JOURNAL OF APPLIED BEHAVIOR ANALYSIS 2011, 44, 859–876 NUMBER 4 (WINTER 2011)

A REVIEW OF RECOMMENDATIONS FOR SEQUENCING RECEPTIVE

AND EXPRESSIVE LANGUAGE INSTRUCTION
ANNA INGEBORG PETURSDOTTIR
TEXAS CHRISTIAN UNIVERSITY

AND

JAMES E. CARR
AUBURN UNIVERSITY

We review recommendations for sequencing instruction in receptive and expressive language

objectives in early and intensive behavioral intervention (EIBI) programs. Several books
recommend completing receptive protocols before introducing corresponding expressive
protocols. However, this recommendation has little empirical support, and some evidence
exists that the reverse sequence may be more efficient. Alternative recommendations include
teaching receptive and expressive skills simultaneously (M. L. Sundberg & Partington, 1998) and
building learning histories that lead to acquisition of receptive and expressive skills without direct
instruction (Greer & Ross, 2008). Empirical support for these recommendations also is limited.
Future research should assess the relative efficiency of receptive-before-expressive, expressive-
before-receptive, and simultaneous training with children who have diagnoses of autism
spectrum disorders. In addition, further evaluation is needed of the potential benefits of
multiple-exemplar training and other variables that may influence the efficiency of receptive and
expressive instruction.
Key words: autism treatment, expressive language, language intervention, listener behavior,
receptive language, tacts, verbal behavior
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _2

Early and intensive behavioral intervention 1998). Each of these books is to a greater or
(EIBI) for children diagnosed with autism lesser extent influenced by a curriculum that
spectrum disorders enjoys substantial empirical was originally developed by O. Ivar Lovaas and
support (for meta-analyses and reviews, see his colleagues at the University of California at
Eldevik et al., 2009; Virués-Ortega, 2010). In Los Angeles (UCLA) and described in detail in
the last 15 years, a number of books and early publications (Lovaas, 1977, 1981). Many
intervention manuals have been published that similarities exist across books. For example, all
describe behaviorally based procedures, and in recommend breaking down complex language
some cases curricula, for teaching various skills skills into basic components, teaching these
in the context of EIBI (Barbera, 2007; Greer skills in a structured format, and using positive
& Ross, 2008; Leaf & McEachin, 1999; reinforcement, discrimination training, and
Lovaas, 2003; Maurice, Green, & Luce, 1996; transfer-of-stimulus-control procedures. How-
Schramm, 2006; M. L. Sundberg & Partington, ever, it may be useful to distinguish between
two general categories of published work of this
A portion of this article is based on a dissertation nature based on subsequent influences: work
submitted by the first author, under the supervision of the that follows the UCLA model, and work that
second author, to the Graduate College at Western
Michigan University in partial fulfillment of the require-
follows the verbal behavior model. UCLA-
ments for the PhD degree in psychology. model publications include the curriculum
Address correspondence to Anna Ingeborg Petursdottir, chapter in Maurice et al. (Taylor & McDo-
Department of Psychology, Texas Christian University, Box nough, 1996), the curriculum manual by Leaf
298920, Fort Worth, Texas 76129 (e-mail: a.petursdottir@
tcu.edu). and McEachin, and the book by Lovaas (2003),
doi: 10.1901/jaba.2011.44-859 all of which address a wide range of language,

859
860 ANNA INGEBORG PETURSDOTTIR and JAMES E. CARR

social, self-help, and preacademic skills. De- recommendation contained therein is evidence
spite updates and influences by more recent based. As Lovaas (2003) pointed out, a
behavior-analytic work, the content remains number of specific recommendations made
fairly close to Lovaas’s (1977, 1981) original in UCLA-model books remain to be investi-
curriculum. Publications that follow the verbal gated, and the same may be true of various
behavior model (Barbera, 2007; Greer & Ross, recommendations that have been introduced
2008; Schramm, 2006; M. L. Sundberg & by verbal behavior model books. In some
Partington, 1998), by contrast, focus primarily cases, different publications make conflicting
on language and communication, and use recommendations, and in these cases, it is
Skinner’s (1957) analysis of verbal behavior important to examine the empirical evidence
as a starting point for program design. In behind each recommendation.
addition, the Greer and Ross (2008) book is In this article, we review recommendations
influenced by current theory and research on regarding the sequencing of language instruc-
derived stimulus relations. These influences are tion protocols that are termed expressive and
reflected in the terminology used in those receptive in the UCLA-model publications.
books, as well as in certain recommendations Expressive protocols are generally those that aim
that concern teaching strategies and curriculum to establish spoken responses by the child
objectives. (however, they can be adapted for use with
A recent survey of EIBI providers (Love, alternative communication systems), whereas in
Carr, Almason, & Petursdottir, 2009) suggests receptive protocols, a child responds non-
that publications that describe both UCLA- vocally to a teacher’s spoken instructions. In
model and verbal behavior curricula are widely the verbal behavior model publications, the
used in EIBI programs. This is not surprising, term expressive is replaced with Skinner’s (1957)
given that books published in both traditions verbal operant taxonomy of speaker behavior.
contain a wealth of helpful information, Instructional targets are described as mands,
recommendations, and suggestions that are tacts, or intraverbals. Greer and Ross (2008) also
based on extensive clinical experience and favor the term listener behavior or listener
empirical data. As a whole, the UCLA model training over receptive language or receptive
has a great deal of empirical support, in that it instruction.
has been used in a number of outcome studies A common recommendation in UCLA-
on the effects of EIBI on intellectual function- model books (Leaf & McEachin, 1999; Lovaas,
ing and other measures (e.g., Lovaas, 1987; 2003; Taylor & McDonough, 1996) is to
Sallows & Graupner, 2005; Smith, Groen, & complete receptive language protocols before
Wynn, 2000). By contrast, no large-scale introducing expressive protocols that address
outcome studies have used curricula derived the same instructional targets. For example,
from the verbal behavior model. This model, before a child is taught to say color names in the
however, has indirect support from research on presence of color stimuli, it is recommended
certain curriculum objectives that depart from that the child be taught to point to the
the UCLA model (see J. E. Carr & Firth, appropriate color stimuli when presented with
2005). Both categories of work also draw on spoken color names. The verbal behavior model
decades of behavior-analytic research on acqui- books, by contrast, typically do not advocate
sition procedures. this sequence. To the extent that these books
Although the content of EIBI publications make sequence recommendations, they vary
is thus firmly grounded in empirical research, depending on the child’s level of verbal
practitioners should not assume that every functioning and the verbal operant that is
 LANGUAGE INSTRUCTION 861

