The big idea
B.21 introduces relational frame theory (RFT) and names its three properties, and G.19 covers equivalence-based instruction and multiple exemplar instruction for naming. That is where the exam stops. This guide continues: how relating is learned and cued, the full range of relations, and what applied work has done with them.
RFT’s central claim (Hayes, Barnes-Holmes, & Roche, 2001) is that relating is itself operant behavior. Children learn, through many reinforced examples, to respond to one event in terms of another according to cues such as “same,” “opposite,” “more than” and “before.” Once that repertoire exists, one sentence can change how a person responds to things they have never encountered. On this account, that is what makes language generative, and also why fear and self-criticism spread so easily.
Breaking it down
Nonarbitrary and arbitrary relating
Many animals learn nonarbitrary relational responding: picking the physically larger of two objects, whatever the objects are.
Arbitrarily applicable relational responding (AARR) is relating controlled by contextual cues rather than physical properties. A young child offered a nickel or a dime picks the nickel, because it is bigger. Later the child learns that the dime is “worth more” and picks it, though nothing about the coins has changed. The cue can now be applied to anything, which is what arbitrarily applicable means. A relational frame is one particular pattern of AARR, such as comparison.
The three properties, more precisely
- Mutual entailment: a relation from A to B lets the learner derive one from B to A. The derived relation need not match. If A is more than B, B is less than A. Equivalence’s symmetry is the special case in which both directions are “same.”
- Combinatorial entailment: two or more relations combine to give new ones. If A is more than B and B is more than C, then A is more than C and C is less than A. Some combinations stay vague: if A differs from B and B differs from C, nothing definite follows about A and C.
- Transformation of stimulus function: the functions of one stimulus change the functions of related stimuli, in line with the relation. If a token buys a snack and a new coupon is “worth more than” the token, the coupon may become more reinforcing without ever being spent.
Contextual control: Crel and Cfunc
A Crel (relational context) specifies the relation: “same as,” “opposite,” “bigger than,” “after.” A Cfunc (functional context) specifies which function transfers. In “This fruit tastes like a lemon,” like is the Crel and tastes is the Cfunc. The fruit may come to evoke puckering, but not reaching for something yellow.
Families of frames
| Frame | Typical cues | Example of a derived relation |
|---|---|---|
| Coordination | same, is, like | “Perro” means dog, so “dog” means “perro” |
| Distinction | different from | A differs from B, but how is unknown |
| Opposition | opposite | The opposite of an opposite is the same |
| Comparison | more, less, faster | The dime is worth more than the nickel |
| Hierarchy | kind of, part of | A tabby is a cat and a cat is an animal, so a tabby is an animal |
| Temporal | before, after | Lunch is after recess, so recess is before lunch |
| Conditional and causal | if-then, because | If the bell rings, recess is over, so the bell signals the end of play |
| Deictic | I-you, here-there, now-then | If I were you, I would see what you see |
Spatial frames (above, near) work the same way.
How frames are learned
RFT treats each frame as a generalized operant shaped by multiple exemplar training: many different stimulus sets, related through the same cue, with feedback, until the pattern of relating follows the cue rather than any particular stimuli. Berens and Hayes (2007) showed this with preschoolers who could not solve arbitrary “more than” and “less than” problems. Reinforced multiple-exemplar training established the skill, and it generalized to new stimuli and new kinds of trial.
Deictic frames and perspective taking
Deictic frames relate events to the speaker’s point of view: I and you, here and there, now and then. Because their meaning shifts with who is speaking, they cannot be learned from physical features. Perspective-taking protocols (McHugh, Barnes-Holmes, & Barnes-Holmes, 2004) test simple relations (“I have a red brick and you have a green brick. Which brick do I have?”), reversed relations (“If I were you and you were me…”) and double reversals. Training them offers a behavioral route to what other fields call theory of mind.
Applied work
- PEAK. The PEAK Relational Training System (Dixon, 2014, and later modules) moves from directly trained skills to equivalence, generalization and transformation of function. Studies report gains for learners with autism, for example deriving “I” relations after training “you” relations (Belisle et al., 2016). Much of the research comes from the developers’ groups; Witts (2018) criticized its methods and Belisle and Dixon (2020) replied. Treat it as promising, with independent replication still limited.
- ACT. Acceptance and commitment therapy (Hayes et al., 2006) builds on RFT’s account of how verbal relations let painful thoughts dominate behavior. Its goal is psychological flexibility: staying in contact with the present and acting on chosen values even when hard thoughts and feelings are present. Behavior analysts apply ACT-based procedures with caregivers, staff and organizations, within their competence (Standard 1.05, Practicing within Scope of Competence).
Rival accounts
Three accounts of emergent relations are still debated. Sidman (2000) argued that equivalence is a basic outcome of the reinforcement contingency, not a separately learned skill. The naming account (Horne & Lowe, 1996; Miguel, 2016) holds that bidirectional naming, speaker and listener behavior joined, makes class formation possible. RFT counts equivalence as one learned frame among many. Palmer (2004) argued that RFT has not stated its core principle clearly enough to test, and that covert verbal behavior might explain the data. Gross and Fox (2009) summarize the controversy.
