Rats were maintained on a 23_5-h water deprivation schedule and trained in CRF operant responding for water_ During the daily _5-h operant session, measures were made of response duration at successive I-min intervals under both reinforcement and extinction conditions, and comparisons were drawn between the following groupings of the data: (1) reinforcement for the first 5 min of responding, (2) reinforcement for the last 5 min of responding. (3) extinction for the first 5 min of responding, (4) extinction for the last 5 min of responding_ The results were that both extinction and satiety increased the mean and variability of response duration such that grouping 4 > 3, 2 > L The results are believed to support a common variability hypothesis for both kinds of behavioral change_ Response duration in operant re~ponding decreases in both mean and variance with continuous reinforcement contengencies and increases in mean and variance with extinction (Hurwitz, 1954; Schwartz & Williams, 1972).Hurwitz (l954) also noted these effects with "extended rewarded practice," an effect he attributed to either fatigue or a reduction in motivation. Subsequent work by Crow, Coop, and Carlock (1977) and Crow, Westveer, and Kass (l976) indicated that the increases in mean and variance of leverpressing duration associated with such extended rewarded practice were not fatigue-induced, but due to satiation.That satiety is closely related to processes of extinction, and that there may be "resistance to satiation" analogous to resistance to extinction has been pointed out previously (Morgan, 1974). and it is possible that the similarity in response duration change occasioned by both extinction and satiety is due to a common element. It has been proposed by the present author (Crow, 1977) that behavioral variability is proportional to the tolerance of the environment to behavioral alteration, and the question arose as to the relevance of this hypothesis to the increased variance seen in satiety and extinction. If both kinds of variance changes are due to common ultimate effects of tolerance changes, a direct comparison of response duration variances should reveal a specific ordinal relation between situations in which either one or both effects were present.
METHOD
SubjectsSixteen female rats 100 to 150 days of age were used. The animals were taken from a basic Long-Evans strain from the Western Washington University colony.
ApparatusThe apparatus has been described elsewhere (Crow et al. , 1976). It consisted of four water-delivery operant chambers 86 equipped to measure response duration of leverpressing at any interval within a drinking session.
ProcedureAll animals were adapted to a 23.5-h water deprivation schedule and trained to a stable level of leverpressing before measures of satiation were begun. After a stable level of response duration had been attained, measures were made of number of responses and average response duration for each minute of the 30-min leverpressing (drinking) session_ The first and last t1ve I-min segmen...