targeted in expressive training. The programs receptive programs can be completed concur-
that these recommendations encompass com- rently with early vocal imitation training.
prise a large proportion of a child’s language Second, receptive training may be easier to
curriculum (e.g., Taylor & McDonough, complete with a child who demonstrates non-
1996). As a result, a practitioner’s decision as compliance, due to the ease of physical prompt-
to which sequence to follow may affect a ing. Third, receptive training may be more
substantial portion of a child’s instructional amenable to the use of errorless training pro-
programming. cedures (e.g., stimulus fading) that enhance an
In the sections that follow, we first review early learner’s success. It is important to note,
empirical evidence of relevance to the receptive- however, that UCLA-model publications con-
before-expressive recommendation. We then tinue to recommend the receptive-before-
review alternative recommendations that have expressive sequence for advanced programs that
been made in a subset of the verbal behavior typically are implemented when the child has
model books (Greer & Ross, 2008; M. L. acquired a basic verbal repertoire, instructional
Sundberg & Partington, 1998) and, finally, control has been established, and it can be
discuss directions for future research. assumed that the child has become a fairly
efficient learner. For example, the recommenda-
THE RECEPTIVE-BEFORE- tion is made for teaching prepositions, cate-
EXPRESSIVE SEQUENCE gories, emotions, letters, and numbers (Leaf &
McEachin, 1999; Taylor & McDonough, 1996).
According to Lovaas (1977), the receptive- The extent to which this sequence enhances
before-expressive recommendation was original- acquisition, compared to the reverse sequence or
ly adopted because it reflected the sequence in other alternatives, needs to be examined.
which typically developing children had been
observed to acquire receptive and expressive Empirical Findings
repertoires. However, Lovaas (1977, 2003) also Although a large body of research exists on
cautioned that a typical developmental sequence teaching receptive and expressive language skills
may not necessarily be the ideal teaching to individuals who have been diagnosed with
sequence, and the sequencing of programs autism or other developmental disabilities, few
ultimately should be based on empirical data studies have attempted to compare directly
on teaching efficiency. In addition, Lovaas the receptive-before-expressive sequence with
(2003) suggested that the reverse sequence other alternatives. We identified nine pub-
may be more beneficial for some children lished studies (Table 1) with clinically relevant
(p. 107), and Leaf and McEachin (1999) stated participants that were designed for the pur-
that ‘‘it is not always necessary to wait for the pose of either comparing (a) a receptive-before-
mastery of receptive labeling before starting expressive sequence to expressive-before-receptive
expressive’’ (p. 209). In spite of these caveats, all or expressive-only training or (b) the efficiency
three UCLA-model publications consistently of receptive and expressive training as an initial
list completion of receptive programs as training condition. The nine studies are meth-
suggested prerequisites for expressive programs. odologically heterogeneous, and a thorough
In the earliest stages of language intervention, description of the specific procedures and
the receptive-before-expressive sequence may be experimental designs employed in each one is
a practical necessity for several reasons. First, beyond the scope of this article. In this section,
it is difficult to complete programs that require we therefore provide only a brief overview of
vocal responses before a child has acquired a methodological similarities and differences, fol-
vocal imitation repertoire, whereas a variety of lowed by a summary of results. In the section that
862 ANNA INGEBORG PETURSDOTTIR and JAMES E. CARR
Table 1
Empirical Evaluations of the Receptive-Before-Expressive Recommendation

Study N Age in years Participant description Summary of results

Cuvo and Riva 20 (a) M 5 16 years (a) Mean IQ 5 44; mean Expressive training took fewer trials than
(1980) 2 months MA 5 5 years, 2 monthsa receptive training.
(b) M 5 4 years (b) mean IQ 5 108; mean Prior receptive training facilitated subsequent
4 months MA 5 5 years 1 montha expressive training, but prior expressive
training rendered subsequent receptive
training unnecessary.
The receptive-before-expressive sequence took
more time and trials than when expressive
training came first.
Similar maintenance in both conditions.
Hupp et al. (1986) 12 5 to 19 Severe mental retardation; Acquisition similar in receptive and expressive
Stage 6 object permanence; conditions, but training did not continue to
matched 2-D to 3-D mastery in either condition.
stimuli; no vocal or Generalization to receptive responding with
signed language; no novel stimuli greater after receptive than
augmentative system expressive training.
Keller and Bucher 6 Children Developmentally disabled; Greater accuracy on receptive tests following
(1979) severe language deficits expressive training than on expressive tests
following receptive training.
Keller and Bucher 5 5 to 10 Borderline to moderate Prior receptive training did not facilitate
(1980) mental retardation; subsequent expressive training.
rudimentary receptive The receptive-before-expressive sequence took
and expressive skills more time and trials than when expressive
training came first.
Miller et al. (1977) 14 M 5 12.67 Mean IQ 5 43.86; mean Prior receptive training did not facilitate
MA 5 4.83 yearsb; subsequent expressive training.
kindergarten arithmetic The receptive-before-expressive sequence took
grade level more time and trials than when expressive
training came first.
Smeets (1978) 1 14 IQ 44c, deaf, mild cerebral Expressive training completed in fewer trials
palsy, blindness in one than receptive training, but only with the
eye; .100 receptive and first few stimulus sets.
expressive labels
Smeets and Striefel 1d 14 IQ 44 c, deaf, mild cerebral Expressive training had a greater effect on
(1976) palsy, blindness in one eye; receptive responding than receptive training
some receptive signing; on expressive responding.
expressive signing rarely Greater accuracy on receptive tests following
observed expressive training than on expressive tests
after receptive training.
Watters et al. 4 10 to 16 Autism; IQ , 45 to 80e; one Prior receptive training did not facilitate
(1981) participant untestable; subsequent expressive training.
existing receptive and The receptive-before-expressive sequence took
expressive vocabulary more time and trials than when expressive
32 to192 words training came first.
Greater accuracy on receptive tests following
expressive training than on expressive tests
following receptive training.
Wynn and Smith 6 4 years 3 months Autism; comprehension age 13 to Variable results within and across participants,
(2003) to 6 years 42 monthsf; expressive language but overall greater accuracy on receptive
4 months age 16 to 42 months f; tests following expressive training than on
vocal-verbal skills ranged from expressive tests following receptive training
one-word requests to four-word
phrases
a
MA 5 mental age; IQ and MA obtained from Peabody Picture Vocabulary Test.
b
Psychological Scales of Infant Development.
c
Leiter.
d
The same participant as in Smeets (1978).
e
WISC/R Performance Scale.
f
Reynell Developmental Language Scales.
 LANGUAGE INSTRUCTION 863

follows, we discuss our conclusions, along with tional sequence) 3 2 (disability) factorial design
collective limitations of the nine studies. We in addition to a multiple baseline design. Miller
refer the interested reader to the original sources et al. (1977) and Hupp et al. also employed
for more detailed information, and encourage between-subjects designs to compare instruc-
readers to evaluate the results of each study in tional conditions. The remaining studies em-
light of their respective methodological ployed within-subject or single-case design
strengths and weaknesses. It should be noted strategies, in which receptive and expressive
that two of the studies (Keller & Bucher, 1979; training to mastery were alternated across
Miller, Cuvo, & Borakove, 1977) were pub- stimuli or stimulus sets, and each condition
lished only as abstracts; thus, minimal proce- was followed by assessment and sometimes
dural information is available. training of the alternate repertoire. At least one
Overview of methods. The participants in two of these studies (Wynn & Smith, 2003) may
studies (Watters, Wheeler, & Watters, 1981; have employed too few instances of training in
Wynn & Smith, 2003) were described as each condition to yield a convincing within-
children with autism, whereas the remaining subject comparison. The results of this study
studies were conducted with children and should therefore be interpreted with caution.
young adults with intellectual disabilities, and Participants in three studies (Smeets, 1978;
no autism diagnoses were specified. As shown in Smeets & Striefel, 1976; Watters et al., 1981)
Table 1, participants were described as func- received instruction in manual signing, whereas
tioning in the moderate to mild range of participants in the remaining studies were
intellectual disability. Most studies provided taught vocal communication. A common
minimal information on existing verbal reper- feature of all nine studies was that the receptive
toires. However, participants generally appear training condition involved establishing con-
to have entered the studies with existing, trol by experimenter-produced signs or vocal-
although in some cases small (e.g., Watters izations over stimulus-selection responses. That
et al., 1981; Wynn & Smith, 2003), receptive is, given a sign or a spoken word, the child
and expressive repertoires. An exception was the selected a visual stimulus from an array of two
study by Hupp, Mervis, Able, and Conroy- or more available stimuli. The expressive
Gunter (1986), in which participants func- training condition typically involved teaching
tioned in the severe range of intellectual the child to emit signs or vocalizations in the
disability and were described as having very presence of visual stimuli. The visual stimuli
limited or no communication skills. Dependent were in all cases three-dimensional objects or
variables included acquisition of the target two-dimensional pictures of objects, and the
receptive and expressive repertoires in terms of resulting repertoires appear to meet Skinner’s
trials to mastery, emergence of one repertoire (1957) definition of the tact as a verbal
following instruction in the other (i.e., effects of response controlled by a nonverbal stimulus.
receptive training on untrained expressive skills An exception is the study by Watters et al.
and vice versa), stimulus generalization, and (1981) in which the participant was presented
maintenance. Experimental designs and other with a picture and a spoken word simulta-
methodological features varied across studies. At neously in expressive training. As a result, the
one extreme, two studies (Smeets, 1978; Smeets signed response could have been under the
& Striefel, 1976) were conducted with a single control of either stimulus or both of them. In
participant, who was the same in both studies. receptive training, the participant similarly was
At the other extreme, Cuvo and Riva’s (1980) presented with both a sign and a spoken word
study included a comparison group of typically in each trial, but stimulus control analyses
developing children and employed a 2 (instruc- revealed that the resulting selection responses
864 ANNA INGEBORG PETURSDOTTIR and JAMES E. CARR

