Two experiments examined the ability of an added stimulus to interfere with
extinction of a target excitatory fear stimulus (a predictor of shock) in
human autonomic conditioning. Both experiments demonstrated disruption of e
xtinction when the added stimulus was inhibitory (a predictor of no shock,
or safety signal). Subjects showed a return of fear when the target stimulu
s was tested alone, on both self-reported shock expectancy and skin conduct
ance measures. The second experiment also demonstrated disruption of extinc
tion when the added stimulus was excitatory. This result suggests that prot
ection from extinction may occur even when the added stimulus is not inhibi
tory. Additional factors that may contribute to protection from extinction
include context-specificity, occasion-setting and external inhibition. The
results highlight the role that concurrent stimuli play in extinction, and
emphasise the need to keep concurrent stimuli as similar as possible to the
desired transfer context in practical applications of extinction such as e
xposure therapy for anxiety. (C) 2000 Elsevier Science Ltd. All rights rese
rved.