Prospective representation: The effects of varied mapping of sample stimuli to comparison stimuli and differential trial outcomes on pigeons’ working memory

Absence of immediate transfer of training of duration symbolic-matching-to-sample in pigeons

“…One possibility is that the use of multiple sets of choice stimuli encourages some form of sample categorization, which induces the subject to differentiate samples according to signal type, as well as duration. The nature of the relationship between samples and choice stimuli has been shown to influence the tendency of pigeons to categorize or .. commonly code" samples (Grant, 1982;Santi & Roberts, 1985;Urcuioli, Zentall, Jackson-Smith, & Steim, 1989;Zentall, Urcuioli, Jagielo, & Jackson-Smith, 1989). In transfer studies that have maintained the same set of response alternatives for two different signal types, the sample-response relationship has been like a many-to-one arrangement-each correet choice alternative has been associatedwith two different sample types of the same duration.…”

Transfer of performance to new comparison choices following differential outcome matching-to-sample

“…If so, transfer of those choices to other samples associated with the same outcomes can be interpreted in two-process terms. To avoid this situation, the new comparison responses (and their differential outcome associations) in Experiment 2 were learned in a task in which (1) each sample stimulus was associated with more than one correct comparison choice (i.e., one-to-many matching; see, e.g., DeMarse & Urcuioli, 1993;Santi & Roberts, 1985;Urcuioli, Zentall, & DeMarse, 1995), (2) each correct response in a pair of choice alternatives yielded a different outcome, and (3) each sample was followed equally often by each outcome. In short, the samples alone did not signal which outcome was scheduled on any given trial.…”

Generalization gradients following training on a symbolic matching-to-sample task

“…The finding that line orientation controlled behavior is theoretically important in light of the fact that matchingto-sample is considered a relatively demanding task (see Honig & Wasserman, 1981), involving both short-and long-term memory processes (Roitblat, 1980;Santi & Roberts, 1985). It has been suggested (Boneau & Honig, 1964;Sutherland & Mackintosh, 1971) that when a task places a high demand on a subject's processing system, irrelevant features, such as a line's orientation, will not gain control of responding.…”