were primarily under the control of the sign establish both expressive and receptive reper-
rather than the spoken word. toires in each condition. Four studies included
The objective of training in most cases was to such analyses. First, Cuvo and Riva (1980)
teach object names; however, in two cases the found that both groups of participants required
objective was to teach coin names or coin values significantly fewer trials to complete expressive
(Cuvo & Riva, 1980; Miller et al.,1977). In Wynn training if they had previously received receptive
and Smith’s (2003) study, four of six participants training than if they had not. Although this
were taught object attributes (i.e., abstract stimulus finding suggested a facilitative effect of receptive
properties shared by a number of training training on expressive training, the facilitative
exemplars, such as long and short). In the study effects of expressive training on receptive
by Smeets (1978), the objective of instruction was training were even greater, because participants
the plural suffix, because object names had been who received expressive training first performed
evaluated with the same participant in a prior study with high accuracy on subsequent receptive
(Smeets & Striefel, 1976). tests, rendering receptive training unnecessary.
Trials to mastery. Four studies included data As a result, the total number of trials required to
on trials to mastery in receptive training when establish both expressive and receptive reper-
there was no prior expressive training with the toires was greater for participants who received
same stimuli, and expressive training when there receptive training first than for those who
was no prior receptive training with the same received expressive training without prior
stimuli. These data provide information on the receptive training. In addition, the receptive-
ease of implementing receptive and expressive before-expressive sequence took twice as many
training as an initial training condition. Cuvo hours to complete as did expressive-only
and Riva (1980) reported a substantially greater training. In contrast to Cuvo and Riva, Keller
mean number of training trials in the initial and Bucher (1980), Miller et al. (1977), and
receptive than in the initial expressive training Watters et al. (1981) found that prior receptive
condition for participants with and without training yielded no savings of expressive training
intellectual disabilities. The participant in trials. In fact, some participants in Watters et al.
Smeets (1978) also required more receptive and one of Keller and Bucher’s participants
than expressive trials to reach criterion with the required more expressive training trials when
first few stimulus sets trained in each condition, they had previously completed receptive train-
but not with subsequent sets. By contrast, two ing than when they had not. All three studies
of the four participants in Watters et al. (1981) were in agreement with Cuvo and Riva that (a)
and three of the four in Keller and Bucher trials to mastery for receptive and expressive
(1980) required substantially more expressive training combined were overall greater when
than receptive trials to reach the mastery participants received receptive training first than
criterion, whereas the remaining participants when they received expressive training first, and
required either a similar number of trials in (b) conducting expressive training first rendered
both conditions or a greater number of trials in receptive training unnecessary.
the receptive condition. Overall, the results did No other studies reported analyses of trials to
not suggest that one training condition was mastery. Sessions to mastery can be derived
consistently more difficult to complete as an from the figures in Wynn and Smith (2003),
initial training condition than the other. but the number of trials was variable across
Of perhaps greater importance are analyses of sessions. The only other measure of acquisition
the facilitative effects of prior receptive training during training was provided by Hupp et al.
on subsequent expressive training, and vice (1986). In this study, receptive and expressive
versa, and the number of trials required to training were not conducted to mastery;
 LANGUAGE INSTRUCTION 865

instead, performance was assessed after a fixed on receptive probes following expressive training
number of training sessions in each condition. across a large number of stimuli, whereas
At that time, accuracy was low in both receptive training rarely produced correct expres-
conditions, and the difference between the sive responses. In the study by Smeets (1978), by
two was not statistically significant; however, contrast, the two training conditions had no
statistical power may have been questionable differential effects for the same participant, in
due to the small number of subjects. that both receptive and expressive training
Emergence of the alternate repertoire. Instead of yielded highly accurate performance on tests of
or in addition to reporting trials to mastery, the untrained repertoire.
several studies evaluated the efficiency of recep- Generalization and maintenance. Few studies
tive and expressive training as an initial training included measures of generalization or mainte-
condition by comparing correct expressive nance of the trained or emergent receptive and
responses after receptive training to correct expressive repertoires. Smeets (1978) assessed
receptive responses after expressive training. generalization of the emergent repertoire to
Although not as informative as trials to criterion, novel exemplars after each instance of receptive
this measure provides an indirect estimate of the or expressive training. Generalization increased
need for training in the untrained condition. across successively trained stimulus sets in both
As previously noted, Cuvo and Riva (1980), conditions. In the Hupp et al. (1986) study,
Keller and Bucher (1980), Miller et al. (1977), receptive generalization was the primary depen-
and Watters et al. (1981) found that expressive dent variable. Thus, generalization of the
training rendered subsequent receptive training trained repertoire to novel stimuli was assessed
unnecessary, because the receptive repertoire following receptive training, whereas emer-
usually emerged to criterion following expressive gence of the alternate repertoire was tested
training. After receptive training, by contrast, with novel exemplars following expressive
expressive training was still necessary in all four training. These researchers found significantly
studies. Of the five remaining studies listed in greater receptive generalization after receptive
Table 1, all except one (Hupp et al., 1986) also than expressive training. However, perfor-
evaluated the effects of expressive training on mance was poor in both conditions, which is
receptive responding, and vice versa. The results not surprising given that training did not
of three studies consistently showed that expres- continue to mastery. Only Cuvo and Riva
sive training was more likely to generate receptive (1980) provided maintenance data. Both expres-
sive and receptive repertoires were maintained at
identification to criterion than receptive training
1- and 4-week follow-ups, regardless of the
was to generate expressive responding. Keller and
training condition.
Bucher (1979) reported criterion performance
on receptive tests following expressive training Conclusion
for five of six participants, but reported criterion None of the nine studies listed in Table 1
performance on expressive tests following recep- provide support for the receptive-before-expres-
tive training for only one participant. In Wynn sive recommendation. If anything, these studies
and Smith’s (2003) study, there was substantial collectively suggest that, when possible, the
variation within and across participants, but 8 of reverse sequence may be more efficient. Specif-
10 instances of expressive training and 4 of 10 ically, it appears that expressive training may
instances of receptive training resulted in reduce or eliminate need for receptive training
criterion performance on tests of the alternate with the same targets (Cuvo & Riva, 1980;
repertoire. Smeets and Striefel’s (1976) partici- Keller & Bucher, 1979, 1980; Miller et al.,
pant consistently responded with high accuracy 1977; Smeets & Striefel, 1976; Watters et al.,
866 ANNA INGEBORG PETURSDOTTIR and JAMES E. CARR

