The effects of cocaine on pigeons' choices between fixed-ratio and geometrically escalating schedules: A preliminary study.

Abstract

Listen

Translate article

Choice patterns over time in situations of diminishing returns have been compared to computations based on the Sums-of-Reciprocals Principle or the Linear Optimality Principle. The present experiment was designed to examine the effect of cocaine on choice in situations of diminishing returns, using the above averaging techniques for reference lines to identify potential invariances. After determining appropriate dosages, pigeons were exposed to a concurrent-chains schedule in which brief initial links led to either of two terminal links. One terminal link consisted of 15 responses (FR15), and the other was a geometrically increasing progressive-ratio schedule (GPR) whose rate of escalation was determined with a multiplier of 1.15. A selection of the FR schedule reset the GPR to its initial response requirement of five. Consistent with earlier research, computations based on the sums-of-reciprocals principal better described choice patterns during baseline (nondrug) conditions. However, patterns shifted towards the reference line generated using the linear optimality averaging principle during the drug conditions. Although neither technique precisely accounted for behavior, it is clear that behavior patterns were systematically altered by cocaine in a way that indicated sensitivity to reinforcement prospects over longer periods of time than the sums-of-reciprocals principle predicts. Marked residual effects on choice were also visible during the return to baseline, as behavior patterns were slow or did not return to those during initial baseline. This study provided a preliminary analysis of the effects of cocaine on choice in situations involving diminishing returns. Key Words: cocaine; resistance to choice patterns; concurrent-chains schedule; persistence; key peck; pigeons. ********** This experiment was an attempt to formulate some elementary principles concerning behavior patterns affected by acute and chronic drug procedures. Both biological and economic traditions include accounts of behavior patterns in homogeneous environments that involve optimization of resources. In the laboratory, investigators of behavior generate reference lines, or potential invariances in net gain, to predict and describe behavior (Nevin, 1984). Two approaches that are often distinguished from one another are short term, or molecular (e.g., Hinson & Staddon, 1983; Navarick, 1979) and long-term, or molar (Baum, 1973). In choice situations, a molecular approach is based on analyzing the moment-to-moment proximity of reinforcers available from each of the available alternatives, and evaluating the sensitivity of behavior to them. A Molar approach involves long term consideration of actions and the consequences of those actions, involving an extended pattern of choice or series of selections, and the degree to which the remote reinforcers influence choice. Experiments that involve choices between fixed-ratio (FR) and progressive ratio (PR) schedules of reinforcement have provided evidence relevant to assessing the degree of molarity characteristic of choice patterns. Initially, the PR response requirement is smaller than the FR response requirement and therefore preferred. In a concurrent-chains procedure, a switch point is the point at which the subject selects the FR alternative after several selections of the PR alternative, resetting the PR to its initial response requirement. Switch points can be compared to reference lines generated by the linear optimality principle (a molar computation) and the sums-of-reciprocals principle (a self-limiting molar computation), allowing the experimenter to identify potential invariances. In their experiments with chimpanzees, Hodos and Trumble (1967) studied schedule selections in a recurring choice situation where the alternatives were a progressive-ratio (PR) and a fixed-ratio (FR) schedule of reinforcement. Both alternatives were concurrently available at every choice point until a response was made, making the non-selected alternative unavailable. …