The few studies that deal with machine translation (MT) as a language learning tool focus on its use by advanced learners, never by beginners. Yet, freely available MT engines (i.e. Google Translate) and MT-related web initiatives (i.e. Gabble-on.com) position themselves to cater precisely to the needs of learners with a limited command of a second language (L2). Anecdotal classroom evidence points to beginners availing themselves of MT help, even against the advice of teachers. In order to find out whether MT could help develop learners’ writing skills in L2, we ran some tests asking participants to write directly into L2 in one instance and into L1 in another, while pre-editing the L1 and post-editing the L2 within the MT’s Tradukka.com interface. Analysis of the output produced has found that MT helps beginners to communicate more, particularly when they had a lesser mastery of the language. The less their mastery of the L2, the greater the difference between the number of words composed with the help of MT and the number of those written directly into L2. It also helped them to communicate better, with blind marking indicating higher quality when writing with MT mediation. Looking at the screen recordings, on the other hand, we found that writing directly into L2 requires more effort, as measured by number of pauses, and involves more engagement with the task, as measured by the number of editing interventions.
- computer-assisted language learning,
- machine translation,
- foreign language teaching and learning,
- second language acquisition,
- computer-assisted translation
Available at: http://works.bepress.com/ignacio_garcia/4/