Posted by: patenttranslator | April 24, 2018

A Big and Joyous Roadkill Party in the “Translation Industry”

Revolutionary, disruptive innovation is the best kind of innovation. That is what we are told over and over again by prophets of disruptive innovation, who are usually well paid by people who already have or intend to become filthy rich from another miracle of innovative disruption.

What the prophets never mention is that just like revolutions, the disruptions that are caused by these innovations come at a great cost.

Some kinds of innovation are not only disruptive, but so destructive that they can kill not only jobs (Uber, Airbnb), but also people (Uber’s self-driving cars) and even most of the life on this planet (the ‘clean diesel’ cars of Volkswagen and other car manufacturers), or at least our way of life.

Much celebrated current examples of disruptive innovation include quasi-magically turning amateur drivers into experienced and safe taxi drivers, or turning regular people who own an apartment in a popular city or a beach town into part-time hoteliers.

Well, into a certain kind of taxi drivers, anyway. Not the kind of taxi drivers that I remember from Japan in the eighties where super-polite taxi drivers wore white gloves and the passengers rested their tired heads on headrests that were thoughtfully enveloped in a perfectly white cover adorned with lace.

I don’t like Uber for a number reasons and I would prefer if taxi drivers and people working in actual hotels could still make a good living. But seeing as I am a cheap guy, I now take Lyft from my house to the airport and vice versa. It’s hard to resist: the taxi ride to the airport in Norfolk used to cost me about 50 dollars plus a tip, the same as from my hotel in Vienna to the airport there, if I remember it correctly. The same ride with Lyft from my house to the airport costs 23 dollars plus a tip.

There is no lace cover on the headrest in Lyft cars, but I usually have a nice conversation with the driver. The last one did not even have time to shave, possibly because he got his driving gig just after he woke up. He had a tiny bit of a foreign accent and when I asked him about it, he eagerly told me his life’s story.

He was in his mid thirties, born in Colombia, and he told me that he just recently married an Asian woman from Laos who was 17 years older than him and had two teenage daughters with her previous husband. He seemed appreciative of my advice on what one can expect from Asian women and from teenage children, two subjects that I know a little bit about, although even after decades of experience, still not much more than what I know about the dark side of the moon.

We had a very nice chat. But let’s get back to the subjects of destructive innovations and roadkill.

Because many innovative innovations can be not only disruptive, but also very destructive, I see similarities between what the destruction that is caused by what machines do to nature and animals, and the kind of havoc that so-called disruptive innovation in the so-called translation industry wreaks upon our world.

Neither the car nor the algorithms are designed to kill animals or written communication, it’s just something that unavoidably happens millions of times a year everywhere where there are roads, or where machine translations are used.

Most of the time nobody profits from the mass slaughter of animals on our roads. In fact, the opposite is often true, because many people die in accidents caused by animals and the animals sometime survive.

Although from the viewpoint of the murdered animals, these accidents are clearly caused by our roads, not by the innocent animals themselves, and this may be how they are paying us back for our treachery.

But since people always try to profit from anything, including the numerous instances of roadkill, a type of cuisine has been developed by us that over time called roadkill cuisine.

According to Wikipedia, “Roadkill cuisine is preparing and eating roadkill, animals hit by vehicles and found along roads. It is a practice engaged in by a small subculture in the United States, southern Canada, the United Kingdom and other Western countries as well as in other parts of the world. It is also a subject of humor and urban legend. Large animals including deer, elk, moose, and bear are frequently struck in some parts of the United States, as well as smaller animals such as squirrels, opossum, armadillos, raccoons, skunks and birds. Fresh kill is preferred and worms are a concern, so the kill is typically well cooked.”

And once you get rid of the worms and maggots, meat is meat and it is ready to be cooked, whether it is fresh meat from a slaughterhouse, or mystery meat from roadkill. The main difference here is that roadkill meat is much cheaper or free. Some restaurants proudly advertise the fact that they serve roadkill meat, such as this one in Seligman, Arizona (You Kill It, We Grill It,) although most probably don’t mention this particular detail.

Just like roadkill, machine translation, the latest disruptive innovation of the “translation industry,” much celebrated in dozens of industry press releases, called “post-editing of machine translations,” is also very cheap.

The industry anticipates that it will be able to generate enormous profits from the roadkill created by well trained algorithms of machine translation. Especially large translation agencies are salivating at the prospect of the additional hundreds of millions of dollars that will end up in their pockets if they are able to put cheap or free algorithms to work instead of expensive to very expensive human translators.

