|
| Empirical Findings and Theoretical Integration |
| Sunday, May 30, 2004 |
| 9:00 AM–10:20 AM |
| Beacon D |
| Area: EAB; Domain: Applied Research |
| Chair: Lanny Fields (Queens College/SUNY) |
| Abstract: . |
| |
| Associative Symmetry in the Pigeon through Arbitrary and Identity Training |
| ANDREA FRANK (University of Iowa), Edward Wasserman (University of Iowa) |
| Abstract: Associative symmetry, performance of a B?A association after training an A?B association, has not seemed possible with nonhuman animals. Using a go/no go successive matching-to-sample procedure, however, we have found that pigeons do show evidence of associative symmetry. Using a single-key go/no gho procedure, our new procedure avoided the spatial location confoundings that have plagued many associative symmetry experiments. We also avoided an equally severe temporal confounding by intermixing the identity matching trials along with the arbitrary matching trials. Interestingly, a second group of pigeons, trained only with the arbitrary matching trials, did not show learning of the symmetrical relation. In addition, a third group of pigeons did not show the symmetrical relation when first trained only with arbitrary matching trials, but did so when identity training trials were later added. These results suggest that, when the discriminative stimuli are shown on a single key and in all temporal locations, pigeons can exhibit symmetrical relations among stimuli. |
| |
| Effects of Training and Testing on the Formation of Linked Perceptual Classes |
| LANNY FIELDS (Queens College/SUNY) |
| Abstract: A linked perceptual class consists of two distinct perceptual classes, the members of which occasion the mutual selection of each other. This occurs after the direct training of at least one conditional discrimination between one member of each class. The documentation of class formation involves the presentation of a wide variety of probes drawn from each of the distinct classes. Responding in a class consistent manner demonstrates the emergence of a linked perceptual class. We will summarize how the formation of linked perceptual classes is influenced by training variables such as the number of conditional discriminations and the particular class members used in the conditional discriminations. In addition, we will consider how the number/variety of probes included in a test block and the programmaticity of probe presentations influences the formation of linked perceptual classes. Finally, we will consider how linked perceptual classes are related to the complex categories that control behavior in natural settings. |
| |
| Stimulus Classes, Relational Stimuli and the Transformation of Stimulus Functions |
| MICHAEL J. DOUGHER (University of New Mexico) |
| Abstract: One of the defining characteristics of stimulus classes or categories is the emergent transformation of the functions of the remaining members of a class when the functions of a subset of the class are modified. This emergent or untrained transformation of functions is conceptually significant because it extends the scope of behavior analysis to novel and complex behaviors. Although an interesting and promising area of research, the relation between stimulus categories and function transformation needs further conceptual as well as empirical attention. One potential problem is that function transformation is often explained by appeal to stimulus classes. The circularity of using the concept of stimulus class to explain a phenomenon that partly defines a stimulus class is problematic and avoids the task of identifying the determinants of the phenomenon. A related problem with explaining function transformation by appeal to stimulus classes is that function transformation among stimuli can be demonstrated in the absence of stimulus classes. The present paper will report a series of experiments that demonstrate function transformation that appears to be based solely on the arbitrary relational properties of stimuli. The paper ends with a conceptual discussion that attempts to integrate these findings with the concept of stimulus classes. |
| |
| Contingencies and Concepts: Interpretation and Analysis of Natural Categories |
| MARK GALIZIO (University of North Carolina, Wilmington) |
| Abstract: Skinner’s Verbal Behavior was primarily an exercise in interpretation. In contrast, the stimulus equivalence approach has generated numerous experimental studies, but relatively little interpretation. To bridge this gap, several experiments will be discussed that use stimulus equivalence methodologies to model effects noted in natural categories such as induction, typicality effects, hierarchical structure and “prior knowledge” effects. Interpretative analysis of natural categories will explore the degree to which contingency-based equivalence classes can account for these and other properties observed by developmental and cognitive psychologists. |
|
| |