See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/275159431

Recommending Intervention for Toddlers With Specific Language Learning

Difficulties

Article in American Journal of Speech-Language Pathology · February 1998

DOI: 10.1044/1058-0360.0701.23

CITATIONS READS
115 2,359

3 authors:

Lesley B Olswang Barbara L Rodriguez

University of Washington University of New Mexico
76 PUBLICATIONS 2,296 CITATIONS 29 PUBLICATIONS 909 CITATIONS

SEE PROFILE SEE PROFILE

Geralyn R Timler
James Madison University
51 PUBLICATIONS 986 CITATIONS

SEE PROFILE

All content following this page was uploaded by Geralyn R Timler on 10 March 2017.

The user has requested enhancement of the downloaded file.

 Clinical Focus

Recommending Intervention for Toddlers With

Specific Language Learning Difficulties: We May
Not Have All the Answers, But We Know a Lot

Lesley B. Olswang
Barbara Rodriguez
Geralyn Timler
University of Washington, Seattle

This paper presents a review of the literature toddlers who exhibit few positive predictors of
designed to identify child behaviors that shape change and many risk factors are more likely to
a profile of toddlers who should receive inter- have a true impairment and need intervention
vention. The review presents empirically docu- than toddlers who exhibit many predictors of
mented predictors of language change and risk change and few risk factors. The review at-
factors for language impairment. It examines tempts to paint a profile of toddlers for whom
research addressing the children having diffi- treatment should be recommended and those
culty learning language and children develop- for whom a watch and see approach should be
ing typically. The argument presented is that followed.

I ntroduction 1996). Although the children are perceived as a cohesive

group, the heterogeneity among them is obvious as one
The Problem
discovers discrepant research findings or one sees these
It has become increasingly clear that most toddlers who children in clinical settings. Certainly the continued debate
are identified as late talkers or as having specific language about diagnosis and prognosis and the apparent heteroge-
impairment (SLI), specific expressive language impairment neity among the children serve as major catalysts for the
(SELI), or slow expressive language development (SELD) question about whether to recommend early intervention.
gradually move within normal limits in language perfor- In addition, parents continue to ask for intervention, and
mance during the first years of school. Research by several clinicians (and researchers) find themselves recommending
teams of investigators, including Paul, Rescorla, Thal, and and providing the services based on their desire to help.
Whitehurst, have contributed to this body of literature, and For example, Paul, who advocates a policy of watch and
the results have been surprisingly similar (see Paul, 1996; see for children with SELD, has stated that if her own
Thal & Katich, 1996, for review). Although this is the case, child’s language was at risk, she would obtain “assistance
recommendations concerning intervention remain confus- now even if the chances were very good that the problem
ing. The confusion seems to reflect wariness about a would be outgrown sooner or later” (Paul, 1996, p. 6). This
“watch and see” (Paul, 1996) approach for such young paper will argue that sufficient data are available for
children, disagreement about the actual diagnosis and clinicians to make informed decisions about whether or not
underlying cause of their delay in learning language, and to recommend early intervention for toddlers who are
prognosis for immediate change. having difficulty learning language. The fact is, we know a
Generally these children are the toddlers who are slow at lot about language development and language delay during
producing first words and word combinations and may the first years of life; decisions regarding whether to
persist in having grammatical errors as syntax emerges. They recommend intervention need not be based on good
are the children who show no other signs of developmental intentions alone, but rather on the data that are available.
delays, such as cognitive, emotional, or sensory problems,
but who are demonstrating difficulty learning language. They
may be identified first as late talkers, or many are diagnosed Early Intervention: Goals and Issues
as having SLI or SELD. The underlying cause of the Early intervention is designed to bring about short-term
language-learning difficulty is not known, but theories change that will influence long-term progress. The primary
abound, many of which seem plausible (see Thal & Katich, objective of early intervention for children with delays in

American Journal of Speech-Language Pathology • Vol. 7 • 1058-0360/98/0701-0023 © American Speech-Language-Hearing

