Deictic Relational Responding and Perspective Taking in Autistic Individuals: A Scoping Review

Perspectives on Behavior Science (2024) 47:107–137
https://doi.org/10.1007/s40614-024-00397-2
REVIEW
Deictic Relational Responding and Perspective‑Taking

in Autistic Individuals: A Scoping Review
Nadine Hempkin1,2 · Maithri Sivaraman3 · Dermot Barnes‑Holmes2
Accepted: 12 January 2024 / Published online: 6 February 2024

© Association for Behavior Analysis International 2024
Abstract
Perspective-taking skills are crucial for successful social interactions and some
autistic individuals seem to demonstrate great difficulty in this area. The concept
continues to generate clinical and research interest across mainstream psychology
and within behavior analysis. Within behavior analysis, relational frame theorists
have argued that deictic relational responding is critically involved in perspective-
taking. We conducted a systematic search of the behavior analytic studies on deic-
tic relational responding and perspective-taking in autistic individuals to highlight
methods used to test perspective-taking and deictic relations, methods to train these
if deficits were observed, and evidence for a relationship between deictic relational
responding and perspective-taking. Seven studies met inclusion criteria and we con-
ducted a descriptive analysis of these studies. We found some variation in the meth-
ods used to test and train perspective-taking through deictic relations. Only three
of the studies attempted to demonstrate a link between deictic relational respond-
ing and perspective-taking. Overall, our review highlighted a need for more research
into deictic relational responding and perspective-taking in autistic individuals, and
we discussed specific areas for future research.
Keywords Deictic relational responding · Perspective-taking · Relational frame

theory · Verbal language development · Autism
Perspective-taking is deemed an important area of study both within mainstream

psychology and in behavior analysis. Broadly defined, perspective-taking involves
inferring another person’s desires and beliefs, in order to interpret their behavior
* Maithri Sivaraman
maithri.sivaraman@tc.columbia.edu
1
Mohammed Bin Rashid Center for Special Education, Operated by the New England Center
for Children, Abu Dhabi, United Arab Emirates
2
School of Psychology, Ulster University, Coleraine, Northern Ireland
3
Teachers College, Columbia University, New York, NY, USA
Vol.:(0123456789)
108 Perspectives on Behavior Science (2024) 47:107–137
and predict what they will do next (Howlin et al., 1999; Sigman & Capps, 1997).
Perspective-taking is considered fundamental to human socialization (Mead, 1934;
Piaget, 1984), permitting an individual to adjust their behavior in accordance with
others. The ability to consider another’s perspective is vital in strengthening social
connections (Galinsky & Ku, 2004; Vescio et al., 2003); establishing and maintain-
ing healthy interpersonal relationships (Hughes & Leekam, 2004); and in the devel-
opment of altruism (Underwood & Moore, 1982).
Although typically developing children show signs of perspective-taking from
early childhood (Carpenter et al., 1998), early studies appear to show that autistic
children demonstrate lower levels of this skill (e.g., Baron-Cohen, 2000; Ozonoff
& Miller, 1995). Given these findings on the difficulties of autistic individuals in
understanding the beliefs and desires of others, a vast body of research has investi-
gated perspective-taking in this population (see Pearson et al., 2013, for a review).
Schwartz Offek and Segal (2022) showed that autistic children’s scores on Theory of
Mind (ToM; Wimmer & Perner, 1983) tests, which are used to assess perspective-
taking skills, were significantly different from the scores of neurotypical peers and
those with a developmental language disorder. On balance, researchers have sug-
gested that the differences in false belief performance (a specific type of perspec-
tive-taking skill) between autistic and neurotypical individuals can be accounted for
by differences in their language level (Loukusa et al., 2014). In any case, as noted
previously, perspective-taking seems critical for successful social interaction and
community involvement. Therefore, it seems important to summarize the findings of
studies targeting this skill in autistic individuals.
The Mainstream Psychology Approach to Perspective‑Taking
Most research into perspective-taking employs concepts and techniques under the
approach referred to as “Theory of Mind” (ToM; Baron-Cohen & Hammer, 1997;
Baron-Cohen, 2000). ToM research describes perspective-taking as having five
levels of understanding of information states when teaching individuals to take the
perspective of one another (Howlin et al., 1999). To summarize, ToM, as articu-
lated by Howlin et al., suggests that perspective-taking skills may be taught across
increasingly complex levels of informational states that progress from simple visual
perspective-taking to understanding false beliefs. Although the ToM model was ini-
tially a descriptive account of perspective-taking, it was subsequently adapted for
use as a teaching tool to train ToM skills to children. Reviews of the developmental
and clinical literature on perspective-taking and ToM, however, generally highlight
the difficulty of ruling out the potential role of broader cognitive concepts and capa-
bilities, of which perspective-taking may be a component. As such, the field of per-
spective-taking has found difficulty in formulating an operational definition of what
these skills involve (Davis et al., 2004).
Perspectives on Behavior Science (2024) 47:107–137 109
Behavior‑Analytic Approach
Initial conceptions of the self as one’s own responder was proposed by Skin-
ner (1974) who suggested that self-knowledge is shaped through an individual’s
knowledge of others and by social contingencies that reinforce the discrimination
of one’s own behavior. Learning to tact one’s own thoughts and feelings is likely
related to responding in accordance with others’ thoughts and feelings (Kanter
et al., 2001; Skinner, 1945). By asking a range of questions about the individual’s
self, the verbal community supports the development of an individual’s ability
to identify and discriminate their own behavior, and thus individuals are better
able to predict and control their behavior as they become more self-aware. Other
accounts within behavior analysis have attempted to explain performance in
ToM tasks as a stimulus control problem involving predictions on others’ behav-
ior based on discriminative stimuli available to the self and to others (Schlinger,
2009; see also Spradlin & Brady, 2008 for a stimulus-equivalence based account
of perspective-taking).
Behavior analytic accounts of perspective-taking have recently been extended
and developed within the framework of relational frame theory (RFT; Hayes
et al., 2001). According to RFT, humans respond relationally to events or objects
when the relation is based on contextually controlled cues through a history
of multiple exemplars (see Hughes & Barnes-Holmes, 2015). In other words,
through extensive verbal interactions and experiences, an individual learns to
derive relations between and among stimuli in a manner that does not rely only
on the physical properties of those stimuli (see Hayes et al., 2001).
There are multiple ways to relate stimuli, and RFT identifies patterns of such
relational responding as relational frames. One such frame is the deictic, which
involves relating oneself in space (Here–There) and time (Now–Then) relative to
others (I–You). The combination of the three relations, that is, I–You, Here–There,
and Now–Then, make up the deictic frame (Barnes-Holmes & Harte, 2022;
Kavanagh et al., 2020). According to RFT, learning to respond in accordance with
the deictic relational frame provides the core behavioral unit that may be involved
in behaviors referred to as ToM in the mainstream psychological literature. In par-
ticular, these three relations combine into a basic deictic relational frame, which
involves locating oneself in time and space relative to another individual. The core
idea is that as children learn to respond in accordance with these deictic relations,
they are essentially learning to relate (verbally) the self to others in the context of
particular times and spaces. And indeed, some research has suggested that deictic
framing ability is predictive of perspective-taking (McHugh et al., 2004a).
On balance, it is important to note that according to RFT, perspective-taking
behaviors may involve more than simply responding in accordance with the deic-
tic frame (of I–You, Here–There, and Now–Then). For example, Kavanagh et al.
(2020) outlined an RFT-based analysis of a False-Belief task that involved deic-
tic framing combined with the relating of relations and the relating of relational
networks. The details of the analysis are not important in the current context, but
we mention it here to emphasize that perspective-taking for RFT is a relationally
complex behavior that extends well beyond the core deictic frame of I–You,
Here–There, and Now–Then. Furthermore, it is important to note that although
the three relations are necessary to develop a perspective of the self with respect
to others, each instance of perspective-taking need not involve all three relations.
With that said, most of the research in this area has focused on the deictic frame
and its involvement in perspective-taking, and thus the current review will con-
centrate here too.
As noted above, perspective-taking for RFT involves (but is not restricted to)
patterns of relational responding in accordance with I–You, Here–There, and
Now–Then (i.e., the “deictic” relations). Such deictic relational responses are
learned from a history of multiple exemplars through asking and responding to
questions such as, “What am I doing now?,” “What did you hear there?,” “Where
were you yesterday” etc. Each time these types of questions are answered, the physi-
cal properties of the environment may change, but the relational properties of I–You,
Here–There, and Now–Then remain persistent across all exemplars (Hayes et al.,
2001). Thus, deictic relations are established by demonstration and cannot be traced
to clearly defined formal (i.e., nonarbitrary) dimensions of the environment; that is,
they occur from the relative perspective of the speaker across a potentially infinite
range of physical contexts (McHugh et al., 2004a).
Most of the empirical research on deictic relational responding has employed
various iterations of a protocol originally developed by Barnes-Holmes (2001). The
original 256-trial protocol targets the three deictic relations, as well as the three lev-
els of relational complexity: simple, reversed, and double reversed. In a simple rela-
tional response, none of the elements are reversed (e.g., “I am standing here, and you
are standing there. Where are you standing?”). In a reversed relational response, one
of the relations is reversed (e.g., “If I were you and you were me, where would you
be standing?”). A correct response reflects this relational reversal (framing condi-
tionally with respect to the I–You relation). In a double reversed relational response,
two relations are reversed concurrently (e.g., “If I were you and you were me and
if here were there and there were here, where would you be standing?”), and a cor-
rect response would appear to require more complex derived relational responding
(McHugh et al., 2004b). Studies have successfully used this protocol to both test for
and train deictic relations in both typically developing individuals (e.g., McHugh
et al., 2004a; Weil et al., 2011) and autistic individuals who demonstrate challenges
in their relational abilities (e.g., Rehfeldt et al., 2007).
Some behavior-analytic studies have also suggested a link between deictic rela-
tional responding ability and subsequent performance on more traditional perspec-
tive-taking tasks (e.g., the Sally-Anne1 task, which is frequently used in research on
1
The Sally-Anne task is a classic cognitive development test used to evaluate the theory of mind in
children. In this task, a child is presented with a story involving two characters, Sally and Anne, and a
physical object, such as a ball. The child is then asked a series of questions about the story, including
where Sally will look for the ball when she returns to the room. The correct answer requires the child
to understand that Sally does not know that Anne has moved the ball to a new location while she was
absent, and therefore will look for it in its original location. The Sally-Anne task is often used to assess
the development of a child’s ability to understand that others can have beliefs, desires, and intentions that
differ from one’s own.
perspective-taking to test for ToM abilities). In one such study, Gómez-Becerra et al.
(2007) presented the Sally-Anne task to 15 children (5 autistic children, 5 children
with down syndrome, 5 typically developing children) aged between 4 and 6 years
old. The researchers found that the autistic children scored the lowest in perspective-
taking tests, followed by the children with Down syndrome. Typically developing
children scored the best across all perspective-taking tests. It should be noted, how-
ever, that in all cases the children’s scores were associated with their language abili-
ties, relational repertoires, and I–You discriminations. That is, the researchers found
that children who scored the best in the perspective-taking tests performed well on
measures of I–YOU relations, he–she discriminations, spatial–temporal relation-
ships, and derived relations (i.e., they performed well on skills considered to be pre-
requisites for perspective-taking; McHugh et al., 2004b; Hayes, 1984).
The results of the Gómez-Becerra et al. (2007) study highlight some important
considerations. First, the results showed that irrespective of group performance,
derived relational responding was associated with perspective-taking, and thus there
appears to be a need to investigate the evidence for an association between perfor-
mance on tests of deictic relations and traditional ToM tasks across studies. In addi-
tion, although the study examined the deictic relations of I–You, and Now–Then, it
did not explicitly define these as deictic relations, nor did it consider the other deic-
tic relation (i.e., Here–There). In addition, other studies (see, for example, McHugh
et al., 2004a) have used the terms deictic relations, deictic frames, and perspective-
taking interchangeably, although there are key technical distinctions between them
within RFT (we will return to this issue in the Discussion). As such, it appears that
the language being used in the literature is not entirely consistent, which may cause
difficulty for practitioners in identifying effective methods for training such skills.
To the best of our knowledge, there has been no previous research summarizing
the literature on perspective-taking and deictic relational responding in autistic indi-
viduals. Given the applied importance of perspective-taking skills and correlational
evidence relating deictic responding to ToM tasks (e.g., McHugh et al., 2004a;
Gómez-Becerra et al., 2007), conducting a scoping review and descriptive summary
of this literature with autistic individuals seems worthwhile. Such an undertaking
may be particularly important considering that some autistic individuals have been
shown to experience challenges with this skill (Rehfeldt et al., 2007).
The Current Review
A small number of behavior-analytic studies have attempted to measure perspective-

