Abstract: How has it happened that the competence-performance distinction has come to be seen as invalid, or if valid, irrelevant, or if relevant, actually harmful to psycholinguistic research? This paper suggest three reasons for the present obloquy of the competence-performance distinction. (a) The grammar of a language does not have an automatic performance interpretation. That is, a model of competence does not contain a specification of a model of performance and does not entail a particular model of performance. (b) Candidate grammars keep changing. (c) In response to these two difficulties, psycholinguistics have attempted to specify performance independently of competence. To the extent that they have been successful and performance is unnecessary and that competence itself is not a useful notion.