Olswang • Rodriguez • Timler
Association 23
language is to bring a child’s performance within normal language change are behavioral characteristics that suggest
limits for his or her chronological age as quickly as that a child who is delayed in language learning will catch
possible. Even if some toddlers outgrow their language up to his or her peers, or behavioral characteristics that
problems, concerns about cumulative effects of the delay suggest that a child is ready to move ahead to the next
and the impact on emotional and behavioral development language milestone. Risk factors are familial or behavioral
have prompted recommendations for intervention. Further, characteristics that suggest that a child is likely to have a
parents often feel the need to “do something” to ease their true language impairment, as opposed to being a late
minds or to have strategies for getting through the day-to- bloomer. Research addressing children having difficulty
day frustrations of living with a child whose language is learning language and children developing typically are
not emerging on schedule. These immediate needs support examined.
the benefits of short-term intervention, whether or not This literature review includes research with children
long-term effects occur. Perhaps these reasons alone who were considered late talkers or late bloomers; children
should make the recommendations for early intervention with specific language impairment, specific expressive
clear, but they don’t. Health and education reform, declining language impairment, slow expressive language develop-
resources, and demands for cost-effective programs have ment, language impairment due to other causes; and
caused practitioners to take a second look at their delivery children who are developing typically. The breadth of
of services. These pressures, coupled with the data from research allowed for a thorough view of predictors and risk
the longitudinal studies suggesting that the majority of the factors of language change in toddlers. The premise
children with SELD outgrow their problems and lack of driving our thinking was as follows: Children who exhibit
agreement about the underlying cause of the disorder, have few positive predictors of change and many risk factors are
prompted the debate about intervention. more likely to have a true impairment and need interven-
In this paper we are viewing intervention as short-term, tion than children who exhibit many predictors of change
intensive involvement with a speech-language pathologist and few risk factors. The review that follows will aggre-
that is designed to bring about immediate, significant gains gate these data in an effort to describe children for whom
in performance. The intervention may involve direct treatment should be recommended versus children for
contact (either one-to-one or in a group) with the child or whom a watch and see approach should be followed.
the parent, but it is by definition a significant expenditure The research presented in this review includes findings
of the clinician’s time for several months. Intervention is that support predominant trends. As with all types of
designed to facilitate or accelerate the change that is likely research, discrepant data or negative findings are available.
to occur on its own (Olswang & Bain, 1991). The treat- The discrepancies may be due to variability across re-
ment may serve to prevent or reduce severity of language search, including subject inclusion and exclusion criteria,
problems in the long run, but the focus is on short-term procedures, or data analysis, or a combination of these
gains. Numerous treatment strategies are known to be factors. The aim of this review is to form a consensus view
successful with toddlers, but no one treatment appears of our knowledge to date. Therefore, this paper presents
significantly more successful than others (see Girolametto, the trends as supported by at least three studies suggesting
Pearce, & Weitzman, 1996; Olswang & Bain, 1991; similar findings. The paper will present predictors of
Weismer, Murray-Branch, & Miller, 1993, for a discussion change first, organized according to speech and nonspeech
of treatment strategies). The question about recommending behaviors, followed by risk factors.
early intervention then is, can treatment serve to trigger
language learning? If a parent calls a clinician and asks Predictors of Change
whether his or her child should be in treatment, the
clinician must be able to make an informed judgment about Language Production
the immediate cost benefits of this recommendation. This The earliest characteristic of a language learning
must be a well-informed decision. Because research has difficulty is most often a delay in the production of first
not directly linked a single diagnosis with a particular words. A 2-year-old with fewer than 50 words is clearly at
treatment, clinicians must make an informed decision on risk for continued delay; the risk grows as a child ages with
the basis of their knowledge of the disorder as it relates to little change in language production (Paul, 1989; Paul &
typical language learning. This involves the clinician Alforde, 1993; Rescorla, Roberts, & Dahlsgaard, 1997;
assessing the child’s performance by identifying behavioral Rescorla & Schwartz, 1990). Specifically, expressive
characteristics of the disorder (i.e., documenting the vocabulary size in relationship to age appears to be a
symptomatology of the problem) and projecting possibili- strong predictor of continued language growth (Fischel,
ties for change in the immediate future. In addition, the Whitehurst, Caulfield, & DeBaryshe, 1989; Olswang,
parents’ needs will influence the clinician’s decision. Parents’ Long, & Fletcher, 1997). However, the picture may not be
beliefs and concerns will need to be weighed against the as clear as to permit saying that expressive vocabulary
child’s performance in making a final recommendation. alone predicts later language learning. Some research
What follows is a review of the literature designed to indicates that expressive vocabulary in relation to compre-
identify child behaviors that shape a profile of toddlers hension is a good predictor. Thal and colleagues (Thal,
who should receive intervention. The review presents Oroz, Evans, Katich, & Leasure, 1995) examined word
empirically documented predictors of language change production and percentage of receptive vocabulary
and risk factors for language impairment. Predictors of produced (i.e., number of words produced divided by the

24 American Journal of Speech-Language Pathology • Vol. 7 • No. 1 February 1998

number of words understood) as measured by the MacArthur the production of general all-purpose verbs and few
Communicative Development Inventories (CDI; Fenson et intransitive and ditransitive verbs.
al., 1993) as a predictor of language learning. Their results
indicated that measures taken at first visits (either 13 or 20
months of age) reliably discriminated the children with Language Comprehension
delays and the children without delays at the second visits Receptive vocabulary in infants and toddlers who are
(20 and 26 months of age). Thal, Tobias, and Morrison developing typically is correlated with later word produc-
(1991) reported that children with delayed comprehension tion (Bates, Benigni, Bretherton, Camaioni, & Volterra,
also had the lowest vocabulary and remained delayed one 1979; Bates, Bretherton, & Snyder, 1988). Several studies
year later. have documented residual expressive language deficits in
The makeup of early vocabulary has also revealed children with receptive language delays at 2 years of age
differences between children who are developing typically (Scarborough & Dobrich, 1990; Thal et al., 1991). Weismer
and those with language impairments. Watkins, Rice, and et al. (1994) found shifting receptive language performance
Moltz (1993) found that children with SLI had less diverse across 3-month assessment intervals, reflecting a need for
verb repertoires than typically developing children. Rice continued monitoring of the late talker. Receptive deficits,
and Bode (1993) further examined verbs and discovered in general, seem to suggest a more severe impairment and
that children with SLI relied on general all purpose verbs a poorer prognosis for change (Bishop & Edmundson,
(GAPs) in their conversations, even though they were 1987; Thal & Tobias, 1992). Several studies have exam-
capable of greater variety. Approximately 8 to 10 percent ined the size of the comprehension-production gap as a
of their verb types accounted for 40 to 50 percent of verb predictor of change. Thal et al. (1991) reported that children
possibilities. GAPs included such words as want, go, who caught up had the greatest comprehension-production
get, do, put, look, make, and got. Recently, Loeb, Pye, gap. However, this was the case when comprehension was
Redmond, and Richardson (1996) investigated verb age appropriate. In another study with SELI children, the
production through a structured elicitation task. They degree of the receptive-expressive gap was correlated to
found that children with SLI tended to label fewer verbs, immediate change in toddlers moving from single words to
particularly low frequency of occurrence verbs, than the multiword productions; the greater the gap, the slower the
age comparison group. Further, they labeled fewer fixed change (Olswang & Bain, 1996). The conflicting data
intransitive verbs compared to same-age peers. Olswang et regarding comprehension suggests caution in drawing
al. (1997) also investigated vocabulary composition as strong conclusions. The consensus suggests that toddlers
possible predictors of immediate change for children with with significant expressive and receptive language delays of
SELI. This study found fairly high correlations between 6 months or more are most at risk for continued language
lexical composition (production of nouns, verbs, and delay. Further, for those children delayed in both compre-
modifiers) and mean length of utterance (MLU) over a 9- hension and production, the larger the comprehension-
week period. On examining verb types, the results indi- production gap, the poorer the prognosis.
cated high correlations for the production of intransitive
and ditransitive1 verb types with subsequent MLU scores,
but not transitive verb types. This study further explored Phonology
the verb production of children who began producing two- Babbled speech has also been examined as a predictor
word utterances during the 9-week study period (“chang- of future language learning. Amount of prelinguistic
ers”) and those who did not (“nonchangers”). They found vocalization appears to be predictive of emerging verbal-
that the changers produced approximately twice as many izations (i.e., meaningful speech) in children developing
single-word verb types as the nonchangers at the begin- typically and those with delays (Camp, Burgess, Morgan,
ning of the study period. The changers and nonchangers & Zerbe, 1987; Kagan, 1971; Rescorla & Ratner, 1996).
produced equivalent numbers of transitive single-word This includes future vocabulary and amount of talking.
verbs. However, the changers produced almost three times Vocalization structure also appears related to later lan-
as many intransitive single words as the nonchangers and guage production. Canonical babbling seems to be most
two times as many ditransitive single words. These results strongly related to emergence of language in children
were found to be statistically significant. Children who developing typically (McCarthren, Warren, & Yoder,
moved on to produce word combinations produced 1996). More specifically, several studies of children with
approximately twice as many intransitive verbs as transi- language learning difficulties have found that size of
tive and almost six times as many ditransitive verbs as consonant inventory in toddlers’ babble appeared related to
transitive verbs. Thus, quantity and variety of vocabulary their future vocabulary growth (Paul & Jennings, 1992;
appears related to language change. In summary, these Rescorla & Ratner, 1996; Whitehurst et al., 1991).
data suggest that a toddler with a small vocabulary in Whitehurst et al. (1991) found that the proportion of
relation to age and a less diverse vocabulary composition, vowel-to-consonant babble was predictive of those
particularly in regard to verbs, is likely to be a good children who did better in language production 5 months
candidate for intervention. Particular red flags would be later. Similarly, research has shown that complexity of
babbling as measured by consonant use correlated with
1
Ditransitive verbs are those that may or may not take a direct object, such emergence of first words (Stoel-Gammon, 1989). Finally,
as the word smell. when children begin to produce their first words, the

