The crib talk reported here of a 2;2-2;4-year-old boy replicates the phenomenon of crib talk reported in previous studies by other investigators. This study adds a corpus of mother-child interaction (MCI) and tests one aspect of the hypothesis that crib talk may enhance production of linguistic forms at a later date. Transcripts of monologues were scored for the presence/absence of two phonologically identical morphemes in obligatory linguistic contexts and in categories of metalinguistic awareness. Differences in form and function between crib talk and MCI are discussed. The discussion of form concentrates on accuracy of morpheme supplied where required, and frequency of forms produced before the point of acquisition. The discussion of function traces metalingual/poetic awareness, that is language play or practice, inquiry about a form, or focus on form or meaning of a word outside its usual context. It is shown that crib talk accuracy with morphemes is superior to that of MCI prior to acquisition. The findings support the hypothesis that some of the language performances required for analytic metalingual operations at a later date have their precursors in the crib. (AMH)