taking; however, only a handful have targeted changes in deictic relational deficits, and
even fewer studies have attempted to extend research findings to autistic children and
adolescents. At the outset we focused on the basic postulate that deictic relations are
very much involved in perspective-taking, specifically targeting the empirical evidence
relevant to teaching these skills to autistic individuals who demonstrated deficits in
perspective-taking abilities. In doing so, we found there were a limited number of stud-
ies that tested/trained deictic relations and investigated the association to perspective-
taking in autistic individuals. Within the limited number of studies that are available in
the literature, we identified significant variation in how studies assessed perspective-

taking, how they trained deictic relations, and conducted evaluations of perspective-
taking posttraining. Thus, our scoping review aimed to identify studies that targeted
perspective-taking and deictic relations in autistic individuals and highlight (1) the
methods that were used to assess perspective-taking; (2) the methods used to train deic-
tic relations if perspective-taking was found to be absent; and (3) whether studies that
trained deictic relations were able to provide evidence for a relationship between deictic
relations and perspective-taking.
Method
We undertook a scoping review of studies that conducted empirical investigations on

deictic relational responding and perspective-taking in individuals with an autism diag-
nosis. We were specifically interested in the methods that were used to test perspec-
tive-taking, the deictic relations that were trained and how these relations were trained,
and whether increases in deictic relational responding influenced perspective-taking
abilities. We followed the Preferred Reporting Items for Systematic Reviews and Meta-
Analyses (PRISMA) guidelines for the scoping review. The PRISMA checklist has
been included as part of the supplementary materials.
Inclusion Criteria
We included articles that reported empirical data on deictic relational responding and
perspective-taking in individuals with a diagnosis of autism. As such, we included stud-
ies in which deictic relations were either assessed and/or trained, and these were either
used as a measure of perspective-taking, or perspective-taking skills were assessed
through another protocol (e.g., ToM tasks). We also included one study in which the
participants had a diagnosis of Asperger’s syndrome, which we considered to be high-
functioning autism based on the most recent definition in The Diagnostic and Statisti-
cal Manual of Mental Disorders (5th ed.; DSM-5; American Psychiatric Association,
2013).
Studies were excluded from the review if they were theoretical articles, unpub-
lished dissertations, or did not include participants with an autism diagnosis (i.e., stated
“developmental delay,” “intellectual disability,” or participants were typically devel-
oping). Studies were also excluded if they only targeted perspective-taking (broadly
defined) but did not train or test for deictic relations.
Search Procedure
We first conducted a database search to identify articles that met our inclusion cri-
teria. We conducted searches on PubMed, Scopus, ScienceDirect, and Wiley Online
Library on November 18, 2022 and included articles that were published until that
date. The search terms included “Deictic relational responding AND autism AND
perspective-taking” OR “Deictic relations AND autism AND perspective-taking”
(some syntaxes varied based on the database used) as well as a filter for the type of
publication (i.e., peer-reviewed journal articles). We did not use any date range for
the studies.
Once eligible studies were identified, we searched their reference lists to iden-
tify additional articles. Finally, we conducted a citation search on Google Scholar
to identify articles that cited the included studies to identify additional articles that
met our inclusion criteria. The first author conducted the searches and the screening.
An independent observer reviewed the full texts against the inclusion criteria. The
agreement between reviewers was 100%.
Secondary Search
A secondary search was conducted with the terms “I–You” and “Here–There” and
“Now–Then” because we observed these terms adopted in the included studies.
This search only revealed one extra study. The use of a secondary search has been
adopted by other recently published reviews (e.g., Padilla et al., 2023).
Outcome Measures and Data Extraction
The first author extracted information on the following areas from the eligible stud-
ies: (1) participant characteristics; (2) the method used to test perspective-taking
including the stimuli used; (3) mastery criterion; (4) which deictic relations were
trained (e.g., I–You, Here–There, Now–Then, and the level of complexity); (5) how
deictic relations were established/trained; (6) the efficacy of training; and (7) what
evidence there was (if any) for a relationship between deictic relations and perspec-
tive-taking. The participant characteristics coded were the demographic variables
including age, gender, and any diagnostic information provided in the study.
An independent observer coded data from 100% of articles (i.e., all included
studies) and interobserver agreement was found to be 100%.
Results
A total of 126 articles were identified through the database search. Following the
removal of duplicates, 96 unique articles were included in the initial screening. Of
these, four were deemed eligible for the review. The reference list search of these
articles resulted in the screening of an additional six articles. Of these, two arti-
cles met our inclusion criteria. Finally, the citation search generated two additional
articles. However, these two articles were unpublished theses, and as such were
not included in the current review. One additional study was identified for inclu-
sion when conducting a secondary search using the terms, “I–You,” “Here–There,”
“Now–Then” along with the primary search terms. Thus, in total, seven studies were
identified for inclusion in the present review. All studies were reported in English.
Figure 1 presents a summary of the reference attrition process. See Tables 1, 2, and
3 for a summarized display of the results described below.
Fig. 1 Reference attrition process
Demographic Characteristics
A total of 30 participants2 were reported across the seven studies included in the
review. Of these, there were 28 male and 2 female participants. All studies included
at least one autistic participant. Twenty-six participants had a diagnosis of autism,
three had a diagnosis of Asperger’s, one had a diagnosis of pervasive developmental
disorder (PDD). Based on the DSM-5 (American Psychiatric Association, 2013) def-
inition, participants with PDD and Asperger’s were regarded in the present review
as having a diagnosis of autism.
The average age of the participants was 9.3. Two studies (Gomez-Becerra et al.,
2007; Lovett & Rehfeldt, 2014) did not include the specific ages for each participant
but both noted the age ranges of the participants.
Methods Used to Test for Deictic Relations
Four of the included studies (Gilroy et al., 2015; Jackson et al., 2014; Lovett &
Rehfeldt, 2014; Rehfeldt et al., 2007) used some extension or modification of the
Barnes-Holmes (2001) protocol to assess perspective-taking abilities. Three of these
studies (Jackson et al., 2014; Lovett & Rehfeldt, 2014; Rehfeldt et al., 2007) used
2
Two studies (Gomez-Becerra et al., 2007; Rehfeldt et al., 2007) also included some participants who
either had Down’s Syndrome or were neurotypical. We did not include these participants in our analysis
because our aim was to evaluate the impact of deictic relational training on autistic individuals.
Table 1 Demographic results from systematic review
Study Aim Diagnosis Total number of participants Age Gender
Gómez-Becerra et al. (2007) a) Tested for the replicability of 5 autism 5 autistics Range: 4–18 years 4M; 1F
studies that show differences
between the perspective-taking
responses of children with autism,
children with developmental dis-
abilities and typically developing
children.
b) Evaluated the results of a modi-
fied False Beliefs Tests and tested
the effects that different prompts
had on the performance of the
participants.
c) Analyzed the potential dif-
ferences in perspective-taking
repertoires of the participants in
relation to other behaviors that
might be related or prerequisites
for the development of functional
equivalent relations, such as
certain level of a verbal repertoire,
or a discriminative repertoire (I/
you/he/she, past/present/future,
present/absent of object or indi-
vidual etc.)
115
Table 1 (continued)
116
Rehfeldt et al. (2007) (a)To determine whether children High functioning Nine autistic/Aspergers Mean: 10 years 1 month (Range: 9M
with high-functioning autism autism, Asperger’s 6 years 8 months to 13 years 4
spectrum disorder and Asper- Syndrome. months)
ger’s syndrome would perform
significantly worse on the Barnes-
Holmes protocol than their
age-matched typically developing
peers.
(b)To determine if performance
on the Barnes-Holmes protocol
improved following specific
reinforcement for responding rela-
tionally. If so, it could be argued
that perspective-taking involves
derived relational responding.
Table 1 (continued)
Jackson et al. (2014) The purpose of this study was to 4 Autism, 1 PDD-NOS Five Mean: 5 years 6 months (Range: 5 5M
replicate and extend the study years to 6 years 1 month)
by Weil et al. with participants
diagnosed with autism.
(Weil et al. (2011) taught deictic
relations to three typically devel-
oping children using the Barnes-
Holmes protocol and assessed any
change in their performance on
second-order false belief tasks.
All participants were able to
reach a mastery criterion of 80
% correct on the protocol, and,
as the participants acquired the

