Relacionar relaciones como modelo analítico‑funcional de la analogía y la metáfora

Full reference: Ruiz, F.J. & Luciano, C. (2012). Relating relations as a functional-analytic model of analogy and metaphor. ACTA COMPORTAMENTALIA, 20(Monographic), 5-31.

Study type: Theoretical-experimental research on analogy and metaphor based on Relational Frame Theory

Background and Objectives

Analogy and metaphor have been principal research topics in cognitive psychology for decades. Researchers have analyzed animal and human behavior in analogue tasks, seeking the biological basis of how organisms understand analogies and create new conceptual connections. However, there is considerable disagreement about what types of interactions produce these cognitive capacities.

The objective of the present work is to present a functional-analytic model of analogy and metaphor developed over the past fifteen years, based on the behavior of relating relations. The authors seek to demonstrate that this model, grounded in Relational Frame Theory (RFT), provides better understanding of analogy and metaphor than previous cognitive models. The work reviews accumulated experimental evidence showing that participants can derive analogies without direct learning, supporting RFT model predictions.

Method

The work consists of extensive literature review on analysis of analogy and metaphor from a functional-analytic perspective based on RFT, integrated with presentation of experimental findings supporting the proposed model.

Participants

Multiple experiments conducted over the past fifteen years by various researchers are described. No single N is provided, as the work reviews results from previous studies. The authors specifically mention:

• Experiment 1 by Barnes, Hegarty and Smeets (1997): Adult participants
• Experiment 2 by Ruiz and Luciano (2011): [number of participants verify in original article]
• Experiment 3 by Stewart, Barnes-Holmes, Roche and Smeets (2002): [number of participants verify in original article]
• Experiment 4 by Carpenter, Smeets, Barnes-Holmes and Stewart (2004): [number of participants verify in original article]
• Additional studies by Pérez and García (2009, 2011a, 2011b) and others

Experimental Design

The work describes several experimental paradigms:

Paradigm 1 - Equivalence-Equivalence: Barnes, Hegarty and Smeets (1997) used a procedure training conditional discriminations with novel stimuli (A1-B1-C1, A2-B2-C2, A3-B3-C3), followed by equivalence test. Stimuli from the two classes were then related to distinct geometric figures of color. The comparison procedure evaluated whether participants could derive analogies.

Paradigm 2 - Analogy from Non-Arbitrary Properties: Stewart, Barnes-Holmes, Roche and Smeets (2002) demonstrated a procedure based on the Relational Evaluation Procedure (REP) in which an infinite number of analogies could be generated without need for training on all relations. Authors showed that with specific training in conditional discriminations with common non-arbitrary properties, adult participants could relate relations of different types.

Paradigm 3 - Quality of Analogies: Carpenter et al. (2002, 2003, 2004) analyzed quality of analogies and conditions facilitating their derivation. They used procedures allowing participants to generate analogies of different types, analyzing what interactions were responsible for their development.

Paradigm 4 - Analogy in Children: Various studies analyzed evolution of analogical reasoning from early ages. Authors mention Carpenter et al. research with 5-9 year old children, showing progressive acquisition of analogical competencies.

Paradigm 5 - Transformation of Functions in Analogy: Ruiz and Luciano (under revision) analyzed how analogies can function as contextual operators altering stimulus disposition to elicit emotional responses, demonstrating transformation of functions through analogical relations.

Materials and Apparatus

Various stimuli were used depending on paradigm:

• Abstract stimuli: novel letters (A, B, C, etc.)
• Geometric figures: triangles, circles, squares in different colors (red, green, blue)
• Computer presentation with matching-to-sample (MTS) procedures
• Feedback: positive or negative according to correctness
• Visual assessment procedures (thermometer) in some studies

Procedure

The general procedure followed these steps:

Phase 1: Training in conditional discriminations with stimulus relation establishment (A-B, A-C, etc.). Participants had to respond correctly across a series of trials achieving criterion (generally 80-90% correct responses).

Phase 2: Test of emergent or derived relations. Evaluation of whether participants could respond to new relations (e.g., B-C) without direct training, based solely on previously trained relations.

