Annual Activity Report 2010

Program Overview

Methodology for the investigation of complex relational learning

1. Categorical-Dimensional Interactions in Symbolic Categories

 

1.A. If equivalence is a model of meaning, a meaningful stimulus should transfer its meaning to abstract stimuli equivalent to it. Three different methods showed transfer of meaning from faces expressing emotions to arbitrary stimuli equivalent to them:

a). Semantic differential: Arbitrary stimuli equivalent to happy or angry faces produce similar semantic differential scores.

b). Semantic priming: Faster recognition of an abstract picture when primed by another abstract picture belonging to the same equivalence class (Bortoloti & de Rose, in press a).

c). Implicit Relational Assessment Protocol (IRAP): This study, using yet another method (IRAP), confirmed that arbitrary stimuli in the same equivalence class are semantically related (Bortoloti & de Rose, in press b).

These studies confirmed that equivalent stimuli may differ in their “degree of relatedness”, as already indicated by results of another research group. Our studies showed additional experimental parameters that affect the degree of relatedness. This is a novel and relevant contribution, which has important implications for the concept of behavioral equivalence and for methodological and practical issues, in the study of symbolic behavior.

1.B (1). Further basic/translational research on reading acquisition: Investigation of recombinative reading generalization with a Miniature Linguistic System alowed the identification and description of critical variables for the expansion of symbolic repertoires by means of recombinative generalization. The development of abstract control by all elements involved in recombination is favored by (a) exposure to a variety of compound stimuli, (b) whose elements are displayed in all possible positions within the learned stimuli, and (c) when training involves superimposition, i.e., the same element is presented as a component of different stimuli (Hanna, Kolsdorff, Quinteiro, Melo, de Souza, de Rose, & McIlvane, 2011). The direct teaching of conditional relations involving the elements to be recombined may accelerate the recombinative repertoire (Hanna, Karino, Araújo, & de Souza, 2010).

1.B (2). The research on textual reading was extended to the acquisition of musical reading: Preliminary results show the acquisition of musical reading in musically illiterate individuals, modeling a system to isolate the acquisition of symbolic relations from past symbolic experiences (Perez & de Rose, 2009; the master’s dissertation of Tauchen; 2010; and several ongoing projects supervised by Elenice Hanna and by Julio de Rose). The procedures developed for these studies constitute important methodological contributions to the field and may be useful for subjects with attentional disorders.

 

2. Selective attention and observing behavior in symbolic tasks

2.A. Subjects must observe the stimuli to learn relations between them. The analyses of results of observing behavior studies could be a useful tool for discovering the reasons for atypically restricted stimulus control and failure to learn relations. Interventions on observing behavior can remediate learning failures and reduce variability of outcome (Dube et al., 2010). The identification of conditions that lead subjects to observe stimuli and learn relations between them resulted in the development of technologies for control of attentional processes and more efficient development of symbolic repertoires.

 

3. Animal Models of Symbolic Behavior and its Prerequisites

During this period the group replicated and extended demonstrations, with non-human subjects, of the properties that define stimulus equivalence and are requisite for symbolic behavior. Improvements in experimental procedures and in teaching approaches are important by-products of these efforts.

3A. Relational Learning in Cebus apella

With Cebus apella the Institute has well-established procedures to generate, consistently and predictably, simple visual discriminations, discrimination reversal, and conditional discriminations by identity, with emergent generalized identity. Generalized identity matching to sample established the concept of sameness, which is an achievement for cebus. The “curriculum” of the Experimental School of Primates is currently qualified to teach IDMTS for all students; this is also an achievement, after years of research in which the main results were variability and a high rate of unsuccess.

The curriculum has also been enriched with discrimination of complex stimuli such as objects, photos, videos and virtual reality environments. An important way station in the EPS program is the development of reliable methods to teach arbitrary matching to sample (Brino, Barros, Galvão, Garotti, Cruz, Santos, Dube, & McIlvane, in press).

A recent study found that Cebus apella exhibits rapid acquisition of complex stimulus relations and emergent performance by exclusion (Brino, Assumpção, Campos, McIlvane, & Galvão, 2010). This performance – a hallmark of language acquisition in humans, is rarely demonstrated in non-human subjects. The results are especially important because the exclusion procedure, superimposed on an identity-matching task, generated arbitrary matching to sample, and can now be used as an alternative to establish this repertoire.

 

 

Relational Learning in Dogs and Bees

3B (1). The laboratories of UFPa and UFSCar developed standardized preparations for studying relational learning with domestic dogs (Canis familiaris) and for identyfing the role of social interactions on the acquisition of pré-symbolic repertories. The first experiments with these subjects demonstraded simple and conditional discrimination learning, emergence of fast mapping (Costa & Domeniconi, in press), and formation of functional classes (Dahás, Brasiliense, Barros, Costa, & Souza, 2010). Ongoing projects aim to extend the generality of the initial results and verify the formation of equivalence classes.

3B (2). Melíponas quadrifasciata learned simple discriminations and showed fast mapping: They chose a novel stimulus when presented with this stimulus and the S- stimulus; but continued to select the S+ when this stimulus was presented simultaneously with the novel stimulus (paper in preparation). In a study conducted in Austrália, one of our doctoral students established, with Apis meliffera, arbitrary relations between line patterns (horizontal/vertical) and colors (yellow/blue), but simmetry tests showed negative results. The procedure is currently been used with the meliponas, in search for generality of the acquisition of aribitrary relations.

The positive results in animal studies of fast mapping with Cebus apella, dogs, and bees suggest an instigating prospect for data integration and document a very important component of relational learning.

 

3C (3). Studies with pigeons were not included in the original project, but two doctoral dissertations contributed with additional data on relational learning, establishing, with these subjects, conditional discriminations that engendered the emergence of the properties that define stimulus equivalence: symmetry (Velasco, Huziwara, Machado, & Tomanari, 2010) and transitivity (paper in preparation).