deictic relations at the reversed
and double reversed levels, scores
on the false-belief tasks increased.
Responding also generalized to
untrained versions of ToM tasks.
One notable limitation to this
study was the possibility of test/
retest effects on ToM assess-
ments.)
117
Table 1 (continued)
118
Lovett and Rehfeldt (2014) (a)Evaluate the use of MEI to teach Asperger’s Three Range: 17–18 years old 2M; 1F
perspective-taking skills to ado-
lescents with Asperger syndrome.
(b)Examine generalization of
perspective-taking skills follow-
ing instruction. Generalization
was assessed using standardized
assessments evaluating ToM per-
formance. As well, generalization
of deictic responding to a more
natural social interaction format
was evaluated.
Gilroy et al. (2015). Evaluate the effectiveness of a ASD Three Mean: 10 years old (Range: 8 years 3M
teaching protocol to establish old to 13 years old).
relational responding in children
diagnosed with an autism spec-
trum disorder using the more
naturalistic story reading approach
from Davlin et al. (2011).
Belisle et al. (2016) The purpose of the study was Autism Three Mean: 14 years (Range: 12 years 3M
to evaluate the efficacy of a old to 18 years old).
relational training procedure in
teaching single-reversal deictic
relational responding to autistic
individuals.
Table 1 (continued)
Barron et al. (2018) The purpose of the current study Autism Two Mean: 6 years old (Range: 6 years 2M
was to evaluate the efficacy of old to 6 years old).
using relational training tech-
niques to teach single reversals of
“Then-Later” and “Here-There”
relations to children with autism.
In addition, the current study
sought to evaluate transfers and
transformations of stimulus func-
tion across both skills.
M Male, F Female, ASD Autism Spectrum Disorder, PDD-NOS Pervasive Developmental Disorder Not Otherwise Specified
119
Table 2 Methods to test and train deictic relations and perspective-taking
120
Study Perspective-taking Tests Deictic Relations Test Deictic Relation(s) Trained How were deictic relations estab-
lished/trained?
Gómez-Becerra et al. (2007) -Sally-Anne Perspective-Taking -Tasks of Discrimination of I/you N/A N/A
Test and I/he–she
- Emotion Faces Test
- Pitu Test (contextual task)-
identifying emotions of others
Rehfeldt et al. (2007). N/A A modified version of the N/A N/A
Barnes-Holmes protocol, as
reported by McHugh et al.
(2004a, b).
Jackson et al. (2014) Tests on five levels of ToM Modified version of the Barnes- I-You, The full version of the Barnes-
throughout study: Holmes protocol (Weil et al. Here-There, Holmes protocol, as reported by
Level 1 (simple visual perspec- 2011) Now-Then McHugh et al. (2004a, b) and
tive-taking). Level 2 (complex used by Weil et al. (2011)
visual perspective-taking).
Level 3 (seeing leads to know-
ing). Level 4 (true belief).
Level 5 (false belief).
Lovett and Rehfeldt (2014) Theory of Mind Inventory A perspective-taking protocol- I–You, Multiple exemplar instruction of
(TOMI) included the three types of Here–There, instructional protocol.
deictic frames (i.e., I–You, Now–Then Included a total of 36 trials
Here–There, and Now-Then) with 12 trials of each simple,
and the three levels of rela- reversed, and double reversed
tional complexity (i.e., simple, relations.
reversed, and double reversed) During instruction automated
used in the McHugh et al. feedback and error correction
(2004a, b) protocol. were delivered.
Table 2 (continued)
Study Perspective-taking Tests Deictic Relations Test Deictic Relation(s) Trained How were deictic relations estab-
lished/trained?
Gilroy et al. (2015) N/A Close to Davlin et al. (2011) pro- Similar to protocol—Davlin Deictic framing training protocol
tocol—which was an extension et al. (2011): with differential feedback
of the Barnes-Holmes deictic I–Character;
framing protocol (McHugh Here–There;
et al., 2004a, b; Rehfeldt et al., Now–Then;
2007) utilizing a story reading
context.
Belisle et al. (2016) N/A PEAK-T assessment- provides a You-I Differential reinforcement:
metric for evaluating an indi- Single-Reversal (You) Deictic
vidual’s relational abilities. Training
Baseline: Single-reversal (You) Mixed Single-Reversal (I and You)
(sYOU) and mixed single- Deictic Frame Training
reversal (I) (sI) relations were All stimuli and procedures adapted
tested for each stimulus set. from PEAK-T
Barron et al. (2018). N/A PEAK-T assessment- provides a Then–Later and Here–There Differential reinforcement:
metric for evaluating an indi- relations. Then–Later and Here–There
vidual’s relational abilities. Training
In the baseline phase, Then– Single Reversal
Later, Here–There, single- Transformation
reversal, and transformation
relations were all probed for
each stimulus set.
N/A = Not Applicable

121
122
Table 3 Outcomes of training

Study Outcomes of Deictic Training Evidence for relationship between deictic relations and perspective-
taking
Gómez-Becerra et al. (2007) N/A Children who scored the best on tasks that contained some deictic
frames (i.e., I–you, now–then), are those who later scored better in
the perspective-taking tests.
Rehfeldt et al. (2007). N/A N/A
Jackson et al. (2014) All participants demonstrated improvements in deictic responding Deictic relational training was taught to the participants, but this was
and met mastery criteria. not sufficient to change more generalized perspective-taking skills,
as measured by a Theory of Mind assessment. Only one out of five
participants demonstrated a clear increase in ToM from the pretest
to the posttest
Lovett and Rehfeldt (2014) All participants demonstrated improvements in deictic responding Little change was observed in ratings on the ToM caregiver report
and met mastery criteria. measure from pretest to posttest.
Gilroy et al. (2015) All participants demonstrated improvements in deictic responding N/A
and met mastery criteria.
Belisle et al. (2016) All participants demonstrated improvements in deictic responding N/A
Barron et al. (2018). All participants demonstrated improvements in deictic responding N/A
N/A = Not Applicable

