Autoshaped acquisition, fixed ratio maintenance extinction, omission, reversal, and choice of different responses: an investigation of biologically prepared responding

Experiments 1 and 2 examined acquisition rates of autoshaped key pecking: (1) as reinforcement occurred independently of pecking and contingent upon pecking, respectively, and (2) in two positions of the floor-key. Neither factor influenced autoshaped responding, nor did variable (1) produce reliable differences in subsequent experiments. However, wall-key pecks were readily autoshaped, but reliable floor-key responding failed to occur in both experiments. These experiments suggested that failure to autopeck the floor-key was due to extraneous stimuli that occurred with key illumination. Such stimuli occurred for floor-key subjects, as their position between the houselight and floor-key shadowed areas surrounding the floor manipulandum. Thus key-light onset illuminated these shadowed areas thereby presenting cues that approximated predictability of the key-light in signaling magazine presentations. This account is consistent with previous research that demonstrated inhibition of autopecking when redundant stimuli occurred during CS presentation.

Fixed ratio schedules (Experiment 3), extinction (Experiment 4), and omission training (Experiment 5) each produced similar session response rates and cumulative response rates within sessions across wall and floor-key groups. Responding also failed to vary in distribution of individual low and peak response rates across sessions and in subjects increasing or decreasing response rates in session to session comparisons in these experiments.

In Experiment 6 former wall and floor-key subjects pecked the floor and wall-key, respectively (reversal condition) to allow within-subject comparisons of response rate on each key. Wall and floor-key response rates were similar in session response rates, cumulative responding within sessions, and in subjects showing daily response rates more rapid than in the previous session. Response rates on the wall and floor keys (during baseline and reversal) also failed to vary reliably when comparisons were made within both wall and floor-key groups between response key conditions. However, individual comparisons of reversal and previous baseline responding tentatively suggested (and Experiment 7 confirmed) that more rapid responding occurred on the floor-key.

In Experiment 7 reinforcement followed any combination of 20 pecks on either or both key(s) thereby allowing within-subject comparisons of floor and wall-key response rates and response allocation within sessions. Across seven sessions average floor-key pecks always exceeded 50 percent of total responses and 15 (Session 1) to 18 (Session 7) pigeons emitted more floor-key pecks per session excepting Session 2 where 13 so responded. Preferred floor-key responding was likely due to the usually more rapid response rates per reinforcement on the floor-key. Individual subject ratios of wall-to-floor-key response rates per reinforcement were correlated with individual ratios of wall-to-floor-key responses. This relation was described by the matching law for ratio schedules. Analysis of the anatomical basis of floor and wall-key responding suggested that floor-key pecking offered a mechanical advantage favoring rapid response execution.

Overall, these experiments failed to confirm instances of biological preparedness to associate stimuli and responses, yet the unsuspected redundant stimuli present in Experiments 1 and 2 may have obscured such an outcome during autoshaping. However, Experiments 6 and 7 lend convincing support to the presence of biological influences in the execution of responding and choice between responses.