ABSTRACT
This chapter concerns the processes involved in picture naming. It suggests that picture naming is mediated by access to at least three representations specifying different forms of stored knowledge about the depicted object. It represents which we will term the structural description, specifies the visual form of the object, most likely described from a canonical perspective. One aspect of a cascade system is that, within any given level of representation, the stored descriptions of many stimuli can be activated in parallel. The discrete stage model and the cascade model make a number of different predictions. To illustrate this, consider the effect of structural similarity between different pictures on picture naming. We now present data relevant to these models, drawn from work with a patient with impaired picture naming and from work with normal subjects. Following the presentation of the data, we discuss their implications for interpreting neurological impairments of picture naming.
