3

Responsivity Education/
Prelinguistic Milieu Teaching
STEVEN F. WARREN, SHELLEY L. BREDIN-OJA, MARTHA FAIRCHILD,
UZBETH H. FINESTACK, MARC E. FEY, AND NANCY C. BRADY

..
ABSTRACT

.,,... Responsivity education!prelinguistic milieu teaching (RE/PMT) is most ap-

propriate for chll,dren who are functioning developmentally between the ages
of approximately 9 and 15 months. Many children with developmental delays
do not reach this developmental period until they are 2 or 3 years of age.
RE/PMT consists of two components: prelinguistic milieu teaching (PMT), 8""0a-e
which is delivered by the clinician to the child, and responsivity education -4- JL.,.
(RE), which is delivered by the clinician to the parents. PMT is designed to - ~Yi e
increase the frequency and complexity of intentional nonverbal commu-J ~.:~.
nicative acts to set the stage for later language learning. Our approach as- ,..,-ce..rf".S
sumes that many parents may not be optimally responsive to the nonverbal
*Co:rrlmunicatlve bids of their children. Consequently, we provide the second
component, RE. In this chapter, we describe the structure of our approach
and its theoretical and empirical bases, and we present several issues that
clinicians must consider to use the intervention effectively.

INTRODUCTION

Throughout the first year of life, the building blocks for language develop-
ment are assembled. Auditory development accelerates, complex babbling ,
emerges, social responsiveness and receptive language skills blossom, and
with the onset of coordinated attention, intentional communication appears,
first in only nonverbal forms (e.g., proto-declaratives such as pointing at

•
Support for much of the research reported in this chapter was provided by grants from the
National Institute of Child Health and Human Development (ROl HD27594; ROl HD34520) and
the Office of Special Education Programs of the U.S. Department of Education (H023C20152; .
H0324C990040). We would like to acknowledge the central role played by Paul Yoder in much of,the
research described here and also acknowledge the support for this research provided by the NICHD
Mental Retardation Research Centers at the University of Kansas and Vanderbilt University.

47
48 Warren et al.

a plane overhead to draw the parent's attention to it and thus create joint
attention). These developmental breakthroughs typically occur by 9 or 10
months of age, well before children utter their first spontaneous words (Bates,
Benigni, Bretherton, Camaioni, & Volterra, 1979). A significant delay in the
emergence of these building blocks of communication is a strong indicator
that the onset of productive language will also be delayed (McCathren, War-
ren, & Yoder, 1996). The basic premise of the two-component intervention
(RE/PMT) approach we present in this chapter is that prelinguistic commu-
nication development establishes the foundation for later language develop-
ment. We further assume that, when prelinguistic communication develop-
ment is delayed or disordered, a carefully targeted and well-implemented
treatment program can help to develop the critical intentional communica-
tion skills necessary for early language intervention to be maximally effec-
tive. In this chapter, we address issues related to identification of candidates
for RE/PMT, discuss the theoretical and empirical bases for the approach,
and describe the practical requirements for this intervention and its key
components. We conclude by presenting our ideas on how children's prog-
ress With prelinguistic intervention can be monitored, how the procedures
can be accommodated to a family's cultural and linguistic differences, and fu-
ture directions in the development and use of the approach.

TARGET POPULATIONS AND ASSESSMENTS FOR

DETERMINING TREATMENT RELEVANCE AND GOALS

RE/PMT is designed for young children who have not yet become frequent,
clear prelinguistic communicators by approximately 12-18 months of age. By
this age, and even earlier for children with severe developmental delays such
as those with Down syndrome, speech-language pathologists (SLPs) can
conclude that some form of interventio.n to facilitate communication devel-
opment is likely to be beneficial, if not necessary. Research on RE/PMT has
focused on children with a mental age of at least 9 months (Warren & Yoder,
1998; Yoder & Warren, 2001; Yoder, Warren, & Hull, 1995). However, the pri-
mary issue in assessment usually is not whether children are too delayed, but
whether they are already too advanced. Applying the intervention with chil-
"'-' k c e~~. t<. dren for whom early language, rather than communication intervention is
liJ rz..S G "'- ? more appropriate would be a serious miscalculation (Yoder & Warren, 2002).
5 ~ rv-e. ~ RE/PMT is appropriate for children who need to increase their frequency
~ k -r~ V1 of gestures and vocalizations. Children·who already are frequent communi- ~
:, c..o::;.I(V)
1
ca. tors will not gain s.ub. stantially from this approach. Consequently, if there]
\~"'ta-""~ S is reliable evidence that a child uses more than 10 words or signs produc-
\.IV"~ 1 .., o t: tively or unders ds more than 7 rds, we would typically recommend
"e VVLf,'l: that intervention focus on expressive vocabulary (using speech or an alter-
~ I"V"' -rift) native communication system), regardless of the child's existing intentional
rorN;,
GA.f \A-communication repertoire. Additionally, RE/PMT is not appropriate for chil-
Responsivity Education!Prelinguistic Milieu Teaching 49

dren who produce more than 1 or at most 2 spontaneous, intentional com-

munication acts per minute in social play with an adult, especially if these
acts typically include a canonical vocalization (i.e., a true consonant plus a
vowel). These criteria for determining which children are too advanced for
RE/PMT also serv.e as the guides for determining when it is time for them to
move on into early language intervention. Yoder and Warren (2002) found
that when RE/PMT was implemented with children who exceeded these cri-
teria, the effects of intervention were minimal.
The Communication and Symbolic Behavior Scales (CSBS™) is an L S 9 .5
assessment instrument that has been developed specifically to describe A..S -t"" ~
the level of a child's prelinguistic communication skills (Wetherby & Prizant, -co c '
1993a, 1993b). This assessment has been independently shown to have very
high reliability and predictive validity (McCathren, Yoder, & Warren, 1999,
2000; Wetherby, Allen, Cleary, Kublin, & Goldstein, 2002). Its primary draw-
back is the length of time it takes to apply and score all of the segments,
which include a~sessments of comprehension and play skills. Obviously, if
norm-referenced information is needed, especially for children who do not
clearly have significant delays in communication development, administra-
tion of the entire test would be desirable (note that the CSBS is appropriate
~nly up to 24 months developmentally). For our purpose of describing chil-
dren's rate of communication and rates of different types of prelinguistic com-
munication, we administer only the communication temptations and shar-
ing books segments as part of our standard assessment. These two segments
provide children with opportunities to request objects, request social inter-
action, comment about novel events, and repair communication breakdowns
within a series of scripted play interactions. We videotape the interactions and
later record and count the number of nonverbal requests (proto-imperatives)
and comments (proto-declaratives) in order to determine whether a child is
likely to benefit most from a RE/PMT (as opposed to a linguistically based in-
tervention) and as a baseline for examining treatment outcomes.
iYAt"SAt-'"t"i"b ~ oJl- V'Y"' t>~l
THEORETICAL BASIS
~ J..o i l d. £:: ~ e .,. vCYC.., ""'e"'1'"
The basic premises and hypotheses underlying RE/PMT stem from a trans-
actional model of social communication development (McLean & Snyder-
McLean, 1978; Sameroff & Chandler, 1975). This model presumes that early
social and communication development are facilitated by bidirectional, re-
~

ciprocal interactions between ~nand their ~t. For example,

a change in the child, such as the onset of intentional communication, may
trigger a change in the social environment. Parents may be more inclined to
repeat and expand the child's messages using words, a response we refer to
as linguistic mapping. These changes then support further development in
the child (e.g., increased communica~ion and vocabulary) and subsequent
changes in the caregivers input (e.g., more complex language interaction
50 Warren et al.

with the child). In this way both the child and the environment change over
time and affect each other in reciprocal fashion as early achievements pave
the way for subsequent development.
A transactional model may be particularly well suited to understanding
social-communication development in young children because caregiver-
child interaction can play such an important role in this process. The period
of early development (from birth to approximately age 3 years) may repre-
sent a unique time during which transactional effects can have a substantial
impact on development. Specifically, the young child's relatively restricted
repertoire during this period may allow any changes in behavior to be espe-
cially salient and observable to caregivers. This in tum may allow adults to
be more responsive to the developing skills of the child than is possible later
in development when children's behavioral repertoires are far more expan-
sive and complex. Durmg this natural window of opportunity, the relation-
ship represented by the transactional model may be employed by a clever
practitioner to multiply the effects of relatively circumscribed interventions
and perhaps alter the very course of the child's development in a significant
way. However, the actions of the practitioner may need to be swift and in-
tense, or they may be muted by the child's steadily accumulating history.
To appreciate the true potential of transactional effects, consider that
an input difference in positive affect expressed by parents toward their child
of 10 events per day (a difference of less than 1 event per waking hour on av-
erage) will result in a cumulative difference of 10,950 such events over a 3-
year period. A child who experiences less positive affect may also experience
cumulatively more negative affect (e.g., "Stop that," "Get out of there," "Shut
your mouth up," "You're a bad baby"). It is easy to conceive of the combina-
tion of these qualitative and quantitative experiential differences contribut.:.
ing to deficits in attachment, exploratory behavior, self-concept, language
development, later school achievement, and so forth.
What evidence do we have that such large cumulative deficits occur
and/or that they play havoc with so.cial and communication development? Al-
though the evidence is mostly correlational, it is nevertheless compelling.
There is substantial evidence that young, typically developing children ex-
; .,. pv.-t ,f perience large differences in terms of the quantity and quality of language
input they receive, and these differences correlate with important indicators
'~· ~y y.
of development later in childhood (e.g., vocabulary size, IQ, reading ability,
vJl O..Y~Y
school achievement) (Feagans & Farran, 1982; Gottfried, 1984; Hart & Risley,
c.Le..-v. 1992; Prizant & Wetherby, 1990; Walker, Greenwood, Hart, & Carta, 1994).
Because young children with developmental delays or sensory disorders often
display low rates of initiation and responsiveness (Rosenberg & Abbeduto,
1993; Yoder, Davies, & Bishop, 1994), they also may experience input that
differs substantially in quantity and quality from the input that high achiev-
ing, typically developing children receive despite the best intentions and efforts
of their caregivers (Brooks-Gunn & Lewis, 1984; Crawley & Spiker, 1983).
51

