Translations have always been a tricky business. Starting with the interpreters in times past and ending with the latest instant translation devices, this field has encountered and survived millennia of hurdles, misuse of terms, bad grammar and plain ignorance, in order to keep bridging the gap between people.

With the turn of the century and the progress of technology, many have attempted to make translations more accessible to people everywhere, by means of various media. Whether it’s a bilingual edition of a novel, an online tool or a pair of fancy headphones, translations seem to turn into the stars of the decade, as e-companies attempt to finally break through the language barriers. However, in order to do this, a lot of resources must be put head to head, corpora must be compiled and time must be spent in order to finely comb through words upon words. This used to be done by employees and then by “verified users”. Today, however, anyone can fill in the uncomfortable shoes that translators have worn for millennia, as we enter the age of “crowdsourcing”. At first, it used to be Google (“Crowdsource”). Then Windows (“Language Community”) adopted a similar strategy. Today, many transnational online businesses have turned to their own audience for translations, proofreading and localization (e.g. “Avira Lingo”, Evernote’s “Zing”, the “Klingon Language Institute”, etc.). It’s so cheap that it’s practically free… but is it any good?

First of all, we must take into account the premises:

Translations are usually considered a luxury service, especially in our current computerized world, in which a great part of the global workforce (i.e. internet users) is able to use at least one form of the English language used in the online environment. Nowadays it’s much easier to communicate with strangers from the other side of the world and, while professional translation services are still required for official matters, these are not deemed as necessary for conducting day-to-day business.

Secondly, a great deal of international companies only want to invest in end-product related activities such as development or marketing. Translations usually end up stuck in a corner until release day, because they are not perceived as part of the core business or as something that directly impacts success, profits, etc. That is, until a poor translation and/or localization ends up being used in marketing, for example, and leads to botched advertising campaigns, followed by terrible reviews and even worse sales.

Lastly, crowdsourcing is free (or extremely cheap) and easily accessible. Also, given that most respondents are already users and/or customers, this tool may kill two birds with one stone: on one hand, crowdsourcing provides free localization; on the other hand, it renders a complete picture of how customers think and talk. This way, a company can understand their personal language and train of thought and, therefore, provide them with highly customized solutions.

Yes, crowdsourcing is a stroke of genius and, yes, there is a kind of safety in numbers. But, at the end of the day, this data collection method still has a big glitch: there is no quality control over the input and, as with all unreliable sources, the final result may end up better or (a lot) worse than expected. For example, no later than this August, Google Translate’s live feature helped to translate the phrase “Supreme Leader” into “Mr. Squidward”, in an article from a North Korean periodical (the “Tongil Sinbo”). Thankfully, this was a little-known occurrence and was quickly corrected. Had the communication channel been more popular, this blunder could have easily led to an international scandal.

Languages are living organisms that change alongside the people that use them. Also, it is preferable to receive input from as many speakers of the same language as possible. But, sometimes, the opinion of the many may still not be the correct one. Each language has polysemantic words, homonyms and synonyms which don’t just work all the time in any given context. Each language has technical terms that may only be employed at certain times.

The upside of crowdsourcing is that it connects the data users (companies) with persons that are not necessarily translators, but have a special understanding of much sought-after fields. The downside of crowdsourcing is that a lot of non-experts are also free to share their (mostly incorrect) opinions. However, since we are not talking about politics, but linguistics, we should lend an ear to those that we know to be in possession of the correct answers. Therefore, the accuracy issue can be easily solved with the help of living and breathing translators. It doesn’t matter if a translation/localization is provided by a crowdsourcing or machine translation tool. As long as the penultimate version ends up in the skilled hands of a translator, the final result can be corrected in due time (if inaccurate) and refined (for a proper localization), so that it may achieve the highest standards in terms of quality and expectations.

Signed: The avid science-fiction reader and interior-design hobbyist.

 

Sources:

Cet article est également disponible en : Roumain Anglais Allemand