PI release as a function of the number of prerelease trials

In the first experiment the magnitude of PI release is examined as a function of the number of trials prior to a category shift. Using the sense-impression dimensions of “round” and “white,” the magnitude of release increased from about 6% after one prior trial to about 60% after four trials. A possible explanation of this effect is discussed, and the second experiment explores some implications of the proposed model. The nature of the relationship between the magnitude of PI release and the contents of the memory trace is discussed, and it is argued that until more is known about the operation of the memory system, it will not be possible to use PI-release data as a source of evidence for the manner in which words are encoded.