The challenges faced by young children who initiate infrequently may be fur-
ther multiplied if their caregiver(s) are or learn to be relatively unresponsive
to their children's communicative efforts (e.g., Hart & Risley, 1995; Saxon,
Colombo, Robinson, & Frick, 2000; Tamis-LeMonda, Bomstein, Baumwell, &
Melstein Damast, 1996).
Caregivers who are unresponsive to their young child's initiations and/or
who often display depressed or negative affect toward the child may repre-
sent a risk factor in terms of the child's emotional, social, and communication
development (Landry, Smith, Miller-Loncar, & Swank, 1997). Unresponsive
caregivers often have children who are insecurely attached (Ainsworth,
. Blehar, Waters, & Wall, 1978), which is a risk factor for poor social-emotional
development (Bomstein, 1989). Furthermore, there is evidence that care-.
givers with low rates of responsivity toward their infants can negate or min-
imize the positive transactional effects of early intervention efforts because
they fail to respond to changes in their child's repertoire being generated by
the intervention (Mahoney, Boyce, Fewell, Spiker, & Wheeden, 1998; Yoder
& Warren, 1998). In. short, the generation of transactional effects likely de-
pends on sensitive, responsive caregivers who notice and nurture the child's
growth.
The generation of strong transactional effects in which the growth of
emotional, social, and communication skills is scaffolded by caregivers can
have a multiplier effect in which a small dose of early intervention may lead
to long-term effects. These effects are necessary when we consider that typ-
ical early intervention by a skilled clinician may represent only 1-2 hours per
week of a young child's potential learning time (Bailey, .Aytch, Odom, Symons,
& Wolery, 1999). Even a relatively intense intervention of 5 hours per week
of intensive interaction would represent just 5% of the child's available social
and communication skill learning time if we assume the child is awake and
learning 100 hours per week. Thus, unless direct intervention accounts for a
large portion of a child's waking hours, transactional effects involving care-
givers are necessary for early intervention efforts to achieve their potential.
In summary, RE/PMT is grounded in the assumption that prelinguistic
skills form the foundation for later language skills. In addition, the transac-
tional model of adult-child interaction serves as a mechanism by which en-
hanced prelinguistic development can serve as a scaffold for communication
and language development. That is, if the child begins to produce more in-
tentional communication acts and/or acts that are more complex (e.g.,
through PMT), parents should respond to those acts in ways that ultimately
will encourage the child to reproduce and revise their acts. Some parents of
children with developmental delays may develop patterns of responding to
their children that are not optimal for their children's communicative devel-
opment. Responsive interaction training may be useful to help them recog1 ~
nize and respond to even small changes in the topography of their children'sJ -t"~ ~;/;.]
communicative acts (Tannock, Girolametto, & Siegel, 1992). These inter-~ YeS-f 1)"'/
~.5~
~)
52 Warren et al.

actions set the stage for more communicative interactions that are higher in
quality and, ultimately, for functional communication using words or signs.

EMPIRICAL BASIS

The initial explorations of the effects of PMT, one of the two key components
of RE/PMT, by Yoder and Warren and their colleagues focused on just a few
children and used single-subject (multiple baseline) designs. These studies
showed that increases in the children's frequency and clarity of prelinguistic
requesting following intervention were correlated with increases in linguis-
tic mapping by teachers and parents who were na.lve as to the specific tech-
niques and goals of the intervention (Warren, Yoder, Gazdag, Kim, & Jones,
1993; Yoder, Warren, Kim, & Gazdag, 1994). In other words, the teachers and
parents of children who increased their use of nonverbal requests increased
their use of contingent responses that repeated, rephrased, or otherwise in-
corporated the presumed meaning of the child's act. Furthermore, children's
intentional requesting targeted in these studies was shown to generalize
across people, settings, communication styles, and time.
Based on the promising results of these initial small intervention stud-
ies, Yoder and Warren (1998, 1999a, 1999b, 2001) conducted a relatively large
(N = 58) longitudinal experimental study of the effects of PMT on the com-
munication and language development of children with general delays in de-
velopment. Fifty-eight children between the ages of 17 and32 months (mean
= 23;·SD = 4) with developmental delays and their primary parent partici-
pated in the study. The children were recruited from three early intervention
centers in Tennessee. Fifty-two of the children had no productive words at
the outset of the study; the remaining six children had between one and five
productive words. All children scored below the lOth percentile on the ex-
pressive scale of the MacArthur-Bates Communicative Development Inven-
tories (CDis} (Penson et al., 1993) and fit the Tennessee definition of devel-
opmental delay (i.e., at least a 40% delay in at least one developmental domain,
or at least a 25% delay in at least two developmental domains).
The children were randomly assigned to one of two treatment groups.
Twenty-eight of the children received PMT; the other 30 children received an
intervention termed responsive small-group (RSG). Treatment sessions
for both groups were 20 minutes per day, 3 or 4 days per week, for 6 months.
PMT represented an adaptation of milieu language teaching (Kaiser, Yoder,
& Keetz, 1992) that aimed to teach the form and functions of requesting and
commenting. It consisted of the following key components: 1) following the
child's attentionallead; 2) building social play routines (e.g., tum-taking inter-
actions such as rolling a ball back and forth); 3) using prompts, such as time
delays (e.g., after rolling the ball back and forth, withholding it until the child
initiated a request to roll it); as well as 4) natural consequences to the child's
acts (e.g., giving the child the desired ball).
/ PH '1 :;:) bE".s-r -tu v c '-"" ·re.lyvrr:;l v~ • --- - . -

'-... R,.S (A -=I 9 ,/ c .,;f t.t ttre-.syo -.,.,s)'v-e...

ResPonsivity Education/Prelinguistic Milieu Teaching 53

RSG represented an adaptation of the responsive interaction approach.

The adult played with the children in a highly responsive manner and com-
mented on what they were doing but never attempted to elicit or prompt any
communication function or form directly. Caregivers were kept. na'ive as to ··!
the specific methods, measures, records of child progress, and child goals
throughout the study. This allowed Yoder and Warren to investigate how
change in the children's behavior as a result of the interventions might affect
the behavior of the primary caretaker and how this, in turn, might affect the
child's development later in time. Data were collected at five points in time
for each dyad: at pretreatment, at posttreatment, and 6, 12, and 18 months
after completion of the intervention.
Although there were no significant main effects of either PMT or RSG,
both interventions had a range of effects on intentional communication de-
velopment among subgroups of the children. The treatment that was most
effective depended on the pretreatmenrlnaternal interaction style and the
·~

education level of the mother (Yoder & Warren, 1998, 2001). For children
with highly responsive and relatively well-educated mothers (i.e., 3-4 years
of college), PMT was effective in fostering generalized intentional communi-
cation development. However, for children with relatively unresponsive and
less well-educated mothers, RSG was relatively more successful in fostering
generalized intentional communication development.
The two interventions differed. along a few important dimensions that
provide a plausible explanation for these effects. PMT uses a child~centered
play context in which verbal or time delay prompts for more advanced forms
of communication are employed as well as social consequences for target re-
sponses, such as specific acknowledgment_(e.g., "That's right") and compli-
ance (e.g., immediately giving the child a toy he or she had requested). RSG
emphasized following the child's attentionallead and being highly responsive
to child initiations while avoiding the use of direct prompts for communica-
tion. Maternal interaction style may have influenced which intervention was
most beneficial because children may develop expectations concerning inter-
actions with adults (including teachers and clinicians) based on their history
of interaction with their primary caretaker. Thus, children with consistently
responsive parents may learn to persist in the face of communication break-
downs, such as might be occasioned by a direct prompt or time delay, be-
cause their history leads them to believe that their communication attempts
will usually be successful. On the other hand, children without this history may
cease communicating when their initial attempt fails. Thus, children of re-
sponsive mothers in the PMT group may have persisted when prompted and
learned effectively in this context, whereas children with unresponsive par-
ents may not have. In contrast, when provided with a highly responsive adult
who virtually never prompted them over a 6-.month period, children of unre-
sponsive mothers showed greater gains than did children of responsive par-
ents receiving the same treatment. For these children, exposure to a highly
54 Warren et al.

responsive adult was a novel experience that generated a high degree of ini-
tiation and responsiveness by them, apparently leading to the treatment re-
sponse that was observed, which eventually washed out during the 12-month
follow-up (Yoder & Warren, 2001).
The effects of maternal responsivity as a mediator and moderator of in-
tervention effects rippled throughout the longitudinal follow-up period. Yoder
and Warren demonstrated that children in the PMT group with relatively re-
sponsive mothers received increased amounts of responsive input from their
mothers in direct response to the children's increased intentional communi-
cation (Yoder & Warren, 2001). Furthermore, the effects of the intervention
with this group were found on the number of intentional communication acts
(Yoder & Warren, 1998) and of requests and comments (Yoder & Warren,
1999b). These became greater with time and significantly affected measures
ofexpressive (i.e., lexical density; expressive scores on the Reynell Devel-
opmental Language Scales [Reynell& Gruber, 1990]) and receptive language
development (i.e., number of semantic relations understood; receptive scores
on the Reynell Scales) 6 and 12 months after intervention ceased (Yoder &
Warren, 1999a, 2001). This fmding contrasts with the results of several early
intervention studies in which the effects were reported to wash out over time
(Farran, 2000).
Finally, two observations from the Yoder and Warren studies support
the prediction of the transactional model that children's early intentional
communication will elicit mothers' linguistic mapping, which in turn will fa-
cilitate children's vocabulary development. First, the amount of responsive
input by the primary caregiver was partly responsible for the association be-
tween intentional communication increases and later language development
(Yoder & Warren, 1999a). Second, there was a significant longitudinal rela-
tionship between maternal responsivity and expressive language develop-
ment (Yoder & Warren, 2001).
The implications of the results achieved by the Yoder and Warren study
(1998, 1999a, 1999b, 2001) are tempered by a more recent efficacy study
(Yoder & Warren, 2002). This study involved 39 prelinguistic toddlers with
developmental delays and their primary parent. As in the previous Yoder and
Warren study, all children scored below the 1Oth percentile on the expres-
sive scale of the CDIs (Fenson et al., 1993) and met the Tennessee definition
for developmental delay. However, in this study 17 of the children (44% of
the sample) had Down syndrome, whereas in the earlier Yoder and Warren
study (1998) only 4 of 58 children had Down syndrome (7%). Half of the
children were assigned randomly to a two-pronged treatment condition. In
this condition, the children received PMT and the primary caretakers went
through a training program intended to ensure that they used a highly re-
sponsive parenting style with their child. Results indicated that the parent-
training. component of the intervention did enhance parent responsivity.
Responsivity Education/Prelinguistic Milieu Teaching 55

