A Pragmatic Eye

Wonder if any of you literally tens of readers had thought “Ah, another blog that has petered out…”?

Truth is, I only have one pair of hands, and apart from the very occasional flurry on Twitter, my spare typing (and thinking) time since June has been taken up with an epic thread on proz (should native_language claims be verified).
I should probably have stuck to my original stance that a discussion was futile as nothing would change, and walked away. But in the event, the discussion actually proved intellectually stimulating in parts (discounting the wummery, trolling, tangents and nonsensical assertions – so, about two thirds of the thread if not more, for which reasons I cannot recommend anyone to plunge into the thread unless they have oodles of spare time and genetically modified tolerance, since I’m sure no ordinary mortal could read the whole thing from scratch without wanting to wrench their eyeballs from their very sockets in frustration at the Olympic-standard asshattery on display in some quarters).

To summarise: proz has a field labelled “native language” (NL from now on). This field is displayed in the profiles of individuals, some of who evidently are not proficient at writing in the language they claim to be native in. Furthermore, this field can be used as a filter by those seeking translation services.

The issues this raises are:
a) it casts doubt on the accuracy of the everyone’s “native language” field, it diminishes the value of the site as a shop-window and thus chisels away at the reputation of anyone on the site (until & unless individuals are given the chance to prove otherwise), and
b) it means people are being considered for jobs for which they do not, in fact, possess one of the required skills, i.e. the ability to write proficiently in the target language.
I don’t use the site to get work; I have used it once or twice to outsource work. My main concern is a), but I accept the undesirability of b).

The assumed reasons or causes for this inaccurate reporting of NL are:
i) mistake about what NL means,
ii) knowing what NL means, but mistakenly considering oneself to be NL standard,
iii) old fashioned lying, to get considered for work.

What I found interesting about the discussion was the element that fixated on “native language” as an objective attribute, as if it were merely hair colour or address. Except of course, it’s not quite that simple; there are many definitions of NL, most posted in the thread (at least, I hope that’s most of them, I don’t think the world needs many more), and proz’s guidance on the matter is very Alice Through the Looking-Glass (“a word… means just what I choose it to mean”). Definitions for the site to use, involving the age a language was acquired, schooling, countries and length of residence, whether the language is actively used now, etc. have been proposed in the thread, with the aim of adding objectivity.

This is laudable; ideally information should be accurate, even if it’s just accuracy for accuracy’s sake. As an aside, even supporters of the idea seem to accept it will be hard to get user acceptance for proz actually verifying this stuff as per the thread title – they just want to ask the questions and leave people to answer as they see fit; it basically just reduces cause i) and does nothing about ii) or iii).

My own stance at this point was straightforward, unsophisticated and arguably something of a circular conclusion. It is dodgy or inadequate writing skills that (almost always) provide the initial evidence that someone who claims an NL is not, in fact, an N of that L. So, taking the thread title at face value, if you want to verify NL, verify the written output. QED. I said as much several dozen times.

This conclusion then led me to ponder that this NL data item is not purely informative, it also serves a purpose, as it also does when we are asked to provide it on, for example, agency registration forms. NL is used to include and exclude translators from consideration for jobs/projects. As is address (less often, for sure). But never hair colour (AFAIK). So NL is not just information for information’s sake. I would contend that NL is used as a target language filter/selection criterion, for instance, because it increases the chances (without guaranteeing it) that the person will be able to write to an acceptable standard in the target language.

In short, it has a function, and within the proz context, quite an important one (given the impact on livelihoods). No matter what label it is given, it is the function that causes people to lie about it (as they do, apparently (I have no direct experience), about hair colour on dating websites) and so it is the ability to meet or perform that function that is important.

So, if you want to “verify” anything, while simultaneously removing causes i), ii) and iii) and resolving issues a) and b), verifying compliance with the function of the field – that of proficiency marker – is surely the way to go? Y’know, like agencies sometimes do when they ask us to do tests? (Since evidently proficient use of claimed NL is one testing criterion, along with the actual translation.)

In the context of the actual discussion in question, I suspect reaching such a conclusion, while intellectually satisfying, is fruitless, as the only changes the site is likely to make are those improving the width of Humpty Dumpty’s Henry’s wallet. In that respect, any improvement along such lines is essentially an intangible feel-good factor from resolving issue a), and if it is tangible, it may even be negative, since the number of people who care enough to actually leave the site if nothing changes is surely fewer than those who pay for membership safe in the knowledge they can claim a competitive advantage they don’t have and Henry will do SFA all about it and who would leave if anything did change. To say nothing of the cost of any verification scheme.

All of which led me to realise (or reaffirmed):

1. It would appear that I probably do not, in theory, have any objection to ‘non-native translation’ (if you see what I mean). Meeting the client’s needs and possession of the right skills to do so are the only factors that count. Where you were born should, in itself, be inconsequential.

2. Function trumps labels.  I think particularly of one of my own working ‘rules’ – always translate section/chapter headings last. Their function is often more important than the actual words themselves when it comes to translating them effectively. As with the NL label in proz, it doesn’t matter what it says, it matters what it does. Extrapolated to (or from, depending where you start) translation, it is broadly equivalent to the underlying idea behind “translate the meaning, not the words”.

3. It is in all our interests to get comprehensive independent credentials as evidence of our abilities rather than rely on commercial websites focused only on quantity not quality and therefore pandering to the lowest common denominator.

4. Threads about proz itself are always, always, going to generate the most traffic, no matter what forum they appear on. (To tell the truth, I knew that before.)

5. I should get out more.

Cripes, if I’d known this was how the post would pan out, I could have given it one of those proper bloody irritating blog post titles like a social media guru would, like “5 Things I Learned This Summer”.

So, anyway, that was my summer – how was yours?