modified protocols that had been reported in the study by McHugh et al. (2004a, b)
and all these studies included trials testing the three types of deictic relations (i.e.,
I–You, Here–There, and Now–Then) and the three levels of relational complexity
(i.e., simple, reversed, and double reversed).
The study by Gilroy et al. (2015) followed the same format as the protocol by
McHugh et al. (2004a, b), but they utilized a naturalistic story-telling approach that
was reported in a study by Davlin et al. (2011). Each instance of deictic relating con-
sisted of two queries, with both forming the basis of the relation (e.g., “You are here
at home, Hercules is by the pillars. If you were Hercules and Hercules, were you. . . .
Where would Hercules be? Where would you be?”
A relatively standardized approach known as Promoting the Emergence of
Advanced Knowledge Transformation Module has been employed to teach relational
skills to autistic children (PEAK-T; Dixon, 2016). Two of the studies included in the
review assessed participants’ deictic relational skills using the PEAK-T assessment
(Belisle et al., 2016; Barron et al., 2018). This assessment provides a metric for eval-
uating an individual’s relational abilities.
The study by Barron et al. (2018) just tested two relations of Then–Later and
Here–There across two levels of complexity (simple and reversed). Although
“Then–Later” is not a conventional deictic relation in the RFT literature, nor
in the original Barnes-Holmes protocol, it is noted to have been included in the
PEAK-T module, and in the study by Barron et al., as a culturally constructed
relational cue that can be heard in everyday language.3 The study by Belisle
et al. (2016) tested the I–You relation across two levels of complexity (simple
and reversed). Both of these studies tested these relations by having the experi-
menter show both sides of a picture card to the participant and holding the card
vertically so that the participant and the assessor could see opposite sides of
the picture card. Questions were used to test the relations (i.e., “If I were you
and you were me, what would you see?”). Presentation of picture cards were
randomized. No reinforcement or feedback was provided to participants during
testing.
In the study by Gómez-Becerra et al. (2007) I–You relations were assessed by
asking each participant informally to evaluate whether he/she could discriminate
between I–You and I–He–She. In addition, incidental trials in the form of games
were used to evaluate these skills. Some of the antecedents presented included:
“Give me the pencil,” “Whose pencil is this?” Spatial relations were assessed using
The Boehm Test of Basic Concepts (Boehm- 3; Boehm, 2000). These research-
ers also assessed skills such as same and different, before and after, first and last,
always and never, and matching. If the participant emitted an incorrect response, up
3
The reader should note that Then–Later is not considered to be a core deictic relation because it does
not necessarily involve a specific perspective (from I–Here–Now); that is, both Here and Now, in the
deictic frame, help to define the perspective of a specific I (i.e., the deictic I). Of course, it is possible,
and even likely, that Then and Later may be responded to from a specific perspective (e.g., “I will do x
then and y later”), but critically a particular perspective is not necessarily involved in a Then–Later rela-
tion because no I, Here, or Now is specified.
to three prompts were provided to produce a correct response. The prompt hierarchy
followed was (1) the placement of the materials was changed, and the antecedent
was presented again; (2) the relevant characteristics of the presented pictures were
described (i.e. “this is a picture of a tree, this is a picture of the same tree, and this is
a picture of a different tree”), and the antecedent was presented again; (3) the size of
the array of pictures presented was reduced.
Overall, four studies tested for deictic relations using some modification of the
Barnes-Holmes (2001) protocol, and two using the PEAK-T assessment. Gómez-
Becerra et al. (2007) used a range of different tests to test for I–You and spatial rela-
tions. As such, there was some variation in terms of the deictic relations that were
assessed, and the level of complexity involved in that testing.
Methods Used to Test for Perspective‑Taking using Traditional

Methods and ToM
Three of the included studies (Jackson et al., 2014; Lovett & Rehfeldt, 2014; Gómez-
Becerra et al., 2007) used methods to test for perspective-taking using traditional
psychological methods and/or ToM. Jackson et al. assessed whether there were any
changes in perspective-taking skills following training and testing on deictic rela-
tions. The participants in their study completed an experimenter-defined test on
all five levels of ToM four times throughout the study; in particular, the ToM tests
increased in complexity from visual perspective-taking (level 1) to a false belief task
(level 5).
The study by Gómez-Becerra et al. (2007) used an adapted Sally-Anne test,
an emotion faces test, and an experimenter-designed “Pitu test” as measures
of perspective-taking. During the adapted Sally-Anne test, participants viewed
three videos each consisting of four phases. Various items were moved by actors
in each video and children were required to answer a set of questions regard-
ing the movement of objects. Videos got progressively more difficult, and the
researchers measured difficulty in terms of the number of objects that were
manipulated in each clip. In the emotion faces test, participants were required
to interpret the emotional states of others by labelling emotions depicted in
pictures, stories, and live representations. During the “Pitu Test,” developed
by the researchers, participants were required to identify the emotional states
of a character named Pitu, who was depicted in various everyday situations.
Each scene contained a contextual cue to allow for the formation of functional
classes between the contextual cue and a specific emotional state (e.g., a flower
was used as a cue to depict the emotional state “sad”). Participants were shown
four flashcards that depicted the character in an everyday situation along with
a contextual cue and were asked related questions. The fourth flashcard was
the test flashcard and was presented with the antecedent “How does Pitu feel
here?”
The study by Lovett and Rehfeldt (2014) used the Theory of Mind Inventory4
(ToMI; Hutchins et al., 2012). The ToMI is a 48-item caregiver-report measure that
evaluates caregivers’ perceptions of the perspective-taking skills of autistic adoles-
cents, including concepts such as false belief, visual perspective-taking, and irony.
Caregivers make a hash mark on an anchored scale, and the placement of the mark
is later measured with a ruler to provide a score ranging between 0 and 20. Higher
ratings on the scale represent a greater degree of ToM or perspective-taking skills
(Lerner et al., 2011).
Overall, three studies employed some methodology to test for perspective-tak-
ing using perspective-taking tasks and ToM. Two of these studies conducted direct
assessments whereas one study used caregiver reports.
Stimuli
All studies used some format to present a scene, and this was followed by ques-
tions regarding the scenario. Rehfeldt et al. (2007) presented questions in an auto-
mated format where participants were required to read the statements and then
answer questions regarding the statements. Lovett and Rehfeldt (2014) also used
an automated format, where a picture was presented on a screen with information
depicting the scene (e.g., activities, locations, feelings of characters), and a ques-
tion was presented at the bottom of the screen which tested deictic relations (i.e.,
I–You, Here–There, and Now–Then). Gilroy et al. (2015) presented scenarios in a
story-telling format and later asked questions regarding the scenario. In the Jackson
et al. (2014) study, the researcher sat across the table from the participant and read
each statement to the participant (e.g., “I have a red brick and you have a green
brick”). The researcher then asked the participant two questions (e.g., “Which brick
do I have?” “Which brick do you have?”). The two studies that used the PEAK-T
assessment (Belisle et al., 2016; Barron et al., 2018) used 2D picture cards which
were followed by related questions to assess I–You relations (Belisle et al., 2016)
and Then–Later, Here–There relations (Barron et al., 2018).
Jackson et al. (2014) used 2D cards to assess simple visual perspective-taking (level
1) and complex visual perspective-taking (level 2), and 3D objects for the other three
levels of the perspective-taking test. The Gómez-Becerra et al. (2007) study used a
video presentation during the Sally-Anne test, 2D pictures during the test for emotions,
and the Pitu test, followed by various questions.
Overall, there was wide variation in the types of stimuli that were used to test deictic
relations (i.e., text, auditory statements, visual) and perspective-taking (e.g., pictures,
objects, videos).
4
The Theory of Mind Inventory (ToMI) is a standardized questionnaire designed to assess an individ-
ual’s understanding and awareness of mental states, such as beliefs, desires, intentions and emotions, in
themselves and others. The ToMI provides researchers and clinicians with a quantitative measure of an
individual’s theory of mind abilities. It consists of a series of scenarios and questions that prompt partici-
pants to infer and explain the mental states of story characters. The ToMI has been widely used in devel-
opmental psychology, cognitive neuroscience, and clinical settings to study various aspects of theory of
mind in both children and adults.
Test Trials
Test trials for deictic relations that utilized some modification of the Barnes-Holmes
(2001) protocol typically involved the presentation of three types of relations (e.g.,
simple, reversed, and double reversed relations). Within each of these three types of
relations were trials that evaluated responding to the three different deictic relations
(I–You, Here–There, Now–Then). These trials were presented as questions which
tested such relations. For example, testing the I–You relation involved asking a question
such as “I have a green brick, you have a red brick. Which brick do I have? Which brick
do you have?”
There were variations in the number of trials that were conducted across studies. In
the Rehfeldt et al. (2007) study, eight trials for the simple relations were included in
the protocol, including two each to test I–You and Here–There, and four trials to test
Now–Then deictic relations. Thirty-six trials were presented for the reversed relations,
including 8 I–You, 12 Here–There, and 16 Now–Then trials. Thirteen trials were pre-
sented for the double reversed relations, including four trials for I–You/Here–There and
nine Here–There/Now–Then trials; there were two questions per trial.
In the study by Lovett and Rehfeldt (2014), the testing protocol and instructional
protocol each consisted of 36 trials and included only one question per trial. Gilroy
et al. (2015) presented 15 trials for simple relations, and 11 trials each for reversed rela-
tions and double reversed relations. The specific content of each probe (e.g., stories
referenced) were kept consistent across participants. In the Jackson et al. (2014) study,
deictic relations were presented across all three levels of complexity for a total of 18 tri-
als (two of each relation type at each complexity level). Data were collected on correct
and incorrect responses for a total of 40 trials for the TOM tests (eight at each of the
five levels).
The two studies that used the PEAK-T assessment (Belisle et al., 2016; Barron et al.,
2018) did not indicate the specific number of trials. Likewise, Gómez-Becerra et al.
(2007) did not specify the number of test trials used to assess deictic relational respond-
ing or perspective-taking.
To summarize, the testing procedures used across studies revealed wide variations
in the test protocol, the types of stimuli used, and in the types and number of trials to
assess deictic relational responding.
Methods Used to Train Deictic Relations
Deictic Relations Trained
Five of the seven studies trained deictic relations. Of the two studies that did not
involve a training component, Rehfeldt et al. (2007) undertook a comparative study
on the perspective-taking abilities between autistic and typically developing indi-
viduals. Likewise, Gómez-Becerra et al. (2007) did not train deictic relations but
tested I–You and Now–Then relations (although not explicitly stated as such), as a
prerequisite for more traditional psychological tests of perspective-taking and com-
pared performance across groups of children.
Of the five studies that involved some form of training, three studies trained all
three deictic relations (i.e., I–You, Here–There, Now–Then; Jackson et al., 2014;
Gilroy et al., 2015). Belisle et al. (2016) trained one deictic relation (I–You), and
Barron et al. (2018) trained two deictic relations (Then–Later, Here–There). In sum,
three of the five studies trained all three relations.
Procedure for Training Deictic Relations
As mentioned previously, five of the seven studies trained deictic relations follow-
ing assessment. Several procedures were tested across studies for their efficacy and/
or utility in training deictic relations. Overall, these procedures included differential
reinforcement using PEAK-T (Belisle et al., 2016; Barron et al., 2018), and multi-
ple exemplar training (Jackson et al., 2014; Gilroy et al., 2015; Lovett & Rehfeldt,
2014). Readers should note, however, that the procedure was termed multiple exem-
plar instruction by Lovett and Rehfeldt (2014).
Training sessions conducted by Belisle et al. (2016) and Barron et al. (2018) used
a similar procedure as the assessments. For example, a card was presented with dif-
ferent pictures on opposite sides, and the participant was shown both sides of the
card. When the card was held vertically, the participant and experimenter saw differ-
ent pictures, and the participant was then asked perspective-taking based questions
(e.g., “If I were you and you were me, what would you see?”). Differential reinforce-
ment was provided contingent on responding. Mixed test probes were conducted
with a novel set of stimuli to test for generalization from training to test stimuli.
In the Lovett and Rehfeldt (2014) study, instruction for each level of relational
complexity was introduced sequentially. Instruction on simple relations was con-
ducted first, and when the participant performed at or above mastery criterion (i.e.,
80% correct responding for three consecutive presentations of the simple relations),
instruction on reversed relations was introduced. Training for reversed and double
reversed relations was conducted in the same manner as simple relations, and differ-
ential reinforcement was provided for responses.
In the Jackson et al. (2014) study, a modified version of the Barnes-Holmes pro-
tocol was utilized for training purposes. Each of the trials consisted of one statement
followed by two questions and both had to be answered accurately for the trial to
be scored as correct. Simple relations were trained first, followed sequentially by
reversed relations, and double reversed relations. Prompts were provided initially
and faded once the mastery criterion was met. No details were provided regarding
the type of programmed feedback used in the study. Gilroy et al (2015) who used a
similar modified version of the Barnes-Holmes protocol, provided differential rein-
forcement during deictic relational training with positive reinforcement provided
contingent on equal or greater accuracy from earlier sessions.
Training Outcomes
In terms of the efficacy of the intervention procedures, all studies reported that all
participants demonstrated improvements in deictic responding and met mastery
criteria (see Table 3 for an overview). The majority of studies used some form of
differential reinforcement and multiple exemplar training. Although there were vari-
ations in the implementation of these procedures, they seemed effective at training
the targeted deictic relational responses in the participants.
Evidence for Relationship between Deictic Relational Responding