However, the pattern of results in terms of various measures of child com-

munication development varied by pretreatment characteristics, and on some
measures the control group achieved growth superior to that of the inter-
vention group. Although the exact reasons for these findings were unclear,
post hoc analysis suggested that the criteria for entry into the study included
children for whom the intervention may have targeted skills that were "too
low." Specifically, the intervention accelerated growth in comments and lexical
density if children began treatment with low frequencies on these measures,
but it appeared to decelerate growth along these dimensions for children
who began treatment with relatively high rates of comments and canonical
vocal communication.
The most recent investigation of the effects of PMT and parent respon-
sivity training has also produced the most positive results yet reported. Fey
et al. (in press) systematically replicated the second Yoder and Warren study
(Yoder & Warren, 2002) with 51 toddlers (average age 26 months at the start
of the study) ·with developmental delays. Children were randomly assigned
to either a combination of PMT and parent responsivity training or to receive
only those services already provided by the community program in which they
were enrolled. Twenty-six of the 51 participants had Down syndrome. Due to
the ambiguous results obtained by Yoder and Warren (2002), three important
procedural modifications were made. First, to be enrolled in the study, all
children were required to have relatively low rates of prelinguistic comment-··
ing and canonical vocalizations because these characteristics were associ-
ated with the most positive outcomes in the second Yoder and Warren study
(2002). Second, clinicians took special care to desist from persistent efforts
to prompt requests when children were nonresponsive to such requests.
Third, when a child produced a clearly intentional act, clinicians responded by
complying with the request and by expanding it rather than modeling some
additional behavior, such as a nonlinguistic vocalization.
Based on data collected at the end of 6 months of intervention, Fey et al.
reported the first "main effect" of PMT plus parent responsivity training. Spe-
cifically, they reported a statistically significant increase in intentional com-
munication relative to the community intervention group. The results ob-
tained for comments and declaratives were not significant, but all adjusted
means were greater for the intervention group. The positive effect on inten-
tional communicative acts was also found to be statistically significant for the
13 subjects with Down syndrome when compared directly with the 13 Down
syndrome subjects in the community intervention group. The clinical signif-
icance of this finding will ultimately depend largely on the extent to which
these early gains are followed by later gains on verbal measures in this on-
going longitudinal investigation.
The initial longitudinal study by Yoder and Warren (1998, 2001) repre-
sents a relatively rare experimental example of children's influence on adults'
56 Warren et al.

use of behavior that in turn fosters the child's further development (Bell &
Harper, 1977). It supports the potential power of the transactional model, at
least during the early period of development when children's behavior reper-
toires are small and their developmental history relatively short. Further-
more, it suggests that RE/PMT can be highly effective with children under
some conditions. Alternatively, the second longitudinal study by Yoder and
Warren (2002) suggests that this approach may be ineffective with children
who have already attained a relatively high level of prelinguistic development.
The recent Fey et al. (in press) study used more conservative entry criteria
to ensure that a prelinguistic intervention was truly appropriate for the chil-
dren's communication levels and also included minor but perhaps important
modifications in the PMT intervention procedures. The main effect on inten-
tional communication resulting from the intervention implemented by Fey
et. al. suggests that these modifications were highly functional. Consequently,
. we present these same recommendations below.

PRACTICAL REQUIREMENTS

As noted, the RE/PMT we have been using and testing involves two compo-
~ nents. First, the interventionist (e.g., an SLP or teacher) must be able to
i~~or~ .. .Stve work on a one-to-one basis directly with the child several times per week
until the child has acquired the necessary skills to be a frequent, clear prelin-
guistic communicator. Depending on the child's developmental profile, this
may take anywhere from a few weeks to more than 6 months. In our clinical
and research experience, the average prelinguistic intervention takes sev-
eral months until the child achieves the exit criteria and goals can be shifted
to productive vocabulary.
The second component is RE for the child's parents. As discussed pre-
(!) Q ~~"~ "''( viously, a relatively high degree of parental respohsivity appears necessary
Pd C.1l to ensure that the direct training of the child's prelinguistic skills has maxi-
-e ~t.c'Jl "mal impact. We have found It Takes Two to Talk-The Hanen Program for
Parents (see Chapter 4; Manolson, 1992) serves as an excellent approach for
helping parents from many backgrounds to establish more responsive inter-
action patterns with their children. In general, we do~dvocate teaching
parents to use PMT procedures. Although these procednfes require the adult
to follow the child's lead and to be sensitive to the form and content of the
child's communicative efforts, they also involve consistent efforts to push the
child to higher levels of communication frequency and complexity. Many par-
ents are reluctant to take on .this role of teacher. Those who do sometimes
find it difficult to separate their direct instruction roles as teachers .and their
highly responsive roles as parents and communication facilitators. Further-
more, this intervention is only appropriate for a few months for most chil-
dren. However, a highly responsive parenting style that will naturally evolve
as children grow and develop is appropriate under most conditions and sup-
Responsivity Education!Prelinguistic Milieu Teaching 57

ports the child's development across a wide range of related domains (Landry,
Smith, Swank, Assel, & Vellet, 2001).

KEY COMPONENTS

Situations termed enabling contexts help to provide an optimal environ-

ment for highly resp~-child interaction and/or the use of spe- .,
cific teaching techniques. These contexts are the same for interventionists
using PMT and for parents who adopt the role of responsive communicator .'
~ in RE. The principles that help to create enabling conte~ts are 1) arrange the

1environment to increase opportunities for communicationt 2) follow the

child's attentionallead, and 3) build social routines in which the child and
adult play predictable roles. These principles are applied at all levels of the
intervention to encourage a high degree of engagement by the child and to
create frequent teaching interactions between the child and the adult. The
basic formats for implementing these principles were developed f~r use in
naturalistic early,language intervention approaches such as milieu teaching
(Warren, 1991) and responsive interaction training (Wilcox & Shannon, 1998).

Arranging the Environment

Children are most likely to initiate communicative acts about things they
need, want, or find novel and interesting (Hart & Risley, 1968). Arranging the
enviromnent so that it naturally supports the need to communicate can in-
crease the frequency of these states, thus giving the child more opportuni-
ties to communicate. This, in turn, gives the interventionist more opportuni-
ties to focus on the clarity and complexity of these acts. For example, adults ';
can place desired items (e.g., food, toys) either out of reach of the child or*P'-""t
in a contextin which adult assistance is necessary to access them. This often -rf, i"~.S
happens quite naturally in homes and child care environments in which the oc-t( o+
child's interests and patterns of action are readily observable. The environ- ~,..,
ment can then be arranged so that the child's expectations are challenged. In
a classroom context, certain toys might be kept in clear plastic containers
with lids on them that children cannot open without adult assistance (see
DVD Clip 12). Crayons might be placed on the floor next to the adult where
the child can see them but cannot easily reach them. Extra cupcakes might
be placed in clear view but beyond the child's reach. Alternatively, contain-
ers and shelves that typically bear the child's toys and objects of interest can
be emptied or filled with new, unanticipated items. Positioning should also
be considered when trying to create enabling contexts. Positioning refers to ~s--rf1> ,j'"J
how an adult places his or her body in relation to the child's body and a focal P · ' · "_j
object. To the extent possible, the clinician should ~d and JJI
focal object at the child's eye level. With infants and toddlers, this may mean~c...i
- ....___.,.__.,. ......
lrT~'l.c.
'\
jOlt'\[..
G..~
58 Warren et al.

the adult will need to lie on the floor with the child or sit on the floor while .
the child sits on a couch or a chair (see DVD Clip 4). This type of close, face-
to-face contact facilitates coordinated joint attention between the adult and
child (MacDonald, 1989). Sitting behinct;;?above ·the child makes this type
of interaction more difficult.

~:- Following the Child's Attentional Lead

-r·n
~~ G-:~tAj Young children attend more closely to objects or events of their choosing,
-r~-c.n~ rather than to objects or events of an adult's choosing (Bruner, Roy, & Ratner,
ll\. ~ 1980). Furthermore, young children have difficulty deploying their attention
on command for longer than very short periods (Goldberg, 1977). Thus, fol-
lowing the child's attentionallead, a universal tenet of virtually all naturalis-
tic early communication and language intervention approaches (Fey, 1986;
Hepting & Goldstein, 1996), is used to sustain the child's interest in activities
and social interaction. In practice, this might mean that the adult plays with
toys or engages in activities of interest to the child (typically selected by the
child from an array of choices) in a manner similar to the child's play. Chil-
dren who are passive and engage in low rates of action, or children who en-
gage in repetitive behavior, can make it _challenging to maintain-this proce-
dure. Adults can easily lapse into directive styles in which they dominate
most interaction episodes with the child. Our experience suggests that if the
goal is to build initiations, then it is far better to simply a9-apt one's behavior
to the child's initiation rate even if it is low. This technique is easier to im-
plement if the adult has made some effort to arrange the environment to in-
crease the likelihood that the child will attend and react spontaneously to
target stimuli in the intervention context.
-ree.II\Ri"'Q~lt} Contingent motor imitation is a technique that can be quite helpful
,Jy with a young child who seldom initiates (Gazdag & Warren, 2000). Contingent
" motor imitation is an exact, reduced, or slightly expanded imitation of the
· ·· child's motor production that is perf93fled by the adult immediately follow-
'· . · ' .. ing the child's motor prod~ion. It represents a specific form of following the
~ ~child's attentionallead. ~simple technique may be used at the start of in-
11 ° "':""" tervention to establish a basic form of turn taking between the child and
t"".:..1. ~ adult that over time can be transformed into interaction and play routines.
Contingent imitation may benefit children ~cause it allows them to regulate
Q> -ru. y- t"'\ - the amount of social stimulation received~ increases the probability that
'"t"A \c=- f~ adult illput will be easily processed and understood (Dawson & Lewy, 1989),
~t may enco)tJiage children to imitate adult behavior on a broader scale (Snow,
1989), ancNf may result in more differentiated play schemes (Dawson &
Adams, 1984).
,"t"c t..k ·t'\ iq LA-e.. Contingent v~ imitation offers many of the same advantages and ben-
':ft ~ efits as contingent motor imitation. It occurs when. adults follow children's
nonverbal vocalizations that are independent of communicative content with
YeS ro ¥\Olf J.......,..., T~"'C-.e- eVe~

1 ,;;- G. Is ~i>t i'rrt-e~~~.