Are these profits going to be realized, or are they already being realized? If we are to believe press releases and propagandistic statistics that are mostly based on “expected” data, there is no question that the profits are in many cases already here and that they will grow exponentially, although only for the smartest among the captains of the “translation industry”, namely those who are clever enough to identify the “sweet spot” in pricing of the roadkill for translations created by computers and algorithms.

But I doubt very much that the roadkill of machine translations will in fact result in a big, joyous, boisterous party for the industry. I think that smaller translation agencies that put all or most of their eggs into the basket of machine translations will in due course go bust, and the big ones, those that are able to put a lot of money into advertising the roadkill of machine translation and selling it to their clients as Angus beef of prime quality, will make some money in the short term, but lose much more money in the long term when their clients realize what kind of meat they are being served.

You can tell pretty easily whether the roadkill meat is good to eat by the way it looks, smells, and whether and how many worms and maggots are in it already.

But for many reasons, which unlike an industry captain, every translator will understand both intuitively and based on experience, it is very difficult to tell where all of the worms and maggots are hidden in roadkill translations created by algorithms.

I believe that these things should not be even called translations, because they are not that. But it is probably too late to try to change the terminology of the “translation industry” at this point, although some people have tried to explain why this terminology is misleading, see for example this blog post analyzing the “evils” of what is called “post-edited machine translations”, or PEMT.

Unlike spoiled meat, the roadkill of machine translations does not have an awful smell, and no amount of post-editing is likely to identify and remove the worms and maggots hidden in it.

The only way to get rid of them is in fact to have the machine translations, or pseudo translation, completely redone by a human translator. But that would then beat the purpose because it would be more expensive as well as slower than not bothering with a machine translation in the first place.

One way for clients to avoid post-edited machine translations, which may be reasonably priced and may look like really good translation, (except that the poorly-paid post-editor  may have created the exact opposite of the meaning of the original text to save time), is to enter into a working relationship with an individual translator or a small, specialized agency.

Because we need both the animals and the roads for our survival, some people, whether they are vegetarians or carnivores, are creating underpasses for animals enabling them to safely cross major roads.

Translators must learn how to avoid the traps of translation industry, which is now salivating at the prospect of a joyous roadkill party in the industry, and create their own bridges, overpasses and underpasses to direct clients who simply can’t use the roadkill that so many translation agencies are trying to sell to their clients, out of ignorance, incompetence and greed.

Nobody will do this for us – we simply need to figure out how to do it on our own.

The other option we have is to specialize in identifying worms and maggots in the roadkill created by machine translation as quickly as possible …. and hope against hope that we will be able to earn enough money in this manner not to become roadkill ourselves.



  1. Brilliant, Steve 🙂 Roadkill versus good, grass-fed beef or lamb is an appropriate comparison for machine pseudo-translation (MpT) versus the real thing. And clearly, those who feel we must join the party have developed a taste for maggots….


  2. “And clearly, those who feel we must join the party have developed a taste for maggots….”

    They just want to sell maggots to their customers maggots for nuggets of gold.

    It’s not like it hasn’t been done before:


  3. It’s not a healthy diet for the customers, that’s for sure. Which does not stop the “translation industry” from salivating over the prospect of expected healthy profits, although I doubt that they will be realized.


  4. Hot off the presses: the latest TAUS newsletter is offering a webinar on “The Redesign of Translation Business from LSP Perspective” which offers a slew of industry bigwigs discussing the “Modern Translation Pipeline”. And what do they squeeze through their “pipeline”? Roadkill with the trimmings? Chopped and minced translators? Or just a never-ending goo of mixed metaphor leftovers?

    Liked by 1 person

    • The product extruded at the end of their business pipeline bears a striking resemblance to that at the end of our respective digestive pipelines…..


  5. My guess would be: all of the above, with a mix of roadkill thrown in for good measure.


  6. I am a long-time reader (nay, a fan!) of your blog. I have been working as a translator and simultaneous interpreter for Korean/English for a decade now. Makes me rejoice reading your realistic analysis of the “translation industry.”

    It upsets me greatly that so much of work contacts (agencies and direct clients) have started asking about “proofreading” work done by “someone else”… except when I see it, it is clearly machine translated. I have stopped taking all proofreading and editing work because of this. The only solace from the debacle is that many clients eventually return.


  7. Thank you so much for your comment.

    It seems to confirm my what I said in my post, namely that industry’s greed may lead to short-term profits, followed by long-term losses.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s


%d bloggers like this: