Association for Behavior Analysis International

The study of complex relational networks within relational frame theory (RFT) has typically received far less attention than the study of individual relational frames. In so far as sophisticated language and cognition involves responding in accordance with complex relational networks, rather than simple frames, it seems important for researchers to increase their focus on the former. Recent conceptual and empirical advances in RFT may help facilitate progress in this regard. The current symposium presents four papers that contribute to this effort. Specifically, the four papers will consider (1) recent developments in the use of a hyper-dimensional multi-level (HDML) framework in applied behavior analyses and the development of complex language abilities; (2) a recent attempt to experimentally model the relating of relational networks; and (3) two recent research programmes examining specific variables critical for relational networking effects produced by the implicit relational assessment procedure (IRAP), one involving stimuli with pre-experimentally established functions and the other involving stimuli that were assumed to be novel before the experimental session. Overall, the current session seeks to highlight some recent ways that advances in RFT have helped facilitate research into complex relational networking and some potentially important variables for the emergence of such behavior.

A basic background in behaviour analysis is assumed

Since its initial proposal, relational and functional contextual control have been critical elements of Relational Frame Theory for explaining human language and cognition. However, while relational contextual control has been extensively investigated, the importance of functional contextual control was arguably fully acknowledged with the emergence of the hyper-dimensional, multi-level (HDML) framework. The HDML explicitly emphasizes the importance of orienting and evoking stimulus properties and suggests new venues for investigation in this regard. Our current focus was analyzing orienting functions as a critical element for the emergence of more elegant and complex relational repertoires. Two children (5y3m with TEA; 3y2m with typical development) were exposed to a gaze-following task, and an eye-tracking device recorded their observational patterns. The task involved observing a human profile directed to one of four objects. The total fixation time and a heat map of the ocular fixation revealed severe differences between the participants. Only the typically developing child successfully demonstrated contextual control initially. These differences were remediated by multiple-exemplar training for functional contextual control. The results observed highlighted elementary patterns of entailed orienting likely critical for the emergence of more complex patterns of relating and suggest possibilities to establish new strategies for building functional contextual control.

This study tested a basic assumption of the differential arbitrarily applicable relational responding effects (DAARRE) model by examining the impact of native versus foreign languages on IRAP performances. The DAARRE model highlights the importance of not only the relational (Crel) properties of stimuli, but also the functional (Cfunc) properties of the stimuli and response options presented in the IRAP trials. Since the native language (NL) is acquired in an emotionally naturalistic and richer context compared to a foreign language (FL), one can expect that a NL word should have stronger orienting and evoking functions than its equivalent (or synonymous) in a FL. Twenty-one Brazilian participants (who speak both Portuguese and English) completed two IRAPs, one in which Portuguese-language stimuli were predominantly used and a second in which English-language stimuli were predominantly used. The IRAP trial-type containing Portuguese-language (i) labels, (ii) targets, and (iii) response options produced the largest effect across all eight IRAP trials -types (four within each IRAP). This result supports the basic assumption of the DAARRE model regarding the overlap of Cfunc and Crel properties within that trial-type for native Portuguese speakers and provides further support for the utility of this conceptual model for interpreting IRAP effects.