Responsf .•ty Education/Prelinguistic Milieu Teaching 59

a partial, exact, or modified vocal imitation. For example, a child might vo-
calize [ga] while holding a plastic ring by her face, making no obvious attempt
to share the act with the adult. In this case, the adult might immediately im-
itate [ga], or [gaga], as a form of vocal play or turn taking (see DVD Clip 1).
This type of vocal imitation (as with motor imitation) allows children to reg-
ulate the amount of social stimulation they receive and may encourage chil-
dren to increase their rate of vocalization and to imitate adult vocalizations
spontaneously (Gazdag & Warren, 2000).

Social Routine Building

Social routines are repetitive, predictable turn-taking games and rituals,
such as Peekaboo and Pat-a-cake. Arranging the environment and following
the child's attentionallead support the development of social routines. Social
routines, in turn, provide an excellent context for facilitating social commu-
nication development. They can be established in the course of daily activi-
ties such as feeding, bathing, and dressing, as well as games and toy play. They
can be unconventional and unique to a given child. The predictable structure
of social routines 'may help children learn and remember new skills. OncJ
children learn predictable roles in a routine, they can devote greateratten- .
tion to analyzing adult models of new ways to communicate (Conti-Ramsden
& Friel-Patti, 1986; Nelson, 1989). Additionally, the effectiveness of models
may be enhanced because slight variations in the routine may create "mod-
erately novel" situations that are particularly salient to young children (Pi-
aget & Inhelder, 1969).
Research with children who are typically developing and with children
who have intellectual impairments has shown that social routines are particu-
larly powerful stimuli for linguistic (Snow, Perlmann, & Nathan, 1987; Yoder &
Davies, 1992) and prelinguistic communication (Bakeman & Adamson, 1984).
Once a social routine is well established with a child, adults can often elicit a
high rate of requests and comments by interrupting or modifying the routine.
Social routines also provide a natural context for modeling these communi-
cation functions and related skills such as turn taking.

Component 1: Prelinguistic Milieu Teaching

The overall goal of PMT is to help children establish and/or increase the fre-
quency, clarity, and complexity of their nonverbal communicative acts. Al-
though it is possible to categorize these acts in many ways, PMT focuses on
two broad types of acts: requests and comments. Requests are instrumental
acts in which the child seeks some object or action. In contrast, comments
are more purely social acts that seek only to share observations and experi-
ences with a partner.
In Table 3.1, we break down the basic goal of increasing the frequency,
clarity, and complexity of the child's nonverbal communicative acts into five
60 Warren et al.

Table 3.1. Prelinguistic milieu teaching procedures ~ 12.. fn">~ (.. ·

~~

1. Establish routines to A. Imitate the child's motor acts

serve as the context for
communicative acts (see B. Imitate the child's vocal acts
DVD Clip 1) c. Interrupt the child's established pattern of actions with an
adult turn, and then wait for the child to take a turn
D. Perform an action the child finds funny or interesting;
pause, then repeat to get more laughter
E. When the child produces one part of the routine, oblige by
performing the act needed to complete it
2. Increase the frequency of A. Recast the child's nonverbal vocalization with a word if the
nonverbal vocalizations child is focused on a clear referent
(see DVD Clips 2-4)
B. During vocal play activities (i.e., when the vocalizations are
not part of a communicative act), model vocalizations with
sounds and word shapes known to be outside the child's
Y}'\u'f e.,_ repertoire

~vr1r'~' c. Model a sound within the child's sound and word shape
repertoire when the vocalizations are not part of a commu-
we) nicative act
D. Imitate the child's spontaneous vocalizations with sounds
and syllable shapes known to be within the child's reper-
l Sc "'-(~\tAr t\j) toire when the vocalizations are not part of a communica~
tive act
E. Imitate the child's spontaneous vocalizations as precisely as
possible when the vocalizations are not part of a commu-
nicative act

3. Increase the frequency Create a need for communication within a routine in which
and spontaneity of co- the child looks at the object, then:
ordinated eye gaze (see
DVD Clips 5-7) A. Provide the child with the desired object or action contin-
gent on looking
B. Verbally prompt for eye gaze
C. Move the desired object to the adult's face to encourage a
more explicit look
D. Intersect the child's gaze by moving the adult's face into the
child's line of regard
E. Once the child complies, explicitly acknowledge the child's
look with fun and well-pleased affect
F. If, after using the methods above, the child fails to produce
the targeted act, provide the child with the desired object
or action

4. Increase the frequency, Create a need for communication within a routine (e.g., by
spontaneity, and range of placing a desired object out of reach), then:
conventional and non-
conventional gestures A. Provide the child with the desired object or action contin-
(see DVD Clips 8-10) gent on the use of a gesture
B. Pretend not to understand by looking and gesturing quizzi-
cally and saying "What?" or "What do you want?"
C. Ask or tell the child to be more specific (e.g., "Show me
which one!" "Which one do you want?")
D. Tell the child, explicitly, to produce a particular gesture
(e.g., "Show me!" "Give it to me!")
E. Model an appropriate gesture
Responsivity Education/Prelinguistic Milieu Teaching 61

F. Once the child complies, verbally acknowledge child's gesture

G. If, after using the methods above, the child fails to produce
the targeted act, provide the child with the desired object
or action
s. Combine components of A. If the child produces one or two components of a communi-
intentional communica- cation act, wait expectantly (i.e., use time delay) to prompt
tion acts. The three com- the second (or third) component
ponents of intentional
B. If the child produces one or two components of a communi-
communication acts are
cation act and does not add another component after the
eye contact with partner,
time delay:
vocalization, and gesture
(see DVD Clips 11-13) • Ask, "what do you want?" or another general prompt
and wait again
• Intersect the child's gaze or use the child's name to
prompt eye gaze
• Model or help the child to produce a gesture
• If the child has produced a communicative act that is fo-
cused clearly on an object, attribute, or event, the clini-
cian should recast the act by producing a word
• If the child produces components yielding a communica-
tive act, the clinician should not produce a nonverbal
model
• Immediately after the child produces the targeted com-
ponent, provide the appropriate consequence and ver-
bal feedback, as described under intermediate goals 1-4
above ·
• If, after using the methods above, the child fails to pro-
duce the targeted act, provide the child with the desired
object or action

intermediate goals. The first intermediate goal is to establish and maintain

social routines (as discussed in the preceding section). The next three in-
termediate goals are to increase the frequency and spontaneity of the three
basic components of .E{elinguistic requesting and commenting acts. These
basic components are nonlinguistic vocalizations, ~ordinated eye gaze, and
(!)conventional and unconventional gestures. Nonlinguistic vocalizations are
defined as either canonical (syllables with a true consonant, e.g., [ba]) or non-
canonical (syllables with only vowel-like sounds or glides, e.g., [a], [m:], [wa]).
Coordinated eye gaze refers to a child's alternating attention between an
object or event of interest and the adult. For example, a child might hear a
noise outside, look toward the window, and then look toward the adult. Un-
conventional gestures are hand movements that are highly contextual.
These include reaches, proximal points (touching an object with index finger
extended), pantomiming an action, giving objects to an adult, and moving
objects toward or away from an adult. Conventional gestures include in-
tentionally communicative acts, such as a headshake or nod, a hand wave
(e.g., hi), an upturned palm (e.g., gimme), or a point to a distal object. Fi-
nally, the fifth intermediate objective is to help the child combine these three
62 Warren et al.

nonlinguistic components into increasingly complex and clear requests and

conunents.
Three principle ~ are used to address all of the intermediate
goals and, ultimately, the overall goal of clear, frequent, intentional conunu-
nication (see Table 3.1). These are prompts, models, and natural conse-
quences.

Prompts Prompts are used to evoke intentional conununication at-

tempts by the child or to evoke specific components of intentional conunu-
nication (i.e., vocalizations, coordinated eye gaze, gestures). They take the
form of time delay, nonverbal prompts, and verbal prompts, all of which can
be used to encourage more frequent and/or complex nonverbal communica-
tion attempts. A time delay for initiation is a type of nonverbal prompt that
often functions as an interruption of an ongoing turn-taking routine. For ex-
ample, if a child and adult were playing a tickling game, the adult might hold
her hands away from the child to interrupt this routine, then look at the child
expectantly until the child initiates a request to continue the routine (see
DVD Clip 10}. Examples of other nonverbal prompts include holding out an
upturned palm (e.g., to get the child to give an object; see DVD Clip 8), mov-
ing an object of the child's interest near or directly in line with the adult's face
(e.g., to get the child to look at the adult), or physically intersecting the
child's line of regard in order for the child to look at the adult (see DVD Clip
6). Verbal prompts for conununication can be open-ended questions (e.g.,
"What?" see DVD Clip 12) or direct commands (e.g., "Show/tell me") in-
tended to elicit communication responses. Verbal prompts also can be used
to elicit a specific component of communication. For example, when a young
child makes a request without eye contact, the directive, "Look at me" or the
child's name may be used to evoke eye contact. If a child does not respond
to a prompt, or reacts aversively to a prompt, it is best not to persist with the
interaction.