1981) and conducting receptive training first because chance responding may inflate correct
may simply increase the overall number of trials responses on the former, particularly if a very
required to establish both repertoires (Cuvo & small stimulus array (e.g., two or three stimuli)
Riva, 1980; Keller & Bucher, 1980; Miller is provided from which to select in the receptive
et al., 1977; Watters et al., 1981). In light of test condition. As a result, comparisons might
these conclusions, it may be surprising that be biased toward showing a greater effect of
the receptive-before-expressive recommendation expressive training on receptive responding than
has survived. However, several limitations of the vice versa. In addition, a high probability of
small existing literature must be highlighted. correct responding by chance during receptive
First, with the exception of Wynn and Smith training could increase the probability of false
(2003), all of the studies in Table 1 were mastery in this condition, creating a further bias
published 20 to 35 years ago. It is possible that against expressive test performance. However,
advances in stimulus control technology and results of studies that employed five or more
reinforcer identification might result in differ- comparison stimuli in the receptive condition
ent outcomes if these studies were replicated (Cuvo & Riva, 1980; Smeets & Striefel, 1976;
today. Second, we were able to locate only two Watters et al., 1981) are consistent with results
studies (Watters et al., 1981; Wynn & Smith, of studies that employed smaller arrays, sug-
2003) in which the participants were described gesting that small arrays did not necessarily
as children with autism. This is a potential skew the results.
In addition to the studies listed in Table 1, a
limitation, given that the receptive-before-
few studies (Connell & McReynolds, 1981;
expressive recommendation has been made in
Holdgrafer & McReynolds, 1975; Petursdottir,
publications that primarily focus on EIBI for
Carr, Lechago, & Almason, 2008; Petursdottir &
this population. In addition, the participants in
Haflidadottir, 2009; Williams & McReynolds,
many studies were substantially older than
1975) have evaluated the effects of expressive and
children typically enrolled in EIBI programs.
receptive training on acquisition of the alternate
Third, few studies provided detailed informa- repertoire in young, typically developing chil-
tion on the participants’ verbal repertoires (e.g,. dren. Holdgrafer and McReynolds (1975) found
existing receptive and expressive skills) at entry inconsistent effects, and Petursdottir et al. (2008)
into the study. As a result, it is not possible to found minimal effects of either training condi-
make any statements, for example, about tion on the alternate repertoire. Results of the
characteristics of children for whom expressive remaining studies, however, were in agreement
training may render receptive training unneces- that receptive repertoires were more likely to
sary. Fourth, all nine studies, with the possible emerge after expressive training than expressive
exception of Watters et al. (1981), addressed repertoires after receptive training. In addition,
tacts of visual stimuli in Skinner’s (1957) Connell and McReynolds (1981) found that
terminology. However, many programs labeled expressive training was more likely than recep-
expressive in EIBI curricula address other types tive training to yield derived expressive and
of relations, such as tacts of stimuli presented in receptive relations. Thus, studies conducted
other sense modalities, mands for reinforcers, with typically developing participants have
and intraverbal responses to spoken words. A yielded results that are consistent with results
fifth limitation is present in studies that com- from studies in which the participants had
pared effects of receptive and expressive training developmental disabilities. This research, how-
on the emergence of the alternative repertoire. ever, suffers from the same limitations as
This limitation consists of difficulty comparing studies conducted with clinically relevant par-
performance on receptive and expressive tests, ticipants, including small stimulus arrays in the
 LANGUAGE INSTRUCTION 867

receptive condition and an emphasis on tacts of of the speaker and the behavior of the listener
visual stimuli, with the exception of two are functionally independent of one another. As
studies that targeted intraverbals in addition a result, they state that ‘‘there is no guarantee
to or instead of tacts (Petursdottir et al., 2008; that by teaching a tact response that a receptive
Petursdottir & Haflidadottir, 2009). response will emerge, or vice versa’’ (p. 158)
and emphasize that each must be established
ALTERNATIVE directly. M. L. Sundberg and Partington make a
SEQUENCE RECOMMENDATIONS clear distinction between expressive programs
that aim to establish three different verbal
The recommendation to teach receptive- operants; mands, tacts, and intraverbals. Thus,
before-expressive targets does not appear to be the question of when to teach expressive skills,
supported by the available evidence. Thus, relative to corresponding receptive targets, is
alternative recommendations that have ap- considered separately for each of these three
peared in EIBI publications warrant consider- verbal operants in the sections that follow.
ation. We are not aware of any publications that Mands. At the very beginning of a child’s
recommend a consistent expressive-before-re- intervention program, M. L. Sundberg and
ceptive sequence for learners who are capable of Partington (1998) place a strong emphasis on
benefiting from expressive instruction, even the establishment of a rudimentary vocal or
though this sequence may have greater empir- signed mand repertoire. If their recommenda-
ical support than the receptive-before-expressive tions are followed, a child may acquire mands
sequence. However, verbal behavior model for several items before he or she receives any
publications sometimes offer alternative recom- receptive training with the same or other items,
mendations that are based on their conceptual or any other expressive training. Empirically, it
foundations as well as practical considerations. is unknown how this recommendation affects
In this section, we review the empirical support future acquisition or the overall outcome of
behind recommendations made in two such the intervention program. However, the rec-
books: M. L. Sundberg and Partington (1998) ommendation to teach mands first may be
and Greer and Ross (2008). The book by M. L. justifiable on practical grounds alone, because a
Sundberg and Partington has enjoyed substan- mand repertoire provides the child with
tial popularity by EIBI providers (Love et al., appropriate means of initiating communication
2009), and its sequence recommendations to access a variety of reinforcers. In addition, the
regarding receptive and expressive language acquisition of effective mands for specific
objectives are more explicit than those made reinforcers can result in a decrease in inappro-
in other verbal behavior model textbooks. Greer priate behavior that has been maintained by the
and Ross, by contrast, do not make specific same reinforcers (e.g., E. G. Carr & Durand,
recommendations regarding the sequence in 1985).
which to introduce early receptive and expres- Tacts. When teaching tacts and related
sive targets. However, their recommendations receptive behavior, M. L. Sundberg and
for establishing advanced receptive and expres- Partington (1998) recommend interspersing
sive language skills are unique in the EIBI tact and receptive training trials for the same
literature and of potential relevance to the items within the same teaching sessions. These
present discussion. sessions also include instructional and mainte-
nance trials for various other instructional
M. L. Sundberg and Partington (1998) targets. No specific rationale is offered for this
Following Skinner (1957), M. L. Sundberg strategy, except for the assumption that tact and
and Partington (1998) assume that the behavior receptive repertoires are functionally independent
868 ANNA INGEBORG PETURSDOTTIR and JAMES E. CARR

of one another and, as a result, each must equal number of receptive and expressive trials
be established through direct contingencies of may not be necessary.
reinforcement. Although some research sup- Intraverbals. Some of the early intraverbal
ports the notion that tact and receptive reper- targets described by M. L. Sundberg and
toires can be acquired independently (e.g., Partington (1998), such as filling in blanks in
Guess & Baer, 1973; Lee, 1981; Wynn & songs, rhymes, and phrases, do not have any
Smith, 2003), the studies listed in Table 1 obvious corresponding receptive instructional
indicate that many children may acquire a targets. As a result, sequencing recommenda-
receptive repertoire without explicit training tions are irrelevant for those targets. However,
if they receive tact training with the relevant many of the more advanced intraverbals
items. At present, little is known about char- described in the book are related to a receptive
acteristics or prior histories that may differen- target. For example, a child might be taught to
tiate these children from children who require respond to phrases such as ‘‘You drive a —’’
explicit training of both repertoires. It is possi- receptively, by pointing to a car, or intraverb-
ble that even for children who can acquire both ally, by saying ‘‘car.’’ By the time these
repertoires as a result of tact training, simulta- programs are introduced, according to M. L.
neous training, as recommended by M. L. Sundberg and Partington’s recommendations,
Sundberg and Partington, is a more efficient the child has already acquired tacts of the target
strategy for achieving the same outcome. How- items (e.g., car). The receptive relations, in
ever, we are not aware of any studies that have which a child selects an item when presented
compared the efficiency of interspersed tact and with spoken instructions that exclude the name
receptive trials with tact-only, receptive-only, or
of the item, are termed receptive by function,
sequential training.
feature, and class (RFFC) by M. L. Sundberg
A few studies have evaluated a similar
and Partington. As with tacts and recep-
interspersal strategy for teaching tacts and
tive discriminations, M. L. Sundberg and
mands. Two studies (Arntzen & Almas, 2002;
Partington emphasize a need for directly train-
Carroll & Hesse, 1987) reported enhanced tact
ing both RFFC and the corresponding intra-
acquisition when tact trials were interspersed
verbals. Although they do not necessarily recom-
with mand trials for the same items, compared
mend a specific sequence, they suggest that it
to tact training alone, whereas Sidener et al.
often may be useful to complete RFFC first
(2010) failed to find the same effect. Similar
(similar to the typical recommendation in UCLA-
strategies might be used to evaluate the
interspersal of verbal operant trials with recep- model publications). The rationale for this sugges-
tive trials. The appropriate ratio of tact to tion is that children may often start tacting the
receptive trials in interspersed training also stimuli that they select during RFFC trials; for
might be investigated. Guess and Baer (1973) example, a child might respond to ‘‘You drive a —’’
found that three of the four participants who by pointing to a picture of a car and also saying
received expressive training with one set of ‘‘car.’’ Thus, establishing the intraverbal relation
instructional targets and receptive training with would only require fading the visual stimulus.
another set failed to acquire the untrained We are not aware of any studies that have
repertoire with one or both sets. In a second demonstrated benefits of a receptive-before-
experiment, expressive and receptive training intraverbal sequence over the opposite sequence
were interspersed with only a few reinforced or other possibilities, such as simultaneous
trials that targeted the alternate repertoire. This instruction. Again, a receptive-before-expressive
modification was sufficient to produce acquisi- sequence may be a practical necessity for
tion of both repertoires, suggesting that an children whose vocal repertoires are nonexistent
 LANGUAGE INSTRUCTION 869

