Chapter 7. Task complexity, modified output, 
and L2 development 
in learner–learner interaction

This chapter examines the effects of task complexity [± reasoning demands] on modified output and the relationship between output modifications and L2 development. Seventy-nine adult English as a Second Language learners were divided into two groups: (1) low reasoning demands; and (2) high reasoning demands; and engaged in two sets of tasks which targeted English past tense and locative prepositions. While learners modified their output using a variety of modification moves, learners who completed high complexity tasks produced more self-repair than those who completed the low complexity ones. Self-repair was related to learning locatives for the high complexity group only as measured by delayed grammaticality judgment and oral post-tests. Pushed output, as well as the total amount of modified output, was related to learning past tense for the low complexity group only as measured by delayed grammaticality judgment post-tests. These findings are discussed in light of the Cognition Hypothesis and show an intricate pattern among level of task complexity, type of target structure, type of modified output, and learning.