Models Models are used to support and enhance the vocal and ges-
tural topography of the child's intentional communication attempts. Vocal
models of sounds that the adult has heard the child produce (e.g., [ba]) can
I.A~~cl r.-1o[ be used during~o19@that is not focused on a clear referent or is not
o.. c. f e c... v otherwise part of a child's communicative act. For example, while the child
v<..~ ~t is banging a stick, the clinician might model [baba]. To ensure that the child
does not misinterpret the adult's nonlinguistic vocalization as an actual label,
it is important that nonlinguistic models are only given in the absence of
clear referents. For example, if the adult models [bababa] while the child
points to a dog, the child may be induced to think that the label for dog is
[ba]. Gestural models are used to encourage the child to use and imitate ges-
tures. For example, when an airplane passes overhead, the adult might point
· :Res:Ponsivity Education/Prelinguistic Milieu Teaching 63

to it as a model for the child to use pointing as an element of commenting

(see DVD Clips 9 and 10 for additional examples).
'
!l
!
l

Providing Natural Consequences PMT is based on milieu teaching

(MT), which is a form of language intervention for children learning words
and early grammatical structures (Warren & Bambara, 1989). It differs most I
i:
dramatically from MT in that the targeted behaviors for PMT are nonverbal ·
acts; in contrast, in MT, the clinician will prompt and provide the child's de-
sired consequences only for verbal acts. Thus, for PMT, nonverbal commu-
nication attempts that are appropriately clear and complex should be conse-
quated in accordance with their intent: child requests should yield the desired
objects and actions, and child comments should result in adult attention to the
child's topic. Continued attention and interaction by the adult are assumed.
These natural consequences may be supplemented with specific ac-
knowledgment and/or by a verbal recast of the child's meaning (i.e.,~ lingl!!s-
tic mapping)~pecific acknowledgment is provided by a smile and comment ttl- _
after the· child produces a targeted intentional communication component. sp ec::.t f,c.
For example, when a child makes eye contact with a caregiver in the course G~.G-Pi)~
of initiating a request, the caregiver might break into a big smile and com-
ment, "You looked at me!" while responding to the child's request (see DVD
Clips 5, ·6, and 7). Frequent use of specific acknowledgment may disrupt the
flow of interaction, and praise statements tend to lose their meaning for re-
cipients if used too frequently. Therefore, these statements should be used
primarily when a child is first acquiring a new behavior and infrequently
thereafter.
0Linguistic mapping occurs when the adult verbalizes the core meaniilg'* .
of the immediately preceding child communication act. For example, a child L.;x • · · :
might point to a photo on a shelf for the adult to see (a comment), and the ~f,J
adult might respond, "Yes, that's Thomas" (see DVD Clip 9). Research with
both typically and atypically developing children has indicated that linguis-
tic mapping can be a powerful contributor to vocabulary development (Nel-
son, 1989). Therefore, we encourage the frequent use of linguistic mapping
as part of adult responses to intentional communication attempts.

Teaching Intentional Communication The specific pririciples and pro--.\'"

cedures we have described should be embedded into ongoing interactiOns f>a.s jc..
and used as dictated by the context and the child's current communication fr-~-plt"...S
goals. Some specific techniques may be used quite frequently (e.g., linguis- "v-e. ...
tic mapping); others, only until the child begins to intentionally use the tar-reet'"""
rt-4
geted skill (e.g., intersection of gaze). However, the enabling procedures of ""'~~
arranging the environment, following the child's attentionallead, and social ~peCAcj~t'1..
routine bUilding are to be used whenever possible and will continue to play ~ IMJ ~
an important tole when efforts shift toward language goals (Warren, 1991). · ~ ~ 0
orce:ouv--
64 Warren et al.

It is important to remember that the frequency and quality of children's en-

c.J.Ut~r<V'.J gagement with objects of interest and their conversational partners (i.e.,
~ (...-<.n their routines) will determine the occasions for the use of any specific tech-
nique. Prompts are used only when motivation to communicate is high (e.g.,
~-(~ tf"Yll\ ·when the child is intently engaged in social interaction). Additional conse-
-f"w,..e... ...,...._.s-e- quences can follow any intentional communication attempt. Specific teach-
-o-f S ye~ ing episodes should be brief, positive, and embedded in the ongoing stream
-t'e~ ~~..So£ interaction. If a child is not responsive to a prompt, it is best to desist and
move on with the interaction. The procedures outlined in Table 3.1 illustrate
a kind of hierarchy, with the techniques requiring the most sophisticated
child responses described first. Children who respond reliably to these high
level techniques may not require the procedures that fall lower in the hier-
archy. On the other hand, when a child does not respond to the higher-level
procedures, the clinician must fall back to the lower steps, providing the child
with more scaffolding. Once the child begins to respond reliably to this level
of assistance, the clinician gradually adopts higher level procedures that
place more responsibility on the child. This general method of utilizing the
steps in Table 3.1 is exemplified in the case example later in this chapter.

Teaching Proto-imperatives It is helpful to first establish social rou-

tines that involve turn taking between the adult and child (e.g., rolling a ball
back and forth, dropping blocks into a container, or playing a musical instru-
ment). Once an initial set of routines has been established and a particular
~ instance of a given routine has gone on for at least two turns, the adult may
.). stop the routine by withholding his or her turn and looking expectantly at the
~ S'loy child (a time delay for initiation). A verbal prompt also might be givenfsuch
~ V'0""1J',.t. as "What?" (to start the activity) or "Do you want this?" (while maintaining
.,... .-t eye contact and holding up an object the child needs to resume the activity).
Y-ro ~f If there is not an appropriate response to the interruption of the routine, or
if the child's response is incomplete (i.e., it is missing a component necessary
to be considered intentional communication), then the adult may provide
further assistance to the child. For example, if the child looks at a toy and
reaches toward it or provides a vocalization, but does not make eye contact
with the adult, then the adult might say, "Look at me," intersect the child's
gaze, or move the toy near the adult's face. The adult also might provide a
gestural model if needed to complete the communication act. Once the child
begins to request across different routines intentionally, then prompts, mod-
" [ els, and specific acknowledgments should be faded outji~
4·~- ~()..'/ ..- should continue as part of the adult's response to the requests, however.
'\............ .A .,./CA. ..... "
<T' . ~
~ llfCA.\l
Teaching Proto-declaratives Proto-declaratives are taught in a decid-
edly different manner from proto-imperatives (Yoder & Warren, 1999b). The
primary motivation for a proto-declarative is recruitment of another's atten-
tion and the sharing of an affective state. Our experience suggests that it is
65

often necessary for young children to first develop a positive relationship

with an adult befpre they will initiate proto-declaratives to the adult fre-
quently. Of course, proto-declaratives may be modeled at any time.
Proto-declaratives are taught by modeling and by providing situations
,;. likely to stimulate their use, or in some contexts by directly prompting with
physical assistance Cetf·• prompting a distal point by physically assisting the
child; see DVD Clip 9). One such situation is the introduction of novel events * i:
i!
or objects. This can take many forms such as adding new toys or items within i r-'t'ftd IAU
routines (which can be done frequently), taking advantage of occasional oc- Vtovel
.~ currences of silly or unusual events (planned or unplanned), a sabotaged rou- ev-e~"t ~
tine, a walk, or a ride in the car. The clinician can model a proto-declarative p..-e..-ce.~a..
by pointing to an object or novel event and directing the child to look. The . Y')"C- .s-ee
clinician should also linguistically map the object or event as a part of this 11/
modeling technique. The clinician can model proto-declaratives concurrent
with or just after the novel event of interest. On occasion, when something
novel occurs, the adult may pretend not to notice. The intent in this case is
to set up a situation in which the child feels the need to direct the adult's at-
tention to the object or event. Yet another approach is to seed an area with
interesting, novel objects and let the child have the run of the room for a few
minutes, but with the adult clearly observing from a short distance away. As
the child discovers novel items in this manner, the child may then comment
to the adult. However, our clinical observation has been that, by far, the ma-
jority of proto-declaratives occur about an object or event that the child has
reason to believe the adult is already attending to. That is, they occur in a
context in which the child and adult are engaged in joint attention (Carpen-
ter, Nagell, & Tomasello, 1998).
To summarize, it is important to remember that PMT procedures should
be embedded into ongoing social and play routines that represent the en-
abling contexts described earlier. Furthermore, procedures should be goal-
driven. Specific procedures may work well with some goals and poorly with
others. Also, clinicians should carefully monitor for and avoid any tendency
to use the procedures in a didactic or directive manner. The effectiveness of
these· procedures depends in part on having a high level of child engagement
that is best maintained by following the child's attentionallead at all times.

Component 2: Responsive Interaction Training l ~IT_)

There is considerable correlational evidence (Landry.et al., 1997; Murray &
Hornbaker, 1997; Smith, Landry, Miller-Loncar, & Swank, 1997) and some ex-
perimental evidence (Landry, Smith, & Swank, in press; Yoder, Warren, Mc-
Cathren, & Leew, ·1998) suggesting that an optimal style of parenting for
promoting social and cognitive competence is one that fosters reciprocal
interactions between parent and child. Such an interaction style allows the
child to have some degree of control over the interaction. Responsive inter-
action techniques (see Chapter 4; Wilcox & Shannon, 1998) are intended to
66 Warren et al.

