Volume 59 Issue 4, June 2003, pp. 567-587

In the first part of the paper, I challenge some basic assumptions underlying the claim that reading is the major source of vocabulary acquisition in L2: the ‘noticing’ assumption, the ‘guessing ability’ assumption, the ‘guessing-retention link’ assumption, and the ‘cumulative gain’ assumption. In the second part, I report on three experiments in which vocabulary gains from reading were compared with gains from word-focused tasks: completing given sentences, writing original sentences, and incorporating words in a composition. Results showed that more words were acquired through tasks than through reading.