Olswang • Rodriguez • Timler 25

percentage of consonants that is correct has been a variable Play
separating children developing language typically from A considerable amount of research has examined the
those with future language learning problems (Paul & relationship between prelinguistic play and later language
Jennings, 1992). In addition, children who were slow to learning. Play appears to serve as a foundation for the content
develop expressive language were less accurate in their of first words and word combinations. The most fruitful areas
production of consonants, less varied in their consonant of investigation have examined combinatorial/thematic play
repertoire, and more restricted in the complexity of syllable (i.e., using toys together in meaningful ways) and symbolic
structure. Stoel-Gammon (1987) has indicated that play (i.e., using objects or toys to stand symbolically for
typically developing toddlers produced 62.1% consonants other objects or toys—pretend play). In both children who
correctly at 21 months and 70.7% consonants correctly at are developing typically and children with language
24 months during conversation. Moreover, the research impairments related to other disorders, combinatorial play
suggests that, at 24 months, a child with a phonetic and symbolic play appear correlated to later language
inventory composed of only four to five consonants and a comprehension and production (Bates et al., 1979; Casby
limited variety of vowels would be viewed as at risk for & Ruder, 1983; Mundy, Sigman, & Kasari, 1990; Mundy,
continued delay (Stoel-Gammon, 1991). Stoel-Gammon Sigman, Kasari, & Yirmiya, 1988). Late talkers seem to
(1991) has indicated that many phonological errors get engage in higher frequency of manipulations, handling,
resolved between 24 and 36 months. She suggests that the and grouping of toys and objects than in combinatorial/
following error patterns persisting until 36 months would thematic play and symbolic play (Rescorla & Goossens,
be cause for concern: numerous vowel errors, widespread 1992). In contrast, toddlers who are slow in their language
deletion of initial consonants, substitution of glottal learning but who demonstrate more elaborate play schemes
consonants or /h/ for a variety of consonants, substitution (closely linked to combinatorial/thematic and symbolic
of back consonants for front ones (particularly velars for play) appear to have a better chance of catching up with
alveolars), and widespread deletion of final consonants. their chronological age peers (late bloomers) (Thal et al.,
Thus, variety and accuracy of sound production are 1991). Weismer, Murray-Branch, and Miller (1994)
important variables to examine in toddlers’ speech; the observed that three of their four late talkers demonstrated a
greater the variety and accuracy, the better the prognosis notable increase in performance on a symbolic play test
for language learning. Therefore, the evidence suggests about the time they achieved a 50-word vocabulary,
that a toddler with few prelinguistic vocalizations, limited followed by an extended surge in word production.
phonetic inventory, restricted syllable structure, vowel Play behavior also appears to be an important building
errors, and less accurate production of consonants is block for early forms of requesting and commenting
presumably at greater risk for continued delay and thus a because object manipulation and interpersonal engagement
good candidate for intervention. are the focus of these communicative functions. Early
interventions involving object manipulation and play have
Imitation been shown to be fruitful for facilitating requesting and
commenting in children with language impairments
In studies of children developing language typically and associated with motor impairments and mental retardation.
those with impairments, evidence has consistently indicated (Olswang & Pinder, 1995; Pinder & Olswang, 1995;
that spontaneous imitations of new linguistic structures Yoder, Warren, & Hull, 1995). In fact, Yoder et al. (1995)
precede the subsequent spontaneous productions of these reported that amounts of combinatorial and symbolic play
same structures (Bloom, Hood, & Lightbown, 1974; in treatment were directly related to improved prelinguistic
Leonard, Schwartz, Folger, Newhoff, & Wilcox, 1979; requesting in children with mental retardation. These
Scherer & Olswang, 1984; Scherer & Olswang, 1989; findings suggest that play is an important instrument for
Weismer et al., 1993). For children with SELI, this has language learning, both for language content and language
been studied directly by Olswang and Bain as part of their use. Specifically, a clinician would be justified in voicing
investigations of dynamic assessment (Bain & Olswang, concern about a toddler whose play consisted primarily of
1995; Olswang & Bain, 1996). In examining toddlers with grouping and manipulating toys, rather than combinatorial
SELI who were producing single-word utterances but or symbolic toy use, or both. A child exhibiting delayed
appeared ready to produce word combinations, success at play behaviors of this type would be an appropriate
producing word combinations with and without models candidate for intervention.
was investigated. Results showed that children who did not
imitate any two-word utterances during the dynamic
assessment did not begin to produce word combinations Gestures
during the 9-week study period. Further, those toddlers Several studies have investigated toddlers’ use of
needing fewer prompts and cues to imitate word combina- representational or communicative gestures as predictors of
tions (i.e., fewer models) were the children who made the change. Thal and Tobias (1992) reported that late bloomers
greatest gains in language production, specifically in (toddlers who caught up to their peers within a year’s time)
combining words (Olswang & Bain, 1996). We conclude used significantly more communicative gestures than age-
that toddlers who do not imitate word combinations, and language-matched controls. Research has also docu-
particularly with a variety of prompts and cues, should be mented that production of sequences of symbolic gestures
considered appropriate candidates for treatment. discriminated those 2-year-olds who outgrew their language