and Perspective‑Taking
Three of the studies in the review examined whether improvements in deictic rela-
tional responding affected performance on traditional perspective-taking measures.
Overall, there was limited evidence that training deictic relations influenced per-
spective-taking as measured through more traditional psychological methods (e.g.,
ToM). For example, the study by Lovett and Rehfeldt (2014) found that following
improvements in deictic responding there was little improvement in ratings on the
ToM caregiver report measure from pretest to posttest for any of the participants.
Likewise, Jackson et al. (2014) found that despite increases in performance on deic-
tic relations, performance on ToM tests did not change significantly for any of the
participants in the experimental group. Gómez-Becerra et al. (2007) did, however,
find that children who scored the best in the perspective-taking tests performed well
on measures of I–You relations, /He–She discriminations, spatial/temporal relation-
ships, and derived relations. Note, however, that this finding was correlational and
thus only suggestive of a relationship between deictic relational responding and tra-
ditional perspective-taking abilities. It should be noted that the studies contained in
the current review do not indicate that simply training deictic relations will lead to
better performance on traditional perspective-taking measures.
Discussion
We conducted a scoping review of the literature on deictic relational responding

and perspective-taking in autistic individuals to characterize the methods used
to assess perspective-taking, methods used to test and train deictic relations and
whether an empirical relationship was demonstrated between deictic relational
responding and perspective-taking in autistic individuals. We identified seven
studies that met our inclusion criteria and found that five of these studies used
various methods to train deictic relations. Three of the studies trained all deictic
relations, one study trained only one deictic relation (I–You), and another study
trained two deictic relations (Here–There and Then–Later; as explained previ-
ously, the latter may not always function as a deictic relation). All five studies
reported positive improvements in participant performance in deictic relational
responding post-training.
In addition, we found that only one of the seven studies in the current review
showed a tentative association between deictic relational responding and perfor-
mance on perspective-taking tasks, and two other studies found that performance
on these tasks was not influenced by improved deictic relational responding. Over-
all, therefore, limited evidence emerged from our review to support a relationship
between deictic relational responding and improved performance on traditional per-
spective-taking tasks.
Some of the results highlighted above warrant additional discussion. First, there
appeared to be some disparities in the deictic relations that were trained. According
to RFT, as argued in the introduction, the combination of the three relations, that is,
I–You, Here–There, and Now–Then, make up the deictic frame. That is, by defini-
tion, the deictic frame involves locating oneself in space and time relative to others.
Therefore, if the aim is to facilitate the emergence of the deictic frame in autistic
individuals, then it seems important for the training to encompass all three relations.
On balance, such training may require multiple sessions conducted across extended
periods of time, which may be challenging to accomplish, resulting in researchers
choosing to train deictic relations individually. In any case, training across all three
relations would seem to be a wise strategy to establish a specific perspective in terms
if I–Here–Now. Other relational responses could certainly be included in the train-
ing, such as Then–Later (in this case ensuring that Then versus Later is responded
to from a specific perspective). It should be noted parenthetically that during the
review process of an earlier version of the current article, a number of points were
raised concerning the concept of deixis itself within RFT, including the definition of
a deictic frame. Given the general importance of this topic, we offer a commentary
on deixis in RFT in Appendix 1, although it may seem somewhat tangential to the
main purpose and focus of the current article (i.e., as a systematic scoping review).
A key finding of the current review is that the training on deictic relations did
not lead to increases in measures of perspective-taking (e.g., ToM) in autistic indi-
viduals. Although this was obtained from only three studies in our review, this find-
ing is supported by other studies with typically developing children. For example,
Montoya-Rodríguez and Molina Cobos (2016) found that despite improved perfor-
mances on deictic relations, the performance on ToM tests did not change for any of
the participants in the experimental group. However, it is important to highlight at
this point that this lack of association between deictic relations and perspective-tak-
ing is by no means a unanimous finding in the literature. For instance, other studies
conducted with typically developing children (e.g., Weil et al., 2011) and individu-
als with schizophrenia (e.g., O’Neill & Weil, 2014) noted that scores on ToM tasks
improved when deictic relational training was implemented. This clearly seems to
be an issue that warrants additional consideration, and in the next section, we will
attempt to offer some potential explanations for the finding in our review.
One possible explanation for the lack of improvement on perspective-taking skills
may simply be a failure in generalization. The format of the deictic relational train-
ing (e.g., statements and instructions that are used, specific behaviors being trained)
is different from that of the perspective-taking measures such as the Sally-Anne task.
Although the deictic relational training involves multiple exemplar training, which
means that a variety of objects or activities are used across trials, this variation may
not be sufficient for some autistic children who have been reported to experience dif-
ficulties with generalization of skills (see, for example, Schreibman, 2000).
In addition, perspective-taking measures such as the Sally-Anne task involve a

