Rats received serial feature-positive discrimination training (A --> X+/X-)
within a conditioned lick suppression preparation (with + representing rei
nforcement and - representing nonreinforcement). In Experiment 1, the featu
re (A) was found to modulate responding to a transfer target CS that itself
had been occasion set with a second occasion setter, but not to a CS that
had been partially reinforced, thereby demonstrating occasion setting with
our preparation. Experiments 2-4 examined the effects on the acquisition of
occasion setting of various types of pretraining exposure to A. Nonreinfor
ced preexposure to A alone did not in any way retard acquisition and/or exp
ression of occasion setting (Experiment 2), but feature-negative discrimina
tion training did (Experiment 3). A feature-irrelevant pseudo-discriminatio
n pretraining procedure produced profound retardation (Experiment 4). The r
esults of Experiments 2 and 3 indicate that a latent inhibition-like effect
for feature-positive occasion setting exists that in some ways is analogou
s to latent inhibition effects in Pavlovian conditioning. The results of Ex
periment 4 indicate that an analogue of Pavlovian learned irrelevance can b
e evidenced for feature-positive occasion setting. The growing parallels be
tween occasion setting and simple Pavlovian conditioning lend support to ac
counts of occasion setting that view it as arising from an interaction of P
avlovian associations rather than from some distinctly different "higher or
der" process. (C) 1999 Academic Press.