by Corey Smith
It’s an idyllic concept: anyone in the world can go to any website and, at the click of a button, translate it into whatever language they choose. WikiLeaks in Farsi? No problem. KnowYourMeme in Russian? Of course. QQ, Tumblr, YouTube and every other social platform known to man in EVERY language? Revolutionary.
This is the goal of Duolingo, a collaborative project seeking to translate the web into the multitude of languages that are spoken around the world. Enthusiasts of TED (which stands for Technology, Entertainment & Design; a set of conferences run by the Sapling Foundation) may have heard of this concept before because one of Duolingo’s co-creators, Luis von Ahn, delivered a talk on web translation at TEDx CMU in January 2012.
It may seem like a lofty task, especially for the relatively small team at Duolingo. How does one organization translate the entire web with only so much manpower, hours and language ability? That’s where everyone else comes in. The point of Duolingo is to “kill two birds with one stone” as von Ahn puts it on Duolingo’s site blog. Duolingo seeks anyone who is looking to learn another language and is willing to test their ability against actual content that needs to be translated on the web.
Also in their blog, Duolingo introduced a main component of their project, a feedback feature. Their software can identify the exact mistakes made in a translation and relay the mistake as well as the correction to the user. In this way, von Ahn claimed the project was more like what would happen in a teacher/student environment.
The success of the project will lie in its robust usage. The more contributors actively practicing their language ability on Duolingo, the more content gets translated. Sentence by sentence. Page by page. Site by site.
Now, what’s the main draw for the user? Well, it’s free.
With the language leader, Rosetta Stone, costing upwards of $500, the playful language adventures of Muzzy clocking in at $200 and piles of expensive phrasebooks collecting dust, this would seem like an attractive option.
However, it is important to keep a few factors in mind before getting lost in the glitz of having an entirely translated web.
- The timeframe for a project such as Duolingo is lengthy and continuous. If anything is apparent, it is that the internet is a fluid outlet for information. Although the exact number is not known, there are well over 140 million registered top-level domain names (TLDs) with thousands being created, deleted or transferred on a daily basis. That’s a lot to keep up with.
- There may be plenty of people out there that fit the appropriate language learning abilities of Duolingo, but access to Duolingo in the first place may become a roadblock. Countries like Myanmar, Cambodia and Niger that have internet penetration rates below 1 percent of the population would be hard pressed to even reach the Duolingo site, let alone amass a large number of capable users able to translate sites into Burmese, Khmer or any of the Niger-Congo languages.
- The same can be said for repressive governments which already block many internet sites like Duolingo for purposes of keeping their citizens away from foreign ideas and information. Duolingo being accepted in places like Iran or Syria where the internet is used as a tool against the citizenry is highly unlikely.
- There is a probability that Duolingo will attract many mono-linguistic users who are looking to begin their acquisition of another language. However, this could present a problem because users with no previous language experience may come up against content that is too difficult for them to translate, despite the precautions Duolingo has taken by splitting up translation into comprehension levels. This could lead to a wave of interest and then disinterest as introductory level users leave and higher-level learners are far and few between.
Duolingo is currently tapping so-called “Masters” of languages to contribute to the feedback feature that will ultimately become the learning tool of the project. It is yet to be divulged how the project will be ensuring their language software will collect information from the web, reach out to websites to create pages in other languages and establish protocols for what websites will be translated.
The site is currently going through beta testing and is slated to launch to the public on June 19, 2012.
(One of Duolingo’s logos is used here for promotional purposes and following fair use and fair comment guidelines.)