relatively naturalistic format in that the test begins with a story followed by ques-
tions. Perhaps the inclusion of more naturalistic test questions or in situ training dur-
ing the deictic relational training phase may remedy this issue. One study included
in the present review that did incorporate a more naturalistic approach to testing
and training deictic relations (Gilroy et al., 2015; see also Davlin et al., 2011 and
Heagle & Rehfeldt, 2006, for other examples) found that participants were able to
both acquire and generalize deictic relational skills in the context of a storybook
and across all three levels of complexity. The researchers did not, however, explore
whether participants performed better on perspective-taking tasks as a result of the
naturalistic deictic training.
There may be other explanations to our findings indicating a weak association
between improvements in deictic relational responding and traditional perspective-
taking tasks. In this next section, we will briefly discuss other conceptual charac-
terizations of perspective-taking within behavior analysis. Barnes-Holmes et al.
(2013) pointed out that responding on the basis of false belief is complex because it
involves combining the deictic relations with logical-not. For instance, consider the
question “Where does Maria think the book is?” from a false belief task. The cor-
rect response to this question involves indicating that I would know where the book
is, and also that Maria would not know because she was not present when the book
was moved. That is, a correct response will require incorporating a combination of
the three deictic relations along with the NOT (i.e., NOT There–Then). Kavanagh
et al. (2020) argued that responding correctly on false belief tasks involves the basic
deictic frame as well as relating relations. That is, at first, my relation to the book
(observing it being placed on the table) is in a frame of coordination with Maria’s
because we are in the same room and observe the same events. My (self) subsequent
observation of the book on a shelf and Maria observing the book on the table are
related by a frame of distinction because Maria and I were in different rooms when
the book was moved. In essence, more recent theoretical accounts of RFT suggest
that false belief tasks may involve more than the basic deictic frame. As such, these
arguments indicate that deictic relational responding alone may not be sufficient to
perform correctly in a false–belief task.
In another account, Guinther (2017, 2018) described a relational triangulation
model for perspective-taking, which did not involve deictic relational responding.
He proposed a triangle in which two corners represent two individuals (and their
body positions) although the third corner represents the target stimuli being attended
to by the individuals. In particular, Guinther used the term triangulated referencing,
which was defined as an individual deriving the relation (e.g., opposition, coordi-
nation) between one’s own and another individual’s behavior towards a particular
stimulus. An important point to be noted is that triangulated referencing was named
as such because of its close alignment with social referencing (Dube et al., 2004) in
which an infant and a caregiver jointly attend to stimuli, and infants look towards a
caregiver to determine how to respond to an unfamiliar stimulus such as a stranger.
Recent developments in RFT have proposed mutually entailed orienting and evoking
(MEO; Barnes-Holmes & Harte, 2022) as a technical concept to describe infant–car-
egiver interactions in which an infant and a caregiver jointly attend to stimuli and the
infant looks back and forth between the stimulus and the caregiver. These authors
argued that early instances of cooperation between infants and caregivers are central
to the development of derived relational responding, and have been said to facilitate
the emergence of a verbal self.
As such, there seems to be broad alignment between the relational triangulation
account of perspective-taking, the Kavanagh et al. (2020) account of relating rela-
tions and the concept of mutually entailed orienting and evoking. In recognizing
areas of overlap, we are not suggesting that there are no differences between these
conceptual analyses, but an adequate review of any such differences, and the poten-
tial implications for research and/or application, is well beyond the scope of the cur-
rent article. In any case, our findings seem to highlight that there could be much
benefit, both theoretically and practically, from exploring the utility of these various
concepts in helping to establish perspective-taking behaviors in autistic and neuro-
typical children.
Another recent account argues that perspective-taking is problem-solving (Tay-
lor & Edwards, 2021) and proposed that responding correctly to false-belief tests
involves recall, and control by stimuli that predict the behavior of others. That is,
individuals are required to recall where the book was when Maria was present in
the room, and responding is under the control of “predictive stimuli” (p. 376) such
as Maria being present in the room and location of the book. The emphasis of this
approach is on the precurrent behaviors (e.g., recall; Skinner, 1966) that are said
to facilitate perspective-taking and then bringing these behaviors under the control
of stimuli involved in a false-belief task. Other researchers (Kavanagh et al., 2020)
have also argued that key relational precursors such as a history of relating events by
coordination (i.e., sameness), distinction (i.e., difference), and temporal order (i.e.,
before/after) with some fluency may be needed for false belief tasks. Children’s per-
formance on these precursors could explain, at least in part, the variation in perfor-
mance on false belief tasks at many different ages. Some autistic children seem to
need training on derived relational responding more broadly (Gómez-Becerra et al.,
2007; see Gibbs et al., 2023, for a review), and may therefore need specific training
on these aforementioned relational precursors and precurrent behaviors to facilitate
perspective-taking (see Gould et al., 2011, for one such study on precursors).
Although the aforementioned characterizations of perspective-taking remain
mostly theoretical, these accounts are a first step towards inspiring more scien-
tific inquiry on the topic of perspective-taking from a behavior-analytic point of
view. In this regard, it may be useful to note that the concept of perspective-tak-
ing is not a technical one within behavior analysis and attempting to render it as
such may not be necessary. Rather, perhaps the research community should focus
on developing the types of technical accounts considered above (e.g., relational
triangulation, mutually entailed orientating, relating relations). These accounts
could then be assessed by determining the extent to which they provide practi-
tioners and applied researchers with robust strategies to establish firm false belief
responding in individuals who may demonstrate difficulties.
In reflecting upon the current work, the sample size of included studies in our
scoping review was predictably small and therefore we are prevented from draw-
ing strong conclusions, particularly concerning the relationship between deictic
relations and perspective-taking. Given that this was a scoping review with a
small number of included studies, we did not assess the included studies for their
experimental rigor or overall quality. Likewise, we did not undertake a meta-
analysis or any other quantitative analyses for the same reasons. We also did not
include any non-English articles in the review.
Despite these issues, our review identified key areas for future research. First,
given the small number of included studies, there seems to be a need to develop
effective training programs for establishing deictic relations that replicate the
positive outcomes reported in the existing literature. Our review identified room
for systematically assessing specific aspects of deictic training protocols (e.g.,
naturalistic scenarios vs. a list of statements), as well as testing disparate training
settings (e.g., trials conducted at a table, trials conducted in a play area). Next,
future research is needed to systematically test the impact of deictic relational
training on perspective-taking tasks. If no collateral improvement in perspec-
tive-taking is observed, then a corroborative theoretical account that supports
empirical findings is warranted. In addition, as with all behaviors, there are likely
going to be individual differences which affect participants’ responses. A huge
challenge is to fine-tune assessments and training protocols to identify these and
address them accordingly. For example, investigating the use of deictic proto-
cols with individuals diagnosed with severe autism and nonvocal forms of com-
munication is needed. Finally, further research is needed on the key relational
precursors and other patterns of relational responding which may contribute to
perspective-taking tasks (see Kavanagh et al., 2020, for an overview of potential
precursors).
In conclusion, our scoping review found that (1) individuals experiencing chal-
lenges in demonstrating deictic relational responding were trained successfully
in the studies that we reviewed; (2) there was some variation in the methods that
were used to test deictic relations and perspective-taking across studies; and (3) we
found insufficient evidence for improvements in perspective-taking following deictic
relational training. Deictic relational responding and perspective-taking have been
argued to be critical for social success and in facilitating the development of the
concept of the self in individuals. Despite the importance of this topic, we identified
a severe paucity in research studies addressing this skill (deictic relational respond-
ing and perspective-taking) in autistic individuals. We hope that our review serves
to highlight the strengths and limitations of the existing literature and inspire much-
needed research on this important topic.
Appendix 1
The three deictic relations (I–YOU, HERE–THERE, and NOW–THEN) have been
referred to as both individual relations and as frames in the RFT literature, including
in the seminal volume on RFT (Hayes et al., 2001). However, it seems more precise
to define the individual elements as relations, but combinations thereof as frames
(see Kavanagh et al., 2020; Barnes-Holmes & Harte, 2022). This is because the con-
cept of the frame typically refers to a minimal relational network that is required to
distinguish a particular class or pattern of AARRing. None of the three individual

