Beliefs about whether or not children receive corrective input for grammatical errors depend crucially on how one defines the concept of correction. Arguably, previous conceptualizations do not provide a viable basis for empirical research (Gold, 1967; Brown & Hanlon, 1970; Hirsh-Pasek, Treiman & Schneiderman, 1984). Within the Contrast Theory of negative input, an alternative definition of negative evidence is offered, based on the idea that the unique discourse structure created in the juxtaposition of child error and adult correct form can reveal to the child the contrast, or conflict, between the two forms, and hence provide a basis for rejecting the erroneous form. A within-subjects experimental design was implemented for 36 children (mean age 5;0), in order to compare the immediate effects of negative evidence with those of positive input, on the acquisition of six novel irregular past tense forms. Children reproduced the correct irregular model more often, and persisted with fewer errors, following negative evidence rather than positive input.