Seeing it in practice
| Setting | What happens | RFT reading |
|---|---|---|
| Clinic | After “worth more” practice with many sets, Ibrahim, 7, trades two stickers for a token he is told is worth more than five stickers. | Comparison frame and transformation of function |
| School | Grace, 9, answers “If I were you, what would I see?” correctly with new pictures. | Generalized deictic reversal |
| Parent coaching | A father who thinks “I’m a failure as a dad” avoids the plan; ACT exercises help him notice the thought and practice anyway. | Loosening a verbal network’s control over behavior |
Common mix-ups
- AARR is not stimulus generalization. Generalization follows physical similarity. AARR follows contextual cues.
- Mutual entailment is not always symmetry. In comparison, the derived relation is the reverse one, less than.
- Correct pronoun use is not proof of deictic framing. The test is deriving relations from another person’s perspective, including reversals.
- RFT is not ACT, and PEAK is not RFT. ACT is a therapy and PEAK a curriculum, both built on the theory.
Exam tip: On the exam, RFT appears as mutual entailment, combinatorial entailment and transformation of stimulus function (B.21). More than, opposite or before point to RFT; symmetry and transitivity point to equivalence.
Quick recap
- RFT treats relating as a learned, generalized operant controlled by contextual cues.
- Crel cues set the relation; Cfunc cues set which function transfers.
- Frames include coordination, distinction, opposition, comparison, hierarchy, temporal, spatial, causal and deictic.
- Multiple exemplar training builds frames; deictic training targets perspective taking.
- PEAK and ACT apply RFT, while Sidman’s account and the naming account remain rivals.
Further reading
- Hayes, S. C., Barnes-Holmes, D., & Roche, B. (Eds.). (2001). Relational frame theory: A post-Skinnerian account of human language and cognition. Kluwer Academic/Plenum Publishers.
- Berens, N. M., & Hayes, S. C. (2007). Arbitrarily applicable comparative relations: Experimental evidence for a relational operant. Journal of Applied Behavior Analysis, 40(1), 45-71.
- Gross, A. C., & Fox, E. J. (2009). Relational frame theory: An overview of the controversy. The Analysis of Verbal Behavior, 25(1), 87-98.
Key terms
- Relational frame theory
- A behavior-analytic account of language and cognition that treats relating stimuli to one another as a learned, generalized operant. Its core properties are mutual entailment, combinatorial entailment and transformation of stimulus function, and it covers relations such as sameness, opposition, comparison and hierarchy, not only equivalence.
- Arbitrarily applicable relational responding
- Responding to one stimulus in terms of another according to contextual cues, such as 'same,' 'more than' or 'opposite,' rather than physical properties. Relational frame theory treats it as a learned, generalized operant, and each pattern of it is called a relational frame.
- Mutual entailment
- In relational frame theory, the derivation of a relation from B to A after a relation from A to B has been learned. The derived relation need not be identical: if A is more than B, then B is less than A. Symmetry in stimulus equivalence is the special case in which both relations are sameness.
- Combinatorial entailment
- In relational frame theory, the derivation of new relations from two or more learned relations that share a stimulus. If A is more than B and B is more than C, then A is more than C and C is less than A. Transitivity in stimulus equivalence is the special case in which every relation is sameness.
- Transformation of stimulus function
- A change in the functions of one stimulus because it has been related to another stimulus that has those functions, in a way that follows the relation. A stimulus related as 'more than' a reinforcer may become more reinforcing without any direct contact.
- Deictic frame
- A relational frame that relates events to the speaker's point of view: I and you, here and there, now and then. Because the meaning of these words shifts with who is speaking, they cannot be learned from physical features, and relational frame theory treats them as the basis of perspective taking.
- Multiple exemplar training
- Teaching with several varied examples of the stimuli, settings or response forms involved in a skill, or rotating one stimulus across different response types, so that responding extends to examples that were never taught.
- Psychological flexibility
- The goal of acceptance and commitment therapy: staying in contact with the present moment and continuing or changing behavior in line with chosen values, even when difficult thoughts and feelings are present. Its account of how thoughts come to dominate behavior draws on relational frame theory.
Check yourself
4 questions on J.2. Choose an answer to see whether you're right and why.
Show the answer
Answer: B. Two trained relations were combined to derive a third, which is combinatorial entailment, and in opposition two opposites combine into sameness. Mutual entailment would be deriving that good is the opposite of zorp from one relation. Transitivity describes equivalence classes, where every relation is sameness, so it cannot produce 'opposite of an opposite.'
Show the answer
Answer: D. A Cfunc specifies which function of one stimulus transfers to the related stimulus, here the odor rather than the color or taste. 'Like' is the Crel: it specifies the relation of coordination. The drink and the vinegar are the related stimuli, not the contextual cues.
Show the answer
Answer: A. RFT treats relational framing as a generalized operant built by multiple exemplar training: many different stimulus sets, related through a cue such as 'worth more,' with feedback. Choosing the larger coin strengthens nonarbitrary relating, which is the problem here. Pairing might change preference but does not teach a relation that transfers to new items.
Show the answer
Answer: D. I and you are deictic relations, defined by the speaker's perspective rather than by physical features, and 'If I were you' asks the learner to reverse them. A tact of location would not require taking another person's point of view. Nothing in the task places the chair and floor in a class-and-member relation, so it is not hierarchy.