or limited. However, this may not be true of research is needed with clinical populations.
children who have acquired the prerequisites for Although the aforementioned studies suggest
RFFC that M. L. Sundberg and Partington that intraverbal training often may generate
(1998) recommend, which include ‘‘50 words, receptive repertoires, it is possible that conduct-
signs, or pictures as mands, tacts, or receptive ing receptive training first might save instruc-
discriminations’’ (p. 174) that occur under a tional effort by reducing the number of
variety of different conditions with short subsequent intraverbal training trials. Alterna-
response latencies. If intraverbals are considered tively, the type of interspersed training that M. L.
to be an instance of expressive language, then Sundberg and Partington (1998) recommend for
the studies listed in Table 1 might provide receptive and tact trials might facilitate acquisi-
tentative support for the opposite expressive- tion of RFFC and intraverbals. These possibil-
before-receptive sequence, but it is important to ities have not been evaluated, and as a result, the
keep in mind that none of these studies actually advantages of having prior receptive train-
targeted intraverbals. We are aware of only ing serve as ‘‘a bridge to the development of
two published studies that have evaluated or some types of intraverbal responding’’ (M. L.
compared the efficiency of both receptive and Sundberg & Partington, p. 206) remain to be
intraverbal training (Petursdottir et al., 2008; demonstrated empirically.
Petursdottir & Haflidadottir, 2009). Both
Greer and Ross (2008)
studies were conducted with typically develop-
The Greer and Ross (2008) book is unique in
ing children, and as previously noted, they
that it emphasizes the establishment of learning
produced mixed results. Petursdottir et al.
histories that may permit a child to acquire new
(2008) simulated RFFC and intraverbal train-
skills without direct reinforcement. Other pub-
ing with nonsense syllables and arbitrary
lications also emphasize the ultimate goal that
stimuli. RFFC generated no increase in in- the child will be able to learn from common
traverbal responding, and intraverbal training classroom contingencies rather than one-on-one
with different participants had minimal effects instruction (e.g., Taylor & McDonough, 1996).
on RFFC performance. The other study Greer and Ross, however, propose specific learn-
(Petursdottir & Haflidadottir, 2009) was con- ing histories, the provision of which is hypoth-
ducted with two slightly older children and a esized to accomplish this goal. Greer and Ross
simplified task. Intraverbal training, as well as share M. L. Sundberg and Partington’s (1998)
tact training, generated receptive responding to assumption that the behavior of speaker and
criterion for both participants, whereas recep- listener are functionally independent in early
tive training usually failed to generate intraver- language acquisition. However, they empha-
bals and tacts to criterion. size the notion that these repertoires may be
In a study with adolescents with intellectual integrated through appropriate histories of rein-
disabilities (M. C. Luciano, 1986), a subset of forcement (a conceptual rationale has been
participants received intraverbal training without described by Greer & Longano, 2010; Greer &
previously acquiring related receptive discrimi- Speckman, 2009). After the child has acquired
nations. These receptive discriminations were basic tact and listener (receptive) repertoires, as
found to be highly accurate after intraverbal well as other prerequisite skills, Greer and Ross
training. Similar findings were reported by both introduce a protocol that is intended to establish
C. T. Sundberg and Sundberg (1990) and naming, which is defined as ‘‘the capacity to
Remington and Clarke (1993). However, none acquire a tact … by simply hearing another
of these studies evaluated the effects of recep- person tact a stimulus’’ (p. 149). Accordingly,
tive training on intraverbals. Thus, additional the goal of this protocol is for the child to be able
870 ANNA INGEBORG PETURSDOTTIR and JAMES E. CARR

to acquire new tacts and listener behavior tacts and listener responses are established
without direct instruction, and ultimately by indirectly by having the child observe another
merely observing another person’s tacts. The person’s tacts. Thus, if the protocol is success-
primary mechanism proposed to accomplish this ful, it may save future instructional effort and
goal is instruction with multiple exemplars. prepare the child to learn new tacts and listener
Greer and Ross’s (2008) multiple-exemplar relations through typical classroom instruction.
instruction (MEI) recommendations may be Greer and Ross describe similar procedures to
summarized as follows. First, the child receives accomplish a variety of other objectives, such as
identity match-to-sample training with five establishing reading and writing skills and
visual stimuli (hereafter termed Set 1) for which enabling a child to acquire mands through tact
he or she has not acquired tacts or listener training.
behavior. During match-to-sample training, the The reader may recognize that Greer and
instructor vocalizes the name of the target Ross’s (2008) protocol shares some similarities
stimulus in each trial; however, correct respons- with M. L. Sundberg and Partington’s (1998)
es do not require control by this vocally simultaneous tact and receptive training recom-
presented stimulus. After the child has acquired mendation, in that both involve interspersing
the identity-matching task, the instructor tact and receptive trials for the same items
probes listener behavior by asking the child to within the same teaching sessions. However,
point to each stimulus given its spoken name. they are not the same recommendation. First, a
In addition, tacts are probed under two teaching session as described by M. L. Sundberg
conditions, with and without an instruction to and Partington includes a greater variety of
tact (e.g., ‘‘What color is this?’’). If listener different instructional targets than an MEI
behavior and tacts are not found to emerge to session as described by Greer and Ross. Second,
criterion level, MEI commences. MEI initially Greer and Ross provide specific guidance on the
is conducted with a set of five new stimuli. Four arrangement of trials within MEI sessions; for
types of trials (identity match-to-sample trials, example, they specify that two instructional
listener trials, and tact trials with and without trials that involve the same stimuli but different
an instruction) are conducted with each target responses (e.g., a tact trial in which the
stimulus, and different trial types with different target response is ‘‘boat’’ and a listener trial in
stimuli are interspersed with one another. After which the target response is pointing to a boat)
the child has met the acquisition criterion for all should never be temporally contiguous. By
four trial types, listener behavior and tacts are contrast, M. L. Sundberg and Partington
probed again with Set 1. If they have not provide no specific suggestions for the order
emerged, MEI is repeated with additional sets, of trials within a session, and their general
until it is demonstrated that the child can approach implies that temporally contiguous
acquire new listener responses and tacts through tact and receptive trials might be desirable at the
the identity match-to-sample procedure alone. beginning of instruction with a new target,
The protocol also may be completed initially because such contiguity might assist with
without any tact trials if a child has already transfer of stimulus control between visual and
acquired a listener, but not a tact, repertoire. vocally presented stimuli. Third, Greer and
The initial goal is then for the child to acquire Ross’s MEI protocol is a temporary intervention
listener responses through the identity match- conducted with a few target items, whereas M.
to-sample procedure, and the protocol later is L. Sundberg and Partington describe a strategy
repeated with tacts included. After the full for teaching a wide variety of tacts and receptive
protocol has been completed, the child receives skills during the course of a child’s intervention
no further tact or listener training. Instead, new program. Accordingly, Greer and Ross’s protocol,
 LANGUAGE INSTRUCTION 871

