Table of Contents
The Core Definition of Relational Frame Theory
Relational Frame Theory, often abbreviated as Relational Frame Theory (RFT), is a comprehensive psychological theory of human language and cognition that emerged from the behavior analytic tradition. At its heart, RFT posits that the unique complexity of human language and high-level cognitive functions—such as reasoning, problem-solving, and metaphor—are attributable to a learned, generalized behavioral mechanism known as derived relational responding. This capability allows humans to arbitrarily relate stimuli to one another based on social and environmental context, rather than solely on their inherent physical properties. The theory is philosophically grounded in functional contextualism, which emphasizes predicting and influencing psychological events (thoughts, feelings, and behaviors) by manipulating variables within their context, offering a naturalistic, non-mentalistic account of complex human behavior.
The core principle of RFT dictates that once an individual is trained to relate two stimuli in a specific way (e.g., A is “bigger than” B), they will automatically and spontaneously derive new, untrained relations (e.g., B is “smaller than” A), provided they have established the necessary generalized relational frame through extensive prior learning. This differs fundamentally from traditional stimulus-response learning, where every single relation must be explicitly taught and reinforced. RFT essentially describes how verbal organisms learn to create vast, interconnected networks of symbolic relations, where the psychological functions of one element in the network can be instantly transformed or transferred to all other related elements. This capacity for arbitrary symbolic relating is viewed as the fundamental behavioral unit of language.
Historical Development and Behavioral Roots
The development of RFT was largely spearheaded by researchers Steven C. Hayes and Dermot Barnes-Holmes, beginning primarily in the 1980s and 1990s, aiming to address perceived gaps in the behavioral analysis of language. The theory serves as an experimental extension and contextualistic interpretation of B.F. Skinner’s radical behaviorism, specifically attempting to provide a more empirically progressive account than the one presented in Skinner’s influential 1957 work, Verbal Behavior. While Skinner’s approach, which categorized verbal behavior into functional units like mands and tacts, proved useful in limited applications such as language training for individuals with developmental disabilities, it failed to generate a robust experimental research program capable of fully explaining the generativity and complexity of human language, reasoning, and logic.
The shift in mainstream psychology away from behaviorism toward cognitivism in the mid-20th century was significantly accelerated by linguist Noam Chomsky’s famous critique of Skinner’s book. Chomsky argued that the sheer speed and creativity of language acquisition demonstrated that language could not be merely learned, necessitating the existence of an innate “language acquisition device.” RFT advocates accepted the challenge posed by Chomsky’s criticism—that behaviorism needed a more powerful explanatory model for complex cognition—and responded by developing the concept of derived relational responding as a generalized operant. This development sought to demonstrate that complex language could indeed be accounted for through environmental interaction and learning, provided the focus was shifted from specific stimulus-response connections to the generalized ability to relate stimuli arbitrarily, thereby bridging the explanatory gap between behaviorism and the observed complexity of human thought.
Derived Relational Responding and Operant Learning
The central mechanism described by RFT is derived relational responding (DRR), which is conceptualized as a generalized operant: a behavioral response class that is reinforced across multiple examples, leading to a pervasive, flexible skill. This generalized operant allows humans to respond to stimuli based on arbitrary relations established by the community (e.g., symbolic relations like “synonymy,” “opposition,” “comparison,” or “hierarchy”), rather than solely on direct experience or physical similarity. The establishment of this operant requires extensive training involving multiple examples, ensuring that the individual responds to the relational context itself, not just the specific items being related.
DRR is characterized by three key properties that distinguish it from simple associative learning. First is mutual entailment, where learning A is related to B implies B is related to A (e.g., if “bigger than” is trained, “smaller than” is derived). Second is combinatorial entailment, where linking A to B and B to C allows the derivation of the relationship between A and C, and C and A (e.g., transitivity). Third, and most critical for explaining cognition, is the transfer of function, where the psychological function (e.g., emotional impact, behavioral consequence) of one stimulus in the network automatically transfers to all other stimuli related to it. These properties allow humans to build vast, intricate networks of symbolic meaning, enabling rapid learning and complex thought processes without requiring direct, exhaustive interaction with every single stimulus or consequence.
RFT in Practice: A Conceptual Example
To illustrate derived relational responding and the transfer of function, consider the process of a child learning the concept of financial value, which is an entirely arbitrary, socially constructed relation. The child does not need to physically experience the consequences of every denomination of money to understand its purchasing power; instead, they learn the relational frame of “more than” and “less than.”
The learning process typically begins with direct training: the child is explicitly taught that a $10 bill is “more than” a $1 bill, and that the $10 bill can purchase a desired toy (the function of “buying power” is attached to the $10 bill). Through pre-established relational frames, the child immediately demonstrates mutual entailment by recognizing that the $1 bill is “less than” the $10 bill, a relation never directly taught. Subsequently, if the child is introduced to a $50 bill and taught only that it is “more than” the $10 bill, they will spontaneously derive, via combinatorial entailment, that the $50 bill is significantly “more than” the $1 bill.
The crucial step for RFT is the transfer of function. Because the $50 bill is now related to the $10 bill in the “more than” frame, the child will immediately assume that the $50 bill possesses the function of “buying power,” and indeed, a greater buying power than the $10 bill, without ever having to test the $50 bill at a store. If the word “debt” is introduced and taught to be “bad” (a negative function) and is related to “spending more than you have,” this negative function will transfer to the entire relational network associated with overspending, influencing future financial behavior based purely on symbolic relations rather than immediate, physical consequences.
Empirical Evidence and Scientific Status
RFT is supported by a significant body of empirical studies that have tested its core tenets concerning the nature of derived relational responding. Research has successfully demonstrated that DRR meets the criteria of an operant class, showing that its establishment depends on the history of reinforcement, the context in which relating occurs, and the use of multiple examples during training to promote generalization. These studies regularly appear in mainstream behavioral journals, indicating a vigorous experimental program.
A particularly important area of empirical support involves RFT’s capacity to account for the interaction between symbolic processes and other forms of learning. Data has shown that derived relational responding can profoundly alter other behavioral processes, such as classical conditioning. For example, if a neutral stimulus is related symbolically to a stimulus that elicits a fear response, the neutral stimulus will often elicit fear, even without direct pairing. RFT proponents point to these findings as evidence that symbolic, relational networks are capable of modifying existing behavioristic interpretations of complex human behavior, validating the theory’s claim that arbitrarily applicable relational responding is a fundamental, transformative learning process.
Therapeutic and Assessment Applications
The practical significance of RFT extends directly into clinical psychology, serving as the theoretical foundation for Acceptance and Commitment Therapy (ACT). RFT provides a functional analysis of language-based suffering, suggesting that many forms of psychological distress stem from “cognitive fusion”—the excessive and rigid application of relational frames (e.g., relating oneself to the arbitrary label “failure,” and experiencing the resulting negative functions). ACT utilizes RFT principles to guide patients in developing psychological flexibility, helping them relate to their thoughts and feelings in a new, non-fused manner, thereby transforming the negative functions associated with their internal verbal behavior.
Beyond psychotherapy, RFT has guided the development of specialized interventions, including programs designed to enhance cognitive and language development in young children with autism and related disorders. Furthermore, RFT has provided the theoretical basis for a specific implicit measure known as the Implicit Relational Assessment Procedure (IRAP). Developed by Dermot Barnes-Holmes, the IRAP is designed to measure the strength and malleability of specific, arbitrary relations between stimuli (e.g., self vs. other; good vs. bad), offering a more granular assessment of implicit symbolic networks than general association tests, and providing valuable data for research into attitudes, biases, and cognitive processes.
RFT’s Place in Cognitive and Linguistic Theories
RFT belongs to the broader subfield of cognitive psychology, specifically functioning as a behavioral theory of language and cognition, placing it in dialogue with competing linguistic models. It contrasts markedly with generative grammar, the nativist approach championed by Noam Chomsky, which argues that language acquisition relies on innate, specialized biological constraints (nativism) to handle the complexity of language input. The generative view posits a Universal Grammar that dictates the range of possible linguistic structures.
RFT, conversely, aligns more closely with emergentist theories of language acquisition, arguing that complex language emerges from the interaction of generalized cognitive learning mechanisms and environmental input. While emergentist theories often focus on statistical learning or chunking, RFT provides a precise functional and behavioral mechanism—the generalized operant of derived relational responding—to explain *how* arbitrary symbolic relations are learned and maintained. By focusing on the functional context and the history of reinforcement rather than innate mental structures, RFT seeks to provide a unified, naturalistic account of the cognitive complexity that has historically driven a wedge between behavior analysis and the broader study of human thought.