If you were following the events in Egypt on Twitter and other social media and are not an Arabic speaker, chances are you were depending on near-real-time translation services provided through communities of volunteers that sprouted up with the speed of a flashmob. Welcome to the Brave New World of crowdsourced translation, a new model for localizing content that’s spreading from the edges of the Web to the business plans of increasing numbers of global companies.
Translation of online content typically falls into two categories: the very high end, where human experts are required to capture precision and subtlety, or the low end, where rough-and-ready machine translation can give readers the gist of a few paragraphs, give or take some comical errors. Now more organizations are experimenting with a third option: crowdsourced translation. This method relies on online communities to localize content into their native languages, distributing the workload across dozens, or perhaps hundreds, of contributors for greater speed, broader consensus on meaning, and lower cost.
Crowdsourced translation has been most successful on sites with strong communities and compelling content. Global Voices, the international community of news-and-events bloggers, depends on members and contributors to translate a large daily volume of multilingual content through its cunning Lingua
program. The TED Open Translation Project aims to create subtitles and transcripts to TEDTalks in more than 40 languages. Even Facebook users got into the act, putting the Facebook interface into languages ranging from the real but little-spoken Frisian dialect of Lower Saxony to fanciful “Pirate” English.
Typical of other massively collaborative communities like Wikipedia, these efforts repay their volunteers in personal satisfaction, social recognition, and the prestige of being affiliated with a worthwhile project -- everything but actual money. They also deliver an extremely valuable service to an organization at much greater speed and accuracy than it could otherwise afford.
Cheaper, faster, better... What’s not to like?
If you’re a professional translator, plenty.
The American Translators Association (ATA) has been warily watching the crowdsourcing trend for some time. They have no problem with volunteer efforts for good causes, but they are concerned about big commercial companies getting into the act. In the ATA’s view, what may look like a boon to multinational content providers and consumers suspiciously resembles Tom Sawyer’s famous scam of getting the kids from the neighborhood to whitewash his fence.
In 2009, the organization publicly took LinkedIn to task for trying to recruit volunteers from its network to localize content, rather than pay professional translators to do the work. “Linkedin is entering hazardous waters by attempting to use volunteers to perform professional services whose results will have a direct impact on the branding, image, and professionalism of a company that claims to be the premier professional networking site on the Web,” said ATA President Jan Stejskal at the time.
Some companies are taking the message to heart. Lionbridge Technologies Inc. , a large localization company based in Massachusetts, has been developing a model that seeks to blend the efficiencies of crowdsourcing with the high caliber of professional translation services. The company recruits specially selected communities of paid translators in its markets, rather than relying on volunteers, according to Nathalie Moliña-Niño, senior strategist behind Lionbridge’s crowdsourcing initiative. It then provides a lot of specialized back-end infrastructure, such as online glossaries and machine-assisted translation, to bolster the productivity of the community.
This rent-a-crowd approach seems like a good compromise, even if it results in lighter workloads for individual translators. Professionals get paid for their skills; organizations can count on high standards and reliable efforts on the kind of ordinary technical and legal projects that consume most localization budgets; and clients can still benefit from greater speed and reduced costs.
Crowdsourced translation not only points the way forward to a more global, multilingual Web, it’s also an interesting template for domesticating mass collaboration for professional, productive use without exploiting the good will of volunteers or compromising the added value of the community.
— Rob Salkowitz is the author of Young World Rising: How Youth, Technology and Entrepreneurship Are Transforming the Global Economy.