Use of a single-code/default strategy by pigeons to acquire duration sample discriminations

Abstract

340Duration sample discriminations typically involve a conditional discrimination in which one comparison stimulus is correct after a short-duration sample (e.g., 2 sec) and the alternative comparison is correct after a long-duration sample (e.g., 10 sec). Following acquisition, memory for these durations may be tested by inserting a delay between the offset of the sample and the onset of the comparisons. Typically, with increasing delays, pigeons' memory for the short (2-sec) sample remains consistently high, but memory for the long (10-sec) sample declines, often below chance (Spetch, 1987). This divergence of retention functions has b been called the choose-short effect, because after long delays, pigeons almost exclusively choose the comparison associated with the short-duration sample.One proposed explanation of the choose-short effect is the subjective-shortening hypothesis (Spetch & Wilkie, 1983; see also Spetch, 1987). According to this hypothesis, over time, memory for the sample duration progressively shortens. Thus, after a long retention interval, memory for a long-duration sample more closely resembles the memory for the short-duration sample experienced during training (when there was no delay). Memory for the short-duration sample also shortens with increasing retention interval; however, it is still more similar to the short-duration samp ple than to the long-duration sample, so matching accuracy does not decrease on these trials.It should be noted, however, that duration sample discriminations do not always yield divergent retention functions. For example, if pigeons are trained on a more comp plex conditional discrimination to choose one comparison r stimulus when the sample is either of short duration or vertical lines and to choose the other comparison stimulus when the sample is either of long duration or horizontal lines, when retention intervals are inserted between the sample and the comparison stimuli, the resulting retention functions are parallel not only on line orientation sample trials, but also on duration sample trials (Grant & Spetch, f 1993). This result suggests some flexibility on the part of pigeons and suggests that the way in which duration samples are represented may depend on the context in which the task is acquired.Divergent retention functions have also been found when the samples were not durations but, rather, the presence versus the absence of a stimulus. For example, Grant (1991) trained pigeons on a conditional discrimination, with food and the absence of food serving as samples. n One comparison was correct following the presentation of food, and the alternative comparison was correct when food had not been presented. During testing with retention intervals, matching accuracy on no-food-sample trials remained highly accurate with increasing retention interval, d and matching accuracy on food-sample trials declined rapidly with increasing retention interval (see also Sherd burne & Zentall, 1993). Thus, retention functions for food d and no-food samples look qui...

Testing the translational-symmetry hypothesis of abstract-concept learning in pigeons

“…If the subjects were defaulting to differ-r r ent responses in the absence of a controlling translationalt symmetry cue, a different bias should appear as the delay t became long enough that the translational-symmetry cue was no longer apparent. Hence, the translational-symmetry hypothesis predicts increases in accuracy on different trials t but decreases in accuracy on same trials-in other words, asymmetrical retention functions (e.g., Singer, Klein, & Zentall, 2006;Wixted & Gaitan, 2004).…”

Coding in pigeons: Multiple‐coding versus single‐code/default strategies

“…However, Singer et al's () findings are not conclusive regarding the animal's coding strategy because they can also be explained by the multiple‐coding hypothesis. According to the latter, during training the pigeons learned a specific code for each of the three samples, “If 2 s, choose C2”, “If 8 s, choose C1”, and “If 32 s, choose C2”.…”

Unraveling sources of stimulus control in a temporal discrimination task

“…During training, the pigeons learned to map three samples (2 s, 6 s, and 18 s) onto two comparisons (red and green hues) such that one comparison-for example, Red-was correct following the 2-s samples and the other comparison, Green, was correct following both the 6-s and 18-s samples. The aim was to test whether the pigeons' performance accorded to a single-code/default rule, a single code for the 2-s sample, and a default rule for all other samples (see also Clement & Zentall, 2000;Singer, Klein, & Zentall, 2006). Because the single code would be triggered by the shortest sample, mapped to Red, the default response rule would be triggered whenever the sample was not 2-s long, that is, "not short," and it would involve the choice of Green, the correct choice following the 6-s and 18-s samples.…”