Two experiments studied the relationship between contextual conditioni
ng, reinforcer quality (food pellets, sugar solution), and the trial-s
pacing (TS) effect. In a contextual conditioning situation, Experiment
1 showed that Feeder responses (nose inserted into the feeder) were m
ore frequent after training with pellets than with the sugar solution,
but Rearing responses (standing on hind legs with front legs on the a
ir or on a wall) were more frequent with the sugar solution than with
pellets. In an autoshaping situation, Experiment 2 showed that when th
e solution is used as the reinforcer, a robust TS effect emerges in le
ver pressing as a result of very low levels of lever-pressing response
s in the massed condition. When food pellets are used as the reinforce
r, no evidence was found of a TS effect. The results of Experiment 2 w
ere analyzed in relation to the levels of Feeder and Rearing responses
, which were shown to be, respectively, noncompetitive and competitive
in relation to lever-pressing. A hypothesis based on response competi
tion can account for most aspects of the present results, which are al
so discussed in relation to models stressing the processing of conditi
oned and unconditioned stimuli and a mediating role for contextual con
ditioning. (C) 1994 Academic Press, Inc.