Is Facebook up for making a big difference to migrants, under-resourced languages and ultimately itself?

Is Facebook up for making a big difference to migrants, under-resourced languages and ultimately itself?

The context

Imagine you’ve decided to learn a new language. Everyone has their own preferred approach depending on their end goal, of course, but as a rule, I’d guess that you’ll be looking for one or two language learning books, you’ll see if the new language is supported by Duolingo or Babbel, and maybe you’ll even check out the available YouTube videos and settle on a few helpful channels. If you’re a more outgoing person, you will probably look for an evening class or a community of native speakers to hang out with and soon afterwards you may also feel brave enough to immerse yourself in authentic content either online or by spending some time in the country. All in all, it’s not going to be easy, but it won’t be impossible, either, with all this choice around and the mass of useful resources. All you need to do is take your pick and if it doesn’t really work out, switch to something else.

Now imagine that you’re a kid who belongs to an ethnic minority that’s discriminated against regularly. Your parents are migrants, moving often without a plan that also includes you getting a proper education. They mean well, but they really don’t know how to succeed in the countries they’re taking you through, not staying long enough for you to learn anything much even if you manage to get a place in school. So you now speak your parents’ rare language that’s hardly ever written down (and for which there isn’t an agreed spelling, nor any schools for you to go to and learn it properly in), and you know some words in some other languages which you can’t read and write in, though – not even to a minimum standard. If you’re a girl, you grow up knowing you’ll have to marry at 16 and have kids and, if by some miracle you get to have a job, your ideal is to be a supermarket check-out lady. If you’re a boy, you have close to zero literacy and numeracy and dream of becoming a famous footballer because you have zero skills apart from holding your own in a fist-fight, and the adults around you do the worst menial jobs available and more often than not really aren’t role models for anything genuinely useful.

Too dramatic? @gr82tweet and I volunteered for a semester back in 2015 to work with 11 Romanian kids between the ages of 9 and 15 living in a local Rroma community and attending a local academy in Leeds, the UK, who were facing this very situation: they could speak Rromani, but not write it, and had very little knowledge of reading and writing in Romanian and English (as well as Italian and Spanish depending on their parents’ route to the UK).

The problem

When we were working with them on both their Romanian and their English, we found out that these kids had been learning to read and write by getting involved in their favourite Facebook groups – they had no other resources! The result was terrible, but their desire to get involved, be heard, and learn is the most important thing to take away from this! The problem with this approach is that, at least in the case of Romanian, learning the language online from posts and comments is a genuine minefield which not even adults are able to able to navigate, let alone unassuming children learners. Missing diacritics, misspelt words, atrociously ungrammatical sentences and wild punctuation are frequently combined in Facebook posts and comments originally written in Romanian (I’ll share with you the science behind how I know this in another post). Here’s an example:

This impacts all language learners, not just the kids I mentioned, which is a major problem because FB has the potential to be the ideal language learning environment thanks to its interactive nature, suite of built-in multimedia tools, and constant stream of relevant, engaging content. BTW, I have no issue with the use of non-standard language on social media, as long as there is a possibility to acquire the standard form, too. For Romanian, though, I don’t know of easily-accessible ones yet and, just like in the case of other languages, too, use of non-standard language alone is unlikely to enable people to break poverty cycles and lead fulfilling, meaningful lives.

Proposed solution

How do we find a sustainable solution to this issue? Well, what is more sustainable than combining the collective enthusiasm of the crowd with the latest tech developments offered by a popular social media platform? So here’s a thought: how about Facebook took advantage of its vast potential to enable its users to better themselves (which, btw, makes perfect business sense!) and, in the case of Romanian (and perhaps other languages, too, but I’ll let the community decide) made available the following:

  • A possibility to spot, classify and correct non-standard RO language; BTW, as such resources grow larger, this may even become an intra-lingual NMT (neural machine translation) post-editing exercise (from non-standard to standard RO), thus speeding up the correction process; needless to say, the inter-lingual NMT engines involving RO will improve their accuracy, too, so genuinely everyone will be a winner!
  • A self-managing, but FB-supported community of Romanian users with different levels of access to validate the corrections because good intentions aren’t unfortunately enough when it comes to writing grammatically-correct language.
  • Tracking of the influence of corrections on individual users’ written language.
  • How these corrections will be presented to the original users, as well as the rest of the community, will be a fun challenge for FB’s and Messenger’s UI designers, and I can’t wait to see the results.

So, what do you folks think about this? Would such a set-up be enough or do we need other tools/functionalities, too? More importantly, any folks working at FB out there willing to give this a shot? You know where to find me: @elearningbakery on Twitter.

I know this makes sense, especially given that FB is playing the long game: users who know how to write better are likely to get better jobs, earn more, and thus be more valuable assets long-term, so let’s get going raising the literacy level of this and any other groups facing the same issues!