.” Don’t get me wrong, technology is great. I love my computer-assisted translation (CAT) tool and I’m all for MT where appropriate, but I hate laziness, and I’ve seen many a translator get lazy over the years.
So, while I sit here contemplating my navel because no one needs a real person who actually is fluent in four languages and has real business knowledge and experience, I wonder how the translation “industry” – I hate calling it an industry; neither “Translator” nor “Language Service Provider” professions I can choose on a form, so how can there be an industry for it? – became synonymous with poor quality and even poorer pricing.
There are a few reasons for this, ranging from fake translators taking advantage of the customer’s ignorance, to salesmen selling things customer’s don’t need. However, the best explanation I’ve seen so far is Chris Durban’s comment on a recent post by Renato Beninatto on LinkedIn. It reminded me of an MT event I attended in Ede in 2011. In his keynote speech, Renato declared to a room of 135 translators that “quality doesn’t matter” (I can still hear the audience’s collective gasp of disbelief) and Jaap van der Meer declared at the end of the event that MT would “relieve translators of their repetitive and stupid tasks.” How cool was that? You no longer had to translate 2,000 words a day to make a living, you could now post-edit 10,000 words a day to make the same amount of money with less work! What an outlook, what a prospect! You could see dollar signs in some of the translators’ eyes and fear in most others’. You see the thing is, translators are not necessarily editors, editors are not necessarily translators, and neither really likes cleaning up other people’s mess, let alone a machine’s. But that’s what two of the visionaries to watch were telling us we’d be doing in the next 10 years, so how could they be wrong?
To a certain degree they were spot on. The translation “industry” they have helped shape is consumed by MT and post-editing, only the latter is the part that often gets forgotten. That would be one reason for the never-ending quality discussion. Seriously, we’ve been going in circles for years. But there’s also a segment that is not part of their chunk of the “industry” – thank God for that – and is under-impressed by what the gurus are selling.
Here’s Chris’s explanation: “Renato, thanks for looping me in. I hadn’t seen your “predictions” ‘lo those many years past (but had seen Marian’s unfortunate quote in the SDL ad), so it was interesting to read.
I always enjoy your energy, but alas here you are once again talking about a chunk of the market (yours) as if it were *the* market. The one and only. Which is happily not the case. (Haven’t we had this conversation before?)
But first: I’m told the GT example you used as an illustration above your article demonstrates the limits of GT, not its advantages. Not my language pair, but a friend offers a correction: “not so bad on the first sentence, where you can get away with the word-for-word, but when the quote pivots cleverly in English, it flops in Russian. So the second sentence should read: “Рассуждая по аналогии, чем более вы уверены в будущем исходе, тем скорее всего вы неправы.”
Again, not my language combination, but one lesson from that poor translation is that it’s fast, but it’s not correct.
I’d add that playing fast and loose with terms like quality is a real issue in the segments I work in. And these days, having seen grown men pull their own heads off trying to define “good enough”/“high quality”, I’m less inclined to let you and the fanboys imply or state outright that quality is a given when you are churning texts out. Or to decide that any translation other than word replacement is “transcreation”. That’s just silly.
On to “productivity”. I know you love metrics, but measuring this in terms of words per day or per hour is probably not the best yardstick to describe the work we do. Surgeons (or plumbers for that matter) sure don’t count the number of patients/customers blasted through per unit of time. Neither should we, and my clients certainly don’t see things that way.
How about: translation comes in so many different forms and performs so many different functions that no one metric alone is useful to define it. Think about it. Translation of corporate strategy memos & crisis communications docs is performed in intense environments with many iterations, and you work in close collaboration with the client. Effective communication is the benchmark, not words per X. Just as translating a slogan and a correspondingly sophisticated ad campaign usually takes weeks, and that’s with a ton of different people (including the client) involved.
Moving right along, a few very attractive clients I’ve signed since September state outright that they are fed up with the thousands-of-words-per-hour crowd. They are looking for a long-term relationship with someone who takes their industry seriously and is prepared to dig in. Their procurement people have signed contracts with two of the large agencies that you lovely guys bow down to, but the top brass (who come looking for my services) simply can’t risk the uneven quality they’re getting. You are wasting their time, they say. And (shh! don’t tell anyone!) their budgets are far higher that what you’re pitching for – ah, but they do require proven skill and serious effort, and 2500 words a day is about right.
So technology is great, but frankly, it’s the easy part. 10,000 or 20,000 words a day @ 0.10 or 0.20 — nothing wrong with that either, if it flies in your segment. But quality and engagement-wise, we’re talking about apples and oranges. I’m pretty sure I’m earning more in my market segments, and I see lots of work crying out for careful hands like mine.
Just wanted to put that out there.”
Renato’s post is well worth reading, as are the comments.