the University of Illinois at Urbana-Champaign. In writing the original version, I had the privilege of working with Professor Charles E. Osgood, who is widely recognized as the founder and one of the leading figures of modern psycholinguistics. I have benefited from ex­ tensive and stimulating discussions with him, not only on this topic but in the development of his theory of language performance in general (see his Lectures on Language Performance, 1980, in this series). However, the re­ sponsibility for the particular formulations of the theory, hypotheses, in­ terpretations, and conclusions found in this work-which have been in­ fluenced, no doubt, by my training as a linguist, rather than as a psychologist-are my own.