W. A. Hunt et al (see record 1972-09232-001) argued that the shape of the relapse curves for smoking, alcohol, and heroin supports a learning conception of relapse in terms of decay of extinction. The present article shows, however, that the relapse curve cannot suggest or support a theory of relapse. Like all survival curves, it reflects the operation of 1 or both of 2 general processes--selection of the surviving individuals and changes in individuals' probabilities of survival--neither of which are specific to relapse. However, by comparing individuals who relapse soon after treatment with those who relapse later or remain abstinent, one may be able to identify the factors influencing probability of survival and to develop a theory of relapse. (5 ref) (PsycINFO Database Record (c) 2006 APA, all rights reserved). © 1979 American Psychological Association.