26 American Journal of Speech-Language Pathology • Vol. 7 • No. 1 February 1998

delays from those who did not (Rescorla & Goossens, 1992; participate in proportionately fewer peer interactions, and
Thal & Bates, 1988; Thal, Tobias, & Morrison, 1991; Thal they have difficulty in gaining entry into peer activities
& Tobias, 1994). (Craig & Washington, 1993). Recently Tretter (1996)
As children move from producing predominantly single examined these conversational skills in three toddlers
words to multiword combinations, they appear to use diagnosed with language impairments. Her results indi-
gestures to support the transition. This change has been cated a varied performance among the children; only one
documented in children developing typically and children of the children demonstrated a preference for responding
with specific expressive delays (Belleville & Tretter, 1994; rather than initiating, reluctance to gain access to a group,
Iverson, Volterra, Pizzuto, & Capirci, 1994; Olswang, and willingness to interact with adults but not peers. This
Johnson, & Crooke, 1992). Two types of gestures occur- child was enrolled in a toddler language group and
ring with single words have been identified in children appeared to benefit from treatment in terms of language
developing typically: complementary and supplementary and conversational growth. Paul, Looney, and Dahm
gestures (Goldin-Meadow & Morford, 1985). Complemen- (1991) examined the socialization of late talkers using the
tary gestures appear to code the same meaning as the single Vineland Adaptive Behavior Scales (Sparrow, Balla, &
word that is produced (e.g., shaking one’s head while Cicchetti, 1984) and found that socialization skills along
saying “no”); thus, the gesture does not expand the with receptive and expressive language were lower than in
meaning of the utterance. Supplementary gestures appear matched normally speaking toddlers. A follow-up of the
to code new or additional meaning in conjunction with the same subjects at age 3 revealed that nearly half of the late
single word (e.g., shaking one’s head while saying “juice” talkers were still behind in expressive communication and
to mean “no juice”); in this case, the gesture expands the socialization. This literature suggests that toddlers who
meaning of the utterance. During the transition, children exhibit socialization problems, including reluctance to
appear to move from using single words predominantly to initiate and participate in conversations with peers, may be
single words plus a complementary gesture, to single of greater concern for a clinician, and they may be likely
words plus supplementary gestures, and finally to word candidates for intervention.
combinations. In two studies designed to explore this
concept, the gestures plus single words were examined in
children with SELI who began producing word combina-
Risk Factors
tions during a 9-week study period (Belleville & Tretter, Heritability
1994; Olswang, Johnson, & Crooke, 1992). In both studies, Research investigating heritability as a risk factor for
the results revealed that children with SELI who began using specific language impairment has yielded fairly consistent
word combinations had more supplementary gestures than results. These studies have reported a higher proportion of
complementary gestures in their repertoires at the beginning relatives with histories of language impairment or learning
of the 9 weeks. These data suggest that children who are disability in families of children diagnosed as SLI than in
using representational gestures in sequences and producing those of typically developing children (Bishop & Edmund-
supplementary gestures along with single words are attempt- son, 1987; Bishop, North, & Donlan, 1995; Lahey &
ing to create more elaborate messages with the tools they Edwards, 1995; Lewis & Thompson, 1992; Paul, 1991;
have. Further, these are the children who appear to be most Tallal, Ross, & Curtiss, 1989; Tomblin, 1989; Tomblin,
likely to make immediate gains in their language production. Hardy, & Hein, 1991; Weismer, Murray-Branch, & Miller,
Therefore, toddlers who exhibit few, if any, representational 1994). Discrepancies in this literature have primarily
gestures, sequences of gestures, and supplementary gestures centered around the definition of SLI. Examination of
plus single words to convey more complex thoughts would related language disorders and family history has yielded
be considered better candidates for intervention. somewhat conflicting results (Spitz, Tallal, Flax, &
Benasich, 1997; Whitehurst, Arnold, et al., 1991), reveal-
Social Skills ing the complexities of this issue. However, the results
consistently support a family history of language impair-
Social skills have been examined along several dimen- ment or learning disability as a risk factor for a toddler
sions. Whitehurst and colleagues (Fischel et al., 1989; who is late in learning to talk. In summary, if one of the
Whitehurst & Fischel, 1994) have discussed the impact of toddler’s parents or siblings demonstrates persistent
specific language delay with the comorbidity of behavior language and learning difficulties, the risk of continued
problems. They suggested that children with behavior language delay in this toddler is increased. This, then,
problems are less likely to outgrow their disorder. Several becomes a risk factor that might alert a clinician to more
studies of preschoolers with language impairment have positively consider intervention for a child.
examined the children’s participation in conversation.
Preschoolers with SLI prefer to initiate interactions with
adults, not peers (Craig, 1993; Hadley & Rice, 1991; Rice, Otitis Media
Sell, & Hadley, 1991). Conversationally, these children Investigations examining the relationship between otitis
tend to be more passive than assertive in interactions (Rice media and language development have yielded inconsistent
et al., 1991). They respond to their communicative results. Teele, Klein, and Rosner (1984) found that
partners, but seldom initiate. Compared to children who are children who had prolonged periods of time with otitis
developing typically, preschool children with SLI tend to media had significantly lower language scores. However, a

