The Cultural Nuance Card Played by Precarious Freelancers

Freelance translators are often heard these days warning users of AI for translation that human translators are needed to provide understanding of cultural nuances. While this assertion is certainly correct if the assumed enemy is raw AI not touched by the vaunted human-in-the-loop, that mode of translation is a strawman argument. Firstly, the human-in-the-loop is likely a desperate former translator trying to survive and willing to try to catch things that have gone wrong because of AI use. There is little evidence that this is not working.

More importantly and more significantly, however, for mainstream freelance translators is the reality that only a tiny portion of translation that supports human translators has anything to do with or is affected by the “cultural nuances” so often mentioned by at-risk translators emphasizing what they alone can bring to the translation process. The subject matter requiring cultural understanding is in the quite limited, in terms of volume, fields of marketing and entertainment.

The boring truth of the translation business is that the overwhelmingly huge portion of translation that is paid for consists of difficult-to-translate and boring (to most people) texts about non-cultural things needed for things in legal, patent and other technical, business/finance, medical, pharmaceutical, and many other fields. There is not a cultural nuance in sight in things like:

  • a 25-page contract for a solar power generation facility being built for a manufacturer,
  • a statutorily required shareholders meeting notification running 40 pages and covering complex business and technical issues facing a major company,
  • 1400 pages of documents recording a police investigation of a sex crime, replete with DNA profiling reports and over 100 pages of photos with lines and arrows and a paragraph on the back telling what each one is (apologies to Arlo Guthrie),
  • a patent for an ion chromatography system, or
  • a complex specification for a system that senses the size of raindrops falling on an automobile windshield and adjusts the speed and operating mode of the windshield wipers.

These are the kinds of translation work that have supported countless freelance translators before the appearance of, and use by translation brokers of AI took it away from professionals.

For the vast majority of translation work that has been available to freelancers from agencies, the cultural nuance argument simply doesn’t work.

Language matters.

It certainly does, and the language used by agencies and even freelance translators discussing selling language services to agencies tells a story of deception or surrender, depending upon which side of the inappropriate terminology in that brokering relationship you’re on.

Collaboration?

Well, hardly ever does this term make sense when a translator sells translation services to an agency. This is particularly the case with the large agencies that control most of the translation-consuming clients and sell most of the language services that are paid for by clients.

Translators don’t collaborate with an agency any more than a chicken laying eggs is collaborating with a poultry farmer who cannot lay eggs themselves, but sells the eggs laid by chickens to egg consumers or to other egg sellers.

It’s actually worse with translation agencies, because, whereas most poultry farmers can probably at least themselves judge the quality of eggs laid by their chickens, many translation-brokering agencies need to enlist and pay yet other people who are not their employees for confirmation of the quality of, and correction of errors in the translations they purchase, because the agencies generally don’t have that ability themselves and need to purchase and resell that ability, again from people who are also not “collaborating” with the agencies, but rather selling them services.

People who think collaboration is an appropriate term in these cases should think about the etymology of that word.

Onboarding?

I don’t mind this term when used to describe what can happen when an employee is hired by a company, but that’s not what goes on when an agency gets a freelancer to agree to a rate and to sign an NDA. Agencies are very seldom hirers of translators, but rather are purchasers of translation services from freelance translators for resale of those services, often with post-purchase processing, which they most often need to outsource as well.

The terms collaboration and onboarding are feel-good terms that agencies hope will compensate at least in part for not treating as professionals the professionals that enable their continuing business.

These terms are also widely adopted by freelancers themselves, perhaps in the hope that they feel good if they reflect reality. They rarely do.

In decades of selling translation and interpreting to consumers and purchasers of those services, I cannot think of a single instance that I could fairly claim to have collaborated with a client. And, of course, I’ve never been onboarded by a client.

Not with a bang but a whimper: Relevance is slipping away from translation organizations.

I have written in more detail about the rather bleak output for freelancers elsewhere, but here are a few disturbing observations of what translation organizations are doing lately.

A number of major translation organizations that say they support and look out for the interests of translators—the ones I have in mind are located on three continents are—to varying degrees, but all clearly to a considerable extent—promoting non-productive belief in delusions by their members.

  • They allow and even encourage freelancer members to think that adopting AI themselves is a strategy for survival, carefully avoiding mention that the use of AI won’t attract clients for freelancers as their agency clients replace them with AI, and that only a very small number of freelancers are able to acquire clients other than agencies, which are well on their way out as purchasers of translation from freelancers. Some of the organizations, amazingly, have even taken to running or sponsoring events that sell AI-related products, teach AI use, or (more surprisingly) discuss post-editing.
  • They either themselves fail to recognize or are afraid that their members will recognize that it is not possible to earn a realistic living by doing the non-translation task of post-editing AI output.
  • They continue to promote the idea that human translators will always be needed (correct, of course), but fail to mention (or fear that their members will themselves realize) that, yes, human translators will always be needed, but only 5 to 10 percent of the current population of freelancers will be needed, and that the net number of actually working translators—post-editing is not translation—will reach that level in the very near future.

The above-noted behavior by translation organizations is uniformed at best and arguably irresponsible. It masks unspoken and unspeakable distress, but also surely is aimed at preserving the relevance of the organizations and of the people running them, in spite of most of their members seeing their own relevance slip away at a pace that defies their efforts to survive.

It is time for translation organizations to get real, face what is happening, and level with their members, rather than feeding them comfortable-sounding pap. If they cannot do that, they should think about other potential trajectories for the organizations, the most suitable one at this point being one that reduces the yearly dues needed to be paid by freelancers to organizations to zero.