ADDING PERSPECTIVE ON THE NEW AGE OF TRANSLATION, Part 1
Machine translation post-editing (MTPE) has become a trend-setter in the translation industry. Praised as great helpers pushing human productivity to new heights, translation engines already process huge translation volumes, their outputs to be “merely refined” by humans. The technological developments being huge indeed, linguists are expected to be happy for getting so great help from the machines. Are they really?
Despite the enthusiasm of big translation industry players who expect huge cost savings on human work, translators, now turned to MT post-editors, often seem to object to the new process. Firstly, their experience indicates a lower MT output quality (meaning more human work) than what financial managers forecast in their productivity sheets. This invariably means linguists making less per hour than before, and despite growing experience with the MTPE process, this financial expectation inconsistency between linguists and clients persists.
Conveyor belt operators of the new age
Poor financial motivation is not the only issue, however. MTPE does not merely mean “you do the translation but more effectively.” The whole character of the profession is changed in essence, turning linguists to conveyor line workers of the Brave New World. An essentially creative task of inventing an idiomatic counterpart of the source text gives way to mechanically checking the machine output, endlessly looking for missing or added words, out-of-context translations, and of course (this being the easiest part) translation pieces that make no sense at all.
Exhausted and demotivated
The linguist is expected not to bother about creating any more – usually they are instructed to use as much of the MT as “usable,” as no one is willing to pay for “unnecessary” man-days. MTPE, however, is no less easy, as it requires humans to invest an inhuman amount of attention. Machines do not make human errors. They make inhuman ones. Maybe also this is the reason why endlessly seeking for machine-made errors could be so tiring.
The more machines, the less happiness?
I wonder whether linguistic students of today are already aware that many of them are not preparing for a creative profession, as their task might be to mechanically search texts, trying to spot where machines got it wrong. Every day, 9 am to 5 pm.
How attractive does that sound, actually?