Today I came across a great new game from Google that would be lots of fun for English-language learners: Spell Up. (Attention Firefox or Safari users: it only works with the Chrome browser.)
A sort of game-ified spelling bee, you hear words, you spell them into your microphone, and if you’re correct you build the words into towers. You also may be asked to pronounce the word, unscramble words, or be given a word missing letters, which you must identify and pronounce. An addition to building towers, you have all the usual game-y bells and whistles, such accumulating points, getting “power ups”, etc.
Some neat features for learners are not only the ability to repeat the word spoken, but to get a translation into another language, or to hear a level-appropriate verbal definition of what the word. There’s also a feature where you can click on any letter of the alphabet and hear how it’s pronounced—a great review for low level and high-level learners alike. When you begin the game, you choose your English level—beginner, intermediate or expert—so learners can play the game at a level appropriate to their language development.
This game draws on Google’s text-to-speech and speech recognition systems, as well as their translation engine and dictionaries. It’s nice to see another useful and educational application of Google’s powerful and wide-ranging technological empire. The speech recognition system is not without its shortcomings, though. You have to spell words out unnaturally slowly, pausing between letters, so that the system can recognize them. On a few occasions I had to repeat letters several times in order to get the system to recognize them, which was very frustrating. Was it my microphone, their system, or was it me? If this happened to me, a native speaker of the North American English I’m assuming their speech recognition engine was modelled after, what will happen in response to my students’ pronunciation?
There’s also the issue of what and whose pronunciation will be recognized, both on the letter level and the word level. Will it recognize a range of regional, social and learner Englishes? When I was asked to pronounce a word, I was told my pronunciation was “pitch perfect”—what will it say to others? Will it judge acceptable a learner English that would be perfectly acceptable in human-to-human interaction? Or will you be able to say simply anything into the microphone and it will tell you you’re correct? These are important questions that will affect the playability of this game. I’ll have to share this game with some of my students and get their report back on it.