relations, however, appear to denote a network of relations (i.e., a frame); in each
case only two elements are specified. Each of the relations may involve some form
of mutual entailment (see below), but a network of relations is not possible with-
out combining it with other relations. In the case of deictic AARRing, it is only
when the I–YOU relation is combined with one or both of the other relations that
at least some of the deictic properties of AARRing may be clearly specified. For
this reason, it seems wise to reserve the term “frame” for patterns of AARRing that
involve combining the I–YOU deictic relation with at least one of the other deictic
relations.5
Before continuing, it seems important to note that the three deictic relations may
each be considered mutually entailed based on how they are employed in natural
language interactions. With respect to I–YOU, for example, if individual A asks
individual B “How are you?”, individual B does not respond “You are fine.” Rather,
they respond “I am fine.” In this sense, a question about “YOU” entails respond-
ing with “I” (i.e., “YOU” is transformed into “I” when the listener responds as a
speaker). A similar issue of entailed transformation applies to the HERE–THERE
relation in that when individual A is at location X, X is HERE and Y is THERE;
but when the individual moves to Y, Y is HERE and X is THERE. In effect, HERE
entails THERE and THERE entails HERE as defined by an individual’s spatial loca-
tion (i.e., “THERE” is transformed into “HERE" and “HERE” is transformed into
“THERE”). Finally, a similar analysis applies to NOW–THEN in that when indi-
vidual A is at time 1, this is NOW; but when they are at time 2, this is NOW and
time 1 is THEN. As such, NOW entails THEN and THEN entails NOW as defined
by an individual’s temporal location (i.e., “NOW” is transformed into “THEN" and
“THEN” is transformed into “NOW”).
As noted above, deictic framing may be defined as responding in accordance with
I-YOU and one or both of the three mutually entailed deictic relations. For example,
if individual A asks individual B “Are you there?” individual B may reply, “Yes,
I am here.” These two statements are relationally coherent within the context of
deictic framing because, as noted above, in such a verbal interaction “you” entails
“I” and “there” entails “here” Likewise, imagine individual A asks individual B,
“If you finish the job by then, let me know”; if individual B does complete the job
by the specified time (then), they may respond at that point with “I just finished
the job now.” In this verbal interaction, “you” again entails “I” and “then” entails
5
This approach is consistent with Barnes-Holmes et al. (2017), who argued that the term deictic, as
used within RFT, refers “to verbal relations that specify an individual as located in a particular space
(e.g., “here” rather than “there”) and at a particular time (e.g., “now” rather than “then”)” (p. 161). As an
aside, the perceptual experience of being located in a particular space and time is, of course, not arbitrar-
ily applicable, as is the case with all other perceptual experiences (e.g., simply experiencing the color red
is not a verbal event in RFT). The arbitrary applicability is found in the relational responding that occurs
when humans learn to communicate with each other about their perceptual experiences by relating them
to arbitrary sounds and other symbols (i.e., having a perspective, or seeing the color red, is not arbi-
trary, but relating these perceptual events to “I”, “me”, “mine” etc., or “red,” “scarlet,” “crimson,” etc. is
arbitrarily applicable, in that there is no formal relationship, for example, between the word “I” and my
ongoing continuous perceptual experience of the world).
“now.” These examples serve to illustrate how the I–YOU deictic relation may be
combined with HERE–THERE and NOW–THEN relations to form a pattern of
relational responding that may be usefully defined as a deictic relational frame or
network. In general terms, therefore, deictic framing involves an individual locating
(or framing) themselves verbally in time and/or space relative to other individuals
(e.g., I am in Belfast at 5pm and Joao is in Bauru at 2pm). In this sense, the deictic
frame thus involves relating relations (see Kavanagh et al., 2020; Barnes-Holmes &
Harte, 2022), which renders the deictic frame somewhat more complex than other
frequently discussed nondeictic frames.6
Similar issues pertaining to the “fuzzy” nature of some of the concepts in RFT
have been noted in relatively recent developments in RFT. For example, in propos-
ing the multidimensional multilevel (MDML) framework, Barnes-Holmes et al.
(2017) noted that “there are many examples of AARRing that do not fit neatly into a
specific level of relational development” (p. 441; the levels were defined as mutual
entailing, relational framing, relational networking, relating relations, and relating
relational networks). The general point being, that as RFT continues to grapple with
the complexity of human language and cognition, some of the established concepts
and categories may need to be seen as relatively fuzzy, and in some cases adjusted or
refined, or even abandoned, in the service of increasing behavioral prediction-and-
influence with precision, scope and depth (see Sivaraman et al. (2023), who defined
listener naming as mutual entailing but speaker naming as relational framing, thus
arguing for a change in the early RFT view of naming as simply mutual entailment).
Data Availability The datasets generated during and/or analysed during the current study are available
from the corresponding author on reasonable request.
Declarations
Conflict of Interest The authors declare no conflict of interest.
References
American Psychiatric Association. (2013). Diagnostic and statistical manual of mental disorders (5th
edn. DSM-5). American Psychiatric Association.
6
Developing a “standard” experimental model of deictic relational responding using a matching-to-
sample (MTS) type-preparation would likely be very challenging (but see Guinther, 2017, 2018, for pos-
sible examples with adult participants). However, it is important to emphasize that RFT is a theoretical
account of human language and cognition, which includes explaining how language is actually used in
the natural environment. In developing this account, it is not a requirement that all conceptual analyses
must yield to a formal MTS-type experimental model. Indeed, it may be that focusing excessively on
MTS preparations will serve to narrow and distort RFT as a "grand theory” of human language and cog-
nition.
Barnes-Holmes, Y. (2001). Analysing relational frames: Studying language and cognition in young chil-
dren [Unpublished doctoral thesis, National University of Ireland, Maynooth]. NUI Campus Reposi-
tory. https://mural.maynoothuniversity.ie/16962/
Barnes-Holmes, Y., Foody, M., Barnes-Holmes, D., & McHugh, L. (2013). Advances in research on deic-
tic relations and perspective-taking. In S. Dymond & B. Roche (Eds.), Advances in relational frame
theory: Research and application (pp. 127–148). New Harbinger.
Barnes-Holmes, D., Finn, M., McEnteggart, C., & Barnes-Holmes, Y. (2017). Derived stimulus rela-
tions and their role in a behavior-analytic account of human language and cognition. Perspectives on
Behavior Science, 41(1), 155–173. https://doi.org/10.1007/s40614-017-0124-7
Barnes-Holmes, D., & Harte, C. (2022). Relational frame theory 20 years on: The Odysseus voyage and
beyond. Journal of the Experimental Analysis of Behavior, 117(2), 240–266. https://doi.org/10.
1002/jeab.733
Baron-Cohen, S. (2000). Theory of mind and autism: A review. International Review of Research in
Mental Retardation, 23, 169–184. https://doi.org/10.1016/S0074-7750(00)80010-5
Baron-Cohen, S., & Hammer, J. (1997). Is autism an extreme form of the" male brain?". Advances in
Infancy Research, 11, 193–218.
Barron, B. F., Verkuylen, L., Belisle, J., Paliliunas, D., & Dixon, M. R. (2018). Teaching “then–later” and
“here–there” relations to children with autism: An evaluation of single reversals and transforma-
tion of stimulus function. Behavior Analysis in Practice, 12(1), 167–175. https://doi.org/10.1007/
s40617-018-0216-1
Belisle, J., Dixon, M. R., Stanley, C. R., Munoz, B., & Daar, J. H. (2016). Teaching foundational perspec-
tive-taking skills to children with autism using the PEAK-T curriculum: Single-reversal “I–You”
deictic frames. Journal of Applied Behavior Analysis, 49(4), 965–969. https://doi.org/10.1002/jaba.
324
Boehm, A. E. (2000). Assessment of basic relational concepts. In B. A. Bracken (Ed.), The psychoedu-
cational assessment of preschool children (3rd ed., pp. 186–203). Lawrence Erlbaum Associates.
Carpenter, M., Nagell, K., Tomasello, M., Butterworth, G., & Moore, C. (1998). Social cognition, joint
attention, and communicative competence from 9 to 15 months of age. Monographs of the Society
for Research in Child Development, 63(4), i–174. https://doi.org/10.2307/1166214
Davis, M. H., Soderlund, T., Cole, J., Gadol, E., Kute, M., Myers, M., & Weihing, J. (2004). Cognitions
associated with attempts to empathize: How do we imagine the perspective of another? Personality
& Social Psychology Bulletin, 30(12), 1625–1635. https://doi.org/10.1177/0146167204271183
Davlin, N. L., Anne Rehfeldt, R., & Lovett, S. (2011). A relational frame theory approach to understand-
ing perspective-taking using children’s stories in typically developing children. European Journal of
Behavior Analysis, 12(2), 403–430. https://doi.org/10.1080/15021149.2011.11434392
Dixon, M. R. (2016). PEAK Relational Training System. Shawnee Scientific Press.
Dube, W. V., MacDonald, R. P., Mansfield, R. C., Holcomb, W. L., & Ahearn, W. H. (2004). Toward a
behavioral analysis of joint attention. The Behavior Analyst, 27, 197–207.
Galinsky, A. D., & Ku, G. (2004). The effects of perspective-taking on prejudice: The moderating role of
self-evaluation. Personality & Social Psychology Bulletin, 30(5), 594–604. https://doi.org/10.1177/
0146167203262802
Gibbs, A. R., Tullis, C. A., Conine, D. E., & Fulton, A. A. (2023). A systematic review of derived rela-
tional responding beyond coordination in individuals with autism and intellectual and developmen-
tal disabilities. Journal of Developmental & Physical Disabilities, 35(6), 1–36. https://doi.org/10.
1007/s10882-023-09901-z
Gilroy, S. P., Lorah, E. R., Dodge, J., & Fiorello, C. (2015). Establishing deictic repertoires in autism.
Research in Autism Spectrum Disorders, 19, 82–92. https://doi.org/10.1016/j.rasd.2015.04.004
Gómez-Becerra, I., Martín, M. J., Chávez-Brown, M., & Douglas Greer, R. (2007). Perspective-taking in
children with autism. European Journal of Behavior Analysis, 8(1), 13–28. https://doi.org/10.1080/
15021149.2007.11434270
Gould, E., Tarbox, J., O’Hora, D., Noone, S., & Bergstrom, R. (2011). Teaching children with autism a basic com-
ponent skill of perspective-taking. Behavioral Interventions, 26(1), 50–66. https://doi.org/10.1002/bin.320
Guinther, P. M. (2017). Contextual influence over deriving others’ true beliefs using a relational triangulation
perspective-taking protocol (RT-PTP-M1). Journal of the Experimental Analysis of Behavior, 108(3),
433–456. https://doi.org/10.1002/jeab.291
Guinther, P. M. (2018). Contextual influence over deriving another’s false beliefs using a relational trian-
gulation perspective taking protocol (RT-PTP-M2). Journal of the Experimental Analysis of Behavior,
110(3), 500–521. https://doi.org/10.1002/jeab.480
Hayes, S. C. (1984). Making sense of spirituality. Behaviorism, 12(2), 99–110.