but not that of M. L. Sundberg and Partington, Conclusion

includes systematic assessment of the emergence The verbal behavior model publications
of tacts and listener behavior without direct generally do not recommend the receptive-
training, and the protocol is discontinued when before-expressive sequence, although M. L.
this outcome is achieved. Sundberg and Partington (1998) suggest that
Training with multiple exemplars has long it may be advantageous in certain cases.
been known to enhance stimulus generalization However, to the extent that these books make
(e.g., Stokes, Baer, & Jackson, 1974). In recent specific sequence recommendations, their ben-
years, a number of studies also have produced efits remain to be investigated empirically.
evidence that histories of multiple-exemplar Greer and Ross (2008) emphasize the use of
training may lead future instruction to produce multiple-exemplar training to overcome func-
various forms of untrained responses (e.g., tional independence of speaker and listener, or
Y. Barnes-Holmes, Barnes-Holmes, Roche, & receptive and expressive, repertoires. Although
Smeets, 2001a, 2001b; Y. Barnes-Holmes, the empirical literature to date (Fiorile & Greer,
Barnes-Holmes, Smeets, Strand, & Friman, 2007; Greer et al., 2005, 2007; Speckman-
2004; Berens & Hayes, 2007; Gomez, Lopez, Collins et al., 2007) is promising, a number of
Martin, Barnes-Holmes, & Barnes-Holmes, questions remain unanswered. First, it is likely
2007; C. Luciano, Becerra, & Valverde, 2004; that children who progress to advanced stages of
Murphy & Barnes-Holmes, 2010). Although language acquisition in programs that do not
employ multiple-exemplar training systemati-
most of these studies have not included an
cally will eventually become capable of acquir-
assessment of untrained vocal responses, a few
ing new receptive and expressive skills without
studies (Fiorile & Greer, 2007; Greer, Stolfi,
explicit instruction. Indeed, expressive training
Chavez-Brown, & Rivera-Valdes, 2005; Greer,
often generated receptive skills for the partici-
Stolfi, & Pistoljevic, 2007; Speckman-Collins,
pants in the studies listed in Table 1, and
Park, & Greer, 2007) specifically have evaluated
receptive training sometimes generated expres-
the multiple-exemplar procedures for tacts and
sive skills. The extent to which multiple-
listener behavior described in Greer and Ross exemplar training results in more rapid achieve-
(2008) with participants with autism spectrum ment of this outcome may need to be
disorders. These studies have demonstrated investigated. In other words, additional research
incidental acquisition of tacts and listener is needed on the extent to which the specific
responses after, but not before, MEI. A potential format of instruction described by Greer and
limitation is that some of the studies did not Ross is necessary to achieve the desired
clearly isolate the effects of MEI from the effects outcome. Although the Greer et al. (2007)
of repeated testing, or the effects of simply study suggested that repeated iterations of
providing additional receptive or expressive sequential receptive and expressive training
instruction. However, one study (Greer et al., might not produce the same effect as MEI,
2007) included a control group that received other possibilities, such as the form of inter-
receptive and expressive instructional trials spersed receptive and expressive training de-
separately, and suggested that the MEI format scribed by M. L. Sundberg and Partington,
produced superior results. In addition, the remain to be evaluated. Second, Greer and Ross
results are consistent with other research on suggested that ‘‘before the child has naming …
multiple-exemplar training that has controlled multiple listener and tact responses must be
for additional testing and instruction (Y. taught directly’’ and ‘‘many children will need
Barnes-Holmes et al., 2001a, 2001b; Gomez extensive tact instruction before they are ready
et al., 2007). to learn the naming capability’’ (p. 158). The
872 ANNA INGEBORG PETURSDOTTIR and JAMES E. CARR

specific prerequisites for beginning multiple- training, and simultaneous receptive and
exemplar training may need to be investigated expressive training with children who have
further. Third, as with research on the recep- autism spectrum disorders. In these studies,
tive-before-expressive recommendation, the ex- it will be necessary to describe participant
isting empirical literature has mostly focused characteristics and existing verbal repertoires
on tacts of visual stimuli as an instance of thoroughly in terms of standardized language
expressive training (but see Greer, Yuan, & assessment scores and in terms of prior
Gautreaux, 2005). acquisition of instructional objectives. A crucial
dependent variable to assess is the total amount
DIRECTIONS FOR FUTURE RESEARCH of instruction required to establish both recep-
tive and expressive repertoires in each condi-
The authors of some of the early studies listed
tion. A few of the studies listed in Table 1
in Table 1 concluded that ‘‘it may be unnec-
attempted to evaluate the receptive-before-
essary to take the time to teach comprehension
expressive recommendation primarily by com-
first on the assumption that it will facilitate
paring the effects of receptive training on
acquisition of production’’ (Cuvo & Riva,
expressive repertoires to the effects of expressive
1980; p. 330), and ‘‘[the results] call into
training on receptive repertoires (Keller &
question the value of prior receptive training in
Bucher, 1979; Wynn & Smith, 2003). How-
language skills for which productive perfor-
mance is desired’’ (Keller & Bucher, 1980, ever, this measure potentially could provide
p. 102). It appears that these conclusions may misleading data if not presented in the context
still apply. If a clinical recommendation is to be of the total amount of instruction required to
made based on the available data, practitioners establish both repertoires. If the emergence of
might consider ignoring suggestions to com- the alternate repertoire is employed as a second
plete new receptive programs before introduc- dependent measure, the receptive training and
ing expressive programs for the same instruc- testing conditions might use large stimulus
tional targets to learners who have existing arrays in order to prevent a bias against the
receptive and expressive repertoires, at least in receptive condition.
the case of expressive protocols that target the As previously noted, the expressive repertoires
tact. Instead, they might consider teaching the that were targeted in most of the studies listed in
tact first, then probing the relevant receptive Table 1 were tacts of visual stimuli according to
response, and teaching it only if necessary. Skinner’s (1957) verbal operant terminology. This
The existing literature, however, is small is understandable, given that tacts of visual stimuli
and dated and has a number of limitations, feature prominently in early expressive lan-
including the absence of known autism diag- guage objectives (e.g., Leaf & McEachin, 1999;
noses for most of the participants. In addition, M. L. Sundberg & Partington, 1998; Taylor &
the relative efficiency of the alternative simul- McDonough, 1996). Due to limitations present
taneous training arrangement described by M. in the existing literature, future research on
L. Sundberg and Partington (1998) needs to be instructional arrangements should continue to
evaluated. Given the ubiquity of receptive and focus on the tact. However, this research also
expressive language objectives in a typical EIBI should be extended to other verbal operants, such
curriculum, additional research is warranted to as the intraverbal, which is a common component
identify the most efficient arrangement of recep- of instructional programs for which the receptive-
tive and expressive instruction. Future research before-expressive sequence has been recom-
should compare the efficiency of receptive-before- mended (Leaf & McEachin, 1999; Lovaas,
expressive training, expressive-before-receptive 2003; Taylor & McDonough, 1996).
 LANGUAGE INSTRUCTION 873

