Charlie Bavington

Professional French to English Translator - Business and I.T.

Bringing a pragmatic eye to meeting your needs

Review: Linguist article on community translation

November 14th, 2013 | Categories: translation issues

I am not sure I have ever been more irritated by an article in The Linguist. I speak of the Oct-Nov 2013 edition of the said organ, entitled “Power of the crowd”, subtitled “Community translation: friend, foe or no big deal”.
Aha, you might think, a balanced article discussing the pros and cons. I shall read on…

(Now the CIoL puts its mag online, you can read the article here , see pp14-15, but I’ll quote the relevant bits here anyway.)

The article opens by mentioning the reaction of translators when Facebook, as a commercial company, started using crowdsourcing in 2007. This reaction was not positive. The article sets out its stall at this point, saying “such reactions were not necessarily warranted back then and they are even less so today.” It seems to me that of the “friend, foe or no big deal” triumvirate, foe has been kicked into touch before the end of the first column in a 2-page spread.

Getting down to brass tacks, the fact that crowdsourcing is sometimes the result of a user community volunteering its services is wheeled out. Sorry, but just because someone volunteers to do something unpaid, it doesn’t make it right to avail yourself of what’s on offer. I take issue on principle with unpaid work being done on behalf of for-profit organisations, whether at the instigation of the organisation itself (as in advertised unpaid internships) or of the “volunteers” and such offers should be politely declined by any for-profit organisation that is not morally reprehensible.
(There seems an unspoken assumption, unless I have misread it, that “community translation” = unpaid. Sometimes crowdsourcing work is paid, albeit typically a pittance. While an improvement on taking people’s work for nothing, the difference in my overall stance is minor.)

There follows a description of the situation with The Economist in China. A bunch of translators is producing a Chinese version of every issue, which apparently the magazine is aware of but did not instigate. While I accept the premise that this is “free publicity” and gives the magazine a head start if it decides to officially launch in China, surely this whole initiative is of borderline legality in terms of copyright if nothing else, and is hardly an upright example of the merits of crowdsourcing? Is this really the best example that could be found for an article of this type? If so, it speaks volumes of an attitude that finds availing oneself of the fruits of others’ labour for little or no payment acceptable.

Next, under the sub-heading of “No such thing as free translation”, the article attempts to justify use of crowd-sourcing by pointing out that there are “financial repercussions”. In summary, firms (e.g. Facebook) have to pay for content management of some kind. As if this somehow means they are absolved from having to pay for the content itself. If you want a thing, as far as I know, the idea is you pay for all of it, not just the bits you can afford. The fact these reprobates have to pay for software and content management does not, in my humble opinion, mean they don’t have to pay (or can pay next to sod all) for the translations.

The next subheading is “Are professionals losing out?” This section proposes that organisations indulging in crowdsourcing often find they need subsequent paid translation work as a result, it “acts as a catalyst that generates demand for professional translation” and that instead of “stealing opportunities for paid translation, in many cases, it creates them”. First, given the author’s undoubted contacts in the industry, it would have been nice to see some names where this has happened, unless it is just a hypothesis. Second, whether hypothetical or rooted in fact, surely the subsequent demand for paid translation would have still been created if the original, crowdsourced, work had been paid for too? Does the fact it was free make any difference to the cause and effect? I don’t see why it would.

The last point made in this section, that it gives new translators an opportunity to learn while gaining experience, is valid to an extent. However, this is an issue that affects every job and profession in the known universe, and while regrettably the solution of getting beginners to work unpaid on behalf of for-profit organisations has gained ground in many areas (unpaid internships, for example), that doesn’t make it right or justified. If a for-profit organisation uses labour from which it gets some value, it should pay for it, full stop.

The article turns next to the issue of quality, claiming that in fact the sheer numbers involved in vetting or ratifying translations ensure crowdsourced content is up to snuff, and furthermore, where the numbers fall short, companies often don’t release crowdsourced translations into the wild in case they’re a bit pants (I paraphrase). We’ve probably all seen amusing screenshots which would tend to suggest this is not always the case, but then the author doesn’t claim it is, and such failings are not the sole preserve of the crowd. I’m happy to accept that the quality difference doesn’t amount to a major factor in our “friend or foe” consideration.

As we hit the last column of the article, the subheading is another question – “Will more companies follow suit?” to which the answer starts “the good news is that yes […] many more companies have followed [Facebook]”. Good news? OK, so we knew the article’s stance, despite the subtitle, from the outset and as we near the end, we might expect conclusions to start being drawn, but thus far, we’ve read precious little “good news”, in my view. The conclusion seems to be that the result has been “more demand for translation in general including for paid translation services”, which could well be true, but we went through that argument two columns ago. Has the well of positive things to say about crowdsourcing run dry?

However, we do actually finally get to some genuine good news, in that “most companies” will not “widely adopt” crowdsourcing, because “The vast majority of the world’s translation projects will not be crowdsourced because that simply isn’t a viable solution.” This happens to chime with what I hope – it also seems to apply to the sort of work I personally do, but this is not an I’m-alright-Jack piece and neither should the resourcefulness of those who see a chance to get something for nothing be underestimated. It also still doesn’t excuse the portion that is crowdsourced, and is no comfort to those working in areas affected by this abomination.

And incidentally, see here for a brilliant post by someone who potentially is affected by crowd-sourcing: http://www.loekalization.com/crowdsourcing.html. While my objections are merely ethical and philosophical posturing about crowdsourcing in its current form, his are practical.

Is there any silver lining? Not all crowdsourcing work is necessarily free. Some of it is paid, after a fashion, such as this derisory offer that coincidentally arrived in my inbox as I starting writing this a while ago.
(In case that link later proves to be dead, in essence the rate offered is $0.01-0.10 per HIT, which is a “Human Intelligence Task (…) used in many kinds of applications, such as data annotation, multimedia, and search engine result evaluation” – you can get paid via PayPal once you have accumulated $50. And yes, I really should unsubscribe from translationdirectory.)

See also this detailed and reasoned description of another crowdsourcing provider: http://linguagreca.com/blog/2013/11/crowd-translation-already-here/.

These particular offers are insulting, yes, but it shows that where there is a will, there could be a way to properly remunerate crowdsourced translation, thereby gaining all the advantages listed above, but without the crucial drawback of being obtained from the unpaid or underpaid efforts of a “community”. As with so much in life, the actual theory of crowdsourcing may not be bad one (the way it meets the apparent need for quick translations of rapidly changing online content, for instance, is undeniable), but the implementation, left in the hands of the morally disadvantaged at the top of so many corporate dungheaps, leaves much to be desired.

Comments are closed.