Stimulus functions of (i) reinforcing stimuli, and (ii) responses, were investigated in terms of their immediate after-effects. In several experiments rats responded on two-component differntial reinforcement of low rate (DRL) schedules. In these
schedules, designed for the present study, a response was reinforced only if preceded by a stated minimum interval of no responding; two such intervals,
(t1 and t2,) were specified depending on whether timing began from ( i) reinforcement, or ( ii ) a non-reinforced response, respectively. The findings were as follows :
1 . The IRTs which occurred following reinforced
and non-reinforced responses were different and were
appropriate to the minimum IRT requirements related to
these two events.
2 . The longer the IRT requirement following a
non-reinforced response the greater was the efiiciency
of performance following reinforced responses.
3. The temporal accuracy of responding was not
affected by the addition of exteroceptive stimuli
to either the t1 or the t2 component.
4 . When reinforcement was omitted from the t2
component the temporal accuracy of performance
deteriorated .
S. The omission of reinforcement following the
first response to occur after each reinforced response
resulted in relatively long pauses following
reinforcement .
6. Response bursts (i) occurred only following
non-reinforced responses, (ii) had a greater
probability of occurrence after responses on the t1
component than after responses on the t2 component,
(iii) occurred with greater probability following t1
responses as the duration of the t2 criterion increased,
(iy) occurred with a relatively high probability when
reinforcement was omitted following some "correct"
responses but declined following extended exposure
to such omissions.
When the concentration of the milk reinforcer was manipulated on four different schedules, namely DRL, fixed-interval, response-initiated fixed-interval and fixed-ratio, the duration of the post-reinforcement pause was found in each case to be an increasing function of the concentration of the preceding reinforcer.
Related changes in response rate were also observed.
These results were discussed with regard to the discriminative and inhibitory properties of reinforcing stimuli and responses in the temporal control of
behaviour. The data were also found to be consistent with frustrative non-reward theory (Amsel, 1958; 1962).