Will Recent Trends in Translation Industry Result in Several Levels of Translation Prices for Different Quality Levels?

The modern globalized world has been with us for several decades now. The basic principle of globalization is simple: capital always goes to wherever the labor force is cheaper. For example US manufacturers first moved their plants from United States to Mexico, and then from Mexico to China and other Asian countries where the labor force was even cheaper.

European and Japanese manufacturers soon followed suit, although unlike their American counterparts, they were smart enough to keep the most important high-tech manufacturing capability at home. As far as I know, most BMWs (the initials stand for Bavarian Motor Works) are still made in Bavaria, and my brother-in-law still works for the Toyota Company in Nagoya.

Although he cannot work as much as he used to when he started working for the company some 30 years because he recently had a stroke, Toyota did not fire him. The company just gave him a less demanding job at a lower pay level and he should be able to retire a few years from now. Compare that to how US automakers dealt with auto workers in Detroit. A former Ford worker from Detroit in his situation would definitely be out work by now and probably facing a mountain of huge medical  bills.

Unlike possibly some of their European and Asian counterparts, most American CEOs simply don’t give a damn anymore about their fellow Americans, which was expressed so eloquently by Mitt Romney recently.

They can’t afford to give a damn if they want to maintain their obscene compensation levels: according to statistics available on Internet but not in The New York Times or Washington Post, the pay ratio of an American CEO to average worker is about 400 : 1, while in Germany it is about 12 : 1, and in Japan about 10 : 1. I don’t know how accurate these statistics are, but given that Washington Post, my local paper, says nothing at all about these pay ratios, they are probably accurate.

It is thus not surprising that the same principle of concentrating profit in the hands of a few people who will then try to control everything to further maximize their profit would be applied also to the translation business.

Many translation agencies have of course already successfully used a relatively new software model to squeeze more profit from freelance translators. If you want to work for a translation agency these days, big or small, you are often expected to agree to be paid only a fraction of the usual rate for what is called “matches” and “fuzzy matches”, namely instances of the same words or same or similar sentences detected by a CAT (Computer-Assisted-Translation) software tool such as Trados, which translators have to purchase on their own for hundreds of dollars if they want to be able to receive work from these agencies.

Some translators, including myself, don’t have to deal with this problem since they don’t work for agencies much if at all, but many are not in a position to say no even to this type of client.

The next phase of what some bloggers started calling hamsterization of translators, which not unlike hamsterization of journalists is a related but not quite identical phenomenon, is likely to be a new requirement that translators incorporate not only CATs, but also machine translation (MT) in  their translations.

This would turn workers who used to consider themselves specialized translators who usually need to have quite a high level of education and skills to be able to translate complicated technical, legal, financial and other documents, into mere post-processors of MT output.

An operator who can convince translators to cooperate with him in this new model in a commercial translation business would then be able to pay translators even less than what they can get these days after the deductions for “matches” and “fuzzy matches” because based on the inexorable logic of this new translation model, there is not much need to employ skilled, educated and experienced translators for mere post-processing. Indeed, that is the economic genius of this model, and you don’t even have to move your translation-producing factories to Mexico and China, unless of course there is more money to be made by doing so.

Since even partially bilingual people can in theory do something like that, human hamsters who have some knowledge of a foreign language should be theoretically able to do this job too because all you really need to do is to be able to “clean up” the MT product.

The question is: will this new model work?

Only time will tell, but I doubt that unlike captains of our industry who found it quite easy to decimate the industrial production capacity of this country by moving their production facilities to wherever the labor is cheaper, captains of the highly fragmented translation industry, whoever they are, will find it as easy to move translating work away from educated, experienced and skilled translators and shift it instead to hamsterized post-processors of MT product.

The problem is, since post-edited machine translations are not really translations (unless we are really talking about retranslations), most clients will be able to tell quite easily the difference between a document that has been translated by a human translator and MT product that has been post-edited by a human hamster who is not likely to be an experienced translator given the odious tasks that he has to perform and the inevitable equally odious level of compensation.

Customers who buy goods manufactured in China at huge stores such as Wal-Mart and Target can’t really afford to go and buy what they need somewhere else. All they can really afford are Wal-Mart and Target products. These products may not be very good, but they are slowly getting better and most importantly, they are always cheaper than comparable products.

Are customers who buy translations on the free market, if there is such a thing, in a similar position?

It would obviously depend on who the customer is, but my guess is that unlike Wal-Mart or Target shoppers, most buyers of translations have quite a few choices.

Our customers can already use MT on their own without having to pay a single penny for machine translation these days. They would probably be willing to pay a little bit extra for MT post-processing if it resulted in a much higher quality of the translation, but to obtain a good translation in this manner is not a likely result. I think the most likely result of these recent trends in the translation industry will be that three levels of translation quality will emerge that will accompanied by three different price levels.

Level 1: Unedited MT Product

This product will probably remain for the most part free. If Google Translate decides to charge for what it is now offering for free, Microsoft Translator and others will be glad to fill in the gap in the market for free MT.

Level 2: Edited MT Product 

This product will be not as expensive as translations crafted by highly educated and experienced human translators. This model is probably already working in some type of arrangements in the multifaceted translation industry. It will be interesting to see the price levels of translations created in this new translation model and whether and for how long this model will survive.

Level 3: Human Translations  

Human translations prepared by highly educated and experienced translators such as this mad patent translator, as opposed to translation produced by subprime or zombie translators. I don’t think that prices will go down for similar types of relatively expensive translations, although some customers may be more resistant to attempts by translators to increase prices as they will be tempted to defect to cheaper translation products supplied by the business model on Level 2.

—————————-

Disclaimer: Some of my predictions turned out to be wildly inaccurate. For example, in 1981 I predicted that the communist system would last for at least another 30 years and decided to get the hell out of Czechoslovakia because life is short and then you die.

The regime collapsed in 8 years. It is probably a mitigating circumstance that the best minds at the CIA also thought that the communist system would survive for quite a few more decades.