This paper is essentially a critical survey of different theoretical approaches to child language in general and to child phonology in particular. The author states her own conviction that language acquisition is different from language retention, that language acquisition is a non-unique process consisting of a multiplicity of devices, that language is essentially habit behavior, that phonology represents a separate language level, and that hearing is not conditioned by articulation. In considering various other theories of the child's acquisition of phonology, the author questions the theoretical foundations of much psycholinguistic research. She is especially concerned that psychological facts and measurements be kept separate from linguistic ones and warns against a tendency which she finds among many researchers to identify method with subject: "The fact that a linguist is able to organize his linguistic data into a coherent system does not necessarily mean that an infant acquires, stores or recalls these data in the same manner as evolved by the linguist." She feels that a great deal more research must be done and that much more empirical evidence, especially in the way of quantitative information, is needed before valid theories can be formed. (FWB)