The Temporal Pattern of Unconstrained Drinking: Rats' Responses to Inversion and Identity Constraints

Abstract

Rats obtained all of their water by licking a metal tube during a series of daily 1-hour sessions. When the tube was freely available throughout, each rat showed the classic temporal pattern of unconstrained drinking: As the session progressed, drinking bouts generally grew shorter and pauses grew longer. In subsequent sessions the tube was opened and closed independently of the rat's behavior, on a schedule that gave the rat a chance to duplicate the exact inverse of its unconstrained baseline pattern. Thus, as the inversion session progressed, the opportunities to drink generally grew longer and the enforced pauses grew shorter. When the rats were forced away from their unconstrained patterns of drinking and pausing, their total time spent drinking consistently fell short of previous values, but total licks and volumetric intake remained at previous levels. The same results occurred under an identity schedule, a series of openings and closings that duplicated the unconstrained pattern of drinking and pausing. The results have implications for theories that assume that instrumental performance under schedule constraint derives from the animal's defense of a measured set-point.Key words: drinking, bouts, pauses, unconstrained pattern, inversion constraint, identity constraint, rats Numerous regulatory models assume that performance under the constraints of a schedule derives from the animal's defense of some set-point for the kinds of behavior controlled by the schedule (Hanson & Timberlake, 1983;Heth & Warren, 1978;Rachlin & Burkhard, 1978;Staddon, 1979;Timberlake, 1980;Timberlake & Allison, 1974). The set-point is commonly defined in terms of the total amount of responding typically observed in the absence of schedule constraint, in constant-duration sessions that allow unrestricted performance of the kinds of behavior under study. Various tests of the set-point status of a particular behavioral total have been performed or proposed (Allison, 1981(Allison, , 1983 The present paper introduces a method of rearranging the component parts of the unconstrained behavior, while forcing no change within any part. The question is whether the animal, forced out of its unconstrained pattern, will still reproduce the unconstrained total. If it does, then the underlying pattern would seem to be irrelevant. If it does not, then the sequential organization of the behavior would seem to be a pertinent system variable with a set-point of its own.As a preliminary step, we recorded the unconstrained behavior of the thirsty rat in a series of daily 1-hr sessions of free access to a water tube. These records provided information about the rat's drinking patterns in the absence of external constraint, including total time spent drinking, the duration of each drinking bout and each pause, and the place of each duration in the entire sequence of bouts and pauses.Next we tested whether the rat would invert its unconstrained sequence of bouts and pauses, if that were the only way it could maintain its total time spent drinking. We did ...

System-specific differences in behavior regulation: Overrunning and underdrinking in molar nondepriving schedules.

“…In paired-baseline, 85% of the total water intake occurred within the first 10 min of the session. These results are typical for the temporal distribution of water intake for deprived rats (Allison et al, 1986; Hatton & Bennett, 1970; Rolls & Rolls, 1982; Gawley et al, 1986). The current constraints reduced this proportion to approximately 25% within the first 10-min bin and forced the rats to consume a larger proportion of water over the remainder of the session.…”

“…It may be that the instigation for drinking is tightly related to the baseline temporal distribution and that local disruptions of this response characteristic effectively reduce the overall instigation for drinking. Allison et al (1986), however, reported that rats maintained baseline water intake when the opportunity to drink was the exact reverse of the baseline temporal pattern of drinking. However, these results are difficult to compare because the rats in that study also had access to food in the experimental chamber, which might have stimulated postprandial drinking.…”

“…Thus, these molar models predict that total baseline responding should be maintained when the terms of the schedule are nondepriving at the molar level (see Timberlake, 1980). In effect, these models assume that motivational systems regulate deviations from response totals rather than from local response characteristics (e.g., burst and interburst durations, temporal distributions of responding—see Allison, Moore, Gawley, Mondloch, & Mondloch, 1986; Dunham, 1977; Gawley, Timberlake, & Lucas, 1986; Premack, 1971; and Timberlake, 1980, 1984, for further discussion of local and molar regulatory processes).…”

Bliss points, stop lines, and performance under schedule constraint.

“…When both models failed in Experiment 3, we were dismayed by the failure but pleased with the power of the test. A feature that the two models have in common is their concern with behavioral totals, rather than the path the individual follows in generating those totals (Allison, Moore, Gawley, Mondloch, & Mondloch, 1986). Perhaps better models can be developed with the help of a more detailed account of the behavioral patterns displayed with both spouts were freely available.…”

Response Deprivation, Reinforcement, and Economics

“…For example, if the contingent requirement is defined in terms of the number of licks at a water spout, the rat may achieve its baseline water intake before reaching its baseline number of water licks (Buxton & Allison, 1990). Symmetrically, if the contingent requirement is defined in terms of volumetric intake, the rat may achieve its baseline number of water licks before reaching its baseline intake (Allison & Buxton, 1992; see also Allison, Moore, Gawley, Mondloch, & Mondloch, 1986). In any event, the suppression of both responses under a nondeprivation schedule suggests that response deprivation may not be sufficient for reinforcement.…”