Hayes, S. C., Barnes-Holmes, D., & Roche, B. (2001). Relational frame theory: A précis. In S. C. Hayes, D.
Barnes-Holmes, & B. Roche (Eds.), Relational frame theory: A post-Skinnerian account of human lan-
guage and cognition (pp. 141–154). Springer. https://doi.org/10.1007/0-306-47638-X_8
Heagle, A. I., & Rehfeldt, R. A. (2006). Teaching perspective-taking skills to typically developing children
through derived relational responding. Journal of Early & Intensive Behavior Intervention, 3(1), 1–34.
https://doi.org/10.1037/h0100321
Howlin, P., Baron-Cohen, S., & Hadwin, J. A. (1999). Teaching children with autism to mind-read: A practi-
cal guide for teachers and parents. John Wiley & Sons.
Hutchins, T. L., Prelock, P. A., & Bonazinga, L. (2012). Psychometric evaluation of the Theory of Mind
Inventory (ToMI): A study of typically developing children and children with autism spectrum dis-
order. Journal of Autism & Developmental Disorders, 42(3), 327–341. https://doi.org/10.1007/
s10803-011-1244-7
Hughes, S., & Barnes-Holmes, D. (2015). Relational frame theory: The basic account. In R. D. Zettle, S. C.
Hayes, D. Barnes-Holmes, & A. Biglan (Eds.), The Wiley handbook of contextual behavioral science
(pp. 129–178). John Wiley & Songs. https://doi.org/10.1002/9781118489857.ch9
Hughes, C., & Leekam, S. (2004). What are the links between theory of mind and social relations? Review,
reflections and new directions for studies of typical and atypical development. Social Development,
13(4), 590–619. https://doi.org/10.1111/j.1467-9507.2004.00285.x
Jackson, M. L., Mendoza, D. R., & Adams, A. N. (2014). Teaching a deictic relational repertoire to children
with autism. The Psychological Record, 64(4), 791–802. https://doi.org/10.1007/s40732-014-0078-z
Kanter, J. W., Parker, C. R., & Kohlenberg, R. J. (2001). Finding the self: A behavioral measure and its clini-
cal implications. Psychotherapy: Theory, Research, Practice, Training, 38(2), 198–211. https://doi.org/
10.1037/0033-3204.38.2.198
Kavanagh, D., Barnes-Holmes, Y., & Barnes-Holmes, D. (2020). The study of perspective-taking: Contri-
butions from mainstream psychology and behavior analysis. The Psychological Record, 70, 581–604.
https://doi.org/10.1007/s40732-019-00356-3
Lerner, M. D., Hutchins, T. L., & Prelock, P. A. (2011). Brief report: Preliminary evaluation of the theory
of mind inventory and its relationship to measures of social skills. Journal of Autism & Developmental
Disorders, 41, 512–517. https://doi.org/10.1007/s10803-010-1066-z
Loukusa, S., Mäkinen, L., Kuusikko-Gauffin, S., Ebeling, H., & Moilanen, I. (2014). Theory of mind and
emotion recognition skills in children with specific language impairment, autism spectrum disorder
and typical development: Group differences and connection to knowledge of grammatical morphology,
word-finding abilities and verbal working memory. International Journal of Language & Communica-
tion Disorders, 49(4), 498–507. https://doi.org/10.1111/1460-6984.12091
Lovett, S., & Rehfeldt, R. A. (2014). An evaluation of multiple exemplar instruction to teach perspective-
taking skills to adolescents with Asperger syndrome. Behavioral Development Bulletin, 19(2), 22–36.
https://doi.org/10.1037/h0100575
McHugh, L., Barnes-Holmes, Y., & Barnes-Holmes, D. (2004a). A relational frame account of the develop-
ment of complex cognitive phenomena: Perspective-taking, false belief understanding, and deception.
International Journal of Psychology & Psychological Therapy, 4(2), 303–324.
McHugh, L., Barnes-Holmes, Y., & Barnes-Holmes, D. (2004b). Perspective-taking as relational responding:
A developmental profile. The Psychological Record, 54(1), 115–144. https://doi.org/10.1007/BF033
95465
Mead, G. H. (1934). Mind, self, and society from the standpoint of a social behaviorist. University of Chi-
cago Press.
Montoya-Rodríguez, M. M., & Molina Cobos, F. J. (2016). Relationship between deictic relational respond-
ing and theory of mind tasks in children: A pilot study. The Psychological Record, 66, 573–587. https://
doi.org/10.1007/s40732-016-0193-0
O’Neill, J., & Weil, T. M. (2014). Training deictic relational responding in people diagnosed with schizo-
phrenia. The Psychological Record, 64(2), 301–310. https://doi.org/10.1007/s40732-014-0005-3
Ozonoff, S., & Miller, J. N. (1995). Teaching theory of mind: A new approach to social skills training for
individuals with autism. Journal of Autism & Developmental Disorders, 25(4), 415–433. https://doi.
org/10.1007/BF02179376
Padilla, K. L., Weston, R., Morgan, G. B., Lively, P., & O’Guinn, N. (2023). Validity and reliability evidence
for assessments based in applied behavior analysis: A systematic review. Behavior Modification, 47(1),
247–288. https://doi.org/10.1177/01454455221098151
Pearson, A., Ropar, D., & de C Hamilton, A. F. (2013). A review of visual perspective-taking in autism
spectrum disorder. Frontiers in human neuroscience, 7, 652. https://doi.org/10.3389/fnhum.2013.00652
Piaget, J. (1984). The attainment of invariants and reversible operations in the development of thinking.
Social Research, 51(1–2), 167–184.http://www.jstor.org/stable/40970935
Rehfeldt, R. A., Dillen, J. E., Ziomek, M. M., & Kowalchuk, R. K. (2007). Assessing relational learning defi-
cits in perspective-taking in children with high-functioning autism spectrum disorder. The Psychologi-
cal Record, 57(1), 23–47. https://doi.org/10.1007/BF03395563
Schlinger, H. D. (2009). Theory of mind: An overview and behavioral perspective. The Psychological
Record, 59, 435–448. https://doi.org/10.1007/BF03395673
Schreibman, L. (2000). Intensive behavioral/psychoeducational treatments for autism: Research needs and
future directions. Journal of Autism & Developmental Disorders, 30(5), 373–378. https://doi.org/10.
1023/A:1005535120023
Schwartz Offek, E., & Segal, O. (2022). Comparing theory of mind development in children with autism
spectrum disorder, developmental language disorder, and typical development. Neuropsychiatric Dis-
ease & Treatment, 2002, 2349–2359. https://doi.org/10.2147/NDT.S331988
Sigman, M., & Capps, L. (1997). Children with autism: A developmental perspective. Harvard University Press.
Sivaraman, M., Barnes-Holmes, D., Greer, R. D., Fienup, D. M., & Roeyers, H. (2023). Verbal behavior
development theory and relational frame theory: Reflecting on similarities and differences. Journal of
the Experimental Analysis of Behavior, 119(3), 539–553. https://doi.org/10.1002/jeab.836
Skinner, B. F. (1945). The operational analysis of psychological terms. Psychological Review, 52(5), 270–
277. https://doi.org/10.1037/h0062535
Skinner, B. F. (1966). What is the experimental analysis of behavior? Journal of the Experimental Analysis of
Behavior, 9(3), 213. https://doi.org/10.1901/jeab.1966.9-213
Skinner, B. F. (1974). About behaviorism. Knopf.
Spradlin, J. E., & Brady, N. (2008). A behavior analytic interpretation of theory of mind. Revista internacional de
psicologia y terapia psicologica/International Journal of Psychology & Psychological Therapy, 8(3), 335.
Taylor, T., & Edwards, T. L. (2021). What can we learn by treating perspective-taking as problem solving?
Perspectives on Behavior Science, 44, 359–387. https://doi.org/10.1007/s40614-021-00307-w
Underwood, B., & Moore, B. (1982). Perspective-taking and altruism. Psychological Bulletin, 91(1), 143.
https://doi.org/10.1037/0033-2909.91.1.143
Vescio, T. K., Sechrist, G. B., & Paolucci, M. P. (2003). Perspective-taking and prejudice reduction: The
mediational role of empathy arousal and situational attributions. European Journal of Social Psychol-
ogy, 33(4), 455–472. https://doi.org/10.1002/ejsp.163
Weil, T. M., Hayes, S. C., & Capurro, P. (2011). Establishing a deictic relational repertoire in young children.
The Psychological Record, 61, 371–390. https://doi.org/10.1007/BF03395767
Wimmer, H., & Perner, J. (1983). Beliefs about beliefs: Representation and constraining function of wrong
beliefs in young children’s understanding of deception. Cognition, 13(1), 103–128. https://doi.org/10.
1016/0010-0277(83)90004-5
Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and
institutional affiliations.
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under
a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted
manuscript version of this article is solely governed by the terms of such publishing agreement and
applicable law.

Deictic Relational Responding and Perspective Taking in Autistic Individuals: A Scoping Review

Uploaded by

Copyright:

Available Formats

Deictic Relational Responding and Perspective Taking in Autistic Individuals: A Scoping Review

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Deictic Relational Responding and Perspective Taking in Autistic Individuals: A Scoping Review

Uploaded by

Copyright:

Available Formats

Perspectives on Behavior Science (2024) 47:107–137

Deictic Relational Responding and Perspective‑Taking

Nadine Hempkin1,2 · Maithri Sivaraman3 · Dermot Barnes‑Holmes2

Accepted: 12 January 2024 / Published online: 6 February 2024

Keywords Deictic relational responding · Perspective-taking · Relational frame

Perspective-taking is deemed an important area of study both within mainstream

The Mainstream Psychology Approach to Perspective‑Taking

The Current Review

A small number of behavior-analytic studies have attempted to measure perspective-

the literature, we identified significant variation in how studies assessed perspective-

We undertook a scoping review of studies that conducted empirical investigations on

Outcome Measures and Data Extraction

Fig. 1 Reference attrition process

Methods Used to Test for Deictic Relations

Study Aim Diagnosis Total number of participants Age Gender

as the participants acquired the

Study Aim Diagnosis Total number of participants Age Gender

N/A = Not Applicable

Table 3 Outcomes of training

N/A = Not Applicable

Methods Used to Test for Perspective‑Taking using Traditional

Methods Used to Train Deictic Relations

Deictic Relations Trained

Procedure for Training Deictic Relations

Evidence for Relationship between Deictic Relational Responding

We conducted a scoping review of the literature on deictic relational responding

In addition, perspective-taking measures such as the Sally-Anne task involve a

distinguish a particular class or pattern of AARRing. None of the three individual

Hayes, S. C. (1984). Making sense of spirituality. Behaviorism, 12(2), 99–110.

You might also like