Olswang • Rodriguez • Timler 27

later study demonstrated that a history of otitis media in repeatedly emerge as part of clusters of variables. Socio-
early childhood did not have a major negative impact on economic status (SES) is one of those characteristics. SES
later language development (Roberts, Burchinal, Davis, seems to be an important variable in predicting children’s
Collier, & Henderson, 1991). More recently, Paul and her development; low-SES families appear to be at higher risk
colleagues (Paul, Lynn, & Lohr-Flanders, 1993) followed a for negative child outcomes (Hart & Risley, 1995; Siegel,
group of late talkers and normally developing toddlers with 1981, 1982).
and without histories of middle ear involvement. They Another parent characteristic, interaction style, has also
found that children who were late to develop expressive been linked to language development, although the
language were at risk for prolonged slow language growth, research results are mixed. Literature on children develop-
at least until age 3, whether they had a history of frequent ing language typically suggests a number of styles that
middle ear involvement or not. Thus, the risk for chronic seem to best facilitate language learning. These include
expressive language delay did not appear to be increased frequently engaging children in reciprocal social interac-
significantly by a history of middle ear involvement. tions, following the child’s lead and participating in
Studies have found no difference in the frequency, contingent interactions, maintaining joint attention around
duration, or timing of otitis media between normally objects and events in the environment, and providing
developing and language delayed children (Bishop & language models in a simplified register (Hart & Risley,
Edmundson, 1986; Lonigan, Fischel, Whitehurst, Arnold, 1995; Tannock & Girolametto, 1992). Generally the
& Valdez-Menchaca, 1992). One of these studies did literature has suggested that the interaction loop between a
suggest that children with a higher frequency of treated parent and child is altered when a child has a language
otitis media between 12 and 18 months were more likely to impairment. This then may alter the parent’s style of
show improvement in their expressive language than interaction. Parents of children with delays often become
children without such a history (Lonigan et al., 1992). less contingently responsive and more controlling in their
These findings suggest that the expressive abilities affected activities and topics (Tannock & Girolametto, 1992). This,
by otitis media improved when otitis media and its coupled with excessive talkativeness and complex utter-
transient effects on hearing were no longer present. ances in speech directed to toddlers who are exhibiting
Increasingly, investigations have documented an delays in language development, may not create the best
association between otitis media and articulation difficul- environment for language learning. Over the years, these
ties. Late-talking children with a history of middle ear behaviors have become the focus of parent-training
involvement before age 3 seemed to have a somewhat programs, which have been shown to be successful for
greater risk for prolonged difficulties with articulation toddlers with specific vocabulary delay (Girolametto et al.,
(Paul et al., 1993). Similarly, Lonigan and colleagues 1996; Girolametto, Pearce, & Weitzman, 1997).
(Lonigan et al., 1992) reported that children who had more Parents bring a host of beliefs to child rearing. These
persistent episodes of otitis media when they were beliefs reflect each parent’s own cultural, religious,
between 18 and 24 months had significantly poorer educational, and socioeconomic background and as such
articulation than children with less experience with otitis influence the way the parent views development and
media. The current evidence suggests that a history of disorders of development. Sometimes beliefs are such that
middle ear involvement places a child at greater risk for they prompt concerns that are more or less consistent with
articulation difficulties. the child’s performance. The day-to-day difficulties of
In summary, prolonged, untreated otitis media places raising a child who is not developing on schedule adds to
a child at greater risk of continued language delay. the challenge of child rearing. When communicative
Further, evidence suggests that toddlers with a history of interactions with a child are interrupted, the interactions
persistent otitis media are at greater risk for difficulties become increasingly burdensome, and this contributes to
with articulation. the degree of the parents’ concerns. Without a doubt, some
intervention is warranted when parents are demonstrating
extreme concerns, such that they are no longer able to
Parent Needs enjoy their child or participate fully in parenting activities.
Finally, in deciding whether to recommend intervention Parents may need information about language development
for children with specific language impairment, clinicians and delay or specific techniques to help foster better
must consider parent needs. In fact, parent variables may interactions with their children, but it is the parents’
outweigh the child’s factors in the decision-making concern that is the focus of the treatment.
process. Two aspects of parent needs are examined here:
parent characteristics and parent concerns. Both contribute
to a parent’s influence on deciding whether to recommend Summary
intervention and to the type of intervention that is consid- This review of the literature suggests several character-
ered most advantageous. istics that indicate whether change in a toddler’s language
Parent characteristics have been examined extensively production is imminent (predictors) and whether the
as they relate to their children’s language and intellectual presence of a true language impairment is likely (risk
development, but results have been anything but clear. Few factors). To some extent, the characteristics can be viewed
isolated characteristics appear to be directly related to as addressing severity of the language problem and,
children’s development. However, a few characteristics therefore, the degree of concern. Unfortunately, the

28 American Journal of Speech-Language Pathology • Vol. 7 • No. 1 February 1998

research to date does not support prioritizing or weighting We would all like to make foolproof clinical decisions.
the characteristics. Essentially, the fewer predictors of If the world were perfect, research would consistently
change demonstrated by a toddler and the greater the point us in the right direction for making decisions about
number of risk factors, the more concerned a clinician intervention recommendations for all toddlers with
should be about language development. Even though we language learning difficulties. This, of course, is not the
are left with a long list of individual characteristics, case, nor is it ever likely to be. This reality must not put us
together they begin to paint a picture of a toddler who is in at a standstill or head us in the direction of making
serious trouble for language development and for whom decisions based solely on good intentions. This article was
intervention would be the most appropriate recommenda- designed to highlight what we know to be true about the
tion. Simply, the more characteristics the toddler—and the development of language in toddlers. Research has
child’s family—demonstrate, the more serious the concern, revealed robust trends about language learning in toddlers
and as a result, the more likely the clinician should be to who are typical and atypical in their language develop-
recommend intervention. Intervention may take several ment. These trends have brought to light characteristics
forms: direct one-to-one treatment with the child, group that allow us to decide whether we should be seriously
treatment with the child, parent training, or some combina- concerned about a toddler’s actual and potential language
tion of these alternatives. Degree of concern may be able to growth. The argument being made from this literature is
help dictate the form of intervention. It follows that the that the magnitude of our concern should directly translate
fewer the characteristics a toddler presents, the more likely to our recommendations. To our way of thinking, and the
a clinician would consider a watch and see approach, thinking of others (Thal & Katich, 1996; Whitehurst &
reviewing the child’s status at 3- to 6-month intervals Fischel, 1994), this is not only a reasonable position, but
(Paul, 1996). also an ethical and intellectually defensible one.
Down the road, a treatment approach may surface that
challenges this perspective. In this scenario, the treatment
Conclusions will most likely be tied to the etiology of language impair-
Table 1 presents a summary of the findings presented in ment. As such, diagnosis will lead directly to treatment,
this article. It highlights predictors and risk factors that and even perhaps, a cure. Recently, Tallal and colleagues
suggest a profile of serious concern for a toddler who is have reported a treatment that is designed to directly
having difficulty learning language. This list, and the address the auditory-processing skills that they believe
profile it paints, is meant as a guide in helping clinicians underlie the language disorder in many children with
decide who to enroll in intervention. specific language impairment (Tallal & Merzenich, 1996;