create just such an optimal style in caregivers. This approach is widely used
in parent training. It Takes 'I\vo to Talk-The Hanen Program for Parents
~ e:......IL ~ [ (Manolson, 1992) is an excellent example of such an approach. Its major goal
Q is to increase the child's social communication skills by enhancing the qual-
ity of interaction between the adult and the child, and it has been shown to
be effective in helping parents of children with developmental delays reach
this goal (see Chapter 4).
Responsive interaction techniques also have been referred to as inter-
active modeling (Wilcox, Kouri, & Caswell, 1991). As with PMT, these tech-
niques require the provision of enabling contexts (e.g., following the child's
attention lead), described earlier in this chapter, to maximize their effec-
tiveness. Linguistic mapping also is strongly encouraged. However, respon-
sive interaction approaches generally discourage the direct elicitation of spe- ·
cific child responses via requests to imitate, or even in some cases the use of
test questions (e.g., "What is that?"). Focused input is provided based on the
child's attentional lead. This input may include models in the form of de,.
scriptive talk or linguistic mapping .
. In our current research, parents receive 8-10 sessions of RE over a 6-
rnonth period. This training is provided by an SLP who has been trained to
conduct The Hanen Parent Training Program. However, our approach repre-
sents an adaptation of The Hanen Program, not an attempt to directly repli-
cate it. The training sessions are conducted in the parent's horne. The initial
goal for these sessions is to develop a sense of trust between the clinician and
the parent. In our experience, if this trust is not established initially, it can be
difficult for the parent to accept the information or apply it to his or her in-
teractions with the child. The clinician and the parent spend some time just
getting to know each other through conversation that does not necessarily
focus on the child or the intervention. Self-disclosure on the clinician's part
and listening to the parent regardless of the topic serve as effective tools to
establish the environment for open and honest communication.
Once the clinician-parent relationship is established, the focus of the
sessions moves to the direct teaching of responsive interaction techniques.
In order to implement these techniques, it is imperative that the parent have
a clear understanding of his. or her child's intentional communication. To il-
lustrate this point, the clinician may point out instances of intentional com-
munication from videotaped PMT sessions with the child. Parents begin RIT
with varied skill levels in terms of how they interact with their child. For ex-
ample, some parents are proficient in following their child's lead but struggle
· with allowing their child adequate time to communicate. Other parents ex-
perience difficulty following their child's lead during play activities, whereas
a few parents are proficient in most of the responsive interaction techniques
before beginning the intervention and need just a little fine tuning. In our ex-
perience, the most difficult technique for many parents to apply is allowing
their child adequate time to communicate. Videotaped sessions of the parent
Responsivity Education!Prelinguistic Milieu Teaching 67

and child playing together are viewed so the clinician and parent can iden-
tify instances of high and low responsivity to child initiations. These viewings
should be interaction and parent driven to the extent possible to maintain
the team-based relationship rather than the teacher-student relationship.
The clinician provides cues to direct the parent to specific instances of high
or low responsivity.

ASSESSMENT METHODS TO SUPPORT ONGOING DECISION MAKING

Children's communication skills will advance to the point where PMT is no

longer the appropriate intervention. To determine a child's readiness to
move from PMT to an early language intervention such as MT, several factors"!,"'~ .. -r-o
must be considere~A decrease in the amount of prompting needed is often VV" o ve..
the first indication that a child may be ready for language intervention. In ad- -t'b ...,.-a,...e
~ clition, the overall rate of intentional communication should be approaching ~x -c
:(, 2 acts per minute in a social interaction play situation. This can be deter:- .S t"~?
t- f>..V ~g the child interact with the parent or another adult and
(0 ~.f ..v-tounting the !!umber.2Lin_!~al communication. acts over a 5- or 10-
~ ~ minute period. These acts should include proto-declaratives as well as proto-
imperatives. Most communicative acts should include all three components
(gazes, gestures, and canonical vocalizations). Our impressionis that children
with high rates of gesture and mutual gaze are appropriate candidates for
language intervention, even if their rates of canonical vocalization are very
low. This observation is consistent with Yoder and Warren's (2002) finding
that children who produced high rates of communication acts prior to inter-
vention responded better without PMT than they did with it. For many such
children with limited phonetic capabilities, this will be the appropriate time
to make a decision whether an augmentative or alternative communication
system would be best for the child and, if so, what kind of system would likely
be most effective.
During the course of PMT, the clinician should also note an increase in
the range of consonants and vowels in the child's phonetic inventory and the
use of a large variety of gestures, including distal point, contact point, give,
reach, wave, as well as other nonconventional gestures. These factors can be
determined during the course of an intervention session as opposed to con-
ducting a lengthy evaluation.

CONSIDERATIONS FOR CHILDREN FROM CULTURALLY

AND LINGUISTICALLY DIVERSE BACKGROUNDS

RE/PMT and the two intervention components described earlier reflect a set
of biases about social communication development and the appropriate roles
of caregivers and practitioners. The acceptability of these procedures, and
hence their ultimate effectiveness, may vary in some cases because of differ-
68 Warren et al.

ences in cultural values and beliefs (Johnston & Wong, 2002; van Kleeck ,
1994}. Early social-cormnunication intervention may be even more susceptible
to problems associated with cultural differences than other forms of interven-
tion for two reasons. First, it often takes place within the family context and
carries an expectation that the caregivers will play an active role and even
adopt a style of interaction with their child that may directly violate some of
their views of appropriate parent-child interaction (Bornstein, 1989}. Sec-
ond, the focus on cormnunication and language development and differences
is inherently one of the most sensitive areas for cross-cultural discourse. A
range of basic SES and ethnic differences is frequently manifested in lan-
guage differences (Heath, 1986}. Furthermore, even a basic goal such as in-
crease the child's rate of communicative initiations can be problematic.
For example, in her study of the Inuit in northern Canada, Crago (1990}
found that "talkativeness" by young children was considered a sign of a
"learning problem" by their parents and was discouraged.
We presume that many if not most potential sources of bias can be lim-
ited or at least identified through the careful collection and consideration of
information on individual family values, beliefs, and desires. This information
then can be used to modify intervention strategies to enhance their accept-
ability and thus their ultimate effectiveness. For example, in the Inuit culture
mentioned previously, it may be most appropriate to involve older siblings or
other caregivers in RE rather than parents. Older youths to whom children
might be expected to speak frequently also might be trained to perform PMT
under an SLP's supervision. A thorough consideration of individual family
differences should be a given with all families, irrespective of their cultural
or ethnic background. Thus, embracing this perspective should place no ad-
ditional burden on practitioners. It is in fact completely congruent with the
notion of individualizing efforts to meet the unique needs of the family and
child, a widely held tenet of early intervention practices in many countries
and cultures (Odom & McLean, 1996).

APPLICATION TO AN INDIVIDUAL CHILD

Bonnie, age 27 months, was born 14 weeks prematurely. Her cognitive skills
were at t~e 13-month level as determined by the Bayley Scales of Infant De-
velopment (Bayley, 1993}. Bonnie's motor skills were also delayed; she was
unable to walk but sat unassisted. She grasped objects in either hand and
transferred objects from one hand to the other. During the initial communi-
cation assessment, Bonnie produced an average of 1 vocalization per minute;
however, these vocalizations were not directed to an adult and typically were
judged to be noncanonical syllables because they did not contain a true con-
sonant. Her rate of canonical vocalizations was 0.51 per minute. Bonnie's
gestures included holding up her arms to request being lifted by an adult and
reaching for objects. When she could not reach an object she wanted, she vo-
69

calized in protest but did not look to an adult for assistance. She averaged
0.21 proto-declaratives and 0.76 proto-imperatives per minute.
Bonnie was enrolled in PMT for four 20-minute sessions weekly. RE
was also provided to her mother 1-2 times per month. The first intermedi-
ate goal for Bonnie was to establish tum-taking routines to serve as a context
for conununication (see intermediate goal 1, Table 3.1). Bonnie enjo"yed
shaking and patting musical instruments, so the clinician began by imitating
Bonnie's actions (specific techniq~ 1A, Table 3.1). The clinician continued
this activity with a variety of toys Ml.til Bonnie noticed the clinician's actions
and began to vary her own actions more frequentl$[9rhe clinician then intro-
duced turn taking by only having one toy available; the clinician played with
the toy in a way she had seen Bonnie play with it, then immediately moved
the toy to lie within her range of grasp. Once Bonnie had played with the toy
for several seconds, the clinician took the toy and played with it for a few sec-
onds. The clinician then placed the toy near Bonnie again (specific technique
1C, Table 3.1).
Once this routine was firmly established, the clinician moved on to in-
termediate goals 2-4 (see Table 3.1), which address the individual compo-
nents of proto-imperatives (i.e., requesting) and proto-declaratives (i.e., com- I. ~
menting). Bonnie did not use alternating gaze yet, so intermediate goal3 was €JfA ~K\J
targeted first. The clinician lifted a desired toy close to her face so that Bon- -e-o-e.
nie did not have to look far to make eye contact (specific technique 3C, Table. a-c-
z......e....
3.1). It was often necessary to intersect Bonnie's line of vision as well (spe-
cific technique 3D, Table 3.1). As soon as Bonnie alternated her gaze from
the toy to the clinician, the clinician praised her for looking and gave her the
toy (specific technique 3E, Table 3.1). Eventually the clinician was able to
hold the toy farther away from her face, requiring Bonnie to look from the
toy back to the adult.
Although alternating eye gaze (i.e., intermediate goal3, Table 3.1) was
targeted during the initial few sessions, once predictable routines had been
established, intermediate goals 2-4 were addressed concurrently during
each successive session. Different established routines allowed for elicitation
of different individual components. To encourage canonical vocalizations
(i.e., intermediate goal2, Table 3.1), the clinician began by imitating Bonnie's
vocalizations, often while she held a toy, such as a Slinky or a tube, up to her
face (specific technique 2E, Table 3.1). This use of the toy made the vocal
play activity more of a game and directed Bonnie's attention to the clinician's
mouth. The clinician vocalized into the Slinky, stacking ring, or cup, then
handed the toy back to Bonnie for her to take a turn. In later sessions, the
clinician followed Bonnie's vocalization with one or more syllables that dif-
fered from Bonnie's by adding consonant sounds. At first, these sounds were
those Bonnie sometimes used (specific technique 3C, Table 3.1). Later, new
sounds not yet in Bonnie's babbling repertoire were added (specific tech-
nique 3B, Table 3.1). To address intermediate goal4 (i.e., use of gestures),
70 Warren et al.

