Modeling Arbitrarily Applicable Relational Responding with the Non-Axiomatic Reasoning System: A Machine Psychology Approach

TL;DR

The paper tackles modeling Arbitrarily Applicable Relational Responding (AARR) by embedding RFT-inspired relational reasoning within the Non-Axiomatic Reasoning System (NARS). It introduces acquired relations as grounding mechanisms that link sensorimotor contingencies to symbolic relations, enabling mutual, combinatorial entailment and transformation of stimulus function within a unified, uncertainty-aware framework. Two theoretical experiments—Stimulus Equivalence/Function Transfer and Opposition/Function Transformation—demonstrate that NARS can derive untrained relations and context-dependent function transformations, mirroring key human cognitive phenomena. This work suggests a viable path for human-like relational cognition in AI, offering a data-efficient, symbolically grounded alternative to purely statistical models and informing future AGI research and embodied reasoning systems.

Abstract

Arbitrarily Applicable Relational Responding (AARR) is a cornerstone of human language and reasoning, referring to the learned ability to relate symbols in flexible, context-dependent ways. In this paper, we present a novel theoretical approach for modeling AARR within an artificial intelligence framework using the Non-Axiomatic Reasoning System (NARS). NARS is an adaptive reasoning system designed for learning under uncertainty. By integrating principles from Relational Frame Theory - the behavioral psychology account of AARR - with the reasoning mechanisms of NARS, we conceptually demonstrate how key properties of AARR (mutual entailment, combinatorial entailment, and transformation of stimulus functions) can emerge from the inference rules and memory structures of NARS. Two theoretical experiments illustrate this approach: one modeling stimulus equivalence and transfer of function, and another modeling complex relational networks involving opposition frames. In both cases, the system logically demonstrates the derivation of untrained relations and context-sensitive transformations of stimulus significance, mirroring established human cognitive phenomena. These results suggest that AARR - long considered uniquely human - can be conceptually captured by suitably designed AI systems, highlighting the value of integrating behavioral science insights into artificial general intelligence (AGI) research.

Figures (5)

• Figure 1: An example scene where the system perceives three different colors at three different locations.
• Figure 2: Task 1 of this paper. Stimulus equivalence and the transfer of function. The necessary pre-training (Phase 1) is excluded from the picture. Picture shows Phases 2-5 of the task. Underlined options indicate correct choices.
• Figure 3: Task 2 of this paper. AARR in accordance with opposition and the transformation of function. The necessary pre-training (Phase 1) is excluded from the picture. Picture shows Phases 2-5 of the task. Underlined options indicate correct choices.
• Figure 4: The two networks trained as part of the first experiment of this paper. Solid arrows represent relations that are explicitly trained. Dashed arrows represent derived relations.
• Figure 5: The network trained as part of the second experiment of this paper. S and O indicate SAME and OPPOSITE, respectively. Left panel shows relations that are explicitly trained. Right panel shows derived relations.