TABLE 1. Predictors and risk factors of language change in toddlers.

Predictors Risk Factors

Speech Nonspeech

Language Production Play Otitis Media

Small vocabulary for age Primarily manipulating and grouping Prolonged periods of untreated otitis
Few verbs Little combinatorial and/or symbolic media
Preponderance of GAPS play
More transitive verbs Heritability
Few intransitive and ditransitive verb Gestures Family member with persistent language
forms Few communicative gestures, symbolic and learning problems
gestural sequences, or supplementary
Language Comprehension gestures Parent Needs
Presence of 6-month comprehension delay Parent characteristics:
Large comprehension-production gap Social Skills Low SES
with comprehension deficit Behavior problems Directive more than responsive
Few conversational initiations interaction style
Phonology Interactions with adults more than peers Parent concern:
Few prelinguistic vocalizations Difficulty gaining access to activities Extreme
Limited number of consonants
Limited variety in babbling structure
Less than 50% consonants correct
(substitution of glottal consonants and
back sounds for front)
Restricted syllable structure
Vowel errors

Imitation
Few spontaneous imitations
Reliance on direct model and prompting in
imitation tasks of emerging language forms

Olswang • Rodriguez • Timler 29

Tallal et al., 1996). These data have been impressive, and Camp, B., Burgess, D., Morgan, L., & Zerbe, G. (1987). A
yet the studies from which they emanate are not without longitudinal study of infant vocalization in the first year.
problems. This line of research may eventually allow Journal of Pediatric Psychology, 12, 321–331.
clinicians to identify groups of children with SLI who will Casby, M., & Ruder, K. (1983). Symbolic play and early
language development in normal and mentally retarded
benefit from a particular form of treatment. That outcome
children. Journal of Speech and Hearing Research, 26,
will lead to other types of diagnoses and treatments and 404–411.
improved confidence in our recommendations. Craig, H. (1993). Social skills of children with specific language
We are not there yet. However, we are extremely knowl- impairment: Peer relationships. Language, Speech, and
edgeable about toddlers with delays in language. They are a Hearing Services in Schools, 24, 206–215.
heterogeneous group, bringing different individual and Craig, H., & Washington, J. (1993). The access behaviors of
family needs. As a result, they need different recommenda- children with specific language impairment. Journal of Speech
tions for intervention. Even with this complexity, we can and Hearing Research, 36, 322–337.
make informed decisions. The first decision is whether to Fenson, L., Dale, P., Reznick, J., Thal, D., Bates, E., Hartung,
bring them into our clinics for our professional help. We may J., Pethick, S., & Reilly, J. (1993). MacArthur Communica-
tive Developmental Inventories. San Diego: Singular Press.
work with the child directly, recommend a preschool
Fischel, J., Whitehurst, G., Caulfield, M., & DeBaryshe, B.
language group, or provide parent training, but we must (1989). Language growth in children with expressive language
decide whether we should watch and see or treat in some delay. Pediatrics, 82, 218–227.
way. The goal of this paper was to share state-of-the-art Girolametto, L., Pearce, P., & Weitzman, E. (1996). Interactive
research that should allow us to make informed clinical focused stimulation for toddlers with expressive vocabulary
decisions about recommending intervention. We may not delays. Journal of Speech and Hearing Research, 39, 1274–
have all of the answers, but we know a lot, and we need to 1283.
start applying the information with confidence. Girolametto, L., Pearce, P., & Weitzman, E. (1997). Effects of
lexical intervention on the phonology of late talkers. Journal
of Speech, Language, and Hearing Research, 40, 338–347.
Author Note Goldin-Meadow, S., & Morford, M. (1985). Gesture in early
child language: Studies of deaf and hearing children. Merrill-
The authors would like to gratefully acknowledge the helpful Palmer Quarterly, 31, 145–176.
comments provided by Luigi Girolametto, Donna Thal, and Marc Hadley, P., & Rice, M. (1991). Conversational responsiveness of
Fey in the final phases of writing this manuscript. speech and language impaired preschoolers. Journal of Speech
and Hearing Research, 34, 1308–1317.
Hart, B., & Risley, T. (1995). Meaningful differences in the
References everyday experience of young American children. Baltimore:
Bain, B., & Olswang, L. (1995). Examining readiness for Paul H. Brookes.
learning two-word utterances by children with specific Iverson, J., Volterra, V., Pizzuto, E., & Capirci, O. (1994). The
expressive language impairment: Dynamic assessment role of communicative gesture in the transition to the two-
validation. American Journal of Speech-Language Pathology, word stage. Cognitive Development, 9, 23–43.
4(1), 81–91. Kagan, J. (1971). Change and continuity in infancy. New York:
Bates, E., Benigni, L., Bretherton, I., Camaioni, L., & Wiley.
Volterra, V. (1979). In E. Bates, L. Benigni, I. Bretherton, L. Lahey, M., & Edwards, J. (1995). Specific language impair-
Camaioni, & V. Volterra (Eds.), The emergence of symbols: ment: Preliminary investigation of factors associated with
Cognition and communication in infancy (pp. 69–140). New family history and with patterns of language performance.
York: Academic Press. Journal of Speech and Hearing Research, 38, 643–657.
Bates, E., Bretherton, I., & Snyder, L. (1988). From first words Leonard, L., Schwartz, R., Folger, K., Newhoff, M., & Wilcox,
to grammar. New York: Academic Press. J. (1979). Children’s imitations of lexical items. Child
Belleville, A., & Tretter, S. (1994). Gestures as precursors to the Development, 50, 19–27.
production of first word combinations by children with Lewis, B., & Thompson, L. (1992). A study of developmental
expressive language impairments. Senior honors thesis, speech and language disorders in twins. Journal of Speech and
University of Washington, Seattle. Hearing Research, 35, 1086–1094.
Bishop, D., & Edmundson, A. (1986). Is otitis media a major Loeb, D., Pye, C., Redmond, S., & Richardson, L. (1996).
cause of specific developmental language disorder? British Eliciting verbs from children with specific language impair-
Journal of Disorders of Communication, 21, 321–338. ment. American Journal of Speech-Language Pathology, 5(4),
Bishop, D., & Edmundson, A. (1987). Language-impaired 17–30.
4-year-olds: Distinguishing transient from persistent impair- Lonigan, C., Fischel, J., Whitehurst, G., Arnold, D., &
ment. Journal of Speech and Hearing Disorders, 52, 156–173. Valdez-Menchaca, M. (1992). The role of otitis media in the
Bishop, D., North, T., & Donlan, C. (1995). Genetic basis of development of expressive language disorder. Developmental
specific language impairment: Evidence from a twin study. Psychology, 28, 430–440.
Developmental Medicine and Child Neurology, 37, 56–71. McCarthren, R., Warren, S., & Yoder, P. (1996). Prelinguistic
Bloom, L. (1993). Developments in cognition. In L. Bloom, The predictors of later language development (pp. 57–75). In K.
transition from infancy to language: Acquiring the power of Cole, P. Dale, & D. Thal (Eds.), Assessment of communication
expression (pp. 214–242). New York: Cambridge University and language. Baltimore: Paul H. Brookes.
Press. Mundy, P., Sigman, M., & Kasari, C. (1990). A longitudinal
Bloom, L., Hood, L., & Lightbown, P. (1974). Imitation in study of joint attention and language development in autistic
language development: If, when and why. Cognitive Psychol- children. Journal of Autism and Developmental Disorders, 20,
ogy, 6, 380–420. 115–128.