*t-oll\ t" £,.."C vs. the clinician employed the enabling context of arranging the environment so
d151""&\ I ~ that desired toys were in sight but out of reach. The clinician then modeled
both contact points and distal points and encouraged Bormie to produce these
pe:. 'f)'\ -r: to request the toys (specific technique 4E, Table 3.1). Proto-declaratives
were also targeted in this way by having a toy perform an unexpected action.
The clinician then modeled a distal point and said, "Look." To encourage the
gesture of a give, toys were placed in clear bags or jars that were difficult to
open. It was often necessary for the clinician to prompt for a give by extend-
ing her hands and asking Bonnie if she needed help (specific technique 4E,
Table 3.1). After 2 months, the clinician no longer needed to model a gesture;
she simply asked Bonnie to show her which toy she wanted (specific tech-
nique 4C, Table 3.1).
Once Bonnie began to readily produce each individual component of an

~o.D-*5
intentional communication act, the ·clinician moved on to intermediate goal
5 (see Table 3.1). For example, when Bonnie alternated her gaze from a de-
vt>I""Y'"' "C> sired toy to the clinician, the clinician prompted for a canonical vocalization
b-rIM~~ by saying "What?" As soon as Bonnie produced a canonical vocalization,
·.the clinician gave her what she wanted and labeled the object ("Oh, you
want the Slinky"). Similarly, if Boi:mie used a.._contact point to request an ob-
ject, the clinician called her name to prompt for an alternating gaze to ac-
company the gesture. The object was supplied only after Bonnie looked up
at the clinician.
After 6 months of PMT, Bonnie was producing intentional communica-
tion acts at a rate of 2.0 per minute to make both proto-imperatives and proto-
declaratives. The number of prompts that were required and the length of
time for waiting during the time-delay technique had decreased greatly. Bon-
nie's rate of canonical vocalizations increased to 1.3 per minute. At this time,
it was determined that she met the criteria to move into language interven-
tion, and her goals shifted to productive word acquisition and use.

FUTURE DIRECTIONS

Relative to early language intervention, very little research has been con-
ducted on RE/PMT. Furthermore, although this approach clearly holds great
potential, we do not yet know its true value or what type of children benefit
the most from it.
We are presently conducting a longitudinal experimental study to de-
termine whether RE/PMT generates a great enough impact on long-term de-
velopment of young children with developmental delays to warrant its wide-
spread clinical and educational application. In this study, we are comparing
two groups of 2-year-olds with developmental disabilities. In one group, the
children receive 6 months of PMT and the parents receive RE. The other
group receives no treatment through our project. Twelve months after entry
into our program, all children receive 6 months of milieu language interven-
71

tion to supplement what they get through the schools. We are interested in
comparing the performance of children in these two groups at each 6-month
mterval in terms of their rates of communication acts and, at the later time
points, their language abilities. We will have the results of this research in 2006.
Meanwhile, Paul Yoder is conducting an analysis of the effects of RE/PMT
with young children with autism. These studies should go a long way toward
indicating the potential of RE/PMT in general. However, there is already clear
evidence of its efficacr for at least some individuals, most notably those that
do not communicate very frequently at the outset and who have highly re-
sponsive parents.
Most of the research on RE/PMT has been conducted by only a handful
of individuals in a few locations. Reliable knowledge of the effects and effec-
tiveness of RE/PMT as well as the development of a full range of specific in-
tervention procedures will require an expanded effort conducted by addi-
tional investigators in different settings with varied populations. Finally, the
question of whether increasing .the intensity of RE/PMT will generate sub-
stantial increases in its effects remains to be answered. The prelinguistic in-
tervention we have described in this chapter involves 60-80 minutes per
week of the cliniCian actually working one on one with the child. The mini-
mal intensity of this intervention should be obvious. A recent report by a
committee of the National Research Council (2001) suggests that young chil-
dren with autism receive 25 hours of direct intervention per week to achieve ·
maximal effects. We have no idea what the optimal intensity of RE/PMT for
children with developmental delays might be, but surely it is more than 80
minutes per week. Consequently,·in 2005 we (Warren, Fey, and Yoder) began
a 5-year longitudinal experimental intervention study of RE/PMT with ran-
dom assignment to high-intensity (5 hours of direct intervention per week)
and low-intensity (1 hour per week) conditions with the support of the Na-
tional Institute of Deafness and Other Communicative Disorders.

RECOMMENDED READING
Wetherby, A.M., Warren, S.F., & Reichle, J. (Vol. Eds.). (1998). In S.F. Warren & J. Reichle
(Series Eds.), Communication and language intervention series: Vol. 7. Transi-
tions in prelinguistic communication. Baltimore: Paul H. Brookes Publishing Co.

REFERENCES
Ainsworth, M.S., Blehar, M.C., Waters, E., & Wall, S. (1978). Patterns ofattachment: A psy-
chological study of the strange situation. Hillsdale, NJ: Lawrence Erlbaum: Associates ..
Bailey, D.B., Aytch, L.S., Odom, S.L., Symons, F., & Wolery, M. (1999). Early intervention
as we know it. Mental Retardation and Developmental Disabilities Research Re-
views, 5, 11-20.
Bakeman, R., & Adamson, L.B. (1984). Coordinating attention to people and objects in
mother-infant and peer-infant interaction. Child Development, 55(4), 1278-1289.
72 Warren et al.

Bates, E., Benigni, L., Bretherton, I., Camaioni, L., & Volterra, V. (1979). The emergence of
symbols: Cognition and communication in infancy. New York: Academic Press.
Bayley, N. (1993). Bay(ey Scales of Infant Development (2nd ed.). San Antonio, TX:
Harcourt Assessment.
Bell, R., & Harper, L. (1977). Child effects on adults. Hillsdale, NJ: Lawrence Erlbaum
Associates.
Bornstein, M.H. (1989). Maternal responsiveness: Characteristics and consequences.
New directions for child development. San Francisco: Jossey-Bass.
Brooks-Gunn, J., & Lewis, M.H. (1984). The development of early visual self-recognition.
Development Review, 4(3), 215-239.
Bruner, J.S., Roy, C., & Ratner, R. (1980). The beginnings of request. In K.E. Nelson (Ed.),
Children's language (Vol. 3, pp. 91-138). New York: Gardner Press.
Carpenter, M., Nagell, K., & Tomasello, M. (1998) . .Social cognition, joint attention, and
communicative competence from 9-15 months of age. Monographs of the Society for
Research in Child Development, 63(4), v-174.
Conti-Ramsden, G., & Friel-Patti, S. (1986). Mother-child dialogues: Considerations of
cognitive complexity for young language learning children. British Journal of Disor-
ders of Communication, 21(2), 245-255.
Crago, M. (1990). Development of communicative competence in Inuit children: Implica-
tions for speech-language pathology. Journal of Childhood Communication Disor-
ders, 13, 73-83.
Crawley, S.B., & Spiker, D. (1983). Mother-child interactions involving two-year olds
with Down syndrome: A look at individual differences. Child Development, 54(5),
1312-1323.
Dawson, G., & Adams, A. (1984). Imitation and social responsiveness in autistic children.
Journal of Abnormal Child Psychology, 12(2), 209-225. ·
Dawson, G., & Lewy, A. (1989). Arousal, attention and the socioemotional impairments of
individuals with autism. In G. Dawson (Ed.), Autism: Nature, diagnosis, and treatment
(pp. 49-74). New York: Guilford Press.
Farran, D. (2000). Another decade of intervention for c)lildren who are low income or dis-
abled: What do we know now? In J. Shonkoff & S. Meisels (Eds.), Handbook of early
childhood intervention (2nd ed., pp. 510-548). Cambridge, England: Cambridge Uni-
versity Press.
Feagans, L., & Farran, D.C. (1982). The language of children reared in poverty: Im-
plications for evaluation and intervention. New York: Academic Press.
Penson, L., Dale, P., Reznick, S., Thai, D., Bates, E., Hartung, J., Pethick, S., & Reilly, J.
(1993). The MacArthur-Bates Communicative Development Inventories: User's
guide and technical manual. Baltimore: Paul H. Brookes Publishing Co.
Fey, M. (1986). Language intervention with young children. Austin, TX: PRO-ED.
Fey, M.E., Warren, S.F., Brady, N.C., Finestack, L., Bredin-Oja, S., Sokol, S., Fairchild, M.,
& Yoder, P.J. (in press). Early effects of prelinguistic milieu teaching and responsivity
education for children with developmental delays and their parents. Journal of Speech,
Language, and Hearing Research.
Gazdag, G.A., & Warren, S.F. (2000). Effects of adult contingent imitation on development
of young children's vocal imitation. Journal of Early Intervention, 23, 24-35.
Goldberg, S. (1977). Social competence in infancy: A model of parent-infant interactions.
Merrill-Palmer Quarterly, 23, 163-177.
Gottfried, A.W. (1984). Home environment and early cognitive development: Integra-
tion, meta-analyses, and conclusions. In A.W. Gottfried (Ed.), Home environment
and early cognitive development: Longitudinal research (pp. 329-342). New York:
Academic Press.
Hart, B., & Risley, T. (1968). Establishing use of descriptive adjectives in the spontaneous
speech of disadvantaged preschool children. Journal ofApplied Behavior Analysis,
1 (2), 109-120.
73

