In 2016, SDL conducted a study called Translation Technology Insights Research surveying 2,784 translation industry players in 115 countries.
The study revealed that 80% of translations sourced through LSPs require rework for terminology inconsistency. It also confirmed that translation quality assessment remains largely subjective: 59% of all respondents—and 41% of LSPs—reported they either do not measure translation quality at all or that they use ill-defined or purely qualitative criteria.
The ordinariness of subjectivity in translation quality assessment could maybe be explained when considering that 60% of respondents also reported of not being aware of any standardized metrics or models for assessing translation quality.
EN 15038 and ISO 17100, as well as ASTM F2575 (let’s forget about ISO/TS 11669), have been largely publicized and most often shown off, too, as a remarkable, outstanding, ultimate achievement of the industry and for the industry, with a pompousness that has always been incommensurable with the actual and factual outcomes.
It is estimated that worldwide there are about 300,000 professional translators and 15,000 LSPs. The sample for SDL’s study could reasonably be considered as statistically significant although maybe strictly not representative, properly distributed, or weighted. Anyway, not only is SDL the world’s largest language service and technology provider, but it is also present throughout the world. The findings from its study can then be considered fundamentally reliable and universally valid.
If most respondents know little or nothing about the formal quality standards used within the industry, it is hardly due to a communication issue. It could not even be due to the existence of many standards. Most possibly, most of those respondents simply do not think it is worth taking the time to know.
On the other hand, many self-styled professional translators maintain that there is no need for technical and statistical details for measuring translation quality, and that only a qualified human translator can appraise it. Also, among those people, many argue that all rating systems are just polling of opinions and are therefore flawed, while many colleagues who claim to be translators are not competent at translation. Except, maybe, for the elusive, coveted and cherished premium segment.
In such a confusion, it is not surprising, then, that translation quality came out as six times more important than price. Price can be quantified, measured and understood, at least as cost, and this also explains why translation is still on no account seen as an investment.
Similarly, it is not surprising that the translations to which most of those dealing with MT have access are mediocre. Probably because the premium segment does not exist, except for price requests, whose large-scale fulfilment is yet to be proved.
Probably, most of the elusive, coveted and cherished premium segment advocates are simply nostalgic for the “good old days.” And maybe the young who might not have lived those “good old days” have just bought into that wishful thinking like a bottle of snake oil liniment.
Well, the news is that there are no “good old days.” Translation is definitely an easy task compared to how it was only twenty-five years ago. It took an entire day then to produce half a dozen of almost neatly typewritten pages. The pay per word or page or line was not higher or anyway commensurate to the effort, and—surprise!— translators were poorly considered even then, and translation almost never hit the news, not even when it was bad. Also, 25 years ago a translator would have had to deal with the queues and times of snail mail or the cost and stress of delivery services. Not to mention the trouble of browsing index cards in libraries, the laziness of librarians and the long waiting for a volume.
Socrates was against the haphazard diffusion of writing for the forgetfulness that comes with it, and indeed he left no writings of his own. Yet, we could have memory of his thought only for the writings of his students Plato and Xenophon. Erasmus of Rotterdam had something against printers, claiming that printing presses would «fill the world with pamphlets and books that are foolish, ignorant, malignant, libelous, mad, impious and subversive.» Yet, he was also one of the most prolific authors of his time. Umberto Eco advised his nephew to memorize all that he could because the Internet cannot replace knowledge nor the computers can replace the brains. Yet, he forgot the right title for the doggerel he mentioned.
The legacy of all these men is outstanding, and, by hook or by crook, they all had to deal with the technology of their times, which anyway was what brought that legacy to us.
Back in the “good old days”, a long project consisted of a 250-page technical specification, patent, user manual, or financial report that would take a month for a human to translate. Today, according to a recent briefing SYSTRAN’s CEO had with Common Sense Advisory, his company wins some very large accounts, including Continental, HP Europe, PwC and Xerox Litigation Services, because they are quite capable of translating millions of words a day as a normal part of their international operational needs.
The Goodle Days are past and gone, a lot of good people have done gone on.