Control of grammar in imitation, comprehension, and production: Problems of replication

A classic study of young children's language ( Fraser, Bellugi, & Brown, 1963) had revealed that comprehension of grammatical contrasts was superior to production of those contrasts. A reanalysis of that experiment indicated that the scoring procedures employed had introduced an artifact such that error scores on the production task were artificially inflated. A subsequent replication of the study showed that comprehension performance was superior to production performance only when the scoring artifact was present. Removal of the scoring artifact revealed no significant differences between scores on the comprehension and production tasks. The results were discussed in terms of their significance for theories of language acquisition.