Future research may reveal individual differ- and, as previously noted, it has some empirical
ences with respect to the most efficient support. However, other variables also have been
arrangement of receptive and expressive trials. proposed to influence this outcome. For exam-
In Wynn and Smith’s (2003) study, although ple, Horne and Lowe’s (1996) naming hypoth-
expressive training overall generated a receptive esis suggests that, during training, the occurrence
repertoire more often than receptive training of certain collateral responses may result in the
generated an expressive repertoire, the reverse emergence of new verbal operants or listener
sequence appeared to be more efficient for one behavior. Horne, Lowe, and Randle (2004)
participant. One of Keller and Bucher’s (1980) reported the results of an unpublished study by
participants similarly appeared to benefit more Bell (1999) in which a requirement to emit
from receptive than expressive training. As a collateral echoic responses during listener train-
result, it may be advisable to develop assessment ing resulted in increased tact emergence, but we
procedures to determine the most efficient are unaware of any published experimental
instructional arrangement for individual learn- demonstrations of this effect. Lowenkron
ers. In addition, researchers might attempt to (1998) similarly has suggested a role of self-
identify variables that influence the efficiency echoic responses during testing in the emergence
of particular arrangements and contribute to of novel listener behavior, and some evidence
individual differences. The identification of exists that self-echoic training may enhance such
such variables might help to determine when outcomes among children diagnosed with autism
to introduce changes into a child’s instructional (Tu, 2006). Future research might explore these
program. For example, when is it necessary to variables further. In addition, several studies
train both receptive and expressive responses have identified variables that affect transfer
related to the same target, and when is a child between verbal operants, such as the establish-
ready to benefit from expressive training alone? ment of autoclitic frames (Hernandez, Hanley,
In addition to the participants’ existing verbal Ingvarsson, & Tiger, 2007) and training re-
repertoires, relevant variables might include, sponses to elements of instructions presented on
for example, the amount of history with recep- test trials (Pérez-González, Herszlikowicz, &
tive and expressive training and prior experi- Williams, 2008). Future research might examine
ence with the specific response topographies whether these variables also affect transfer
required in receptive and expressive trials (Lee, between speaker and listener repertoires.
1981). Finally, it may be worthwhile to evaluate
For children who have progressed beyond the other suggestions regarding sequences of in-
earliest stages of language acquisition, the structional programs that have been recom-
potential benefits of multiple-exemplar training mended in EIBI publications. For example,
should be evaluated further. Prerequisites for Leaf and McEachin (1999) and Taylor and
beginning multiple-exemplar training should be McDonough (1996) suggest completion of
examined, and alternative means of achieving the visual–visual match-to-sample protocols prior
outcome of integrated speaker and listener to the introduction of receptive or expressive
repertoires might be explored. At a more basic programs that target categorization, and Greer
level of research, researchers should continue to and Ross (2008) suggest prerequisites for many
search for variables that influence the emergence of the instructional programs presented in their
of verbal operants after receptive training and book. Although EIBI already has sufficient
receptive repertoires after verbal operant training. empirical support to be considered the treat-
Training with multiple exemplars is one such ment of choice for children diagnosed with
variable that has been proposed (e.g., D. Barnes- autism spectrum disorders (Eldevik et al., 2009),
Holmes, Barnes-Holmes, & Cullinan, 2000), continued research on specific components of
874 ANNA INGEBORG PETURSDOTTIR and JAMES E. CARR

programming can only serve to enhance its Fiorile, C. A., & Greer, R. D. (2007). The induction of
naming in children with no prior tact responses as a
efficiency. function of multiple exemplar histories of instruction.
The Analysis of Verbal Behavior, 23, 71–87.
Gomez, S., Lopez, F., Martı́n, C. B. B., Barnes-Holmes,
REFERENCES Y., & Barnes-Holmes, D. (2007). Exemplar training
Arntzen, E., & Almas, I. K. (2002). Effects of mand-tact and a derived transformation of function in accor-
versus tact-only training on the acquisition of tacts. dance with symmetry and equivalence. The Psycho-
Journal of Applied Behavior Analysis, 35, 419–422. logical Record, 57, 273–294.
Barbera, M. L. (2007). The verbal behavior approach: How Greer, R. D., & Longano, J. (2010). A rose by naming:
to teach children with autism and related disorders. How we may learn how to do it. The Analysis of
London: Jessica Kingsley. Verbal Behavior, 26, 73–106.
Barnes-Holmes, D., Barnes-Holmes, Y., & Cullinan, V. Greer, R. D., & Ross, D. E. (2008). Verbal behavior
(2000). Relational frame theory and Skinner’s Verbal analysis: Inducing and expanding complex communica-
Behavior: A possible synthesis. The Behavior Analyst, tion in children with severe language delays. Boston:
23, 69–84. Allyn & Bacon.
Barnes-Holmes, Y., Barnes-Holmes, D., Roche, B., & Greer, R. D., & Speckman, J. (2009). The integration of
Smeets, P. M. (2001a). Exemplar training and a speaker and listener responses: A theory of verbal
derived transformation of function in accordance with development. The Psychological Record, 54, 449–488.
symmetry. The Psychological Record, 51, 287–308. Greer, R. D., Stolfi, L., Chavez-Brown, M., & Rivera-
Barnes-Holmes, Y., Barnes-Holmes, D., Roche, B., & Valdes, C. (2005). The emergence of the listener to
Smeets, P. M. (2001b). Exemplar training and a speaker component of naming in children as a
derived transformation of function in accordance with function of multiple exemplar instruction. The
symmetry: II. The Psychological Record, 51, 569–603. Analysis of Verbal Behavior, 21, 123–134.
Barnes-Holmes, Y., Barnes-Holmes, D., Smeets, P. M., Greer, R. D., Stolfi, L., & Pistoljevic, N. (2007).
Strand, P., & Friman, P. (2004). Establishing Emergence of naming in preschoolers: A comparison
relational responding in accordance with more-than of multiple and single exemplar instruction. European
and less-than as generalized operant behavior in Journal of Behavior Analysis, 8, 119–131.
young children. International Journal of Psychology Greer, R. D., Yuan, L., & Gautreaux, G. (2005). Novel
and Psychological Therapy, 4, 531–558. dictation and intraverbal responses as a function of
Berens, N. M., & Hayes, S. C. (2007). Arbitrarily multiple exemplar instructional history. The Analysis
applicable comparative relations: Experimental evi- of Verbal Behavior, 21, 99–116.
dence for a relational operant. Journal of Applied Guess, D., & Baer, D. M. (1973). An analysis of
Behavior Analysis, 40, 45–71. individual differences in generalization between
Carr, E. G., & Durand, M. V. (1985). Reducing behavior receptive and productive language in retarded chil-
problems through functional communication train- dren. Journal of Applied Behavior Analysis, 6,
ing. Journal of Applied Behavior Analysis, 18, 311–329.
111–126. Hernandez, E., Hanley, G. P., Ingvarsson, E. T., & Tiger,
Carr, J. E., & Firth, A. M. (2005). The verbal behavior J. H. (2007). A preliminary evaluation of the
approach to early and intensive behavioral interven- emergence of novel mand forms. Journal of Applied
tion for autism: A call for additional empirical Behavior Analysis, 40, 137–156.
support. Journal of Early and Intensive Behavior Holdgrafer, G., & McReynolds, L. (1975). An experi-
Intervention, 2, 18–27. mental analysis of comprehension and production in
Carroll, R. J., & Hesse, B. E. (1987). The effects of children’s acquisition of morphological rules. Human
alternating mand and tact training on the acquisition Communication, 5, 45–62.
of tacts. The Analysis of Verbal Behavior, 5, 55–56. Horne, P. J., & Lowe, C. F. (1996). On the origins of
Connell, P. J., & McReynolds, L. (1981). An experimen- naming and other symbolic behavior. Journal of the
tal analysis of children’s generalization during lexical Experimental Analysis of Behavior, 65, 185–241.
learning: Comprehension or production. Applied Horne, P. J., Lowe, C. F., & Randle, V. R. L. (2004).
Psycholinguistics, 2, 309–332. Naming and categorization in young children: II.
Cuvo, A. J., & Riva, M. T. (1980). Generalization and Listener behavior training. Journal of the Experimental
transfer between comprehension and production: A Analysis of Behavior, 81, 267–288.
comparison of retarded and nonretarded persons. Hupp, S. C., Mervis, C. B., Able, H., & Conroy-Gunter,
Journal of Applied Behavior Analysis, 13, 315–331. M. (1986). Effects of receptive and expressive training
Eldevik, S., Hastings, R. P., Hughes, J. C., Jahr, E., of category labels on generalized learning by severely
Eikeseth, S., & Cross, S. (2009). Meta-analysis of mentally retarded children. American Journal of
early intensive behavioral intervention for children Mental Deficiency, 90, 558–565.
with autism. Journal of Clinical Child and Adolescent Keller, M. F., & Bucher, B. D. (1979). Transfer between
Psychology, 38, 439–450. receptive and productive language in developmentally
 LANGUAGE INSTRUCTION 875

