The analysis of the pre-choice behaviors in an operant conditioning task led to the observation that pigeons often produced anticipatory pecks that were directed at the location of their next response. Despite the possible utility of this behavior for understanding basic behavioral processes in animal learning and the widespread use of touchscreen displays to present pictorial stimuli, there has been little evaluation of the spatial distribution of touchscreen responding. So, we sought to investigate the mechanisms that account for this anticipatory behavior, whether this behavior changes over time, and how general this phenomenon might be. To answer these and other related questions, we report in a series of eleven studies and two re-analyzed datasets a detailed characterization of this anticipatory discrimination behavior in both pigeons and rats.
In the first chapter, we review relevant literature related to the phenomenon of anticipatory behavior and prospective coding. In the second chapter, we outline a basic three-link discrimination paradigm, which we adapted from a procedure originally developed to study spatial anticipation in autoshaping. This simple procedure afforded us the ability to measure responses during a task that engages prospective processing.
In the third chapter, we evaluate two possible mechanistic explanations for this anticipatory behavior; namely, that animals are motivated to produce anticipatory responses because of a shorter temporal route to reinforcement or because of the spatial and temporal contiguity of the stimuli used in the task. In the fourth chapter, we evaluate several spatial parameters that might importantly influence the distribution of these anticipatory responses. In the fifth chapter, we re-evaluate data from two previously published projects to assess the generality of the observed phenomenon and to evaluate the possibility that the anticipatory responses are a fractional reproduction of the terminal response.
Finally, in the sixth chapter, we discuss the implications for the presented work in several fields. We also sketch a computational framework for the presented data using a Dynamic Field Theory model, attempting to show how the prospective representation of an upcoming spatial location might guide anticipatory behavior.