30 American Journal of Speech-Language Pathology • Vol. 7 • No. 1 February 1998

Mundy, P., Sigman, M., Kasari, C., & Yirmiya, N. (1988). speech and language impaired children. Journal of Speech and
Nonverbal communication skills in Down syndrome children. Hearing Research, 34, 1299–1307.
Child Development, 59, 235–249. Roberts, J., Burchinal, M., Davis, B., Collier, A., & Henderson,
Olswang, L., & Bain, B. (1991). Intervention issues for toddlers F. (1991). Otitis media in early childhood and later language.
with specific language impairments. Topics in Language Journal of Speech and Hearing Research, 34, 1158–1168.
Disorders, 11, 69–86. Scarborough, H., & Dobrich, W. (1990). Development of
Olswang, L., & Bain, B. (1996). Assessment information for children with early language delay. Journal of Speech and
predicting upcoming change in language production. Journal Hearing Research, 33, 70–83.
of Speech and Hearing Research, 39, 414–423. Scherer, N., & Olswang, L. (1984). Role of mother’s expansions
Olswang, L., Johnson, G., & Crooke, P. (1992, November). in stimulating children’s language production. Journal of
Single to multi-words: A journey with SLI children. Paper Speech and Hearing Research, 27, 387–396.
presented at the annual convention of the American Speech- Scherer, N., & Olswang, L. (1989). Using structured discourse
Language-Hearing Association, San Antonio, TX. as a language intervention technique with autistic children.
Olswang, L., Long, S., & Fletcher, P. (1997). Verbs in the Journal of Speech and Hearing Disorders, 54, 383–394.
emergence of word combinations in young children with Siegel, L. (1981). Infant tests as predictors of cognitive and
specific expressive language impairment. European Journal of language development at two years. Child Development, 52,
Disorders of Communication, 32, 15–33. 545–557.
Olswang, L., & Pinder, G. L. (1995). Preverbal functional Siegel, L. (1982). Reproductive, perinatal, and environmental
communication and the role of object play in children with factors as predictors of the cognitive and language develop-
cerebral palsy. Infant-Toddler Intervention: The Transdisci- ment of preterm and full-term infants. Child Development, 53,
plinary Journal, 5, 277–300. 963–973.
Paul, R. (1989, April). Profiles of toddlers with delayed Sparrow, S., Balla, D., & Cichetti, D. (1984). Vineland Adaptive
expressive language development. Paper presented at the Behavior Scales (survey form). Circle Pines, MN: American
biennial meeting of the Society for Research in Child Guidance Service.
Development, Kansas City, MO. Spitz, R., Tallal, P., Flax, J., & Benasich, A. (1997). Look
Paul, R. (1991). Profiles of toddlers with slow expressive language who’s talking: A prospective study of familial transmission of
development. Topics in Language Disorders, 11, 1–13. language impairments. Journal of Speech, Language, and
Paul, R. (1996). Clinical implications of the natural history of Hearing Research, 40, 990–1001.
slow expressive language development. American Journal of Stoel-Gammon, C. (1987). Language Production Scale. In L.
Speech-Language Pathology, 5(2), 5–21. Olswang, C. Stoel-Gammon, T. Coggins, & R. Carpenter
Paul, R., & Alforde, S. (1993). Grammatical morpheme (1987). Assessing prelinguistic and early linguistic behaviors
acquisition in 4-year-olds with normal, impaired, and late in developmentally young children. Seattle: University of
developing language. Journal of Speech and Hearing Washington Press.
Research, 36, 1271–1275. Stoel-Gammon, C. (1989). Prespeech and early speech develop-
Paul, R., & Jennings, P. (1992). Phonological behavior in ment of two late talkers. First Language, 9, 207–224.
toddlers with slow expressive language development. Journal Stoel-Gammon, C. (1991). Normal and disordered phonology in
of Speech and Hearing Research, 35, 99–107. two-year-olds. Topics in Language Disorders, 11, 21–32.
Paul, R., Looney, S., & Dahm, P. (1991). Communication Tallal, P., & Merzenich, M. (1996, November). New treat-
and socialization skills at ages 2 and 3 in “late-talking” ments for language-impaired children: Integrating basic and
young children. Journal of Speech and Hearing Research, clinical research. Miniseminar presented at the annual
34, 858–865. convention of the American Speech-Language-Hearing
Paul, R., Lynn, T., & Lohr-Flanders, M. (1993). History of Association, Seattle, WA.
middle ear involvement and speech/language development in Tallal, P., Miller, S., Bedi, G., Byma, G., Wang, X., Nagarijan,
late talkers. Journal of Speech and Hearing Research, 36, S., Schreiner, C., Jenkins, W., & Merzenich, M. (1996).
1055–1062. Language comprehension in language-learning impaired
Pinder, G. L., & Olswang, L. (1995). Development of commu- children improved with acoustically modified speech. Science,
nicative intent in young children with cerebral palsy: A 271, 81–84.
treatment efficacy study. Infant-Toddler Intervention: The Tallal, P., Ross, R., & Curtiss, S. (1989). Familial aggregation
Transdisciplinary Journal, 5, 51–70. in specific language impairment. Journal of Speech and
Rescorla, L., & Goossens, M. (1992). Symbolic play develop- Hearing Disorders, 54, 167–173.
ment in toddlers with specific expressive language impair- Tannock, R., & Girolametto, L. ( 1992). Reassessing parent-
ment. Journal of Speech and Hearing Research, 35, 1290– focused language intervention programs. In S. Warren & J.
1302. Reichle (Eds.), Causes and effects in communication and
Rescorla, L., & Ratner, N. (1996). Phonetic profiles of toddlers language intervention (pp. 49–79). Baltimore: Paul H.
with specific expressive language impairment (SLI-E). Brookes.
Journal of Speech and Hearing Research, 39, 153–165. Teele, D., Klein, J., Rosner, B., & The Greater Boston Otitis
Rescorla, L., Roberts, J., & Dahlsgaard, K. (1997). Late- Media Study Group. (1984). Otitis media with effusion
talkers at 2: Outcomes at age 3. Journal of Speech, Language, during the first three years of life and development of speech
and Hearing Research, 40, 556–566. and language. Pediatrics, 74, 282–287.
Rescorla, L., & Schwartz, E. (1990). Outcome of toddlers with Thal, D., & Bates, E. (1988). Language and gesture in late talkers.
specific expressive language delay. Applied Psycholinguistics, Journal of Speech and Hearing Research, 31, 115–123.
11, 393–407. Thal, D., & Katich, J. (1996). Predicaments in early identifica-
Rice, M., & Bode, J. (1993). GAPs in the verb lexicons of tion of specific language impairment: Does the early bird
children with specific language impairment. First Language, always catch the worm? In K. Cole, P. Dale, & D. Thal (Eds.),
13, 113–131. Assessment of communication and language (pp. 1–28).
Rice, M., Sell, M., & Hadley, P. (1991). Social interactions of Baltimore: Paul H. Brookes.