disabled children [Abstract]. Journal of Applied Petursdottir, A. I., & Haflidadottir, R. D. (2009). A
Behavior Analysis, 12, 311. comparison of four strategies for teaching a small
Keller, M. F., & Bucher, B. (1980). The influence of foreign-language vocabulary. Journal of Applied Be-
receptive training on rate of productive language havior Analysis, 42, 685–690.
acquisition in mentally retarded children. Behavior Remington, B., & Clarke, S. (1993). Simultaneous
Research of Severe Developmental Disabilities, 1, communication and speech comprehension: Part II.
93–103. Comparison of two methods of overcoming selective
Leaf, R., & McEachin, J. (1999). A work in progress. attention during expressive sign training. Augmenta-
Behavior management strategies and a curriculum for tive & Alternative Communication, 9, 49–60.
intensive behavioral treatment of autism. New York: Sallows, G. O., & Graupner, T. D. (2005). Intensive
DRL. behavioral treatment for children with autism: Four-
Lee, V. L. (1981). Prepositional phrases spoken and heard. year outcome and predictors. American Journal on
Journal of the Experimental Analysis of Behavior, 35, Mental Retardation, 10, 417–438.
227–242. Schramm, R. (2006). Educate toward recovery: Turning the
Lovaas, O. I. (1977). The autistic child. New York: Irvine. tables on autism. A teaching manual for the verbal
Lovaas, O. I. (1981). Teaching developmentally disabled behavior approach to ABA. Retrieved from Lulu.com
children. The ME book. Austin, TX: Pro-Ed. Sidener, T. M., Carr, J. E., Karsten, A. M., Severtson, J.
Lovaas, O. I. (1987). Behavioral treatment and normal M., Cornelius, C. E., & Heinicke, M. R. (2010).
educational and intellectual functioning in young Evaluation of single and mixed verbal operant
autistic children. Journal of Consulting and Clinical arrangements for teaching mands and tacts. The
Psychology, 55, 3–9. Analysis of Verbal Behavior, 26, 15–30.
Lovaas, O. I. (2003). Teaching individuals with develop- Skinner, B. F. (1957). Verbal behavior. Acton, MA:
mental delays. Basic intervention techniques. Austin, Copley.
TX: Pro-Ed. Smeets, P. M. (1978). Establishing generative performance
Love, J. R., Carr, J. E., Almason, S. M., & Petursdottir, A. and cross modal generalization of the manual plural
I. (2009). Early and intensive behavioral intervention sign in a severely retarded deaf girl. British Journal of
for autism: A survey of clinical practices. Research in Communication Disorders, 13, 49–57.
Autism Spectrum Disorders, 3, 421–428. Smeets, P. M., & Striefel, S. (1976). Acquisition and cross
Lowenkron, B. (1998). Some logical functions of joint modal generalization of receptive and expressive
control. Journal of the Experimental Analysis of signing skills in a retarded deaf girl. Journal of Mental
Behavior, 69, 327–354. Deficiency Research, 20, 251–259.
Luciano, C., Becerra, I. G., & Valverde, M. R. (2007). Smith, T., Groen, A. D., & Wynn, J. W. (2000). Randomized
The role of multiple-exemplar training and naming trial of intensive early intervention for children with
in establishing derived equivalence in an infant. pervasive developmental disorder. American Journal on
Journal of the Experimental Analysis of Behavior, 87, Mental Retardation, 105, 269–285.
349–365. Speckman-Collins, J., Park, H. S., & Greer, R. D. (2007).
Luciano, M. C. (1986). Acquisition, maintenance, and Generalized selection-based auditory matching and
generalization of productive intraverbal behavior the emergence of the listener component of Naming.
through a transfer of stimulus control procedures. Journal of Early and Intensive Behavior Intervention, 4,
Applied Research in Mental Retardation, 7, 1–20. 412–429.
Maurice, C., Green, G., & Luce, S. (Eds.). (1996). Stokes, T., Baer, D. M., & Jackson, R. L. (1974).
Behavioral intervention for young children with Programming the generalization of a greeting
autism: A manual for parents and professionals. Austin, response in four retarded children. Journal of Applied
TX: Pro-Ed. Behavior Analysis, 7, 599–610.
Miller, M. A., Cuvo, A. J., & Borakove, L. (1977). Sundberg, C. T., & Sundberg, M. L. (1990). Comparing
Teaching naming of coin values—Comprehension topography-based verbal behavior with stimulus
before production versus production alone [Abstract]. selection-based verbal behavior. The Analysis of Verbal
Journal of Applied Behavior Analysis, 10, 735–736. Behavior, 8, 31–41.
Murphy, C., & Barnes-Holmes, D. (2010). Establishing five Sundberg, M. L., & Partington, J. W. (1998). Teaching
derived mands in three adolescent boys with autism. language to children with autism or other developmental
Journal of Applied Behavior Analysis, 43, 537–541. disabilities. Pleasant Hill, CA: Behavior Analysts,
Pérez-González, L. A., Herszlikowicz, K., & Williams, G. Inc.
(2008). Stimulus relations analysis and the emergence Taylor, B. A., & McDonough, K. A. (1996). Selecting
of novel intraverbals. The Psychological Record, 58, teaching programs. In C. Maurice, G. Green, & S.
95–129. Luce (Eds.), Behavioral intervention for young children
Petursdottir, A. I., Carr, J. E., Lechago, S. A., & Almason, with autism: A manual for parents and professionals
S. M. (2008). An evaluation of intraverbal training (pp. 63–177). Austin, TX: Pro-Ed.
and listener training for teaching categorization skills. Tu, J. C. (2006). The role of joint control in the manded
Journal of Applied Behavior Analysis, 41, 53–68. selection responses of both vocal and non-vocal
876 ANNA INGEBORG PETURSDOTTIR and JAMES E. CARR

children with autism. The Analysis of Verbal Behavior, Williams, G. C., & McReynolds, L. V. (1975). The
22, 191–207. relationship between discrimination and articulation
Virués-Ortega, J. (2010). Applied behavior analytic training in children with misarticulations. Journal of
intervention for autism in early childhood: Meta- Speech and Hearing Research, 18, 401–412.
analysis, meta-regression, and dose-response meta- Wynn, J. W., & Smith, T. (2003). Generalization between
analysis of multiple outcomes. Clinical Psychology receptive and expressive language in young children
Review, 30, 387–399. with autism. Behavioral Interventions, 18, 245–266.
Watters, R. G., Wheeler, L. J., & Watters, W. E. (1981). The
relative efficiency of two orders for training autistic Received September 30, 2010
children in the expressive and receptive use of manual Final acceptance March 8, 2011
signs. Journal of Communication Disorders, 14, 273–285. Action Editor, Michael Kelley
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.