I’m blogging in English because English is what I speak best. Other members of the Sharek961 team could blog in Arabic or French (I could too, but you wouldn’t want to read it!), but those people are busy with other key functions - getting the designs just right, reaching out to local organizations, identifying semantic markers to bring you the most relevant RSS… so for this one member of the Sharek961 team, with my limited dev skills, it’s blogging. And in English.

But how does blogging in English match our mission as a project for Lebanon, an often multilingual but overwhelmingly Arabic-speaking country? The honest answer is that it doesn’t. And it’s one of the issues we’ve had to confront as we’ve developed this site: because we are working with partners inside and outside of Lebanon, and because our team is both Lebanese from Lebanon and Lebanese from the diaspora, our primary working language for this project is English.

What’s the big deal? Why does it matter if we’re in Arabic?

One: From the beginning we’ve been committed to delivering a bilingual site. Although internet access remains limited across Lebanon, and according to Meedan (via Google) only about 0.4% of the web is Arabic language content, Lebanon is an Arabic speaking country first and foremost, whereas English and French tend to reflect generational or socio-economic biases. Therefore, we feel that it is essential to make Sharek961 accessible in Arabic, in order to reach the broadest possible linguistic community.

Two: In Lebanon, English speakers are disproportionately concentrated in urban areas, so if we were to outreach and publish only in English, we would be consciously underserving the non-urban population. As political affiliations may fluctuate with urban versus rural settings, in addition to reinforcing socio-economic biases, we could trip over political biases too. As a project committed to neutrality, we must exist, publish, and outreach in Arabic.

Three: We want to avoid being just another project for the diaspora or the elite. Initiatives like Sharek961 have become, frankly, sexy; people like foreign countries doing neat things with technology (we’ll leave the analysis of why to the disciples of Edward Said). We agree with George’s post on the Meedan blog; too many projects are designed for an international community, rather than the local one they purport to serve. There are legitimate reasons for making your project accessible to an international audience - raising your profile, promoting your cause on a bigger stage - but for Sharek961, there is no success unless it is local success.

How should we handle content on a bilingual site?

Now that we’ve established that we need to exist as a bilingual project, what’s the best way to do this? Do we translate all content, so that it is consistent and equally accessible to both Arabic and English speakers? Or do we allow for a natural bifurcation, allowing Arabic speakers to address the issues most relevant to them, while English speakers identify their own priorities?

If we translate one-for-one, it’s labor intensive, whether by automated machine translation or human volunteers. By indiscriminately translating all content, we add a haphazard element to content that undermines context and meaning. If we selectively translate ‘good’ content, by the simple act of choosing we add subtle editorial to what we’ve pledged would be a neutral project. If we don’t translate, we risk depriving our audience of important information that might be sourced by only one of the linguistic communities.

Our initial preferred way of addressing translation was to offer all content in the original to both language groups. At first we planned to offer a fully bilingual site interface, with English and Arabic presented simultaneously, so all users would see the same thing. Posts, reports and content in Arabic and English would go up side by side. Arabic speakers could read Arabic, English speakers could read English, and our bilingual or semi-bilingual audience would have both. It wasn’t ideal, but it was egalitarian.

We discussed adding capacity for crowd-sourced translation at a later point - but we had no way of verifying the accuracy of those translations. For example - someone sends in Arabic: “Zaatar W Zeit is the best late night food!” and the translation ends up reading in English “Zaater W Zeit? You crazy? Barbar is much better!” As Sharek961 expects to handle reports a good deal more sensitive than the best late-night food spots in Beirut, we were worried about the significant potential for abuse. Without close moderation, something we are not currently equipped to provide, this could pose more problems than it solved.

Which is where Meedan comes in… but that’s a subject for the next blog post.