The second principle of Memrise’s approach is to remind users systematically. Using an algorithm developed by neuroscientist and cofounder Greg Detre, the app is designed so “plants,” or words, wilt when not tended to. The user interface tells users which plants are wilting, a problem they can remedy by “watering,” or repeated testing. Reminders pop up when a user is most likely to forget new words, rather than at random intervals.
The final Memrise principle is adaptive testing, which means that questions vary in difficulty according to the user’s performance. “Other language sites get this wrong,” says Cooke. “It’s really important that you test these memories at the right time and in the right way.”
Memrise isn’t the only social language-learning site on the Web. Others, like LiveMocha and Babbel, take a simpler community-based approach, in which users depend on other users for evaluation. These sites also have minor game components, offering points for achievements. But many users give up on learning a language remarkably quickly, says Cooke, and he believes that the learning techniques employed are partly to blame. “No other app uses more than one or two of these memory principles,” he says, referring to the three principles behind Memrise. Most rely solely on “non-choreographed” testing, he says, and fail to encourage users to recall newly acquired words.
Memrise is currently focused on getting users to memorize words, rather than teaching a deeper understanding of a language through grammar lessons or speaking. “It seems to work relatively well for teaching vocabulary,” says Luis von Ahn, a professor at Carnegie Mellon University and co-creator of a game-based language-learning website called Duolingo. “But that’s only a small part of learning a language.”