Association for Behavior Analysis International

RFT can be traced back to a 1984 conference presentation on rule-governed behavior. The study of rule-governed behavior in RFT has tended to be overshadowed by the study of relational frames. However, adequate analyses of rule-governed behavior and related complex phenomena, such as prejudice and political polarisation, arguably require the study of increasingly complex relational networks. A more intense focus in this regard has begun to emerge in recent years. This work has been underpinned, to some extent, by the emergence of a hyper-dimensional, multi-level (HDML) framework for conceptualising research in RFT generally. Specifically, the HDML has encouraged analyses of increasingly complex relational networks along four dimensions, including coherence and complexity. The four papers in this symposium will consider some of this work. Specifically, the four papers will consider (1) the impact of relational coherence vs incoherence in establishing rule-following for speakers from different experimentally established groups; (2) attempts to explore the impact of relative degrees of relational coherence on rule-following and speaker preferences; (3) recent experimental analyses of prejudice as involving transformations of functions through complex hierarchical relational networks; (4) how recently offered concepts within RFT may help advance conceptual analyses of complex social behaviors relevant to political polarisation.

A basic background in behaviour analysis is assumed

Previous experiments have revealed that different levels of relational coherence can affect rule-following behaviors. In addition, relational coherence may be under the control of conditional discriminations. For example, different speakers may generate different rule-following responses in the same individual with the same rule by manipulating the levels of relational coherence associated with the speaker. This presentation aims to address this research topic by presenting two experiments. Experiment 1 attempts to replicate previous experimental results exploring the impact of speakers established as relatively coherent (100%, 50%, 0% coherent) on participant rule-following. Experiment 2 is designed to perform a parametric analysis of different levels of relational coherence (90%, 80%, 70%, and 60%) in three different speakers and assess how participants’ speaker preference is affected. Overall, the initial results have been replicated, and the parametrical analyses involved in Experiment 2 have so far indicated that variability in rule-following increases as relational coherence decreases. The implications of these findings for developing a more detailed perspective of the variables that affect rule-following are considered, as are some next steps for the analysis of this complex human behavior.

Prejudice often leads to regrettable social effects, such as segregation and mistreatment. Derived verbal events underlie this phenomenon, regulating our interactions in the social environment and establishing functions to stimuli without a history of direct learning. A notable gap exists in the experimental study of more complex derived stimulus relations investigating racial bias. The present study investigated if hierarchical relations could affect biased responses. Twelve undergraduate students were trained to respond according to two contextual cues: “includes” and “belongs to.” These cues were used to form three hierarchical levels (top, middle, and bottom). Then, negative or positive functions were established, pairing adjectives to one of the middle-level category labels. The effects of the functions directly paired were inquired for all arbitrarily derived hierarchical levels, characters, and extra non-arbitrary related stimuli, checking for generalization of the responses. In the first experiment, five participants responded according to hierarchical relations, showing transitive and asymmetrical class containment and unilateral property induction. In a second experiment, with a broader testing phase, all participants responded according to hierarchical relations. Additionally, further undesirable effects of prejudice were observed on generalization based on physical traits or color.

There is a growing concern about political polarization in current societies. De Rose and Rocha (in press) suggested that RFT may provide useful elements to analyze the phenomenon of polarization. Political positions may be related in accordance with different patterns of relational responding, such as difference, hierarchy, comparison, or opposition. When positions are related in frames of opposition, transformation of functions result in highly polarized affective functions of the opposing opinions and attitudes, so that if one has positive valence the other will necessarily have negative valence. Recent developments in RFT may contribute to advancing this analysis, taking into account the orienting, evoking, and motivating functions involved in any given act of relating. As such, negative evoking functions may impair the mere contact with the arguments of opposed positions, and thus contribute to confirmation bias and motivated reasoning, to the point that political discussions and search for consensus becomes impossible. This presentation will elaborate on how current RFT developments may lead to refinements in the analysis of political polarization based on frames of opposition as derived relational networks and their differential functional properties.