Olswang • Rodriguez • Timler 31

 Thal, D., Oroz, M., Evans, D., Katich, J., & Leasure, K. (1995, Weismer, S., Murray-Branch, J., & Miller, J. (1994). A
June). From first words to grammar in late-talking toddlers. prospective longitudinal study of language development in
Paper presented at the annual Symposium on Research in late talkers. Journal of Speech and Hearing Research, 37,
Child Language Disorders, Madison, WI. 852–867.
Thal, D., & Tobias, S. (1992). Communicative gestures in Whitehurst, G., Arnold, D., Smith, M., Fischel, J., Lonigan,
children with delayed onset of oral expressive vocabulary. C., & Valdez-Menchaca, M. (1991). Family history in
Journal of Speech and Hearing Research, 35, 1281–1289. developmental language delay. Journal of Speech and
Thal, D., & Tobias, S. (1994). Relationships between language Hearing Research, 34, 1150–1157.
and gesture in normally developing and late-talking toddlers. Whitehurst, G., & Fischel, J. (1994). Early developmental
Journal of Speech and Hearing Research, 37, 157–170. language delay: What, if anything, should the clinician do
Thal, D., Tobias, S., & Morrison, D. (1991). Language and about it? Journal of Child Psychology and Psychiatry, 35,
gesture in late talkers: A 1-year follow-up. Journal of Speech 613–648.
and Hearing Research, 34, 604–612. Whitehurst, G., Fischel, J., Lonigan, C., Valdez-Menchaca,
Tomblin, B. (1989). Familial concentration of developmental M., Arnold, D., & Smith, M. (1991). Treatment of early
language impairment. Journal of Speech and Hearing expressive language delay: If, when, and how. Topics in
Disorders, 54, 287–295. Language Disorders, 11, 55–68.
Tomblin, B., Hardy, J., & Hein, H. (1991). Predicting poor- Yoder, P., Warren, S., & Hull, L. (1995). Predicting children’s
communication status in preschool children using risk factors response to prelinguistic communication intervention. Journal
present at birth. Journal of Speech and Hearing Research, 34, of Early Intervention, 19, 74–84.
1096–1105.
Tretter, S. (1996). Identification of language impairment and Received April 16, 1997
disability in preschoolers with specific expressive language Accepted October 30, 1997
impairment. Master’s thesis, University of Washington,
Seattle. Contact author: Lesley B. Olswang, University of Washington,
Watkins, R., Rice, M., & Moltz, C. (1993). Verb use by Department of Speech and Hearing Sciences, 1417 NE 42nd
language-impaired and normally developing children. First Street, Seattle, WA 98105-6246
Language, 13, 133–143.
Weismer, S., Murray-Branch, J., & Miller, J. (1993). A
comparison of two methods for promoting productive Key Words: language intervention, late talkers, specific
vocabulary in late talkers. Journal of Speech and Hearing language impairment, slow expressive language develop-
Research, 36, 1037–1050. ment, toddler language delay

32 American Journal of Speech-Language Pathology • Vol. 7 • No. 1 February 1998

View publication stats