Luis Von Ahn of Captcha / ReCaptcha fame, talks about his new crowdsourcing start up – Duolingo. (here’s the N.Y. Times article I tweeted previously) He envisions harnessing the large “desire” to learn English which exists, proposing a model whereby learners don’t pay but rather get paid for learning English.

How? Well, by translating the web.

I’m not too convinced, for two big reasons.

1. There remains the fact, despite the illusion out there, that overt crowdsourcing is participated in by many millions. In fact, the crowd is rather small (think of how few are those that edit wikipedia, for example). Clay Shirky has recently noted this problem when it comes to more intensive forms of collaboration/effort. (go here for great discussion and video lecture) We also have reports of massive opting out of crowdsourcing. Learning in this way takes significant individual motivation and I’m not sure this will overcome that hurdle.

2. Is translation a way to learn a language? He makes big pronouncements about this as “language learning” but to me, it is a return to an old and thrown into the dustbin of history methodology. Translation helps learners learn English and “mediate” between languages but it is far from being a “method” of language acquisition. See Scott Thornbury’s fine blog for more on this subject. Reminded me of this great blog post and comment thread by Kirsten Winkler: How Online Language Teaching Start Ups lack educational expertise.

What do you think? Is this headed for success? Will it solve a valid problem (students needing to learn English but not having the means or proper environment)?