Correct responses, error analyses, and theories of word production

Nickels and Howard (2004) argued that effects of word length on correct responses of a case series of nine aphasic individuals could be accounted for entirely in terms of number of phonemes, and suggested that effects of complexity on correct responses may have been overstated in the literature. In a commentary on this paper, N. Martin disputed some of Nickels and Howard's claims focusing on methodological issues concerning the differences between analysis of errors and correct productions, and the implications of this study for theoretical accounts of word production. Here we respond to those points raised.