“…Comparing children's performance in studies with systematically different numbers and duration of training and testing trials can help us to understand whether children require a certain training regimen to acquire or improve their referential communication skills. Although children's initial communicative attempts are usually not sufficiently informative and often include a communicative failure, some early training studies demonstrated that children can repair their communicative failures by (1) directing their attention to the contrastive features of objects in referential settings (e.g., Asher & Wigfield, 1981;Lefebvre-Pinard & Reid, 1980), (2) observing adult models who are competent in referential communication (Whitehurst, 1976;Whitehurst, Sonnenschein, & Ianfolla, 1981), (3) experiencing (rather than observing) communicative breakdowns of their own (Deutsch & Pechmann, 1982;Robinson & Robinson, 1985;Sonnenschein & Whitehurst, 1984), and (4) getting explicit feedback about the reason(s) for the inadequacy of their communicative message (Sonnenschein & Whitehurst, 1984;Whitehurst, 1976;Whitehurst et al, 1981). A recent study (Matthews et al, 2007) found that the best way to boost children's referential performance was by enabling them to experience communicative breakdown and repair.…”