THE DEVELOPMENT OF FIXED‐RATIO PERFORMANCE UNDER THE INFLUENCE OF RIBONUCLEIC ACID¹

Abstract

The transition from fixed-ratio 1 performance (every response reinforced) to fixed-ratio 30 performance (every thirtieth response reinforced) was studied in nine pigeons. These were divided into three treatment groups given daily oral doses of saline, or 250 mg/kg/day or 500 mg/kg/day of yeast ribonucleic acid. Detailed computer-assisted analyses of how fixed-ratio behavior develops revealed the following typical sequence. After the transition, the first few ratios typically were emitted without long interresponse times within the ratio. Steady responding then ceased, and numerous long interresponse times occurred, with no systematic relationship to ordinal position within the ratio. Gradually, a new pattern evolved, characterized by a consistently long post-reinforcement time, a border region of the next few interresponse times within which the mean interresponse time monotonically decreased, and short interresponse times within the last 80% of the ratio. Long interresponse times were eliminated from this last section of the ratio without regard to proximity to reinforcement. Various analytical procedures suggested that the final pattern can be conceived, in part, as the shaping of a reliable response topography. The group of three pigeons given 250 mg/kg/day of yeast ribonucleic acid responded at higher rates than the saline and 500 mg/kg/day groups. The latter group, in contrast to the saline and lower dose groups, which continued to increase their rates, reached a rate asymptote very early.

Steady‐state Performance on Fixed‐, Mixed‐, and Random‐ratio Schedules

“…Thus, Weiss and Gott (1972) found that the peaks of IRT distributions from pigeons on a FR 30 schedule sometimes shifted when subjects were administered amphetamines or other drugs. Furthermore, in several studies two or more peaks have been observed in IRT frequency distributions from FR schedules (Gott & Weiss, 1972;Mazur & Hyslop, 1982;Weiss & Gott, 1972). Gott and Weiss (1972), for example, found several peaks at multiples of about .3 sec, and they reported that the multiple peaks were the result of one or more ineffective pecks occurring between two recorded pecks.…”

“…Furthermore, in several studies two or more peaks have been observed in IRT frequency distributions from FR schedules (Gott & Weiss, 1972;Mazur & Hyslop, 1982;Weiss & Gott, 1972). Gott and Weiss (1972), for example, found several peaks at multiples of about .3 sec, and they reported that the multiple peaks were the result of one or more ineffective pecks occurring between two recorded pecks. Yet regardless of the shape of a particular IRT distribution, the results of Mazur and Hyslop (1982) and of the present study suggest that changing the ratio size or the type of ratio schedule will not substantially alter the shape of that distribution.…”

Serial properties of behavior and their chemical modification

“…This is apparent from an experiment with pigeons trained on fixed-ratio schedules, in which recording the behavior in real time permitted a close examination of the acquisition process. Gott and Weiss (1972) traced the development of mature ratio performance in several different ways, beginning with the transition from continuous reinforcement to FR (fixed-ratio) 30. As Ferster and Skinner (1957) first demonstrated, the transition process is ragged.…”

“…Second, the proportion of very short IRTs increased. Gott and Weiss (1972) termed these "nibbles," because they were produced by rapid beakoscillations, not pecks. Third, the mode at about 600 msec was almost totally eliminated.…”

“…But temporal patterning cannot be neglected even in behavioral accounts of ratio performance. The "molecular" model of ratio performance described by Mazur (1982), for example, could be expanded to incorporate the findings of Gott and Weiss (1972) on the distribution of IRTs, especially the intrusion of pauses. Further, the differential response to drugs of the different IRT clusters indicates that molecular accounts probably need to deal with variable topographies .…”