ABSTRACT
In a well-known paper, Bowerman discusses the development of the use of novel causatives in her own children. Both children, according to Bowerman, at first produced only conventionally available intransitive and transitive forms. Since Bowerman published her analyses, other analysts have disagreed both about the nature of the generalization that children capture and about the productivity of the causative, both at the outset and later. Maratso's concluded that the chronology Bowerman reports makes it likely that some kind of productive process is going on but that it may be a highly marginal productivity. Both the naturalistic data and our experimental data indicate children for whom the causative becomes productive. But the data also indicate children for whom the causative never becomes productive, or for whom the productivity of the construction becomes marginal at best.
