More is not always better: a response to Roelofs, Meyer, and Levelt

In a series of papers we have argued that the distinction between lemma and lexeme levels of representation in lexical access may be unnecessary. We pointed out that the evidence cited in support of this view is not incompatible with alternative accounts that do not assume a lemma level of representation. Furthermore, we argued that there are neuropsychological observations and results from tip-of-the-tongue (TOT) experiments that appear to be problematic for the lemma/lexeme distinction. Roelofs et al. [Roelofs, A., Meyer, A.S., Levelt, W.J.M., 1998. A case for the lemma/lexeme distinction in models of speaking: comment on Caramazza and Miozzo (1997). Cognition 69, 219-230.] have challenged our conclusions by attempting to demonstrate that (1) a model that incorporates the lemma/lexeme distinction can account for the putatively problematic neuropsychological and TOT data; (2) there are other data that appear to be problematic for a type of model that does not include a lemma level of representation. In the present paper we respond to these criticisms by showing (1) that the neuropsychological and TOT data still represent a challenge for the lemma/lexeme distinction, and (2) that the other evidence cited by Roelofs et al. is not incompatible with lexical theories that assume only one lexical layer.