Hart, B., & Risley, T. (1992). American parenting of language-learning children: Persist-
ing differences in family child interactions observed in natural home environments. De-
velopmental Psychology, 28(6), 1096-1105.
Hart, B., & Risley, T. (1995). Meaningful differences in the everyday experience of
young American children. Baltimore: Paul H. Brookes Publishing Co.
Heath, S.B. (1986). Taking a cross-cultural look at narratives. Topics in Language Dis-
orders, 7(1), 84-94.
Hepting, N.H., & Goldstein, H. (1996). What's "natural" about naturalistic language inter-
vention? Journal of Early Intervention, 20, 250-265.
Johnston, J.R., & Wong, M. (2002). Cultural differences in beliefs and practices concern-
ing talk to children. Journal of Speech, Language, and Hearing Research, 45(5),
916-926.
Kaiser, A.P., Yoder, P.J., & Keetz, A. (1992). Evaluating milieu teaching. In S.F. Warren &
J. Reichle (Series & Vol. Eds.), Communication and language intervention series:
Vol. 1. Causes and effects in communication and language intervention (pp. 9-48).
Baltimore: Paul H. Brookes Publishing Co.
Landry, S.H., Smith, K.E., Miller-Loncar, C.L., & Swank, P.R. (1997). Predicting cognitive-
language and social growth curves from early maternal behaviors in children at varying
degrees of biological risk. Developmental Psychology, 33(6), 1040-1053.
Landry, S.H., Smith, K.E., & Swank, P.R. (in press). Responsive social contexts: The origins
of early communication and independent problem solving. Developmental Psychology.
Landry, S.H., Smith,' K.E., Swank, P.R., Assel, M.A., & Vellet, S. (2001). Does early re-
sponsive parenting have a special importance for children's development or is consis-
tency across early childhood necessary? Developmental Psychology, 37(3), 387-403.
MacDonald, J, (1989). Becoming partners with children: From play to conversation.
San Antonio, TX: Special Press.
Malloney, G., Boyce, G., Fewell, R.R., Spiker, D., & Wheeden, C.A. (1998). The relation-
ship of parent-child interactions to the effectiveness of early intervention services for
at-risk children and children with disabilities. Topics in Early Childhood Special Edu-
cation, 18(1), 5-17.
Manolson, A. (1992). It Takes Two to Talk. Toronto: The Hanen Centre.
McCathren, R.B., Warren, S.F., & Yoder, P.J. (1996). Prelinguistic predictors oflater lan-
guage development. In S.F. Warren & J. Reichle (Series Eds.) & K.N. Cole, P.S. Dale, &
D.J. Thai (Vol. Eds.), Communication and language intervention series: Vol. 6. Ad-
vances in assessment of communication and language (pp. 57-76). Baltimore:
Paul H. Brookes Publishing Co.
McCathren, R.B., Yoder, P.J., & Warren, S.F. (1999). Prelinguistic pragmatic functions as
predictors of later expressive vocabulary. Journal ofEarly Intervention, 22, 205-216.
McCathren, R.B., Yoder, P.J., & Warren, S.F. (2000). Testing predictive validity of the com-
.. , munication composite of the Communication and Symbolic Behavior Scales. Journal
;y of Early Intervention, 23, 36-46.
rz, McLean, J.E., & Snyder-McLean, L. (1978). A transactional approach to early lan-
guage training. Columbus, OH: Charles E. Merrill.
nt Murray, A.D., & Hornbaker, A.V. (1997). Maternal directive and facilitative interaction
styles: Associations with language and cognitive development of low risk and high risk
S. toddlers. Development and Psychopathology, 9(3), 507-516.
National Research Council. (200 1). Educating children with autism. Committee on Ed-
ra- ucational Interventions for Children with Autism. Division of Behavioral Social Sciences
mt and Education. Washington, DC: National Academy Press.
rk: Nelson, K.E. (1989). Strategies for first language teaching. In M.L. Rice & R.L. Schiefel-
busch (Eds.), The teachability of language (pp. 263-31 0). Baltimore: Paul H. Brookes
ous Publishing Co.
is, Odom, S.L., & McLean, M.E. (1996). Early intervention/early childhood special edu-
cation recommended practices. Austin, TX: PRO-ED.
74 Warren et al.

Piaget, J., & Inhelder, B. (1969). The psychology ofthe child. New York: Basic Books.
Prizant, B., & Wetherby, A.M. (1990). Toward an integrated view of early language and
communication development artd socioemotional development. Topics in Language
Disorders, 10(4), 1-16.
Reynell, J.K., & Gruber, C.P. (1990). The ReynellDevelopmental Language Scales, U.S.
Edition. Los Angeles: Western Psychological Services.
Rosenberg, S., & Abbeduto, L. (1993). Language and communication in mental re-
tardation: Development, processes, and intervention. Hillsdale, NJ: Lawrence Erl-
baum Associates.
Sameroff, A.J., & Chandler, M.J. (1975). Reproductive risk and the continuum of care-
taking casualty. In F.D. Horowitz, E. Hetherington, S. Scarr-Salapatek, & G. Siegel
(Eds.), Review of child development research (Vol. 4, pp. 187-244). Chicago: Uni-
versity of Chicago Press.
Saxon, T., Colombo, J., Robinson, E.L., & Frick, J.E. (2000). Dyadic interaction profiles in
infancy and preschool intelligence. Journal of School Psychology, 38, 9-25.
Smith, K.E., Landry, S.H., Miller-Loncar, C.L., & Swank, P.R. (1997). Characteristics that
help mothers maintain their infants' focus of attention. Journal of Applied Develop-
mental Psychology, 18(4), 587-601.
Snow, C.E. (1989). Imitativeness: A trait or a skill? In G.E. Speidel & K.E. Nelson (Eds.),
The many faces of imitation in language learning. New York: Springer-Verlag.
Snow, C.E., Perlmann, R., & Nathan, D. (1987). Why routines are different: Towards a
mUltiple factors model of the relation between input and language acquisition. In K.E.
Nelson & A. van Kleeck (Eds.), Child language (Vol. 6, pp. 65-97). Hillsdale, NJ:
Lawrence Erlbaum Associates.
Tamis-LeMonda, C.S., Bomstein, M.H., Baumwell, L., & Melstein Damast, A. (1996). Re-
sponsive parenting in the second year: Specific influences on children's language and
play. Early Development and Parenting, 5(4), 173-183.
Tannock, R., Girolametto, L., & Siegel, L.S. (1992). Language intervention with children
who have developmental delays: Effects of an interactive approach. American Jour-
nal on Mental Retardation, 97, 145'-160.
van Kleeck, A. (1994). Potential cultural bias in training parents as conversational part-
ners with their children who have delays in language development. American Journal
of Speech-Language Pathology, 31, 67-78.
Walker, D., Greenwood, C., Hart, B., & Carta, J. (1994). Prediction of school outcomes
based on early language production and socioeconomic factors. Child Development,
65(2 Spec No), 606-621.
Warren, S.F. (1991). Enhancing communication and language development with milieu
teaching procedures. In E. Cipani (Ed.), A guide for developing language compe-
tence in preschool children with severe and moderate handicaps (pp. 68-93).
Springfield, IL: Charles C Thomas.
Warren, S.F., & Bambara, L.M. (1989). An experimental analysis of milieu language inter-
vention: Teaching the action-object form. Journal of Speech and Hearing Disorders,
54(3), 448-461.
Warren, S.F., & Yoder, P.J. (1998). Facilitating the transition to intentional communica-
tion. In S.F. Warren & J. Reichle (Series Eds.) & A.M. Wetherby, S.F., Warren, & J. Re-
ichle (Vol. Eds.), Communication and language intervention series: Vol. 7. Tran-
sitions in prelinguistic communication (pp. 365-385). Baltimore: Paul H. Brookes
Publishing Co.
Warren, S.F., Yoder, P.J.,Gazdag, G.E., Kim, K., & Jones, H.A. (1993). Facilitating prelin-
guistic communication skills in young children with developmental delay. Journal of
Speech and Hearing Research, 36(1), 83-97.
Wetherby, A., Allen, L., Cleary, J., Kublin, K., & Goldstein, H. (2002). Validity and reliabil-
ity of the communication and symbolic behavior scales developmental profile with very
75

young children. Joumal of Speech, Language, and Hearing Research, 45(6),

1202-1218.
Wetherby, A.M., & Prizant, B.M. (1993a). Communication and Symbolic Behavior
Scales™ (CSBS). Baltimore: Paul H. Brookes Publishing Co.
Wetherby, A.M., & Prizant, B.M. (1993b). Profiling communication and symbolic abilities
in young children. Joumal of Childhood Communication Disorders, 15(1), 23-32.
WJ.lcox, M., Kouri, T., & Caswell, S. (1991). Early language intervention: A comparison of
classroom and individual treatment. American Joumal of Speech-Language Pathol-
ogy, 1, 49-62.
Wilcox, M., & Shannon, M. (1998). Facilitating the transition from prelinguistic to lin-
guistic communication. In S.F. Warren & J. Reichle (Series Eds.) & A.M. Wetherby, S.F.
Warren, & J. Reichle (Vol. Eds.), Communication and language intervention series:
Vol. 7. Transitions in prelinguistic communication (pp. 385-416). Baltimore: Paul
H. Brookes Publishing Co.
Yoder, P.J., & Davies, B. (1992). Do children with developmental delays use more frequent
and diverse language in verbal routines? American Joumal on Mental Retardation,
97(2), 197-208.
Yoder, P.J., Davies, B., & Bishop, K. (1994). Reciprocal sequential relations in conversa-
tions between parents and children with developmental delays. Joumal of Early In-
tervention, 18(4), 362-379.
Yoder, P.J., & Warren, S.F. (1998). Maternal responsivity predicts the extent to which
prelinguistic intervention facilitates generalized intentional communication. Joumal
ofSpeech, Language, and Hearing Research, 41(5), 1207-1219.
Yoder, P.J., & Warren, S.F. (1999a). Maternal responsivity mediates the relationship be'-
tween prelinguistic intentional communication and later language. Joumal of Early
Intervention, 22, 126-136.
Yoder, P.J., & Warren, S.F. (1999b). Self-initiated proto-declaratives and proto-imperatives
can be facilitated in prelinguistic children with developmental disabilities. Joumal of
Early Intervention, 22,337-354.
Yoder, P.J., & Warren, S.F. (2001). Relative treatment effects of two prelinguistic commu-
nication interventions on language development in toddlers with developmental delays
vary by maternal characteristics. Journal of Speech, Language, and Hearing Re-
/ search, 44(1), 224-237.
V Yoder, P.J., & Warren, S.F. (2002). Effects ofprelinguistic milieu teaching and parent re-
sponsivity education on dyads involving children with intellectual disabilities. Joumal
of Speech, Language, and Hearing Research, 45(6), 1158-1174.
Yoder, P.J., Warren, S.F., & Hull, L. (1995). Predicting children's response to prelinguistic
communication intervention. Joumal of Early Intervention, 19(1), 74-84.
Yoder, P.J., Warren, S.F., Kim, K., & Gazdag, G.E. (1994). Facilitating prelinguistic com-
munication skills in young children with developmental delay. II: Systematic replication
and extension. Journal of Speech and Hearing Research, 37(4), 841-851.
Yoder, P.J., Warren, S.F., McCathren, R.B., & Leew, S. (1998). Does adult responsivity to
child behavior facilitate communication development? In S.F. Warren & J. Reichle (Se-
ries Eds.) & A.M. Wetherby, S.F. Warren, & J. Reichle (Vol. Eds.), Communication
and language intervention series: Vol. 7. Transitions in prelinguistic communi-
cation (pp. 39-58). Baltimore: Paul H. Brookes Publishing Co.