Abstract
According to a model of inter-theoretic relations advocated by Patricia S. Churchland, psychology will need to revise its theories so as to fit them for "smooth reduction" to the neurosciences, and this will lead to the elimination of reference to intentional contents from psychology. It is argued that this model is ambiguous; on one reading it is empirically implausible, on the other its methodology is confused. The connectionist program NETtalk, far from exemplifying the model as Churchland claims, suggests a theoretical rationale for employing relations to intentional contents in psychology