Phase 3: In some studies, evaluation of common properties between stimuli. Participants had to identify what physical or abstract properties stimuli from different relations shared.

Phase 4: In function transformation studies, evaluation of whether emotional or functional properties of one stimulus could transform through analogical connections. For example, if stimulus A was aversive and A related analogically to B, B could acquire aversive properties without direct exposure.

Analysis

Data were typically presented as:

• Percentages of participants correctly deriving analogies (typically 70-90% in representative studies)
• Reaction times compared between trials (similar-similar vs. distinction-distinction)
• Accuracy analysis of analogies proposed by participants

Results

The literature review results consistently show that:

1. Within-domain analogy derivation: Adult participants with typical development are capable of deriving analogies relating relations of same type (coordination, distinction, etc.) without direct training need. The percentages of participants showing this capacity ranged approximately 70-90% on the first analogy test, consistently selecting response options implying relating relational networks with common physical property.

2. Between-domain analogy derivation: In later studies (Barnes et al. 2005; Carpenter et al. 2002, 2003), participants demonstrated ability to extend this capacity to more complex analogies where relation between the two stimulus series was not based simply on topography but on more abstract relational functions. In the example (Gil, Luciano, Ruiz and Valdivia-Salas 2012), participants identified analogies with common physical properties (color, shape) more precisely and quickly than analogies requiring abstraction of functional properties.

3. Dependence on common formal properties: Stewart et al. (2001) and subsequent findings showed that 90% of participants consistently selected response options implying relating relational networks with common physical property. In the present example, the correct response involved choosing N1O1 since the relational network they belonged to contained combinatorial similarity relation between them. More detailed analysis showed participants identified analogies with common physical properties between networks most of the time (90% of participants in primary example), while when analogy required changing selection criteria for analogy subjects, test response followed evolutionary pattern similar to that found with acquisition of classical analogical reasoning tests.

4. Generalizability across limited trainings: Ruiz and Luciano (2011) found that training participants in analogy test format where only three conditional discriminations with equivalence class formation was conducted could be sufficient to facilitate abstraction of new relation class. Participants subsequently demonstrated competence deriving this new class in presence of completely novel stimuli, demonstrating that generative character of relational learning allows repertoire of relations applicable to new stimuli to emerge through training in multiple examples.

5. Function transformation through derived relations: Important finding showed that 90% of participants, after training in arbitrary relations between stimuli that had developed histories of contrasted interaction, could consistently respond to subsequent trials implying relating relations with specific topography yet sharing discriminative functional properties.

Discussion and Conclusions

The analysis presented demonstrates that a model based on RFT emphasizing behavior of relating relations provides more parsimonious and precisely predictive explanation of analogy than previous cognitive approaches. Particularly, demonstrated capacity of participants to derive analogies without direct training in all implicated relations, based solely on application of previously established relational frame, suggests that complex underlying cognitive structures need not be assumed for analogy reasoning.

The authors underscore that RFT research has opened new dimensions for analogy investigation. The distinction between common formal and common functional properties, capacity to derive new relations through function transformation, and possibility of abstracting properties in everyday life have been aspects previously without adequate exploration.

Limitations: The authors acknowledge that despite promising findings, research on analogy from RFT perspective is still in early stages. There is need for more rigorous tests demonstrating whether derived analogies function as contextual operators in function transformation and whether this can be documented under controlled research conditions before age 12, which has been limitation in previous studies.

Significance and Contribution

This work contributes significantly to understanding fundamental processes of analogy and metaphor by providing a coherent theoretical framework (RFT) explaining how humans generate and understand analogies without assuming complex cognitive operations or underlying structures. It integrates experimental findings demonstrating the generative capacity of analogical understanding: organisms can extract generalities through training in multiple examples and apply them to completely novel contexts. It opens possibilities for understanding how metaphor functions in therapeutic change and cultural knowledge transmission, with relevance for both basic cognitive process researchers and clinicians and educators interested in facilitating conceptual learning and cognitive flexibility.