By Matt Harrington, BCBA · Behaviorist Book Club · April 2026 · 12 min read

Relational Frame Theory in Practice: Teaching Deictic and Temporal Relations to Children with Autism

Overview & Clinical Significance

The capacity to understand relations like 'then-later' and 'here-there' may appear deceptively simple, but for children with autism who have not acquired these deictic and temporal frames, they represent a significant barrier to participation in everyday social and instructional contexts. Conversations depend on shared reference points in time and space. 'We will do that later,' 'Put it over there,' and 'Remember what happened then?' are all relational statements that require the listener to track shifting perspectives and referents. Children who cannot reliably respond to or use these relations face compounding difficulties in academic settings, social exchanges, and following multi-step instructions.

Barron and colleagues (2019) demonstrated the utility of relational training procedures derived from Relational Frame Theory for teaching these relations to children with autism. Using single-reversal procedures and testing for transformation of stimulus function, they showed that training on a limited set of exemplars produced generalization to novel stimuli — a finding with direct implications for how practitioners design intervention targets in this domain.

For BCBAs working with children who have language and cognition goals, this course addresses a gap that traditional verbal behavior assessment tools may not fully capture. Mand, tact, and intraverbal repertoires do not map cleanly onto deictic and temporal relational responding, which requires flexibility across shifting frame contexts. Understanding the RFT framework behind this training approach helps practitioners identify when a child is struggling with relational rather than topographic aspects of a language target.

Background & Context

Relational Frame Theory, developed by Hayes and colleagues, proposes that human language and cognition are fundamentally relational in nature — that the capacity to respond to stimuli in terms of their relations to other stimuli, rather than just their physical properties, underlies the full range of human symbolic behavior. Derived relational responding is the process by which novel stimulus relations are established through history with other relations rather than through direct training on every possible exemplar.

Deictic frames — relations like I/you, here/there, and now/then — are particularly important because they require the speaker or listener to adopt a perspective that shifts with context. Unlike spatial relations defined by absolute dimensions, 'here' and 'there' are defined relative to the speaker's current position. 'Then' and 'later' similarly require tracking the speaker's temporal reference point. Children with autism often have particular difficulty with perspective-taking, and RFT research suggests that difficulty with deictic frames may be a functional contributor to that difficulty rather than just a correlated deficit.

The single-reversal procedure used by Barron and colleagues tests whether trained relations can be transformed when the frame reverses — that is, whether a child who has learned to respond correctly to 'here' under training conditions can also respond correctly when the spatial reference shifts. Transformation of stimulus function refers to the process by which the behavioral function of one stimulus is altered through its relational connection to another stimulus that already has a trained function. Both concepts are foundational to understanding why this training approach produces generalization efficiently.

Clinical Implications

The primary clinical implication concerns target selection. When practitioners assess language and communication for children with autism, temporal and deictic relational responding should be included as specific evaluation targets rather than assumed to develop automatically once basic mand and tact repertoires are in place. A child may have extensive tact and intraverbal repertoires while still failing to reliably use 'later' and 'then' as temporal reference markers in conversation.

Once a deficit in relational responding is identified, the intervention approach demonstrated by Barron and colleagues provides a structured training procedure. The use of single reversals allows practitioners to test flexibility within trained frames efficiently, rather than requiring training on every possible exemplar. This has practical implications for programming efficiency: rather than creating dozens of exemplar sets for every deictic relation, practitioners can design training procedures that test for and build transformation across a smaller set of trained instances.

Code 2.09 of the BACB Ethics Code requires behavior analysts to consider outcomes that are meaningful to the client and their family. Teaching temporal and deictic relational responding addresses outcomes that are directly meaningful in everyday contexts — following directions, participating in family conversations, understanding classroom instructions, and narrating personal experience. Positioning these targets explicitly in terms of their social validity strengthens the ethical and practical case for prioritizing them within a comprehensive treatment program.

Ethical Considerations

Code 2.01 requires behavior analysts to provide effective treatment based on a scientific foundation. The RFT-based training approach described in this course meets that standard — it is theoretically grounded, supported by single-case experimental data, and produces generalization through a principled mechanism rather than rote training on unlimited exemplars. Practitioners who implement this approach are acting consistently with the scientific knowledge requirement.

The ethical weight of this topic also involves target prioritization. Children with autism typically have multiple areas of need, and treatment time is finite. Arguing for the inclusion of temporal and deictic relational targets requires practitioners to articulate clearly why these targets are clinically significant and how they connect to broader functional outcomes. The evidence from Barron and colleagues supports that argument: these relations are foundational to conversational reciprocity, academic instruction following, and social participation.

Code 2.14 requires that behavior analysts continually evaluate treatment effectiveness. For relational training targets, this means monitoring not just whether the trained relations are acquired but whether transformation of function is occurring — that is, whether the training is producing the generalized relational flexibility it is designed to produce. Progress monitoring should be designed to detect transformation effects, not just performance on trained exemplars.

Assessment & Decision-Making

Assessing temporal and deictic relational responding requires careful attention to how probes are designed. Because these relations are defined relative to shifting reference points, probe items must vary the reference point systematically to distinguish between a topographic response to a specific stimulus configuration and a genuinely flexible relational response. A child who always says 'here' when an object is placed near them may be responding topographically rather than relationally.

The assessment procedure derived from the Barron and colleagues research involves establishing baseline responding on single reversals before training, then implementing training and testing for transformation of stimulus function with novel stimuli. This design allows practitioners to identify whether the trained relations are producing generalized relational flexibility or whether additional training is needed on additional exemplar sets.

Practitioners should also consider the verbal behavior level at which a child is operating more broadly before implementing RFT-based training. Children who are still at early echoic or limited mand stages may not yet have the prerequisite relational responding history to benefit from deictic frame training. Functional assessment of the verbal behavior repertoire — including intraverbal responding that requires tracking speaker perspective — helps determine when this training is appropriate and how to sequence it within a broader language program.

What This Means for Your Practice

Integrating RFT-based relational training into practice begins with expanding the conceptual vocabulary practitioners use when assessing and describing language targets. 'Deictic relational responding,' 'transformation of stimulus function,' and 'single-reversal probe' are concepts that allow clinicians to specify precisely what a child needs to learn and why the training procedure is designed the way it is. That precision matters for treatment planning, team communication, and progress monitoring.

For programs serving children with autism who have intermediate to advanced verbal behavior repertoires, temporal and deictic targets offer a productive area for expanding the sophistication of language programming. The training procedures are efficient, the generalization data are promising, and the functional outcomes — conversational participation, instruction following, perspective-taking — are directly meaningful to families and classrooms.

Finally, this course illustrates the value of engaging with the RFT literature as a source of clinical innovation. Many behavior analysts trained primarily in verbal behavior analysis have limited exposure to the RFT research base, but the two frameworks are complementary rather than competing. Practitioners who understand both are better positioned to design interventions that address the full range of language and cognition targets that children with autism require.