Translating vozmob (2)

Ultimately, we’re hoping that projects like ours can help bridge conversations across communities that seldom talk together (but often do scream at each other). We believe that a useful starting point is for people to tell their own stories, so that others can begin to understand who they are rather than simply judge them by how they are portrayed. Thus the vozmob platform, making it possible for immigrant workers in L.A. to tell their stories in their own words, pictures, sounds and videos, using a readily available tool – the mobile phone in their pocket. That storytelling is of course immensely worthwhile in itself, as storytellers develop self and collective awareness, and build literacy skills. But for it to have a real impact beyond their community, the stories need to be accessible to people who don’t understand Spanish. This is not only a translation issue – we can learn to tell stories with pictures and sounds and video, as several stories on vozmob have done quite effectively, or with maps, and we’re interested in exploring those possibilities. But in many cases, translation becomes crucial – translation of text (SMS, MMS’s text fields, or text entered into the vozmob blog from a computer), but also translation of speech (story told into a phone, speeches or interviews recorded with a phone), soundtracks from videos, or texts that appear in photos. I mention all these different modes because each poses distinct challenges, each is amenable to different technical solutions, and to different translation processes. Here is a (non-exhaustive) list of things we’ve tried, thought about, or would like to explore:

– Text: our CMS currently allows any registered blogger to translate any blog post and its title, as written text. That can serve as well to transcribe speech, soundtracks or text on images, and to translate those transcriptions. So far, we’re relying on volunteers to do this (see examples at http://vozmob.net/translated or http://vozmob.net/traducido), and we have no ‘quality control’ process – no good way to note whether a translator/translation is good or not, to provide feedback, suggestions or edits to translators in a systematic way. We’ve talked about teaming up with a Spanish language class or with an ESL class offered by IDEPSCA, and about crowd-sourcing translations, which could be improved collectively on a wiki blog node. Any pointer to groups that do this well would be appreciated. We’re interested in both tools and process.

– Sounds: All we’ve done so far is the occasional transcription/translation into text mentioned above. An obvious improvement would be subtitles, in the original language and/or translated. We’re aware of sites like dotsub, Suggestions about doing something like that would be welcome as well (any dotsub-like plugin we could use with Drupal?). Also, we haven’t yet talked about dubbing, but that might be an interesting avenue to explore, especially since we soon will be able to send content back to phones on which reading text or subtitles may not be easy.

– Content management / context: Drupal’s own interface is mostly translated in Spanish (less thoroughly so in Portuguese or French). We’ve set up our site so that administrators can contribute translations back to the drupal mother ship when they run into something that’s missing. We’ve also used a somewhat haphazard approach to bilingual tags – e.g. we have used both ‘mujeres’ (http://vozmob.net/mujeres) and ‘women’ (http://vozmob.net/taxonomy/term/75) as tags, not really consistently. There is probably a drupal way to connect them… we should look into it.

– Machine translation and Text-to-speech is something we haven’t really looked into yet, except for occasionally generating first drafts of translations using google’s translate tool (with mixed success, partly because the original Spanish text often includes colloquial expressions, mis-spellings, etc).

I’m probably forgetting some aspects and will let others chime in. But generally, we’re interested in thinking through good ways to conduct these bilingual (or multilingual) multimedia conversations, we’d like to be in touch with others who are wrestling with this and who have found approaches that work, pointers to useful tools, etc. And of course, we always welcome volunteer translators – all they need to do is sign up on the site and get going.

(related post: https://vozmob.wordpress.com/2008/12/08/translating-vozmob-content/)

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s